Area Wide Calculation of Traffic Induced CO2 Emission ... - Springer Link

KSCE Journal of Civil Engineering (2012) 16(3):450-456 DOI 10.1007/s12205-012-1525-5

Transportation Engineering

www.springer.com/12205

Area Wide Calculation of Traffic Induced CO2 Emission in Seoul Im Hack Lee*, Seungjae Lee**, Jin Soo Park***, and Shin Do Kim**** Received March 28, 2011/Revised May 24, 2011/Accepted June 21, 2011

···································································································································································································································

Abstract To reduce emissions of Greenhouse Gas (GHGs) from automobiles, real on-road emission data should be required to help policy makers establish standards for reducing GHG emissions. Since general studies on calculating CO2 emissions from transportation sources have been based on petroleum consumption data, it has been difficult to analyze spatial activities and characteristics of transportation sources. In this study, we propose a method for calculating CO2 emissions from on-road transportation sources in Seoul. We focused on CO2 emission calculation by applying real traffic flow data and analyzed base emissions from the main roads and local streets. Because the emissions were calculated using a 1 km×1 km grid cell format, these data can be applied to other compatible transportation data sets for air pollution analysis, modal shift analysis, etc in the transportation sector. Keywords: GHG, V.K.T., CO2, emission, local streets ···································································································································································································································

1. Introduction On-road transportation is a significant source of Greenhouse Gas (GHG) emissions and accounts for about 22% of the regional GHG emissions in Seoul. Since the advent of the philosophy of Environmentally Sound and Sustainable Development (ESSD) in the late 1980s and Rio Declaration on the Environment and Development in 1992, a new paradigm has emerged in transport policy planning and making for sustainable transport systems capable of satisfying the needs of transportation services while minimizing the side effects (KOTI, 2007). This paper provides estimates of GHG emissions from on-road vehicles in Seoul. One of the reasons that we produce air pollutant emission inventories is to apply them as input data for computer modeling of various kinds. Many studies have revealed limitation for interpreting and addressing complicated phenomena of the society from a single discipline. The trends of recent research, therefore, tend to carry out integrated urban modeling with other disciplines (Urbansim, delta etc.) such as environmental, transportation, urban engineering and GIS parts, so as to identify current issues, forecast future situations, and to apply resolvable scenarios to develop alternative policy measures. In applying input data from air pollution emissions inventory to models, uniform grid formats are used as a rule. Therefore, it is necessary to establish emission database of the uniform grid format in which spatial analysis can be performed if the modeling input data are to be applied to GHG emission inventories of

the transportation part. As climate change has become a major global issue, the estimation method of GHG emissions in the environmental sector required some changes from the top-down method using simply statistic data to the one with which spatial/temporal analyses were possible and applicable to specific conditions of the central and local governments. This study created a format of the GHG emission database of the transportation sector by 1km x 1km grid to use as input data for integrated models that interpret urban climate change problems. The merit of this database format is that it can be applied for environmental pollution models such as Calpuff and CMAQ as well as for the urban integrated models. To reduce emissions of Greenhouse Gases (GHGs) from automobiles, actual on-road emission data should be required to help policy makers establish standards for reducing GHG emissions. In this study, we propose a method for calculating CO2 emissions from on-road transportation sources in Seoul. The category of the on-road transportation sources includes all types of light-duty vehicles such as automobiles and light trucks, and heavy-duty vehicles such as trailers and buses. These vehicles are operated with many types of gaseous and liquid fuels. Vold (2005) applied the framework to Greater Oslo based on output from a land use and transport model. In this paper, the principle for determination of the housing rents in MEPLAN, TRANUS and DELTA, was to adjust the housing rents such that the share of residents in the zones stay within upper and lower.

*Doctoral Candidate, Dept. of Environmental Engineering, University of Seoul, Seoul 130-743, Korea (E-mail: [email protected]) **Professor, Dept. of Transportation Engineering, University of Seoul, Seoul 130-743, Korea (E-mail: [email protected]) ***Doctor, Climate and Air Quality Research Department, National Institute of Environmental Research Environmental Research Complex, Incheon 404708, Korea (E-mail: [email protected]) **** Professor, Dept. of Environmental Engineering, University of Seoul, Seoul 130-743, Korea (Corresponding Author, E-mail: [email protected]) − 450 −

Area Wide Calculation of Traffic Induced CO2 Emission in Seoul

This pricing mechanism is also used in Noth et al. (2003) said that UrbanSim simulates the development of urban areas, including land use, transportation, and environmental impacts, over periods of 20 or more years, and its purpose is to aid urban planners, residents, and elected officials in evaluating the long-term results of alternate plans, particularly as they relate to such issues as housing, business and economic development, sprawl, open space, traffic congestion, and resource consumption. Simmonds et al. (2010) said that the MEPLAN/TRANUS approach takes the interactions between activities as the key variables, and these are predicted, and then the location of activities is calculated from the total levels of interaction, and the patterns of interaction are also factored, from persons to trips, to give transport demand matrices, and this approach may be called the ‘interaction-focused’ approach, since the central feature is that the predicted interactions determine the location of activities, and he said the alternative approach, represented particularly in URBANSIM and DELTA, first predicts the location of landusing activities, and then models the interactions between those located activities, so this can obviously be called the “locationfocused” approach. Sum studies executed Calpuff air dispersion model and CMAQ model with road emissions (Yim, 2010; Sokhi, 2006). Several studies have uncovered meaningful relationships between vehicle traffic data and GHG emissions. Smit (2008) analyzed ‘the magnitude and direction of the effect is a function of emission model (type), shape of the composite emission factor curve and change in the joint distribution of (sub)-network Vehicle Kilometers Traveled (VKT) and speed.’ Scholl (1996) provided a comparative analysis of the changes in energy use and CO2 emissions from passenger transport in nine OECD countries. Hu et al. (2010) discussed technology and policy initiatives needed to deal with energy challenges related to road transportation development by making comparisons with the following foreign experiences: promoting the development and dissemination of alternative fuels, such as ethanol, methanol and bio-diesel, and clean vehicles, strengthening regulations related to vehicle fuel economy and emission, improving traffic efficiency and facilitating public transport development. Piecyk et al. (2010) determined the baseline trends in logistics and supply chain management and associated environmental effects. Factors affecting freight transport demand, truck fuel consumption and related CO2 emissions are classified into six categories in different levels of logistical decision-making. In the studies above, vehicle flow rates were used to estimate air pollutant emissions, although they did not consider local street emissions. In general, estimation of vehicle flow rates is carried out in main roads while local street emissions are not calculated. However, local street emissions are not negligible. Thus, this paper produces estimation methods of greenhouse gases from local streets using traffic rates that hadn’t been considered in other studies.

Vol. 16, No. 3 / March 2012

2. Method and Data In the base year 2004, approximately fifteen million vehicles were registered in Korea with approximately 3.1 people per vehicle (KNSO, 2007). The fuel sources used around the end of 2005 include gasoline (41.7% of total fuel consumption), diesel (47.9%), LPG (9.0%), and Compressed Natural Gas (CNG) (1.2%). Fig. 1 shows that most of the liquid and gaseous fossil fuels such as gasoline and butane are consumed for transportation in Seoul. The fundamental methodologies for estimating GHG emissions from road vehicles have not changed since the publication of the 1996 IPCC Guidelines, except that the emission factors now assume full oxidation of the fuel. Estimated emissions from road transport can be based on two independent sets of data: fuel sold and vehicle kilometers. If these are both available, a comparison of the two can validate their accuracy (IPPC, 2006). The methodology determines GHG emissions based on the basic equation: GHG Emissions = Numbers of Vehicles × Avg. VKT × Unit Emission factors

(1)

Where, vehicle types are sub-divided into a number of different categories ranging from small passenger vehicles, to lightduty trucks, to heavy-duty trucks to various specialized vehicle classifications. As part of the research on estimating vehicular GHG baselines, the required information on the number of kilometers driven by various vehicle types and in order to contribute to this study, we undertook VKT analysis in the fall of 2009 with traffic data of each fundamental unit government (called “Gu”), Seoul Development Institute (SDI) and Seoul Metropolitan Police Agency (SMPA). Each ‘Gu’ calculated the traffic rate in its local boundary and we merged the traffic rate data in Fig. 2. The high resolution-level data for the numbers of vehicles are recent and of high accuracy. The SDI has released transportation network data for the capital region of Korea. The assigned daily traffic flows are broken down by hour, using the monitored traffic counts. The hourly traffic counts data were collected over a week at 4 different road classifications: cordon line (39 locations), Han River bridges (18 locations), arterial roads (34 locations), and central business district roads (26 locations). The traffic counts have been directed by the SMPA

Fig. 1. Analysis of Energy Consumption Balance in Seoul using an Energy Flow Chart

− 451 −

Im Hack Lee, Seungjae Lee, Jin Soo Park, and Shin Do Kim

Fig. 4. The Result of Traffic Flows in Main Roads

Fig. 2. Traffic Flow Gathering Points in Seoul

and are reported on the Seoul City Government’s website (Seoul City Government, 2008)

3. Result Transportation is the movement of people and goods from one location to another, and traffic flow is the interactions between vehicles and infrastructure such as highways, signal, and traffic control devices. Although we should use real time traffic flow data to obtain road CO2 emissions, we conducted statistical calculations based on road traffic flow data from 25 “Gu”s with 723 points, from SDI and SMPA systems. The traffic flow data of the main roads were calculated as shown in Fig. 3: 1) calculate the sum of the traffic flow data from the A direction to the B direction, and the sum of the traffic flow data from the B direction to the A direction separately, and 2) calculate the sum of traffic flow data of both directions. These data were then inputted with GIS Database (DB) format. Figure 4 shows the traffic flow result of GIS DB. The characteristic of traffic flow in Seoul was that there was more land in the south of the Han River than north of it. There was a lot of land in the southwest in Yeongdeungpo-gu, in the southeast in Gangnamgu and Seocho-gu, and in the north in Jung-gu and Jongno-gu.

The result is a link-specific array representing the number of vehicles that drive at a particular mean speed and that abide by the set of rules within the boundary conditions discussed before. To illustrate the results, four traffic flow distributions are shown in Fig. 5. In this figure, gray lines represent grid, and sections mean roads. This array is then multiplied by the link length to arrive at an array of total VKT according to vehicle class, resulting in VKT values per 1km/ h speed and VKT values per traffic situation. Finally, these VKT values are multiplied by the appropriate composite emission factors in order to compute the total link emissions. The distribution of traffic flow data by vehicle type is very important when road emissions are calculated. General emissions have been calculated by using the government registration data of cars because of the complexity and the insufficiency of the vehicle type data. However, the emissions are based on the government registration data of cars and can vary greatly from real data. Therefore, we used real counting data to estimate the vehicle types, and the results for the various vehicle type mixing ratios are shown in Fig. 6. The contribution of passenger cars was about 80% in Gangnamgu, Seocho-gu and Songpa-gu, in which the reliance ratio of local

Fig. 3. Calculation Method of Traffic Flows

Fig. 5. Calculation Method of VKT − 452 −

KSCE Journal of Civil Engineering


Fig. 6. Vehicle Type Mixing Ratios of Each Sector in Seoul

public finance is high and there are many wealthy households. On the other hand, the contribution of taxis and motorbikes was relatively high in Jung-gu, Jongno-gu and Dongdaemun-gu, in which the commercial functions are active. The results of Fig. 6 were used to calculate the annual kilometers by each vehicle type, by each main road, and the annual transportation CO2 emissions in Seoul, which totaled 8.1 million ton-CO2/yr with contribution ratios of 38.0% from gasoline fuel, 45.8% from diesel fuel and 16.2% from butane fuel. The highranking sectors of annual CO2 emissions were Gangnam-gu, Seocho-gu, Songpa-gu and Yeongdeungpo-gu, in which the residential and commercial functions are vigorous, whereas the low-ranking sectors such as Dobong-gu and Gangbuk-gu had relatively low levels of commercial function or small residential populations Because the emission input data format for the air environmental modeling system generally consist of a 1 km×1 km set (Kannari et al., 2007; Baldasano et al., 2008; Tuia et al., 2007; Kluizenaar et al., 2001), we processed the CO2 emission DB by 1 km ×1 km resolution format (697 meshes in Seoul) with the GIS extraction method (Mena, 2003) and the results are shown in Fig. 7. The results of this figure were the same as those analyzed in Fig. 4, in that the high-ranking emission grid cells were those with high-ranking traffic flows. As aforementioned in chapter 2, (Method and Data) we calculated the on-road CO2 emissions by the second method based on the fuel sold data. The data of consumed petroleum (gasoline,

Fig. 7. The Result of CO2 Emission at Main Roads Vol. 16, No. 3 / March 2012

diesel, butane) were gathered by the Korea National Oil Corporation, with 646 petroleum stations for gasoline and diesel and 52 for butane. The high-ranking orders of petroleum stations per unit area were in Dongdaemun-gu, Yeongdeungpo-gu and Gangnam-gu, and the low-ranking orders of butane stations per unit area were in Dongjak-gu, Jung-gu, Jongno-gu and Yongsangu because of the safety problem. Figure 8 shows the petroleum sale record-based annual CO2 emissions in Seoul. The high-ranking orders of petroleum sale records per unit area were in Yangcheon-gu, Gwangjin-gu and Gangnam-gu, with a total of 10.2 million ton-CO2/yr, comprised of 37.1% for gasoline fuel, 38.5% for diesel fuel and 24.4% for butane fuel. In the next step, we compared the CO2 emissions according to traffic flow with those according to the petroleum sale record, and obtained a result of a 21% difference in the Seoul boundary. For detailed analysis of the reason for this 21% gap, we compared the traffic flow-based emissions and the petroleum sale record-based emissions for each sector unit, and the results are shown in Fig. 9. Sectors with values over 1.0 had more traffic flow-based CO2

Fig. 8. Petroleum Sale-Based CO2 Emissions in Seoul

Fig. 9. Ratios of Traffic-Based CO2 Emission over Fuel SaleBased CO2 Emission in Each Sector − 453 −


emissions than petroleum sale record-based emissions, while values less than 1.0 represented the opposite case. The sectors located in the center of Seoul or with no butane stations had values more than 1.0, but general sectors had values lower than 1.0. The total Seoul-wide average was 0.79, which indicated a gap of 21% between the traffic flow-based CO2 emissions and the petroleum sale record-based emissions. To explain this 21% gap, we examined the calculating system for the traffic flows. The road definitions in the “Korean Minister of Land, Transport and Maritime (KMLTM)” are sorted into “Expressway”, “National Highway”, “Special Municipal Road/Metropolitan Municipal Road”, “Provincial Road”, and “Municipal Road”. In upper roads, traffic flows can be counted at “Expressway”, “National Highway”, and “Provincial Road” by KMLTM, and can be counted partially or cannot be counted at “Special Municipal Road/Metropolitan Municipal Road” and “Municipal Road”. In Seoul in 2005, the lengths of roads were 22 kilometers for “Expressway”, 169 kilometers for “National Highway”, and 7,854 kilometers for “Special Municipal Road”, and 6,252 kilometers of the last category were roads narrower than 12 m called “local streets” that had no traffic flow data. Therefore, we considered that the emissions from local streets have to be considered and we formulated a plan to calculate these missing data. We divided the roads in Seoul into two groups: main roads (red lines) with counted traffic flows and local streets (black lines) without. The results are shown in Fig. 10. In the GIS analysis results of the Seoul roads, the lengths of main roads and local streets were about 1,200 kilometers and 6,252 kilometers, respectively. We attributed the 21% gap between the traffic flowbased CO2 emissions and the petroleum sale record-based emissions to the missing local streets-based emissions. The high-class sectors having long local street lengths were Gangnam-gu, Songpa-gu and Seocho-gu, and the low-class sectors having long local street lengths were Jongno-gu, Jung-gu and Yongsan-gu. If the in-out traffic conditions in the regional boundaries in Seoul are different, then they can act as interference factors that affect the calculated emissions in each sector. When the in-traffic flows are larger than the out-traffic flows, traffic flow-based emission can be overestimated. When the out-traffic flows are

larger than in-traffic flows, traffic flow-based emission can be underestimated. Therefore, we analyzed the traffic volume equilibriums at 12 boundaries between Seoul and other cities and the results are shown in Fig. 11. The high-ranking cities of traffic flows along with Seoul were Seongnam, Goyang, Incheon and Anyang. The number of daily out-traffic flows from Seoul was 1,639,244 cars/day and that of

Fig. 10. View of Main Roads and Local Streets in Seoul

Fig. 13. Total CO2 Emissions from 697 Meshes in Seoul − 454 −

Fig. 11. View of Traffic Volume Equilibriums at 12 Boundaries

Fig. 12. Amount of CO2 Emissions in Local Streets



in-flows to Seoul was 1,706,179 cars/day. As the difference of traffic flow was lower than 3 %, we assumed that the 21% gap occurred because of local streets-based emissions and calculated those CO2 emissions in each 1 km × 1 km grid cell in proportion to the length of local streets in its boundary. In the Seoul boundary, the annual local street-based CO2 emission was 2.1 million ton-CO2/yr. In the above results, we analyzed that the main road-based emission was 8.1 million ton-CO2/yr and the local street-based emission was 2.1 million ton-CO2/yr. The total CO2 emission by the 1 km x 1 km grid cell format in Seoul is shown in Fig. 13.

4. Discussion Vehicles using fossil fuels as power sources should release GHG emissions. Estimating GHG emissions from vehicles is crucial for national and local governments to develop GHG emission inventories in this regard. Nevertheless, due to limits of technology in considering characteristics of vehicles of various kinds, we had used only Tier 1 method estimating GHG emissions based on fuel consumption until about a decade ago. In recent years, as a way of considering real road conditions, the need for Tier 3 Method has increased to facilitate local governments to establish GHG reduction measures by individual street conditions. However, it was identified that we were not applying the whole of the Tier 3 method because in most cities including Seoul the actual traffic flow rates of local streets were not included in the target items for counting traffic flow rates although they are an important factor in GHG emission inventories. In Korea, the tasks for Tier3 method for GHG source inventory estimation on on-road part that were carried out by the Ministry of Environment (MoE) have been transferred to the Ministry of Land, Transport and Maritime Affairs (MLTM) since January 2011. The Tier3 method the Ministry of Environment (MoE) applied for traffic flow calculation was used with annual average data of travel distance by vehicle from odometers the data obtained from annual vehicle tests conducted on a regular basis at the vehicle registration agencies of each local government of Korea. Strictly speaking, however, the way Tier 3 was used above is not perfectly well applied, because they are not reflecting actual road conditions. By the method MoE applied for emission calculation, for example, if vehicles registered at the registration agency located in Seoul drive more often on the roads of Gyeonggi-do or Incheon, GHG emissions will still be regarded as released on Seoul roads although actual driving was not done on them. Under this circumstance, MLTM currently in charge of the inventory task based on Tier 3 has not yet developed appropriate methodology for estimating GHG emission on the road part. In this study, we decided to develop a GHG emission method that can reflect conditions of both main roads and local streets. The on-road CO2 emissions measured in this study were comVol. 16, No. 3 / March 2012

Table 1. Comparison of On-Road CO2 Emission Estimates Item

This study

NIER (Tier1)

KOTI (Tier1)

MLTM (Tier2)

2006

2007

2007

10.6

9.6

10.3

Time

2006

Main road emission

8.1

Local street emissions

2.1

Traffic flow counting

available

not

not

not

Spatial analysis

available

not

not

not

pared with those from other researches. The spatial boundary was limited to Seoul, and the time boundaries were 2006 for the MoE and 2007 for MLTM. A total of four research cases were used as targets for comparison. The first case is a Tier 1 CO2 emission estimates conducted by the National Institute of Environmental Research (NIER), which is the affiliated organization of the Ministry of Environment. The second one is a Tier 1 emission estimates by the Korea Transport Institute (KOTI), the affiliated organization of the Ministry of Land, Transport and Maritime Affairs, and the third one is the result of Tier 2 carried out by the Ministry of Land, Transport and Maritime Affairs. Those data above are presented in table 0.0. The comparison above considers traffic flows of not only main roads but local streets, and facilitates spatial analysis of CO2 emissions. Accordingly, this study result will be useful to produce database of substantive on-road CO2 emissions, which will eventually help local governments to establish low-carbon policy measures implemented according to Sustainable Transportation Logistics Development Act of the country.

5. Conclusions Since general studies on calculating CO2 emissions from transportation sources have been based on petroleum consumption data, it has been difficult to analyze spatial activities and characteristics of transportation sources. In this study, we calculated CO2 emissions by applying real traffic flow data, and analyzed the main road- and local street-based emissions to be compared. Since the emissions were calculated using a 1 km × 1 km grid cell format, these data can be applied to other compatible transportation data sets for air pollution analysis, modal shift analysis, etc in the transportation sector.

Acknowledgments This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST) (No. 2009-413-D00001).

References Baldasano, J., Güereca, L. P., López, E., Gassó, S., and Guerrero P. J. (2008). “Development of a high-resolution (1 km × 1 km, 1 h)

− 455 −


emission model for Spain: The High-Elective Resolution Modelling Emission System (HERMES).” Atmospheric Environment, Vol. 42, No. 31, pp. 7215-7233. Hu, X., Chang, S., Li, J., and Qin, Y. (2010). “Energy for sustainable road transportation in China: Challenges, initiatives and policy implications.” Energy, Vol. 35, Issue. 11, pp. 4289-4301. Intergovernmental Panel of Climate Change National Greenhouse Gas Inventories Program (2006). “2006 IPCC guidelines for national greenhouse gas inventories.” IPCC, Vol. 3, pp. 3-10 Kannari, A., Tonooka, Y., Baba, T., and Murano, K. (2007). “Development of multiple-species 1 km × 1 km resolution hourly basis emissions inventory for Japan.” Atmospheric Environment, Vol. 41, No. 16, pp. 3428-3439. Kluizenaar, Y. D., Aherne, J., and Farrell, E. P. (2001). “Modelling the spatial distribution of SO2 and NOx emissions in Ireland.” Environmental Pollution, Vol. 112, No. 2, pp. 171-182. KNSO (2007). Korea national statistical office, Republic of Korea. KOTI (2007). A study for the sustainable urban transport system: Phase 1 land development pattern and transportation network system for energy saving, The Korea Transport Institute. KOTI (2009). Investigation of greenhouse gas emission for transportation part, pp. 107-122. Mena, J. B. (2003). “State of the art on automatic road extraction for GIS update: A novel classification.” Pattern Recognition Letters, Vol. 24, No. 16, pp. 3037-3058. Noth, M., Borning, A., and Waddell, P. (2003). “An extensible, modular

architecture for simulating urban development, transportation, and environmental impacts.” Computers, Environment and Urban Systems, Vol. 27, Issue 2, pp 181-203. Piecyk, M. I. and McKinnon, A. C. (2010). “Forecasting the carbon footprint of road freight transport in 2020.” Int. J. Production Economics, Vol. 128, Issue 1, pp. 31-42. Scholl, L., Schipper, L., and Kiang, N. (1996). “CO2 emissions from passenger transport.” EnergvPolicy, Vol. 24, No. 1, pp. 17-30. Simmonds, D. (2011). “Alternative approaches to spatial modeling.” Research in Transportation Economics, Vol. 31, Issue 1, pp. 2-11. Smit, R., Poelman M., and Schrijver J. (2008). “Improved road traffic emission inventories by adding mean speed distributions.” Atmospheric Environ., Vol. 42, No. 5, pp. 916-926. Sokhi, R. S. (2006). “Prediction of ozone levels in London using the MM5–CMAQ modelling system.” Environmental Modelling & Software, Vol. 21, Issue 4, pp. 566-576. Tuia, D., Eicker, M. O., Zah, R., Osses, M., Zarate, E., and Clappier, A. (2007). “Evaluation of a simplified top-down model for the spatial assessment of hot traffic emissions in mid-sized cities.” Atmospheric Environment, Vol. 41, No. 17, pp. 3658-3671. Vold, A. (2005). “Optimal land use and transport planning for the Greater Oslo area.” Transportation Research Part A: Policy and Practice, Vol. 39, Issue 6, pp 548-565. Yim, S. (2010). “Use of high-resolution MM5/CALMET/CALPUFF system: SO2 apportionment to air quality in Hong Kong.” Atmospheric Environment, Vol. 44, Issue 38, pp. 4850-4858.

− 456 −
