Comparison of Passenger Rail Energy Consumption with Competing Modes (2015)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

53 C H A P T E R 4 This chapter presents a discussion of MMPASSIM models of the energy consumption and GHG emissions of selected passenger rail services in the United States. The services selected for the case studies represent the variety of commuter, regional intercity and long-distance intercity passenger rail operations found across the country. Single-train simulations were conducted to provide baseline energy consumption and GHG emissions values for each case study service and illustrate the variation in these metrics across different types of passenger rail operations. For each passenger rail service developed as a case study, information about the rail route (e.g., vertical grade profile, horizontal curvature, station stops and timetable speed limits) was col- lected from railroad track charts or publicly available datasets and publications. Baseline energy intensity, GHG emissions and performance metrics for each case study service were established by simulating an appropriate train consist on each route under typical operating conditions. 4.1 Passenger Rail Systems and Services Evaluated As indicated by previous research and the published benchmarks of passenger rail energy efficiency summarized in Chapter 2, rail mode efficiency varies according to the type of service, length of route, average speed and train consist. To capture this potential range of efficiency across passenger rail service in the United States, a spectrum of commuter, regional intercity, long-distance intercity and high-speed rail (HSR) services were selected for evaluation. Selection of case studies was also influenced by • key corridors for ridership and train density; • operations that had recently received upgrades or improvement projects; and • systems with sufficient data available from publicly available sources. Five commuter rail systems, nine regional intercity systems, two long-distance intercity sys- tems, and one HSR system are analyzed via MMPASSIM single-train simulation case studies. 4.1.1 Commuter Rail Services The five commuter rail case study services analyzed are primarily diesel-electric systems that use bi-level coaches characterized by high seating density. Operating characteristics for the com- muter rail systems were obtained from annual reports to the National Transit Database (NTD) and train schedules published on operator websites. 4.1.1.1 Minneapolis: Metro Transit Northstar Metro Transit in Minneapolis, MN, operates the Northstar commuter service over a distance of 40 miles between Big Lake, MN, and Minneapolis. This service uses diesel-electric locomotives Single-Train Simulation of Passenger Rail Energy Efficiency

54 Comparison of Passenger Rail Energy Consumption with Competing Modes and bi-level coaches with high seating density. A typical train consists of one locomotive and three passenger coaches with a total of 426 seats. As a new-start operation serving 10 stations, trains operate only during peak periods and in the dominant direction of commuter traffic (Metro Transit 2014). Route data for the Metro Transit Northstar was obtained from BNSF track charts (BNSF 2005). 4.1.1.2 Chicago: Metra BNSF The Metropolitan Rail Corporation (Metra) is the commuter rail division of the Regional Transportation Authority in Chicago, IL. Metra serves 241 stations on 11 different commuter rail lines (Metra 2014). For this study, the 38-mile BNSF commuter rail line between Aurora, IL, and Chicago’s Union Station was selected for analysis. Metra BNSF service operates with bi- level coaches and two different models of diesel-electric locomotives. One of the busiest Metra lines, this service operates frequently during peak and non-peak hours, using express schedul- ing patterns over different zones where trains only make stops at designated blocks of stations. The selected train schedule is an inbound morning peak run that includes seven station stops (Metra 2012). Local trains operate with a single locomotive, but some express trains operate with two locomotives. The case study train consist includes one locomotive and six passenger coaches with a total of 870 seats. Route data for the Metra BNSF line was obtained from BNSF track charts (BNSF 2005). 4.1.1.3 Seattle: Sounder South Sound Transit in Seattle, WA, operates its Sounder commuter rail route between Seattle, WA, and Tacoma, WA, covering 40 miles with seven stops (Sound Transit 2014). This service oper- ates with diesel-electric locomotives and bi-level coaches with high seating density. A typical train consists of one locomotive and four passenger coaches with a total of 568 seats. This service operates only during peak periods and mostly in the dominant direction of commuting traffic. Route data for the Sounder South line was obtained from BNSF track charts and FRA Improved Passenger Equipment Evaluation Program (IPEEP) report route tables (Bachman et al. 1978). 4.1.1.4 Los Angeles: Metrolink Orange Metrolink, the Southern California Regional Rail Authority, operates the Orange com- muter rail line from the greater Los Angeles region south through Orange County, CA. This service operates during peak and non-peak periods with diesel-electric locomotives and bi-level coaches. A typical train consists of one locomotive and four passenger coaches with a total of 568 seats. This service makes 15 station stops on the 86 miles between Oceanside and Los Angeles, CA (Metrolink 2014). The same route is also simulated with a Pacific Surfliner train consist (as described in the section on regional intercity rail services) that makes nine station stops over the same route segment. Route data for the Metrolink Orange Line was obtained from Union Pacific track charts, BNSF track charts and FRA IPEEP route table data (Bachman et al. 1978). 4.1.1.5 Washington, DC: MARC Penn Line Maryland Area Regional Commuter (MARC) operates the Penn Line service in the Washing- ton, DC and Baltimore, MD, metropolitan areas. The Penn Line serves 13 stations over 77 miles, using both diesel-electric and electric locomotives on a portion of Amtrak’s Northeast Corridor (Maryland Transit Administration 2013). Diesel-electric locomotives are limited to 79 miles per hour, whereas electric locomotives can travel at 125 miles per hour. Bi-level coaches with high seating density are used on both diesel and electric services. Two different case study train consists were examined; both use five passenger coaches with a total of 650 seats, but one uses an electric locomotive and the other uses a diesel-electric locomotive. Route data for the MARC Penn Line was obtained from FRA IPEEP route tables and updated with more recent timetable speed data (Bachman et al. 1978; Amtrak 2008).

Single-Train Simulation of Passenger Rail Energy Efficiency 55 4.1.2 Regional Intercity Rail Services The nine regional intercity passenger rail service case study routes range from approximately 125 miles to 450 miles in length and cover various regions of the United States. Operating char- acteristics for each route were obtained from public Amtrak train schedules. 4.1.2.1 Oklahoma City, OK–Fort Worth, TX: Heartland Flyer Amtrak operates the Heartland Flyer service over a distance of 206 miles between Oklahoma City, OK, and Fort Worth, TX. The route has seven station stops. The service typically operates using one diesel-electric locomotive and three bi-level coaches with a total of 252 seats. Amtrak has tested alternative biodiesel fuels on this route, using a retrofitted diesel-electric locomo- tive (Shurland et al. 2012). Route data for the Heartland Flyer was obtained from BNSF track charts (BNSF 2005). In the past, the service has operated in a push-pull configuration with a non-powered control unit (NPCU). The service now sometimes uses a single locomotive and no NPCU, requiring extra distance at the terminals (3.5 miles total) to turn the train consist at a nearby wye. 4.1.2.2 Charlotte–Raleigh, NC: Piedmont The North Carolina Department of Transportation (NC DOT), in cooperation with Amtrak, operates the 173-mile Piedmont service between Charlotte and Raleigh, NC. A typical train includes one diesel-electric locomotive and four single-level coaches with a total of 336 seats. This service operates two round trips daily with nine station stops. Route data for the Piedmont was obtained from Norfolk Southern track charts (Norfolk Southern 2002). 4.1.2.3 San Jose–Auburn, CA: Capitol Corridor The Capitol Corridor Joint Powers Authority, in partnership with Amtrak, operates the Capi- tol Corridor service between Auburn and San Jose, CA. This service uses diesel-electric locomo- tives and bi-level coaches to offer seven round trips daily on the 168-mile, 17-station route. A typical train consists of one locomotive and four passenger coaches with a total of 360 seats. Route data for the Capitol Corridor was obtained from Union Pacific track charts. 4.1.2.4 New York City–Buffalo, NY: Empire Service Amtrak operates the Empire Service between New York City and Albany, NY, with some of the 18 daily round trips extending to Buffalo, NY. This route has recently been upgraded to emerging HSR service, with trains operating at a maximum of 110 miles per hour on portions of the route (FRA 2009). The Empire Service uses dual-mode locomotives with diesel-electric and electric traction capabilities. Third-rail electric traction is used when entering tunnels to access New York Penn Station. The case study for NCRRP Project 02-01 considers a train covering 438 miles and serving 15 stations between New York City and Buffalo. The study train includes one dual-mode locomotive and five single-level passenger coaches with a total of 416 seats. Route data for the Empire Service was obtained from FRA IPEEP route tables and updated with more recent timetable speed data (Bachman et al. 1978; New York State DOT 2014). 4.1.2.5 Chicago, IL–Detroit, MI: Wolverine Amtrak operates the Wolverine service over a distance of 258 miles and covering 14 stations between Chicago, IL, and Detroit, MI. This service uses diesel-electric locomotives with single- level coaches. A typical train includes one locomotive, four coaches with a total of 336 seats, and a baggage car. This route has recently been improved to support three round trips daily at 110 miles per hour; however, accurate speed limit and track chart data reflecting recent track speed upgrades were not available for use in NCRRP Report 3. Route data for the Wolverine service was obtained from FRA IPEEP route tables (Bachman et al. 1978).

56 Comparison of Passenger Rail Energy Consumption with Competing Modes 4.1.2.6 Portland, OR–Seattle, WA: Cascades Amtrak’s Cascades service between Eugene, OR, and Vancouver, British Columbia, includes service between Portland, OR, and Seattle, WA. This service uses lightweight, tilting, single-axle, articulated single-level Talgo 12-car trainsets with a diesel-electric locomotive and control cab car. Each trainset has 340 seats. The case study examined for NCRRP Report 3 considers a train covering 343 miles and serving eight stations between Portland, OR, and Seattle, WA. Route data for the Cascades was obtained from FRA IPEEP route tables and BNSF track charts and updated with more recent timetable speed data (Bachman et al. 1978). 4.1.2.7 Los Angeles–San Diego, CA: Pacific Surfliner Amtrak operates the Pacific Surfliner service between San Luis Obispo and San Diego, CA, which includes a 127-mile segment between Los Angeles and San Diego. This portion of the route has 15 stations. The service runs 12 daily round trips using diesel-electric locomotives and bi-level coaches. A typical train consists of one locomotive and four passenger coaches with a total of 336 seats. Route data for the Pacific Surfliner was obtained from Union Pacific track charts, BNSF track charts and FRA IPEEP route tables (Bachman et al. 1978). 4.1.2.8 Chicago–Quincy, IL: Illinois Zephyr Amtrak’s Illinois Zephyr service operates over a distance of 258 miles between Chicago and Quincy, IL. This service uses diesel-electric locomotives and single-level coaches, operating two round trips daily and serving 10 stations. A typical train consists of one locomotive, four coaches with a total of 336 seats, and a baggage car. Route data for the Illinois Zephyr was obtained from BNSF track charts. 4.1.2.9 New York City–Washington, DC: Northeast Regional Amtrak operates the Northeast Regional, a higher-speed rail service on the Northeast Corridor, using electric locomotives and single-level coaches and traveling at a maximum speed of 125 miles per hour. This service offers 18 round trips per day between Boston, MA and Washington, DC. The case study examined for NCRRP Report 3 considers one of the trains covering the 226 miles between New York City’s Penn Station and Washington DC, serving 10 stations. The case study train includes one electric locomotive and six single-level passenger coaches with a total of 504 seats. Route data for the Northeast Corridor was obtained from FRA IPEEP route tables and updated with more recent timetable speed data (Bachman et al. 1978; Amtrak 2008). 4.1.3 Long-Distance Intercity Rail Services Long-distance intercity passenger rail services are operated on a limited number of routes by Amtrak. Two different long-distance intercity services were developed as case study routes. Operating characteristics for both services were obtained from public train schedules and route guides published by Amtrak. 4.1.3.1 Chicago, IL–Los Angeles, CA: Southwest Chief Amtrak operates daily Southwest Chief service, covering the distance of 2,251 miles between Chicago, IL, and Los Angeles, CA, with 33 station stops. The service typically operates using two diesel-electric locomotives and eight bi-level passenger railcars, including coaches, sleeping cars, a lounge and a dining car. Given that some of the railcars do not have revenue seats and the sleep- ing cars have limited occupancy, the train carries a maximum of 364 passengers. Route data for the Southwest Chief was obtained from BNSF track charts (BNSF 2005). 4.1.3.2 Chicago, IL–Denver, CO: California Zephyr Amtrak operates daily California Zephyr service between Chicago, IL, and Emeryville, CA (just outside San Francisco). The service typically operates using two diesel-electric locomotives

Single-Train Simulation of Passenger Rail Energy Efficiency 57 and eight bi-level passenger railcars, including coaches, sleeping cars, a lounge and a dining car. Given that some of the railcars do not have revenue seats and the sleeping cars have limited occu- pancy, the train carries a maximum of 364 passengers. This case study considers the portion of the route covering the 1,038 miles between Chicago, IL, and Denver, CO, and serving 16 stations. Route data for the California Zephyr was obtained from BNSF track charts. 4.1.4 High-Speed Rail Services Currently, the only HSR services in the United States operate on the Northeast Corridor and do not exceed 150 mph for extended periods of time. Thus, simulation of the Amtrak Acela does not offer the best illustration of HSR energy efficiency for comparison to benchmarked HSR systems in Europe and Asia. Also, the research team was unable to obtain train resistance coeffi- cients appropriate for the Acela. Although similar in overall appearance and shape to other HSR trainsets with known train resistance coefficients, the higher tare weight of the Acela compared to European and Asian trainsets does not allow these coefficients to be translated directly to the Acela service. 4.1.4.1 Los Angeles–Fresno, CA: California HSR A new, dedicated high-speed (220-mph) rail service between San Francisco and Los Ange- les, CA, is currently being developed. This service will use electric multiple-unit trainsets with high-capacity single-level coaches, similar to European HSR equipment with known train resis- tance characteristics. Preliminary horizontal and vertical geometry data for the 292-mile section between Fresno and Los Angeles, CA, was made available in technical memorandums released during the planning stages of the project (CHSRA 2010; CHSRA 2012). Using this information, the Fresno–Los Angeles segment of the proposed California High-Speed Rail (CAHSR) route was developed into a hypothetical case study to illustrate the HSR capabilities of the MMPASSIM tool. The CAHSR route is assumed to operate with an 11-vehicle electric trainset using distributed power with 446 seats, serving three stations in express service (i.e., with one intermediate stop). 4.1.5 Summary and Ridership Although many of the services developed as single-train case studies may operate different train consists on a daily and seasonal basis, the baseline configurations (Table 4-1) were devel- oped with public data to represent average service conditions on each route. The number of stops listed in Table 4-1 includes the origin and terminating stations for the train run. Average speed is calculated as a final output of MMPASSIM and considers posted passenger train speed, dwell at station stops, and delay for train meets and other unscheduled stops specified in the case study input data. Additional detail on MMPASSIM input parameters for each case study is provided in Appendix E. Ridership and load factor data are required for each case study service to calculate energy and GHG emissions intensities per passenger-mile and passenger trip (Table 4-2). Load factors for the commuter rail system are based on analysis of the 2012 NTD. Amtrak operations are assigned average load factors (provided by Amtrak) based on the type of service provided. The load factor for CAHSR matches the higher load factor commonly used to study the feasibility of HSR systems. 4.2 Baseline Single-Train Simulation Results This section presents baseline single-train simulations of the passenger rail case studies for NCRRP Project 02-01 that were completed using the MMPASSIM spreadsheet tool. Results of each simulation describe the energy intensity and GHG emissions intensity of one round-trip

58 Comparison of Passenger Rail Energy Consumption with Competing Modes Route Origin—Destination Locomotive(s) Trailing Railcars Seats a Stations b Average Speed (mph) Commuter Rail Metro Transit Northstar Big Lake–Minneapolis, MN 1 x 3,600 hp Diesel 3 bi-level 426 10 40 Metra BNSF Aurora–Chicago, IL 1 x 3,150 hp Diesel 6 bi-level 870 7 22 Sounder South Seattle—Tacoma, WA 1 x 3,600 hp Diesel 4 bi-level 568 7 39 Metrolink Orange Oceanside–Los Angeles, CA 1 x 3,600 hp Diesel 4 bi-level 568 15 37 Pacific Surfliner c Oceanside–Los Angeles, CA 1 x 3,600 hp Diesel 4 bi-level 336 9 41 MARC Penn Line Perryville, MD–Washington, DC 1 x 3,100 hp Diesel 5 bi-level 650 13 43 MARC Penn Line d Perryville, MD–Washington, DC 1 x 3,221 kW Electric 5 bi-level 650 13 50 Regional Intercity Heartland Flyer + NPCU Oklahoma City, OK—Fort Worth, TX 1 x 4,100 hp Diesel 3 bi-level 252 7 42 Heartland Flyer Oklahoma City, OK—Fort Worth, TX 1 x 4,100 hp Diesel 3 bi-level 252 7 43 Piedmont Charlotte—Raleigh, NC 1 x 3,600 hp Diesel 4 single-level 336 9 47 Capitol Corridor Auburn—San Jose, CA 1 x 3,600 hp Diesel 4 bi-level 360 17 41 Empire New York—Buffalo, NY 1 x 4,100 hp Dual-mode e 5 single-level 416 15 58 Wolverine Chicago, IL—Detroit, MI 1 x 4,100 hp Diesel 4 single-level, 1 baggage 336 14 57 Cascades Portland, OR—Seattle, WA 1 x 3,600 hp Diesel 12-car Talgo, 1 cab car 340 8 47 Pacific Surfliner Los Angeles—San Diego, CA 1 x 3,600 hp Diesel 4 bi-level 336 15 40 Illinois Zephyr Chicago—Quincy, IL 1 x 4,100 hp Diesel 4 single-level, 1 baggage 336 10 62 Northeast Regional New York City—Washington, DC 1 x 3,221 kW Electric 6 single-level 504 10 68 Long-Distance Intercity Southwest Chief Chicago, IL—Los Angeles, CA 2 x 4,100 hp kW Diesel 8 bi-level 364 33 53 California Zephyr Chicago, IL—Denver, CO 2 x 4,100 hp Diesel 8 bi-level 364 16 65 High-Speed Rail California HSR Fresno—Los Angeles, CA 1 x 9,266 kW EMUf 11-car trainset 446 3 179 aAssumed as typical operations and consist configuration for each case. bCitations for this column can be found in each route’s respective entry in Appendix E. cPacific Surfliner intercity train used as a commuter service. dMARC Penn Line service uses both diesel-electric and electric consists. eDual-mode refers to dual propulsion—a diesel-electric locomotive that can also draw electricity from third-rail or overhead electricity supply lines. fElectric multiple-unit. Table 4-1. Characteristics of passenger rail service case study routes. Route Load Factor Metro Transit Northstar 0.24 Metra BNSF 0.27 Sounder South 0.29 Metrolink Orange 0.25 MARC Penn Line 0.30 Amtrak Regional/State-Supported 0.42 Amtrak Northeast Regional 0.53 Amtrak Long-Distance 0.63 California HSR (CAHSR) 0.60 Table 4-2. Ridership assumptions for evaluated passenger rail services.

Single-Train Simulation of Passenger Rail Energy Efficiency 59 rail movement of the specified train consist. Round-trip simulations were used to average out any directional bias with respect to differences in origin and destination elevation, track profile gradient and relative sequence of speed restrictions and station stops on each route. The results of the single-train simulations consider the passenger rail mode in isolation; the output metrics include only the energy consumed and GHG emissions produced by the passen- ger rail trip and associated fuel and energy production. Later chapters of this report incorporate the energy consumption and GHG emissions of the access and egress legs of passenger rail trips into the analysis. Although the simulated case study services represent typical train consists and operating pat- terns on each route to provide average energy consumption and GHG emissions metrics, one round trip cannot capture the performance of all train movements occurring as part of each passenger rail service. For any given service, certain trains may operate with longer or shorter train lengths and may potentially make a different number of station stops. Wind direction can alter train resistance and resulting performance. Crew members’ differing operating styles can alter throttle settings and locomotive duty cycles. Interference from other rail traffic may vary daily, weekly or seasonally. Small differences in the output metrics are not as important as their relative order of magnitude. Ridership also fluctuates with time of day and day of the week. Because the results expressed per passenger-mile are calculated on the basis of average load factors, actual trains with higher or lower ridership will exhibit varying energy intensity and GHG emissions performance. The seat-mile statistics present a practical maximum best-case performance as if all available seats on the train were occupied. 4.2.1 Commuter Rail The results of the baseline single-train simulations for the commuter rail services (Table 4-3) are comparable to the energy intensities of diesel-electric locomotive-hauled commuter rail sys- tems analyzed in the literature review. In 2011, the energy intensity of similar systems ranged from 423 to 578 Btu per seat-mile, with an average of 486 Btu per seat-mile. Because of their low average load factor, the commuter rail systems exhibit a great dispar- ity between the per seat-mile and per passenger-mile metrics. With the majority of ridership concentrated during peak hours and traveling at much higher load factors, however, the typical passenger commuting by rail is likely to experience lower energy intensity and GHG emissions intensity, approaching the per seat-mile values for their respective train run. Commuter rail Route /seat-mi (Btu) (lb-GHG) /passenger-mi (Btu) (lb-GHG) Travel Time (hrs) Average Speed (mph) Metro Transit Northstar 554 0.098 2,308 0.408 2.02 40 Metra BNSF 423 0.075 1,511 0.267 3.44 22 Sounder South 435 0.077 1,740 0.308 2.07 39 Metrolink Orange 366 0.065 1,462 0.259 4.72 36 Pacific Surfliner * 578 0.102 1,377 0.244 4.24 41 MARC Penn Line 478 0.085 1,593 0.282 3.62 43 MARC Penn Line (electric) 565 0.091 1,883 0.302 3.05 50 *Pacific Surfliner intercity train used as a commuter service between Oceanside and Los Angeles, CA. Table 4-3. Baseline single-train simulations of commuter rail services.

60 Comparison of Passenger Rail Energy Consumption with Competing Modes operations that allow standing passengers can even perform at per passenger intensity levels below the per seat-mile statistics during peak periods. Two interesting sub-comparisons can be made between particular pairs of simulated opera- tions. The first is between the Metrolink Orange and the Pacific Surfliner. Although the Amtrak Pacific Surfliner regional intercity service is targeted toward longer-distance travelers, a pas- senger commuting between Oceanside, CA, and Los Angeles, CA, can potentially use the Pacific Surfliner as an alternative to the Metrolink Orange commuter rail service. Interestingly, although the higher seating density of the Metrolink commuter train allows it to achieve lower energy consumption and GHG emissions per seat-mile for this trip, the higher average load factor of the Pacific Surfliner allows the intercity service to achieve lower energy and GHG emissions intensities per passenger. The MARC Penn Line was simulated with both electric and diesel-electric propulsion. Because the electric train can travel at 125 miles per hour whereas the diesel-electric train is limited to 79 miles per hour, the electric train consist has higher energy consumption and GHG emissions metrics than the diesel-electric train. 4.2.2 Regional Intercity Rail The baseline simulation results for regional intercity services (Table 4-4) are generally more energy efficient and less GHG emissions intense than many of the benchmarks collected dur- ing the literature review for this study. Analysis of the NTD found the energy intensity of the regional Downeaster and Keystone/Pennsylvanian services to be 921 and 859 Btu per seat-mile, respectively. The Amtrak short-haul services also are more efficient than the VIA Rail corridors east and west of Toronto, at 1,046 and 1,156 Btu per seat-mile, respectively (see Table 2-6). The VIA Rail corridors use single-level coaches exclusively, however, and operate at higher speeds than do many of the Amtrak routes. These operating conditions may contribute to the VIA trains’ higher energy and GHG emissions intensities compared to the simulated regional inter- city corridors. The results of the Heartland Flyer simulations offer a comparison between operation with and without an NPCU. The NPCU adds weight and train resistance without increasing ridership or the number of seats, which results in a 19% increase in energy intensity and an 18% increase in GHG emissions. Route /seat-mi (Btu) (lb-GHG) /passenger-mi (Btu) (lb-GHG) Travel Time (hrs) Average Speed (mph) Heartland Flyer with NPCU 583 0.103 1,388 0.246 9.74 42 Heartland Flyer 488 0.086 1,162 0.206 9.67 43 Piedmont 681 0.120 1,621 0.287 7.33 47 Capitol Corridor 561 0.099 1,335 0.236 8.14 41 Wolverine 431 0.076 918 0.162 9.07 57 Cascades 585 0.103 1,393 0.246 14.46 47 Empire 380 0.067 904 0.160 15.11 58 Pacific Surfliner 613 0.109 1,461 0.258 6.42 40 Illinois Zephyr 431 0.076 916 0.162 8.27 62 Northeast Regional 512 0.064 966 0.121 6.60 68 Table 4-4. Baseline single-train simulations of regional intercity rail services.

Single-Train Simulation of Passenger Rail Energy Efficiency 61 4.2.3 Long-Distance Intercity Rail Results of baseline simulation results for long-distance intercity services (Table 4-5) are com- parable to the intercity averages from Mittal of 1,000 Btu per seat-mile. The results are slightly lower than the long-distance averages of VIA Rail Canada services (1,717 and 1,431 Btu per seat-mile for Eastern and Western services, respectively). However, the VIA routes operate with single-level railcars and with more sleeping, lounge and food-service cars than the simulated Amtrak long-distance intercity services. Despite the benefit of bi-level passenger coaches, the presence of food-service and sleeping cars necessitates an additional locomotive that increases energy and GHG emissions intensities per seat-mile compared to regional intercity and commuter rail service. Both the Southwest Chief and California Zephyr were simulated with identical train consists. Thus, the differences in their energy intensity and GHG emissions intensity metrics illustrate the amount of variation that can be created by different route speed profiles, track geometry and station stopping patterns. 4.2.4 High-Speed Rail The baseline simulation results for the California HSR simulation (Table 4-6) far exceed the range of energy intensity values for international HSR services collected during the literature review. It is important to reiterate that, in MMPASSIM, the energy reported for electric locomo- tives includes the regional fuel mix consumed in generating electricity, whereas most operators report the metered kWh consumed by trains and do not include the fuel used to generate that electricity. To compare operator reports against an MMPASSIM prediction, the referenced SYSTRA number must be adjusted to allow for the locomotive losses from wheel to pantograph, the on- board hotel power consumption, the acceleration/braking profile and the gradient profile, and include the fuel consumed in generating the electricity. The research team took the number of 14.36 kWh/train-kilometer (or 245 kJ/seat-mile) from Table 2 in a SYSTRA desk study of carbon impacts (SYSTRA 2011). This number represents the energy consumed at the wheels for a constant speed of 300 kmh, and estimated what it would be if it included other losses and fuel consumed in electricity generation. First, applying the Califor- nia fuel mix would bring the SYSTRA number of 245 kJ/seat-mile to 421 Btu/seat-mile, which is 72% of the simulation output in Table 4-6. The SYSTRA report also indicates results from a full simulation of the TGV-Réseau (SYSTRA 2011, p. 24) as being 22 kWh/train-kilometers and Route /seat-mi (Btu) (lb-GHG) /passenger-mi (Btu) (lb-GHG) Travel Time (hrs) Average Speed (mph) Southwest Chief 864 0.153 1372 0.243 84.8 53 California Zephyr 711 0.126 1128 0.200 31.7 65 Table 4-5. Baseline single-train simulations of long-distance intercity rail services. Route /seat-mi /passenger-mi Travel Time Average Speed (Btu) (lb-GHG) (Btu) (lb-GHG) (hrs) (mph) California HSR 585 0.050 975 0.084 3.3 179 Table 4-6. Baseline single-train simulations of an intercity HSR service.

62 Comparison of Passenger Rail Energy Consumption with Competing Modes also compares the resistive energy at the wheels (SYSTRA 2011, p.10) for the TGV-Réseau as being 16.25 kWh/train-kilometers (as indicated in comparison with the AGV-11 in Table 2-20 of this report). This ratio of 16.5/22 = 75% is close to the 72% ratio between our simulation of the AGV and the SYSTRA report’s energy at the wheels scaled to include fuel consumption. It does not vali- date the AGV parameters we estimated, but does indicate that MMPASSIM’s simulation results are close to those of other simulations when adjusted to provide a like-for-like comparison. On a per passenger-mile basis, the simulated California HSR compares favorably to all but the most energy-efficient simulated regional intercity passenger rail services while simultaneously increasing average speed by a factor of three to four. Given its use of relatively clean electricity generation sources in California, the simulated Cali- fornia HSR route has the lowest GHG emissions intensity of all simulated passenger rail services per seat-mile and passenger-mile. 4.2.5 Discussion Ranking the simulated passenger rail services by energy intensity per passenger-mile high- lights the influence of load factor on the results (see Figure 4-1). Although the possible energy intensity of the passenger rail services clusters around 500 kJ/seat-mile, the energy intensity per passenger-mile ranges from under 1,000 Btu to about 2,300 Btu depending on the service’s load factor. The commuter rail services, which have the lowest average load factor, exhibit the highest energy intensity. As will be further demonstrated in subsequent chapters, combinations of high load factor, stop spacing and train length make several of the regional intercity passenger services the most efficient passenger rail operations. Figure 4-1. Energy intensity of simulated passenger rail services (Btu).

Single-Train Simulation of Passenger Rail Energy Efficiency 63 The GHG emissions intensity per passenger-mile exhibits similar trends (Figure 4-2). Wider variation is seen in the per seat-mile metric, however, because of the electrified operations on several routes. The top two routes in terms of pounds of greenhouse gas (lb-GHG) per passenger- mile are both electrified routes, and the system that ranks third uses dual-mode operation. The California HSR case study route ranks at the top for GHG emissions and near the top for energy intensity on a per passenger-mile basis. Lightweight equipment, optimized align- ment design, high load factor and relatively clean sources of electricity all combine to offset the increased energy demands of operation at speeds near 200 mph. The ratio of GHG emissions to energy consumed for each simulated route is relatively con- stant for all diesel locomotive-powered case studies, as exhibited by the linear form of the plot in Figure 4-3. This result is expected, given that the diesel locomotive combustion process that Figure 4-2. GHG emissions intensity of simulated passenger rail services (lb-GHG). Energy Intensity (Btu/seat-mi) Em iss io ns (l b-G HG /se at- mi ) Figure 4-3. Ratio of GHG emissions to energy consumed by simulated passenger rail services.

64 Comparison of Passenger Rail Energy Consumption with Competing Modes controls the emissions per unit of energy produced is essentially the same for all case study routes using diesel propulsion. The electric traction case studies fall below this relationship, however, because their emissions are controlled by the source generation of electricity. The dual-mode New York–Buffalo (Empire Service) route falls very near the diesel-electric trend because the majority of the trip is spent in diesel mode. The other two electric cases are farther below the diesel-electric trend, indicating that they involve cleaner sources of electricity and produce less overall GHG emissions per unit of energy consumed than diesel-electric traction.