WORK PROGRAM

MBTA SFY 2019 National Transit DAtabase:
data Collection and Analysis

June 21, 2018

Proposed Motion

The Boston Region Metropolitan Planning Organization (MPO) votes to approve this work program.

Project Identification

Unified Planning Work Program (UPWP) Classification

Agency and Other Client Transportation Planning Studies and Technical Analyses

Project Number 14356

Client

MBTA
Client Supervisor: Robert Guptill

Project Supervisors

Principal: Katie Pincus
Manager: Steven Andrews

Funding Source

Future MBTA Contract

Schedule and Budget

Schedule: 17 months from notice to proceed

Budget: $136,574

Schedule and budget details are shown in Exhibits 1 and 2, respectively.

Relationship to MPO Work

This studyis supported in full with non-MPO funding. Committing MPO staff to this project will not impinge on the quality or timeliness of MPO-funded work.

Background

For many years, in support of the MBTA’s National Transit Database (NTD) submittals to the Federal Transit Administration (FTA), Central Transportation Planning Staff (CTPS) has produced passenger-miles traveled and unlinked trip estimates for the MBTA’s bus and trackless trolley modes. In state fiscal year (SFY) 1996, the scope of the analysis was expanded to include the heavy rail and light rail transit modes. In SFY 2000, the scope was expanded further to include the MBTA commuter rail mode. Since SFY 2001, CTPS has also produced passenger-miles traveled and unlinked trip estimates for the MBTA’s purchased service bus routes (that is, routes for which the MBTA contracts with a private carrier to provide the service). Beginning in SFY 2014, at the request of the FTA, the MBTA reclassified some of its bus routes as part of the rapid-bus mode for NTD reporting. Beginning in SFY 2017, the MBTA and CTPS decided to use a rolling three-year average for some of the intermediate factors used to derive passenger-miles traveled and unlinked trips.

The table on the next page summarizes the data CTPS collects and processes to estimate the average trip length, passenger-miles traveled, and unlinked trips for each mode, as well as the sources of these data.

In SFY 2014, the FTA began to allow the MBTA to use its automatic passenger counter (APC) data for reporting to the NTD. In SFY 2016, the MBTA directed CTPS to conduct passenger counts on APC-equipped buses for verifying APC data. The MBTA uses these counts, along with its APC data, to estimate the unlinked passenger trips and passenger-miles traveled on its directly operated bus mode and non-gated portion of the rapid-bus mode. CTPS will continue to collect these data in SFY 2019.

In SFY 2019, some of the purchased service bus and all of the directly operated trackless trolley routes will have APC-equipped vehicles. For these routes, the MBTA will use APC data to estimate unlinked trips and passenger-miles traveled. As with directly operated buses, CTPS will conduct ridechecks on APC-equipped purchased service buses for verifying APC data. CTPS will use full-route ridechecks to estimate total passenger-miles traveled and unlinked passenger trips for non-APC-equipped purchased service bus routes.

Data and Sources for Unlinked Passenger Trip
and Average Trip Length Calculations

Note: ODX data, if used, would replace transit trip survey data to estimate transfer factors and average trip lengths.

MB = Motorbus; RB = Rapid Bus; TT = Trackless Trolley; HR = Heavy Rail; LR = Light Rail;
CR = Commuter Rail; DO = Directly Operated; PS = Purchased Service (Purchased Transportation, PT, in the NTD); AFC = Automated Fare Collection; APC = Automatic Passenger Counter; ODX = Origin-Destination-Transfer Model

Objectives

The objectives of this project are to develop estimates of passenger-miles traveled and unlinked trips for the following MBTA directly operated transportation modes: bus, rapid bus, trackless trolley, heavy rail, and light rail. CTPS will also develop an estimate of the average trip length per passenger for the commuter rail mode and estimates of passenger-miles traveled and unlinked trips for contracted MBTA local bus service. Finally, CTPS will review the MBTA’s APC-derived passenger-miles traveled and boarding estimates.

CTPS will collect the data upon which these estimates will be based in a variety of ways:

• Ridechecks on a sample of APC-equipped buses on the directly operated bus and trackless trolley modes, the non-gated portion of the directly operated rapid-bus mode, and the purchased service bus mode
• Full-route ridechecks, including farebox fare-mix surveys, on the purchased service bus mode for the routes without APC-equipped buses
• Transit trip surveys on the heavy rail mode, light rail mode, and gated portions of the rapid-bus mode to determine origin-destination information (transfer rates and average trip lengths)
• Faregate noninteraction, farebox noninteraction, and rear-door entry surveys from stations or Green Line and Mattapan Line vehicles equipped with automated fare collection (AFC) technology
• Inferred origin-destination information from AFC data, if available from the MBTA or its partners, to determine origin-destination information (transfer rates and average trip lengths)
• Commuter rail ridership data from passenger counts conducted by the MBTA or its contractors, or from the MBTA’s mobile ticketing (mTicket) vendor

Work Description

Task 1  Develop Sampling Plans

For the directly operated bus mode and the non-gated portion of the rapid-bus mode, CTPS will develop a sampling plan for conducting ridechecks on a selection of 60 scheduled trips on each mode over the course of a single quarter during SFY 2019. The selection of quarters will be determined by CTPS’s staffing availability.

For the directly operated trackless trolley mode, CTPS will develop a sampling plan for conducting ridechecks on a selection of 60 scheduled trips over the course of a single quarter during SFY 2019. The selection of quarters will be determined by CTPS’s staffing availability.

For the purchased service bus mode, CTPS will develop a sampling plan for conducting ridechecks on a selection of 60 APC-equipped trips to verify the accuracy of the APC data. CTPS will also develop a sampling plan for conducting full-route ridechecks, including fare-mix surveys, on the non-APC equipped routes. CTPS will perform the ridechecks over the course of a single quarter during SFY 2019. The selection of quarters will be determined by CTPS’s staffing availability.

For the heavy rail mode, light rail mode, and gated portions of the rapid-bus mode, CTPS will develop a sampling plan for passenger surveys that will ensure that surveys are conducted at a random selection of stations over the course of an entire year and that the results represent all days of the week and all service periods. CTPS will also conduct noninteraction surveys at the stations.

For light rail service at surface stops, onboard observations are necessary because not all passengers interact with fare-collection equipment when boarding Green Line and Mattapan High-Speed Line vehicles. CTPS will conduct counts of passengers who do not interact with the farebox. For each car, it may be necessary to have two ridecheckers: one to count the number of rear boardings and the other to count the number of passengers boarding through the front door by category—those who interact with the farebox and those who do not interact with the farebox (the latter include flash-pass trips, children, and fare evaders). CTPS will develop a sampling plan that will ensure that these observations are conducted on surface light rail over the entire year for all days of the week and all service periods.

For the commuter rail mode, CTPS may obtain and analyze four potential data sources: conductor audits, data from the MBTA’s mobile ticketing vendor, Keolis Commuter Service’s passenger counts, and CTPS’s passenger counts collected as a part of a separate project. No direct data collection is planned for commuter rail.

CTPS will collect as much data as possible through electronic means, using hand-held mobile devices that support CTPS-developed applications.

Products of Task 1

• Sampling plan for SFY 2019 directly operated bus and rapid-bus ridechecks
• Sampling plan for SFY 2019 directly operated trackless trolley ridechecks
• Sampling plan for SFY 2019 purchased service bus ridechecks and fare-mix surveys
• Sampling plan for SFY 2019 passenger surveys
• Sampling plan for SFY 2019 faregate noninteraction counts and surface light rail observations

Task 2  Collect Data

The ridecheck assignments generated by the sampling plans created in Task 1 for the directly operated bus mode, the non-gated portion of the rapid-bus mode, the trackless trolley mode, and the purchased service bus mode will be executed by CTPS. CTPS will classify how purchased service bus passengers pay for their trips. For heavy rail and light rail, CTPS will conduct passenger surveys at each of the survey locations, and will conduct counts of the number of passengers passing through faregates—including those who do not interact with the faregates—at survey locations in stations that have faregates. Along Green Line and Mattapan Line surface routes, CTPS will conduct onboard counts of passengers, including those who do not interact with the farebox.

CTPS will perform ridechecks, passenger surveys, and passenger counts, using mobile devices where practicable. The data collected on ridechecks will be uploaded directly to CTPS’s bus ridership information database, and these data will be checked for completeness and accuracy. Passenger survey results and passenger count data will be uploaded directly to a different database, and these data also will be checked for completeness and accuracy.

The MBTA will provide CTPS with detailed AFC data for the heavy rail, light rail, trackless trolley, and gated portions of the rapid-bus network; monthly and annual fare revenue reports for the purchased service bus routes; and APC data for the purposes of verifying directly operated bus, directly operated trackless trolley, the non-gated portion of rapid-bus, and APC-equipped purchased service bus unlinked passenger trips and average trip length. If the MBTA concludes that data from the MBTA’s origin-destination-transfer model (ODX) is sufficient for NTD reporting purposes, the MBTA will provide CTPS with relevant ODX output in addition to AFC data.

Products of Task 2

• APC data for unlinked passenger trips and passenger-miles traveled on the directly operated bus mode, the non-gated portion of the rapid-bus mode, the trackless trolley mode, and the purchased service bus mode
• Ridecheck data for a selection of trips on the directly operated bus mode, the non-gated portion of rapid-bus mode, the trackless trolley mode, and the purchased service bus mode with AFC-equipped buses for verifying APC data
• Full-route ridecheck data for the purchased service bus mode without APC-equipped buses
• AFC data for total boardings for directly operated heavy rail, light rail, trackless trolley, and gated portions of the rapid-bus mode
• Transit trip survey results
• Noninteraction data for faregates at stations and for fareboxes on the directly operated surface light rail and trackless trolley modes
• Potential ODX transfer factors and average passenger trip length for heavy rail, light rail, and gated portions of the rapid-bus mode

Task 3  Process Ridecheck, Passenger Survey, and Passenger Count Data

CTPS will process the ridecheck, passenger survey, and passenger count data, including data on passenger noninteraction with faregates and fareboxes.

Product of Task 3

Processed ridecheck, passenger survey, and passenger count data

Task 4  Estimate Passenger-Miles Traveled and Unlinked Trips

Subtask 4.1  Estimate Passenger-Miles Traveled and Unlinked Trips for
Directly Operated Services

For the MBTA’s directly operated bus mode, non-gated portion of the rapid-bus mode, and trackless trolley mode, CTPS will select some bus trips that have APC-equipped buses to conduct passenger counts for use in APC data verification. CTPS will then provide the results of these counts to the MBTA as reports of vehicle passenger loads and trip summary statistics. The MBTA will use these CTPS-produced results in estimating passenger-miles traveled and unlinked passenger trips. CTPS will then assess the MBTA’s estimates of passenger-miles traveled and unlinked passenger trips.

CTPS will obtain AFC faregate passenger counts from the MBTA, which will provide information about the total number of passengers boarding at subway stations on the heavy rail, light rail, or rapid-bus mode. CTPS will then estimate the factors that account for the number of transfers between modes, based on the origin-destination passenger surveys conducted in Task 2 and the origin-destination surveys conducted in the preceding two years. In addition, CTPS will develop a faregate noninteraction factor from the observations at station survey locations, and will apply the factor to the AFC faregate counts to estimate the total number of unlinked heavy rail and light rail trips attributable to subway linked trips.

For light rail surface stops, CTPS will use counts of boarding passengers who do not interact with the farebox to develop a farebox noninteraction factor. CTPS will apply this factor to the AFC farebox counts of the total number of passengers on surface light rail. CTPS will then apply additional factors to account for transfers made to other light rail or heavy rail lines, which will generate estimates of the total of unlinked light rail and heavy rail riders attributable to light rail surface boardings.

For the heavy rail and light rail modes, CTPS will convert the origin-destination data generated by the passenger surveys and the processed AFC data into estimates of the average passenger-miles traveled per transit mode. The average passenger-miles traveled per passenger will be multiplied by the total number of passengers to yield estimates of the total number of passenger-miles traveled for each mode.

If the MBTA decides to use ODX as the basis of reporting to the NTD, CTPS will adjust the methodology as needed to incorporate the ODX outputs.

For the commuter rail mode, CTPS will use one or more of the sources of ridership counts described in Task 1 as the basis for estimating unlinked passenger trips. In the past two years, because conductor audit data were not available, CTPS obtained anonymized origin-destination data from the MBTA’s mobile ticketing vendor. That dataset was sufficient for estimating the average passenger-miles traveled per trip. In addition, in SFY 2019, CTPS will examine Keolis’s passenger counts; this dataset should also be sufficient for calculating the average passenger-miles traveled per trip. However, CTPS does not produce estimates of unlinked passenger trips or total passenger-miles traveled for commuter rail, and CTPS does not expect to conduct any direct observations of the commuter rail mode.

Subtask 4.2  Estimate Passenger-Miles Traveled and Unlinked Trips for
Purchased Service Bus Mode

For the purchased service bus mode, CTPS will produce estimates of passenger-miles traveled and unlinked trips using revenue data from the MBTA for non-APC-equipped routes and output from CTPS’s database of bus ridership information. CTPS will generate estimates of the average farebox deposit, and will then estimate the average trip length per passenger based on ridecheck observations. The total unlinked trips will be estimated by dividing the average farebox deposit by the total revenue. The total passenger-miles traveled will be calculated by multiplying the total unlinked trips by the average trip length per passenger. Considering changes to the equipment servicing the Winthrop bus routes, CTPS will work with the MBTA to develop procedures to estimate passenger-miles traveled and unlinked trips using APC data.

CTPS will combine the results from APC-equipped and non-APC-equipped routes into a single set of values for the purchased service bus mode.

Product of Task 4

Estimates of passenger-miles traveled and unlinked trips, including a summary by service day of week, for all MBTA modes, except commuter rail

Task 5  Document Results

CTPS will document the results of Task 4 and the methodology of the study in three technical memoranda: one for the purchased service bus mode, one for directly operated services, and one for the commuter rail mode. The technical memoranda will describe the data-collection and analysis processes and present a summary of the results. In addition, CTPS will document the results of Task 4 in a summary table presenting the data for the MBTA’s directly operated modes by the service day of week. The MBTA has requested that CTPS transmit a draft copy of the memoranda and table by October 15, 2019, and a final version by October 31, 2019.

Products of Task 5

Three technical memoranda and one table

Task 6  Assist with the Compliance Audit

The FTA requires that an independent auditor review and verify the MBTA’s estimates of directly operated bus and rail passenger-miles traveled and unlinked trips. As the agency responsible for these estimates, CTPS will provide any materials and assistance necessary for the audit.