- Email: [email protected]
Shifting fuels, downsizing or both? The Swedish example
Transportation Research Part D 18 (2013) 62–69
Contents lists available at SciVerse ScienceDirect
Transportation Research Part D journal homepage: www.elsevier.com/locate/trd
Shifting fuels, downsizing or both? The Swedish example Frances Sprei a,b,⇑, Sten Karlsson b a b
Stanford University, Stanford, CA, USA Chalmers University of Technology, Gothenburg, Sweden
a r t i c l e
i n f o
Keywords: Specific fuel consumption Downsizing Service attributes Performance attributes Diesel cars Flex-fuel cars
a b s t r a c t The paper looks at changes in Sweden’s new car fleet between 2002 and 2010. Between 2002 and 2007 consumer amenities such as acceleration capacity and passenger space continued to increase while fuel consumption steadily decreased. During these years the main technological and market change was a shift toward diesel and flex-fuel ethanol vehicles. After 2007 the average weight and power of the vehicles were more or less constant, while fuel consumption decreased by 13% between 2007 and 2010. The developments after 2007 suggest that 77% of the technological development between 2002 and 2010 resulted in reductions in fuel consumption compares to previous years when 35% of any technology change resulted in a net reductions. The shift can partly be attributed to the increased share of diesels and an engine downsizing. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Between 1985 and 2002, only 65% of the technological development allowing more efficient fuel consumption and lower CO2 emissions in the Swedish car fleet was offset by improvements in consumer ‘‘amenities’’ such as larger vehicles and improved acceleration capacity. Since 2002 there has been increased public and governmental awareness and willingness for action. The result has been a policy shift towards reducing the carbon impact of the new car fleet. The annual weight based registration tax changed in 2006 to be based on CO2 emissions. There has been a rebate program for ‘‘environment-friendly cars’’, i.e., vehicles emitting less than 120 g CO2/km, flex-fuel cars,1 gas driven cars and electric vehicles. Flex-fuel and gas driven cars have also received parking subsidies in about 40 municipalities and are exempted from congestion charging in Stockholm. Carbon and energy taxing of fossil fuels has continued to increase and hybrids, flex-fuel cars and gas driven cars are given reduction on the fringe taxing for company provided cars. Here we investigate the new car market from 2002 to 2007, as well as to 2010.
2. Method Data for the calculations are based two sources; sales statistics per vehicle model for 2007 and vehicle model characteristics with over 50 parameters per model for 1575 car models on sale during 2007. The former were collected from the Swedish Road Administration,2 and vehicle model parameters from the private company Autograph-bilfakta AB. Additional data was collected from car manufacturers and other databases when this was needed. In total 97% of the cars registered are covered by ⇑ Corresponding author at: Stanford University, Stanford, CA, USA. E-mail address: [email protected] (F. Sprei). Flex-fuel cars can be driven on any mixture of pure gasoline and E85 (85% gasoline and 15% ethanol). 2 The Swedish Road Administration was in 2010 transformed into the Swedish Transport Administration. Both names are used here, depending on when data was gathered. 1
1361-9209/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.trd.2012.09.006
F. Sprei, S. Karlsson / Transportation Research Part D 18 (2013) 62–69
63
the data. Privately imported cars and cars registered as light trucks are not included in the database. Data from1975, 1985, 1995 and 2002 are from similar databases constructed for these years (Sprei et al., 2008). Data for 2010 are based on the EU-monitoring database and the Swedish Transport Administration (2011). These databases do not cover the same number of parameters, as in previous years, thus not allowing the same analysis. They do, however, cover the same share of new registered vehicles. Sales-weighted averages and distributions are calculated to give an overview of the new car market. This was done for: all the cars sold, gasoline cars, diesel cars and flex-fuel cars. Gasoline cars include flex-fuel vehicles that can be driven with ethanol or gas. Fuel consumption (FC) is given in gasoline equivalents,3 and for flex-fuel cars is based on their pure gasoline consumption; because of the lower energy density of ethanol in E854 consumption in l/100 km will be higher for it. The effect of service and performance attributes on fuel consumptions is calculated stepwise. First, we calculate implied values for mass5 (Ms), engine capacity (DVs) and aerodynamic drag resistance (CDAs) due to increased acceleration and passenger compartment space (measured by index6 I). Changes in FC are computed from fuel consumption sensitivities derived from an analytical expression relating fuel consumption and driving cycle characteristics (Eqs. (4) and (5)). The implied fuel consumption due to improved service attributes, considering no technical improvement is:
FC s ¼ FC 02 þ DFCðDM s Þ þ DFCðDVsÞ þ DFCðDC D As Þ
ð1Þ
The change in mass DMa does not necessarily correspond to the implied change, DMs, due the combined effect of the adoption of light materials, more accessories and weight changes resulting from safety improvements. A second implied fuel consumption (FCsm) is calculated based on DMa:
FC sm ¼ FC 02 þ DFCðDMa Þ þ DFCðDV sm Þ þ DFCðDC D As Þ
ð2Þ
Relying on actual changes in both engine capacity (DVa) and aerodynamic resistance (DCDAa) an alternative implied fuel consumption can be derived (Eq. (3)) by successively adjusting to actual values to get inferred effects of changes in attributes, between 2002 and 2007, and can also be compared with the actual fuel consumption in 2007 (FC07). Both FC02 and FC07 are sales weighted averages of the reported fuel consumption based on the New European Driving Cycle where diesel values have been recalculated to gasoline equivalents.
FC sa ¼ FC 02 þ DFCðDMa Þ þ DFCðDV a Þ þ DFCðDC D Aa Þ
ð3Þ
To calculate fuel consumption sensitivity regarding the attributes we use an analytical expression developed by An and Ross (1993). We assume that fuel consumption may be approximated by a linear function of the engine power output:
Pfuel ¼ aN þ bP b
ð4Þ
where Pfuel is the rate of fuel use [kW], a is the engine friction characteristics [kJ/rev], N is the engine speed [rps, revolutions per second], b is the inverted thermal efficiency of the engine [–], and Pb is the brake power output [kW]. The variables N and Pb are then expressed in terms of vehicle and test drive cycle characteristics:
FC TDC ¼ a0 V
1 t st 1 1 1 Pace þb N pwr ð1 tbr Þ þ cNidle t br þ q C D Akv 2r þ C R Mg þ bM v 2p n þ 2 v r 1 t st e 2 vr 1 tst
ð5Þ
The vehicle parameters are: FCTDC: fuel energy consumed per unit distance of the test drive cycle [J/m]; a0 : a/V; V: cylinder displacement volume [l]; Npwr: average engine speed during the test drive cycle for the time when the engine delivers power [rps]; Nidle: average engine speed when the engine is idling [rps]; e: drive train efficiency; CD: air drag coefficient; A: frontal surface area [m2]; CR: rolling resistance coefficient; M: vehicle mass in test drive cycle [kg]; M: effective mass, M plus corrections for effects due to rotational inertia [kg]; Pace: average power requirement for vehicle accessories [kW]. r : average running speed [m/s]; v 2p : the root-mean-square peak velocity over the driving Driving cycle parameters are: v cycle [m/s]; tbr: the percentage of time for brake use while the car is running; tst: time that the vehicle is stopped in percent [s]; n: the number of stops per unit distance; c: compensation parameter for cold start; k: approximation parameter to handle air resistance (set to 1 or 2); b: adjustment parameter to better capture braking energy. Further, q is the air density [kg/ m3], and g is acceleration due to gravity [m/s2]. Parameters a0 and b are calculated for gasoline cars and diesel cars separately. For ease of exposition, analysis is a percentage-based decomposition of fuel consumption. The drawbacks are that the numerical results depend on method, base year, and order of calculation steps. Since the new car fleet is heterogeneous there are also uncertainties based on assumptions made and segment chosen to calculate relationships; although these are likely small. Our computations do not imply any conclusions on causality between technological development and improved service parameters. 3
Energy values used: 9.04 MW h/m3 for petrol and 9.96 MW h/m3 for diesel. E85 consists of 85% ethanol and 15% gasoline. 5 With mass we mean curb weight plus the 75 kg for the driver’s weight. 6 The index was developed by Autograph-bilfakta AB. The index is a sum in mm of nine passenger volume length measures, all adhering to the international SAE-standard. 4
64
F. Sprei, S. Karlsson / Transportation Research Part D 18 (2013) 62–69
Table 1 Sales-weighted averages for gasoline cars, diesel cars and all cars sold in 2002 and 2007. Gasoline
Passenger space index (mm) Acceleration time 0–100 km/h (s) Weight (kg) Maximum power (kW) Cylinder volume (cc) CD frontal area (m2)
Diesel
All
2002
2007
% Change
2002
2007
% Change
2002
2007
% Change
9124 10.16 1406 101.2 1956 0.670
9147 10.20 1415 103.9 1883 0.695
0.25 0.39 0.63 2.8 3.73 3.67
9277 10.80 1610 102.2 2184 0.688
9402 10.00 1636 110.5 2105 0.728
1.34 7.42 1.58 8.1 3.65 5.89
9133 10.21 1417 101.2 1969 0.671
9233 10.13 1490 105.9 1958 0.706
1.09 0.75 5.12 4.7 0.55 5.16
3. Trends in the car market 3.1. Overview Traditionally the Swedish new car fleet has been dominated by cars with gasoline engines and manual gear boxes. During the last years a change toward a more differentiated new car fleet, considering engine choice, has emerged. Between 2002 and 2007 manual gasoline vehicles went from having almost 80% of the market share to less than 50%. Diesel engines have had approximately 5% of the market share, while in 2007 the share was over 30%. The penetration of diesel cars in the market has occurred earlier in Europe (Schipper et al., 2002) and in 2007 the share in EU27 was roughly 50% (European Commission, 2009). Vehicles that can be driven on an alternative fuel such as ethanol or gas were another novelty on the Swedish car market. In 2007 12% of the sold cars could be driven on an alternative fuel, mainly E85. The large share of flex-fuel vehicles is unique for Sweden in a European context. Since the seventies there has been a continuous trend of enhancement of important car parameters The average new car weighs now almost 1500 kg, an increase by 30% since 1975. Heavier cars have been a trend in Europe as well and the yearly increase has fluctuated around 1% (European Commission, 2009). The increase of the car weight in Sweden is coupled with larger passenger space: 9233 in the measured index compared to 8607 in 1975. Performance of the vehicles has increased the most in relative terms since 1975 and average maximum power is now over 100 kW. Improved acceleration capacity has led to shorter acceleration times and an average new sold vehicle in Sweden now accelerates from 0 to 100 km/h in 10.1 s compared to 15.2 s in 1975. At the same time as the average car has become larger and more powerful, fuel consumption has decreased with 28% since 1975. Fuel consumption remained more or less unaltered for the new car fleet until 2004 and has since then decreased, coinciding with a larger share of diesel cars. 3.2. Attributes by fuel type 2002–2007 Minor increases in parameters occurred between 2002 and 2007 in gasoline and alternative fuelled vehicles (Table 1).7 Passenger volume space and weight increased with less than 1%, while aerodynamic resistance increased with about 3%. Maximum power increased with almost 3% but the average acceleration time was almost unaltered. From a technical development point of view we can note that the engine capacity was reduced. Fuel consumption for gasoline was steady from 2000 to approximately 2004 and decreased slightly thereafter, Fig. 2. 3.3. Diesel cars The average values of all the attributes of diesel cars are higher than the average gasoline car and the sales-weighted average of all new cars sold. In average a diesel car was in 2007 16% heavier than a gasoline car, 3% larger in passenger space index and had a 6% more powerful engine. This has not always been the case, especially not for performance related attributes such as maximum power and acceleration time. The relative increase for diesel cars the last years has also been higher compared to gasoline cars (including flex-fuel and gas cars) for all parameters and most markedly for the performance related ones. Size related parameters such as passenger space index and weight have increased between 1% and 2%, between 2002 and 2007. The average cylinder volume of diesel cars has decreased with about the same relative magnitude as for all the gasoline cars. Between 2000 and 2003 the average fuel consumption for diesel cars actually increased, coinciding with an increase in average attribute values. First after 2005 we see a lowering of average fuel consumption, Fig. 2. Diesel cars are most often found in the upper segments of the market. The share of diesels in the premium, large and SUV segments of the market in 2007 was 41%, while for small cars the share was only 13%. The concentration of diesel in the upper segments is not exclusive for Sweden: Meyer and Wessely (2009) find the same development in Austria, and Schipper et al. (2002) for a number of European countries. 7
For vehicles that can only been driven by gasoline there has not been any increase at all.
F. Sprei, S. Karlsson / Transportation Research Part D 18 (2013) 62–69
65
Diesels engines are generally more energy efficient than gasoline engines even if the difference is diminishing through technologies such as turbocharging, direct gasoline injection, etc. The average diesel car is at the same time larger and has a more powerful engine compared to the average gasoline car. In order to better understand the differences in fuel consumption we perform some calculations of average fuel consumption, using Eq. (5), 8 and shifting the characteristics of the diesel cars with those of the gasoline cars. Given engine characteristics of an average gasoline engine, but weight and cylinder volume for an average diesel, the fuel consumption for the average diesel car would increase by 11%. If instead we would use an average diesel engine, but the weight and cylinder volume of an average gasoline car, the average fuel consumption would decrease by 20%. The larger share of diesel cars in the Swedish new car fleet can be seen as a combination of policy and market push. While diesel cars have not received as clear policy advantages as flex-fuel cars have, there have been signals that they are more on the level. Between the 1st of July 2006 and 31st of December 2007 diesel cars with particle filters received a one-time reduction of the annual registration tax of 6000 SEK ($850). In the formula to calculate the yearly registration tax based on CO2 emissions the tax for diesel cars is multiplied with a factor of 3.3 compared to the gasoline car with the same emissions (to compensate for lower fuel tax). Despite this factor the reform of the yearly registration tax still implied a lowering of taxes for many diesel car models. Since 2007 this factor has also been lowered, while the tax on diesel fuel has increased. The objective is to equalise the taxation of diesel and gasoline vehicles (Swedish National Tax Agency, 2011). In parallel to these tax reforms there has been a shift in the market: there are now more diesel models available and the performance of some diesel cars even exceeds most gasoline cars. 3.4. Flex-fuel cars Flex-fuel cars, i.e., cars that can be driven on a blend of ethanol and gasoline, do not differ significantly from pure gasoline cars or the average car when it comes to their size and weight, but have on average more powerful engines. The difference is mainly determined by the models available on the market. While the number of models has been increasing there were still no small flex-fuel cars available and very few in the middle segment (six models). From 2004 the average fuel consumption for flex-fuel cars has increased mainly due to the introduction of more vehicle models in the upper segment with higher fuel consumption. In 2007 the average fuel consumption for flex-fuel cars was higher than for gasoline cars. The explanation is again found in the lack of availability of low-consuming models and not a worse efficiency per se. In comparison with gasoline cars with the same power and weight, the fuel consumption does not deviate. Policy pressure concerning fuel efficiency in flex-fuel cars has been weak (Sprei, 2009). The rebates and subsidies for these cars have either not had any specification on fuel efficiency or such a generous one that all car models have been eligible. Flex-fuel cars, despite higher fuel consumption represent a possibility to reduce carbon emission. The environmental gain is however dependent on the cars being fuelled with E85 and not gasoline. The actual E85 refuelling rate has been very dependent on the relative price difference between E85 and gasoline, e.g., sales of E85 plummeted at the end of 2008 when gasoline prices fell (Swedish Petroleum Institute, 2009). The Swedish EPA and Road Administration have also reduced the presumed CO2 saving potential of flex-fuel cars compered to gasoline cars,9 from 56% to 22% mainly due to the decreasing sales of E85 fuel (Swedish Road Administration, 2009). 3.5. Attribute trends 2007–2010 The shift of the car fleet from a dominance of gasoline cars to diesel cars has continued into 2010, when 50% of new cars sold were diesel fuelled. The share of flex-fuel cars was in 2010 roughly the same as in 2007, however this is a fall in respect to the 19% of the sales in 2009 (Swedish Road Administration, 2010). From the dataset available a levelling out of the attributes seems to have happened. The average car was slightly heavier (1501 kg), however the increase is so small that it is within the error term. Both gasoline cars and diesel cars had on average somewhat lower weights. However, since diesel cars were still roughly 16% heavier than gasoline cars and there has been a continuous shift toward diesels, this has not manifested in a decrease in the average weight of the fleet. Cylinder volumes have decreased in both diesel and gasoline cars. In total the average cylinder volume decreased by 9% from 1958 cc in 2007 to 1788 cc in 2010. While the dataset does not cover maximum power in 2010, according to ACEA data the average power of the cars sold in Sweden, January through August in 2010 was 102 kW (ACEA, 2011). There is an observable shift is the distribution of CO2 emissions (Fig. 1). Noticeable is the hump below 120 g CO2/km, which is the upper limit for being labeled as an environmental friendly car and be exempted from the annual circulation tax for 5 years. There were in 2010 over 170 models having emissions lower than 120 g CO2/km, and 20 of these even had emissions under 100 g CO2/km. Between 2002 and 2007 cars with emissions under 120 g CO2/km became more mainstream cars; this development has continued to 2010 and has stretched into cars with emissions under 100 g CO2/km. The 8 The relationship between weight and passenger space (based on regression analysis) is similar for gasoline cars and diesel cars and thus weight is used as a proxy for passenger space. The relationships between power and cylinder volume are also similar for the engine types and thus the cylinder volume in Eq. (5) can be seen an approximation for maximum engine power. 9 This reduction potential is used to calculate the climate impact of new sold cars in Sweden (see http://www.naturvardsverket.se/bilindex).
66
F. Sprei, S. Karlsson / Transportation Research Part D 18 (2013) 62–69
0.5 0.4 0.3 0.2 0.1 0 0
100
200
300
400
500
600
g CO2/km 2010
2007
2002
Fig. 1. Distribution of tail-pipe CO2 emissions (g/km) for new cars sold in 2002, 2007 and 2010.
variety of models under 100 g CO2/km is rather large, including Volvo V50 (a station wagon considered a family car) weighing 1500 kg and with a power of 85 kW.
3.6. Engine downsizing Estimates for saved fuel consumption presume that mass is kept constant, which has not been the case for the Swedish new car fleet. So while specific power (SP) for the average car has increased from 51 kW/l in 2002 to 54 kW/l in 2007, the question is how this technical downsizing has been used. For a given car weight, the engine becomes smaller (an absolute downsizing) if the relative increase in specific power outweighs the relative increase in the power to weight ratio. Fig. 2 gives for the available car models for 2002 and 2007 the distribution of specific power (SP) with car power to weight ratio (PWR), which is roughly proportional to acceleration capacity. The vertical axis, specific power [kW/l], corresponds to technical downsizing and the horizontal axis to a downsizing in performance such that the combined downsizing is maximized for a shift toward the upper left corner in the figure. We see that for available models, generally higher specific power also implies higher performance (PWR) even if a certain bifurcation can be seen for car-models over 130 kW/ton. Sprei and Karlsson (2008) found that the line corresponding to the points where SP is numerically equal to PWR, that is, one litre of cylinder volume per ton of car, to be a rough lower ‘limit’ for downsizing among the models then available (2002 and older). This is still the case for 2007 (see Fig. 2). Only a limited number of models cross the line and thus a full downsizing, i.e., combining engines with high specific power and low performance cars, still represented an unexploited saving potential in the new car fleet. The balance of weight and maximum power between 2007 and 2010, combined with reduced cylinder volume has resulted in a downsizing of engines, with roughly 20% of the car models in 2010 having an engine capacity per weight 1 l/ ton, the lowest value in the previous years. In Fig. 3 we can see, for 2010 compared to previous years, a shift toward smaller engine capacity per weight both in the models available and in cars actually sold.
Fig. 2. Distribution of specific power with car power to weight ratio for car models available in 2002 and 2007. Note: The straight line where SP is numerically equal to PWR is plotted.
F. Sprei, S. Karlsson / Transportation Research Part D 18 (2013) 62–69
67
Fig. 3. Distributions of engine capacity per car weight (l/ton) for car models available on the market (left) and cars sold (right) for 2002, 2007 and 2010.
4. Attribute trends on fuel consumption 4.1. Developments 2002–2007 Comparing the effect of changes in passenger space on weight (DMs) is compared with actual change (DMa), we find that gasoline cars have not increased much in size and the effect on weight is minor. The increased size of the diesel cars would have implied an increase in mass by roughly 6% while the actual increase in weight was less than 2%. For all the cars larger passenger space would have implied a 6% increase in weight while the actual change was 5%. For all three groups of cars there is a net reduction of weight, which is the balance between an increase of weight due to other services such as more comfort, improvements in safety and more/heavier accessories, and the weight savings achieved from the application of lighter materials and design. The data does not render possible a further analysis of this balance or an explanation of why the net mass decrease is larger for diesel cars. In terms of implied changes in engine capacity, (DVs and DVsm), gasoline cars have seen little change in service attributes (passenger space, weight and acceleration capacity), which also gives a small change in implied engine capacity (plus 2%). In reality the average engine capacity decreased by almost 4%. The changes in diesel vehicles would result in an almost 14% larger engine capacity. Half of this increase can be attributed to larger vehicles and half to improved acceleration capacity. Even for this group of vehicles the actual change is 4%. For all cars sold, increased service parameters would have implied an increased engine capacity by 6% (almost all from larger vehicles) while the actual reduction was 1%. Observed technological improvement, specifically increased specific power, has in this case resulted in an actual reduction of engine capacity instead of an increase.
Fig. 4. Fuel consumption relative FC in 2002 due to larger passenger space, improved acceleration and air resistance allowing for mass and engine specific power changes.
68
F. Sprei, S. Karlsson / Transportation Research Part D 18 (2013) 62–69
There was no systematic difference in aerodynamic resistance between small cars and larger cars, implying that even if larger cars might have larger frontal area they were designed more aerodynamically, i.e., with a lower CD. From 2002 to 2007 the average CDA has increased and to be able to take this into consideration in our calculations, DCDAs is set equal to the change in the actual CDA (DCDAa). Taking into consideration all the implied effects of increased passenger space, acceleration and aerodynamic resistance we find that the fuel consumption (FCs) should have increased to 8.7 l/100 km, Fig. 4. Net reductions in mass give a slightly lower FC of 8.6 l/100 km (FCsm). The increase in specific power and the related decrease in engine capacity reduce FC to 8.4 l/ 100 km (FCsa). FCdies, 8.0 l/km, is calculated by presuming that the average diesel car maintains its attributes but has an average gasoline engine instead. The difference between FCsa and FCdies is a balance term containing improvements that the used method and data cannot identify. It can include technical improvements and changes such as higher thermal efficiency, reduced pump losses, reduced gear ratios, and lower rolling resistance. The difference between the actual value in 2007 and FCdies is the effect of the increased share of diesel engines. The actual value in 2007 (all fuel consumptions in gasoline equivalents) was 7.5 l/100 km. FCs is 6% higher than FC in 2002 and actual FC in 2007 was 8% lower suggesting that only 43% of the technological improvements in fuel consumption were directed toward increased services and performance. This runs counter to previous periods. A major part of the reduction in fuel use is due to a shift to diesel engines because, given the same service attributes as in 2007, but with gasoline engines only, the reduction in fuel consumption would have been only 2.4%, hence only 70% of the technological development would have been offset by ameliorated consumer amenities. 4.2. Use of the technological development in 2007–2010 The dataset for 2010 does not allow for as detailed analysis as in the previous years. What can be observed is that there has been a shift toward lower fuel consumption. Average fuel consumption in 2010 in gasoline equivalents was 6.5 l/100 km, 13% lower than in 2007. From the data available we can infer that the consumer amenities have hardly increased, given that passenger space index has followed the development of weight (which was almost the case for all cars between 2002 and 2007). Considering that maximum power seems to have stalled and the power to weight ratio has remained fairly constant or slightly decreased, it is plausible to say that acceleration capacity is also similar. Applying this, 77% of the utilized technological development has resulted in actual savings in fuel consumption between 2002 and 2010.10 We have seen that the shift to diesel has been the single largest factor behind lower fuel consumption in the period 2002– 2007, Fig. 4. When using Eq. (5) with the estimated a0 and b coefficients for gasoline cars in 2007, combined with the characteristics of the average new car in 2010, we calculate an implied fuel consumption of 7.5 l/100 km. This gives an estimate of the fuel consumption in 2010, had there not been any diesel engines. This estimated fuel consumption is most probably an overestimate since we can expect that the average gasoline engine has become more efficient between 2007 and 2010. But still, an implied fuel consumption of 7.5 l/100 km suggests that 57% of the development has resulted in lower fuel consumption. While this is smaller than the change including the further shift to diesel, this is still a substantial trend break compared to earlier results. 5. Conclusions Swedish car buyers have shifted toward diesel and flex-fuel cars. Sprei and Wickelgren (2011) found that car buyers often wanted at least the same size and performance as previous cars, but with lower fuel consumption, and thus shifting fuels becomes an attractive option. The increased penetration of diesel cars has been coupled with improvements in performance of these vehicles implying that a shift to diesel cars has not meant worse driveability, rather the opposite. There are environmental gains in terms of fuel consumption and CO2 emissions in switching to diesel, but there are risks of ‘‘rebound’’ effects leading to longer driving distances due to lower running costs (Schipper, 2009), although these are difficult to determine.11 Acknowledgments Funding from the AES program at the Swedish Energy Agency and The Swedish Research Council Formas is gratefully acknowledged. Thanks to Bilfakta AB, Swedish Road Administration and Bengt Ogren at Volvo Car Corporation for providing data, and Johan Torén and Linus Helming for gathering it. References ACEA, 2011. New Passenger Car Registrations – Breakdown by Specification Average Power. (accessed 02.09.11). An, F., Ross, M., 1993. A Model of Fuel Economy and Driving Patterns. SAE Paper 930328. 10
This is a large shift compared to Sprei et al. (2008) finding for the development between 1985 and 2002. For flex-fuel cars fuel rates have been found important; in 2009 the sales of E85 decreased despite the increased sales of flex-fuel vehicles (Svenska Petroleum and Biodrivmedel Institutet, 2011). 11
F. Sprei, S. Karlsson / Transportation Research Part D 18 (2013) 62–69
69
European Commission, 2009. Report from the Commission to the European Parliament and the Council – Monitoring the CO2 Emissions from New Passenger Cars in the EU: Data for The Year 2008. COM/2009/0713 Final. Meyer, I., Wessely, S., 2009. Fuel efficiency of the Austrian passenger vehicle fleet – analysis of trends in the technological profile and related impacts on CO2 emissions. Energy Policy 37, 3779–3789. Schipper, L., Marie-Lilliu, C., Fulton, L., 2002. Diesels in Europe. Analysis of characteristics, usage patterns, energy savings and CO2 emission implications. Journal of Transport Economics and Policy 36, 305–340. Schipper, L., 2009. Automobile Fuel; Economy and CO2 Emissions In Industrialized Countries: Troubling Trends through 2005/6. Embarq, World Resource Institute. Svenska Petroleum and Biodrivmedel Institutet, 2011. Förnybara drivmedel. (accessed 26.10.11). Sprei, F., Karlsson, S., 2008. The role of market and technical downsizing in reducing carbon emissions from the Swedish new car fleet. Energy Efficiency 1, 107–120. Sprei, F., Karlsson, S., Holmberg, J., 2008. Better performance or lower fuel consumption. Technological development in the Swedish new-car fleet 1975– 2002. Transportation Research Part D: Transport and Environment 13, 75–85. Sprei, F., 2009. Vilka Styrmedel har ökat Personbilarnas Energieffektivitet i Sverige? FRT 2009:1. Physical Resource Theory. Chalmers University of Technology, Sweden. Sprei, F., Wickelgren, M., 2011. Requirements for change in new car buying practices – observations from Sweden. Energy Efficiency 4, 193–207. Swedish National Tax Agency, 2011. Fordonsskatt. (accessed 26.10.11). Swedish Petroleum Institute, 2009. Statistics Sweden. Volumes. Deliveries fuel, per month. (accessed 28.12.09). Swedish Road Administration, 2009. Uppdaterade Reduktionsvärden för Etanol – Och Gasfordon Till Bilindex, 2009–10-06. Swedish Road Administration, Borlänge. Swedish Road Administration, 2010. Bilindex 2010 – Bilaga Län Och Kommuner. Swedish Road Administration, Borlänge. Swedish Transport Administration, 2011. Bilindex 2011 – Bilaga Län Och Kommuner. Swedish Transport Administration, Borlänge.
Contents lists available at SciVerse ScienceDirect
Transportation Research Part D journal homepage: www.elsevier.com/locate/trd
Shifting fuels, downsizing or both? The Swedish example Frances Sprei a,b,⇑, Sten Karlsson b a b
Stanford University, Stanford, CA, USA Chalmers University of Technology, Gothenburg, Sweden
a r t i c l e
i n f o
Keywords: Specific fuel consumption Downsizing Service attributes Performance attributes Diesel cars Flex-fuel cars
a b s t r a c t The paper looks at changes in Sweden’s new car fleet between 2002 and 2010. Between 2002 and 2007 consumer amenities such as acceleration capacity and passenger space continued to increase while fuel consumption steadily decreased. During these years the main technological and market change was a shift toward diesel and flex-fuel ethanol vehicles. After 2007 the average weight and power of the vehicles were more or less constant, while fuel consumption decreased by 13% between 2007 and 2010. The developments after 2007 suggest that 77% of the technological development between 2002 and 2010 resulted in reductions in fuel consumption compares to previous years when 35% of any technology change resulted in a net reductions. The shift can partly be attributed to the increased share of diesels and an engine downsizing. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Between 1985 and 2002, only 65% of the technological development allowing more efficient fuel consumption and lower CO2 emissions in the Swedish car fleet was offset by improvements in consumer ‘‘amenities’’ such as larger vehicles and improved acceleration capacity. Since 2002 there has been increased public and governmental awareness and willingness for action. The result has been a policy shift towards reducing the carbon impact of the new car fleet. The annual weight based registration tax changed in 2006 to be based on CO2 emissions. There has been a rebate program for ‘‘environment-friendly cars’’, i.e., vehicles emitting less than 120 g CO2/km, flex-fuel cars,1 gas driven cars and electric vehicles. Flex-fuel and gas driven cars have also received parking subsidies in about 40 municipalities and are exempted from congestion charging in Stockholm. Carbon and energy taxing of fossil fuels has continued to increase and hybrids, flex-fuel cars and gas driven cars are given reduction on the fringe taxing for company provided cars. Here we investigate the new car market from 2002 to 2007, as well as to 2010.
2. Method Data for the calculations are based two sources; sales statistics per vehicle model for 2007 and vehicle model characteristics with over 50 parameters per model for 1575 car models on sale during 2007. The former were collected from the Swedish Road Administration,2 and vehicle model parameters from the private company Autograph-bilfakta AB. Additional data was collected from car manufacturers and other databases when this was needed. In total 97% of the cars registered are covered by ⇑ Corresponding author at: Stanford University, Stanford, CA, USA. E-mail address: [email protected] (F. Sprei). Flex-fuel cars can be driven on any mixture of pure gasoline and E85 (85% gasoline and 15% ethanol). 2 The Swedish Road Administration was in 2010 transformed into the Swedish Transport Administration. Both names are used here, depending on when data was gathered. 1
1361-9209/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.trd.2012.09.006
F. Sprei, S. Karlsson / Transportation Research Part D 18 (2013) 62–69
63
the data. Privately imported cars and cars registered as light trucks are not included in the database. Data from1975, 1985, 1995 and 2002 are from similar databases constructed for these years (Sprei et al., 2008). Data for 2010 are based on the EU-monitoring database and the Swedish Transport Administration (2011). These databases do not cover the same number of parameters, as in previous years, thus not allowing the same analysis. They do, however, cover the same share of new registered vehicles. Sales-weighted averages and distributions are calculated to give an overview of the new car market. This was done for: all the cars sold, gasoline cars, diesel cars and flex-fuel cars. Gasoline cars include flex-fuel vehicles that can be driven with ethanol or gas. Fuel consumption (FC) is given in gasoline equivalents,3 and for flex-fuel cars is based on their pure gasoline consumption; because of the lower energy density of ethanol in E854 consumption in l/100 km will be higher for it. The effect of service and performance attributes on fuel consumptions is calculated stepwise. First, we calculate implied values for mass5 (Ms), engine capacity (DVs) and aerodynamic drag resistance (CDAs) due to increased acceleration and passenger compartment space (measured by index6 I). Changes in FC are computed from fuel consumption sensitivities derived from an analytical expression relating fuel consumption and driving cycle characteristics (Eqs. (4) and (5)). The implied fuel consumption due to improved service attributes, considering no technical improvement is:
FC s ¼ FC 02 þ DFCðDM s Þ þ DFCðDVsÞ þ DFCðDC D As Þ
ð1Þ
The change in mass DMa does not necessarily correspond to the implied change, DMs, due the combined effect of the adoption of light materials, more accessories and weight changes resulting from safety improvements. A second implied fuel consumption (FCsm) is calculated based on DMa:
FC sm ¼ FC 02 þ DFCðDMa Þ þ DFCðDV sm Þ þ DFCðDC D As Þ
ð2Þ
Relying on actual changes in both engine capacity (DVa) and aerodynamic resistance (DCDAa) an alternative implied fuel consumption can be derived (Eq. (3)) by successively adjusting to actual values to get inferred effects of changes in attributes, between 2002 and 2007, and can also be compared with the actual fuel consumption in 2007 (FC07). Both FC02 and FC07 are sales weighted averages of the reported fuel consumption based on the New European Driving Cycle where diesel values have been recalculated to gasoline equivalents.
FC sa ¼ FC 02 þ DFCðDMa Þ þ DFCðDV a Þ þ DFCðDC D Aa Þ
ð3Þ
To calculate fuel consumption sensitivity regarding the attributes we use an analytical expression developed by An and Ross (1993). We assume that fuel consumption may be approximated by a linear function of the engine power output:
Pfuel ¼ aN þ bP b
ð4Þ
where Pfuel is the rate of fuel use [kW], a is the engine friction characteristics [kJ/rev], N is the engine speed [rps, revolutions per second], b is the inverted thermal efficiency of the engine [–], and Pb is the brake power output [kW]. The variables N and Pb are then expressed in terms of vehicle and test drive cycle characteristics:
FC TDC ¼ a0 V
1 t st 1 1 1 Pace þb N pwr ð1 tbr Þ þ cNidle t br þ q C D Akv 2r þ C R Mg þ bM v 2p n þ 2 v r 1 t st e 2 vr 1 tst
ð5Þ
The vehicle parameters are: FCTDC: fuel energy consumed per unit distance of the test drive cycle [J/m]; a0 : a/V; V: cylinder displacement volume [l]; Npwr: average engine speed during the test drive cycle for the time when the engine delivers power [rps]; Nidle: average engine speed when the engine is idling [rps]; e: drive train efficiency; CD: air drag coefficient; A: frontal surface area [m2]; CR: rolling resistance coefficient; M: vehicle mass in test drive cycle [kg]; M: effective mass, M plus corrections for effects due to rotational inertia [kg]; Pace: average power requirement for vehicle accessories [kW]. r : average running speed [m/s]; v 2p : the root-mean-square peak velocity over the driving Driving cycle parameters are: v cycle [m/s]; tbr: the percentage of time for brake use while the car is running; tst: time that the vehicle is stopped in percent [s]; n: the number of stops per unit distance; c: compensation parameter for cold start; k: approximation parameter to handle air resistance (set to 1 or 2); b: adjustment parameter to better capture braking energy. Further, q is the air density [kg/ m3], and g is acceleration due to gravity [m/s2]. Parameters a0 and b are calculated for gasoline cars and diesel cars separately. For ease of exposition, analysis is a percentage-based decomposition of fuel consumption. The drawbacks are that the numerical results depend on method, base year, and order of calculation steps. Since the new car fleet is heterogeneous there are also uncertainties based on assumptions made and segment chosen to calculate relationships; although these are likely small. Our computations do not imply any conclusions on causality between technological development and improved service parameters. 3
Energy values used: 9.04 MW h/m3 for petrol and 9.96 MW h/m3 for diesel. E85 consists of 85% ethanol and 15% gasoline. 5 With mass we mean curb weight plus the 75 kg for the driver’s weight. 6 The index was developed by Autograph-bilfakta AB. The index is a sum in mm of nine passenger volume length measures, all adhering to the international SAE-standard. 4
64
F. Sprei, S. Karlsson / Transportation Research Part D 18 (2013) 62–69
Table 1 Sales-weighted averages for gasoline cars, diesel cars and all cars sold in 2002 and 2007. Gasoline
Passenger space index (mm) Acceleration time 0–100 km/h (s) Weight (kg) Maximum power (kW) Cylinder volume (cc) CD frontal area (m2)
Diesel
All
2002
2007
% Change
2002
2007
% Change
2002
2007
% Change
9124 10.16 1406 101.2 1956 0.670
9147 10.20 1415 103.9 1883 0.695
0.25 0.39 0.63 2.8 3.73 3.67
9277 10.80 1610 102.2 2184 0.688
9402 10.00 1636 110.5 2105 0.728
1.34 7.42 1.58 8.1 3.65 5.89
9133 10.21 1417 101.2 1969 0.671
9233 10.13 1490 105.9 1958 0.706
1.09 0.75 5.12 4.7 0.55 5.16
3. Trends in the car market 3.1. Overview Traditionally the Swedish new car fleet has been dominated by cars with gasoline engines and manual gear boxes. During the last years a change toward a more differentiated new car fleet, considering engine choice, has emerged. Between 2002 and 2007 manual gasoline vehicles went from having almost 80% of the market share to less than 50%. Diesel engines have had approximately 5% of the market share, while in 2007 the share was over 30%. The penetration of diesel cars in the market has occurred earlier in Europe (Schipper et al., 2002) and in 2007 the share in EU27 was roughly 50% (European Commission, 2009). Vehicles that can be driven on an alternative fuel such as ethanol or gas were another novelty on the Swedish car market. In 2007 12% of the sold cars could be driven on an alternative fuel, mainly E85. The large share of flex-fuel vehicles is unique for Sweden in a European context. Since the seventies there has been a continuous trend of enhancement of important car parameters The average new car weighs now almost 1500 kg, an increase by 30% since 1975. Heavier cars have been a trend in Europe as well and the yearly increase has fluctuated around 1% (European Commission, 2009). The increase of the car weight in Sweden is coupled with larger passenger space: 9233 in the measured index compared to 8607 in 1975. Performance of the vehicles has increased the most in relative terms since 1975 and average maximum power is now over 100 kW. Improved acceleration capacity has led to shorter acceleration times and an average new sold vehicle in Sweden now accelerates from 0 to 100 km/h in 10.1 s compared to 15.2 s in 1975. At the same time as the average car has become larger and more powerful, fuel consumption has decreased with 28% since 1975. Fuel consumption remained more or less unaltered for the new car fleet until 2004 and has since then decreased, coinciding with a larger share of diesel cars. 3.2. Attributes by fuel type 2002–2007 Minor increases in parameters occurred between 2002 and 2007 in gasoline and alternative fuelled vehicles (Table 1).7 Passenger volume space and weight increased with less than 1%, while aerodynamic resistance increased with about 3%. Maximum power increased with almost 3% but the average acceleration time was almost unaltered. From a technical development point of view we can note that the engine capacity was reduced. Fuel consumption for gasoline was steady from 2000 to approximately 2004 and decreased slightly thereafter, Fig. 2. 3.3. Diesel cars The average values of all the attributes of diesel cars are higher than the average gasoline car and the sales-weighted average of all new cars sold. In average a diesel car was in 2007 16% heavier than a gasoline car, 3% larger in passenger space index and had a 6% more powerful engine. This has not always been the case, especially not for performance related attributes such as maximum power and acceleration time. The relative increase for diesel cars the last years has also been higher compared to gasoline cars (including flex-fuel and gas cars) for all parameters and most markedly for the performance related ones. Size related parameters such as passenger space index and weight have increased between 1% and 2%, between 2002 and 2007. The average cylinder volume of diesel cars has decreased with about the same relative magnitude as for all the gasoline cars. Between 2000 and 2003 the average fuel consumption for diesel cars actually increased, coinciding with an increase in average attribute values. First after 2005 we see a lowering of average fuel consumption, Fig. 2. Diesel cars are most often found in the upper segments of the market. The share of diesels in the premium, large and SUV segments of the market in 2007 was 41%, while for small cars the share was only 13%. The concentration of diesel in the upper segments is not exclusive for Sweden: Meyer and Wessely (2009) find the same development in Austria, and Schipper et al. (2002) for a number of European countries. 7
For vehicles that can only been driven by gasoline there has not been any increase at all.
F. Sprei, S. Karlsson / Transportation Research Part D 18 (2013) 62–69
65
Diesels engines are generally more energy efficient than gasoline engines even if the difference is diminishing through technologies such as turbocharging, direct gasoline injection, etc. The average diesel car is at the same time larger and has a more powerful engine compared to the average gasoline car. In order to better understand the differences in fuel consumption we perform some calculations of average fuel consumption, using Eq. (5), 8 and shifting the characteristics of the diesel cars with those of the gasoline cars. Given engine characteristics of an average gasoline engine, but weight and cylinder volume for an average diesel, the fuel consumption for the average diesel car would increase by 11%. If instead we would use an average diesel engine, but the weight and cylinder volume of an average gasoline car, the average fuel consumption would decrease by 20%. The larger share of diesel cars in the Swedish new car fleet can be seen as a combination of policy and market push. While diesel cars have not received as clear policy advantages as flex-fuel cars have, there have been signals that they are more on the level. Between the 1st of July 2006 and 31st of December 2007 diesel cars with particle filters received a one-time reduction of the annual registration tax of 6000 SEK ($850). In the formula to calculate the yearly registration tax based on CO2 emissions the tax for diesel cars is multiplied with a factor of 3.3 compared to the gasoline car with the same emissions (to compensate for lower fuel tax). Despite this factor the reform of the yearly registration tax still implied a lowering of taxes for many diesel car models. Since 2007 this factor has also been lowered, while the tax on diesel fuel has increased. The objective is to equalise the taxation of diesel and gasoline vehicles (Swedish National Tax Agency, 2011). In parallel to these tax reforms there has been a shift in the market: there are now more diesel models available and the performance of some diesel cars even exceeds most gasoline cars. 3.4. Flex-fuel cars Flex-fuel cars, i.e., cars that can be driven on a blend of ethanol and gasoline, do not differ significantly from pure gasoline cars or the average car when it comes to their size and weight, but have on average more powerful engines. The difference is mainly determined by the models available on the market. While the number of models has been increasing there were still no small flex-fuel cars available and very few in the middle segment (six models). From 2004 the average fuel consumption for flex-fuel cars has increased mainly due to the introduction of more vehicle models in the upper segment with higher fuel consumption. In 2007 the average fuel consumption for flex-fuel cars was higher than for gasoline cars. The explanation is again found in the lack of availability of low-consuming models and not a worse efficiency per se. In comparison with gasoline cars with the same power and weight, the fuel consumption does not deviate. Policy pressure concerning fuel efficiency in flex-fuel cars has been weak (Sprei, 2009). The rebates and subsidies for these cars have either not had any specification on fuel efficiency or such a generous one that all car models have been eligible. Flex-fuel cars, despite higher fuel consumption represent a possibility to reduce carbon emission. The environmental gain is however dependent on the cars being fuelled with E85 and not gasoline. The actual E85 refuelling rate has been very dependent on the relative price difference between E85 and gasoline, e.g., sales of E85 plummeted at the end of 2008 when gasoline prices fell (Swedish Petroleum Institute, 2009). The Swedish EPA and Road Administration have also reduced the presumed CO2 saving potential of flex-fuel cars compered to gasoline cars,9 from 56% to 22% mainly due to the decreasing sales of E85 fuel (Swedish Road Administration, 2009). 3.5. Attribute trends 2007–2010 The shift of the car fleet from a dominance of gasoline cars to diesel cars has continued into 2010, when 50% of new cars sold were diesel fuelled. The share of flex-fuel cars was in 2010 roughly the same as in 2007, however this is a fall in respect to the 19% of the sales in 2009 (Swedish Road Administration, 2010). From the dataset available a levelling out of the attributes seems to have happened. The average car was slightly heavier (1501 kg), however the increase is so small that it is within the error term. Both gasoline cars and diesel cars had on average somewhat lower weights. However, since diesel cars were still roughly 16% heavier than gasoline cars and there has been a continuous shift toward diesels, this has not manifested in a decrease in the average weight of the fleet. Cylinder volumes have decreased in both diesel and gasoline cars. In total the average cylinder volume decreased by 9% from 1958 cc in 2007 to 1788 cc in 2010. While the dataset does not cover maximum power in 2010, according to ACEA data the average power of the cars sold in Sweden, January through August in 2010 was 102 kW (ACEA, 2011). There is an observable shift is the distribution of CO2 emissions (Fig. 1). Noticeable is the hump below 120 g CO2/km, which is the upper limit for being labeled as an environmental friendly car and be exempted from the annual circulation tax for 5 years. There were in 2010 over 170 models having emissions lower than 120 g CO2/km, and 20 of these even had emissions under 100 g CO2/km. Between 2002 and 2007 cars with emissions under 120 g CO2/km became more mainstream cars; this development has continued to 2010 and has stretched into cars with emissions under 100 g CO2/km. The 8 The relationship between weight and passenger space (based on regression analysis) is similar for gasoline cars and diesel cars and thus weight is used as a proxy for passenger space. The relationships between power and cylinder volume are also similar for the engine types and thus the cylinder volume in Eq. (5) can be seen an approximation for maximum engine power. 9 This reduction potential is used to calculate the climate impact of new sold cars in Sweden (see http://www.naturvardsverket.se/bilindex).
66
F. Sprei, S. Karlsson / Transportation Research Part D 18 (2013) 62–69
0.5 0.4 0.3 0.2 0.1 0 0
100
200
300
400
500
600
g CO2/km 2010
2007
2002
Fig. 1. Distribution of tail-pipe CO2 emissions (g/km) for new cars sold in 2002, 2007 and 2010.
variety of models under 100 g CO2/km is rather large, including Volvo V50 (a station wagon considered a family car) weighing 1500 kg and with a power of 85 kW.
3.6. Engine downsizing Estimates for saved fuel consumption presume that mass is kept constant, which has not been the case for the Swedish new car fleet. So while specific power (SP) for the average car has increased from 51 kW/l in 2002 to 54 kW/l in 2007, the question is how this technical downsizing has been used. For a given car weight, the engine becomes smaller (an absolute downsizing) if the relative increase in specific power outweighs the relative increase in the power to weight ratio. Fig. 2 gives for the available car models for 2002 and 2007 the distribution of specific power (SP) with car power to weight ratio (PWR), which is roughly proportional to acceleration capacity. The vertical axis, specific power [kW/l], corresponds to technical downsizing and the horizontal axis to a downsizing in performance such that the combined downsizing is maximized for a shift toward the upper left corner in the figure. We see that for available models, generally higher specific power also implies higher performance (PWR) even if a certain bifurcation can be seen for car-models over 130 kW/ton. Sprei and Karlsson (2008) found that the line corresponding to the points where SP is numerically equal to PWR, that is, one litre of cylinder volume per ton of car, to be a rough lower ‘limit’ for downsizing among the models then available (2002 and older). This is still the case for 2007 (see Fig. 2). Only a limited number of models cross the line and thus a full downsizing, i.e., combining engines with high specific power and low performance cars, still represented an unexploited saving potential in the new car fleet. The balance of weight and maximum power between 2007 and 2010, combined with reduced cylinder volume has resulted in a downsizing of engines, with roughly 20% of the car models in 2010 having an engine capacity per weight 1 l/ ton, the lowest value in the previous years. In Fig. 3 we can see, for 2010 compared to previous years, a shift toward smaller engine capacity per weight both in the models available and in cars actually sold.
Fig. 2. Distribution of specific power with car power to weight ratio for car models available in 2002 and 2007. Note: The straight line where SP is numerically equal to PWR is plotted.
F. Sprei, S. Karlsson / Transportation Research Part D 18 (2013) 62–69
67
Fig. 3. Distributions of engine capacity per car weight (l/ton) for car models available on the market (left) and cars sold (right) for 2002, 2007 and 2010.
4. Attribute trends on fuel consumption 4.1. Developments 2002–2007 Comparing the effect of changes in passenger space on weight (DMs) is compared with actual change (DMa), we find that gasoline cars have not increased much in size and the effect on weight is minor. The increased size of the diesel cars would have implied an increase in mass by roughly 6% while the actual increase in weight was less than 2%. For all the cars larger passenger space would have implied a 6% increase in weight while the actual change was 5%. For all three groups of cars there is a net reduction of weight, which is the balance between an increase of weight due to other services such as more comfort, improvements in safety and more/heavier accessories, and the weight savings achieved from the application of lighter materials and design. The data does not render possible a further analysis of this balance or an explanation of why the net mass decrease is larger for diesel cars. In terms of implied changes in engine capacity, (DVs and DVsm), gasoline cars have seen little change in service attributes (passenger space, weight and acceleration capacity), which also gives a small change in implied engine capacity (plus 2%). In reality the average engine capacity decreased by almost 4%. The changes in diesel vehicles would result in an almost 14% larger engine capacity. Half of this increase can be attributed to larger vehicles and half to improved acceleration capacity. Even for this group of vehicles the actual change is 4%. For all cars sold, increased service parameters would have implied an increased engine capacity by 6% (almost all from larger vehicles) while the actual reduction was 1%. Observed technological improvement, specifically increased specific power, has in this case resulted in an actual reduction of engine capacity instead of an increase.
Fig. 4. Fuel consumption relative FC in 2002 due to larger passenger space, improved acceleration and air resistance allowing for mass and engine specific power changes.
68
F. Sprei, S. Karlsson / Transportation Research Part D 18 (2013) 62–69
There was no systematic difference in aerodynamic resistance between small cars and larger cars, implying that even if larger cars might have larger frontal area they were designed more aerodynamically, i.e., with a lower CD. From 2002 to 2007 the average CDA has increased and to be able to take this into consideration in our calculations, DCDAs is set equal to the change in the actual CDA (DCDAa). Taking into consideration all the implied effects of increased passenger space, acceleration and aerodynamic resistance we find that the fuel consumption (FCs) should have increased to 8.7 l/100 km, Fig. 4. Net reductions in mass give a slightly lower FC of 8.6 l/100 km (FCsm). The increase in specific power and the related decrease in engine capacity reduce FC to 8.4 l/ 100 km (FCsa). FCdies, 8.0 l/km, is calculated by presuming that the average diesel car maintains its attributes but has an average gasoline engine instead. The difference between FCsa and FCdies is a balance term containing improvements that the used method and data cannot identify. It can include technical improvements and changes such as higher thermal efficiency, reduced pump losses, reduced gear ratios, and lower rolling resistance. The difference between the actual value in 2007 and FCdies is the effect of the increased share of diesel engines. The actual value in 2007 (all fuel consumptions in gasoline equivalents) was 7.5 l/100 km. FCs is 6% higher than FC in 2002 and actual FC in 2007 was 8% lower suggesting that only 43% of the technological improvements in fuel consumption were directed toward increased services and performance. This runs counter to previous periods. A major part of the reduction in fuel use is due to a shift to diesel engines because, given the same service attributes as in 2007, but with gasoline engines only, the reduction in fuel consumption would have been only 2.4%, hence only 70% of the technological development would have been offset by ameliorated consumer amenities. 4.2. Use of the technological development in 2007–2010 The dataset for 2010 does not allow for as detailed analysis as in the previous years. What can be observed is that there has been a shift toward lower fuel consumption. Average fuel consumption in 2010 in gasoline equivalents was 6.5 l/100 km, 13% lower than in 2007. From the data available we can infer that the consumer amenities have hardly increased, given that passenger space index has followed the development of weight (which was almost the case for all cars between 2002 and 2007). Considering that maximum power seems to have stalled and the power to weight ratio has remained fairly constant or slightly decreased, it is plausible to say that acceleration capacity is also similar. Applying this, 77% of the utilized technological development has resulted in actual savings in fuel consumption between 2002 and 2010.10 We have seen that the shift to diesel has been the single largest factor behind lower fuel consumption in the period 2002– 2007, Fig. 4. When using Eq. (5) with the estimated a0 and b coefficients for gasoline cars in 2007, combined with the characteristics of the average new car in 2010, we calculate an implied fuel consumption of 7.5 l/100 km. This gives an estimate of the fuel consumption in 2010, had there not been any diesel engines. This estimated fuel consumption is most probably an overestimate since we can expect that the average gasoline engine has become more efficient between 2007 and 2010. But still, an implied fuel consumption of 7.5 l/100 km suggests that 57% of the development has resulted in lower fuel consumption. While this is smaller than the change including the further shift to diesel, this is still a substantial trend break compared to earlier results. 5. Conclusions Swedish car buyers have shifted toward diesel and flex-fuel cars. Sprei and Wickelgren (2011) found that car buyers often wanted at least the same size and performance as previous cars, but with lower fuel consumption, and thus shifting fuels becomes an attractive option. The increased penetration of diesel cars has been coupled with improvements in performance of these vehicles implying that a shift to diesel cars has not meant worse driveability, rather the opposite. There are environmental gains in terms of fuel consumption and CO2 emissions in switching to diesel, but there are risks of ‘‘rebound’’ effects leading to longer driving distances due to lower running costs (Schipper, 2009), although these are difficult to determine.11 Acknowledgments Funding from the AES program at the Swedish Energy Agency and The Swedish Research Council Formas is gratefully acknowledged. Thanks to Bilfakta AB, Swedish Road Administration and Bengt Ogren at Volvo Car Corporation for providing data, and Johan Torén and Linus Helming for gathering it. References ACEA, 2011. New Passenger Car Registrations – Breakdown by Specification Average Power.
This is a large shift compared to Sprei et al. (2008) finding for the development between 1985 and 2002. For flex-fuel cars fuel rates have been found important; in 2009 the sales of E85 decreased despite the increased sales of flex-fuel vehicles (Svenska Petroleum and Biodrivmedel Institutet, 2011). 11
F. Sprei, S. Karlsson / Transportation Research Part D 18 (2013) 62–69
69
European Commission, 2009. Report from the Commission to the European Parliament and the Council – Monitoring the CO2 Emissions from New Passenger Cars in the EU: Data for The Year 2008. COM/2009/0713 Final. Meyer, I., Wessely, S., 2009. Fuel efficiency of the Austrian passenger vehicle fleet – analysis of trends in the technological profile and related impacts on CO2 emissions. Energy Policy 37, 3779–3789. Schipper, L., Marie-Lilliu, C., Fulton, L., 2002. Diesels in Europe. Analysis of characteristics, usage patterns, energy savings and CO2 emission implications. Journal of Transport Economics and Policy 36, 305–340. Schipper, L., 2009. Automobile Fuel; Economy and CO2 Emissions In Industrialized Countries: Troubling Trends through 2005/6. Embarq, World Resource Institute. Svenska Petroleum and Biodrivmedel Institutet, 2011. Förnybara drivmedel.