Feature:　Railways and The Environment (part 2)
Effects of High-Speed Transportation Systems on Environmental Improvement in Japan
Haruo Ishida and Seiji Iwakura

In Japan, the use of motor vehicles still shows no signs of decreasing and road traffic continues to rise in terms of both urban and inter-city transport. As a result, a reduction in carbon dioxide (CO₂) emissions by the transport sector is far from being achieved. Similarly, local air pollution caused by nitrogen oxides (NOx) and suspended particulate matter (SPM) is still common and many urban areas fail to meet government air quality standards. Meanwhile, public pressure for environmental improvement has intensified in recent years. The Japanese government agreed to reduce emissions of gases causing global warming at the UN Convention on Climate Change (COP3) held in Kyoto in December 1997. In July and August 1998, suits were filed against the government and highway authorities in Osaka and Kawasaki, respectively, alleging their liability for local air pollution caused by exhaust gases. Clearly, drastic and urgent actions are expected from the transport sector toward environmental improvement, especially in the management of urban and inter-city transport.
This article reviews major environmental issues facing Japan and analyzes the current status of environmental improvement policies undertaken by the Ministry of Transport (MOT), along with technology to control pollution at source. In particular, evaluation of various environmental parameters is becoming important, especially analytical systems to quantify the effects of policies and programmes. Major issues related to such quantitative assessment are identified and analyzed. Finally, we describe a recently developed analytical framework to evaluate the effects of environmental improvement measures on inter-city transport, including upgrading of rail service levels, increasing charges for using automobiles, and promoting low-pollution vehicles. The evaluation results for various programmes are presented.

Global Warming and Local Air Pollution

Figure 1 shows the breakdown of CO₂ emissions by country in 1992. Japan ranks fourth in absolute amount of CO₂ emissions. Per capita emissions are around 50% of the USA and Canada, at the same level as the UK and Italy, and four times the level of China. The ratio of emissions to GNP is equivalent to that of Italy, 50% that of the USA, Canada, and the UK, and 5% that of China. By industry, the transport sector accounts for 18.9% of the total CO₂ emissions in Japan, and is the key to effective reduction of emissions. Within the transport sector, passenger transport produces the dominant (64.4%) share of emissions, of which automobile transport accounts for 87.5%. The share of emissions by the transport sector has been rising each year, explaining why CO₂ emissions by the transport sector grew 90% between 1973 and 1994, in contrast to the 20% increase by all sectors. Table 1 shows the environmental impacts predicted by the Environment Agency on Japan caused by doubling the present CO₂ concentration; the impacts are extensive.
Table 2 summarizes the impacts of local air pollutants on people. The NOx level has been gradually declining, but it is still higher in major urban areas than exhaust emission standards. In 1991, the carbon monoxide (CO) levels at all air quality monitoring stations throughout Japan met the emission standards. The level of sulphur oxides (SOx) declined by 82.4% between 1970 and 1989, due to reduction of the sulphur content of fuels, wider use of desulphurization technology, and improved fuel efficiency. Progress with SPMs is rather slow; in 1991, only 50% of air quality monitoring stations and just 30% of exhaust gas monitoring stations met the environmental standards. Diesel vehicles are estimated to produce 20% to 40% of SPM emissions.

Figure 1: National CO₂ Emissions in 1992
Table 1: Impact on Japan of Doubling of Atmospheric CO₂ Levels
Table 2: Health Impacts of Air Pollutants

CO₂ Emissions by The Transport Sector and Traffic Demand Trends

Despite these apparent improvements, passenger traffic in the transport sector has been shifting continuously from energy-efficient railways to energy-inefficient systems such as the private motor vehicles. Figure 2 shows the changes in modal share of inter-city passenger transport (passenger-km) between 1955 and 1994. Motor vehicles have increased their share at an accelerating rate since 1965, from 32% to 67%. For trips of 300 km or more, air transport has gained a considerable share due to growth of personal income and the increase in the number of local airports.
As shown in Fig. 3, the use of motor vehicles in urban transport has risen rapidly. Clearly, motor vehicles have gained in popularity for commuting and private trips. Table 3 shows that energy consumption per passenger-km by road transport in urban areas during rush hours is about 23 times that of railways, not to mention the resultant chronic congestion and air pollution. In contrast, Japanese railways mainly use electricity for power, which causes less air pollution.
There is now a serious need to reduce the environmental impacts of the transport sector by developing a well-coordinated system that combines different modes according to their advantages. For example, the railway can carry large numbers of people over short to medium distances with high energy efficiency, air has the overwhelming advantage of speed over long distances, and roads are convenient for door-to-door continuity.

Table 3: Energy Consumption of Different Transport Modes
Figure 2: Modal Share of Inter-city Passenger Transport (passenger-km)
Figure 3: Modal Share in Metropolitan Tokyo

Environmental Improvement Policies in Transport Sector

The Japanese MOT has been studying and developing a package of both technical and institutional policies to address the issue of global warming (Table 4). The strategy with the highest priority is emission control at source for motor vehicles. However, the importance of other policy measures, including diversion of traffic to energy-efficient systems, control of motor vehicle use by area, date and/or time zone, public campaigns about the need to reduce CO₂ emissions, and technological cooperation with other countries, are all recognized. More precisely, short- and medium-term measures up to 2010 include changing the automobile tax system, promoting the ‘eco-drive’ concept, constructing and upgrading public transport systems, promoting their use, and streamlining distribution systems. These measures will evolve into long-term measures, including development of zero-emission vehicles (ZEV).
Recently, a variety of R&D efforts are underway to control emissions at source. New technology is being developed in the following main areas: (1) improving engine thermal efficiency; (2) reducing acceleration resistance by cutting vehicle weight; (3) reducing rolling resistance; (4) reducing energy loss in drive system, such as transmission; and (5) modifying engine design to burn alternative fuels. Table 5 compares the major features of new motor vehicles powered by conventional and alternative fuels. Compared with present vehicles, cars powered by fuel cells, electricity and natural gas have excellent prospects for reducing CO₂ emissions, but they still have to overcome problems of limited driving range.
The railway industry is also making extensive efforts to become more environment friendly. For example, since 1992, JR East has been setting the goals shown in Table 6. In the fields of energy saving and reduction of CO₂ emissions, the major focus is on upgrading railcar design and performance as well as on power generation systems. Weight reduction, use of regenerative braking, and introduction of variable voltage variable frequency (VVVF) circuits are major parts of the effort to achieve higher energy efficiency. For example, JR East's latest Series 209 EMU has been improved as shown in Table 7 to achieve a 50% reduction in energy consumption compared to its Series 103 predecessor. R&D into power generation involves development of more efficient and cleaner energy sources. In addition, state-of-the art technologies are being developed, such as waste heat recycling and commercialization of solar power generation. The 16 leading private railway companies are increasingly using energy-saving vehicles (Table 8).

Table 4: Global Warming Countermeasures Proposed by Ministry of Transport
Table 5: Comparison of Major Features of Alternative-Fuel Cars
Table 6: JR East Environmental Improvement Measures (at 1996)
Table 7: JR East Equipment Performance Upgrades
Table 8: Energy-Efficient EMUs Introduced by Large Private Railway Companies

Assessing Effects of Environmental Improvement Measures

Generally, quantitative analysis conducted to assess the effects of environmental improvement measures by the transport sector is designed to determine costs and benefits. Two indicators: physical, such as emitted amount of pollutant, and economic, such as damage to the national economy caused by environmental degradation, are used. However, a major problem is how to establish reliable and agreed base values for pollutant emissions and social costs. Comparison of base values currently used in Japan shows considerable variation among published data; the difference between the values is as large as 200% for CO₂, 400% for NOx, 300% for SOx, and 1000% for CO (Table 9). This is caused by a number of factors, including differences in vehicle age, model, and fuel in different studies, as well as differences in statistical analysis methods. In any case, the fact remains that the most vital and important set of data for policymaking are not available.
A similar situation exists in quantifying the social costs of air pollution. Although much research has been conducted in the USA and Europe, there is no agreed uniform method to measure social costs. Moreover, the measurement conditions vary considerably between countries, resulting in great variation of social costs. Similarly, there are practically no useful data on social costs available in Japan, especially no measured data related to the social costs of global warming, which can be used in the transport sector. To address this shortcoming, we applied the contingent valuation method to measure the social cost defined as the amount that Japanese people would be willing to pay for government policy measures to counter both the positive and negative effects on the Japanese environment and society (Table 1), caused by a rise in air temperature of 1 to 2.5C resulting from a doubling in the level of atmospheric CO₂. A questionnaire prepared according to the US National Oceanic and Atmospheric Administration (NOAA) guidelines was used to survey 1620 people in Tokyo, Nagoya, Osaka, Sendai, Okayama, and Kumamoto. The survey showed that respondents would only pay ¥8235 per year to prevent global warming. This figure was then used to determine the social cost per tonne-carbon, which came to ¥8320/t-C.
Although we hope that base value emissions will soon be made more precise and unified, current quantitative analysis methods and data must allow for some different possibilities, and failure to do so could result in incorrect policy decisions.

Table 9: Comparison of Pollutant Emissions Base Values by Transport Mode

Policy Analysis System LETS

The Low Emission Transportation System Simulator (LETS) is designed primarily to assess the effects of environmental improvement measures for inter-city transport systems. It is based on the widely used demand forecast model with some customization, and examines the macro relationships between traffic demand and environmental impacts. The system programme runs on a PC and produces simulation results in 2 minutes. It measures the above effects by combining traffic data (passenger-km) obtained from the demand forecast model with base values for pollutant emissions and social costs. The result can be used to estimate user benefits that are consistent with the demand forecast.
Figure 4 shows the fundamental structure of LETS. The demand forecast sub-unit for inter-city transport analyzes demand changes caused by development of high-speed transport networks or increases/decreases in fares and taxes. Another sub-unit quantifies changes in pollutant emissions from different transport modes, the incidence of traffic accidents, etc. Multiplying traffic data (passenger-km) from the demand forecast sub-unit by emission base values for respective transport modes, gives the environmental pollution loads and other data. A third sub-unit measures changes in environmental pollution loads caused by development of high-speed transport systems in monetary terms to determine the external economic effects of such development. The use of monetary units allows various environmental pollutant loads (e.g., CO₂, NOx) to be treated uniformly as social costs that can then form the basis of cost/benefit analysis.
The traffic demand forecast sub-unit uses a four-stage estimation method based on three types of trip (business, tourism, other) and three transport modes including air, rail, road (excluding bus).
Existing demand forecast models tend to underestimate environmental impacts because they fail to take induced traffic into account. To address this shortcoming, LETS estimates induced traffic by linking different models at various stages through accessibility variables. As a result, it can express induced traffic volumes and changes in selection of destinations due to reduced travel time caused by increased railway operating speeds. Moreover, LETS can analyze how traffic is affected by decreases in disposable income caused by fare increases. Figure 5 shows the concept of the induced demand generation process. The size of the circles represents traffic volume, and the changes in circle size show that traffic distribution patterns change with the improvement of transportation systems. In the diagram on the left, trips starting in zone O are divided equally between three destinations: A, B, and C. As improvement in the transport system reduces the travel time between zones O and A, more traffic is induced in zone O (shown by a black outer circle), leading to a significant increase in the number of trips to zone A. The increase in total traffic volume spills over to zone C, although the transport system from zone O to C has not been improved, but trips from zone O to zone B decrease because the transport to zone B is less convenient than the other two destinations.

Figure 4: LETS Components
Figure 5: Diagram Showing Induced Demand and Changing Traffic Distribution

Analysis of Policy Measures and Effects

This section describes the LETS simulation results based on the Japanese transport networks shown in Fig. 6. The environmental improvement effects and socioeconomic impacts of various policy measures are compared assuming that GDP grows at an annual rate of 2% while the number of automobiles per household grows at an annual rate of 1%. Table 10 summarizes the analyzed policy options, option levels, and forecast results for 2010.
If no environmental improvement measures are taken under the assumed conditions, total traffic volume expressed as passenger-km will grow by 47% between 1990 and 2010, and generated traffic volume will grow by 43% for the same period. The modal shares for the generated traffic volume are 67% for motor vehicles, 26% for rail, and 7% for air. Pollutant emissions will increase by 45% for CO₂, 47% for NOx, 42% for SOx and 49% for CO.
Table 10 also shows the rates of change in 2010 when improvement measures are taken. First, improvement of rail services to encourage a demand shift to rail, which causes relatively small environmental pollution loads, does not always contribute to environmental improvement because it induces additional demand at some improved service level (e.g., a significant increase in train speed). Nevertheless, pollutant emissions per trip are always reduced by a demand shift to rail.
Generally, any economic measure such as increases in fares, tolls, taxes, etc., is most effective in producing environmental improvements, but the resultant increase in travel costs may negatively impact the utility (satisfaction) level of transport users to some extent.
However, it should be noted that LETS does not account for recycling of tax revenues. Consequently, effective investment of taxes generated by economic measures must be taken into account in the future.
Introduction of low-polluting cars is very effective in improving the environment, suggesting the need for purchase incentives.
Although it is well known that railways are clearly more environment friendly than other transport modes, their main advantage is their high transport efficiency, which should be the primary basis for deciding new construction.
The simulation results show that although strict policy measures (higher option levels) may not stabilize CO₂ emissions at the 1990 level, meaningful improvement can reasonably be expected by combination of pollution control measures at source (e.g., promotion of low-pollution cars) with effective measures to encourage demand shift to low-pollution transport modes such as rail.
A further analysis examined the effectiveness of two policy combinations: 1. Increasing operating speeds of inter-city lines along with increasing highway tolls; and 2. Increasing operating speeds of inter-city lines and creation of the carbon tax. The main purpose of this analysis is to find how to achieve environmental improvement without affecting the utility level of transport users. Figure 7 shows the relationship between social costs (related to global environment, local environment and traffic accidents) and economic effects. Introduction of the carbon tax on its own contributes to increased economic loss, but if it is combined with faster inter-city rail services, positive effects are produced. This clearly suggests that some options can achieve both environmental improvement and higher utility, depending upon taxation rates, traffic conditions, and other relevant factors.
The above analyses are useful for preliminary evaluation of various policy measures. However, at the same time, the relatively low precision of base values for pollutant emissions and social costs requires careful interpretation of the results. For example, the highest CO₂ base value for automobile emissions is 70% higher than the lowest value. When the former value is used to evaluate the economic effects of the carbon tax, users face a higher tax burden because the tax is imposed on the basis of the weight of emitted carbon. The reverse is true when the lowest base value is used. The difference has an impact on the results related to traffic demand, user benefits, and carbon tax revenues, thereby influencing policy options.

Table 10: Effect of Environmental Improvement Measures in 2010
Figure 6: Japanese Transport Networks as of 1998
Figure 7: Economic Effects of Policy Combinations

Conclusion

As a concluding remark, we would like to point out major issues to be addressed for effective evaluation and implementation of policy options.

•	Many policy measures in the transport sector tend to evoke trade-offs between improvement of convenience/economic development and environmental preservation/improvement. This is attributable primarily to lack of assessment of possible measures, and poor communication with affected parties. To ensure timely selection and implementation of effective measures, policy options (including scenarios) and evaluation results must be made available for broad-based public discussion.
•	To encourage open and fruitful discussion, formal communication routes should be instituted between policy-makers and the public.
•	Since discussion of environmental policy tends to be dominated by intransigent positions based on different opinions, efforts should be made to develop a common viewpoint based on quantitative analysis using objective data. More precisely, base values for pollutant emissions and social costs should be unified through research and extensive field measurement, and a national consensus should be established on what values to use as the basis for discussion. Finally, forecasting models must be made more accurate.