Energy Use of Subway Networks in the Sustainability era

Transcription

1 Energy Use of Subway Networks in the Sustainability era Sybil Derrible, Ph.D. Candidate, Department of Civil Engineering, University of Toronto CORS-INFORMS International Conference Toronto, ON June 15,

2 Rationale Current Work: Networks Effects Coverage, Directness, Connectivity per capita Boardings y = Ln(x) R 2 = Coverage Area / Total Area 2

3 Rationale Networks Effects - Applied TTC Network in 25 years? Getting to Carbon Neutral: - Best Practices - Rules of thumb 3

4 Context - In the battle towards reducing GHG emissions, cities are of paramount importance (Copenhagen climate conference, December 2009) - Transportation is one of the core issues - Due to their environmental benefits, public transit systems are likely to grow - It is clear that t public transit is most often more environmentally sustainable than private automobiles, but: What is the impact of network growth on energy use of subway networks? 4

5 Framework What does energy use mean? - A larger network will emit more because it carries more riders, is it more sustainable or less? - If we account for passengers: should we use energy per ride (boarding) or per passenger kilometres travelled (PKT)? - Note: talk about energy use, not emissions of tco2e because the energy grid is not known, and does not add information about the sustainability of subway networks Energy-use per PKT (MJ/PKT) Streetcar LRT Subway Commuter Rail Toronto Montreal Ottawa Calgary 0.25 Vancouver European Avg North-American Avg prefer to use energy use per ride (total boardings or boardings per capita) 5

6 Framework What directly influences energy use? - Size of the network (route length) Tokyo (292km) vs. Toronto (69km) - Number of operating subway/metro units (wagons) London (3900) vs. Athens (216) - Vehicle Technology 6

7 Framework How to use network characteristics - Absolute Values (as collected) Subway City Route length (km) Metro Units Toronto Montreal Relative Values (route length by population, metro units by population) PKT per track-km Streetcar LRT Subway Commuter Rail Toronto 13,309,671 15,364,738 1,969,227 Montreal 42,702, ,966 Ottawa 3,369,765 Calgary 10,007,934 Vancouver 19,130, ,175 European Avg 3,383,406 26,339,997 5,875,131 North-American Avg 7,935,681 16,166,881 1,817,752 prefer absolute although h consider differences during the analysis 7

8 Dataset and Sources Dataset: - Originally planned to study 19 subway networks in the workd - Narrowed down to 15 cities due to data availability Sources: - Route length: from each individual transit authorities (2008 data) - Energy Use and Metro Units: Millennium Cities Database (1995 data) Note: consider systems that have undergone few changes only 8

9 Results - partial City Boardings (million) Route length (km) Metro Units Total Energy Use in TJ Energy per unit (MJ/km) Toronto Montreal Chicago Washington DC San Francisco Mexico City London Paris Lyon Berlin Athens Stockholm Tokyo Osaka Seoul

10 Vehicle Technology Average = Seoul Toronto Montreal Chicago Washington DC San Francisco Mexico City London Paris Lyon Berlin Athens Stockholm Tokyo Osaka Should we take it in account? We will see that network characteristics are more relevant 10 Energy use per wagon km

11 Boardings and Network Characteristics Impact of Route Length ings per capita Boardi Route Length (km) 11

12 Boardings and Network Characteristics Impact of Route Length Total Bo oardings (million) y = e x R 2 = Route Length (km) 12

13 Boardings and Network Characteristics Impact of Number of Metro Units ings per capita Boardi Metro Units (wagon) 13

14 Boardings and Network Characteristics Impact of Number of Metro Units Total Boa ardings (million) y = 0.149x R 2 = Number of Metro Units (wagons) 14

15 Network Characteristics Route Length vs. Number of Metro Units London Paris Tokyo Number of metro units (wagon n) Osaka Mexico City Berlin Seoul Chicago Montreal Lyon Athens Toronto Stockholm Washington DC San Francisco y = x R 2 = Route length (km) 15

16 Energy Use and Network Characteristics Energy and Route Length Mexico City Tokyo Seoul London Total Energy Use (TJ) Osaka Washington DC Paris Toronto Stockholm Chicago Berlin Montreal San Francisco LyonAthens Route Length (km) 16

17 Energy Use and Network Characteristics Energy and Number of Metro Units Mexico City Tokyo Seoul London Total Energy Use (TJ) Washington DC Chicago Osaka Paris Toronto Berlin San Francisco Stockholm Montreal Lyon Athens Metro Units (wagons) 17

18 Energy Use and Boardings Energy and Boardings Is there a Relationship? Tokyo 2500 Seoul Total Bo oardings (million) Paris London Mexico City 1000 Osaka Berlin 500 Montreal Stockholm Washington DC Athens Toronto Chicago Lyon San Francisco Total Energy Use in TJ 18

19 Energy Use and Boardings Seoul 19 Tokyo Osaka Sto tockholm Montreal Chicago Washing ngton DC San Fra rancisco Mexi xico City London Paris Lyon Berlin Athens Toronto T Energy Use per rid e (MJ)

20 Energy Use and Boardings 7.79 Perhaps the San Francisco BART is too long or network design could be optimized Seoul 20 Montreal Chicago Washingto gton DC San Fran ancisco Mexic ico City London Paris Lyon Berlin Athens Stoc ckholm Tokyo Osaka Toronto To En ergy Use per ride (MJ)

21 Energy Use and Boardings 7.79 Perhaps Chicago is too long, similar to San Francisco 5.90 But still, it is has long as Mexico city, are there any other reasons Montreal Chicago Washingto gton DC San Fran ancisco Mexic ico City London Paris Lyon Berlin Athens Stoc ckholm Tokyo Osaka Seoul Toronto To En ergy Use per ride (MJ)

22 Energy Use and Boardings 7.79 Are these networks underutilized? 5.90 (Characterized as Regional Coverage and Directness) 4.81 What about Network Design? Seoul Montreal Chicago Washingto gton DC San Fran ancisco Mexic ico City London Paris Lyon Berlin Athens Stoc ckholm Tokyo Osaka Toronto To En ergy Use per ride (MJ)

23 Energy Use and Boardings 7.79 Could these networks be overutilized? The smallest systems (Athens and Lyon are along with the largest (Tokyo, Paris). What about their design? Montreal Chicago Washingto gton DC San Fran ancisco Mexic ico City London Paris Lyon Berlin Athens Stoc ckholm Tokyo Osaka Seoul Toronto To En ergy Use per ride (MJ)

24 Energy Use and Boardings 7.79 What about these networks? Montreal Chicago Washingto gton DC San Fran ancisco Mexic ico City London Paris Lyon Berlin Athens Stoc ckholm Tokyo Osaka Seoul Toronto To En ergy Use per ride (MJ)

25 Conclusion - We looked at 15 subway networks in the world to investigate the impact of network size on energy use - Evidently, Total Energy Use of Subway systems is highly correlated with network size - However Energy Use per Ride seems to be relatively independent of network size - Emphasis is put on NETWORK DESIGN since it is correlated with Ridership - New projects should therefore focus on network design in order to favour ridership, which will in turn minimize energy use per ride 25

26 Questions Thank You Questions? Sybil Derrible 26