Appendix C Operating Plans and Ridership Forecasting Report

Fixed Guideway Study

Appendix C

Operating Plans and Ridership Forecasting Report

Fixed Guideway Study

Operating Plans

and

Ridership Forecasting

Report

Central Oklahoma Transportation and Parking

Authority

Prepared by

Manuel Padron & Associates

Under subcontract to

Carter-Burgess

TABLE OF CONTENTS 1.0 Introduction ____________________________________________________________ 1 1.1 Purpose_______________________________________________________________1 1.2 Project Scenarios_______________________________________________________1 2.0 Operating Plans _________________________________________________________ 2 2.1 No Action Scenario _____________________________________________________2 2.2 Enhanced Bus Scenario _________________________________________________3 2.3 Bus Rapid Transit Scenario _____________________________________________16 2.4 Commuter Rail Scenario ________________________________________________19 2.5 System Plan __________________________________________________________20 3.0 Operating Statistics ______________________________________________________ 29 3.1 Bus Operating Statistics for 2030_________________________________________29 3.2 Rail Operating Statistics for 2030_________________________________________30 3.0 Forecasting Methods and Revisions to the OCARTS Model _____________________ 32 3.1 OCARTS Planning Area_________________________________________________32 3.2 Trip Generation _______________________________________________________32 3.2 Trip Generation _______________________________________________________33 3.3 Trip Distribution _______________________________________________________35 3.3 Trip Distribution _______________________________________________________35 3.4 Pathbuilding and Modal Choice __________________________________________36 3.5 Development of Modal Share Targets _____________________________________38 3.6 Highway Assignment___________________________________________________41 4.0 2030 System-wide Ridership Forecasts _____________________________________ 42 4.1 Ridership by Transit Mode and Trip Purpose _______________________________43 4.2 Transit Modes of Access and Egress _____________________________________44 APPENDIX A : SUPPLEMENTAL OPERATIONS SUMMARIES_____________________47 APPENDIX B : SUPPLEMENTAL RIDERSHIP SUMMARIES _______________________68

LIST OF TABLES AND FIGURES

Figure 1. COTPA 2030 No Action System Plan... 2

Figure 2. COTPA 2030 Enhanced Bus Plan... 4

Figure 4. Routes 7 and 12... 5

Figure 5. Existing Routes 11 and 14 ... 6

Figure 6. Modified Routes 11 and 14 ... 7

Figure 7. Existing Routes 23, 1, and 19 ... 8

Figure 8. Modified Route 23 (A & B)... 8

Figure 9. New Route L1 ... 9

Figure 10. New Routes E1 and F1 ...10

Figure 11. New Route E2 ...11

Figure 12. Modified Route 15 and New Routes F2 and F4...12

Figure 13. Modified Route 24 and New Route F3...13

Figure 14. Downtown Streetcar (no Intermodal alignment)...14

Figure 15. Tinker AFB Commuter Routes ...15

Table 1. Tinker AFB Commuter Route Park and Rides ...16

Figure 16. BRT Routes in the BRT Scenario...17

Table 2. BRT Stations in the BRT Scenario ...18

Figure 17. CRT Routes in the CRT Scenario ...19

Table 3. Commuter Rail Operating Characteristics in the CRT Scenario...20

Figure 18. The COTPA FGS System Plan ...21

Table 4. Commuter Rail Operating Characteristics in the System Plan ...22

Figure 19. CRT Routes in the System Plan...22

Table 5. CRT Route Summary ...23

Table 6. CRT Station Connections ...23

Figure 20. BRT Routes in the System Plan ...24

Table 7. BRT Station Connections ...25

Figure 21. Downtown Streetcar in the System Plan ...26

Table 8. Downtown Streetcar Station Connections ...27

Figure 22. New Route L2 ...28

Table 9. COTPA 2003 Operating Statistics ...29

Table 10. COTPA 2030 Bus Operating Statistics ...30

Table 11. COTPA 2030 Downtown Streetcar Operating Statistics...31

Table 12. COTPA 2030 Commuter Rail Operating Statistics...31

Figure 23. OCARTS Traffic Analysis Zones ...32

Table 13. Revised Trip Attraction Equations ...34

Table 14. Trip Generation Comparison ...35

Table 15. Trip Distribution Comparison ...36

Figure 24. General Form of the Modal Split Function ...37

Table 16. Transit HB Work Calibration Data and Results ...39

Figure 25. Transit Non-Work Model Targets and Results...40

Table 17. Transit Linked Trip Comparison...40

Table 18. Year 2000 Transit Assignment Results...40

Table 20. System-Wide Summary of Ridership Forecasts and Highway Assignments ...43

Table 21. System-wide Summary of Transit Forecasts by Transit Mode ...43

Table 22. System-Wide Transit Ridership by Trip Purpose ...44

Table 23. System-Wide Daily Modes of Access or Egress ...45

Appendix A: Operating Plans ...48

Appendix A: Operating Plans (cont’d) ...49

Appendix A : Streetcar Run Time Estimates ...50

Appendix A : Streetcar Run Time Estimates (cont’d)...51

Appendix A: Commuter Rail Run Time Estimates ...52

Appendix A: Commuter Rail Run Time Estimates (cont’d) ...53

Appendix A: Commuter Rail Run Time Estimates (cont’d) ...54

Appendix A: Commuter Rail Run Time Estimates (cont’d) ...55

Appendix A: Weekday Bus Operating Statistics ...56

Appendix A: Weekday Bus Operating Statistics (cont’d) ...57

Appendix A: Weekday Bus Operating Statistics (cont’d) ...58

Appendix A: Weekday Bus Operating Statistics (cont’d) ...59

2030 Enhanced Bus Plan (cont’d) ... Error! Bookmark not defined. Appendix A: Weekday Bus Operating Statistics (cont’d) ...60

Appendix A: Weekday Bus Operating Statistics (cont’d) ...61

Appendix A: Weekday Bus Operating Statistics (cont’d) ...62

2030 CRT Plan ...62

Appendix A: Weekday Bus Operating Statistics (cont’d) ...63

2030 CRT Plan (cont’d)...63

Appendix A: Weekday Bus Operating Statistics (cont’d) ...64

2030 System Plan...64

Appendix A: Weekday Bus Operating Statistics (cont’d) ...65

Appendix A: Rail Operating Statistics...66

Appendix A: Rail Operating Statistics (cont’d) ...67

APPENDIX B : SUPPLEMENTAL RIDERSHIP SUMMARIES...68

1.0 Introduction

1.1 Purpose

This report documents operating plans and travel demand forecasts for five scenarios considered in the Fixed Guideway Study (FGS) for the Central Oklahoma Transportation & Parking Authority (COTPA). Operating plans were developed in conjunction with COTPA staff and other local stakeholders. Travel demand forecasts were made for each scenario using the Association of Central Oklahoma Governments’ (ACOG) transportation planning model. The forecasts were for the year 2030 and model results were used to evaluate operational efficiencies, ridership potential and cost-effectiveness. The goal of these analyses was the development of an effective, cost-feasible transit system plan to serve the future demand of travelers in the metropolitan Oklahoma City area.

1.2 Project Scenarios

In the course of the FGS, five unique scenarios were analyzed for the year 2030. These include:

° No Action Plan – this scenario applies the current COTPA operating plan to the future year in order to assess the costs, impacts, and ridership of a “do nothing” approach.

° Enhanced Bus Plan – this scenario attempts to meet future travel demand through a system of local, express, and feeder bus that more than doubles the service levels COTPA currently operates. A downtown streetcar is included.

° Bus Rapid Transit (BRT) Plan – this option expands from the enhanced bus plan to add a network of bus rapid transit routes along the major arterial corridors of the region.

° Commuter Rail (CRT) Plan – this scenario modifies the BRT plan by replacing bus lanes with Commuter Rail service along the north-south (Edmond to Norman) and east-west (Yukon to Mideast-west City) corridors.

Results of the above scenarios were examined in detail and combined with feedback from COTPA and community stakeholders, led to the development of a final system plan scenario:

° System Plan – this scenario incorporates elements of all four previous scenarios and includes a three-spoked commuter rail network with downtown hub, integrated bus rapid transit, a downtown streetcar, and enhanced bus service.

Essentially, three sections follow this introduction. The first section provides details on the bus and rail operations assumed for each of the five scenarios. The next section describes the assumptions on which the forecasts are based and identifies revisions to the OCARTS model that were made to improve the forecasts. The last section presents the system-wide results of the forecasts. The report also has appendices that supplement the materials presented in each of the sections.

2.0 Operating Plans

2.1 No Action Scenario

The No Action Plan typically provides a useful starting point for analysis of future year scenarios by providing a reference for the costs, impacts, and ridership the transit operator would

expected to observe in an environment that included only existing and committed transportation projects.

In the case of COTPA, there is currently no dedicated source of funding in place to support transit improvements, such as new routes, increased frequencies, or greater coverages. Thus, for planning purposes in this study, the current system of routes with minimal adjustment is appropriately considered the 2030 No Action Plan (Figure 1).

The No Action network operates throughout the Oklahoma City metropolitan area and includes twenty-two local bus routes, three express bus routes, six rubber-tire trolley routes, and six CART (Cleveland Area Rapid Transit) bus routes operating within the City of Norman.

Local routes typically operate at a daily frequency of thirty or sixty minutes, while some operate on a staggered schedule of thirty then forty-five minutes throughout the day. Express service runs primarily in the peak periods. In general, route service is offered 5:30 a.m. to 8:00 p.m. Monday through Saturday. Appendix A provides descriptions and frequencies for the routes in the No Action Plan.

2.2 Enhanced Bus Scenario

The Enhanced Bus scenario seeks to build on the No Action Plan by improving local and express bus service, adding a downtown streetcar, and creating a commuter bus network to Tinker Air Force Base. This plan also serves as a background network for the bus rapid transit and commuter rail scenarios. Appendix A provides descriptions and headways for all routes in this scenario.

Developing this plan included the compilation of several sources for information and research and heavily considers COTPA staff’s vision and goals for enhanced bus service. Chief among sources is the COTPA Transit Development Program prepared by The Goodman Corporation (October 1996), figures from the COTPA Long Range Plan prepared by Multisystems, Inc. (date unknown), COTPA Tinker AFB employment data (January 2003), COTPA’s FY 2002-03 Route

Performance Report (June 2003), the Metro Transit Route System Modification Plan [DRAFT]

prepared by McDonald Transit Associates, Inc. (November 2004) and the report’s associated boardings by stop and line data (observed May 2004), and Metro Transit Weekly Ridership data (December 2004).

The first major goal of the enhanced bus plan is to increase service throughout the core COTPA region to a level that can reliably and sustainably serve the population at large (and specifically the transit-dependent community). Core local routes have more frequent service - often at fifteen-minute intervals in the peak periods. All offpeak service now operates at thirty-minute intervals or less. Express routes operate frequently in peak periods and minimally in offpeak periods and include new service to outer communities.

The second objective of the plan is to restructure route alignments to better reflect the travel patterns and needs of COTPA transit riders. This restructuring includes the addition and modifications of several routes to shift the COTPA system from a primarily radial one to a blend of radial and grid systems that provide both adequate crosstown and CBD-directed services. Care is given to insure that most travel requires zero or one transfers, often at locations that could develop into future transit mini-hubs. Figure 2 presents the Enhanced Bus plan for the COTPA region.

Figure 2. COTPA 2030 Enhanced Bus Plan

Following are a list of route-by-route changes made to the 2030 No Action network to develop the 2030 Enhanced Bus network:

Route 2

Consistent high-performing route serving low-income communities. Increase peak frequency from 30 minutes to 15 minutes.

Route 3

Consistent high-performing route serving low-income communities. Increase peak frequency from 30 minutes to 15 minutes.

Route 4

Average-performing route serving low-income communities. Increase peak and offpeak frequencies from 60 minutes to 30 minutes to improve access to the rest of the transit system.

Eliminate deviation to Blackwelder and Georgia Avenues; ridecheck shows this segment to be unproductive.

Route 5

Consistent high-performing route through attractive Western Avenue corridor. Increase headways from 30 minutes to 15 minutes all day to Quail Springs Mall (5A). Reduce trips continuing on to Mercy Hospital to every 60 minutes (5B).

Routes 7 and 12

Consistent high-performing routes through attractive May Avenue corridor (Figure 3). Combine routes to become single crosstown route on May Avenue with 15 minute all day service

(previous service to CBD can be accessed through single transfer or walk to Routes 8, 10, 16, or 38). Eliminate unproductive deviation on Route 7 to Villa Avenue (ridecheck). Add deviation from May Avenue west on SW 15th Avenue to Dell, south on Grand Avenue, east on SW 25th Avenue back to May (Figure 4).

Figure 4. Routes 7 and 12

Existing Routes 7 and 12. Combined Routes 7/12

Route 8

Good-performing route serving low-income communities. Increase frequencies from 60 minutes to 30 minutes all day to provide steady access to the rest of the transit system. Reroute onto Northwest Expressway to provide local service on important corridor. Eliminate deviation to NW 50th Avenue; segment is covered by modified Route 29.

Route 9

Extend western terminus to loop through NW 10th Avenue and Council Road (existing Route 23 alignment) and eliminate deviation on Villa Avenue and Main Street (9A). Increase frequencies from 60 minutes to 30 minutes all day to provide steady access to the rest of the transit system. Alternate peak pattern (9B) continues south on Council to employment center at SW 44th

Avenue.

Route 10

Slightly increase frequencies from 37 minutes to 30 minutes all day on this average-performing route.

Route 11

Restructure this low-performing route (Figure 5) from CBD-oriented to crosstown route through attractive SW 29th Street corridor at 15-minute peak and 30-minute offpeak frequencies. Trim western loop to run from SW 18th Street south on Meridian Avenue and east onto SW 29th Street. Deviate south on Portland, east on SW 36th, and north on Independence back to SW 29th. On eastern end, loop north from SW 29th onto High, west on SW 15th, south on Central, east on SW 22nd, and back to High Avenue (Figure 6).

Route 14

Improve access of average-performing route (Figure 5) to the heart of southern corridor by rerouting to stay on SW 44th Street to Western Avenue and then travel north into CBD.

Eliminates deviation through low-income community that will be served by modified Routes 11 and 20. Adjust frequencies to 30 minutes all day (Figure 6).

Figure 6. Modified Routes 11 and 14

Route 13

Average-performing route connecting low-income communities with shopping destinations. Increase frequencies from 37 minutes all day to 15 minutes peak and 30 minutes offpeak to provide steady access to the rest of the transit system. Extend southeast loop to I-240 and Shields Boulevard.

Route 16

Average-performing route should experience greater ridership as primary CBD-oriented route in southwest. Extend to SW 134th Avenue & Pennsylvania Avenue. Adjust frequencies to 30 minutes all day.

Route 18

Truncate northern end of route at Wilshire Boulevard to eliminate unproductive low-density segment. Deviate to NW 63rd, Western, Grand, and I-44. Adjust frequencies to 30 minutes all day.

Route 20

Reroute average-performing route from Walker Avenue to Robinson Avenue into CBD using existing Route 14 alignment. Adjust frequencies to 30 minutes all day.

Route 22

Consistent high-performing route serving low-income communities. Increase peak frequency from 30 minutes to 15 minutes.

Routes 23, 1, and 19

Combine consistent high-performing Route 23 with under-performing Routes 1 and 19 (Figure 7) to create enhanced crosstown Route 23 through attractive NW 23rd Street corridor. Western terminus begins in Bethany with loop through NW 36th Street & Rockwell Avenue (service to Reno Avenue & Council Road taken over by modified Route 9) before continuing east on NW 23rd Avenue. Increase frequencies from 30 minutes to 15 minutes all day through Capitol complex, ending at current eastern terminus of Route 1 in low-income community between Bryant Avenue and Coltrane Road (23A). Selected trips continue on NE 23rd Avenue to Green Pastures community every 60 minutes (23B) (Figure 8).

Figure 7. Existing Routes 23, 1, and 19

Figure 8. Modified Route 23 (A & B)

Route 29

Under-performing route serves as Meridian Avenue crosstown. Increase frequencies from 70 minutes to 30 minutes all day to provide reliable level of service through retail and airport corridor. Extend northern end of route up Meridian to loop at NW 63rd Street & Tulsa Avenue. Eliminate loop to Portland Avenue.

Route 38

Consistent high-performing route serving low-income communities and productive NW 10th Street corridor. Increase frequencies from 37 minutes all day to 15 minutes peak, 30 minutes offpeak.

Route 40

Eliminate this low-performing route traversing south Walker corridor. Service will be enhanced on Routes 11, 13, 14, and 20 which can be accessed by the low-income communities this route captures.

Route L1 (new service)

Add new crosstown service on MacArthur Boulevard at 30-minute daily frequency. Service begins at NW 126th Avenue and Rockwell Avenue, travels south on Rockwell and east on Hefner to MacArthur. Route terminates at Reno Avenue, with a deviation to Carol between NW 16th and NW 10th (Figure 9).

Figure 9. New Route L1

Express Route 15

Adjust frequencies to operate at average 30 minutes in the peak periods and 120 minutes in the offpeak. Reroute existing to exit I-40 at Air Depot Boulevard, past Tinker AFB, before

terminating at Reno Avenue & Midwest Boulevard in Midwest City.

Express Route 24

Adjust frequencies to operate at average 30 minutes in the peak periods and 120 minutes in the offpeak. Reroute to stay on I-35 into CBD (no Shields Boulevard deviation). Add stop at 4th Street in Moore.

Express Route 37

Adjust frequencies to operate at average 30 minutes in the peak periods and 120 minutes in the offpeak. Add stop at Broadway and Memorial.

Express Route E1 (new service)

Establish new Yukon Express service from Cornwell Avenue & Main Street in Yukon, west on Main Street, south on Garth Brooks Boulevard, and east on I-40 into CBD. Set frequencies at

average 30 minutes in the peak periods and 120 minutes in the offpeak (Figure 10). Stop at Council and I-40.

Figure 10. New Routes E1 and F1

Express Route E2 (new service)

Establish new Northwest Express service from Council Road & Britton Road on Northwest Expressway, east on I-44, and south on I-235 into CBD. Set frequencies at average 30 minutes in the peak periods and 120 minutes in the offpeak (Figure 11). Stop at Penn Square Mall.

Figure 11. New Route E2

Trolley Route 50 (Oklahoma Spirit Blue Line)

Eliminate and replace with new Downtown Streetcar.

Trolley Route 51 (Oklahoma Spirit Orange Line)

Restore service to January 2005 levels (60 minutes all day).

Trolley Route 52 (Oklahoma Spirit Red Line)

Restore service to January 2005 levels (5 minutes midday).

Trolley Route 31 (The Eddy Gold Line)

Increase peak service to 15 minutes.

Trolley Route 32 (The Eddy Green Line)

Add reverse loop.

Trolley Route 33 (The Eddy Maroon Line)

Increase peak service to 15 minutes.

Feeder Route F1 (new service)

Establish Yukon Circulator through Cornwell Avenue, Main Street, Garth Brooks Boulevard, and Vandament Avenue at 30-minute peak and 60-minute offpeak frequency (Figure 10 above).

Feeder Routes F2 and F4 (new service)

Establish Midwest City Circulator (F2) from Douglas Boulevard west on Reno Avenue, south on Air Depot Boulevard, and west on Adair Boulevard to Rose State College. Continue east on SE 29th Street, north on Midwest Boulevard, east on NE 23rd Street, and south on Douglas back to Reno. Establish Del City Circulator (F4) from Rose State College west on SE 29th Street, north on Bryant Avenue, east on SE 15th Street, south on Sunnylane Road, east on SE 44th Street, and north on Sooner Road back to 29th. Both routes run at 15-minute peak and 30 minute

Figure 12. Modified Route 15 and New Routes F2 and F4

Feeder Route F3 (new service)

Establish Moore Circulator from Main Street north on Broadway Avenue, east on 12th Street, north on Eastern Avenue, west on 27th Street, south on Santa Fe Avenue, east on 4th Street, north on Eastern Avenue, and west on Main back to Broadway. Service runs at 15 minutes in the peak and 30 minutes in the offpeak (Figure 13).

Figure 13. Modified Route 24 and New Route F3

Cleveland Area Rapid Transit (CART) Routes

Maintain current route alignments for these six routes in the Norman and Oklahoma University areas. Increase frequencies on five Norman circulator routes to 15 minutes in the peak and 30 minutes in the offpeak. For the OU Shuttle, add a reverse direction loop and increase frequency to 15 minutes all day both directions.

Downtown Streetcar

The downtown streetcar serves as a circulator providing frequent, direct service between the Metro Transit Center, the Oklahoma City CBD, Cox Convention Center, Bricktown, Oklahoma University Medical Center, and St. Anthony’s Medical Plaza. In future scenarios, the streetcar will also serve as a connection to commuter rail and bus rapid transit service and the proposed I-40 redevelopment boulevard.

The streetcar operates bi-directionally at a 15-minute frequency all day. It is planned to run in the middle of the street in shared right-of-way with automobile traffic and be powered by

electrified overhead light rail technology. Delay time of 160% of run time in the CBD (Metro Transit Center station to Convention Center station) and 60% elsewhere is assumed to account for the shared ROW and turning movements. Station dwell times are 20-seconds.

In the Enhanced Bus scenario, the route is 4.94 miles in length and makes 27 stops (Figure 14). Estimated running time to complete the loop is 33.7 minutes for an average speed of 8.8 mph. Appendix B presents the station-to-station run times for the streetcar as it exists in both the Enhanced Bus/BRT/CRT scenarios and the System Plan.

Figure 14. Downtown Streetcar (no Intermodal alignment)

Tinker AFB Commuter Routes

The enhanced bus plan also includes a network of commuter routes serving Tinker Air Force Base from points across the Oklahoma City metropolitan area (Figure 15). These routes each operate long-haul express service from two or three park and ride locations to Tinker AFB, with two inbound trips in the early morning (between 5:00 a.m. and 6:30 a.m.) and two outbound trips in the early afternoon (between 2:30 p.m. and 4:00 p.m.). Table 1 presents the park and ride locations where each of these commuter routes would stop.

Table 1. Tinker AFB Commuter Route Park and Rides

Route PnR Locations

T1 Edmond 2

nd _{Street & Broadway, 15}th _{Street & Broadway,} Memorial Road & Broadway

T2 Northwest Council Road & Northwest Expressway, MacArthur Avenue & Northwest Expressway, Penn Square Mall

T3 Yukon Main Street & Cornwell Avenue, Garth Brooks Road & Vandament Avenue, Council Road & I-40

T4 Mustang Mustang Road & 74th Street, Pennsylvania Avenue & I-240

T5 Norman Gray Street & University Boulevard, Main Street & 24 th

Avenue, 4th Street (Moore) & I-35

T6 Shawnee Shawnee

T7 Choctaw NE 23rd Street & Choctaw Avenue, Douglas Road & Reno Avenue

2.3 Bus Rapid Transit Scenario

The Bus Rapid Transit scenario modifies the Enhanced Bus plan by adding a layer of BRT service and adjusting other bus service accordingly. The project team in conjunction with COTPA staff and community stakeholders developed the concept for this plan.

A network of eleven BRT routes across the region operate in designated BRT lanes along arterial roads with limited stops (roughly every mile), upgraded transit facilities, and park and ride availability where feasible (Figure 16). BRT routes experience a time discount of thirty percent over non-BRT service to account for right-of-way and signaling advantages. Some BRT service (Routes B5 and B11) extends from the BRT right-of-way to make additional stops in mixed-flow traffic.

All express bus service and local route 22 are replaced with BRT service in this scenario. Overlapping service on local routes 3, 5, 23, and 38 is scaled back from fifteen to thirty minutes. Other local and feeder service exists primarily as in the Enhanced Bus scenario, with

connections made to proximate BRT stations. The downtown streetcar and Tinker network continue to operate in this plan. Appendix A provides descriptions and headways for all routes in this scenario.

Figure 16. BRT Routes in the BRT Scenario

All BRT routes operate at 30-minute frequencies in the peak, and most drop to hourly service in the offpeak period. Some routes (B2, B7, B9 & B10) continue to operate at 30-minutes

throughout the day. Parking at BRT stations was assumed in locations where land use, travel patterns, and community support suggested it. Table 2 details the BRT stations by route, and identifies stations that include parking facilities.

Table 2. BRT Stations in the BRT Scenario

Route Stations (PnR in italics)

B1 Edmond

2nd & Broadway, 15th & Broadway, Memorial & Broadway, Memorial & Western, Britton & Western, NW 63rd & Western,

NW 23rd & Classen, Metro Transit Center

B2 Kilpatrick

Mercy Health Center, Quail Springs Mall, Memorial & Western, Britton & Western, NW 63rd & Western, NW 23rd & Classen, Metro Transit Center

B3 Northwest

Kilpatrick & NW Expressway, Council & NW Expressway, Rockwell & NW Expressway, Brookside & NW Expressway,

Meridian & Northwest Expressway, Integris Baptist Hospital, Penn

Square Mall, NW 23rd & Classen, Metro Transit Center

B4 Meridian

Kilpatrick & NW Expressway, Council & NW Expressway, Rockwell & NW Expressway, Brookside & NW Expressway,

Meridian & Northwest Expressway, NW 23rd & Meridian, Reno & Meridian, SW 15th & Meridian, SW 29th & Meridian, SW 54th & Meridian, Will Rogers Airport

B5 Reno

Main & Cornwell1, Garth Brooks & Vandament1, Sara & Reno,

Council & Reno, MacArthur & Reno, Meridian & Reno, May &

Reno, Western & Reno, Metro Transit Center

B6 SW 59th FAA, SW 54

th _{& Meridian, May & SW 59}th_{, Penn & SW 59}th_, Shields & SW 59th, Sooner & SW 59th, GM/Tinker

B7 Moore-OUMC 4

th _{& Telephone, 27}th _{& Shields, I-240 & Shields, SW 59}th _& Shields, SW 29th & Shields, OUMC, Lottie & NW 23rd

B8 Norman

University & Boyd, Main & Flood, Main & 24th, Main & 36th, Tecumseh & Telephone, 19th & Telephone, 4th & Telephone, 27th & Shields, I-240 & Shields, SW 59th & Shields, SW 29th & Shields, Reno & Shields, Metro Transit Center

B9 MWC/Tinker SE 29

th _{& Midwest, Midwest & Reno, Sooner & Reno, MLK &} Reno, Shields & Reno, Metro Transit Center

B10 NW 23rd

Meridian & NW 23rd, Villa & NW 23rd, Classen & NW 23rd,

Broadway & NW 23rd, Lincoln & NW 23rd, Lottie & NE 23rd, MLK & NE 23rd, MLK & NE 50th

B11 MLK Grand & NE 63

rd1_{, MLK & NE 63}rd1_{, MLK & NW 50}th_{, MLK & NE} 23rd, MLK & Reno, Shields & Reno, Metro Transit Center 1. Stop is made during BRT mixed-flow service.

2.4 Commuter Rail Scenario

The Commuter Rail Plan adds yet another mode of transit service onto the BRT scenario. Local bus service is adjusted to interact with two commuter rail lines: a North-South route operating from Edmond to Norman and an East-West route operating from Midwest City & Tinker AFB to Yukon (Figure 17). The project team in conjunction with COTPA staff and community

stakeholders developed the concept for this plan.

Figure 17. CRT Routes in the CRT Scenario

The commuter rail service seeks to provide peak-oriented service from outer suburban communities into downtown Oklahoma City along existing freight and/or passenger rail lines. Service is bi-directional with 30-minute peak and 60-minute offpeak frequencies. Double-tracking is assumed along the entire north-south corridor, while the east-west corridor assumes some single-track segments. A delay of ten seconds per mile is incorporated into the travel time

calculation for the east-west line to account for single-tracking. Station dwell times are 60-seconds.

In the Commuter Rail scenario, the north-south CRT consists of an Edmond to OKC segment and a Norman to OKC segment that join in downtown Oklahoma City at Santa Fe station. There are a total of nine full stations on the line and an additional special events station at OU. The east-west CRT consists of a Yukon to OKC segment and a Midwest City/Tinker to OKC segment that converge downtown at Union station. The line has eight total stations. Table 3 documents the travel time, distance and average speed along each segment, and Appendix C presents the station to station run times for commuter rail as it exists in both the CRT scenario and the System Plan.

Table 3. Commuter Rail Operating Characteristics in the CRT Scenario

Segment Distance (miles) Travel Time (min) Average Speed (mph)

Edmond 13.91 20.6 40.5

Norman (OU) 17.81 (19.06) 23.1 (26.3) 46.2 (43.5)

Yukon 13.70 22.1 37.2

MWC/Tinker 9.63 17.2 33.6

Five BRT routes (B1, B5, B7, B8, & B9) are discontinued in this scenario and replaced by the new commuter rail service. Underlying enhanced bus service is rerouted to connect to

commuter rail stations where appropriate. Given the lack of a common downtown CRT station between the two lines, a bus feeder route (F5) is added to circulate riders between Santa Fe station and Union station. Appendix A details operating assumptions by route for the CRT scenario.

2.5 System Plan

The System Plan seeks to combine the most successful elements of the previous four scenarios to create the most effective solution for the transit riders in the COTPA region. The concept for this plan was developed through continued coordination with COTPA staff and community stakeholders as well as extensive analysis of the ridership, costs, and logistics of the prior scenarios.

The System Plan includes three commuter rail lines, three BRT lines, a downtown streetcar, enhanced bus service, and Tinker commuter bus (Figure 18). The plan also incorporates improved connectivity between transit modes throughout the region, most notably a new

downtown intermodal transit station where commuter rail, BRT, and local service combine within the proposed I-40 redevelopment corridor. Appendix A details the route-by-route operating assumptions for the System Plan.

Figure 18. The COTPA FGS System Plan

Commuter Rail

Commuter rail in the System Plan consists of three segments: the Edmond corridor, the Norman corridor, and the Midwest City/Tinker AFB corridor. Alignment characteristics for the three segments are the same as in the CRT scenario, except the three branches would converge on a common station in downtown Oklahoma City, the new Intermodal Station within the proposed I-40 redevelopment area. Operating characteristics for each corridor are presented in Table 4.

Table 4. Commuter Rail Operating Characteristics in the System Plan

Segment Distance (miles) Travel Time (min) Average Speed (mph)

Edmond 14.18 20.8 40.8

Norman (OU) 17.54 (18.79) 22.9 (26.1) 45.9 (43.2)

MWC/Tinker 9.33 17.8 31.4

A total of four routes would operate along the three spokes of the commuter rail network. The first (C1) would operate from Edmond to Norman, the second (C2) would operate from Edmond to Midwest City/Tinker, and the third (C3) would run from Norman to Midwest City/Tinker. Headways on each line would be 60-minutes in the peak and 120-minutes in the offpeak for a combined headway of 30-minutes peak and 60-minutes offpeak within each corridor. A fourth train (C4) provides 60-minute peak direction rush service between the 4th Street station and the NW 63rd Street station (Figure 19).

Most station locations remain the same from the CRT scenario; however, the Crossroads Mall station shifts to just north of the mall complex to connect to the BRT network at SW 59th Street. The Lincoln station shifts eastward to Martin Luther King to balance station distances and provide a connection to transit riders in Northeast OKC. Appendix C provides station-to-station CRT travel times in the System Plan. Table 5 below summarizes the CRT routes depicted above.

Table 5. CRT Route Summary

Route Stations Distance (miles) Travel Time (min)

C1 Edmond to Norman 9 31.72 43.7

C2 Edmond to MWC/Tinker 8 23.51 38.6

C3 Norman to MWC/Tinker 8 26.87 40.7

C4 4th St to NW 63rd St 5 14.39 20.4

The CRT system incorporates great connectivity with the rest of the transit network. Table 6 summarizes the transit connections available at each commuter rail station.

Table 6. CRT Station Connections

Station CRT Route Parking Connecting Transit

OU (Special Events) C1, C3 Yes N11, N32

Norman C1, C3 Yes N10

Tecumseh C1, C3 Yes ---

4th Street C1, C3, C4 Yes 16, F3

SW 59th Street C1, C3, C4 Yes 20, B6

Intermodal C1, C2, C3, C4 No 13, 18, 52, B5A, B5B, Streetcar

NW 23rd Street C1, C2, C4 No 23A, 23B

NW 63rd Street C1, C2, C4 Yes 5A, 5B, 18

Memorial C1, C2 Yes L2

Edmond C1, C2 Yes 31, 32, 33

MLK C2, C3 No 22

Sooner C2, C3 Yes F2, F4

Bus Rapid Transit

Three BRT routes are proposed in the System Plan: B3-Northwest, B5-Reno, and B6-SW 59th (Figure 20). Of the eleven original BRT routes, four are replaced by commuter rail service (B1, B7, B8, & B9) and four are returned to better-performing local service (B2, B4, B10, & B11).

Figure 20. BRT Routes in the System Plan

Route B3 retains the same stations and alignment as in previous scenarios. B5 is rerouted to terminate at the new downtown intermodal center, and a second pattern of the route is added. B5A continues to run to Yukon in mixed-flow traffic and alternates service with B5B serving Will Rogers Airport. B6 is modified to have the BRT lane terminate at the SW 59th Street Commuter Rail station. The route continues on to Sooner Road in mixed flow traffic. All three routes operate at 30-minute peak and 60-minute offpeak frequencies.

Like commuter rail, the BRT routes seek to interact with the rest of the transit network. Table 7 summarizes the transit connections available at each BRT station.

Table 7. BRT Station Connections

Station BRT Route Parking Connecting Transit

Kilpatrick B3 Yes --- Council/NW B3 Yes 29 Rockwell B3 Yes 29 Brookside B3 No 8, 29, L1 Meridian/NW B3 No 29 Integris Baptist B3 No 7/12, 10

Penn Square Mall B3 Yes 4, 8

Classen B3 No 5A, 5B, 23A, 23B

Metro Transit B3 No 2, 3, 4, 5A, 5B, 8, 9A, 9B, 10, 14,

16, 20, 22, 38, Streetcar

Sara B5A No ---

Council/Reno B5A Yes 9A, 9B

MacArthur B5A No 9A, 9B, 38, L1

Meridian/Reno B5A/B No 9A, 9B, 29, 38

Fairgrounds B5A/B No 7/12, 9A, 9B

Western B5A/B No 9A, 9B, 14, 16

Intermodal B5A/B No 13, 18, 52, CRT, Streetcar

15th Street B5B No 11, 29, 38 29th Street B5B No 11, 29 54th Street B5B, B6 No 29 Airport B5B No 29 FAA B6 No --- May B6 Yes 7/12 Pennsylvania B6 No 16 Shields B6 No 20 59th Street B6 Yes 20, CRT

Downtown Streetcar

The downtown streetcar exists in similar form as in prior runs, utilizing the same operating train running characteristics and headways. The alignment and stations are slightly modified; the streetcar diverts south from the Convention Center station to the Downtown Intermodal Center, and the route is diverted to run straight up Oklahoma rather than looping to Stiles (Figure 21).

Figure 21. Downtown Streetcar in the System Plan

In the System Plan, the route is 4.85 miles in length and makes 24 stops. Estimated running time to complete the loop is 31.0 minutes for an average speed of 9.4 mph. Appendix B presents the station-to-station run times for the streetcar as it exists in both the Enhanced Bus/BRT/CRT scenarios and the System Plan. Table 8 catalogs the transit connections available at each Streetcar station.

Table 8. Downtown Streetcar Station Connections

Station Connecting Transit

Metro Transit 2, 3, 4, 5A, 5B, 8, 9A, 9B, 10, 14, 16, 20, 22, 38, B3

Memorial 2, 3, 4, 5A, 5B, 9A, 9B, 22, 38

Robert S Kerr 2, 3, 9A, 9B

Park 2, 3, 9A, 9B, 52 Convention Center 52 Santa Fe --- Intermodal 13, 18, 52, B5A, B5B, CRT Ballpark 52 Sheridan 52 Deep Deuce --- Flat Iron 22 PHF --- Lincoln 3, 18 OUMC South --- OUMC East --- OUMC North --- OUMC West 18 10th Street 3 Oklahoma --- Broadway 2 Harvey 38 St. Anthony's 4, 5A, 5B, B3 Walker 4, 5A, 5B, B3

Enhanced Bus

Enhanced bus routes in the System Plan retain the same goals developed in the previous scenarios: frequent, effective bus service connecting transit riders throughout the COTPA area to higher transit modes and ultimately to employment, shopping, and other destinations across the region. Most routes utilize headways and alignments developed in the Enhanced Bus Plan. Applicable routes continue to make connections to BRT and CRT stations. A few routes are modified in this scenario to take advantage of new transit features, and one new route is added.

Local Routes 4, 8, 10 & 38

The downtown alignments of these routes are simplified to bring them directly to the Metro Transit Center and a connection to the downtown streetcar instead of circulating around the CBD.

Local Routes 13 & 18 and Oklahoma Spirit Red Line Trolley

These routes are rerouted to terminate at the new Downtown Intermodal Center located within the proposed I-40 redevelopment zone. They are part of a southern CBD transfer hub that connects riders from the west (B5), southwest (Route 13), and northeast (Route 18) to the commuter rail network, and vice versa.

Local Route 16

Extend along SW 134th Street into the commuter rail station in Moore, providing another transit option to this community.

Local Route 29

Utilizes the more successful BRT alignment that begins up the Northwest Expressway.

Local Route L2 (new service)

This new route moves riders across the Kilpatrick corridor every half-hour, beginning at NW 126th Avenue and Rockwell Avenue, traveling east on Memorial Road and terminating at the Memorial CRT Station (Figure 22).

Figure 22. New Route L2

Feeder Route F5 and Oklahoma Spirit Orange Line Trolley

These routes are eliminated. The commuter rail circulator (F5) is unnecessary with a common downtown commuter rail station, and the Orange Line Trolley is replaced by BRT service between downtown and the airport corridor (B5B).

3.0 Operating Statistics

An operating statistics model was developed to forecast the annual vehicle miles, annual

vehicle hours, and peak vehicles by transit mode for each scenario. This model considers route length, route time (including layover), peak and offpeak period frequencies, and service span to derive daily statistics. For bus statistics, an annualization factor derived from actual data is applied to scale miles and hours up to an annual level. For rail statistics, annual data is calculated by considering the number of weekdays, Saturdays, and Sundays for which service is provided.

The model was calibrated to COTPA’s actual 2003 service statistics as reported in the National Transit Database (NTD). Route length and time were derived from the travel demand model, and route frequencies and service span were extracted from COTPA’s published bus

schedules. Weekday peak period operation was approximated from 6:30 a.m. to 8:30 a.m. and 3:30 p.m. to 6:00 p.m., and offpeak operation was estimated to occur from 5:30 a.m. to 6:30 a.m., 8:30 a.m. to 3:30 p.m., and 6:00 p.m. to 8:00 p.m. Appendix D tabulates daily trips, vehicles, miles, and hours by route calculated for 2003.

An annualization factor of 279 was calculated from the NTD report and applied to daily statistics. Annual vehicle miles and hours were estimated to approximately one percent of NTD values, and peak vehicles were projected to within three percent, or two vehicles. Table 9 presents the operating statistics for 2003.

Table 9. COTPA 2003 Operating Statistics

Actual Model Difference

Peak Buses 80 78 -2.5%

Annual Revenue Hours 236,235 239,180 1.2%

Annual Revenue Miles 3,783,765 3,737,016 -1.2%

3.1 Bus Operating Statistics for 2030

The calibrated operating statistics model was applied to the future year forecasts. As in the calibration, route length and time were derived from the travel demand model. For the No Action Plan, assumptions and results were very similar to those produced in the validation run, with small variations (around one percent) as a result of slight changes to the service plan and increased congestion in the model.

Beginning with the Enhanced Bus Plan and in all subsequent scenarios, service hours were expanded, from 5:00 a.m. to 10:00 p.m., with peak service from roughly 6:00 a.m. to 9:00 a.m. and 4:00 p.m. to 7:00 p.m. Saturday service hours are likewise increased, and Sunday service is added. Expansion of weekend service leads to the annualization factor increasing to 300 for these scenarios. Table 10 summarizes annual bus operating statistics for the 2030 scenarios, and Appendix D tabulates daily bus trips, vehicles, miles, and hours by route.

Table 10. COTPA 2030 Bus Operating Statistics

No Action Enhanced

Bus BRT CRT System Plan

Peak Buses 79 161 152 135 145

Annual

Revenue Hours 250,229 595,059 585,176 527,868 577,989

Annual

Revenue Miles 3,703,721 9,471,228 9,947,461 8,846,713 9,370,805

The Enhanced Bus plan more than doubles the service available in the No Action, through a combination of greatly increased coverage, frequencies, and service span. The BRT option enjoys a slight decrease in statistics over the Enhanced Bus Plan by running a network of more efficient BRT routes in place of express and some local service. The CRT sees a further

decrease by replacing bus operations with rail, while the System Plan, which blends a variety of bus and rail services, results in bus operating stats between CRT and BRT.

3.2 Rail Operating Statistics for 2030

Rail operating statistics were calculated for the four options in which rail was present, beginning with the Enhanced Bus scenario. Route lengths and times were taken from the run time

calculations shown in Appendices B and C. Saturday and Sunday service were assumed to approximate weekday offpeak service. To reach annual figures, rail services in the COTPA region were assumed to operate 255 weekdays, 52 Saturdays, and 58 Sundays/Holidays each year. Appendix E tabulates daily rail trips, vehicles, miles, and hours by route calculated for the 2030 operating plans.

Travel demand model output for streetcar and commuter rail was analyzed to determine the number of cars required to carry the maximum passenger volumes along each line. In all scenarios, the downtown streetcar can be served with a single car. Annual train and car statistics for the downtown streetcar are presented in Table 11. Note that the alignment variation in the System Plan results in slightly reduced train- and car-miles.

Table 11. COTPA 2030 Downtown Streetcar Operating Statistics No

Action

Enhanced

Bus BRT CRT System Plan

Peak Trains/Cars n/a 5 5 5 5

Annual Revenue Train-

and Car-Hours n/a 27,725 27,725 27,725 27,725

Annual Revenue Train-

and Car-Miles n/a 219,138 219,138 219,138 215,146

Analysis of commuter rail passenger volumes indicate that the Norman to Edmond corridor, particularly the Moore to downtown Oklahoma City segment, require more than 1-car trains. In the CRT scenario, this demand is met by operating 2-car trains on the Edmond to Norman CRT route (C1) while the Yukon to MWC/Tinker CRT route (C2) uses 1-car trains. In the System Plan, the 4th Street to NW 63rd Street Rush Train (C4) would operate 2-car trains to meet demand, while the other three lines (C1, C2, & C3) can suffice with 1-car trains. Table 12 documents annual commuter rail train and car statistics.

Table 12. COTPA 2030 Commuter Rail Operating Statistics

No Action Enhanced

Bus BRT CRT System Plan

Peak Trains n/a n/a n/a 7 7

Annual Revenue

Train-Hours n/a n/a n/a 24,763 23,520

Annual Revenue

Train-Miles n/a n/a n/a 778,958 623,810

Peak Cars n/a n/a n/a 11 8

Annual Revenue

Car-Hours n/a n/a n/a 30,883 25,050

Annual Revenue

Car-Miles n/a n/a n/a 973,084 645,826

The System Plan’s three-spoked alignment with one rush train requires the same number of peak trains as the CRT scenario’s four-pronged network; however, the former accounts for fewer train-hours and –miles due to shorter running times and distances. A significant advantage is gained in cars statistics for the System Plan by confining the 2-car train to the shorter rush train rather than having to place it on the full north-south line.

4.0 Forecasting Methods and Revisions to the OCARTS

Model

4.1 OCARTS Planning Area

The Oklahoma City planning area encompasses all of Oklahoma County and portions of five surrounding counties including; Canadian, Cleveland, Grady, Logan and McClain counties. The OCARTS model uses 878 geographic areas or traffic analysis zones (TAZs) to represent existing and future land uses in the region. The magnitude and character of the region’s trip making activities depends on the land use and socioeconomic make-up of each TAZ. Figure 23 illustrates the OCARTS TAZ structure.

4.2 Trip Generation

The OCARTS model employs a common four-step modeling process, wherein the initial step is trip generation. Trip generation models estimate the volume and types of trips made by people during a typical 24-hour weekday. The process estimates trip productions and trip attractions for each TAZ in the region.

Trip generation models usually make estimates for home-based and non-home based trips, with sub-categories of three or more trip purposes. The OCARTS trip generation model estimates person trips for five trip purposes:

° Home-Based Work (HBW),

° Home-Based Shopping (HBSHOP),

° Home-Based Social-Recreational HBSR),

° Home-Based Other (HBO), and

° Non-Home Based (NHB).

As all trips have a beginning and an end, two computations are made for each TAZ during trip generation. One computation estimates the number of trips that begin at a TAZ. In model speak, these are called productions. The other computation estimates how many trips will be attracted to or end at a TAZ and are called attractions.

After discussing the OCARTS model’s attraction equations with ACOG modeling staff, it was agreed that some changes to would enhance model performance in relation to the study’s objectives. The revised attraction equations appear in Table 13. For HBW trips, the revised equations compute work attractions for three income categories, depending on the location of a TAZ (i.e., CBD vs. elsewhere). In addition, attractions are formulated on the basis of three income categories that are generalized as low, medium or high. The income divisions were set up to be consistent with 2000 Census Journey-to-Work divisions. The income assumptions are as follows:

° Low – Under $30,000,

° Medium – Between $30,000 and $75,000, and

° High – Over $75,000.

The computation of HB non-work (HBNW) and NHB attractions have a similar CBD versus elsewhere distinction. The revised equations stem from research found in NCHRP Report 365. It merits some mention that the NCHRP report only provides for an aggregate computation of HBNW trips (i.e., a single HBNW trip purpose). As such, the attraction result is factored into the three HBNW trip purposes of the OCARTS model after the initial aggregate computation. The factors reflect percentages of the unadjusted OCARTS model, which are as follows:

° HB Shopping – 21%

° HB Social-Recreational – 27%

° HB Other – 52%

Although ACOG currently accomplishes the trip generation step using a spreadsheet, a script or control file was created for this study. The control file pivots off the same input variables used in

employment totals. A computation for internal-external and truck trips was also placed into the script.

Table 13. Revised Trip Attraction Equations

Table 14 compares trip generation results from the unadjusted OCARTS model with results that include revisions discussed in the preceding paragraphs. Although differences occur with the revisions, they stem from rounding and balancing algorithms in the software.

Home-Based Work Trips 1 Attracted to the CBD

Low Income Work Attractions = 0.107 * Retail Employment + 0.198 * Other Employment Medium Income Work Attractions = 0.661 * Retail Employment + 0.394 * Other Employment High Income Work Attractions = 0.171 * Retail Employment + 0.312 * Other Employment Attracted Elsewhere

Low Income Work Attractions = 0.977 * Retail Employment + 0.262 * Other Employment Medium Income Work Attractions = 0.289 * Retail Employment + 0.524 * Other Employment High Income Work Attractions = 0.028 * Retail Employment + 0.188 * Other Employment Home-Based Non-Work Trips 2

Attracted to the CBD

Total TAZ Attraction = 2.0*Retail Employment + 1.441*Other Employment+0.5*Occupied Households Attracted Elsewhere

Total TAZ Attraction = 9.0*Retail Employment + 1.441*Other Employment+0.5*Occupied Households with: 21% allocated as HB Shopping

27% allocated as HB Social-Recreational 52% allocated as HB Other

Non-Home Based Trips 2

Attracted to the CBD

Total TAZ Attraction = 1.2*Retail Employment + 1.047*Other Employment+0.5*Occupied Households Attracted Elsewhere

Total TAZ Attraction = 4.1*Retail Employment + 1.047*Other Employment+0.5*Occupied Households

Notes: 1) HBW equations based on 2000 Census Journey to Work Data. Income definitions:

Low = Under $30,000 Med = $30,000 to $75,000 High = Over $75,000

2) HBNW and NHB equations based on NCHRP Report 365.

Table 14. Trip Generation Comparison (Total Weekday Trips by Trip Purpose)

Trip

Purpose _{OCARTS Revised Difference OCARTS Revised Difference}

HB Work 583,668 585,805 0.4% 770,863 782,363 1.5% HB Shop 373,399 366,445 -1.9% 494,036 493,349 -0.1% HB Soc-Rec 492,765 483,363 -1.9% 657,584 656,758 -0.1% HB Other 932,839 915,154 -1.9% 1,256,936 1,257,055 0.0% NHB 1,293,266 1,269,228 -1.9% 1,734,183 1,734,606 0.0% Total 3,675,937 3,619,995 -1.5% 4,913,602 4,924,131 0.2% Volume Difference (55,942) 10,529 Jun-06 Year 2030 Year 2000

4.3 Trip Distribution

While trip generation models estimate the number of trips produced by or attracted to each TAZ, these trips are not yet going anywhere per se. This is the business of the 2nd step in the four-step modeling process, trip distribution, wherein a prediction is made of how many trips are likely to be exchanged between specific zone pairs. The products of trip distribution are trip tables or matrices containing specific zone-to-zone values representing person trips for each trip purpose.

The OCARTS model employs a standard gravity model during trip distribution. In forecasting travel demand, gravity theory maintains that travel from one zone to another grows more attractive depending on the number of trips at the origin and destination points (i.e., TAZs) and less attractive depending on the time and costs associated with making the journey. With HBW trips, for example, large residential areas are most likely to exchange trips with large

employment areas. However, the further apart a pair of TAZs are, the less likely trips will be exchanged.

All gravity models employ “friction” factors (FF’s) as a means for calibrating model results to observed trip length statistics. Revised FF’s were needed for the distribution step because changes to the OCARTS attraction equations created different interaction between TAZs. The revised FF’s were adjusted iteratively until average trip lengths for each trip purpose were similar to the unadjusted OCARTS model. The adjustment process was halted when average travel times appeared reasonably close to the unadjusted OCARTS result. Table 15 provides a comparison.

Table 15. Trip Distribution Comparison (Average Trip Lengths in Minutes)

Trip

Purpose OCARTS Revised Difference OCARTS Revised Difference

HB Work 17.2 17.8 Low income 17.1 17.3 Medium income 17.4 18.0 High income 18.0 18.6 Average HB Work 17.2 17.4 1.1% 17.8 18.0 0.8% HB Shop 9.7 10.0 2.7% 10.3 10.4 1.5% HB Soc-rec 14.7 14.9 1.4% 15.0 15.4 2.4% HB Other 10.5 10.6 1.4% 11.1 11.2 1.2% NHB 11.0 10.5 -4.2% 11.8 10.9 -7.6% Jun-06 Year 2000 Year 2030

4.4 Path building and Modal Choice

After person trips are distributed between the traffic zones, the next process in the 4-step model chain, mode choice, estimates the means by which people are expected to travel (e.g.,

automobile, bus, train, etc.) At the onset of the FGS, the OCARTS model did not have a modal choice step because transit usage is less than 1% of total daily trip making in the Oklahoma City region. As such, this became perhaps the biggest challenge in the effort to produce travel demand forecasts for the study’s scenarios.

In formulating modal choice, the predominant approach in use today is a logit model. The overall choice process weighs a variety of characteristics including those of the travelers, the journeys and the transportation systems. Traveler characteristics typically include income and sometimes auto ownership. Characteristics of the journey depend on the trip purpose and time of day when the trip occurs. For simplicity, it is common for choice models to assume all work trips occur during peak periods and non-work trips occur in off-peak times. The characteristics of the transportation system are assembled through pathbuilding processes. Pathbuilding is also called skimming because during the process, coded transportation networks are skimmed to extract the times and costs associated with the journey.

Although logit models are the preferred approach to estimating mode choice, the models are somewhat complex and require on-board survey data and considerable time to calibrate. Neither time nor on-board survey data was available and a simpler approach based on Quick

Response1 techniques was used to estimate travel market shares for the FGS scenarios. The methods of Quick Response yield a market share estimate for specific zone pairs a ratio of impedance for alternate travel modes. The theoretical form of the Quick Response model is similar to the gravity model used in trip distribution:

1

NCHRP Report 187, Quick-Response Urban Travel Estimation Techniques and Transferable Parameters, TRB 1978

Theoretical Model Form:

Tijm = Pi Aj e-bIijm j m Aj e-bIijm

where; Tijm = trips from i to j by mode m Pi = productions from zone i Aj = attractions to zone j

e = natural logarithm (2.71828..) b = trip purpose-specific constant Iijm = impedance from i to j by mode m

Fortunately, the theoretical form can be simplified to a less complicated expression. Below is an example of the theoretical model, simplified to compute transit market shares using an

impedance ratio of transit impedance (It) over auto impedance (Ia).

An example application of the simplified expression yields the curves shown below in Figure 24. Note that the value of the exponent (b) varies by trip purpose and governs the general shape of the curve.

Figure 24. General Form of the Modal Split Function

1

1 +

[ ]

I

I

b

% of Transit

Market Share

=

Impedance Ratio (Transit/Auto)

I

I

ACOG provided network line code for COTPA bus service. Line code parameters for service frequency and general alignments were checked against COTPA’s published schedules (Year 2004) and modified accordingly. As is a common practice, transit lines were coded for peak and midday service periods. This approach to coding stems largely from convenience, whereby peak impedance can be used to estimate modal shares for the work trip purpose and off-peak impedance for other purposes.

After the transit networks were coded, pathbuilding processes were scripted (TP Plus) to create impedance tables for input to the modal choice model. Bus travel times are computed on a link-by-link basis, wherein congested highway travel times of a particular link is factored depending on the highway link’s facitility type and area type. In testing the pathbuild process, the factors were adjusted iteratively until the model yielded average bus travel speeds within 5% of actual statistics.

Impedance variables in the ratio (i.e., I transit / I auto) are extracted from the highway and transit networks during pathbuilding processes. Results of the processing reflect the “best” or shortest

time path between zone pairs. For transit, the pathbuilding process yields matrices with the following values:

° In-vehicle time or ride time,

° Out-of-vehicle time (walking and waiting),

° Drive access time, and

° Fares.

Path matrices for automobile travel include the variables:

° Congested travel time by auto,

° Distance,

° Terminal time by area type, and

° Average parking costs for downtown traffic zones.

Before the mode choice model evaluates the impedance ratio, an impedance utility is computed for two competing paths (i.e., auto versus transit). The as follows:

Impedance = In-Vehicle Travel Time (IVTT) +2.5*Out-of-Vehicle Time (OVT) + Costs (in minutes)

In the above expression, the 2.5 OVT factor reflects the perception that the time spent out of a vehicle (auto or transit) is roughly twice as onerous as the ride time (IVTT). Automobile costs are assumed as $0.30 per mile plus parking. Transit costs include fares and driving cost in the case of park-ride. Costs are converted to minutes by summing all costs and dividing each zone’s average annual household income (Year 2000 Census statistics) by 120,000 annual minutes.

Income adjustment factors were applied to the transit impedance calculation to calibrate the model using targets from 2000 Census Journey-to-Work (JtW) data. Table 16 below shows the work trip data used to calibrate the model, model results and the income adjustment factors. According to 2000 JtW statistics for the OCARTS study area, 1.2% of low-income work trips were made by transit. As the universe of trips in the 2000 JtW data was 35,125, an aggregate modal share target for modeling low-income transit trips was multiplying the model’s HBW person trip population (257,129) by 1.2% to yield an target of 3,097 of transit riders. Targets for the medium and high-income categories were derived in a similar fashion.

Ridership is often generalized as being either linked trips or “total” boardings. The values in the table are linked trips or sometimes-called “first” boardings and are different from total system boardings because they do not reflect transfer activity. Moreover, linked trip estimates for the basis for cost-effectiveness evaluations.

The JtW statistics merit some additional discussion, where Table 16 shows a total of 630 HBW transit riders, of which 423 (67%) are low income. Medium and high-income categories make-up 26% and 6% of the total transit riders respectively. These base-year relationships remain largely unchanged in the 2030 forecast results because the ability to attract so-called “choice” riders from upper income households to transit depends on a host of factors beyond the forecasting methods put into the OCARTS model for this study. Congested auto travel times are a

predominant factor in the model’s computation of auto versus transit trips. Moreover, the OCARTS model predicts auto congestion will change appreciably over the 30-year horizon. Further discussion of this is provided in upcoming sections.

Table 16. Transit HB Work Calibration Data and Results

Income % by Person Transit Model Adjustment Income * by Auto by Transit Transit Trips Work Targets Result Factors

Low 35,125 423 1.2% 257,129 3,097 3,142 0.31

Medium 77,336 170 0.2% 375,709 826 938 0.60

High 41,097 37 0.1% 152,399 137 110 1.00

Total 153,558 630 785,237 4,060 4,190

* Income categories assume: Low = Under $30,000 Med = $30,000 to $75,000

High = Over $75,000 Jun-06

Daily Person Trips

Year 2000 Census Journey-to-Work Year 2000 OCARTS Model

Targets for non-work transit person trips were derived from 2004 COTPA data and the daily work trip target noted in Table 16 (i.e., 4,060). The computations and model results are shown in the figure below:

Figure 25. Transit Non-Work Model Targets and Results

9,500 COPTA Daily Riders Dec 04

x 1.25 COPTA transfer daily transfer ratio = 1.25 7,600 Estimated daily linked Trips

- _{4,060 HB W ork Target}

3,540 Non-W ork Targets (7,600 total daily minus 4,060 HB W ork) 2,990 84% HB Non-W ork

550

16% NHB

2,975 Model HBNW (adjustment factor = 1.13)

552

Model NHB (adjustment factor = 3.75)

Jun-06

The comparison below suggests the calibrated model was yielding a reasonable estimate of daily transit trips by trip purpose as compared to the targeted values. As noted earlier in this report and also to study management teams, the mode split model, its calibration and validity tests are decidedly aggregate. A fully disaggregate calibration and validation effort requires comprehensive household and on-board surveys, neither of which is presently available.

Table 17. Transit Linked Trip Comparison (Validation Targets vs. Model Results)

Trip Validation Year 2000

Purpose Targets Model Person Trips %

HB Work 4,060 4,190 130 3%

Non-Work 3,540 3,527 -13 0%

Daily 7,600 7,717 117 2%

Jun-06

Model Differences

After producing person trip estimates, a TP Plus script for assigning the trips was developed to estimate daily boarding activity. Table 18 below compares December 2004 COTPA boarding statistics to results of the Year 2000 transit assignments. As shown, results suggest the model overstates boardings for the system as a whole by 1,048 boardings or 11%. Given the level of effort in getting to this point in the processing stream, this aggregate result was considered reasonable in terms of forecasting existing and the scenarios of the FGS.

Table 18. Year 2000 Transit Assignment Results

Transit COPTA Model

Service Dec 2004 Year 2000 Boardings %

Local Bus & Trolley 9,321 10,478 1,157 12%

Express Bus 298 189 -109 -37%

Totals 9,619 10,667 1,048 11%

Jun-06

4.6 Highway Assignment

As has been mentioned earlier, the OCARTS model stream did not have a formal mode choice step at the beginning of this study. Directly after trip distribution, person trips were transformed to origins and destinations, converted to auto trips using auto occupancy factors and then assigned to the highway network. When the modal choice step was put into the model stream, additional matrix processing was needed to account for the estimated transit riders. In general, the added processing includes a step to subtract the transit person trip tables from daily person trip tables.

Table 19 compares aggregate results from highway assignment runs with the unadjusted OCARTS model versus results from the revised model. Daily vehicle-miles for the unadjusted model totaled 29.62 million for the year 2000. With all of the revisions in place, the 2000 highway assignment yields 28.62 million vehicle miles, 1 million (3%) less than the unadjusted OCARTS result. Differences in the year 2030 assignment results are similar to those of the base year 2000.

It is also interesting to note that congested speeds do not show significant degradation between the base year 2000 and the year 2030 horizon. This result suggests that horizon year highway improvements provide enough additional capacity to keep daily congestion in the year 2030 roughly the same as in the base year 2000. Moreover, increases in transit ridership over the 30-year horizon will be more directly related to growth in low-income population and improvements to transit service than declining highway congestion.

Table 19. Highway Assignment Results

Undjusted Revised Undjusted Revised

OCARTS OCARTS Absolute % OCARTS OCARTS Absolute %

Daily Vehicle-Miles 29.62 million 28.62 million 1.00 million 3% 44.64 million 43.94 million 0.70 million 2%

Daily Vehicle-Hours 768,000 738,000 30,000 4% 1,164,000 1,138,000 26,000 2%

Congested Speed 38.6 38.8 0.2 1% 38.4 38.5 0.1 0%

Jun-06

Year 2000 Year 2030