TRANSPORT AND THE URBAN ENVIRONMENT Patrick Moriarty and Damon Honnery Department of Mechanical Engineering, Monash University-Caulfield Campus, 900 Dandenong Road, Caulfield East, Vic 3145, Australia Summary This paper examines ways of reducing transport-related emissions in the large cities of Australia and Asia. Although vehicle ownership in Asian cities is very low by Australian standards, high population densities and low levels of road space per vehicle lead to high levels of traffic congestion and air pollution. Attempting to provide more urban road space per vehicle will not help, as there exists a strong latent demand for car travel. The best way to cut urban transport’s environmental costs is to use electric rail transport. High congestion in Asian cities means that own-right-of-way rail will usually be faster than street vehicular travel--public or private. An analysis of Australian urban transport systems shows rail is only about twice as energy efficient as car travel. In Asian cities, however, existing rail travel is up to 10 times as energy efficient. Higher train seat occupancy rates and lower car fuel efficiencies, both a result of high population density and congestion, result in disproportionate reductions in air pollution and greenhouse gas emissions relative to private travel. Keywords: Asian cities, Australian cities, greenhouse gas emissions, electric rail transport, urban car travel, urban environment, urban density. 1. Introduction Traffic-related air pollution is a major cause of urban environmental deterioration, not only in the heavily car-oriented Australian cities, but also in the large cities of Asia, despite their low vehicle ownership levels (World Resources Institute (WRI) 1996, United Nations (UN) 1997). Further, many countries in Asia are encouraging domestic car industries (Taylor 1997), so car ownership levels there seem set to rise in the future. Ballew and Schnorbus (1993), for example, used South Korea’s experience of growth in income and car ownership as a model for China. On this basis they projected China’s car population to reach 36 million in the year 2010, noting that in South Korea “vehicle sales and all consumer durables exploded once per capita income reached approximately $3500 a year”. More generally, Japan has been used as a model for industrialising countries, especially in Asia. The main aim of this paper is to examine the environmental consequences of rising car ownership in the large cities of East and Southeast Asia, and to explore alternative solutions. We find that by comparison with Australia’s largest cities (Sydney and Melbourne), the densely-populated Asian cities have far less road space per vehicle, leading to higher traffic congestion and air pollution. Strong latent demand for cars in Asia means that attempts to solve congestion by road-building will seldom work. Instead, the combination of high densities and low traffic speeds mean that electric rail transport can provide competive travel times and local and global emissions which are only a fraction of those for car travel. 2. Density Comparison of Asian and Australian Cities Asia already contains some of the world’s largest cities. In most developing Asian countries, the size of large cities is set to increase even further, since the degree of urbanisation is presently low, but is projected to increase rapidly as industrialisation and economic growth occur (World Bank 1997). For example, the urban population of China is projected to increase from its 1995 level of 30% to 51% by 2020, representing an increase from 369 million to 756 million urban residents. Indonesia’s urban population is expected to reach 151 million, and the Philippines, 71 million by 2020 (UN 1997). The largest Asian cities are also much larger than those in Australia, a difference which will widen with further rapid urbanisation in Asia. Projected populations of the largest cities in Asia and Australia in 2015 are given below in Table 1. Table 1.Urban population projections for 2015. City Tokyo Shanghai Jakarta Beijing Tianjin Manila Seoul Bangkok Osaka Sydney* Melbourne* Pop (million). 28.7 23.4 21.2 19.4 17.0 14.7 13.1 10.6 10.6 4.7 4.0 * Assuming 1% annual compound growth 1996-2015. Sources: Australian Bureau of Statistics (ABS) 1996, World Bank 1997. Another crucial difference between the large cities of Asia and Australia is their much higher population densities, as shown in Table 2. It is very difficult to get comparable statistics for urban population density, hence their availability has partly dictated which cities to include. Here, the table presents the ‘urban area’ density for Sydney and Melbourne, as defined by the Australian Bureau of Statistics (ABS), as well as the density of the ABS-defined Inner Areas. Comparable figures for urban area density would be the density for the fully built-up areas of Singapore and Hong Kong, and for Tokyo, the density of the population in the ‘densely inhabited districts’, which covers about 87% of the four-prefecture metropolitan area population (Japan Statistical Association 1998), all shown in Table 2. The 23 wards of Tokyo, and the Tongdaemun district of Seoul, can be compared to the Inner Area density of the Australian cities. It is clear that densities in Asian cities are up to an order of magnitude higher than in Australia’s densest two cities. Less-detailed data indicate high densities for other major Asian cities, for example up to 50 000 persons/sq.km in the densest parts of Shanghai in 1992 (Xin 1996) Another difference between Australian and Asian cities is in the density changes over time. For example, the urban area densities of Sydney and Melbourne have fallen 45% and 28% respectively since 1947, as innercity residents settled at much lower densities in caroriented suburban areas (Moriarty 1996). The experience of Japanese cities is in marked contrast: the proportion of the Japanese population living at high densities has steadily risen from 43.7% in 1960 to 64.7% in 1995, although the fraction at extremely high densities has fallen (Japanese Statistical Assocation, 1998). This is hardly surprising given that only a small fraction of the country is suitable for settlement, and the resulting extraordinarily high costs of urban land in Japan. Table 2. Population density of various cities, 1994. City Description of area Melbourne Urban area -Inner city Urban area -Inner city Entire area -HK Is. And Kowloon Urbanised area -23 wards Entire area -Built up areas Met. City area -Tongdaemun district Sydney Hong Kong Tokyo Singapore Seoul * * Population density (persons/sq.km) 1 750 2 565 2 090 3 455 5 720 26 130 10 960 16 580 4 690 7855 17 015 29 275 1995 data. Sources: ABS census results 1991 and 1996; Japan Statistical Association 1998; Dept. of Statistics, Singapore 1996; Howlett 1996; Korea National Statistical Office 1996. For developing Asian countries, the total population in cities of one million or greater is expected to grow rapidly, given the present low levels of urbanisation. Future prospects for decreases in urban densities in the cities of developing Asia do not look promising. Rapid industrialisation in Japan, Taiwan and South Korea was at the expense of grain land, so that by 1993, these countries were importing 77%, 74% and 68% respectively of their grain. China is similarly losing about 1% of its cropland for grain annually, as it rapidly industrialises. But Brown and Kane (1995), from whom the above figures are drawn, use US Department of Agriculture projections to show that China (let alone other Asian countries), cannot follow the grain-importing route of the early Asian industrialised countries. Some time around 2010, there will simply not be sufficient grain to import. Asian countries will therefore not be able to release much land for urbanisation. Future urban population growth in most Asian countries will thus most likely need to be at even higher densities than occurs at present. 3. Comparative Transport Data Cities in developing Asian countries usually have higher per capita incomes than do rural areas, which helps explain why urban population growth is expected to be so rapid. Because of these higher incomes, car ownership also tends to be much higher. In South Korea and Thailand, for example, about half of all registered cars operate in the capital city (World Bank 1997). In other words, demand for cars will be highest in Asian cities, which unfortunately are where congestion and air pollution are worst (WRI 1996). Transport-related data for a number of countries in the region (and their major cities where data is available) are shown in Table 3. Data at the city level is preferable, but is not usually available. The far higher urban densities found in Asian cities have important consequences for present levels of car ownership (as shown in Table 3), traffic congestion, and air pollution. Australian cities, despite their higher car ownership levels, have a much smaller number of vehicles (excluding two-wheelers) per km of road. The contrast in road space per vehicle is actually much greater than shown, since Asian cities have narrower roads, have far greater proportions of motorbikes and bicycles, and in addition, some cities also have animal drawn vehicles. The result is much higher average travel speeds in Australian cities. For example in inner Melbourne in 1994, peak hour speeds of well over 20km/hr were typical, with between-peak speeds of 30km/hr (DJA Maunsell 1995). Much higher speeds_about 45km/hr_were typical of the city overall. In contrast, in many Asian cities, peak-hour speeds fell during the 1980s (WRI 1996, Koshi 1989). The peak-hour average for Asian cities is now estimated to be 16km/hr, with peak-hour speeds inside the Middle Ring Road of Bangkok only 9km/hr (Phiu-Nual 1996, World Bank 1997). Evidently, cars in many Asian cities spend much of their time stationary in traffic jams; in Bangkok, the equivalent of 44 days per year! (World Bank 1997). Table 3. Selected transport characteristics of various cities/countries, 1994. Country Australia -Sydney SD -Melbourne SD Hong Kong Indonesia Japan -Tokyo prefecture -Osaka prefecture Malaysia Singapore Sth. Korea Thailand Cars per 1000 pop. GDP per capita ($US 1994 PPP) Cars per $m. GDP 461 425 510 54 10 341 252 263 128 116 115 21 18 120 20 425 19 420 21 000 3 600 21 140 30 000 24 500 8 440 21 900 10 330 6 970 25.5 20.8 26.3 2.7 2.7 16.2 8.4 10.7 15.1 5.3 11.2 2.9 Motor vehicles per km of road 13 71 75 289 N.A. 57 267 275 49 152 100 73 Sources: as for Table 1, and ABS 1998, UN 1996a, World Bank 1996. Two important consequences follow from these very low and often deteriorating traffic speeds in Asian cities. The first is that fuel consumption, and thus traffic-related air pollution and greenhouse gas emissions per vehicle-km, are much higher than they would be if traffic flowed more freely. The second is that fixed-rail public transport with its own right of way will usually be faster, often even at off-peak times. These shorter travel times for rail, (together with zero time and money parking costs), can go a long way toward overcoming the inherent advantages of private car travel, such as flexibility, privacy, weather protection, and door-to-door travel. Nor does it seem possible to remove rail’s travel speed advantage by providing new road space, the ameliorative approach to these problems adopted by cities in Australia. Although Asian urban roads are already very congested, there is substantial latent demand for car ownership and travel, as evidenced by the low values of cars per unit of GDP compared with Australia (whose value is typical for OECD countries), as shown in Table 3. (The GDP values are expressed in Purchase Parity Prices, as these best compare actual purchasing power for different countries.) Further evidence for latent car demand is the keen bidding for import licences in Singapore. It is very likely that any attempt to increase road space (never an easy task in densely populated cities), will merely result in higher car ownership and use, with no improvement in traffic speeds. Indeed, this is already happening: in many Asian cities, car ownership and road provision are growing, while as already mentioned, traffic speeds remain very low and are often falling. 4. Electric Rail Transport as a Solution Rail transport, with its own right of way, we argue, offers the best means of economically reducing transport related air pollution and greenhouse gas emissions in Asian cities, without the need to compromise travel times. In Australian cities, emission reductions would also occur, but a trade-off would have to be made with travel speeds. This section develops in detail the arguments to support these assertions. First, electric rail is far more land-use efficient than car travel. With practical minimum headways of 90120 seconds, 30-40 trains per hour can be accomodated on one track. For 8-car trains, and 100 seats per car, a seated passenger volume of 32 000 is possible. With passengers standing, much higher volumes are possible. Lines in Moscow, New York and Tokyo carry 50 000 or more per line, while Hong Kong’s Mass Transit Railway attains up to 80 000 passengers per hour (Black 1995, White 1995). In contrast, a single lane on a multi-lane freeway can carry no more than about 2 300 cars per hour, even under ideal conditions. As traffic density (that is, vehicles per km of road lane) increases past an optimum value, the carrying capacity decreases. On ordinary arterial roads, with cross-traffic, the figure will be even lower (Wright 1996). In many Asian cities the presence of large numbers of bicycles and even animaldrawn vehicles, will further reduce traffic volumes. In summary, congestion on heavy rail services decreases travel speeds somewhat, because of the need for greater stopping times at stations, but greatly increases landuse efficiency (as measured by passenger-km per lane per hour). For vehicular road travel, on the other hand, congestion not only greatly increases travel times, but actually reduces land-use efficiency. Under the congested conditions typical of large Asian cities, heavy rail is up to two orders of magnitude more landuse efficient than private car travel. Different transport modes can also be compared on the basis of energy use or pollution emissions per passkm. For Australia, the 1994/5 energy intensity of passenger car travel was 2.67 MJ/passenger-km (Apelbaum 1997). The fuel intensity (in litre/100km) for cars is not known for most Asian countries, but the national figure for Japan is slightly higher than that for Australia, at just under 12 litre/100km (International Energy Agency 1997). Car occupancy rates are also assumed to be similar to that for Japan and Australia, at 1.6 persons/vehicle. For simplicity, it is therefore assumed that for all countries in the region, secondary energy intensity can be taken as 2.67 MJ per pass-km, the Australian value. The CO2 equivalent emissions for the full fuel cycle are also assumed to be the same as Australia’s, 0.225 kg/pass-km. This figure, derived from Apelbaum, assumes 3.08 kg CO2 equivalent emissions per litre of fuel. The processes and energy requirements for converting oil in the ground to petrol/diesel in the fuel tank are assumed to vary little from country to country (Moriarty 1994). For large urban areas, congestion could increase this figure significantly: in 1988 in Bangkok, fuel intensity was 17.0litre/100km, and in Tokyo, even with its large numbers of ‘kei’(mini) cars, 13.3 litre/100km (Riley 1994, Hayashi 1996). In Australia, however, car fuel consumption in 1995 was only 6% higher in urban compared with non-urban areas (Apelbaum 1997). For electric rail transport, CO2 equivalent emissions per pass-km can be conveniently expressed as the product of the following four factors: 1. CO2 equivalent emissions per MJ of primary fuel 2. MJ primary energy/MJ secondary energy, i.e. the energy efficiency of electricity generation 3. MJ secondary energy/seat-km 4. Seat-km/passenger-km, i.e. the inverse of seat occupancy rate. As regards the first factor, with the important exception of China, the power station fuel mix is far less fossil fuel-based than is the case for Australia, where fossil fuels, predominantly coal, forms 90.7% of the input fuel (UN 1996b). Emissions of CO2 will therefore be less, especially for countries with large shares of hydro or nuclear power. In addition, efficiencies for coal power stations are usually lower than for those using other fuels (factor 2). There is no reason why the third factor should be different in Asian as compared with Australian cities, but the fourth factor has much lower values in Asian cities, as is evident from the high figures given above for the Tokyo and Hong Kong rail systems. In Australian cities, electric rail transport in 1994/5 required 0.41 MJ of electrical energy (secondary) per pass-km, very similar to that for trams (0.39 MJ/passkm). When account is taken of the full fuel cycle, emissions are 0.093 kg CO2 equivalent per pass-km for urban rail, and 0.109 kg/pass-km for trams (Apelbaum 1997, Bureau of Transport and Communication Economics 1996). These values are both less than half that for cars, given above. For Asian cities, as discussed above, electric rail’s advantage will be even greater. In 1992, Hong Kong’s heavy rail transit system consumed only 0.18 MJ of electrical energy per passkm, less than half that for Australia (Moriarty and Mees 1995). Full fuel cycle emissions per pass-km could thus be nearly an order of magnitude less than that for car travel, especially when Asian traffic congestion was considered. In both Asia and Australia, the prospects for further large reducions in electric rail emissions are much better than for car emissions. Electric power generation has shown continuous efficiency improvements, with much scope for further gains. Any shift to non-carbon fuels would also lower CO2 equivalent emissions. Existing rail rolling stock can be made more efficient. For example, retrofitting the entire Hong Kong heavy rail transit fleet with GTO thyristor chopper controls was anticipated to drop electricity consumption by a third by 1995, when due for completion (Moriarty and Mees 1995). Secondary energy efficiency can also be dramatically improved by the use of lighter carriages (White 1995). Finally, just as for land-use efficiency, over-crowding of rail increases energy efficiency, but decreases it for car travel. In summary, electric rail transport can play an important role in improving the environment in large Asian cities. For the largest, such as present-day Tokyo, extensive electric rail systems are essential. For Sydney and Melbourne, a significant shift to urban rail seems harder to justify_and even harder to achieve. Urban growth rates are low, urban density is falling, and congestion (and air pollution) are of far less concern than in Asian cities. But two other problems will give the already extensive electric rail systems in these cities a far greater role than at present. Australian self-sufficiency in oil will decline rapidly over the next decade, and global oil depletion will probably lead to much higher oil prices by 2010 (Hatfield 1997, Ivanhoe 1997, Moriarty 1997). An effective response to rising levels of greenhouse gases will also be needed by then, and Australia has high per capita emissions (World Bank 1997). Increased use of electric rail transport can tackle both problems. Reductions in CO2 equivalent emissions will be enhanced by patronage increases because seat occupancy rates, and thus energy efficiency, will rise (Moriarty and Beed 1992). 5. Conclusions Despite much lower levels of car ownership than Australian cities, large Asian cities already have serious traffic-induced air pollution problems. The biggest Asian cities are already far larger than Australia’s, and in the developing economies are growing much more rapidly, mainly because of industrialisation combined with existing low levels of urbanisation. They are also up to an order of magnitude denser than Australian cities. In Australia, it was possible in the post-war era to accommodate rising car ownership by low density suburbanisation. This approach is less feasible in industrialising Asian countries because of rapid urban growth, lower income levels, and the need to preserve agricultural land. Major urban road building programs, even if economically and politically feasible, will not usually relieve congestion, because of the high latent demand for car travel in Asian cities. The present and future congestion in Asian cities means that electric heavy rail travel, which is very space-efficient, will often be faster door-to-door than car travel, even at off-peak times. The high patronage levels on rail services in densely populated cities means that rail’s energy efficiency will be far higher than that for car travel on congested roads. Electric rail can not only save much valuable land in these dense cities, but can also greatly reduce transport air emissions—both local and global. References Apelbaum Consulting Group 1997, The Australian Transport Task, Energy Consumed and Greenhouse Gas Emissions, Commonwealth of Australia, Canberra. Australian Bureau of Statistics (ABS) 1996, Census results. Also 1991 census. ABS 1998, (Various years) Commonwealth and State Yearbooks. Ballew P.D. and Schnorbus R.H. 1993, ‘Assessing global auto trends’, Economic Perspectives, March/April, pp.15-26. Federal Reserve Bank of Chicago. Black A. 1995, Urban Mass Transportation Planning, McGraw-Hill, NY. Brown L. and Kane H. 1995, Full House: Reassessing the Earth’s Population Carrying Capacity, Earthscan Publications, London. Bureau of Transport and Communication Economics 1996, Transport and Greenhouse: Costs and Options for Reducing CO2 Emissions, BTCE Report No. 94, AGPS, Canberra. Department of Statistics, Singapore 1996, Yearbook of Statistics Singapore 1995, Ministry of Trade and Industry. DJA Maunsell 1995, Melbourne City Link Traffic and Revenue Forecasts, DJA Maunsell. Hatfield, C.B. 1997, ‘Oil back on the global agenda’, Nature, 387:121. Hayashi Y. 1996, ‘Economic development and its influence on the environment: urbanisation, infrastructure and land use planning systems’. In Transport, Land-Use and the Environment (eds. Hayashi Y.and Roy J.), Kluwer Acadamic Publishers, Dordrecht. Howlett R. (ed.) 1996, Hong Kong 1996, Govt. Printer, HK. International Energy Agency (IEA) 1997, Indicators of Energy Use and Efficiency, OECD/IEA, Paris. Ivanhoe L.F. 1997, ‘Get ready for another oil shock’, The Futurist, 31(1):.20-23 Japan Statistical Association 1998, Japan Statistical Yearbook 1998, Statistics Bureau, Tokyo. (Also earlier editions). Koshi M. 1989, ‘Major tasks for the future of road transport in Japan’, in OECD Road Transport Research 20th Anniversary Seminar, Challenges and Opportunities for Tomorrow, OECD, Paris. Moriarty P. 1994, ‘Can alternative fuels reduce greenhouse emissions?’ Int. J. of Vehicle Design, 15(1/2):1-7. Moriarty P. 1996, ‘Can urban density explain personal travel levels?’ Urban Policy and Research, 14 (2):109-18. Moriarty, P 1997, ‘The future for Australian personal passenger travel’, Road and Transport Research, 6(4):114-22. Moriarty P. and Beed C. 1992, ‘Policy options for reducing urban transport greenhouse emissions’, Road and Transport Research, 1(2):76-86. Moriarty P. and Mees P. 1995, ‘Reducing transport’s environmental impacts: alternative approaches’, Transactions of the International Symposium on Energy, Environment and Economics, Univ. Melbourne 20-24th Nov., pp.639-44. National Statistical Office 1996, Korea Statistical Yearbook 1996, Rep. Of Korea. (Also earlier editions). Phiu-Nual K. 1996, ‘Congestion and pollution in a rapidly expanding city of South-East Asia: the case of Bangkok’. In Transport, Land-Use and the Environment (eds. Hayashi Y.and Roy J.), Kluwer Acadamic Publishers, Dordrecht. Riley, R. 1994, Alternative Cars for the 21st Century, SAE Inc. Warrendale, PA. Taylor A. 1997, Danger: rough road ahead, Fortune, March 17, pp.37-42. United Nations 1996a, Statistical Yearbook, 41st Issue, 1996, U.N., N.Y. United Nations 1996b, 1994 Energy Statistics Yearbook, U.N., N.Y. United Nations 1997, Economic and Social Survey of Asia and the Pacific 1997, U.N., N.Y. White P. 1995 Public Transport: Its planning, management and operation, (3/e), UCL Press, London. World Bank 1996, World Development Report 1996, Oxford U.P., NY. World Bank 1997, World Development Indicators 1997, Oxford U.P., NY. World Resources Institute 1996, World Resources 1996-7, Oxford University Press, NY. Wright P. 1996, Highway Engineering, (sixth ed.), John Wiley and Sons, NY. Xin C. J. 1996, ‘Bicycle transportation in Shanghai: status and prospects’, Transportation Research Record 1563, pp.8-15.