Department of Transportation – Rail Division

Oregon Rail Study Appendix I

Wilsonville to Salem Assessment

Prepared by: Parsons Brinckerhoff Team Parsons Brinckerhoff Simpson Consulting Sorin Garber Consulting Group Tangent Services Wilbur Smith and Associates

April 2010

Table of Contents EXECUTIVE SUMMARY...... 1 INTRODUCTION...... 3

WHAT IS COMMUTER RAIL? ...... 3

GLOSSARY OF TERMS...... 3

STUDY AREA...... 4

WES COMMUTER RAIL...... 6

OTHER PASSENGER RAIL SERVICES IN THE CORRIDOR ...... 6 OUTREACH WITH RAILROADS: PNWR AND BNSF ...... 7

PORTLAND & WESTERN RAILROAD...... 7

BNSF RAILWAY COMPANY ...... 7 ROUTE CHARACTERISTICS...... 8

DATA COLLECTION ...... 8

FREIGHT RAIL OPERATIONS ...... 10

LAND USE CHARACTERISTICS...... 10

TRACK RIGHT OF WAY AND CONDITIONS...... 18

TRANSIT FACILITIES AND SERVICES ...... 22

ROUTE CONSIDERATIONS IN NORTH AND CENTRAL SALEM...... 26 OPERATING PLAN ASSUMPTIONS ...... 28

ALIGNMENT ...... 28

OPERATING SCHEDULE...... 28

TRAIN EQUIPMENT ...... 29

STATIONS ...... 30

MAINTENANCE FACILITY OPTIONS ...... 41 DATA ANALYSIS...... 42

RIDERSHIP FORECASTING ...... 42

POLICY ASSESSMENT ...... 49

CONCEPTUAL CAPITAL COST ESTIMATES...... 50

CONCEPTUAL OPERATING COST ESTIMATES ...... 53 RESULTS OF THE FEASIBILITY ASSESSMENT ...... 53

COST AND USER BENEFIT ANALYSIS ...... 53

FINANCING PLAN ...... 54

GOVERNANCE ISSUES...... 55 SUMMARY OF KEY FINDINGS...... 57 APPENDIX A: LESSONS LEARNED FROM WES...... 59 APPENDIX B: OPERATING ASSUMPTIONS...... 72 i APPENDIX C: RIDERSHIP FORECASTING...... 83 APPENDIX D: COST ESTIMATES...... 93

List of Tables Table 1: Operating Scenarios ...... 29 Table 2: Daily Ridership Projections by Scenario ...... 46 Table 3: Ridership Projections Comparison ...... 47 Table 4: Projected Wilsonville-to-Salem Extension Ridership vs. Current Ridership Data ...... 48 Table 5: Conceptual Capital Cost Estimate ...... 52 Table 6: Conceptual Annual Operating Cost Estimates ...... 53

List of Figures Figure 1: Study Area Including Existing WES Alignment...... 5 Figure 2: Industrial Areas in the Corridor ...... 12 Figure 3. Existing Land Use at Wilsonville Station Area ...... 13 Figure 4. Existing Land Use at the Possible Woodburn Station Area...... 14 Figure 5. Existing Land Use at the Possible Keizer Station Area ...... 15 Figure 6. Existing Land Use at the Possible North Area ...... 16 Figure 7. Existing Land Use at the Possible Central Salem Station Area...... 17 Figure 8: Existing Conditions and Identified and Operations Issues (North Portion) ...... 20 Figure 9: Existing Conditions and Identified Track and Operations Issues (South Portion)...... 21 Figure 10: SMART Central at Wilsonville Station ...... 23 Figure 11: Cherriots Route Map (Keizer, north and central Salem)...... 25 Figure 12: Front Street NE & Norway Street NE ...... 26 Figure 13: North Salem ...... 27 Figure 14: Maple St. NE at Spruce St. NE...... 27 Figure 15: Salem Riverfront Downtown Area...... 28 Figure 16: Station Locations and Passing Options...... 32 Figure 17: Possible Woodburn Station Area...... 35 Figure 18: Woodburn Comprehensive Plan in Possible Station Area ...... 36 Figure 19: Possible Keizer Station Area...... 37 Figure 20. Possible Salem Station Options...... 39 Figure 21. Photos of the Alignment in Salem (Photos locations noted on Figure 20) ...... 40 Figure 22: Existing WES Maintenance Facility...... 41 Figure 23: Peak Period Volumes on I-5 vs. Capacity ...... 45

ii Executive Summary Commuter rail or “interurban” rail service began declining in the U.S. with the advent of the automobile before World War II. By the end of the 1950s, commuter rail no longer existed in Oregon. Through much of the past 20 years there have been varying levels of interest in reinstituting commuter rail service in Oregon, resulting in a number of commuter rail studies initiated during that timeframe. In the mid 1990s, Washington County and the greater Portland metropolitan area began seriously considering a commuter rail service in suburban Washington County, which culminated with the opening of the Beaverton-to-Wilsonville Westside Express Service (WES) commuter rail line in early 2009 over trackage operated by Portland & Western Railroad (PNWR).

Currently, 70,000 to 90,000 passenger vehicles (approximately 90,000 to 120,000 person trips) travel each weekday on Interstate-5 (I-5) between the southern Portland metropolitan area (Wilsonville) and Salem with peak travel during the morning and evening commute periods. By 2030 these levels are expected to nearly double to 130,000 to 160,000 passenger vehicles per day (approximately 150,000 to 200,000 person trips). Commuter rail service provides an opportunity to potentially divert some of these vehicle trips from the state highway system to a parallel rail line, but also faces challenges. In addition to the routine financial, engineering and environmental hurdles a project of this size encounters, acquiring the railroad’s approval is a surmountable challenge on its own.

In order to assess the challenges and opportunities of extending commuter rail from Wilsonville to Salem, this study examines information collected through a variety of sources including interviews with the analysts and decision-makers who developed and implemented WES service. This analysis provides an understanding of the project planning and construction process associated with WES as well as insight into the possible extension.

Extending commuter rail service to Salem has the potential to attract 3,000 to 4,000 riders per day by 2030, approximately 15 percent of whom are using the existing WES route1 (these riders are traveling to points north of Wilsonville and are currently driving to the existing Wilsonville park-and-ride station), with a slight benefit to congestion on I- 5 and Highway 99E between the south Portland metropolitan area and central Salem. Station locations considered in this analysis include: Wilsonville, Woodburn, Keizer, and North Salem or Central Salem. Capital construction cost for the extension in 2008 dollars is estimated to be $327-$387 million, which includes mitigation projects to offset current and future impacts to PNWR’s freight operations both along the extension segment as well as the existing

1 Based on discussions with TriMet and URS staff.

1 WES corridor. Depending on the operating scenario analyzed, annual operations and maintenance costs for the extension could add $5.5 million to over $6.9 million to the $2.25 million annual operating costs for the WES being incurred by TriMet. This report concludes that extending commuter rail to Salem is technically feasible, but faces many political and financial challenges. In order for the extension to become reality, these challenges will need to be successfully met:  There are many rail-served industrial users in the rail corridor who rely on PNWR to maintain efficient freight service. The introduction of passenger rail stations and transit friendly land uses, such as residential and commercial development, adjacent to the rail line may negatively impact PNWRs freight service by driving out industrial customers. A balance between encouraging commuter rail ridership and limiting the impacts to freight service must be found to benefit the duel use of the corridor.  Funding for the WES project was generated from four sources: Federal Transit Administration (FTA), Washington County, TriMet, and the State of Oregon. It is unlikely that either Washington County or TriMet would be a major funding partner in the extension project. Funding within Marion County has not been identified. In addition, achieving FTA funding at comparable levels to WES might prove more difficult due to the land use patterns at the possible station locations. The auto-oriented characteristics at the Keizer and Woodburn station locations will not measure well against the FTA criteria supporting transit oriented land uses and transportation policies.  A key lesson learned from the WES project was the importance of having dedicated political champions who continue to support the project even in times of controversy. For the extension to move forward, an equivalent level of support from Marion County leaders and city leaders in Wilsonville, Donald, Woodburn, Keizer and Salem will be necessary. The report is not intended to recommend whether or not commuter rail service should be implemented, but rather to provide a preliminary assessment to generate discussion about the potential advantages, disadvantages, and feasibility of the extension. Further, this study does not serve as the FTA required Alternatives Analysis/Draft Environmental Impact Statement (AA/DEIS) process (included in cost estimate). An AA/DEIS detailed analysis, which addresses the criteria and requirements for developing commuter rail projects, including additional stakeholder outreach, would be necessary to determine whether further development of the extension is warranted. The AA/DEIS effort could take about up to 2-3 years to complete, with a preliminary estimated cost ranging between $3 and $5 million.

2 Introduction This preliminary feasibility study examines extending commuter rail to Salem through interviews with the analysts and decision-makers who developed and implemented the WES service. The report is not intended to recommend whether or not commuter rail service should be implemented, but rather to provide a preliminary assessment and to generate discussion about the potential advantages and disadvantages of the extension. A more detailed analysis, which addresses the criteria and requirements for developing commuter rail projects, including additional stakeholder outreach, would be necessary to determine whether further development of the extension is warranted.

What is Commuter Rail? Commuter rail transit provides passenger rail service along railroad tracks between cities, suburbs, central business districts (CBD), and other employment and residential areas. Service is often limited to the morning and evening commute hours and generally provides an appealing transportation alternative for employees working daytime, weekday hours (typically Monday through Friday, for work shifts that typically start between 6:00 and 9:00 a.m. and end between 3:00 p.m. and 6:30 p.m.), and who live and work in the corridor vicinity or can use connecting transit service. Unlike transit, which generally operates on exclusive-use track with stops located every one-quarter mile to one mile, commuter rail typically operates on the same track as freight rail operations and typically provides station stops in key cities and towns along a corridor, typically no closer that 2-4 miles between stations.

Commuter rail transit also serves a distinctly different purpose than intercity passenger service (e.g., ). While commuter rail transports people to and from their daily employment locations during periods, intercity passenger rail service usually connects major population centers, such as Portland and , over greater distances. Intercity rail also operates throughout the day and stops less frequently than commuter rail, although some commuters do use intercity rail for work purposes.

Glossary of Terms The following terms are used within this report:

 AA/DEIS: Alternatives Analysis/Draft Environmental Impact Statement: Federally required process to determine the alignment which would result in the best configuration and least impact to the people and environment in the corridor.  DMU: Diesel-: Self-propelled passenger coaches, similar to the vehicles used for WES, where a diesel-powered unit provides both power and passenger seating within the same vehicle and can pull at least one additional unpowered car.  EMU: Electrical-multiple unit; Self-propelled passenger coaches powered by electricity, instead of internal combustion. The powered unit can operate singly or with multiple powered cars and may also pull unpowered trailer cars.

3  Locomotive hauled : “Traditional” types of passenger rail trains where a locomotive (which does not carry passengers) pulls one or more passenger cars along the route.  OE Line: The former Oregon Electric route over which PNWR operates, between Salem and Portland.  Quiet Zone: In compliance with the Federal Horn Rule,2 quiet zones provide crossing improvements and signalization requirements that allow trains to no longer blow their warning whistles at road crossings. The rule was established in 2005 and is administered by the Federal Railroad Administration (FRA) in response to congressional action. Quiet zones provide opportunities for communities to mitigate the effects of noise.

Study Area The Wilsonville-to-Salem commuter rail extension study area extends 29 miles along the OE line alignment from Wilsonville to Salem.3 Station locations include Wilsonville, Woodburn, Keizer and Salem. Two separate station locations were studied in Salem, one in north Salem and one in central Salem. The alignment is shown in Figure 1.

In 2009 it is estimated that over 600,000 people reside, and nearly 300,000 people are employed, within the study area shown in Figure 1. By 2030 both of these figures will almost double to nearly one million residents and 500,000 employees within the study area.4 Currently, 70,000 to 90,000 passenger vehicles (approximately 90,000 to 120,000 person trips) travel each weekday on I-5 between the south Portland metropolitan area (Wilsonville) and Salem with peak travel during the morning and evening commute periods. By 2030 these levels are expected to nearly double to 130,000 to 160,000 passenger vehicles per day (approximately 150,000 to 200,000 person trips).

2 Federal Railroad Administration. Final Rule on the Use of Locomotive Horns at Highway­Rail Grade Crossings. 2005. http://www.fra.dot.gov/us/Content/1318. October 1, 2009. 3 PNWR has expressed need for additional facility upgrades along the existing WES alignment for freight operations, as discussed later in Capacity Analysis section within the Data Analysis chapter. 4 Estimates are based on the June 2009 County‐level Long Term Forecast provided by Global Insight. Estimated 2009 and 2030 corridor population and employment are based on the relationship of population within the corridor as compared to the county.

4 Figure 1: Study Area Including Existing WES Alignment

5 WES Commuter Rail WES Commuter Rail began service between the and the Wilsonville Transit Station on February 2, 2009. Three intermediate stations are located along the 14.7 mile alignment: Hall/Nimbus Station; and Tualatin Station. WES operates on the former OE line from 5:20 to 9:20 a.m. and 3:30 to 8:00 p.m. Monday through Friday, with limited service on holidays.5 Service during these peak commuting hours is generally every 30 minutes for a total of 16 round trips each day - eight round trips in the morning and eight round trips in the evening.

TriMet, the Portland-metro regional transit authority, designed and constructed all of the commuter rail facilities and currently sponsors the WES service. Funding for WES was provided from four sources including: the State of Oregon, Washington County, TriMet and federal funding through the FTA’s Section 5309 New Starts program.

Service is provided via three self-propelled DMU and one trailer; each of the three self-propelled rail cars seats 74 passengers, while the one trailer seats 80. Trains are operated by PNWR crews. A maintenance facility was constructed at the Wilsonville Station and is currently near maximum capacity; however land and right of way are available to expand the maintenance facility in the future. Currently, 1,200 passengers commute on WES each day, about 500 fewer than anticipated prior to opening; approximate daily capacity is 4,200 passengers.6

Other Passenger Rail Services in the Corridor Extension of commuter rail from Wilsonville to Salem could provide a regional passenger rail service that complements the and intercity passenger routes. Amtrak Cascades operates two round trips per day on the (UPRR) mainline between Eugene and Portland, with stops in Albany, Salem and Oregon City before continuing on to Seattle and Vancouver B.C. The Coast Starlight is an interstate Amtrak train using the UPRR and BNSF Railway Company (BNSF) mainline providing daily passenger rail service between Seattle and Los Angeles with Oregon stops in Portland, Salem, Albany, Eugene, Chemult and Klamath Falls.

ODOT is currently studying the feasibility of shifting intercity passenger rail service from the UPRR mainline to the OE route, which could result in a potential scenario of freight, intercity passenger rail, and commuter rail service sharing the same corridor. Both of these intercity rail scenarios have been incorporated into this feasibility assessment.

5 From TriMet website, http://trimet.org/schedules/r203.htm. Accessed October 5, 2009. 6 John Griffiths, TriMet. Personal Communication. December 17, 2009.

6 Outreach with Railroads: PNWR and BNSF Privately owned railroads—like the owners of the OE line—can decide who, when, where, and if a commuter rail operation may exist on their system. That is unlike Amtrak, which the railroads must allow to operate over their networks. Therefore, the importance of a good relationship with host railroads cannot be underestimated. Information provided by the railroads was used in this feasibility assessment, specifically to conduct operating and capacity assessments and to analyze costs for construction, rail operations, maintenance and insurance.

There are three separate right of way owners along the OE line from Wilsonville to Salem: the State or Oregon owns the right of way between Wilsonville to just north of Keizer and BNSF owns the stretch from North of Keizer to Salem. The operating rights of the ODOT-owned right of way are owned by PNWR. The sections owned by BNSF are leased to PNWR. BNSF and PNWR must agree in order to allow commuter rail on their system in order for it to happen. Both PNWR and BNSF were interviewed to discuss their willingness to entertain extending commuter rail to Salem.

Portland & Western Railroad PNWR’s experience with WES has left them with valuable lessons that they are intent on not repeating. WES has unintentionally reduced their freight capacity between Beaverton and Wilsonville. PNWR cannot operate freight trains concurrently with the WES passenger trains. As a result, PNWR has had to refrain from operating freight trains through the WES corridor during hours that commuter trains operate and some shipments have moved to truck. Given the impact to current operations and concern for the future, PNWR does not support expansion of WES on its system at this time. See Appendix A for lessons learned from WES.

BNSF Railway Company BNSF owns the railroad track structure and right of way from Perkins Road crossing north of Keizer through Salem. BNSF representatives were asked about the concept of a commuter rail extension to Salem, their experience with other commuter rail projects, and how commuter rail would be different from intercity passenger rail.7 Key issues identified in this interview include:

 The commuter rail operating plan must be developed early in the project planning process because frequency and bi-directional operation significantly affect infrastructure plans and double tracking needs, resulting in significant impacts to capital costs.

7 DJ Mitchell, BNSF AVP Passenger Operations, Eddie Hale, BNSF Manger, Shortline Relations. Personal Interview. April 21, 2009.

7  BNSF representatives agreed that for both Intercity and commuter rail operation of passenger trains requires significant attention to capacity improvements to ensure that freight trains can adequately navigate the corridor.  BNSF will require indemnification similar to that already in place for PNWR. BNSF described this requirement as being similar to the “sovereign immunity” enjoyed by many local government entities. Other commuter rail authorities have satisfied this requirement through contractual provisions and insurance, or alternatively, through the enactment of statutory protections. BNSF also expressed a possible willingness to sell the corridor to ODOT at market value, which would likely reduce or eliminate the requirement to indemnify BNSF against potential liability.  BNSF representatives encouraged early engagement of railroad and public agencies in the commuter rail planning process to ensure that challenges are identified early and solutions are developed without compromising either freight or passenger services. Route Characteristics This section of the assessment describes the data sources used to collect information about the existing route and then describes the land uses, transit service, right of way, and track condition characteristics of the OE rail line between Wilsonville and Salem.

The existing OE rail line between Wilsonville and Salem passes through a diversity of land uses, ranging from farm land to single family residential housing in suburban communities to vibrant downtowns with a mix of uses, urban densities and existing transit services. Track conditions along the alignment are adequate to serve its current purpose as low density freight railroad, but would need to be rebuilt to support passenger rail service.

Data Collection Information about the existing conditions within the study area was collected through a variety of sources:

 Railroad Track Charts: Track charts depicting the entire alignment were reviewed to identify the existing infrastructure and land area within the railroad right of way.  Population and Employment Statistics: The 2000 U.S. Census population and journey-to-work statistics, and the June 2009 County-level Long Term Forecast by Global Insight were used to develop estimates of regional and corridor population and employment statistics.  Freight Operations: Information about freight train operations, including the number of trains per day (current and future), running speed, switching operations, etc. was gathered from interviews with PNWR staff, consultant team experience with the PNWR system and customers along the route, and observations during various site visits.

8  Site Visits: Site visits were conducted at each of the conceptual station sites and other significant locations along the route to gain a first-hand understanding of the track conditions and existing land uses along the alignment.  Land Use Plans: Zoning and comprehensive plans were reviewed for jurisdictions along the alignment. These include: Washington and Marion Counties; the cities of Wilsonville, Donald, Woodburn, Keizer, and Salem; and the community of Hopmere/Brooks. Additionally, the Keizer Station Plan8 was referenced as it includes a location for a possible commuter rail station.  Mapping: Aerial photos along the alignment were viewed through a geographic information system (GIS) program to gather an understanding of the terrain within the study corridor. Interviews with Washington County and TriMet: In addition to the railroad interviews described in the previous section, interviews were held with TriMet and Washington County staff, the other key players in the implementation of WES commuter rail. These agencies provided data, reports and other information specific to commuter rail for consideration in this preliminary feasibility analysis.  Commuter Rail Ridership (Current and Future Forecasts): Current WES ridership, travel flow between Wilsonville and Salem, and other systems for comparison purposes were gathered from a variety of sources: o TriMet ridership reports for WES, which showed approximately 1,200 riders per day as of June 2009. o American Public Transit Association (APTA), commuter rail ridership reports for the third quarter of 2008, for national commuter rail systems9 This data was used to compare the Wilsonville-to-Salem extension projected ridership with that of established systems with similar characteristics for the purposes of gauging how well the proposed project would compete for FTA funding. o Interviews with agencies in the study area identified previous Wilsonville- to-Salem commuter rail studies, which included preliminary ridership forecasts and cost estimates from several different sources. These reports were used to inform and validate passenger ridership forecasts and conceptual cost estimates. Previous studies include: . Metro: High Capacity Transit Study, Draft Detailed Corridor Evaluation (2009).

8 Keizer Station Plan, as amended July 2007, City of Keizer: http://www.keizer.org/commdev/KS/Keizer%20Station%20Plan.pdf. 9 http://www.apta.com/research/stats/ridership/riderep/documents/08q3cr.pdf

9 . Salem-Keizer Area Transportation Study (SKATS): Commuter Rail Sketch-Level Analysis White Paper (2002). . TriMet: WES Salem Extension: Market Analysis and Ridership Assessment (URS, 2009). . HDR. Commuter Rail Feasibility between Wilsonville Salem/Keizer. (PowerPoint Presentation, 2009).

Freight Rail Operations Currently, PNWR operates an average of two northbound and two southbound freight trains per day in the study corridor, but at times PNWR operates more. Freight trains average 25 mph along the corridor. PNWR performs switching operations at a number of locations along the corridor: Hopmere, West Woodburn, and north Salem are the most frequent locations for switching operations. By 2030, the number of freight trains is expected to double.

PNWR indicates that they have reduced or eliminated freight service during the morning and afternoon weekday peak periods in the WES corridor, and have curtailed midday service as well, in order to provide capacity for commuter rail service. PNWR is under agreement with TriMet to operate the WES trains and to ensure that they have priority on the PNWR track.

Land Use Characteristics Wilsonville Station is the current southern terminus of WES commuter rail and the starting point of the extension studied in this assessment (see Figure 1). The station is located on the west side of I-5 on Barber Street. Properties abutting the station to the north, east and south are zoned Planned Development Industrial (PDI), which is intended to provide opportunities for a variety of industrial operations and associated uses. The existing developments on these properties include industrial uses such as distribution centers and manufacturing plants. The land to the west of the station is zoned Residential Agriculture Holding Industrial (RAHI) and was developed as a transit transfer station and park-and-ride lot when the Wilsonville Station was built (see Figure 3).

Further south on the alignment, Donald is an incorporated city located between Wilsonville and Woodburn with just over 1,000 residents.10 There are existing freight customers in Donald on the OE line. The rail line intersects with Main Street and is surrounded by single family residential and commercial zoning and is developed as such.

The city of Woodburn is located entirely east of the alignment (see Figure 4), and the railroad is adjacent to Woodburn’s westward urban growth boundary (UGB). The

10 Portland State University. Population Research Center. 2008 Oregon Population Report.

10 existing land use zoning along the alignment is primarily Rural Residential and Primary Agriculture. The city’s comprehensive plan map shows that the land within the UGB near the alignment is designated for Low Density Residential (LDR) uses, which permit developments such as single family residences, group homes and child day care. Lands to the west of the alignment, outside of the UGB, are zoned Exclusive Farm Use (EFU) on the Marion County comprehensive plan map. Just west of Woodburn are several existing agricultural based rail-served industrial customers.

Woodburn Company Stores, an auto-oriented, regional shopping destination with approximately 100 retail outlet stores, is located in the northwest corner of the city, about one mile driving distance east from the alignment.

Another industrial area, Loganville, just north of Woodburn, has active agricultural related shippers.

South of Woodburn, Hopmere is an unincorporated community outside of an established UGB. It is adjacent to the community of Brooks which is east of the alignment; between the two communities, there is a combined population of fewer than 400 people.11 The route travels along Unincorporated Community Industrial (UCI) and Acreage Residential (AR) zoned parcels. UCI zoning allows for more intensive industrial development than rural designations outside unincorporated communities. Hopmere has been identified as an industrial area serving agricultural and construction product needs.

11 Brooks‐Hopmere Community Plan. 2007. http://www.co.marion.or.us/PW/Planning/zoning/ruralcommunityplans/brookshopmere. Accessed June 22, 2009.

11 Figure 2: Industrial Areas in the Corridor

12 Figure 3. Existing Land Use at Wilsonville Station Area

13 Figure 4. Existing Land Use at the Possible Woodburn Station Area

14 Figure 5. Existing Land Use at the Possible Keizer Station Area

15 Figure 6. Existing Land Use at the Possible North Salem Station Area

16 Figure 7. Existing Land Use at the Possible Central Salem Station Area

17 The city of Keizer is adjacent to the city of Salem with an estimated 2008 population of over 36,000.12 The alignment runs through the eastern section of the city where a planned development, called Keizer Station, is under development (see Figure 5). The Keizer Station Plan was developed under the “activity center design plan” guidelines of the Keizer Development Code, which calls for diverse land uses including service commercial, industrial, housing, retail, office and public uses and identifies the location of a possible commuter rail station.13 Development of a portion of the Keizer Station area to the east of the alignment has occurred over the past three to five years, and the development plan identifies additional phases not yet built. This recent development is a portion of the larger Keizer Station Center development that includes a mix of commercial and mixed uses including Target, Lowes, and Cost Plus, among others. Currently there are residential uses to the west of the alignment; however, the Keizer Station Plan calls for this area to be redeveloped as a Retail Service Center, which includes a mix of commercial and low density residential uses.

Continuing south, the OE line runs through the heart of Salem, which has a population of over 154,500.14 In the northern part of Salem the line passes through an active industrial area and then heads toward the and downtown Salem (see Figures 6 and 7). Zoning districts along the southern portion of the Salem alignment range from single and multi-family residential to general and neighborhood commercial. The final portion of the alignment ends along Front Street in Salem’s Central Business District zone.

Track Right of Way and Conditions The alignment from Wilsonville Station to Salem could support passenger train speeds up 110 mph, if the track is rebuilt, except where curvature requires reduced speeds. The existing right of way is large enough for two mainline tracks; a few locations where there is less than 50 feet of right of way may require that the existing track be shifted to accommodate a second track for commuter rail operations.

Both buried and overhead utilities are located within the right of way in the Wilsonville city limits and at other locations along the corridor. Utilities may need to be relocated to accommodate an additional track.

The single-track Willamette River Bridge near Wilsonville was built in 1976 and is in good condition. This study assumes the bridge would remain as a single track. The alignment also crosses the 570-foot Lake Labish concrete bridge upon entering Keizer city limits.

12 Portland State University. Population Research Center. 2008 Oregon Population Report. 13 Keizer Station Plan. 2007. http://www.keizer.org/commdev/KS/Keizer%20Station%20Plan.pdf. Accessed June 23, 2009. 14 Portland State University. Population Research Center. 2008 Oregon Population Report.

18 Track through north Salem is located within Front Street with limited separation from vehicular and pedestrian traffic along with several at-grade crossings. Improving speed would require a shift of the mainline to separate the railroad track from the street and to eliminate conflicts between different uses at intersections. Speeds through Salem would be limited to less than 40 mph and a quiet zone would help reduce impacts on adjacent neighborhoods.

19 Figure 8: Existing Conditions and Identified Track and Operations Issues (North Portion)

20 Figure 9: Existing Conditions and Identified Track and Operations Issues (South Portion)

21 Transit Facilities and Services Many of the communities along the OE line are served by transit services with existing facilities near the alignment.

Together, the transit transfer station and the park-and-ride lot west of the Wilsonville Station are called SMART (South Metro Area Regional Transit) Central at Wilsonville Station (SCWS). SCWS is a transfer station between WES, all SMART routes, Salem- Keizer Transit Agency, Cherriots, and (CAT) as shown in Figure 10. SCWS includes a 400 space park-and-ride lot with 48 free bicycle lockers.

SMART service includes five bus routes and one shuttle providing radial service to areas around the south metro area from SCWS. The Cherriots, jointly with SMART, provide bus service Route 1X between downtown Salem and SCWS. Frequency of Route 1x is hourly during the morning commute and every half hour during the evening commute. Route 2X, operated by SMART, provides service from SCWS to Barbur Boulevard Transit Center every half hour during the morning and evening peak commuting times, and hourly during non-peak periods.

SMART and CAT provide joint service via Route 3, the Purple Line, to Canby. Service is hourly Monday through Friday with a break in service between 9:00 and 10:30 a.m. and between 1:30 and 4:00 p.m. SMART Route 4 is a Wilsonville circulator providing service nearly every half hour during the day except between 9:30 a.m. and 2:30 p.m. when service is every 40 to 60 minutes. SMART Route 5 provides service to the Commerce Circle, which is also served by TriMet bus 96. Frequency of service is every half hour between 5:30 and 10:00 a.m. as well as between 3:30 and 7:00 p.m. with no mid-day service. SMART Route 6 provides service to Canyon Creek and the SMART office located in downtown Wilsonville east of I-5 and north of Elligsen Road; frequency is every half hour between 6:30 and 10:00 a.m. and 3:30 and 7:30 p.m. with no mid-day service.

Further south, the Woodburn Public Works Department’s Transit Division provides hourly bus circulation service throughout Woodburn.

Chemeketa Area Regional Transportation System (CARTS) is a bus transportation partnership between Marion, Polk and Yamhill counties. Route 10 provides service between Woodburn and Salem with departures from Woodburn Transit Station at 6:40 a.m., 9:00 a.m., 3:15 p.m., and 5:10 p.m. At this time, there is no transit route connection to the possible Woodburn commuter rail station location.

22 Figure 10: SMART Central at Wilsonville Station

Source: SMART. http://www.ridesmart.com/Index.aspx?page=128. Accessed August 13, 2009.

23 In Keizer, Cherriots provides transit service from the Keizer Station development as shown in Figure 11 with four bus routes (4, 15, 18, and 19) circulating to the Salem Transit Mall, Chemekta Community College, or throughout Keizer. Route 4, Keizer Station Express, provides service between the Salem Transit Mall and Keizer Station every half hour during the weekdays. Route 15, Keizer Station/CCC, provides service between the Keizer Station and Chemekta Community College every half hour on weekdays. Route 18, Keizer Crosstown Loop, provides service between the Keizer Station and Windsor Island Road every half hour on weekdays. Route 19, River Rd N/Keizer Station, provides service between Keizer Station and Salem Transit Center every half hour on weekdays.

At the terminus of the commuter rail extension in Salem, the alignment could be served by all downtown Salem-oriented Cherriots bus routes.

24 Figure 11: Cherriots Route Map (Keizer, north and central Salem)

Source: Cherriots. http://www.cherriots.org/Downloads/cherriots_system_map.pdf. Accessed January 12, 2010.

25 Route Considerations in North and Central Salem A prior study suggested that the initial phase of the commuter rail extension should terminate at Keizer or north Salem due to challenges of extending commuter rail through north and central Salem.15 Although terminating the extension in north Salem would reduce immediate impacts to adjacent communities and relieve some design issues, it would force the majority of Salem-bound riders to transfer to another transit mode before reaching central Salem’s employment and commercial core and would likely reduce ridership.

Through Salem, the current alignment passes near a number of pedestrian areas, schools, parks and office parks. Planning for increased rail service near any of these existing uses requires sensitivity to neighborhood and community needs and safety, an issue which would need to be addressed in the AA/DEIS process. The following photos illustrate existing uses and environment along the alignment. Mitigation measures would need to be developed to address the pedestrian and community impacts that would result from the commuter rail extension.

Figure 12 shows the PNWR track occupying Front Street in central Salem. There is no separation of uses from the tracks along this portion of the alignment. Passing through a mix of commercial and industrial uses, to meet FTA and FRA design requirements, the commuter rail line would likely need to be separated from Front Street vehicular traffic.

Figure 12: Front Street NE & Norway Street NE

15 HDR. Commuter Rail Feasibility between Wilsonville Salem/Keizer. (PowerPoint Presentation, 2009).

26 The alignment passes near Salem-area schools, such as Oregon School for the Deaf shown in Figure 13. The school grounds are fenced and crossing gates are installed for vehicles and pedestrian crossings to ensure safety of school children, drivers, and pedestrians. This site would need additional consideration the AA/DEIS process to ensure the school would not receive disproportionate impacts from the implementation of commuter rail.

Figure 13: North Salem

Figure 14 shows a pedestrian path crossing in a residential neighborhood in northeast Salem. Again, extra effort would need to be made to make the alignment of the commuter rail extension safe for pedestrian crossings.

Figure 14: Maple St. NE at Spruce St. NE

27 Figure 15 shows the large residential tower on Front Street across from Salem’s Riverfront Carousel. This tower is in the heart of downtown Salem close to retail shops and numerous restaurants. The possible central Salem terminus of the alignment would be located nearby.

Figure 15: Salem Riverfront Downtown Area

Operating Plan Assumptions The following section describes the assumptions used to develop the alignment, schedule, train equipment, station locations, terminus, and maintenance facilities analyzed for the possible commuter rail extension in this feasibility assessment. A more detailed summary of the operating assumptions including schedules and capacity analysis is contained in Appendix B.

Alignment Extend commuter rail along the OE line from where WES commuter rail currently operates, south to Salem where it would terminate in central Salem near the riverfront.

Operating Schedule Eight operating scenarios were studied to capture the impacts of different combinations of commuter train frequency, Amtrak train frequency and freight train frequency.

As detailed in Table 1, it was assumed that the existing WES schedule retained 30- minute morning and evening peak period headways under both a 30-minute and 60- minute Salem operating , and all scenarios include hourly mid-day service which is in addition to the current WES schedule. Under the 30-minute headway scenarios all trains would travel between Beaverton and Salem. Under the 60-minute

28 headway model every other train would travel to Salem while the others would maintain the current service to Wilsonville and then turn back to Beaverton.

The scenarios represent a range of operating strategies but are by no means the only options. A more detailed operating analysis, which would be required for a subsequent alternatives analysis and engineering study for the extension, could be undertaken during an AA/DEIS process to examine operating scenarios that could potentially reduce capital and operating costs (compared to 30-minute service along the entire corridor) without substantially reducing ridership.

Table 1: Operating Scenarios

Scenario Schedule Intercity Passenger Freight Traffic Rail (Amtrak) Existing Existing Expected Schedule Wilsonville Wilsonville Schedule2 Future growth 30-Min Peak + to Salem 30- to Salem 60- RT/Day on 6 RT/Day Existing (100% new Midday Min Peak Min Peak UPRR on OE Line Traffic increase) Hourly Service 1 X X X X

2 X X X X

3 X X X X

4 X X X X

5 X X X X

6 X X X X

7 X X X X

8 X X X X

Train Equipment An analysis was undertaken to estimate the number of trainsets needed to operate service under each scenario. Additionally, the analysis examined equipment and locomotive types (DMU, EMU, and locomotive hauled train sets; all defined in the glossary of terms in the Introduction of this report).

The current WES equipment, three DMUs and one unpowered trailer car, is not sufficient to maintain current schedules and add service to Salem. Additional train equipment ( or other powered units as well as unpowered cars) would be needed to extend commuter rail to Keizer or Salem.

29 For this analysis the number of additional train sets is dependent upon desired train frequency. The 30-minute peak headway scenarios require five additional train sets and the 60-minute peak headway scenarios require four additional train sets.

WES currently uses DMUs built by Colorado , which went out of business during train car delivery for WES. At this time, there is no other U.S. DMU car manufacturer to engineer and construct WES-compatible commuter rail cars for a service extension.16 Potential vendors would likely be able to match specifications of the existing WES trains, such as the equipment used for signaling, communicating, propulsion, heating and cooling, and passenger comfort. However, economical procurement of additional train equipment would be dependent on finding other partner commuter operations from around the country that could pool orders to generate the critical mass necessary for competing vendors to bid on engineering and construction of components and achieve cost savings from the ability to competitively bid train equipment.

Types of train sets purchased around the nation vary among transportation agencies. For this preliminary feasibility assessment, efforts were made to find options that would maintain the look and feel of existing WES trains for the extension, which is desirable from operational continuity. During the project design process, equipment options would be analyzed in a detailed equipment procurement plan.

Another train technology option for the commuter rail extension would be locomotive hauled train sets. Procurement of locomotive hauled train sets would be easier than purchasing custom ordered trains matching WES, as there are a number of manufacturers as well as current procurement contracts by other agencies that could be used. Other agencies that might be candidates to share in purchasing similar locomotive hauled trains for commuter service include Sound Transit in Seattle (Sounder); Utah Transit Authority (Front Runner); and Trinity Rail Express in the Fort Worth-Dallas area. However, existing WES infrastructure would require significant modification to facilitate locomotive hauled train sets. These modifications would include a stronger Beaverton Creek bridge at Beaverton Transit Center and significant changes to all existing WES station platforms to accommodate both the vertical and horizontal requirements of locomotive hauled trains. Because of significantly different equipment design and specifications, it would be very costly to make modifications necessary to operate both DMU and locomotive hauled trains along the corridor.

Stations This section provides detailed descriptions of the station locations and terminal station options studied in this assessment, as detailed in Figure 16. It was assumed that all scenarios considered would use the same stations, except in Salem.

16 U.S. Railcar, LLC, a group of private investors affiliated with Value Recovery Group, Inc. (VRG) of Columbus, OH, announced in June 2009 that they have acquired the DMU manufacturing rights and will resume manufacturing this type of commuter rail equipment late in 2009.

30 Multiple options for a southern station terminus were studied: Keizer, north Salem, and central Salem. In reviewing the current and projected commute trip patterns in the Salem/Keizer metropolitan area, based on studies provided by the Mid- Council of Governments (MWVCOG) and the distances between stations, it was determined that if the extension terminated in north Salem there would be no need for a Keizer commuter rail station. If the extension terminated in central Salem, a Keizer station would be necessary to accommodate area demand, while a north Salem station would be eliminated.

31 Figure 16: Station Locations and Passing Siding Options

32 Wilsonville Station Area The commuter rail extension would begin at the existing Wilsonville Station, where WES Commuter Rail service currently terminates. The existing Wilsonville commuter rail station includes a park-and-ride parking lot with approximately 400 spaces and a transfer station. Currently, the park-and-ride lot is approximately 40 to 45 percent full on a daily basis with vehicles parking to use WES as well as the SMART 1X bus service to Salem17. With the extension to Salem and the forecasted ridership (discussed below in the Data Analysis section of this report), this lot would fill to 85 percent of its current capacity by 2020 and be over capacity by 2030. According to TriMet, there is available land to construct approximately 250 additional spaces, which should be sufficient to accommodate the demand in year 2030. Thus, additional property would likely not be needed for vehicle parking.

The expansion of TriMet’s existing maintenance and storage facilities at Wilsonville would be required if this facility is used to provide maintenance for additional trains. During planning for WES, adequate land was acquired to expand the existing maintenance and storage facility, and it was built so that it could be enlarged in the future. A separate train storage facility is also envisioned at the south end of the route in Salem.

Woodburn Station Area Woodburn is the next city along the alignment large enough to support a station. As shown in Figure 17 and Figure 18, the Woodburn Station would be sited along the western UGB expansion amended to the city of Woodburn Comprehensive Plan in 2008.18 This station would be located adjacent to low density residential land use to the east and farm land to the west. The Woodburn Comprehensive Plan suggests the station be located on the north side of OR 219 but the existing industrial use at that location leads this study to locate the station to the south, just outside of the UGB.

The AA/DEIS process would likely need to study consistency with Oregon Planning Goals, especially Goal 2 (Land Use Planning), Goal 3 (Agricultural Land) and Goal 12 (Transportation). Locating a station outside of a current UGB would likely require a UGB expansion request accompanied by a station area plan, which would need to be prepared and submitted through the Marion County and Oregon state planning process (including the Department of Land Conservation and Development [DLCD] and ODOT) before an urban facility such as a commuter rail station and accompanying uses would be allowable in this location. This land use plan would need to provide for compatible land uses around the station that would not interfere with freight train operations or

17 Email exchange with Steve Allen of SMART and Young Park of TriMet, August 11, 2009. 18 City of Woodburn, October 2008 Update. http://www.woodburn‐ or.gov/communitydevelopment/planning/compplanupdate/08CompPlanMap.pdf.

33 public facilities such as a platform, bus transfer area and park-and-ride. The station area would eventually be annexed into the city of Woodburn, and the city would need to develop a local street system to provide for buses, bicycles, pedestrians and auto drop- off and pickup. CARTS could provide bus service between the commuter rail station and the Woodburn Company Stores outlet mall and other destinations in the city; however, this would require rerouting of the existing city bus or adding specific bus service to city destinations.

The Woodburn Transportation System Plan (TSP) from 2005 supports extending commuter rail to Salem with a stop in Woodburn: “Current discussion focus on extending the commuter rail planned between Wilsonville and Beaverton down to Salem. If this occurs, the City should see a passenger stop. This stop could occur west of Butteville Road, north of Oregon 219. If this stop is established, the intracity fixed route transit system should incorporate a stop at the rail station.”19

The station location described in the TSP is not ideal because a rail-served industrial use is located at this site. The station location should not displace existing freight rail uses to accommodate passenger operations and thus should be located south of OR 219. An operating nursery is located on the east side of the PNWR track on the south side of OR 219, which would be impacted by a station located there, while the west side is agricultural use and may experience fewer impacts if a station were to be located there. A station area plan should be undertaken to further address this issue; if the station is sited south of OR 219, the city’s TSP would need to be updated to reflect the slightly different station location to the south of OR 219.

The railroad right of way at the Woodburn station location could accommodate a platform but other facilities, such as a rail siding to separate freight and commuter trains, as well as a bus loading area and a park-and-ride lot would require additional land acquisition.

19 City of Woodburn. 2005. Woodburn Transportation System Plan. http://www.ci.woodburn.or.us/communitydevelopment/planning/compplanupdate/(10‐31)06TSP.pdf. Accessed December 12, 2009.

34 Figure 17: Possible Woodburn Station Area

35 Figure 18: Woodburn Comprehensive Plan in Possible Station Area

Keizer Station Area The Keizer station would be located within the city of Keizer adjacent to the Keizer Station retail development at approximately Lockhaven Drive and Chemawa Road, as shown in Figure 19 and as identified by the Keizer Station Plan.20 The Keizer Station would likely require acquisition of additional right of way for station facilities as it does not appear the developer of Keizer Station has dedicated land as of yet to the city of Keizer for a commuter rail stop.

This station location has been advanced for this study in large part because the city of Keizer instituted zoning to promote transit-oriented development in the vicinity, specifically development that would support commuter rail. The Keizer Station Plan provides for a commuter station within the plan area, at approximately Lockhaven Drive and Chemawa Road. The AA/DEIS process under FTA would require additional local

20 Figure 4, Page 29 of the Keizer Station Plan document, as amended by the City of Keizer in June 2007 (http://www.keizer.org/commdev/KS/Keizer%20Station%20Plan.pdf).

36 analysis of the station area including a “siting assessment” to determine if this is the most appropriate location for a commuter rail station. The Keizer Community Development Code provides development criteria for transit stations, requiring adequate auto parking areas, passenger waiting areas, bicycle parking, and creation of facilities to transfer to bus routes.21 Recent development around this station area was primarily designed for automobile access and is not considered in this report as transit- or pedestrian-oriented in its layout.

If the Keizer Station is selected as the southern terminus of the commuter rail line, commuters to and from Salem would be required to transfer to buses. While this would avoid operating and capital cost issues associated with extending the route into Salem, the transfer would increase travel time for commuter rail passengers and make the mode less attractive. The forecasting work conducted for this feasibility assessment estimated a 20 percent reduction in ridership compared to a central Salem terminus.

Figure 19: Possible Keizer Station Area

21 Keizer Development Code. 2009. http://www.keizer.org/commdev/publications/code‐web.pdf. Accessed June 23, 2009.

37 Salem Station Area Extending commuter rail service on the OE line presents an option to terminate the commuter rail line in north Salem or the Salem central business district (shown in Figure 20). A possible station location adjacent to Salem’s Riverfront Carousel was studied as a possible terminus for commuter rail because it is located within walking distance of Salem’s urban core. It would be adjacent to a currently ongoing redevelopment of the former Boise Cascade site into a mixed use, urban development area, which could further encourage improved transit access.

The Central Salem Station, shown in Figure 20, would provide convenient access to this large regional employment center. Because of constraints presented by the Willamette River waterfront, buildings, parks and adjacent streets, it may not be possible to provide for on-site passenger parking (there are parking facilities within walking distance of the station which passengers could use) or locate a train maintenance or storage facility in this area. A maintenance facility for the south end of the extension would be better located in the north Salem industrial area, where it would be more compatible with adjacent land uses as well as PNWR’s train operations. However, two or three storage tracks at the central Salem station for turning trainsets should not be problematic for surrounding development and would be preferable to multiple daily shuttle movements if trains were parked elsewhere in Salem between runs.

The existing railroad track through central Salem would require upgrading throughout the north part of downtown to provide separation from vehicular, bicycle and pedestrian traffic. Given the experience from the WES project with train noise, it is assumed that if the Central Salem Station becomes the terminus, a quiet zone would be established throughout Salem.

A North Salem Station, shown in Figure 20, was also studied as a terminal station location. This station would be located near the Salem Parkway and Pleasant View Drive NE. The existing and zoned land uses in the vicinity are General Industrial, Industrial Commercial, and Public Amusement (which is a form of recreational and open space) for the Claggett Creek Natural Area. If this station becomes the southern terminus of the commuter rail line, commuter rail riders destined for central Salem would need to make transfers to bus lines, reducing the number of potential riders due to the added travel time of a forced transfer.

In addition, PNWR and BNSF representatives expressed concern over potential non- industrial redevelopment of parcels adjacent to the North Salem Station, as well as the need to preserve existing industrial access if a commuter rail station were built in this location. A North Salem Station could conceivably attract requests for zoning and comprehensive plan changes to the detriment of the existing industrial area.

38 Figure 20. Possible Salem Station Options

39 Figure 21. Photos of the Alignment in Salem (Photos locations noted on Figure 20)

40

Maintenance Facility Options WES has a maintenance and storage facility at Wilsonville Station, as shown in Figure 22. This facility is operating near capacity. However, it was constructed and land was acquired to accommodate expansion should commuter rail be extended to Salem in the future. In addition, a storage facility would be needed near the Salem terminus to house trains between service hours.

Figure 22: Existing WES Maintenance Facility

41 Data Analysis The eight scenarios described in the Operating Plan Assumptions section were analyzed for ridership forecasts, capacity, capital requirements and operational costs. The general assumptions and findings are summarized in this section; a more detailed summary of the ridership forecasting process is included in Appendix C.

Ridership Forecasting There is no existing travel demand model that can provide ridership forecasts for the entire study area22. While the Statewide Model (SWIM) does encompass the Portland- Salem area, it cannot precisely estimate or forecast trips under 50 miles in length. Since all of the trips using existing WES, and the extension, would be less than 50 miles in length, they could not be accurately modeled by SWIM. Thus, a simplified, sketch- planning model was developed for this assessment, which created order-of-magnitude ridership comparisons for different scenarios.

A number of passenger rail-based sketch planning models were examined for input to this model including:

 “Sketch Model to Forecast Heavy-Rail Ridership”, Len Usvyat, Linda Meckel, and Mary Dicarlantonio, presentation to Transportation Research Board, 2009.

 Cambridge Systematics, “Transit Market Research Models: PSRC Transit Database and Sketch Planning Tool – Phase I Project”, prepared for the Puget Sound Regional Council, June 2008 and used for the “BNSF Eastside Corridor Commuter Rail Feasibility Study” for Sound Transit, by Parsons Brinckerhoff, December 2008.

 “Commuter Rail in Oregon,” background paper for the Oregon Transportation Plan Update, publication date unknown.

Using these methods, as well as travel characteristics and variables from Metro, MWVCOG, and SWIM demand forecast models, projected commuter rail ridership in this corridor is primarily driven by the following factors:

 Parking costs at destination: Commuters are more likely to use transit (commuter rail) when parking costs at their intended destination are relatively high.

 Rail line accessibility: Commuters are more likely to use transit when a park-and- ride is located within three to five miles, via driving or public transit, of their

22 Note: It is likely FTA will require a formal travel demand forecasting model be developed for use in the AA/DEIS process. This model would need to be capable of producing FTA‐compliant travel user benefit output.

42 originating point (typically place of residence), and with a rail station within a ¼ to ½ mile walk of their final destination (workplace) (or with connecting transit having a stop within ¼ mile of their final destination).

 Travel time savings: Where travel time savings using commuter rail (as compared to auto in-vehicle travel) are greater than five minutes over the trip duration, commuters are more likely to consider taking transit.

 Reliability of travel time: Knowing that a transit trip can be made within a certain travel time, or the expectation that commuters can arrive at their destination within a few minutes of a desired time, encourages use of transit.

 Employment and population densities: A higher density of employment at the destination end (workplace), and, to a lesser extent, higher population density at the trip origin (home end) encourages use of transit.

 Availability of automobiles: Commuters without an automobile available at their trip origin (home) are more likely to use transit.

The majority of the trips occurring on the existing WES line are home-based work trips, in which the traveler commutes from home to work in the morning peak hours and returns from work to home in the afternoon or evening peak hours.

A sketch-planning ridership model was developed to help provide a range of expected ridership levels with the extension of commuter rail to Salem. The model uses baseline estimates for a 30-minute operating frequency scenario, and then “pivots” from this baseline estimate depending on options being studied. The model was calibrated to existing conditions (2009) using the following factors:

 Journey-to-work data from the 2000 Census, extrapolated to estimate Year 2009 conditions based on 2009 population and employment estimates from IHS/Global Insight, which provides county-level projections for the statewide economic projections as well as for statewide modeling.

 Existing ridership information from TriMet (WES), Cherriots, SMART, and CARTS, along with an estimate of vanpools that could switch to commuter rail if it were available.

 Existing travel times and volumes on I-5 between the south Portland metropolitan area (Wilsonville), western Portland metropolitan area (Beaverton, Tualatin and Hillsboro), and Salem.

Year 2009 data or estimates reflect the economic downturn when Oregon’s unemployment rate exceeded 12 percent. Transit agencies, including TriMet and Cherriots, indicated ridership was down as much as 15 percent in 2009. Therefore, future-year projections were adjusted by a factor of 1.15 to account for the calibration- year economic conditions that are assumed to not exist in 2020 or 2030 planning years.

43 For this assessment, it is assumed that the extension would be open to traffic by 2020. The year 2030 is included for long-range forecasting, also called the “planning horizon year.”

Using this information, transit mode shares between counties along the WES and Wilsonville-to-Salem routes were calibrated. The model assumes that commuter rail mode shares of the south Portland metropolitan area to Salem/north Marion County travel market will be affected by I-5 travel time and thus will increase for future-year scenarios.

Traffic growth trends on I-5 were obtained from the Transportation Planning Analysis Unit (TPAU).23 Figure 23 compares existing and future traffic volumes on I-5 near Woodburn and Keizer. The “approximate peak period capacity” is a relative indicator of the level of congestion on I-5 during a typical weekday – it assumes that each direction of I-5 has a peak hour capacity of 6,000 vehicles (three lanes, each with a capacity of 2,000 vehicles per hour), and each direction is at peak operations approximately two hours of the day; thus, the peak period capacity is roughly 12,000 vehicles per peak period, which is equivalent to approximately 15,000 persons.

Currently, the segment of I-5 between the south end of Portland and Salem is operating near speed limit conditions (60-65 mph) during the morning and afternoon peak travel periods, although segments of the corridor north of Wilsonville and into the Tualatin area do currently operate at or over capacity during peak periods. By 2020, this segment of I-5 will not be able to adequately accommodate demands during portions of peak weekday periods, and by 2030, it will likely be operating at or over capacity during much of the a.m. and p.m. peak periods.

Under these operating conditions, travel times between south Portland and Salem will increase between 2009 and 2030 by up to 15 minutes per trip, resulting in a trip which takes 45 to 50 minutes today increasing to well over an hour by 2030. Travel time along the corridor will become less predictable as road conditions change and the potential for accidents or incidents increase due to “saturated” traffic conditions. These conditions in turn are expected to make commuter rail more attractive to commuters.

23 Email exchange between Sine Adams of PB and Christina McDaniel‐Wilson, Transportation Analyst with TPAU, July 2009.

44 Figure 23: Peak Period Volumes on I-5 vs. Capacity

I-5 Peak Period/Peak Direction Volumes Approximate “Peak Period 18000 Capacity”

16000

14000

12000

10000 Near Woodburn

8000 Near Keizer

6000

4000

2000

0 Existing (2009) 2020 2030

Based on the projections shown in Figure 23, commuter rail/transit mode shares for trips between Washington and Marion Counties (reflecting trips on WES and the extension to Salem) would increase from 1-3 percent of all commute trips in 2009 to 3-5 percent of all commute trips by year 2030, which is almost doubling the transit mode share.

For the 2020 and 2030 future year scenarios, the following assumptions were made and used to adjust the baseline estimates upward or downward accordingly:

 Where commuter rail service terminates in Keizer or North Salem, commuter rail ridership would decrease by approximately 20 percent due to the additional transfer needed to get to central Salem, the primary destination of most Salem- bound workers. Many of these commuters would not be willing to transfer to a local bus in Keizer or northern Salem to reach central Salem.

 Existing WES service would remain at 30-minute headways.

 The model was calibrated to estimate Wilsonville-to-Salem ridership for 30- minute headways. Where service was reduced to 60-minute headways, ridership is estimated to decrease by 10 percent.

 Adding midday service would increase ridership by 15 to 20 percent.

45 The daily ridership projections under these various scenarios are shown in Table 2.

Table 2: Daily Ridership Projections by Scenario

Wilsonville– Projected Ridership to­Salem Southern Midday Scenario Headway Terminus Service 2020 2030

60 minutes Existing Central Salem No Existing = 600 riders24 (buses)

1 30 minutes Central Salem No 2,920‐3,570 3,240‐3,960

North Salem or 2 30 minutes No 2,190‐2,670 2,430‐2,970 Keizer

3 60 Minutes Central Salem No 2,630‐3,210 2,910‐3,560

4 30 Minutes Central Salem Yes 3,440‐4,210 3,820‐4,670

5 60 Minutes Central Salem Yes 3,100‐3,790 3,440‐4,200

North Salem or 6 60 Minutes Yes 2,320‐2,840 2,580‐3,150 Keizer

Previously-completed commuter rail studies for the Wilsonville-to-Salem corridor were reviewed prior to the initiation of this preliminary feasibility assessment. Each of the studies provides ridership forecasts for the extension of commuter rail from Wilsonville to Salem. These forecasts were used to prepare and for comparison with the ridership forecasts assembled for this preliminary feasibility assessment:

 Metro High Capacity Transit Study, Draft Detailed Corridor Evaluation (2009): Nelson-Nygaard, commissioned by Metro, used peer evaluation of comparable commuter rail facilities to arrive at ridership estimates on a per-capita basis. In general, the estimates show that under the existing conditions, 2,800 people per day would use commuter rail for their trip between Beaverton and Salem. Projecting this estimate to 2020, approximately 3,900 people per day would use the commuter rail alignment, equating to a 40 percent increase in ridership by 2020.

24 Combination of ridership on the SMART/Cherriots 1X route, as reported by SMART (approximately 400‐ 500 riders per day), plus an estimate of vanpools operating in the corridor. If adjusted for economic conditions, this is estimated to be approximately 700 bus/vanpool riders per day.

46  Salem-Keizer Area Transportation Study (SKATS) Commuter Rail Sketch-Level Analysis White Paper (2002): This paper estimates that five to 10 percent of current travelers between Salem and Portland during peak periods within the corridor would ride the commuter rail extension. SKATS’ study did not include actual ridership forecasts.  TriMet WES Salem Extension: Market Analysis and Ridership Assessment (URS, 2009): This memorandum summarizes initial research on trips between Salem and the Portland area to present an assessment of the potential ridership market that could use a peak period commuter service. It estimates that 2,100 to 3,000 people would ride commuter extension rail to the central Salem terminus in 2015. This analysis used 2000 Census journey-to-work data to assist with commute trip and mode share forecasting.  HDR Commuter Rail Feasibility between Wilsonville Salem/Keizer (PowerPoint Presentation 2009): This study, prepared with TriMet, estimates that by 2020, 2,500 to 3,800 new riders (in addition to current WES riders) would ride the commuter rail extension to the north Salem terminus. TriMet identified the north Salem station as the most viable terminus to avoid issues similar to those found with WES along Lombard Street in Beaverton. However, it was noted that as momentum for the project builds, extending the rail to downtown Salem would likely become a desired option by the regional community. Table 3 compares 2020 ridership projections from the various sources, including the model developed for this feasibility study. Please note that each estimate used different assumptions, the results are do not provide not an accurate side by side comparison but rather offer an at-a-glance comparison of the range of estimates.

Table 3: Ridership Projections Comparison

Source Current Metro26 TriMet/URS27 TriMet/HDR28 Average feasibility assessment25

2020 2,920 – 3,570 3,900 3,000 ‐ 3,60029 2,500 – 3,800 Approx. 3,400 Ridership

25 Assumes service terminates in central Salem. 26 Estimates based Between Beaverton and Salem. 27 Estimate extrapolated to 2020. Assumes service terminates in central Salem. 28 Assumed service terminates in north Salem. 29 Estimate extrapolated to 2020 based on URS 2015 projections.

47 Table 4 compares the projected Wilsonville-to-Salem ridership levels with current ridership data for other similar commuter rail lines.

Table 4: Projected Wilsonville-to-Salem Extension Ridership vs. Current Ridership Data30

Year 2020 Current

Altamont Front Commuter Wilsonville­ WES Runner Rail Runner Express to­Salem Beaverton to Ogden­Salt­ Albuquerque, Stockton to San Line Extension Wilsonville, OR San Diego, CA Lake, UT NM Jose, CA

Ridership 3,400 (avg.) 1,200 5,100 7,900 2,700 3,700

Corridor 620,000 440,000 900,000 659,000 396,000 997,000 Population

Riders per Day per 340 150 567 1,199 682 371 100,000 Population

Ridership Conclusions The ridership forecast model developed for this feasibility study yields results that fall within the range of projections developed for prior studies. While there are obvious differences due to the modeling details and assumptions, the model outputs were compared with general expectations for non-auto travel, which produced ridership results within a reasonable range.

The future-year ridership projections indicate that the commuter line would attract a range of different trips based on the type of service, station locations, and connections provided under the various scenarios. When compared with other commuter rail lines around the country, ridership projected for the Wilsonville-to-Salem extension is within the range of other operations. However, caution is expressed here as Table 4 compares 2020 Wilsonville-to-Salem ridership with existing ridership from other lines; it is likely that the 2020 projections (if they were available) for the other lines would be substantially higher, resulting in the Wilsonville-to-Salem data appearing at the low end of ridership levels.

30 Sources: Metro/Nelson‐Nygaard for Regional High Capacity Transit Study; American Association (APTA), third quarter 2008 statistics.

48

Capacity Analysis Capacity for the commuter rail extension is crucial since both freight and commuter rail service would operate on the same track, both with high levels of demand. As mentioned earlier, during the discussions with the railroads, it became apparent that if commuter rail service were added along the corridor, additional passing sidings and/or double tracking would be needed at multiple locations along the alignment, allowing freight and commuter trains to pass each other as necessary. While commuter rail must provide timely and quick service to its riders, freight rail service requires flexibility and consistency to provide service to its customers.

Based on discussions with PNWR, all scenarios would require a continuous second track the length of the existing WES corridor between Tualatin and Wilsonville. This would mitigate current freight operational issues that PNWR is experiencing and would help overcome expected capacity constraints if intercity passenger service (Amtrak) were shifted to the OE line.

To maintain access for freight rail service to PNWR’s industrial customers with the Wilsonville-to-Salem commuter rail extension, many locations near industrial hubs would require passing sidings or auxiliary drill tracks. A minimum of four passing sidings would be needed under the 30-minute headway scenarios, plus the second track between Wilsonville and Tualatin. However, with the 60-minute service headway scenario from Wilsonville to Salem, only two locations would require passing sidings, plus the second track between Wilsonville and Tualatin.

Policy Assessment Because the commuter rail extension to Salem would be a major transportation project which has the potential to result in changes to land use and transportation policies along the corridor, an assessment of the potential policy implications of the proposed commuter rail extension is provided in this section. This analysis will need to be addressed in more detail during the AA/DEIS process required by FTA.

Since Wilsonville to Salem, is outside of TriMet’s service area, a new regional entity will likely need to be established to oversee the project development, administer funding, and eventually own and operate the system. Marion County, Woodburn, Keizer, and Salem, at a minimum, would need to take policy action to establish a regional transit entity. A more detailed discussion is included in the Governance section in this study.

Guidance in the Oregon Statewide Planning Goals applies to the proposed commuter rail corridor as well as stations along the corridor, including:

 Goal 2 Land Use Planning: “Major Revisions: major revisions include land use changes that have widespread and significant impact beyond the immediate area, such as quantitative changes producing large volumes of traffic; a qualitative change in the character of the land use itself, such as conversion of residential to industrial use; or a spatial change that affects large areas or many

49  Goal 3 Agricultural Lands: “Agricultural lands shall be preserved and maintained for farm use, consistent with existing and future needs for agricultural products, forest and open space and with the state's agricultural land use policy expressed in ORS 215.243 and 215.700”. This may apply to the rural, agricultural land in the possible Woodburn station area.

 Goal 12 Transportation: “The number and location of major transportation facilities should conform to applicable state or local land use plans and policies designed to direct urban expansion to areas identified as necessary and suitable for urban development. The planning and development of transportation facilities in rural areas should discourage urban growth while providing transportation service necessary to sustain rural and recreational uses in those areas so designated in the comprehensive plan.” This would likely apply to the possible Woodburn station area.

 Goal 12 Transportation: “Lands adjacent to major mass transit stations …and other major air, land and water terminals should be managed and controlled so as to be consistent with and supportive of the land use and development patterns identified in the comprehensive plan of the jurisdiction within which the facilities are located.” This would apply to the industrial areas and freight rail operations in the Woodburn and north Salem station areas.

As discussed earlier, due to existing industrial uses at the city-suggested Woodburn station location, this report studied the possible station location just outside the established UGB. As such, it will require compliance with Oregon’s land use and transportation planning laws and rules, specifically, development of the station will require approval of a UGB expansion. The process would need to study the potential land use, agricultural, and transportation impacts of the station and parking facilities.

Conceptual Capital Cost Estimates The conceptual capital cost estimate for the commuter rail extension to Salem was developed using unit cost information from the ”Portland to Eugene Intercity Passenger Rail Feasibility Study,” costs from the WES project acquired from TriMet, the Lessons Learned from WES report (Appendix A), and relevant data from other recent commuter rail projects. Mitigation measures (passing sidings double tracking, station design, crossovers, etc.) were assumed over the length of the corridor and varied under each scenario. Although no preliminary engineering analysis was conducted, a set of

50 engineering assumptions was made using information gathered from track charts, aerial mapping and numerous site visits. In addition, the Portland to Eugene Intercity Passenger Rail Assessment provides cost data related to bridge inspection and replacement. The cost estimation memorandum in Appendix D summarizes the conceptual capital cost assumptions and details.

Major cost elements for the extension include:

 Train equipment (number of new trainsets needed)  Stations at Woodburn, Keizer and Salem  Maintenance and support facilities  Signal/communication and crossing warning upgrades  Track upgrade and/or replacement  Passing sidings, and double track (Tualatin to Wilsonville)

Other key points used in calculating estimated capital costs include the following:

 Based on the railroad outreach, PNWR would likely require mitigation (a second track) for the existing WES line between Wilsonville and Tualatin.

 BNSF would consider selling its right of way outright, but this cost is not included in the estimates.

 There is a cost “premium” for administration of an FTA grant, multiple agency agreements and funding, and construction of improvements under rail traffic, which are factored into the contingency costs.

 It is assumed that the existing Willamette River Bridge on the OE line will remain in its current condition and single-track configuration, and will not be upgraded to allow for faster, passenger train operation.

 WES costs were approximately $11 million per mile (all costs included). The cost estimates for the commuter rail extension range from $11.3 million to $13.3 million per mile, of which approximately $1.0 million to $1.5 million per mile is due to mitigation measures along the existing WES corridor.

 A “Quiet Zone” is assumed to be included as part of the route through Salem.

 The conceptual capital cost estimate is provided in Table 5 below.

51

Table 5: Conceptual Capital Cost Estimate COST ELEMENT (RAIL ASSUMPTIONS) UNIT COSTS 60­MINUTE 30­MINUTE (Per mile or FREQUENCY FREQUENCY per each) SCENARIO SCENARIO Track‐ New with Subgrade; including: welded rail, $2,500,000 $15,000,000 $25,000,000 concrete ties, and new turnouts 3‐5 miles of sidings along the 30 mile corridor, plus another 3‐5 miles for a second or double‐track between Tualatin and Wilsonville

Track‐Replace/Upgrade to as much as 79 MPH (FRA $1,000,000 $30,000,000 $30,000,000 Class 4)

Track – Replace ties for track upgrades $500,000 $15,000,000 $15,000,000

Double Crossovers (1 location in the double‐tracked $2,000,000 $2,000,000 $2,000,000 section)

Stations (Woodburn, Keizer, Salem) $5,000,000 $15,000,000 $15,000,000

Signals & Communications (does not include $1,000,000 $25,000,000 $25,000,000 crossings)

Maintenance, storage and support facilities $10,000,000 $8,000,000 $10,000,000

Crossings – All public crossings (including crossing $400,000 $12,000,000 $12,000,000 signal upgrade)

Bridges‐Replace timber with concrete $4,800 $24,000,000 $24,000,000

Equipment $5,000,000 ‐ $30,000,000 $50,000,000 $6,000,000

SUBTOTAL $176,000,000 $208,000,000 Preliminary Engineering and Permitting Services $ 22,880,000 $ 27,040,000 (13%) Construction Enginee r ing (8%) $14,080,000 $16,640,000

Contingency (50%)31 $88,000,000 $104,000,000

FTA/Multi‐Agency Administration (15%) $26,400,000 $31,200,000

TOTAL $327,360,000 $386,880,000

31 Accounts for design components not readily identified at this level of detail, as well as for estimated right of way acquisition. Also includes costs for establishing a Quiet Zone through Salem.

52 Conceptual Operating Cost Estimates Conceptual operating costs were developed for the Wilsonville-to-Salem commuter rail extension using information from TriMet, the Lessons Learned from WES analysis (Appendix A), and other recent similar commuter rail projects. The approximate $5.5 – 6.9 million per year operating cost estimate was established using the assumed costs shown in Table 6. Appendix D provides a summary and detail of capital and operating costs.

Table 6: Conceptual Annual Operating Cost Estimates

Operational Element Estimate for Wilsonville­to­Salem Commuter Rail Extension

Transit Agency Staff $400,000 – 500,000

Railroad Maintenance Work and Support Staff $1,750,000 – 2,250,000

Commuter Train Operations $1,400,000 – 1,700,000

Performance Incentive (for on‐time performance by PNWR crews) Not included in this assumpti o n.

Insurance $2,000,000 – 2,500,00032

TOTAL $5,550,000 – 6,950,000

Results of the Feasibility Assessment The analysis completed for this assessment can be used to evaluate the feasibility of extending the commuter rail service from Wilsonville to Salem. This assessment considers the costs and benefits of the extension; environmental, right of way and land use considerations; the ability to accommodate freight and passenger rail traffic; possible sources of financing; and governance issues. These findings are provided for ODOT and other stakeholders to determine whether additional analyses and/or advancement of the Wilsonville-to-Salem extension concept are warranted.

Cost and User Benefit Analysis Reduction in passenger vehicle traffic on adjacent and parallel highways (i.e., I-5 and OR 99E and 99W), resulting from use of commuter rail, would improve traffic flow on those systems for other traffic including truck traffic. With the extension of commuter

32 TriMet assumes the insurance to be a less, a marginal increase over current rate: Tuck Wilson, Neil MacFarlane, Jeff Lowe, Ken Kirse, Tamara Lesh, and Thomas Heilig. Personal Interview. April 28, 2009.

53 rail to Salem, there would be a slight reduction in congestion, up to three percent on I-5 during the morning and evening peak commuting periods between 2015 and 2030.

While not included in the travel demand forecast model developed for this assessment, both OR 99E and 99W provide similar access to regional destinations as I-5 and are experiencing congestion during peak travel times as well. Vehicle trips on OR 99E and 99W may shift to I-5 to take advantage of any improved traffic flow that could emerge due to a diversion of I-5 trips to commuter rail. Therefore, much of the reduction in congestion resulting from commuter rail could be realized on OR 99E and 99W rather than on I-5. Although the study area would need to be more completely evaluated in order to quantify the overall improvements in travel time and reductions in delay resulting from extension of commuter rail, it is anticipated that the overall travel improvements in the Wilsonville-Salem corridor (i.e., I-5, OR 99E and 99W) would be very limited.

As discussed, many of the commuter rail stations would provide opportunities to connect to other transit systems throughout the region, providing greater opportunities for the transit-dependent population in the region. The mobility of transit-dependent individuals is significant because it is a primary variable used by the FTA in evaluating proposed transit projects for funding under its New Starts program.

Financing Plan Interviews with the railroads, TriMet, and Washington County identified 1) the need for creative funding solutions to implement the expansion of commuter rail service to Salem, and 2) the need for a political champion(s) to communicate the funding request message locally and to FTA for additional analysis.

The existing WES line was funded from four primary sources: State of Oregon (approximately 25%); Washington County and local cities (approximately 20%); TriMet (approximately 15%); and, FTA (approximately 40%).33

If the Wilsonville-to-Salem extension were able to reproduce the same financing formula as WES for its $357 million capital construction cost (average of the two cost options shown in Table 5) the following allocation of funding would result (in 2009 dollars):

 FTA: $143 million34  Regional/transit agency: $54 million  State of Oregon: $89 million  Local (Marion County and cities of Salem, Keizer and Woodburn): $71 million.

33 Based on information contained in the Washington County Commuter Rail Fact Sheet, TriMet,October 2008. 34 Typical FTA funding ranges from 40‐60% of project costs.

54 Currently, there are no grants for public transit investments available from the FRA. The most likely federal funding source would be FTA. In order to qualify for FTA funding, lead agencies need to show that financial backing is in place and that the proposed investment demonstrates a balance between forecasted ridership and total construction and operational costs. Requests for FTA funding compete on a national scale using the FTA’s New Starts criteria.

FTA New Starts funds are allocated competitively based on several elements that are individually rated. FTA then assigns an overall rating to each project proposal and evaluates awards of funding based on side-by-side comparison of all projects nationwide, and the expected funding available each fiscal year as apportioned by Congress. On January 13, 2010, a New Starts policy shift was announced that will impact how New Starts projects are scored.35 The general process described above will most likely remain, however proposed future New Starts guidance will “be based on livability issues, such as economic development opportunities and environmental benefits, in addition to cost and time saved, which are currently the primary criteria.”36

Governance Issues Extension of commuter rail to Salem would involve many jurisdictions - including the State of Oregon, Washington, Marion and Clackamas Counties, the cities of Wilsonville, Woodburn, Keizer and Salem, the PNWR and BNSF railroads, and the transit agencies throughout the corridor – TriMet, SMART, CAT, and Cherriots, the Salem-Keizer Area Transit District. These agencies would all need to work together to ensure agreement on the investment.

Interviews with representatives of Washington County, the railroads and TriMet, indicated that, in large part, the successful implementation of WES was the result of visible, consistent, and local champions throughout the planning and construction of the project. Although there are several known individuals who have indicated strong support for extension of commuter rail from Wilsonville to Salem, none of these known proponents reside in the counties or cities that would potentially contribute funds to the project. A similar coalition of proponents from the Wilsonville to Salem area local agencies has not yet emerged.

35 Federal Transit Administration. 2010. New Starts Policy Shift. http://www.fta.dot.gov/planning/planning_environment_11045.html. Accessed January 19, 2010.

36 Federal Transit Administration. 2010. News Releases: Obama Administration Proposes Major Public Transportation Policy Shift to Highlight Livability. http://www.fta.dot.gov/news/news_events_11036.html. January 19, 2010.

55 A number of potential governance structures for the Wilsonville-to-Salem commuter rail extension were examined and are assessed below.

Single Transit Agency with Operating Agreements This option entails an entity, such as TriMet, to develop operating agreements with other agencies outside of their established district. This would mean that TriMet would enter into agreements with Marion County, SMART, and Cherriots to operate the commuter rail line. This structure is similar in nature to current discussions TriMet is having with C-TRAN (the Vancouver, Washington-based transit operator) and other entities in southwest Washington State regarding extension of light rail to Vancouver. The current approach being discussed is for TriMet to operate the light rail service with operating and funding agreements with C-TRAN. This type of structure exists elsewhere around the country for similar circumstances.

In the interview, the TriMet representative expressed interest in transferring ownership and operation of WES to another entity as part of an extension project. If TriMet is not interested in assuming this role, another agency would need to be identified. The transit agencies in the Wilsonville to Salem corridor are much smaller than TriMet and less experienced with projects of this size.

New Regional Transit Agency This option envisions a new regional commuter rail authority, similar to Sound Transit in the Puget Sound region of Washington, which would act as the lead agency to secure funding for and implement commuter rail extension to Salem. As discussed, the Wilsonville-to-Salem extension bisects numerous jurisdictions. Identification of a single regional agency to lead the pursuit would be useful to develop a unified approach among the numerous parties involved. This option was determined to be feasible.

Amtrak Amtrak has the capability and some authority to contract for commuter rail operations but not governance of commuter rail systems. Amtrak’s statutory authority pertains only to “national passenger rail transportation” such as its interstate network, meaning Amtrak would not have the same priority rights for commuter rail as it does for intercity rail operated over a freight rail line. Amtrak currently operates commuter service in a few locations around the country including California (Capitol Corridor, Pacific Surfliner) and Illinois (several lines which terminate in Chicago).37

37 Amtrak operated commuter rail in seven locations as of 1997 but does not provide governmental oversight or governance of any corridor. Since then, regional operators have taken over commuter rail operations from Amtrak and the regional entities are providing governance as well. Source: United State General Accounting Office. 1998. Intercity Passenger Rail, Issues Associated With a Possible Amtrak Liquidation. http://www.gao.gov/archive/1998/rc98060.pdf,

56 Although it is feasible for Amtrak to operate commuter rail, Amtrak taking on governmental oversight (governance) is not considered feasible.

ODOT or Another State Agency There are a number of state department of transportation agencies (DOTs) around the country which fund or operate commuter rail systems. Examples include: Minnesota (Northstar Commuter Rail), Connecticut (the Metro-North corridor out of New York City), and New Mexico (Rail Runner). Also, a number of state DOTs are currently involved in the development of commuter rail systems, including: Connecticut (New Haven- Hartford-Springfield commuter rail) and Florida (Central Florida commuter rail). In all cases, the DOTs needed to substantially increase staff to plan, build, fund and administer commuter rail projects and ongoing operations.

At this time, ODOT does not operate any transit services and does not have appropriate staffing or financial capacity to take on a regional commuter rail operation. ODOT does not have taxing authority in place to fund capital and operating expenses and ODOT is constitutionally prohibited from using gas tax revenue for non-highway purposes.

This governance option was determined to be unworkable under ODOT’s revenue structure. Summary of Key Findings Extending commuter rail from Wilsonville to Salem presents opportunities and challenges that would need to be addressed before project implementation, including the required AA/DEIS process under FTA. This preliminary assessment does not make recommendations about whether this investment is warranted but does outline key considerations and opportunities for further investigation.

Key findings of this high level assessment are as follows:

 Concerns over freight capacity have led PNWR to not support extending commuter rail to Salem at this time.  The capital cost for the Wilsonville-to-Salem commuter rail extension project is estimated at $327 million to $387 million.  Operating costs are estimated to be $5.5 million to $6.9 million plus per year. Insurance costs, one of the primary components of operating costs, may be marginal if TriMet’s existing insurance policy can cover extended service to Salem. However, if a different entity assumes ownership and operating responsibility for commuter rail, that entity may be required to provide a new insurance policy.  The extension project has the potential to attract 3,000 to 4,000 riders per day by 2030, which would slightly reduce congestion on I-5 and OR 99E between the south Portland metropolitan area (Wilsonville) and Salem. Terminating the extension in central Salem would result in up to 20 percent more ridership than a terminus in Keizer or north Salem. However, even with rail investment, highway

57  A key lesson learned from the WES project was the importance of having dedicated political champions who continue to support the project even in times of controversy. For the extension to move forward, an equivalent level of support from Marion County leaders and city leaders in Wilsonville, Donald, Woodburn, Keizer and Salem will be necessary.  Ridership on the extension is forecast to be moderate at best when compared to other commuter rail projects nationwide; this may make it difficult for the project to compete for federal funding, especially through FTA where projects compete on a national basis for finite resources.  Funding for the project faces several obstacles. First, funding sources from local communities would be needed in any matching grant pursuit. The most likely source of Federal funding would be from FTA Section 5309 New Starts which is highly competitive.  Station locations will require careful planning to not negatively impact PNWR’s rail freight business.  A new transit entity would be the most feasible governance structure for the commuter rail extension to include the existing WES line. The new entity would be responsible for capital funding, operations and operating costs, and administration of the system.

58 Appendix A: Lessons Learned from WES

Introduction and Purpose Oregon’s first commuter rail transportation system, known as the Westside Express Service (WES), began operation on February 2, 2009 between the cities of Beaverton and Wilsonville. WES represents the culmination of efforts spanning more than a decade to bring rail mass transit along the 15-mile route parallel to congested Interstate- 5 (I-5) and Oregon State Highway 217, providing connections to other mass transit routes, including the regional system at Beaverton. The institution of this commuter rail service required complete reconstruction of trackage, purchase of three self powered railcars and one unpowered trailer, and construction of five stations, four facilities, and a maintenance facility. The reported cost of these improvements total $166 million to date.1 Major project sponsors included Washington County, Tri-County Metropolitan Transportation District of Oregon (TriMet), the State of Oregon and the cities of Beaverton, Tigard, Tualatin, and Wilsonville. Prior to the inauguration of WES, this route had not seen scheduled passenger rail service since the early 1930’s. WES operates on a rail corridor controlled by the Portland & Western Railroad (PNWR), and has a 50 year agreement to conduct train operations between Beaverton and Wilsonville. Ownership of the right of way is split with TriMet owning the portion from Beaverton to Tigard, which it purchased from the Union Pacific Railroad; from Tigard to Wilsonville, the right of way is owned by the State and is managed by the Oregon Department of Transportation’s (ODOT) Rail Division. WES operation is unique in several aspects. First, WES uses self propelled (DMU) railcars which are used in very few other US passenger rail operations. Secondly, WES is characterized as suburb-to-suburb system, which is different from other US commuter rail systems. It does not extend to and from the central business district of the metropolitan region it serves, although it does provide connections with the MAX light rail system at its north end in Beaverton which connects to the central business district of Portland. WES is also one of the few US systems that is hosted by a as opposed to one of the seven large national railroads, commonly referred to as Class I railroads. ODOT is studying the feasibility of expanding the WES operation south from Wilsonville to Salem, a distance of approximately 30 miles. To assist with analyzing the feasibility of this extension, a series of interviews with various stakeholders involved in the recent WES project were conducted, including PNWR, BNSF Railway Company (BNSF), TriMet, and Washington County. These interviews provided valuable “lessons learned” information on aspects of the WES project that can be used for the Wilsonville to Salem study as well as other future commuter rail studies in Oregon. This report reviews lessons learned on the following topics:  Capacity for Freight and Passenger Operations;  Scalability;

1 “Washington County Commuter Rail Project History,” TriMet website, http://www.trimet.org/about/history/wes.htm. Additional costs for quiet zones and new equipment are not included.

59  Equipment Selection;  Cost Efficiency and Cost Control;  Signal and Communications Systems;  Liability and Indemnity Challenges;  Environmental and Land Use Issues; and,  Safety.

Capacity for Freight and Passenger Operations Ensuring that a given rail route can accommodate capacity for current and future train operations, both commuter and freight, is a significant challenge that is deserving of a robust and disciplined analysis. The speed and frequency of passenger operations is often in conflict with traditional freight movements. A proper capacity assessment and investment program will address both startup conditions for the new services and potential future demands such as increased freight or passenger demand, challenges in equipment of advances in signaling and train control technologies. Discussions between WES stakeholders about capacity design began back in the late 1990’s. At that time, PNWR representatives felt that significant portions of the route would need to be double-tracked to ensure the ability to operate passenger and freight trains simultaneously. Early conceptual designs reflected this view and included substantial amounts of double track. As the project progressed, changes in Federal Transit Administration (FTA) funding and evaluation processes negatively impacted the cost-benefit calculations for WES, resulting in a reduction in the amount of federal funding available for the project. Project planners looked to reduce costs by scrutinizing the amount of double track required. Unlike larger railroads, PNWR had no internal capability to model rail capacity impacts that a proposed commuter rail operation might produce, so they relied on the information produced by TriMet and Washington County consultants. Recognizing the importance of robust and expanded freight service, TriMet conducted extensive modeling of freight and commuter service during the design phase and updated it during construction. However, in the later stages of project construction, changes to track and signal system design created further impacts that were never included in the modeling. This series of events led to limited capacity and inadequate time to solve signal performance and reliability issues in the field prior to the initiation of regular commuter service. In the interview, PNWR indicated that the completed corridor between Beaverton and Wilsonville does not have adequate capacity to accommodate current freight train operations. As a result, they have had to refrain from operating freight trains through the commuter rail corridor during hours that commuter trains operate. Although the agreement between TriMet and PNWR recognize that priority be given to the commuter trains, it was not intended to preclude or limit the freight operation altogether during commute hours. This assertion has been reinforced by interviews with Knife River Corporation (KRC), a major shipper of aggregates by rail through the WES corridor. KRC has stated that they have had to shift some volumes back to truck due to the impact of the WES operation on the transit times for their shipments.

60 In addition to the capacity shortage, PNWR regrets the elimination of industrial sidings and spurs that reduced access to potential freight customers and has made it clear that they will not agree to removal of existing industrial sidings or spurs in the future. PNWR representatives stated that with current reduced traffic levels resulting from the economic downturn, impacts to their freight operations are manageable, however it has increased their operating costs. Once the freight levels return to normal, PNWR fears that the bottleneck presented by the commuter operation will have a significant negative impact on their ability to handle freight traffic and that no capacity exists for any increases in freight traffic beyond recent historical levels. Given the impact to current operations and concern for the future, PNWR does not support expansion of WES on their system. Recognizing the capacity limitations in the corridor, TriMet assumed adding capacity by double tracking the length of the corridor from Beaverton to Salem in a 2009 feasibility study. Lessons Learned about Capacity for Freight and Passenger Operations Historically, many freight railroads have been very reticent about allowing commuter trains to use their freight network due to concern about loss of system capacity (as well as liability issues, which will be examined later in this paper). Most recent new US commuter rail operations have been constructed on lines with very little freight traffic or after significant capacity expansion as dictated by exhaustive modeling efforts. However, it would appear that the modeling undertaken for WES was not sufficient to identify impacts to the then existing freight traffic, and the impacts of changed signal system designs were not modeled at all. A robust operations modeling process will be required by PNWR to address capacity constraints on the current and future operations. Formal corridor modeling should incorporate the chosen train control/signaling technologies to ensure that the capacity impacts of the chosen system(s) are valued and incorporated into the procurement process. Local railroad personnel must be part of the analysis process, and public transportation authorities must better understand unintended impacts on freight rail movements including industrial switching operations. Failure to adequately provide for freight traffic defeats the purpose of commuter rail as the highway capacity created by moving people from automobiles to trains will be consumed with freight loads displaced from trains to trucks.

Scalability One of the major advantages of commuter rail over bus transit is the ability to accommodate growth in ridership and in passenger capacity without substantial additional investment, as well as the reliability of the trip (compared to buses traveling in mixed flow on congested highways). However, in this case several previous design decisions will likely impact the ability of the WES line to handle growth of either commuter rail ridership or freight traffic on the corridor. As discussed earlier, PNWR stated that the current corridor does not have adequate track capacity to accommodate its current freight operation without impacting the ability to operate freight trains at any time of the day. Growth of freight traffic would only exacerbate this situation. Current ridership of WES, approximately 1200 passengers per day, is roughly half of the projected ridership, so consumes less than half of the existing ridership capacity. Due

61 to the peak nature of the commuter travel pattern, as ridership increases in the future, some trains will likely fill up before others, forcing some commuters to modify travel patterns to ensure ability to access the train. Additionally, as has been recently experienced by TriMet, there is no spare passenger equipment for WES. Thus, when one or more units are required to be taken offline for maintenance, it impacts service, requiring shuttle buses to run, often with increased delays. Although ridership capacity is not an issue today, when future growth requires additional seats, the first likely step will be to add a second car to the all train sets. Currently only one of the three train sets operates with a second trailer car. Additionally, TriMet is currently in the process of procuring used self-propelled passenger cars to have spare equipment when needed. WES is currently operated only during the morning and evening peak commute times, but TriMet’s Shared Use Agreement with PNWR allows for the operation of 16 additional trains during midday without any additional infrastructure construction. PNWR representatives indicated that this would further impinge on their freight operations and make maintenance of the track and signals more difficult and costly because these activities are currently conducted during the midday hiatus in commuter train operation. Longer term growth in passenger ridership would either require operation of longer trains with more than two cars by purchase of additional DMUs or require switching to higher capacity locomotive hauled trains. Either option would necessitate significant infrastructure modifications including:  Upgrading of the Beaverton Creek bridge at the Beaverton Transit Center (BTC) to accommodate the weight of a conventional locomotive.  Lengthening of stub tracks at BTC and Wilsonville.  Lengthening of station platforms and adjustment of platform height (currently, only two-car trains can be adequately handled on all WES platforms).  Expanding the maintenance facility to accommodate longer trains and/or different types of train equipment.

Lessons Learned about Scalability Growth of passenger ridership in the short to medium term (5-10 years) will be accommodated by filling the current empty seats and then by expanding one-car trains to two cars requiring purchase of additional DMUs. Even modest growth of rail freight volumes will be constrained unless additional double track and improved signaling is added to the route. Future extension of the commuter rail operations should ensure that track capacity can accommodate growth in freight train and/or commuter train frequency without substantial additional infrastructure. Further, planners should ensure that transition to locomotive hauled trains is not precluded as an option for increasing ridership capacity, as several commuter systems, such as Trinity Rail Express (TRE) in Dallas/Fort Worth in 2001, have found that this transition eventually needs to occur. TRE operates both DMU’s and locomotive hauled trains in the same corridor.

62 Equipment Selection There are three basic types of commuter rail equipment in use today in the US. The dominant equipment remains train sets of passenger coaches hauled by conventional locomotives2. New train sets can be purchased from several different manufacturers and re-manufacturers that refurbish older train sets for commuter service. Coaches can be single level cars seating up to 100 passengers or a bi-level design with over 140 seats in an upper and lower level. (EMU) train sets are used in several commuter operations - including the Chicago area, the Northeast, and the Denver area (planned). These trains are electrically self powered coaches, not pulled or pushed by locomotives. They can also be single or bi-level design, however, unlike the diesel-electric locomotives, EMUs are propelled by electrified overhead catenary or wayside power rails called typically called “third rail.” Diesel Multiple Unit (DMU) cars are a third type of train set used for commuter rail operations, which can also be either single levels or bi-levels. In recent years very few US systems have used DMU cars, although they are heavily used in other parts of . Due to Federal Railroad Administration (FRA) crashworthiness regulations, DMU cars that operate on trackage shared with freight trains must meet very stringent structural standards, which for the most part preclude the use of equipment designs from other countries. Similar to EMUs, these trains use cars powered by on-board diesel engines and are not pulled or pushed by locomotives like traditional trains. Early in the scoping process for the WES operation, Washington County and their consultants concluded that DMU equipment would be most suitable for the Beaverton to Wilsonville route. Based on the projected ridership, the significantly higher weight and excess capacity of conventional locomotive hauled train sets would have been less efficient. Additionally, the shorter length of a two-car DMU train set reduced the required length of platforms and station tracks at Beaverton and Wilsonville, thereby reducing the cost for these items. During the preliminary engineering phase of the project, DMU cars were not being produced by any manufacturer for US operations, although Colorado Railcar (CRC) had a prototype on a demonstration tour around the country. As a result of the potential interest in using DMU cars on the planned WES operation, CRC’s demonstration car was exhibited at various locations on the future WES corridor in November 2002. When TriMet solicited proposals for DMU equipment, CRC provided the only compliant proposal, and in 2005, TriMet entered into an agreement with CRC to construct three power cars and one unpowered trailer car for delivery in late 2008. Originally, TriMet planned to acquire an additional as reserve capacity for maintenance and servicing requirements, but the order for the fourth power car was ultimately eliminated as a cost saving measure. By 2007 it became apparent that the scheduled completion date was likely to be delayed and that CRC was experiencing financial problems. In early 2008, TriMet stepped in to take an active role in managing CRC finances to ensure that suppliers were paid and to avoid further schedule delays so that the cars could be delivered in

2 Such as the system in the Puget Sound region as well as a number of systems on the East Coast.

63 summer of 2008. Shortly thereafter, in December 2008 CRC ceased operations and was liquidated. The original fleet of three DMUs and one unpowered trailer has proven to be inadequate. Recognizing this, TriMet began discussions with potential suppliers of DMU cars. In the meantime, TriMet purchased two used diesel-powered passenger coaches from the Railroad to use as spares in the event of future equipment outages on the system.3 These cars will presumably alleviate the service disruptions that occur when one of the CRC built power cars requires maintenance. The spare cars were built in the 1950’s by a different manufacturer (Budd) with vastly different subsystems than the CRC DMUs requiring additional spare parts and maintenance practices. Lessons Learned about Equipment Selection The selection of CRC as the supplier of DMU cars has been one of the most visible negative public relations issues in the WES project. The basic premise of selecting DMU cars for cost efficiency was a sound decision; however purchasing rail cars of an unproven design from a small manufacturer was risky, and in this particular case it resulted in substantial added costs and schedule delay to the project. Although the cars were ultimately delivered and are now operating, they are orphans without substantive technical support, and sourcing additional cars if ridership increases or the system is expanded could prove to be challenging. The lack of a spare car for maintenance rotation will potentially impact public and political support for the operation. There have been occasional service reductions resulting from failures of one of the three power cars. Presumably the purchase of two spare cars from the will alleviate this problem, although the fact that they are much different than the CRC DMUs will complicate maintenance logistics. TriMet has indicated that any expansion of service will require an evaluation of equipment technology and availability. If DMU cars are to be used for any expansion, careful consideration should be given to partnering with other entities in order to generate a larger order of equipment. This would reduce the unit cost of each DMU and also make it more likely that established rail equipment manufacturers would be interested in supplying the equipment.

Cost Efficiency and Cost Control The WES project has attracted a fair amount of media attention due to budgetary issues. An article in newspaper from December 2008 reported that the $133 million project budget was exceeded by $34 million, representing a 26% cost overrun of the original budget.4 Although much of the attention has been focused on issues with the DMU railcars, other parts of the project experienced budget overruns as well. Both TriMet and Washington County indicated that the original cost budget was overly optimistic, and the project scope was not as well vetted as it could have been. A number of third party issues were not anticipated and resulted in unforeseen increases to project costs. These included access mitigation to a gas station in Beaverton, private

3 “TriMet looks to purchase used trains to help deal with WES disruptions”, Oregonian, October 28, 2009. http://www.oregonlive.com/news/index.ssf/2009/10/after_morning_disruptions_trim.html 4 “Westside Express deal cost TriMet millions”, Oregonian, Dec 14, 2008.

64 grade crossing access issues, parking disputes with nearby stores in Tualatin, and impacts to construction work windows resulting from PNWR freight trains operating through construction areas. Washington County attributed some of the cost overrun to the three years of delay resulting from challenges in meeting the FTA cost-benefit analysis process. As construction material costs between 2005 and 2008 experienced some of the highest inflation in recent history, it seems clear that this inflation significantly impacted final construction costs. In addition, there were flaws in the cost estimation process during the early project development stages of WES. Original estimates of $80 million for the entire system were used in the Environmental Assessment (EA). These early estimates were completed without any participation by TriMet or ODOT, each of whom had substantial requirements that were not incorporated into the early estimates. The actual final cost was $166 million. Lessons Learned about Cost Efficiency and Control Although TriMet has long been a recognized leader in light rail mass transit systems, the WES project represents the first venture within the state into commuter rail where tracks are shared with freight trains. In addition, the non-traditional suburb-to-suburb service provided by WES does not fit into well defined processes used by federal agencies to evaluate these types of projects. It is perhaps inevitable that this project would face some significant challenges as stakeholders became educated in the unique issues surrounding heavy commuter rail, including the requirements of new regulatory authorities, and development of a new interface with a regional freight railroad. Project sponsors have indicated that the development of project scope and design was not as robust as it should have been. Future expansion of the system should take this experience into account and ensure that greater rigor is placed into planning, operations simulation, cost estimation, and engineering processes to avoid the surprises experienced by the WES project. In the future, all operating and responsible parties need to have opportunities to fully vet their concerns and requirements. The issues raised need to be combined with an independent risk assessment of all project elements and the needs of all project participants. A continuous risk assessment for WES was completed by the FTA, however it did not include requirements outside of FTA’s purview such as FRA’s safety regulations.

Signal and Communications Systems During the design process for WES, TriMet and PNWR decided to install a coded cab signal system to enhance the safety of train operations on the shared corridor between Beaverton and Wilsonville. Coded cab signal technology is meant to prevent unsafe operation of trains using older track circuit technology rather than a global positioning system (GPS). While the base technology for this type of system has been around for decades, the hardware and some features of the WES system are unique. PNWR indicated that the vendor’s design engineers did not fully understand the operation of mixed freight and passenger operations and designed the system to be overly restrictive, which has impacted the capacity of the system and impinged on PNWR’s ability to operate freight trains during commuter operation time windows. Signal system design changes occurred near the end of the engineering design process, resulting in limited time for testing and calibration prior to implementation.

65 PNWR and TriMet continued to work through performance and reliability issues of the signal system “in real time” after the regular commuter operations began. The crash of a Commuter train and a Union Pacific freight train near Chatsworth, California in September 2008 focused national attention on the issue of commingled freight and passenger operations. As a result of this incident, the US Congress enacted the Rail Safety Improvement Act (RSIA) of 2008 in October, including as one of its major provisions a mandate to adopt (PTC) technology, which typically uses GPS-based signals, by 2015.5 The current signaling system will have to be modified to comply with the RSIA, however TriMet believes the current system to be 75 percent compliant.. PNWR has expressed their opposition to applying for a waiver to the PTC requirement and plan to follow UPRR and BNSF’s lead in compliance. One possible approach to compliance could be the adoption of an overlay of PTC technology in addition to the current cab signal system. The design of this system could allow some capacity improvement to address the design limitations of the current cab signal system. Lessons Learned about Signal and Communications Systems While the technology and regulation of signaling and communications systems continue to evolve it is essential that adequate time be budgeted to design, install and calibrate these systems. It is possible that before the WES service is expanded, the existing cab signal system will be modified to meet RSIA requirements. However, the fundamental issues of ensuring that system design accommodates freight as well as passenger operation will remain and planners and engineers will need to ensure that the challenges of abbreviated implementation schedules are not repeated.

Liability and Indemnity Challenges Addressing potential liability risks and obtaining the appropriate insurance are challenges that virtually all new commuter rail operations must resolve, and these become even more difficult where the commuter operation shares tracks owned or controlled by a freight railroad. Freight railroads generally view the imposition of passenger trains in a shared corridor as an exponential increase in risk over what they would normally experience when solely operating freight trains because the potential for injury and the resulting insurance and damage claims increase many-fold when passengers are involved in any type of incident. According to a recent Government Accounting Office (GAO) report, freight railroads have taken varied approaches regarding liability and indemnity provisions in access contracts.6 Many recent agreements have required the commuter rail authority to fully indemnify and protect the freight railroad against all costs and claims arising from incidents involving commuter trains, which often includes protection for the freight railroad even against incidents that are caused by their own negligence. The reasoning, sometimes referred to as the “but for” philosophy, is that if an incident occurs involving freight operations but does not involve a passenger train, the cost including claims would be much lower. Indemnification against a freight railroad’s own negligence has

5 Pub. L. No. 110‐432 6 GAO‐09‐282

66 become controversial in other states where public officials have balked at providing this type of indemnity. In the case of WES, PNWR also insisted on full indemnification. The negotiations over liability, indemnification, and insurance proved to be some of the most challenging aspects of establishing the Shared Use Agreement between TriMet and PNWR, which contains all contractual underpinnings for the WES operation over PNWR’s trackage. Ultimately, TriMet provided insurance coverage up to $200 million through a combination of self insurance and externally purchased insurance policies. The cost of these policies is large, and TriMet indicated these policies may be applied to an expanded system without a significant increase in premium cost. For potential claims above $200 million, TriMet and PNWR are relying on the Amtrak Reform and Accountability Act of 1997 (ARAA)7 which caps potential passenger damage claims at $200 million; however, PNWR has concerns that this statutory liability cap has never been tested in a court of law and that it may not hold in a major incident. According to the GAO report cited earlier, this concern is shared by other freight railroads around the country. BNSF Railway Company (BNSF), owner of seven miles of the Oregon Electric rail line from Keizer to Salem over which a WES extension would operate, indicates that any expansion of the WES commuter rail line will require indemnification of BNSF similar to that already in place for PNWR. BNSF describes this requirement as being similar to the “sovereign immunity” enjoyed by many local government entities, and they state that other commuter rail authorities have satisfied this requirement by contractual provisions and insurance or alternatively by the enactment of statutory protections. BNSF also expressed a potential willingness to sell the corridor at market price to the eventual commuter rail system entity, which would likely reduce or eliminate the requirement to indemnify BNSF against potential damages. Lessons Learned about Liability and Indemnity Challenges The insurance and indemnity structure currently in place for the WES operation is consistent with other commuter operations around the country with the liability protection required by PNWR being provided through a combination of self insurance and external insurance policies. The cost of the current insurance is $1.5 million a year, roughly 30% of WES’ annual operating costs8, but the system could potentially be expanded without a substantial increase in the insurance costs. If the corridor from Keizer to Salem is not purchased from BNSF as part of the expansion, BNSF will also require indemnification and insurance coverage by the commuter operator. Alternatively, additional statutory liability protection could be enacted by the Oregon legislature which would reduce the need for external insurance. This may prove to be politically challenging to enact, especially if the freight railroads insist on protection against their own negligence.

Environmental and Land Use Issues The WES project involved reconstruction of rail lines that have been an active transportation corridor for almost 100 years resulting in potential environmental issues

7 Pub. L. No. 105‐134, 111 Stat. 2570 (1997). 8 HDR, “Commuter Rail Feasibility between Wilsonville and Salem/Keizer,” April 2, 2009.

67 much different than those faced when constructing a brand new light rail route or street car line. Due to requirements of the FTA funding arrangements, Washington County and TriMet conducted an EA for the proposed route as part of the preliminary engineering process, which highlighted environmental issues including noise impacts that would require permitting and/or mitigation. Railroads that are engaged in interstate commerce have enjoyed a historical preemption of local environmental and land use regulations. However, as a public agency, TriMet is not exempt from these processes. Therefore, TriMet went through the permitting processes as all public projects do. As a result of TriMet’s approach, very few controversies arose regarding environmental aspects of the project, with the two major exceptions being noise impacts and the station location in Tualatin. The prospect of WES creating increased train noise was a concern raised by local residents in Tualatin prior to the commencement of construction activities. At the request of the City of Tualatin, TriMet revisited the original EA for the project and conducted a recalibration of the noise impacts analysis which reconfirmed that the projected noise impacts were not significant enough to require any mitigation under FTA regulations. These findings did not mollify local residents, and there continue to be a number of noise complaints about the greater frequency and noise volume of train horns being sounded at the closely spaced grade crossings in the Tualatin area and the fact that many of these horns are sounded in early morning weekday hours. It should be noted that train horns are regulated by the FRA, and not FTA, so the only way to avoid the horn noise is to create a “quiet zone,” which requires all grade crossings within the zone to meet significantly upgraded standards to ensure motorist safety. Rules to establish quiet zones were established by the FRA in 2005, after WES project planning and design was underway but not complete. TriMet is actively engaged with the City of Tualatin attempting to implement a quiet zone by upgrading grade crossings, and TriMet has also filed two waiver requests with the FRA to allow the use of quieter horns that are below the required minimum standard decibel level. In November 2009 PNWR filed a response to the FRA opposing the waivers. The question of where to locate the Tualatin station arose as a topic of debate early in the planning process for the WES project, so the City of Tualatin along with Washington County looked at various options, solicited public comment, and held open town hall meetings to ensure that public feedback was considered in the selection process. Out of this process, it was determined that the former railroad station grounds near Tualatin’s central business district would be used for station platforms and a small adjacent park and ride facility. Representatives of the adjacent shopping center voiced objections to the selected location, primarily centered on the possibility of commuter rail passengers using shopping center parking stalls, so modifications were made to the station design to address the concerns and objections subsided. As TriMet began the pre-construction permitting process for the station, concerns by the adjacent shopping center resurfaced and resulted in delays to construction. Ultimately the permits were issued and the station constructed; however, the delays created schedule and construction cost impacts for TriMet.

68 Lessons Learned about Environmental and Land Use Issues Although WES is in an existing rail corridor with existing train noise, the noise complaints that have plagued the WES project were at least partly due to lack of understanding by local leaders and land use planners about the impacts that a heavy rail transportation corridor can create in a residential setting. To ensure that noise impacts are minimized to the greatest extent possible, it is important that the engineering and budget for quiet zones be considered early in the process. Regional land use planners must take into account the noise and vibration that can emanate from a rail corridor when developing long term land use plans, such as comprehensive plans and zoning regulations. New residential development should not be directly located alongside rail corridors. Where development exists or cannot be avoided near these corridors, natural sound barriers such as berms or vegetation should be used to mitigate noise impacts. Understanding that existing heavy rail corridors with little traffic today may experience increases of freight or passenger related traffic in the future is imperative.

Safety TriMet, PNWR, ODOT and the local jurisdictions along the line worked closely together to design safety measures into the WES project design. Nearly all frequently used grade crossings, including privately accessed crossings, were upgraded with flashing lights and warning gates and infrequently used private crossings were protected with conventional locked gates. TriMet and PNWR analyzed the corridor in consultation with local officials to determine locations where pedestrian trespassing was likely to occur, and these locations received right-of-way fencing and additional signage to discourage pedestrian use of the tracks. In preparation for the commencement of WES operations, TriMet conducted a comprehensive public education campaign with communities along the route, including school visits, to raise general awareness about the pending increase in train traffic and train speed. TriMet also actively engaged local law enforcement personnel to raise the profile and awareness regarding the seriousness of trespassing violations and TriMet and PNWR’s commitment to take legal action against trespassers. Lessons Learned about Safety In the first nine months of service WES has experienced no injury accidents. Given the significant increase in train frequency and train speed, the lack of injuries is testimony to the effectiveness of the safety plan enacted by the project stakeholders. Any expansion of the commuter operation will need to ensure that the proactive steps taken in the WES project are adapted and implemented to ensure public safety is protected on the commuter rail extension.

Conclusions The WES system has been alternatively praised and vilified during the relatively brief time it has been in operation. It took many years and required significant efforts by the various proponents to bring to fruition and represents a significant departure from previous mass transportation modes used in the Portland Metro area. The service is now providing a safe and quick transportation alternative for the residents and workers

69 in the southwest regions of the Portland metropolitan area. In reviewing the project with stakeholders, several key areas were highlighted and lessons learned during the planning, design, construction and operations of WES should be taken into account if the service is to be extended to Salem.  Capacity for Freight and Passenger Operations: Freight rail capacity must be protected. A more robust modeling effort is warranted in any future commuter rail to ensure that current and future freight traffic is not displaced by the commuter operation. Formal corridor modeling should incorporate the chosen train control/signaling technologies to ensure that the capacity impacts of the chosen system(s) are valued and incorporated into the procurement process. Failure to adequately provide for freight traffic defeats the purpose of commuter rail as the highway capacity created by moving people from automobiles to trains will be consumed with freight loads displaced from trains to trucks.  Scalability: Future expansion of the commuter rail service should ensure that track capacity can accommodate short-term to medium-term growth (5-15 years) in either freight train or commuter train frequency without substantial infrastructure additions. Careful consideration must be given to any infrastructure design decisions that could have impact on future capacity expansion.  Equipment Selection: Purchasing equipment from a small manufacturer proved risky. On-going technical support and future equipment orders should be considered when weighing car manufacturers. Whatever equipment is used in the future, careful consideration should be given to partnering with other entities in order to generate a larger order which would increase the likelihood that established rail equipment manufacturers would be interested in supplying the equipment. Adequate rolling stock should be acquired up front to provide spares for normal maintenance requirements.  Cost Efficiency and Cost Control: The early budget estimates for WES was overly optimistic, and the project scope was not well vetted. A number of third party issues were not anticipated and resulted in unforeseen project cost increases. Future expansion of the commuter rail system should take this experience into account and ensure that greater rigor is placed into planning, cost estimation, and engineering processes to avoid the surprises experienced by the WES Commuter Rail project.  Signal and Communications Systems: The signal system is a key component of system, affecting both the safety and overall capacity of the system. Careful consideration should be given in future expansions of WES to ensure that the technology is robust and that implementation schedules allow for proper fine tuning of the system. Such systems should meet FRA requirements for PTC.  Liability and Indemnity Challenges: The liability and indemnity structure was negotiated between TriMet and PNWR and any expansion will require additional negotiations. The current insurance costs $1.5 million, consuming 30% of the annual WES operating budget. It is unknown whether the current coverage will satisfy PNWR or BNSF on an expansion or if they will require additional coverage or statutory liability protection.

70  Environmental and Land Use Issues: Quiet zones should be considered early in the design process. Where development exists or is planned near existing rail corridors, consideration should be given to requiring natural sound barriers such as berms or vegetation to mitigate current and future impacts.  Safety: Careful attention was paid to safety during the design process resulting in crossing improvements, right-of-way fencing and a comprehensive public awareness campaign. Expansion of the commuter operation will need to ensure that similar proactive steps are taken to ensure protection of public safety.

71 Appendix B: Operating Assumptions

Summary and Methodology For the WSK extension, eight operating scenarios were developed and analyzed for ridership forecasts, capacity, capital requirements and operational costs. Operating assumptions were developed based on existing WES service, different assumptions with regard to track capacity, freight operations, and whether or not Intercity Passenger Rail service would be shifted to the Oregon Electric line.

WES Commuter Rail began service between the Beaverton Transit Center and the Wilsonville Transit Station on February 2, 2009. Three intermediate stations are located along the 14.7 mile alignment: Hall/Nimbus Station; Tigard Transit Center and Tualatin Station. WES operates on the former OE line from 5:20 to 9:20 a.m. and 3:30 to 8:00 p.m. Monday through Friday, with limited service on holidays.1 Service during these peak commuting hours is generally every 30 minutes for a total of 16 round trips each day - eight round trips in the morning and eight round trips in the evening.

Service is provided via three self-propelled DMU railcars and one trailer; each of the three self-propelled rail cars seats 74 passengers, while the one trailer seats 80. Trains are operated by Portland & Western Railroad (PNWR) crews. A maintenance facility was constructed at the Wilsonville Station and is currently near maximum capacity; however land and right of way are available to expand the maintenance facility in the future.

Equipment was not analyzed as part of the operations analysis. It was assumed that different types of trainset equipment would still result in similar operating conditions. This appendix focuses on the process to develop the different operating conditions that were analyzed in the Commuter Rail Feasibility report.

Assumptions The consultant team has worked with agency to draft the following operating plan assumptions:  Station Location: analyzed with new stations at Woodburn, Keizer or north Salem, and central Salem.  Train Equipment: performance equal to TriMet DMUs.  Schedule: two schedules were modeled: 30 minute peak period and 60 minute peak period. Both schedules were modeled with and without assuming that intercity passenger rail service exists on the same corridor and a 100 percent increase in freight traffic resulting in eight scenarios.  Freight Operations: Information about freight train operations, including the number of trains per day (current and future), running speed, switching operations, etc. was

1 From TriMet website, http://trimet.org/schedules/r203.htm. Accessed October 5, 2009.

72 gathered from interviews with PNWR staff, consultant team experience with the PNWR system and customers along the route, and observations during various site visits. Currently, PNWR operates an average of two northbound and two southbound freight trains per day in the study corridor, but at times PNWR operates as many as four freight trains in each direction each day. Freight trains average 25 mph along the corridor. PNWR performs switching operations at a number of locations along the corridor: Hopmere, West Woodburn, and north Salem are the most frequent locations for switching operations. By 2030, the number of freight trains is expected to double. PNWR indicates that they have reduced or eliminated freight service during the morning and afternoon weekday peak periods in the WES corridor, and have curtailed midday service as well, in order to provide capacity for commuter rail service. PNWR is under agreement with TriMet to operate the WES trains and to ensure that they have priority on the PNWR track.  Intercity Passenger Rail: a separate study was prepared that examined the feasibility of shifting Intercity Passenger Rail (Amtrak) service to the Oregon Electric line. By 2030, up to six round-trip intercity passenger trains per day were assumed to be using the OE line (the same line as for this commuter rail study). Intercity passenger stations would be located at central Salem, Wilsonville or Tualatin, and downtown Portland. It was assumed that Amtrak trains would take priority over all other trains, including PNWR freight trains and commuter rail trains. For this analysis, it was assumed that the commuter rail and intercity schedules would accommodate both passenger train operations and that bypass tracks would be built at key locations along the corridor to enable intercity trains to bypass commuter rail and freight trains that may be traveling at the same time, which was assumed to be an infrequent occurrence.

Transit Facilities and Services Many of the communities along the OE line are served by existing transit services with facilities near the alignment.

Together, the transit transfer station and the park-and-ride lot west of the Wilsonville Station are called SMART (South Metro Area Regional Transit) Central at Wilsonville Station (SCWS). SCWS is a transfer station between WES, all SMART routes, Salem- Keizer Transit Agency, Cherriots, and Canby Area Transit (CAT). SCWS includes a 400 space park-and-ride lot with 48 free bicycle lockers.

SMART service includes five bus routes and one shuttle providing radial service to areas around the south metro area from SCWS. The Cherriots, jointly with SMART, provide bus service Route 1X between downtown Salem and SCWS. Frequency of Route 1x is hourly during the morning commute, and every half hour during the evening commute. Route 2X, operated by SMART, provides service from SCWS to Barbur Boulevard Transit Center every half hour during the morning and evening peak commuting times, and hourly during non-peak periods.

73 SMART and CAT provide joint service via Route 3, the Purple Line, to Canby. Service is hourly Monday through Friday with a break in service between 9:00 and 10:30 a.m. and between 1:30 and 4:00 p.m. SMART Route 4 is a Wilsonville circulator providing service nearly every half hour during the day except between 9:30 a.m. and 2:30 p.m. when service is every 40 to 60 minutes. SMART Route 5 provides service to the Commerce Circle, which is also served by TriMet bus 96. Frequency of service is every half hour between 5:30 and 10:00 a.m. as well as between 3:30 and 7:00 p.m. with no mid-day service. SMART Route 6 provides service to Canyon Creek and the SMART office located in downtown Wilsonville east of I-5 and north of Elligsen Road; frequency is every half hour between 6:30 and 10:00 a.m. and 3:30 and 7:30 p.m. with no mid-day service. Further south, the Woodburn Public Works Department’s Transit Division provides hourly bus circulation service throughout Woodburn.

Chemeketa Area Regional Transportation System (CARTS) is a bus transportation partnership between Marion, Polk and Yamhill counties. Route 10 provides service between Woodburn and Salem with departures from Woodburn Transit Station at 6:40 a.m., 9:00 a.m., 3:15 p.m., and 5:10 p.m. At this time, there is no transit route connection to the possible Woodburn commuter rail station location.

In Keizer, Cherriots provides transit service from the Keizer Station development with four bus routes (4, 15, 18, and 19) circulating to the Salem Transit Mall, Chemekta Community College, or throughout Keizer. Route 4, Keizer Station Express, provides service between the Salem Transit Mall and Keizer Station every half hour during the weekdays. Route 15, Keizer Station/CCC, provides service between the Keizer Station and Chemekta Community College every half hour on weekdays. Route 18, Keizer Crosstown Loop, provides service between the Keizer Station and Windsor Island Road every half hour on weekdays. Route 19, River Rd N/Keizer Station, provides service between Keizer Station and Salem Transit Center every half hour on weekdays. At the terminus of the commuter rail extension in Salem, the alignment could be served by all downtown Salem-oriented Cherriots bus routes.

Alignment The proposed alignment would extend commuter rail along the OE line from where WES Commuter Rail currently operates, south to Salem where it would terminate in central Salem near the riverfront (options also explored terminating the commuter rail line at Keizer or in north Salem) as shown on Table B-5.

Operating Speeds Based on current WES travel speeds and times, and assuming that the extension to Salem would provide track improvements to increase speeds, as well as sidings to enable trains to pass by traveling in opposite directions so as to minimize delays, the following were assumed station-to-station travel times and average speeds that were assumed in this analysis.

74 Table B-1: Station-to-Station Average Operating Speeds and Times

Beaverton TC to Hall/Nimbus to Tigard TC to Tualatin to Wilsonville to Woodburn to Hall/Nimbus Tigard TC Tualatin Wilsonville Woodburn Keizer Keizer to Salem Miles between 2.5 miles 2.3 miles 3.5 miles 6.2 miles 13.3 miles 14.1 miles 5.1 miles Avg. speed 37 mph 37 mph 37 mph 37 mph 60 mph 60 mph 37 mph Travel-time 5 min 6 min 6 min 10 min 14 min 14 min 10 min Operating Schedule Eight operating scenarios were studied to capture the impacts of different combinations of commuter train frequency, Amtrak train frequency and freight train frequency. All scenarios used the operating speeds and times assumptions captured in Table B-1.

As detailed in Table B-2, it was assumed that the existing WES Commuter Rail schedule retained 30 minute morning and evening peak period headways under both a 30 minute and 60 minute Salem operating headway, and all scenarios include hourly mid-day service, which is in addition to the current WES schedule. Under the 30 minute headway scenarios all trains would travel between Beaverton and Salem. Under the 60 minute headway model every other train would travel to Salem while the others would maintain the current service to Wilsonville and then turn back to Beaverton.

The scenarios represent a range of operating strategies but are by no means the only options. A more detailed operating analysis, which would be required for a subsequent alternatives analysis and engineering study for the extension, could be undertaken during an AA/DEIS process to examine operating scenarios that could potentially reduce capital and operating costs (compared to 30 minute service along the entire corridor) without substantially reducing ridership.

Although commuter rail trains could operate up to 90 mph under appropriate track improvements and tangent conditions south of Wilsonville to north of Keizer, it is more likely that the presence of other trains, the effects of positive train control, and the frequency of crossings may limit the maximum operating speeds to between 70 and 79 mph, which is what was factored into the operations assumptions. If trains could achieve a 90 mph maximum speed, the travel time between Keizer and Woodburn may be reduced by 2 to 3 minutes.

75 Table B-2: Operating Scenarios

Scenario Schedule Intercity Freight Traffic Passenger Rail (Amtrak) Existing Schedule Expected 30-Min Peak Wilsonville Wilsonville Existing Future + new to Salem to Salem Schedule 6 RT/Day growth Midday 30-Min 60-Min 2RT/Day on OE Existing (100% Hourly Peak Peak on UPRR Line Traffic increase) Service 1 X X X X

2 X X X X

3 X X X X

4 X X X X

5 X X X X

6 X X X X

7 X X X X

8 X X X X

Operating Schedules Two sets of schedules were developed: one with 30 minute headways along the entire corridor between Beaverton and Salem, the other with 60 minute headways between Wilsonville and Salem, retaining the current 30 minute headways between Wilsonville and Beaverton. Schedules were set up so that trains could “turn around” (change direction) at the Beaverton and Salem termini allowing sufficient time for engineer breaks and passenger deboarding and boarding.

Table B-3 shows a possible operating schedule assuming 30 minute headways along the entire Beaverton-to-Salem corridor, while Table B-4 shows 30 minute headways between Beaverton and Wilsonville, and 60 minute headways between Wilsonville and Salem.

Other commuter rail systems were analyzed to determine if it was possible to reduce headways below 30 minutes. In almost all instances, 30 minutes was the lowest headway; thus, the minimum 30 minute headway was maintained. For midday operations, 60 minute headways were assumed.

76 Table B-3: 30 Minute Headway Schedule

OPTION 2: 30-Minute Frequency North of Wilsonville, 60-minute frequency, Wilsonville to Salem (total travel-time 1hr, 12 minutes) SOUTHBOUND Beaverton TC Hall/Nimbus Tigard TC Tualatin Wilsonville Woodburn Keizer Salem

5:56 AM 6:01 AM 6:07 AM 6:13 AM 6:23 AM 6:37 AM 6:51 AM 7:01 AM 6:46 AM 6:51 AM 6:57 AM 7:03 AM 7:13 AM 7:27 AM 7:41 AM 7:51 AM 7:36 AM 7:41 AM 7:47 AM 7:53 AM 8:03 AM 8:17 AM 8:31 AM 8:41 AM 8:26 AM 8:31 AM 8:37 AM 8:43 AM 8:53 AM 9:07 AM 9:21 AM 9:31 AM 9:16 AM 9:21 AM 9:27 AM 9:33 AM 9:43 AM 9:57 AM 10:11 AM 10:21 AM 4:03 PM 4:08 PM 4:14 PM 4:20 PM 4:30 PM 4:44 PM 4:58 PM 5:08 PM 4:53 PM 4:58 PM 5:04 PM 5:10 PM 5:20 PM 5:34 PM 5:48 PM 5:58 PM 5:43 PM 5:48 PM 5:54 PM 6:00 PM 6:10 PM 6:24 PM 6:38 PM 6:48 PM 6:33 PM 6:38 PM 6:44 PM 6:50 PM 7:00 PM 7:14 PM 7:28 PM 7:38 PM 7:23 PM 7:28 PM 7:34 PM 7:40 PM 7:50 PM 8:03 PM 8:18 PM 8:28 PM

NORTHBOUND Salem Keizer Woodburn Wilsonville Tualatin Tigard TC Hall/Nimbus Beaverton TC 5:04 AM 5:14 AM 5:28 AM 5:42 AM 5:52 AM 5:58 AM 6:04 AM 6:09 AM 5:54 AM 6:04 AM 6:18 AM 6:32 AM 6:42 AM 6:48 AM 6:54 AM 7:00 AM 6:44 AM 6:54 AM 7:08 AM 7:22 AM 7:32 AM 7:38 AM 7:44 AM 7:49 AM 7:34 AM 7:44 AM 7:58 AM 8:12 AM 8:22 AM 8:28 AM 8:34 AM 8:39 AM 8:24 AM 8:34 AM 8:48 AM 9:02 AM 9:12 AM 9:18 AM 9:24 AM 9:29 AM 3:11 PM 3:21 PM 3:35 PM 3:49 PM 3:59 PM 4:05 PM 4:11 PM 4:16 PM 4:01 PM 4:11 PM 4:25 PM 4:39 PM 4:49 PM 4:55 PM 5:01 PM 5:06 PM 4:51 PM 5:01 PM 5:15 PM 5:29 PM 5:39 PM 5:45 PM 5:51 PM 5:56 PM 5:41 PM 5:51 PM 6:05 PM 6:19 PM 6:29 PM 6:35 PM 6:41 PM 6:46 PM 6:31 PM 6:41 PM 6:55 PM 7:09 PM 7:19 PM 7:25 PM 7:31 PM 7:36 PM

Table B-4: 30/60 Minute Headway Schedule

Beaverton TC Hall/Nimbus Tigard TC Tualatin Wilsonville Woodburn Keizer Salem 1a 5:56 AM 6:01 AM 6:07 AM 6:13 AM 6:23 AM 6:37 AM 6:51 AM 7:01 AM 2b 6:26 AM 6:31 AM 6:37 AM 6:43 AM 6:53 AM - - - 4b 6:56 AM 7:01 AM 7:07 AM 7:13 AM 7:23 AM 7:37 AM 7:51 AM 8:01 AM 3b 7:26 AM 7:31 AM 7:37 AM 7:43 AM 7:53 AM - - - 2d 7:56 AM 8:01 AM 8:07 AM 8:13 AM 8:23 AM 8:37 AM 8:51 AM 9:01 AM 1c 8:26 AM 8:31 AM 8:37 AM 8:43 AM 8:53 AM - - - 3d 8:56 AM 9:01 AM 9:07 AM 9:13 AM 9:23 AM 9:37 AM 9:51 AM 10:01 AM 4d 4:03 PM 4:08 PM 4:14 PM 4:20 PM 4:30 PM 4:44 PM 4:58 PM 5:08 PM 2f 4:33 PM 4:38 PM 4:44 PM 4:50 PM 5:00 PM - - - 1e 5:03 PM 5:08 PM 5:14 PM 5:20 PM 5:30 PM 5:44 PM 5:58 PM 6:08 PM 3f 5:33 PM 5:38 PM 5:44 PM 5:50 PM 6:00 PM - - - 2h 6:03 PM 6:08 PM 6:14 PM 6:20 PM 6:30 PM 6:44 PM 6:48 PM 7:08 PM 4f 6:33 PM 6:38 PM 6:44 PM 6:50 PM 7:00 PM - - - 3h 7:03 AM 7:08 PM 7:14 PM 7:20 PM 7:30 PM 7:44 PM 7:58 PM 8:08 PM

NORTHBOUND Salem Keizer Woodburn Wilsonville Tualatin Tigard TC Hall/Nimbus Beaverton TC 2a 5:04 AM 5:14 AM 5:28 AM 5:42 AM 5:52 AM 5:58 AM 6:04 AM 6:09 AM 4a - - - 6:12 AM 6:22 AM 6:28 AM 6:34 AM 6:39 AM 3a 6:04 AM 6:14 AM 6:28 AM 6:42 AM 6:52 AM 6:58 AM 7:04 AM 7:09 AM 2c - - - 7:12 AM 7:22 AM 7:28 AM 7:34 AM 7:39 AM 1b 7:09 AM 7:19 AM 7:33 AM 7:47 AM 7:57 AM 8:03 AM 8:09 AM 8:14 AM 3c - - - 8:12 AM 8:22 AM 8:28 AM 8:34 AM 8:39 AM 4c 8:09 AM 8:19 AM 8:33 AM 8:48 AM 8:58 AM 9:04 AM 9:10 AM 9:15 AM 2e 3:11 PM 3:21 PM 3:35 PM 3:49 PM 3:59 PM 4:05 PM 4:11 PM 4:16 PM 1d - - - 4:19 PM 4:29 PM 4:35 PM 4:41 PM 4:46 PM 3e 4:11 PM 4:21 PM 4:35 PM 4:49 PM 4:59 PM 5:05 PM 5:11 PM 5:16 PM 2g - - - 5:19 PM 5:29 PM 5:35 PM 5:41 PM 5:46 PM 4e 5:16 PM 5:25 PM 5:40 PM 5:54 PM 6:04 PM 6:10 PM 6:16 PM 6:21 PM 3g - - - 6:19 PM 6:29 PM 6:35 PM 6:41 PM 6:46 PM 1f 6:16 PM 6:26 PM 6:40 PM 6:54 PM 7:04 PM 7:10 PM 7:16 PM 7:21 PM

77 Train Equipment An analysis was undertaken to estimate the number of trainsets needed to operate service under each scenario. Additionally, the analysis examined equipment and locomotive types (DMU, EMU, and locomotive hauled trainsets; all defined in the glossary of terms in the Introduction of this report).

The current WES equipment, which includes three DMUs and one unpowered trailer car, is not sufficient to maintain current schedules and add service to Salem. Additional train equipment (locomotives or other powered units as well as unpowered cars) would be needed to extend commuter rail to Keizer or Salem.

For this analysis the number of additional trainsets is dependent upon desired train frequency. The 30 minute peak headway scenarios require five additional trainsets and the 60 minute peak headway scenarios require four additional trainsets.

WES currently uses DMUs built by Colorado Railcar, which went out of business after train car delivery for WES. At this time, there is no other U.S. DMU car manufacturer to engineer and construct WES-compatible commuter rail cars for a service extension.2 Potential vendors would likely be able to match specifications of the existing WES trains, such as the equipment used for signaling, communicating, propulsion, heating and cooling, and passenger comfort. However, economical procurement of additional train equipment would be dependent on finding other partner commuter operations from around the country that could pool orders to generate the critical mass necessary for competing vendors to bid on engineering and construction of components and achieve cost savings from the ability to competitively bid train equipment.

Types of trainsets purchased around the nation vary among transportation agencies. For this preliminary feasibility assessment, efforts were made to find options that would maintain the look and feel of existing WES trains for the extension, which is desirable from both operating and funding perspective. During the project design process, equipment options would be analyzed in a detailed equipment procurement plan.

Another train technology option for the commuter rail extension would be locomotive hauled trainsets. Procurement of locomotive hauled trainsets would be easier than purchasing custom ordered trains matching WES, as there are a number of manufacturers as well as current procurement contracts by other agencies that could be used. Other agencies that might be candidates to share in purchasing similar locomotive hauled trains for commuter service include Sound Transit in Seattle (Sounder); Utah Transit Authority (Front Runner); and Trinity Rail Express in the Fort Worth-Dallas area. However, existing WES infrastructure would require significant modification to facilitate locomotive hauled trainsets. These modifications would include a stronger Beaverton Creek bridge at Beaverton Transit Center and significant changes

2 U.S. Railcar, LLC, a group of private investors affiliated with Value Recovery Group, Inc. (VRG) of Columbus, OH, announced in June 2009 that they have acquired the Colorado Railcar DMU manufacturing rights and will resume manufacturing this type of commuter rail equipment late in 2009.

78 to all existing WES station platforms to accommodate both the vertical and horizontal requirements of locomotive hauled trains. Because of significantly different equipment design and specifications, it would be very costly to make modifications necessary to operate both DMU and locomotive hauled trains along the corridor.

Stations Multiple options for a southern station terminus were studied: Keizer, north Salem, and central Salem. In reviewing the current and projected commute trip patterns in the Salem/Keizer metropolitan area, based on studies provided by the Mid-Willamette Valley Council of Governments (MWVCOG) and the distances between stations, it was determined that if the extension terminated in north Salem there would be no need for a Keizer commuter rail station. If the extension terminated in central Salem, a Keizer station would be necessary to accommodate area demand, while a north Salem station would be eliminated.

Operations in Station Areas The commuter rail extension would begin at the existing Wilsonville Station, where WES Commuter Rail service currently terminates. The existing Wilsonville commuter rail station includes a park-and-ride parking lot with approximately 400 spaces and a transfer station. Currently, the park-and-ride lot is approximately 40 to 45 percent full on a daily basis with vehicles parking to use WES as well as the SMART 1X bus service to Salem. With the extension to Salem and the forecasted ridership, this lot would fill to 85 percent of its current capacity by 2020 and be over capacity by 2030. According to TriMet, there is available land to construct approximately 250 additional spaces, which should be sufficient to accommodate the demand in year 2030. Thus, additional property would likely not be needed for vehicle parking.

The expansion of TriMet’s existing maintenance and storage facilities at Wilsonville would be required if this facility is used to provide maintenance for additional trains. As discussed above, during planning for WES, adequate land was acquired to expand the existing maintenance and storage facility, and it was built so that it could be enlarged in the future. A separate train storage facility is also envisioned at the south end of the route in Salem.

The railroad right of way at the Woodburn station location could accommodate a platform but other facilities, such as a rail siding to separate freight and commuter trains, as well as a bus loading area and a park-and-ride lot would require additional land acquisition.

The Keizer station would be located within the city of Keizer adjacent to the Keizer Station retail development at approximately Lockhaven Drive and Chemawa Road. The Keizer Station would likely require acquisition of additional right of way for station facilities as it does not appear the developer of Keizer Station has dedicated land as of yet to the city of Keizer for a commuter rail stop.

If the Keizer Station is selected as the southern terminus of the commuter rail line, commuters to and from Salem would be required to transfer to buses for the final leg of

79 their trip. While this would avoid operating and capital cost issues associated with extending the route into Salem, the transfer would increase travel time for commuter rail passengers and make the mode less attractive. The forecasting work conducted for this feasibility assessment estimated a 20 percent reduction in ridership compared to a central Salem terminus.

Extending commuter rail service on the OE line presents an option to terminate the commuter rail line in north Salem or the Salem central business district. A possible station location adjacent to Salem’s Riverfront Carousel was studied as a possible terminus for commuter rail because it is located within walking distance of Salem’s urban core. It would be adjacent to a currently ongoing redevelopment of the former Boise Cascade site into a mixed use, urban development area, which could further encourage improved transit access.

The Central Salem Station would provide convenient access to this large regional employment center. Because of constraints presented by the Willamette River waterfront, buildings, parks and adjacent streets, it may not be possible to provide for on-site passenger parking (there are parking facilities within walking distance of the station which passengers could use) or locate a train maintenance or storage facility in this area. A maintenance facility for the south end of the extension would be better located in the north Salem industrial area, where it would be more compatible with adjacent land uses as well as PNWR’s train operations. However, two or three storage tracks at the central Salem station for turning trainsets should not be problematic for surrounding development and would be preferable to multiple daily shuttle movements if trains were parked elsewhere in Salem between runs.

The existing railroad track through central Salem would require upgrading throughout the north part of downtown to provide separation from vehicular, bicycle and pedestrian traffic. Given the experience from the WES project with train noise, it is assumed that if the Central Salem Station becomes the terminus, a quiet zone would be established throughout Salem.

A North Salem Station was also studied as a terminal station location. This station would be located near the Salem Parkway and Pleasant View Drive NE. The existing and zoned land uses in the vicinity are General Industrial, Industrial Commercial, and Public Amusement (which is a form of recreational and open space) for the Claggett Creek Natural Area. If this station becomes the southern terminus of the commuter rail line, commuter rail riders destined for central Salem would need to make transfers to bus lines, reducing the number of potential riders due to the added travel time of a forced transfer.

In addition, PNWR and BNSF Railway Company representatives expressed concern over potential non-industrial redevelopment of parcels adjacent to the North Salem Station, as well as the need to preserve existing industrial access if a commuter rail station were built in this location.

80 Track Capacity Assumptions Based on the train schedules and operating scenarios, as well as potential station locations, it was determined that sidings would be needed along the corridor, varying depending on train frequency. Sidings were generally located away from station areas, to reduce the possibility of passengers attempting to get to a far-side train while not watching for an oncoming near-side train, and vice-versa for departing a train. If a north Salem station becomes the southern terminus, a siding may be necessary in the station area to enable freight and intercity trains to pass by trains stopped at this station.

Thus, the sidings generally were set to allow trains to dwell and wait for oncoming trains to pass by. Sidings would be long enough to allow for freight trains as well as intercity passenger trains to use them.

The following were determined to be needed based on the different frequencies:

 For all scenarios, a siding north of Wilsonville along the existing WES line is necessary to restore freight service, especially if midday commuter rail service is initiated.  For 60 minute headways, one and possibly two additional sidings would be necessary, one located between Wilsonville and Woodburn, and possibly a second located between Woodburn and Keizer.  For 30 minute headways, as many as four additional sidings would be needed, including the two under the 60 minute scenarios, as well as two additional sidings: one in north Salem, the other a second siding between Woodburn and Keizer. These are shown in Table B-5.

81 Table B-5: Station Locations and Passing Siding Options

82 Appendix C: Ridership Forecasting

Summary and Methodology The eight scenarios described in the Operating Plan Assumptions section were analyzed for ridership forecasts, capacity, capital requirements and operational costs. The detailed passenger forecasting process is summarized in this appendix; for this assessment, it is assumed that the extension would be in operation by 2020. The year 2030 is included for long-range forecasting, also called the “planning horizon year.”

There is no existing travel demand model that can provide ridership forecasts for the entire study area1. While the Statewide Model (SWIM) does encompass the Portland- Salem area, it cannot precisely estimate or forecast trips under 50 miles in length. Since all of the trips using existing WES, and the extension to Salem (called here WSK for Wilsonville-Salem-Keizer), would be less than 50 miles in length, they could not be accurately modeled by SWIM. Thus, a simplified, sketch-planning model was developed for this assessment, which created order-of-magnitude ridership comparisons for different scenarios.

A number of passenger rail-based sketch planning models were examined for input to this model including:

 “Sketch Model to Forecast Heavy-Rail Ridership”, Len Usvyat, Linda Meckel, and Mary Dicarlantonio, presentation to Transportation Research Board, 2009.  Cambridge Systematics, “Transit Market Research Models: PSRC Transit Database and Sketch Planning Tool – Phase I Project”, prepared for the Puget Sound Regional Council, June 2008 and used for the “BNSF Eastside Corridor Commuter Rail Feasibility Study” for Sound Transit, by Parsons Brinckerhoff, December 2008.  “Commuter Rail in Oregon,” background paper for the Oregon Transportation Plan Update, publication date unknown.  “Determining the Ridership Potential of Commuter Rail Routes”, David H. Kaplan (Kent State University, OH) and Brian P. Holly (Oregon Economic and Community Development Department, Salem, OR). Using these methods, as well as travel characteristics and variables from Metro, the Mid-Willamette Valley Council of Government’s (MWVCOG), and SWIM demand forecast models, projected commuter rail ridership in this corridor is primarily driven by the following factors:

 Parking costs at destination: Commuters are more likely to use transit (commuter rail) when parking costs at their intended destination are relatively high. Parking

1 It is likely FTA will require a formal travel demand forecasting model be developed for use in the AA/DEIS process. This model would need to be capable of producing FTA‐compliant travel user benefit output.

83 costs were assumed in central Salem but nowhere else. Parking costs in central Salem were assumed to be similar to today’s costs (adjusted for inflation).  Rail line accessibility: Commuters are more likely to use transit when a park-and- ride is located within three to five miles, via driving or public transit, of their originating point (typically place of residence), and with a rail station within a ¼ to ½ mile walk of their final destination (workplace) (or with connecting transit having a stop within ¼ mile of their final destination). The forecasting work assumed the existing WES stations, plus stations at Woodburn, Keizer, and two Salem options: North Salem or Central Salem (but not both).  Travel time savings: Where travel time savings using commuter rail (as compared to auto in-vehicle travel) are greater than five minutes over the trip duration, commuters are more likely to consider taking transit. In this analysis, travel times were compared to equivalent trips using I-5.  Reliability of travel time: Knowing that a transit trip can be made within a certain travel time, or the expectation that commuters can arrive at their destination within a few minutes of a desired time, encourages use of transit.  Employment and population densities: A higher density of employment at the destination end (workplace), and, to a lesser extent, higher population density at the trip origin (home end) encourages use of transit.  Availability of automobiles: Commuters without an available automobile at their trip origin (home) are more likely to use transit. For this high-level of modeling, there were no adjustments made to ridership based on automobile availability. For FTA-compliant forecasting, the forecast model will need to incorporate this component. The majority of the trips occurring on the existing WES line are home-based work trips, in which the traveler commutes from home to work in the morning peak hours and returns from work to home in the afternoon or evening peak hours.

Model Development Journey-to-Work (JTW) data from the 2000 Census was used to establish baseline, county-to-county commute patterns. Table C-1 summarizes the JTW results for the study area; only those cells highlighted in yellow are considered trips occurring with the study area that could potentially use the commuter rail extension to Salem.

Table C-1: Year 2000 US Census Journey-to-Work Trips, Commuter Rail Study Area

Trip to Workplace in: Trip from Residence in: Clackamas Marion Multnomah Washington Yamhill Clackamas 81,893 4,346 55,962 19,590 620 Marion 5,546 99,984 3,789 4,598 1,393 Multnomah 26,323 1,515 262,646 32,587 687 Washington 12,789 1,541 52,610 155,583 1,922 Yamhill 1,575 1,561 2,374 6,653 24,593

84 Since the information in Table C-1 is for the year 2000, and existing ridership including that on the WES is from 2009, JTW trip-making needed to be adjusted to estimated 2009 conditions. This was done by using county population and employment estimates from Global Insight data, which is used by various state entities including ODOT and the Office of Economic Analysis, as shown in Table C-2 for 2000 and 2009 conditions as well as future-year estimates.

Table C-2: County Population and Employment Estimates and Projections2

Population Clackamas Marion Multnomah Washington Yamhill 2000 340392 286021 663482 449672 85468 2009 386332 319376 728769 540046 100361 2020 442485 363731 804347 663723 117800 2030 497015 407663 855025 790108 131749

Employment Clackamas Marion Multnomah Washington Yamhill 2000 134081 124588 456561 226205 29149 2009 145431 129851 433592 240176 31412 2020 170550 149344 491891 300324 37970 2030 190016 165466 522980 353826 42121 The rate of increase between 2000 and 2009 was averaged by county for population and employment, and the JTW 2000 data was adjusted upward by this average growth rate, as shown in Table C-3 (only potential commuter rail trip connections are shown).

Table C-3: Estimated 2009 Commute Trips between Study Area Counties

Trip to Workplace in: Trip from Residence in: Clackamas Marion Multnomah Washington Yamhill Clackamas 4,731 21,517 Marion 107,926 3,915 5,008 1,528 Multnomah 1,622 Washington 14,615 1,728 Yamhill 1,730 2,521 7,438 The current (2009) WES ridership level is approximately 1,200 passengers per day for transit passengers traveling between Wilsonville and Beaverton (with connections into downtown Portland via MAX). Additionally, based on interviews with Cherriots, SMART, CARTS, and Metro, there are approximately 600 riders per day traveling between Wilsonville and Salem, 400 on transit and 200 using long-distance vanpool.3

2 http://www.ihsglobalinsight.com/

3 Information supplied from Valley Vanpool and Cherriots courtesy of URS and TriMet.

85 Since traffic-zone level mode split information was not available, commuter rail trip capture rates were estimated for each county-to-county interchange. The rates were adjusted upward or downward slightly until the model estimated existing ridership within 10 percent of the total WES and bus ridership totals (1,200 WES and 600 bus/vanpool riders). The resulting existing mode share estimates are shown in Table C-4.

Table C-4: Existing Mode Shares with Commuter Rail Study Area

Estimated WES/WSK Mode Shares Clackamas Marion Multnomah Washington Yamhill 2.0% 0.7% 0.2% 4.0% 3.0% 0.5% 2.0% 0.7% 3.0% 0.6% 0.5% The base-year model yielded the following WES and bus/vanpool daily trip-making within the study area, as shown in Table C-5 (yellow highlights indicate trips on SMART/CARTS plus vanpools).

Table C-5: Model Estimation of Existing Commuter Rail/Bus/Vanpool Trips in Study Area

Trip to Workplace in: Trip from Residence in: Clackamas Marion Multnomah Washington Yamhill Clackamas 95 151 Marion 162 157 150 8 Multnomah 32 Washington 102 52 Yamhill 15 37 This table gives a rough approximation of what ridership would look like if commuter rail were extended from Wilsonville to Salem today (2009).

When the base year model was calibrated, it was noted that opening-year ridership on the WES line was lower than the 1,700 riders per day expected from prior studies. This was attributed to the severe economic downturn of 2008-2010. It is expected that the forecast years of 2020 and 2030 would reflect more “normal” economic circumstances. Thus, a 15 percent economic adjustment factor will be applied for future-year modeling results, which reflects the approximate transit trip-making decreases experienced by TriMet and other transit agencies in 2009 compared to previous years.

Applying the Model for Future Forecasts The model described is a pivot-point model; in other words, it uses baseline estimates from the model calibration, as well as a 30 minute commuter rail operating frequency scenario, and then “pivots” up or down from this baseline estimate depending on options being studied.

86

Using this information, future-year commuter rail transit mode shares between counties along the WES and WSK routes were calibrated. The model assumes that commuter rail mode shares of the south Portland metropolitan area to Salem/north Marion County travel market will be affected by I-5 travel time and, thus, will increase for future-year scenarios due to highway congestion.

Traffic growth trends on I-5 were obtained from the Transportation Planning Analysis Unit (TPAU).4 Table C-6 compares existing and future traffic volumes on I-5 near Woodburn and Keizer. The “approximate peak period capacity” is a relative indicator of the level of congestion on I-5 during a typical weekday – it assumes that each direction of I-5 has a peak hour capacity of 6,000 vehicles (three lanes, each with a capacity of 2,000 vehicles per hour), and each direction is at peak operations approximately two hours of the day. Thus, the peak period capacity is roughly 12,000 vehicles per peak period, which is equivalent to approximately 15,000 persons.

Currently, the segment of I-5 between the south end of Portland and Salem is operating near speed limit conditions (60-65 mph) during the morning and afternoon peak travel periods, although segments of the corridor north of Wilsonville and into the Tualatin area currently operate at or over capacity during peak periods. By 2020, this segment of I-5 will not be able to adequately accommodate demands during portions of peak weekday periods, and by 2030, it will likely operate at or over capacity during much of the a.m. and p.m. peak periods.

Under these operating conditions, travel times between south Portland and Salem will increase between 2009 and 2030 by up to 15 minutes per trip, resulting in a trip, which takes 45 to 50 minutes today, increasing to well over an hour by 2030. Travel time along the corridor will become less predictable as road conditions change and the potential for accidents or incidents increase due to “saturated” traffic conditions. These conditions in turn are expected to make commuter rail more attractive to corridor commuters.

4 Email exchange between Sine Adams of PB and Christina McDaniel‐Wilson, Transportation Analyst with TPAU, July 2009.

87 Table C-6: Peak Period Volumes on I-5 vs. Capacity

I-5 Peak Period/Peak Direction Volumes Approximate “Peak Period Capacity” 18000

16000

14000

12000

10000 Near Woodburn

8000 Near Keizer

6000

4000

2000

0 Existing (2009) 2020 2030

Based on the projections shown in Table C-6, commuter rail/transit mode shares for trips between Washington and Marion Counties (reflecting trips on WES and the extension to Salem) would increase from a mode share of 1 to 3 percent of all commute trips in 2009, to a transit mode share between 3 to 5 percent of all commute trips by year 2030, almost doubling the transit mode share. Based on analysis conducted by TriMet and Metro, it is estimated that approximately 15 percent of WSK commuter rail riders are those who would board the existing WES line and continue south of Wilsonville, while the remainder would be new riders attracted by the existence of the WSK extension. Future-year commuter rail transit mode shares (baseline forecast) are shown in Table C-7:

Table C-7: Future Year Baseline Commuter Rail Mode Shares

Potential WES/WSK Mode Shares Clackamas Marion Multnomah Washington Yamhill 2.0% 0.7% 0.3% 5.0% 3.0% 0.5% 3.0% 0.7% 5.0% 0.6% 0.5%

The baseline WSK operations alternative assumptions are:

88  30 minute headways along the entire corridor (WES plus WSK), no transfer between trains  New stations at Woodburn, Keizer and Central Salem  No midday service (reflecting existing operations)  Current parking costs (adjusted to keep up with inflation) assumed in central Salem and Portland, no out-of-pocket parking costs elsewhere in the study area Because this model is a relatively high-level pivot point estimating tool, resultant ridership was adjusted upward and downward by 10 percent to establish a range; that range is what is being reported in the report. If and when a Federal Transit Administration compliant alternatives analysis is undertaken, a more detailed travel demand model will be required.

The resultant baseline daily ridership forecasts are:

 Year 2020 (opening year): 2,920 to 3,570 riders per day  Year 2030 (horizon year): 3,240 to 3,960 riders per day.

Impacts of Different Operations Assumptions The following are pivot (adjustment) factors to the model based on varying operating scenarios:  Station Location: where commuter rail service terminates in Keizer or North Salem, commuter rail ridership would decrease by approximately 20 percent due to the additional transfer needed to get to central Salem, the primary destination of most Salem-bound workers. Many of these commuters would not be willing to transfer to a local bus in Keizer or northern Salem to reach central Salem.  Schedule: where service was reduced to 60 minute headways, ridership is estimated to decrease by 10 percent, reflecting that many transit users must make the trip regardless of service level (called captive riders, typically those without an automobile or alternative transportation available to make the trip). Adding midday service would increase ridership by an estimated 15 to 20 percent, which is approximately the amount of ridership experienced by mid-size transit agencies during midday periods.  Travel Time Savings: reflected in the increase in transit mode shares for the baseline alternative. No further adjustment is necessary.  Reliability of Trip and Travel Time: reflected in the increase in transit mode shares. No further adjustment is necessary.  Parking Costs: there was no modeling of varying parking costs.  Population and Employment Density, Auto Ownership: reflected in the increase in transit mode shares. No further adjustment is necessary. The daily ridership projections under these various scenarios are shown in Table C-8.

89 Table C-8: Daily Ridership Projections by Scenario

Projected Ridership WSK Southern Midday Scenario Headway Terminus Service 2020 2030

60 minutes Existing Central Salem No Existing = 600 riders5 (buses)

1 30 minutes Central Salem No 2,920-3,570 3,240-3,960

North Salem or 2 30 minutes No 2,190-2,670 2,430-2,970 Keizer

3 60 Minutes Central Salem No 2,630-3,210 2,910-3,560

4 30 Minutes Central Salem Yes 3,440-4,210 3,820-4,670

5 60 Minutes Central Salem Yes 3,100-3,790 3,440-4,200

North Salem or 6 60 Minutes Yes 2,320-2,840 2,580-3,150 Keizer

Other Studies Previously-completed commuter rail studies for the WSK corridor were reviewed prior to the initiation of this preliminary feasibility assessment. Each of the studies provides ridership forecasts for the extension of commuter rail from Wilsonville to Salem. These forecasts were used to prepare and for comparison with the ridership forecasts assembled for this preliminary feasibility assessment:

 Metro High Capacity Transit Study, Draft Detailed Corridor Evaluation (2009): Nelson-Nygaard, commissioned by Metro, used peer evaluation of comparable commuter rail facilities to arrive at ridership estimates on a per-capita basis. In general, the estimates show that under the existing conditions, 2,800 people per day would use commuter rail for their trip between Beaverton and Salem. Projecting this estimate to 2020, approximately 3,900 people per day would use the commuter rail alignment, equating to a 40 percent increase in ridership by 2020.  Salem-Keizer Area Transportation Study (SKATS) Commuter Rail Sketch-Level Analysis White Paper (2002): This paper estimates that five to 10 percent of current travelers between Salem and Portland during peak periods within the

5 Combination of ridership on the SMART/Cherriots 1X route, as reported by SMART (approximately 400‐500 riders per day), plus an estimate of vanpools operating in the corridor. If adjusted for economic conditions, this is estimated to be approximately 700 bus/vanpool riders per day.

90 corridor would ride the commuter rail extension. SKATS’ study did not include actual ridership forecast numbers.  TriMet WES Salem Extension: Market Analysis and Ridership Assessment (URS, 2009): This memorandum summarizes initial research on trips between Salem and the Portland area to present an assessment of the potential ridership market that could use a peak period commuter service. It estimates that 2,100 to 3,000 people would ride commuter extension rail to the central Salem terminus in 2015. This analysis used 2000 Census journey-to-work data to assist with commute trip and mode share forecasting.  HDR Commuter Rail Feasibility between Wilsonville Salem/Keizer (PowerPoint Presentation 2009): This study, prepared with TriMet, estimates that by 2020, 2,500 to 3,800 new riders (in addition to current WES riders) would ride the commuter rail extension to the north Salem terminus. TriMet identified the north Salem station as the most viable terminus to avoid issues similar to those found with WES along Lombard Street in Beaverton. However, it was noted that as momentum for the project builds, extending the rail to downtown Salem would likely become a desired option by the regional community. Table C-9 compares 2020 ridership projections from the various sources, including the model developed for this feasibility study. Please note that each estimate used different assumptions, the results are do not provide not an accurate side by side comparison but rather offer an at-a-glance comparison of the range of estimates.

The SKATS study was reviewed but did not contain actual ridership forecasts; it did estimate 5 to 10 percent of the commuters in the corridor would use commuter rail if it were available.

Table C-9: Ridership Projections Comparison

Source Current Metro7 TriMet/URS8 TriMet/HDR9 Average feasibility assessment6

2020 2,920 – 3,570 3,900 3,000 - 3,60010 2,500 – 3,800 Approx. 3,400 Ridership

6 Assumes service terminates in central Salem.

7 Estimates based Between Beaverton and Salem.

8 Estimate extrapolated to 2020. Assumes service terminates in central Salem.

9 Assumed service terminates in north Salem.

10 Estimate extrapolated to 2020 based on URS 2015 projections.

91 Table C-10 compares the projected WSK ridership levels with current ridership data for other similar commuter rail lines.

Table C-10: Projected WSK Extension Ridership vs. Current Ridership Data11

Year 2020 Current

Altamont WES Front Rail Commuter Beaverton to Coaster Runner Runner Express WSK Wilsonville, San Diego, Ogden-Salt- Albuquerque, Stockton to Line Extension OR CA Lake, UT NM San Jose, CA

Ridership 3,400 (avg.) 1,200 5,100 7,900 2,700 3,700

Corridor 620,000 440,000 900,000 659,000 396,000 997,000 Population

Riders per Day per 340 150 567 1,199 682 371 100,000 Population

Ridership Conclusions The ridership forecast model developed for this feasibility study yields results that fall within the range of projections developed for prior studies. While there are obvious differences due to the modeling details and assumptions, the model outputs were compared with general expectations for non-auto travel, which produced ridership results within a reasonable range.

The future-year ridership projections indicate that the commuter line would attract a range of different trips based on the type of service, station locations, and connections provided under the various scenarios. When compared with other commuter rail lines around the country, ridership projected for the WSK extension projections is within the range of other operations. However, caution is expressed here as Table C-10 compares 2020 WSK ridership with existing ridership from other lines; it is likely that the 2020 projections (if they were available) for the other lines would be substantially higher, resulting in the WSK data appearing at the low end of ridership levels.

11 Sources: Metro/Nelson‐Nygaard for Regional High Capacity Transit Study; American Public Transport Association (APTA), third quarter 2008 statistics.

92

Appendix D: Cost Estimates

Summary and Methodology This Appendix presents the assumptions relied on to prepare the capital and operating cost estimates for extending WES commuter rail service from Wilsonville to Salem (WSK). Both estimates were conducted at a conceptual level. Conceptual capital costs were developed using data gathered from current unit cost information presented in the Portland to Eugene Intercity Passenger Rail Assessment1 the WES project (TriMet), the Lessons Learned from the WES project report, and other relevant and available data from recent, similar commuter rail projects.

Track Right of Way and Conditions The alignment from Wilsonville Station to Salem could support passenger train speeds up to 110 mph, if the track is rebuilt, high- protection is provided, and private crossings are grade-separated or removed, except where curvature requires reduced speeds. The existing right of way is large enough for two mainline tracks; a few locations where there is less than 50 feet of right of way may require that the existing track be shifted to accommodate a second track for commuter rail operations.

Both buried and overhead utilities are located within the right of way in the Wilsonville city limits and at other locations along the corridor. Utilities may need to be relocated to accommodate additional track.

The single-track Willamette River Bridge near Wilsonville was built in 1976 and is in good condition. This study assumes the bridge would remain as a single track. The alignment also crosses the 570-foot Lake Labish concrete bridge upon entering Keizer city limits.

Track through north Salem is located within Front Street with limited separation from vehicular and pedestrian traffic along with several at-grade crossings. Improving speed would require a shift of the mainline to separate the railroad track from the street and to eliminate conflicts between different uses at intersections. Speeds through Salem would be limited to less than 40 mph and a quiet zone would help reduce impacts on adjacent neighborhoods.

Maintenance Facility Options WES has a maintenance and storage facility at Wilsonville Station, which is operating near capacity. However, it was constructed and land was acquired to accommodate expansion should commuter rail be extended to Salem in the future. In addition, a storage facility would be needed near the Salem terminus to house trains between service hours. Thus, there was no assumption for a second maintenance facility but

1 Oregon Department of Transportation, Portland to Eugene Intercity Passenger Rail Assessment , June 2010.

93

costs were included to expand the capacity of the current maintenance facility as well as build additional storage facilities.

Conceptual Capital Cost Estimate The conceptual capital cost estimate for the commuter rail extension to Salem was developed using unit cost information from the Portland to Eugene Intercity Passenger Rail Assessment costs from the WES project acquired from TriMet, the Lessons Learned from WES report (Appendix A), and relevant data from other recent commuter rail projects. Mitigation measures (passing sidings double tracking, station design, crossovers, etc.) were assumed over the length of the corridor and varied under each scenario. Although no preliminary engineering analysis was conducted, a set of engineering assumptions was made using information gathered from track charts, aerial mapping and numerous site visits. In addition, the Portland to Eugene Intercity Passenger Rail Assessment provides cost data related to bridge inspection and replacement. The cost estimation memorandum in Appendix E summarizes the conceptual capital cost assumptions and details.

Major cost elements for the extension include:

 Train equipment (number of new trainsets needed), which varies depending on the operating scenarios (frequency of trains)

 Stations at Woodburn, Keizer and Salem

 Maintenance and support facilities

 Signals/communications and crossings warning upgrades

 Track upgrade and/or replacement

 Passing sidings, and double track (Tualatin to Wilsonville)

Other key points used in calculating estimated capital costs include the following:

 Based on the railroad outreach, PNWR would likely require mitigation (a second track) for the existing WES line between Wilsonville and Tualatin.

 BNSF would consider selling its right of way outright, but this cost is not included in the estimates.

 There is a cost “premium” for administration of an FTA grant, multiple agency agreements and funding, and construction of improvements under rail traffic, which are factored into the contingency costs.

 As noted above, it is assumed that the existing Willamette River Bridge on the OE line will remain in its current condition and single-track configuration, and will not be upgraded to allow for faster, passenger train operation.

94

 WES costs were approximately $11 million per mile (all costs included). The cost estimates for the commuter rail extension range from $11.3 million to $13.3 million per mile, of which approximately $1.0 million to $1.5 million per mile is due to mitigation measures along the existing WES corridor.

 A “Quiet Zone” is assumed to be included as part of the route through Salem.

The conceptual capital cost estimate is presented on the following page.

95

Table D-1: WSK Conceptual Capital Cost Estimate

UNIT COSTS 60-MINUTE 30-MINUTE COST ELEMENT (RAIL ASSUMPTIONS) (Per mile or FREQUENCY FREQUENCY per each) SCENARIO SCENARIO Track- New with Subgrade; including: welded rail, concrete ties, and new turnouts OE line: assume 3-5 miles of sidings along the 30 $2,500,000 $15,000,000 $25,000,000 mile corridor, plus another 3-5 miles for a second or double-track between Tualatin and Wilsonville, depending on operating scenario

Track-Replace/Upgrade to as much as 79 MPH $1,000,000 $30,000,000 $30,000,000 (FRA Class 4)

Track – Replace ties for track upgrades $500,000 $15,000,000 $15,000,000

Double Crossovers (1 location in the double- $2,000,000 $2,000,000 $2,000,000 tracked section)

Stations (Woodburn, Keizer, Salem $5,000,000 $15,000,000 $15,000,000

Signals & Communications (does not include $1,000,000 $25,000,000 $25,000,000 crossings)

Maintenance, storage and support facilities $10,000,000 $8,000,000 $10,000,000

Crossings – All public crossings (including crossing $400,000 $12,000,000 $12,000,000 signal upgrade)

Bridges-Replace timber with concrete $4,800 $24,000,000 $24,000,000

Equipment $5,000,000 - $30,000,000 $50,000,000 $6,000,000

SUBTOTAL $176,000,000 $208,000,000 Preliminary Engineering and Permitting Services (13%) $ 22,880,000 $ 27,040,000

Construction Engineering (8%) $14,080,000 $16,640,000

Contingency (50%)2 $88,000,000 $104,000,000

FTA/Multi-Agency Administration (15%) $26,400,000 $31,200,000

TOTAL $327,360,000 $386,880,000

2 Accounts for design components not readily identified at this level of detail. Also includes costs for establishing a Quiet Zone through Salem.

96

Conceptual Operating Cost Estimate Conceptual operating costs were developed for the Wilsonville-to-Salem commuter rail extension using information from TriMet, the Lessons Learned from WES analysis (Appendix A), and other recent similar commuter rail projects. Currently, the WES line’s annual operating costs are approximately $2.25 million (TriMet interview in April 2009). Based on this discussion and an analysis of commuter rail projects in operations elsewhere in the United States3, the estimated additional operating costs for the commuter rail extension to Salem were pro-rated based on TriMet’s current costs and compared to the operating costs from other studies, resulting in the approximate $5.5 – 6.9 million per year operating cost estimate shown in Table D-2.

Table D-2: Conceptual Operating Cost Estimate

Operational Element Estimate for Wilsonville-to-Salem Commuter Rail Extension

Transit Agency Staff $400,000 – 500,000

Railroad Maintenance Work and Support Staff $1,750,000 – 2,250,000

Commuter Train Operations $1,400,000 – 1,700,000

Performance Incentive (for on-time performance by PNWR crews) Not included in this assumption.

Insurance $2,000,000 – 2,500,0004

TOTAL $5,550,000 – 6,950,000

3 Based on “Austin‐San Antonio Commuter Rail Study” (22‐30 cents per passenger‐mile) and “Regional Commuter Rail Connectivity Study” for Houston/Galveston Region, Kimley‐Horn Associates, 2008 ($84 per train mile including operating, administrative, insurance and other costs, adjusted to $90 per train mile for 2009 costs).

4 Estimated as approximately half of the totaled other operational costs. Per interview with TriMet, insurance is assumed to be a small, marginal increase over current rate: Tuck Wilson, Neil MacFarland, Jeff Lowe, Ken Kirse, Tamara Lesh, and Thomas Heilig. Personal Interview. April 28, 2009.

97