Ridership Forecasting Technical Report Prepared By: VHB

APPENDIX L

Ridership Forecasting Technical Report

Prepared by: VHB I-90 Allston Interchange Project Draft Environmental Impact Report Ridership Forecasting Technical Report

TABLE OF CONTENTS

INTRODUCTION ...... 1 METHODOLOGY ...... 1 Key Assumptions ...... 1 Transit Capacity Analysis Approach ...... 3 Commuter Rail ...... 5 Rapid Transit ...... 6 MBTA Local Bus and Shuttle Bus ...... 6 RESULTS ...... 6 Ridership ...... 6 Transit Capacity ...... 8 SUMMARY OF IMPACTS ...... 11 Ridership ...... 11 Transit Capacity ...... 12

ATTACHMENT A CTPS Methodology and Assumptions of Regional Travel Demand Modeling Memorandum ATTACHMENT B CTPS Model Validation Process Memorandum ATTACHMENT C CTPS Transit Network Assumptions Memorandum ATTACHMENT D CTPS Land Use Assumptions Memorandum ATTACHMENT E CTPS Transit Crowding Analysis

i I-90 Allston Interchange Project Draft Environmental Impact Report Ridership Forecasting Technical Report

INTRODUCTION

This Ridership Forecasting Technical Report has been prepared in support of the Draft Environmental Impact Report (DEIR) for the I-90 Allston Interchange Project (the Project). The purpose of this report is to document the methodology used to determine the ridership forecasts and associated transit capacity impacts for MBTA services and other transit services to be operating at the proposed West Station and in the study area to support the planning, environmental review, and preliminary design efforts for the Project. Ridership forecasts were developed for the future No Build and Build Alternatives, for the project’s 2025 Opening Year and 2040 Design Year.

METHODOLOGY

Ridership forecasts and transit capacity impact analyses for the Project were developed using standard transportation planning industry practice for the evaluation of transportation systems and infrastructure. The analysis was based on the development of existing conditions and future year travel demand forecasts provided by the Central Transportation Planning Staff (CTPS). CTPS is the staff to the Metropolitan Planning Organization (MPO) for the Boston Region and works with the communities within the region to address issues relating to transportation, land use, and economic development. CTPS develops and maintains the regional transportation demand model. CTPS’s method of travel demand forecasting follows the traditional four steps of trip generation, trip distribution, modal split, and travel assignment, and uses EMME software to run the travel demand model. A detailed summary of the travel demand model methodology utilized by CTPS is provided in Attachment A. Details of the process used to validate the regional travel demand model for the Project are contained in Attachment B. For the Project, CTPS provided a calibrated baseline, or existing condition, and developed No Build and Build Alternative ridership forecasts for the 2025 Opening Year and 2040 Design Year. CTPS also developed forecasts for vehicle loading on commuter rail, rapid transit, local bus, and shuttle bus services operating in the study area to analyze potential passenger crowding impacts. To support the ridership modeling and transit capacity analyses, the Project: • Provided CTPS with inputs for the development of ridership estimates, including future transit service levels and future land use data in the project vicinity. Descriptions of the service level and land use assumptions used for the modeling efforts are contained in Attachment C and Attachment D, respectively. • Reviewed outputs from CTPS and provided feedback to clarify and refine results.

KEY ASSUMPTIONS

The 2040 travel demand forecasts provided by CTPS assume the implementation of several transportation projects by 2040, consistent with the currently adopted Long-Range Transportation Plan (LRTP) of the Boston Region MPO, Charting Progress to 2040 (2015). For a full listing of the transit-related projects in the region included in the CTPS model for the 2040

1 I-90 Allston Interchange Project Draft Environmental Impact Report Ridership Forecasting Technical Report

Design Year, please see Attachment C. The future improvements that are likely to have the greatest influence within or adjacent to the Project study area include:

• Increased Service on the Worcester Main Line • Green Line Extension to College Avenue and Union Square in Somerville The 2040 No Build Alternative does not include the proposed West Station. Also, no significant changes in MBTA bus services within or adjacent to the project area are expected to occur by 2040 in the No Build Alternative.

The 2040 Build Alternatives assume the opening of the proposed West Station, which would be a new stop on the MBTA’s Worcester/Framingham Line (Worcester Line) commuter rail service. MassDOT developed a proposed commuter rail schedule which meets the MBTA’s Service Delivery Policy1 of providing at least three morning and three afternoon peak period, peak direction train stops at the proposed West Station. The MBTA’s service policy also includes provision of 180-minute headways at a minimum during off-peak periods. The proposed Worcester Line schedule is included in Attachment C, and assumes that some Worcester Line trains would stop at the new Boston Landing Station at New Balance as well as at the proposed West Station. Several commuter rail peak period trains would run express through both stations. New Urban Rail train service was not incorporated within the ridership model horizon, because a future build year is not determined and service frequency has not been defined.

The 2040 Build model scenarios tested various interchange and roadway network options, including the base interchange alternative, an interchange alternative that includes a new roadway connection to Commonwealth Avenue, and an alternative that restricts portions of Stadium Way for transit vehicles only (southbound between Western Avenue and Cambridge Street South).

Bus service to West Station was also modeled, including the addition of new shuttle lines assumed to be operating in the project area. The shuttle bus assumptions incorporated into the CTPS model vary depending on the modeled scenario, as follows:

• In the 2025 No Build, 2025 Build and 2040 No Build alternatives, one new shuttle bus route between Harvard Square and Barry’s Corner in Allston was assumed to operate at 10-minute headways during the daytime and 20-minute headways during the Night Period2. • In the 2040 Build base interchange alternative and the alternative with transit restrictions on Stadium Way, three new shuttle routes were assumed, including: o One new Harvard Shuttle route between Harvard Square and West Station, operating at 5-minute headways during the AM and PM Peak Periods, 15-minute headways during the Midday Period3, and 20-minute headways during the Night Period. o One new Kendall Shuttle route between Kendall Square and West Station via Central Square, operating at 5-minute headways during the AM and PM Peak

1 Massachusetts Bay Transportation Authority. Service Delivery Policy. June 2, 2010. www.mbta.com 2 Night Period: 6:00 PM to 6:00 AM 3 Midday Period: 9:00 AM to 3:00 PM

2 I-90 Allston Interchange Project Draft Environmental Impact Report Ridership Forecasting Technical Report

Periods, 15-minute headways during the Midday, and 20-minute headways during the Night Period. o One new LMA Shuttle route between Ruggles Station and West Station via the Longwood Medical Area, operating at 5-minute headways during the AM and PM Peak Periods, 15-minute headways during the Midday, and 20-minute headways during the Night Period. • In the 2040 interchange alternative with a new roadway connection to Commonwealth Avenue, two new shuttle routes were assumed: o One new Harvard Shuttle route between Harvard Square and West Station, operating at 5-minute headways during the AM and PM Peak Periods, 15-minute headways during the Midday, and 20-minute headways during the Night Period. o One new Kendall-LMA Shuttle between Kendall Square and West Station via Ruggles Station and the Longwood Medical Area, operating at 5-minute headways during the AM and PM Peak Periods, 15-minute headways during the Midday, and 20-minute headways during the Night Period. Bus destinations and headways used in the CTPS model were developed based on coordination between CTPS, MassDOT, and Harvard University representatives. The operator for these shuttle routes is not determined at this time, and could be operated by the MBTA or by others. The model included minor diversions of the existing MBTA local bus Routes 64 and 66 in the vicinity of West Station, but did not consider potential stops by intercity buses travelling via I-90 to the South Station bus terminal. The model also excluded construction of any parking facilities at or near the station. Curbside taxi and "kiss-and-ride" drop-off accommodations at West Station were included in the model. The ridership model assumed that land development within the “project area” would amount to approximately 7 million square feet. For the purposes of the CTPS modeling, the project area includes the Beacon Park Yard, the Harvard University Enterprise Zone, the Harvard Business School area and the Harvard athletic fields/recreational area (CTPS TAZs 238, 244, 245 and 246). More details on the land use assumptions used to develop ridership data are contained in Attachment D. TRANSIT CAPACITY ANALYSIS APPROACH

A description of the approach and specific assumptions used in the transit capacity analysis is presented in the following sections. The MBTA’s Service Delivery Policy states that its purpose is “to ensure that the MBTA provides quality transit services that meet the needs of the riding public,” which is consistent with the MBTA’s enabling legislation and other external mandates. Vehicle load standards, as detailed in the Service Delivery Policy, define the levels of crowding that are acceptable by time period and mode of transportation. The Service Delivery Policy outlines weekday time periods of service for vehicle loading standards, as shown in Table 1. The CTPS model uses different time period definitions, as outlined in Table 2. For purposes of the transit capacity analysis, the AM and PM Peak periods from the CTPS model were analyzed, as these are the periods that represent the highest ridership and levels of passenger loading within transit vehicles. From a comparison of Table 1 and Table 2, the CTPS model AM Peak period is comprised of the MBTA Service Delivery

3 I-90 Allston Interchange Project Draft Environmental Impact Report Ridership Forecasting Technical Report

Policy Early AM and AM Peak periods. The CTPS model PM Peak Period is comprised of portions of the MBTA Service Delivery Policy Midday School and PM Peak periods.

Table 1 – MBTA Service Delivery Policy Weekday Time Period Definitions

Time Period MBTA Policy Definition Early AM 6:00 AM – 6:59 AM AM Peak 7:00 AM – 8:59 AM Midday Base 9:00 AM – 1:29 PM Midday School 1:30 PM – 3:59 PM PM Peak 4:00 PM – 6:29 PM Evening 6:30 PM – 9:59 PM Late Evening 10:00 PM – 11:59 PM Night/Sunrise 12:00 AM – 5:59 AM Source: MBTA. Service Delivery Policy. 2010.

Table 2 – CTPS Model Time Period Definitions

Time Period CTPS Model Definition AM Peak 6:00 AM – 9:00 AM Midday 9:00 AM – 3:00 PM PM Peak 3:00 PM – 6:00 PM Night 6:00 PM – 6:00 AM Source: CTPS (see Attachment A).

The MBTA Service Delivery Policy vehicle load standards used in this analysis are summarized in Table 3 by mode and time period. It is important to note that the vehicle load standards represent average maximum loads over the particular time period, expressed on a per-car basis. Because the vehicle load standards represent averages, depending on scheduling constraints and passenger peaking characteristics (including the different loadings among individual cars on a train), it is possible for some individual trips or vehicles to exceed the vehicle load criteria, even though the average load may comply with the Service Delivery Policy standards. It is also important to note that the capacity analyzed per the MBTA vehicle load standard is not the absolute maximum capacity; rather, it is a conservative interpretation based on the MBTA’s Service Delivery Policy. The absolute maximum number of passengers on a transit vehicle (also referred to by the MBTA as “crush capacity”) is greater than the vehicle load standards dictated by the MBTA’s Service Delivery Policy.

4 I-90 Allston Interchange Project Draft Environmental Impact Report Ridership Forecasting Technical Report

Table 3 – MBTA Service Delivery Policy Weekday Vehicle Load Standards

MBTA Service Delivery Policy Corresponding CTPS Model MBTA Policy Vehicle Load Standard Mode Time Period Time Period (Passengers/Seats)

Commuter Early AM & AM Peak AM Peak 110% Rail Midday School & PM Peak PM Peak Early AM & AM Peak AM Peak 225% Green Line Midday School & PM Peak PM Peak Early AM & AM Peak AM Peak 140% Bus Midday School & PM Peak PM Peak Source: MBTA. Service Delivery Policy. 2010.

The transit capacity analysis was focused on the Peak Load Point (the location of maximum utilization of a transit line, or the station-to-station segment with the highest passenger loads). By analyzing the passenger loads at the highest demand segments, and confirming that there is available capacity, it would stand to reason that there is excess capacity on the remainder of the line. The Peak Load Points, and their associated projected passenger volumes, were identified from the CTPS travel demand model for the No Build and Build Alternatives. Peak period passenger volumes were then factored for the peak hour, when the highest ridership occurs. The peak hour/peak period ratios were developed by CTPS based on empirical data for each mode and direction. The peak hour passenger demands were then compared to the system capacity using the MBTA’s Service Delivery Policy vehicle load standards. Peak hour maximum load capacities were calculated based on assumed average operating headways4, car seated capacities established from the 2014 MBTA Ridership and Service Statistics (the “Blue Book”), the number of cars per trainset (where applicable), and the vehicle load standards per MBTA policy. Capacity assumptions used in the analysis for each mode are summarized below. Commuter Rail

The commuter rail capacity analysis considered the MBTA’s Worcester Line service, which operates through the project area and will serve the proposed West Station. The assumed 2025 Opening Year and 2040 Design Year commuter rail operating headways were calculated from the proposed future year schedule provide in Attachment C. Six-car trainsets on the Worcester Line were assumed for the future year transit capacity analysis. Car seated capacities were established as 185 passengers per car, or 1,110 passengers per six-car trainset, based on data from the 2014 MBTA Blue Book.

Using the vehicle load standards defined in the MBTA’s Service Delivery Policy (shown in Table 3), with the trainset seated capacity and the number of trips per peak hour, the Policy

4 The headway is the elapsed time between one vehicle and then next traveling in the same direction.

5 I-90 Allston Interchange Project Draft Environmental Impact Report Ridership Forecasting Technical Report maximum load capacity for the peak hour was established. Attachment E contains the maximum load capacity calculations for the Worcester Line. Rapid Transit

The rapid transit capacity analysis considered the MBTA’s Green Line B Branch service, which operates in the vicinity of the project area. Based on the proposed 2025 Opening Year and 2040 Design Year rapid transit operating headways, which were assumed to equal existing headways, the number of Green Line trips provided during the AM and PM peak hours was estimated. Three- car trainsets on the Green Line were assumed for the future year analysis. Car seated capacities were established as 45 passengers per car, or 135 passengers per three-car trainset, based on data from the 2014 MBTA Blue Book.

Using the vehicle load standards defined in the MBTA’s Service Delivery Policy (shown in Table 3), with the trainset seated capacity and the number of trips per peak hour, the Policy maximum load capacity for the peak hour was established. Attachment E contains the maximum load capacity calculations for the Green Line B Branch. MBTA Local Bus and Shuttle Bus

The bus transit capacity analysis considered MBTA local bus routes operating in the vicinity of the Project (Routes 57/57A, 64, 66, and 70/70A), as well as proposed new shuttle bus routes (Harvard – Barry’s Corner Shuttle, Harvard Shuttle, Kendall Shuttle, LMA Shuttle, and a West Station-LMA-Kendall Shuttle). Based on the proposed 2025 Opening Year and 2040 Design Year operating headways, the number of bus trips provided during the AM and PM peak hours was calculated. Future year headways for MBTA local bus routes were assumed to equal existing headways. Headway assumptions for proposed new shuttle services were developed based on coordination between CTPS, MassDOT, and Harvard University representatives. Seating capacity for MBTA local buses was established from the 2014 MBTA Blue Book, equal to 39 passengers for Routes 57/57A, 64, and 70/70A, and equal to 57 passengers for Route 66. Proposed new shuttle bus services were assumed to operate with vehicles providing the same capacity as a standard MBTA city bus, with a seated capacity of 39 passengers.

Using the vehicle load standards defined in the MBTA’s Service Delivery Policy (shown in Table 3), with the bus seated capacity and the number of trips per peak hour, the Policy maximum load capacity for the peak hour was established. Attachment E contains the maximum load capacity calculations for the bus and shuttle routes analyzed.

RESULTS

RIDERSHIP

The ridership forecasting results are summarized in the following sections. Given the inputs and modeling assumptions described above, as well as the roadway networks developed for the

6 I-90 Allston Interchange Project Draft Environmental Impact Report Ridership Forecasting Technical Report modeled scenarios, the daily commuter rail and bus shuttle ridership at West Station was projected for the 2025 Opening Year and 2040 Design Year alternatives, as summarized in Table 4. In the 2040 Design Year, daily commuter rail ridership at West Station is projected to total approximately 220 to 270 boardings, depending on the roadway network option. As compared to the commuter rail projections, a substantially higher number of bus shuttle boardings are projected at West Station, ranging from approximately 650 to 2,300 daily boardings, depending on the roadway network option. The 2040 alternative resulting in the highest number of daily transit boardings at West Station was found to be the option with transit restrictions on Stadium Way, which was forecast to have 270 commuter rail boardings and 2,300 bus shuttle boardings for a total of 2,570 transit boardings at the station. However, this is only 70 total boardings more (+ 2.8 percent) than the base interchange alternative which is projected to have 2,500 total boardings and does not include user restrictions on Stadium Way southbound. Table 4 Daily West Station Ridership Projections Summary

Source: CTPS. Results rounded to nearest 10.

Table 5 summarizes the ridership projections for the commuter rail stations, MBTA bus routes, and other bus shuttle routes in the study area under each analyzed network option. It is notable that once West Station is opened in 2040, projected demand at Boston Landing Station is expected to decrease, suggesting that some Boston Landing riders would divert to West Station upon opening. Together, the two stops offer an approximate doubling of total ridership from the 2040 No Build Boston Landing demand (200 daily boardings) to the combined 2040 Build Boston Landing/West Station demand (380 to 430 daily boardings, depending on the network option). However, the combined ridership for the two stations is anticipated to be much lower than the combined demand projected at the three Newton stations (Auburndale, West Newton, and Newtonville).

7 I-90 Allston Interchange Project Draft Environmental Impact Report Ridership Forecasting Technical Report

Table 5 Daily Transit Ridership Projections Summary

Source: CTPS. a Results rounded to nearest 10. b Results rounded to nearest 100.

TRANSIT CAPACITY

The results of the transit capacity analysis are summarized in the following sections, including potential impacts of the ridership forecasts on the future capacity on commuter rail, rapid transit, and local bus routes operating in the vicinity of the Project. Detailed analysis tables are provided in Attachment E.

The analysis compared the projected peak hour demand from the CTPS travel demand model to available capacity using the conservative loading standards from the MBTA’s Service Delivery Policy. Volume-to-capacity (V/C) ratios greater than 1.0 indicate that the average demand during the peak hour would exceed available capacity as defined by MBTA policy. Table 6 and Table 7 summarize the results of the 2025 Opening Year transit vehicle loading analysis for the AM and PM peak hours, respectively. Results for the 2040 Design Year transit vehicle loading analysis for the AM and PM peak hours, respectively, are summarized in Table 8 and Table 9.

8 I-90 Allston Interchange Project Draft Environmental Impact Report Ridership Forecasting Technical Report

Table 6 2025 Opening Year Vehicle Loading Summary for AM Peak Hour

Source: CTPS (see Attachment E). a IB = inbound; OB = outbound.

Table 7 2025 Opening Year Vehicle Loading Summary for PM Peak Hour

Source: CTPS (see Attachment E). a IB = inbound; OB = outbound.

9 I-90 Allston Interchange Project Draft Environmental Impact Report Ridership Forecasting Technical Report

Table 8 2040 Design Year Vehicle Loading Summary for AM Peak Hour

Source: CTPS (see Attachment E). a IB = inbound; OB = outbound. Volume-to-capacity (V/C) ratios greater than 1.0 indicate that the average demand during the peak hour would exceed available capacity as defined by the MBTA’s Service Delivery Policy.

Table 9 2040 Design Year Vehicle Loading Summary for PM Peak Hour

10 I-90 Allston Interchange Project Draft Environmental Impact Report Ridership Forecasting Technical Report

In 2025, the Build Alternative would not result in crowding impacts to commuter rail, rapid transit, local bus or bus shuttle routes that would exceed the MBTA’s Service Delivery Policy maximum load that would be greater than the impacts anticipated in the No Build Alternative. In the 2025 Opening Year No Build Alternative and the Build Alternative, loading on the MBTA’s surface Green Line B Branch and Route 66 are projected to exceed Service Delivery Policy capacity in the inbound direction during the AM Peak Hour. During the PM Peak Hour, loading on the MBTA’s Route 64 is projected to exceed Service Delivery Policy capacity in the inbound direction in both the 2025 No Build and Build scenarios. In all of these cases, overcrowding is projected in the No Build Alternative and is not a result of the proposed project. There would be no additional impacts to passenger crowding in the 2025 Opening Year as a result of the Project. In the 2040 Design Year, the Build Alternative would not result in crowding impacts to commuter rail, rapid transit, local bus or bus shuttle routes that would exceed the MBTA’s Service Delivery Policy maximum load that would be greater than the impacts anticipated to occur in the No Build Alternative. During the AM Peak Hour, projected growth is anticipated to result in passenger crowding on the MBTA’s surface Green Line B Branch (inbound), Route 66 (inbound) and Route 70/70A (outbound) under the No Build Alternative. Passenger crowding on the MBTA’s surface Green Line B Branch (inbound), Route 66 (inbound) and Route 70/70A (outbound) in the 2040 Build Alternatives is projected to be approximately equal to or less than overcrowding anticipated in the No Build Alternative. During the PM Peak Hour, growth in the No Build Alternative is anticipated to result in passenger crowding on the MBTA’s Route 64 (inbound and outbound), and Route 70/70A (inbound). In the Build Alternatives, passenger crowding on the Route 64 is projected to be less than overcrowding anticipated in the No Build Alternative. On the Route 70/70A, passenger crowding during the PM Peak Hour in the Build Alternatives is anticipated to be slightly higher than crowding anticipated in the No Build Alternative, exceeding capacity by approximately five to six percent as compared to one percent in the No Build Alternative. However, in both the No Build and Build Alternatives, projected overcrowding on the Route 70/70A that would exceed the MBTA’s Service Delivery Policy maximum load is due to growth in the No Build Alternative, and is not a result of the proposed project.

SUMMARY OF IMPACTS

RIDERSHIP

The proposed project would result in approximately 220 to 270 daily commuter rail boardings at West Station in the 2040 Design Year, depending on the selected roadway network option. As compared to the commuter rail projections, a substantially higher number of bus shuttle boardings are projected at West Station, ranging from approximately 650 to 2,300 daily boardings, depending on the roadway network option. West Station is projected to divert some ridership from Boston Landing Station. Together, the two stops offer an approximate doubling of total ridership from the 2040 No Build Boston Landing demand (200 daily boardings) to the combined 2040 Build Boston Landing /West Station demand (380 to 430 daily boardings, depending on the roadway network option).

11 I-90 Allston Interchange Project Draft Environmental Impact Report Ridership Forecasting Technical Report

TRANSIT CAPACITY

A comparison of vehicle loading projections between the No Build and Build Alternatives shows that any passenger overcrowding that would exceed the MBTA’s Service Delivery Policy maximum load is due to growth in the No Build Alternative, and is not a result of the Project. Therefore, no additional mitigation measures would be required to address capacity constraints.

Attachment A

METHODOLOGY AND ASSUMPTIONS OF CENTRAL TRANSPORTATION PLANNING STAFF REGIONAL TRAVEL DEMAND MODELING

Central Transportation Planning Staff

State Transportation Building Ten Park Plaza, Suite 2150 Boston, Massachusetts 02116

May 8, 2017

INTRODUCTION

The regional travel forecasting model set of the Central Transportation Planning Staff (CTPS) is based on procedures that have evolved over many years at CTPS. It follows the traditional four- step travel-modeling process of trip generation, trip distribution, mode choice, and trip assignment and is implemented in the EMME software package. This modeling process is employed to estimate present and future average weekday transit ridership and average weekday highway traffic volumes, primarily on the basis of demography and the characteristics of the transportation network. The model set simulates travel on the entire eastern Massachusetts transit and highway systems. When the model set estimates future travel, the inputs include forecasts of demography and projections of the impacts of transit and highway improvements.

This document describes in detail the model set developed for the Allston Interchange Traffic Study. The model set will be referred to as “the model” for brevity.

CTPS 3 May 8, 2017

DESCRIPTION OF THE MODEL

OVERVIEW OF THE FOUR STEPS

The CTPS Modeling Area is comprised of 101 cities and towns that make up the Boston Region Metropolitan Planning Organization (MPO) area, along with 63 eastern Massachusetts communities outside of the MPO area. The first step is trip generation, in which the total number of trips produced by model area residents is calculated using demographic and socioeconomic data and trip production rates estimated from household travel survey data. Similarly, the number of trips attracted to different types of land use, such as employment centers, schools, hospitals, shopping centers, etc., is estimated using land use data and trip generation rates obtained from household travel surveys. This information is produced at the level of disaggregated geographic areas known as transportation analysis zones (TAZs). All calculations are performed at the TAZ level.

In the second step, trip distribution, the model combines each trip-end production with a trip-end attraction to form a trip, resulting in the distribution of the trips produced in each TAZ throughout the region. Trips are distributed based on transit and highway travel times, distances, and costs between TAZs, on the relative attractiveness of each TAZ, as measured by the number of trips attracted to that TAZ, and on the travel patterns recorded in household travel surveys.

Once the number of trips for each purpose between TAZ pairs is determined, mode choice (the third step), allocates the trips among the available modes of travel. These include walk, single- occupancy vehicle (SOV), high-occupancy vehicle (HOV), and transit. The latter is subdivided by access mode: walking to transit or driving to transit. To determine the proportion of trips to allocate to each mode, the model takes into account the travel times and distances, number of transfers required, parking availability the costs associated with each option. The household travel survey is utilized to determine the influence of these variables in the selection of travel modes.

After estimating the number of trips by mode for each purpose for all possible TAZ combinations, the model assigns trips to their respective specific routes in trip assignment in the fourth and final step. All possible highway routes or transit paths between two TAZs are considered in the path selection process.

A schematic representation of the modeling process is shown in Figure 1.

CTPS 4 May 8, 2017

FIGURE 1 The Four-Step Demand Modeling Process

CTPS 5 May 8, 2017

MODEL STRUCTURES AND INPUTS

Modeled Area

Figure 2 presents the CTPS model area, which is divided into five concentric rings for model estimation and calibration purposes. These rings will be referred to in subsequent elaboration of the modelling process.

Zone System

The modeled area is divided into 2,727 internal TAZs. There are 124 external stations around the periphery of the modeled area that allow for travel between the modeled area and adjacent areas of Massachusetts, New Hampshire, Rhode Island and beyond.

Transportation Networks

There are two types of network: transit and highway. Both are integrated in EMME. The highway network comprises express highways, principal and minor arterials, and local roadways. The transit network comprises commuter rail lines, rapid transit lines, bus lines (MBTA and private carriers), free shuttles, and boat lines. The model identifies the frequency of train and bus service, routing, travel time, and fares for all lines.

 Highway Network: The regional highway network contains in excess of 40,000 links and 15,000 nodes. It is fairly dense in the study area, although typical to modeled networks, it does not include all local and collector streets. Each link is coded with the appropriate free-flow speed, number of lanes, and lane capacity per hour. Functional class (expressway, arterial, collector, local road) is coded, as are various geographic flags useful for summarizing emissions. Another code is used to distinguish links open only to HOVs from all other links.

 Transit Network: The transit network represents all regional transit agency bus, rail, and boat services in eastern Massachusetts, as well as free shuttle busses and private express buses. Most-likely travel paths are built through the network, and the resulting impedances (travel times, distances, and costs) are used as inputs to the trip distribution and mode choice models. After mode choice, transit trip tables by time of day are assigned to the network travel paths.

CTPS 6 May 8, 2017

FIGURE 2 CTPS Modeled Area and Ring Boundaries

CTPS 7 May 8, 2017

Major Data Inputs

CTPS’s travel model underwent a major revision in 1993, and several important data sources were used in that revision. More recent data became available during the past decade, which have been used to bring the model up to 2012 conditions. These major data items underlying the model are as follows:

 Household Travel Survey: In 1991, CTPS conducted a household travel survey. The survey took the form of an activity-based travel diary that each subject household filled out for one weekday. Approximately 4,000 households, generating some 39,000 weekday trips, were represented in the final database. The data were used to estimate new models for trip generation, auto ownership, trip distribution, and mode choice.

 External Cordon Survey: Also in 1991, a survey of automobile travelers bound for the modeled area from adjacent areas was undertaken. Survey results were used in trip generation and distribution to update estimates of external trips.

 Site-Level Employment Database: Employment estimates for 2000 were taken from a single, unified regional employment database based on employment data from the Department of Employment and Training, supplemented with extensive research by CTPS. Aggregate employment data for the year 2009 were used to update this database for use for the base-year analysis in the regional model version used for this study.

 2000 U.S. Census: Various files from the 2000 US Census (population, households, and group quarters) were used in the model estimation and calibration processes. In addition, Census Journey-to-Work information was incorporated into the model at several stages of model development. Aggregate population estimates for the year 2009 were used to update this data for this study.

 Ground Counts: Transit ridership and highway traffic volume data representing early 1990s conditions were amassed into a database and used to calibrate the components of the travel model. Updated counts and volumes, collected between 2005 and 2012, have been used for model validation.

 On Board Transit Passenger Survey: CTPS surveyed passengers on all MBTA transit modes in an effort spanning the years 2008-2010. Data from this survey, specifically for transit service in the study area, were used to validate and calibrate components of trip distribution and mode choice for the model.

 Updated Employment, Population, and Gas Prices: The trip tables produced by the trip generation and trip distribution models represent the year 2009. These are brought up to 2012 by applying the changes in the Boston area Consumer Price Index, Boston area gasoline price, Eastern Massachusetts town residential populations, Eastern Massachusetts town residential workers, and Eastern Massachusetts town employment. The square root of the ratio of 2011 Eastern Massachusetts town residential populations to 2009 is applied to all the rows of the trip tables (with the exception of the square root

CTPS 8 May 8, 2017

of the ratio of the 2012 Eastern Massachusetts town residential workers to 2009 applied to all the rows of the home-based work trip table). The square root of the ratio of 2012 Eastern Massachusetts town employment to 2009 is applied to all the columns of the trip tables. The 2012 CPI and gas price are used to adjust the auto operating costs and the effective level of other costs in the mode choice and assignment models.

Analysis Year

The base year is 2012, the opening year 2025 and the design year 2040.

Time-of-Day Considerations

The mode choice and transit assignment steps of the modeling process are conducted on the basis of time periods. These are comprised of an AM peak period (6:00 AM–9:00 AM), a midday period (9:00 AM–3:00 PM), a PM peak period (3:00 PM–6:00 PM), and a nighttime period (6:00 PM–6:00 AM). The trip generation model, however, is based on daily trips. The trip distribution model considers two time periods: peak (the AM peak and PM peak periods) and off-peak.

The trip volumes produced by the trip generation model are split into peak and off-peak period trips. The trip tables produced by the trip distribution model are then split into the four time periods defined above. Next, the highway vehicle trips and transit person trips created by the mode choice model are converted from production/attraction format to an origin/destination format, based upon factors created from the data collected in the 1991 Household Travel Survey. It evolved that the PM peak period spread into the nighttime period between 1991 and 2012, so 15% of the PM peak period trips are moved to the nighttime period before the mode choice model is run.

The final trip tables created for each time period reflect observed levels of congestion on the highway system. The results of the four assignments are summed to obtain average weekday traffic (AWDT) results.

Population, Household, and Employment Forecasts

Households and employment by type are major inputs to the travel model process, being the variables upon which trip generation is performed. The forecasts for the region were developed by combining household and employment forecasts produced independently by the seven regional planning agencies (RPAs) in eastern Massachusetts: the Central Massachusetts Regional Planning Commission (CMRPC), Merrimack Valley Planning Commission (MVPC), Metropolitan Area Planning Council (MAPC), Massachusetts Regional Planning Commission (MRPC), Northern Middlesex Council of Governments (NMCOG), Old Colony Planning Council (OCPC), and Southeastern Regional Planning and Economic Development District (SRPEDD). The regional control totals mandated by Massachusetts Department of Transportation (MassDOT) for the future years were observed in the forecasts. Forecasts for the 101 cities and towns that make up the MAPC area (also the Boston Region MPO area) were developed by MAPC by allocating the community developments by using the land use model,

CTPS 9 May 8, 2017

Cube Land. This land use forecast was part of the Boston Region MPO’s Long Range Transportation Plan (LRTP), which was endorsed in 2015 and amended in 2016.

The land use forecast in the Allston interchange study area was superseded by Harvard University’s ten-year Institutional Management Plan. A separate memorandum documenting the land use assumptions in the study area is available.

CTPS 10 May 8, 2017

THE FOUR STEPS OF THE TRAVEL DEMAND MODELING PROCESS

TRIP GENERATION

The first step in the travel forecasting process is performed by the model set’s trip generation model. This model uses socioeconomic characteristics of the region’s population and information about the region’s transportation infrastructure, transportation services, and geography to predict the amounts of travel that will be produced by and attracted to each of the TAZs within the region.

The trip generation model is composed of seven parts:

 Base-year detailed inputs  Future-year inputs  Estimation of detailed input requirements for future years  Estimation of detailed socioeconomic characteristics  Estimation of vehicle ownership  Estimation of trip productions and attractions  Balancing of trip productions and attractions

A description of each of these parts is presented below.

Base-Year Detailed Inputs

The base-year inputs required for the trip generation model are presented in Table 1. Although the base-year trip generation model produces results for 2009, these results are subsequently factored using a Fratar process following the trip distribution model procedures to produce 2012 trip tables.

CTPS 11 May 8, 2017

TABLE 1 Trip Generation Model: Base-Year Input Requirements

Data Source Geographic Level Population 2009 estimates TAZ (census block) Group Quarters Population 2009 estimates TAZ (census block) Household Size, Income, Workers, Vehicles 2009 estimates TAZ (census block) Population Age 2009 estimates City or town Basic, Retail, Service Employment 2009 estimates TAZ Public K-12 Employment 2009 estimates TAZ Private K-12 Employment 2009 estimates TAZ College Employment 2009 estimates TAZ Resident Workers 2009 estimates TAZ (census block group) Dorm Population 2009 estimates TAZ (census block) Labor Participation Rate by Age Group Bureau of the Census Region Land Area CTPS regional database TAZ Geographical Ring CTPS regional database TAZ Public Use Microdata Areas CTPS regional database Public Use Microdata Areas External Trip Productions and 1991 External Travel Survey, Attractions 2000 U.S. Census External station External Growth Factors RPA and CTPS forecasts External station Transit Walk Access Factor Transit network TAZ

External Attraction and Production Terminal Times 1991 External Travel Survey External station

Future-Year Inputs

The future-year inputs required for the trip generation model, some of which are the same as for the base year, are:

 Total TAZ households  Total TAZ population  Total TAZ group quarters population  Total community population by age  TAZ employment in basic industries  TAZ retail trade employment  TAZ employment in service industries

CTPS 12 May 8, 2017

 Regional labor participation rates  External trip production and attraction growth factors  Transit walk-access factors

Estimation of Detailed Input Requirements for Future Years

Various procedures are used to prepare the trip generation model input data for future years. The variables that are estimated for the future years are listed below.  Households by household size  Households by income quartile  Resident workers  Households by workers per household  School employment (K-12 and college)  Dorm population  External person trips  Attraction and production terminal times

Demographic and employment information is major inputs to the travel modeling process. The future year information was developed in cooperation with the Boston Region MPO and MassDOT for the MPO’s Long Range Transportation Plan (LRTP).

Household Size

The change in TAZ average household size is implied in the base-year inputs and future-year forecasts (total population minus group quarters population divided by total households). The distribution of future-year households by household size is estimated by the following procedure.

First, the future-year households are distributed among the household size categories in the same proportions as in the base year. It is then assumed that all households capable of making the implied change (households of two or more for household size reductions; all households for household size increases) will have the same probability of changing in size by one person. This probability of changing is set equal to the extent needed to match the forecasted change in household size, and the resulting distribution of households by household size is used for the future-year scenario.

For example, in the base year the numbers of 1-person, 2-person, 3-person, 4-person, 5-person, and 6+-person households are, respectively, 100, 200, 50, 25, 10, and 5, with a total household population of 835. This represents an average household size of 2.141. If there were 780 future- year households, they would initially be distributed as 200, 400, 100, 50, 20, and 10 1-person, 2- person, 3-person, 4-person, 5-person, and 6+-person households, respectively.

However, if the future-year average household size were 2.000, then the households with 2 or more persons would have a 19 percent [(2.141 - 2) * 780/580] probability of dropping in size by one. The resulting distribution would thus be estimated as follows:

276 1-person households [200 + (.19 * 400)]

CTPS 13 May 8, 2017

343 2-person households [400 – (.19 * 400) + (.19 * 100)] 90.5 3-person households [100 – (.19 * 100) + (.19 * 50)] 44.3 4-person households [50 – (.19 * 50) + (.19 * 20)] 26.2 5+-person households [20 – (.19 * 20) + 10]

In the case of TAZs with no households in the base year, the proportional distribution of households by household size at the community level is used for the base year in these calculations.

Household Income

The future-year distribution of households by household income quartile is estimated by assuming that the proportional distribution of households by income quartile remains constant within each TAZ. In the case of TAZs with no households in the base year, the proportional distribution of households by household income at the community level is used for the base year.

Resident Workers per Household

The change in the number of resident workers at the community level is obtained by multiplying the base-year and future-year estimates of over age 15 population by labor force participation rates by age cohort. Dividing the base-year and future-year estimates of community-level resident workers by the base year and future-year numbers of households in the community, respectively, produces estimates of the base year and future year average workers per household. All of the TAZs within each community are assumed to have the proportional change in workers per household implied by these base-year and future-year community-level estimates. Multiplying the resultant estimate of resident workers per household by the forecasted number of households yields the forecasted number of resident workers by TAZ.

Household Workers

The future-year number of households per TAZ within each category of number of workers per household is estimated by using workers-per-household distribution curves developed by CTPS from the 1990 U.S. Census. These curves, summarized in Table 2 below, indicate a default percentage distribution of households for the base-year and future-year TAZ estimates of average workers per household. The proportional changes in the default number of households within each category of workers per household implied by this comparison are applied to the actual base-year TAZ distribution of households to obtain the distribution of households by workers per household to be used for the future scenario. The average number of workers per household at the community level is used for the base year in TAZs with no households in the base year.

For example, if the average number of workers per household changes from 1.7 to 1.8, the default distribution of households among the categories 0-worker, 1-worker, 2-worker, and 3+- worker would change from 7%, 32%, 45%, and 16% to 5%, 29%, 47%, and 19%, respectively. If the actual base-year distribution of households among those categories is 8%, 31%, 44%, and 17%, the changes in the default distributions indicate a future-year distribution of households of

CTPS 14 May 8, 2017

6%, 28%, 46%, and 20% 0-worker, 1-worker, 2-worker, and 3+-worker households, respectively.

School Employment

 K-12

The level of employment in schools providing education up to the 12th grade is assumed to be proportional to the number of community residents of ages 5-19.

 College

The level of employment at all colleges and technical schools within the region is assumed to be proportional to the number of regional residents of ages 20-24.

TABLE 2 Workers per Household Diversion Curves

Avg. Workers Households by Number of Workers per HH 0 1 2 3+ Total <=.45 58% 40% 2% 0% 100% .45 - .55 52% 46% 2% 0% 100% .55 - .65 47% 46% 6% 1% 100% .65 - .75 43% 46% 10% 1% 100% .75 - .85 38% 46% 13% 3% 100% .85 - .95 34% 46% 16% 4% 100% .95 - 1.05 30% 45% 20% 5% 100% 65% - (35% * 60% - (16% * (36% * Avg (15% * Avg 1.05 - 1.65 Avg Wrk/HH) Avg Wrk/HH) Wrk/HH) - 15% Wrk/HH) - 10% 100% 1.65 - 1.75 7% 32% 45% 16% 100% 1.75 - 1.85 5% 29% 47% 19% 100% 1.85 - 1.95 4% 26% 48% 22% 100% 1.95 - 2.05 3% 22% 48% 27% 100% 2.05 - 2.15 2% 18% 49% 31% 100% 2.15 - 2.25 1% 14% 49% 36% 100% 2.25 - 2.35 1% 10% 49% 40% 100% 2.35 - 2.45 1% 4% 50% 45% 100% 2.45 - 2.55 1% 4% 50% 45% 100% > 2.55 0% 5% 50% 45% 100%

CTPS 15 May 8, 2017

Dorm Population

The dorm population within a TAZ is assumed to remain the same proportion of the total group quarters population within a TAZ.

External Person Trips

Base-year external person trips are adjusted to produce traffic volumes at the external stations that match the observed counts for the base year. These base-year external person trips are then adjusted according to growth factors for the vehicle volumes at each external station. These growth factors are presently based upon an analysis of historical trends.

Attraction and Production Terminal Times

The attraction and production terminal times (the time it takes to travel between a vehicle and the trip origin or destination) are estimated through the application of a model developed at CTPS. This model first estimates terminal times as a function of household density (see Table 3). An alternative estimate of the production and attraction terminal times for each TAZ is based on employment density ranges (see Table 4). For regional modeling, the larger of the two estimates is assigned to a TAZ. Several TAZs with regionally unique characteristics (locations of major generators such as airports or large colleges) were assigned terminal times in the base year different from those estimated by the terminal-time model. In these cases, the model is used to estimate changes in terminal times.

TABLE 3 Household Terminal Time

Household Density Production Attraction (HH per acre) (minutes) (minutes) 0 – 5 1 1 5 – 10 2 2 10 – 15 3 3 15 – 25 4 4 > 25 5 5

CTPS 16 May 8, 2017

TABLE 4 Employment Terminal Time

Employment Density Production Attraction (employees per acre) (minutes) (minutes) 0 – 5 0 1 5 – 10 1 2 10 – 25 2 3 25 – 50 3 4 50 – 100 4 5 100 – 200 5 6 > 200 6 7

Estimation of Detailed Socioeconomic Characteristics

A three-way distribution of the households within each TAZ by household size, income, and workers is required in order to estimate the distribution of households by vehicle ownership levels. While this is available from the U.S. Census at the subregional level, such distributions at the TAZ level are estimated through iterative proportional fitting techniques. Using the appropriate subregional matrix as a seed, the cell values are adjusted through 10 iterations to match row and column totals to the estimated TAZ-level totals in order to produce an estimated three-way distribution of households for each TAZ.

Estimation of Vehicle Ownership

Base year households are distributed by vehicle ownership based on data from the 2000 U.S. Census. The distribution of future scenario households by vehicle ownership is estimated through the application of a set of models developed by CTPS.

The CTPS vehicle ownership model was estimated as a set of four multinomial logit disaggregate choice models, one for each of four income categories, in which the decision maker was the household unit and the set of alternatives was the ownership, by the household, of 0, 1, 2, or 3-or-more vehicles. In this model, households are segmented into four income categories, since income is believed to be the most significant variable in vehicle-ownership choice. Other variables included in the model are household size, workers per household, household density, employment density, household location, and transit walk-access factors. The data set used to estimate this model contained 3,504 observations. Once estimated, the model was validated to observed vehicle ownership data. The models, one for each household income quartile, are presented in Table 5.

Estimation of Trip Productions and Attractions

CTPS 17 May 8, 2017

The number of trip productions and trip attractions within a TAZ are estimated through the application of a set of models developed at CTPS. These models estimate the number of trip productions and attractions as a function of household size, workers per household, vehicles per household, income, household location, number of households, basic employment, retail employment, college employment, school employment, and service employment. The trip production models for the home-based purposes [home-based work (HBW), home-based work- related (HBWR), home-based personal business (HBPB), home-based social-recreational (HBSR), home-based school (HBSC), and home-based pick-up/drop-off (HBPD) are presented in Table 6, and the trip production models for the non-homebased purposes [non-home-based work (NHBW) and non-home-based other (NHBO)] and the trip attraction models for all purposes are presented in Table 7.

CTPS 18 May 8, 2017

TABLE 5 Summary of Vehicle Ownership Model

HH Workers HHs Employ High- Low- Transit Walk- Constant Size per HH per Acre per Acre Density Density Ring01 Accessibility Low-Income Household Model 0 Vehicles -0.0474 -0.1692 -0.1312 0.0239 0.7136 1 Vehicle 2 Vehicles -3.139 0.6182 0.4414 -0.0424 3+ Vehicles -5.074 0.7968 0.6927 -0.2232

Medium-Low-Income Household Model 0 Vehicles -1.573 -0.1874 -0.3417 0.05 0.5716 0.5392 1 Vehicle 2 Vehicles -1.745 0.5202 0.4279 -0.0627 -0.0334 -0.0056 3+ Vehicles -5.101 0.7371 1.112 -0.0627 -0.0693

Medium-High-Income Household Model 0 Vehicles -2.63 0.0459 0.7704 1 Vehicle 2 Vehicles -1.223 0.6609 0.2377 -0.0391 0.4026 -0.5962 -0.0054 3+ Vehicles -4.572 0.7899 1.289 -0.0779 -1.223 -0.0073

High-Income Household Model 0 Vehicles -2.793 0.0349 1 Vehicle 2 Vehicles 0.5049 0.3475 0.2688 -0.06 -0.0154 -0.0074 3+ Vehicles -3.807 0.5717 1.628 -0.136 -0.0468 -0.0077

High-Density = 1 if HH/acre > 6 or Employ/acre > 7 Low-Density = 1 if HH/acre < 0.5 and Employ/acre< 0.7 Ring01 = 1 if TAZ is in Ring 0 or Ring 1 Transit Walk-Accessibility = Portion of TAZ within walk-access distance of transit service

Balancing of Trip Productions and Attractions

Connecting a trip production with a trip attraction of the same trip purpose forms a trip. As a result, the number of productions and attractions for each trip purpose must be equal. In order to achieve this, the trip productions and attractions are balanced.

For most trip purposes, the number of regional attractions is the least reliable estimate. Therefore, the normal balancing procedure is to set the total number of regional attractions equal to the difference between the grand total of productions and the total number of external attractions.

CTPS 19 May 8, 2017

TABLE 6 Home-Based Trip Production Rate Models

Home-Based Work Trip Production Rates Home-Based Personal Business Trip Production Rates Workers HH Vehicles per HH Workers HH Vehicles per HH per HH Size 0 1 2+ per HH Size 0 1 2 3+ 1 1 0.94 1.17 1.11 0 1 1.19 1.95 2.11 2.87 1 2 1.01 1.23 1.18 0 2 2.91 3.32 3.50 4.24 1 3 1.15 1.38 1.32 0 3 3.29 3.70 3.88 4.62 1 4 1.48 1.70 1.65 0 4 4.16 4.58 4.73 5.49 1 5+ 1.56 1.78 1.71 0 5+ 1.56 4.71 4.87 5.63

2 2 2.47 2.66 2.47 1 1 0.50 1.01 1.20 1.27 2 3 2.64 2.81 2.61 1 2 1.85 2.35 2.55 2.62 2 4 2.68 2.84 2.64 1 3 2.25 2.82 3.04 3.11 2 5+ 2.83 2.99 2.79 1 4 2.52 2.91 3.08 3.13 1 5+ 2.55 2.93 3.15 3.23 3+ 3 2.72 3.14 3.68 3+ 4 2.75 4.02 4.55 2 2 1.04 1.50 1.63 2.12 3+ 5+ 2.88 4.15 4.68 2 3 1.40 1.87 1.99 2.48 2 4 2.37 2.83 2.95 3.45 2 5+ 2.44 2.91 3.03 3.52

3+ 3 1.43 1.96 2.24 2.49 3+ 4 2.00 2.75 3.14 3.49 HB Work-Related Trip Production Rates 3+ 5+ 2.34 3.20 3.67 4.08 HH Workers per HH Size 1 2 3+ 1 0.12 2 0.10 0.18 3 0.10 0.20 0.28 4 0.18 0.23 0.35 5+ 0.21 0.29 0.41

CTPS 20 May 8, 2017

TABLE 6 (cont.) Home-Based Trip Production Rate Models

Home-Based School Trip Production Rates HB Social/Recreational Trip Production Rates HH Household Income Quartile HH Workers per Household Ring Size Low Med-low Med-high High Size 0 1 2 3+ 0 & 1 1 0.20 0.12 0.08 0.06 1 0.88 0.70 0 & 1 2 1.22 0.56 0.28 0.26 2 1.79 1.13 1.17 0 & 1 3 1.82 1.42 0.51 0.51 3 1.79 1.49 1.68 2.24 0 & 1 4 2.53 1.82 1.77 1.72 4 2.02 1.95 2.14 2.87 0 & 1 5+ 5.07 4.05 3.04 2.53 5+ 3.58 3.50 3.85 3.94

2 1 0.15 0.03 0.02 0 2 2 0.41 0.18 0.13 0.05 2 3 1.30 0.92 0.35 0.25 2 4 2.01 1.55 1.47 1.19 HB Pick-up/Drop-off Trip Production Rates 2 5+ 2.57 2.28 2.11 2.06 HH Vehicles per Household Size 0 1 2 3+ 3 & 4 1 0.01 0.04 0.04 0.02 1 0.04 0.04 0.04 0.04 3 & 4 2 0.06 0.25 0.05 0.04 2 0.10 0.22 0.13 0.13 3 & 4 3 0.54 0.41 0.41 0.41 3 0.30 0.41 0.36 0.28 3 & 4 4 0.90 1.07 1.02 0.97 4 0.36 0.58 1.07 0.42 3 & 4 5+ 1.35 2.53 2.24 1.85 5+ 0.85 1.73 1.58 1.08

TABLE 7 Trip Attraction Rates and Non-Home-Based Trip Production Rates

Basic Retail Service Employment Households Employment Employment College K-12 Other Production Rate Models Non-Home-Based Work 0.07 0.47 1.78 1.86 0.93 0.93 Non-Home-Based Other 0.57 1.74 2.49 0.28 0.28 Attraction Rate Models Home-Based Work 1.42 1.64 1.23 1.23 1.23 Home-Based Work-Related 0.06 0.35 0.27 0.08 0.08 Home-Based Personal Business 1.25 4.17 Home-Based Social/Recreational 1.28 1.34 1.13 Home-Based School 3.30 9.25 Home-Based Pick-Up/Drop-Off 0.13 0.04 0.04 0.04 4.25 0.04 Non-Home-Based Work 0.11 0.32 2.36 1.85 0.79 0.79 Non-Home-Based Other 0.59 1.91 2.01 0.22 0.22

More information is available about regional patterns for home-based work (HBW) trips during the base year. In order to produce base year home-based work trip ends that reflect the observed patterns, the following changes are made as part of the base year balancing procedure:

 Total regional HBW attractions are adjusted to match the base year ratio of total regional HBW attractions to total regional HBW productions with the estimated ratio from the updated 2009 employment and workforce estimates (1.0063).  Total external HBW attractions are adjusted to match the base year ratio of total external HBW attractions to total regional HBW productions with the estimated ratio from the updated 2009 employment and workforce estimates (.0474).  Total external HBW productions are set equal to the difference between the grand total of HBW attractions and the regional HBW productions.

In addition, forecasts of future regional employment (the determinant of home-based work trip regional attractions) are available, so the estimates of future external HBW productions and attractions are less reliable than the estimates of future regional HBW productions and attractions. The model assumes that the number of external HBW productions will satisfy the forecasted employment within the region, so the HBW external productions are set equal to the difference between the total HBW attractions and the regional HBW productions.

CTPS 22 March 11, 2013

TRIP DISTRIBUTION

The trip distribution model performs the second step in the travel forecasting process. It combines the estimated trip productions and trip attractions prepared by the trip generation model (combining the HBW and HBWR purposes into a new HBW purpose) into:  an interregional vehicle trip table and an intraregional pick-up/drop-off vehicle trip table, to be used as input into the highway assignment model; and,  intraregional person trip tables to be used as inputs into the mode choice model.

The trip distribution model is made up of three components: a set of internal-external trip distribution models and two sets of intraregional trip distribution models (one for peak travel periods and the other for non-peak travel periods). An overview of the model is presented below.

Internal-External Trip Distribution

Internal-external trip distribution refers to a process in which all internal and external average weekday (AWD) trip ends (trip productions and attractions) are combined into trips using AWD highway impedances, but only the trips with one end in an internal zone and the other end in an external zone are retained. The resultant internal-external/external-internal trip tables are used as inputs to the highway assignment model. The remaining trip ends are used as inputs to the intraregional trip distribution model.

The model includes a separate process for each of seven trip purposes: home-based work, home- based personal business, home-based social/recreational, home-based school, home-based pick- up/drop-off, non-home-based work, and non-home-based other. The process undertaken for each purpose consists of the following five steps:

 Convert highway travel times from time period origin-destination format to AWD production-attraction format  Estimate and apply gamma functions to create an initial trip table estimate  Initiate a three-dimensional balancing process, adjusting the initial trip table to match trip productions, trip attractions, and a trip-length frequency distribution  Create internal-external/external-internal vehicular trip tables  Create intraregional person trip table productions and attractions

Each of these steps is described below.

Conversion of Highway Travel Times

Estimates of highway travel times are prepared using the highway assignment model on an origin-destination basis for each time period. In order to use these estimates with the trip productions and attractions from the trip generation model, the estimates from origin TAZ to destination TAZ and from destination TAZ to origin TAZ produced by the highway assignment are combined for each trip purpose based upon temporal directional factors developed for each trip purpose from the latest regional household travel survey.

CTPS 23 May 8, 2017

Estimation and Application of Gamma Functions

Interregional gamma functions are estimated using linear regression fitting to reflect the relationship between base-year highway travel time estimates and survey trip tables. These functions are used to provide an estimate of the number of trips within each cell of the trip table for a future scenario based upon the highway travel times for that future scenario.

The resultant trip table is referred to as the seed trip table. A trip length frequency distribution is imposed upon the seed trip table by dividing the table into classes of zone pairs. The zone pairs within each class connect a common pair of districts (forming an “interchange”) and fall within a designated range of trip lengths (or “class”). A separate gamma function is used for each interchange. The number of interchanges and classes used for each trip purpose is presented in Table 8.

TABLE 8 Number of Interchanges and Classes Used for Each Trip Purpose

Internal-External Intraregional Peak Intraregional Non-Peak Trip Purpose Interchanges Classes Interchanges Classes Interchanges Classes HBW 36 250 36 250 36 250 HBPB 36 250 34 233 36 247 HBSR 35 247 33 225 36 242 HBSC 24 229 16 216 16 225 HBPD 25 243 4 49 4 51 NHBW 36 250 36 250 36 249 NHBO 25 246 33 227 36 249

Three-Dimensional Balancing

The seed trip table is adjusted through an iterative process in order to match its subtotals as closely as possible to the estimated trip productions, trip attractions, and trip length frequency distribution. Each iteration consists of adjusting all the cells within a dimension (row, column, or class) by the factor needed to match the sum of that dimension to the estimated subtotal in that dimension (productions for row, attractions for column, trip length range trips for class) and then performing the same calculations for the other two dimensions. Since there is more confidence in trip production estimates than in the trip attraction or trip length frequency estimates, the iterative process ends with an exact matching of the trip table production totals to the input trip productions for each purpose.

Internal-External Trip Tables

The portions of the resultant trip table connecting external stations and regional TAZs are saved and adjusted for use in the highway assignment model. Vehicle occupancy data from the latest

CTPS 24 May 8, 2017

external travel survey are used to convert the person trips to vehicle trips. Temporal and directional factors from the latest external travel survey are then used to convert the trips from one matrix of AWD trips from production zone to attraction zone to four matrices of time period trips from origin zone to destination zone.

Intraregional Productions and Attractions

The portions of the resultant trip table connecting a pair of regional TAZs are summed by TAZ of production and TAZ of attraction for use in the Intraregional Trip Distribution Model. Data from the latest household travel survey are used to split these trip production and trip attraction files into peak period and non-peak period files.

Intraregional Trip Distribution (Peak and Non-Peak)

Intraregional trip distribution refers to a process in which all peak period and non-peak period intraregional trip ends are separately combined into trips using composite impedances from the mode choice model. The resultant peak and non-peak intraregional trip tables are used as inputs to the mode choice model and highway assignment model.

The model includes a separate process for each of seven trip purposes: home-based work, home- based personal business, home-based social/recreational, home-based school, home-based pick- up/drop-off, non-home-based work, and non-home-based other. Similar to the Internal-External Trip Distribution Model, the process undertaken for each purpose consists of the following three steps:

 Convert composite impedance estimates from time period to peak and non-peak format  Estimate and apply gamma functions to create an initial trip table estimate  Initiate a three-dimensional balancing process, adjusting the initial trip table to match trip productions, trip attractions, and a trip-length frequency distribution

The results of these steps are then processed to final form in the following two steps:

 Create pick-up/drop-off vehicular trip tables  Create intraregional person trip tables

The five steps are described below.

Conversion of Composite Impedances

Estimates of purpose-specific composite impedances are prepared using the mode choice model for origin-destination TAZ pairs for each time period. In order to use these with the intraregional trip productions and attractions from the Internal-External Trip Distribution Model, the composite impedance estimates produced by the mode choice model are adjusted for production- attraction TAZ pairs for each trip purpose by temporal factors for each trip purpose from the latest regional household travel survey.

CTPS 25 May 8, 2017

Estimation and Application of Gamma Functions

Intraregional gamma functions are estimated using linear regression fitting to reflect the relationship between base year composite impedance estimates (the combined utilities for all modes from the mode choice models) and survey trip tables. These functions are used to provide an estimate of the number of trips within each cell of the trip table for a future scenario based upon the composite impedances for that future scenario.

Three-Dimensional Balancing

The seed trip table is adjusted through an iterative process to match its subtotals as closely as possible to the estimated trip productions, trip attractions, and composite impedance range frequency distribution. This process is the same as the one used in the Internal-External Trip Distribution Model. Since there is more confidence in trip production estimates than in the trip attraction or trip length frequency estimates, the iterative process ends with an exact matching of the trip table production totals to the input trip productions for each purpose.

Pick-Up/Drop-Off Vehicular Trip Tables

Since all trips for the home-based pick-up/drop-off purpose are assumed to be vehicular trips, the resultant trip tables for that purpose are converted directly to vehicular trip tables so that they can be used in the highway assignment model. Vehicle occupancy data from the latest household travel survey are used to convert the person trips to vehicle trips. Temporal and directional factors from the latest household travel survey are then used to convert the trips from matrices of peak period and non-peak period trips from production zone to attraction zone to matrices of time period trips from origin zone to destination zone.

Intraregional Person Trip Tables

The resultant trip tables for the other purposes are then prepared. Data from the latest household travel survey are used to split these peak period and non-peak period files into person trip tables for each time period. These trip tables are then used as inputs to the mode choice model.

CTPS 26 May 8, 2017

MODE CHOICE

Overview

Mode choice is the third step in travel demand forecasting and in CTPS’s regional travel demand model. It is the process in which the trips from distribution are split between various available modes of the transportation network.

CTPS developed multinomial logit mode choice models by trip purpose using the latest Household Travel Survey data, travel impedances obtained from highway and transit networks, 1990 and 2000 U.S. census data, and a variety of other data sources. The mode choice models estimate modal splits for four trip purposes: HBW, Home-based Other (which includes HBPB and HBSR), HBSc, and NHB. These models have been calibrated and validated. The mode choice models are applied, by purpose, to the intraregional person trip tables that result from the trip distribution model.

The mode choice models split the trips for each purpose among six modes: 1) walk-access transit, 2) drive-access transit, 3) single-occupancy vehicles, 4) high-occupancy vehicles with two or more persons (high-occupancy vehicles with two persons only for the HBW trip purpose), 5) high-occupancy vehicles with three or more persons (for the HBW trip purpose only), and 6) a pure walk mode. The stations used in the execution of drive-access transit trips are identified using a special component of the mode choice model: a station choice model. Specific sub-mode selection (i.e., local bus, express bus, light rail, commuter rail, rapid transit, bus rapid transit, and boat) occurs during the transit assignment process.

The mode choice models estimate mode splits for intraregional trips only (trips contained within the model boundaries). They estimate mode shares for both inter-zonal trips (from one zone to another zone) and intra-zonal trips (from and to the same zone); however, intra-zonal trips are only split between the walk and auto modes.

Factors based upon the latest household travel survey are used to divide the trip tables produced by the trip distribution models into two trip tables: one for the trips made from production TAZ to attraction TAZ, the other for the trips made from attraction TAZ to production TAZ. The mode choice models are applied to these trip tables in two stages: first for the trips made from production TAZ to attraction TAZ (using the origin-destination input matrices), then for the trips made from attraction TAZ to production TAZ (using the inverse of the origin-destination input matrices).

Variables

The following are brief descriptions of the variables the mode choice models use to estimate mode splits:

Nest coefficient: Represents the degree of interactivity between the modes within the nest and other modes or nests. The value ranges between 0 and 1, with 1 indicating that switches to and

CTPS 27 May 8, 2017

from other modes are as likely as switches to and from modes within a nest. A value of 0 indicates there would be no switching between the nest modes and other modes.

In-vehicle travel time (IVTT): Represents time spent in the modal vehicle during a given trip.

Out-of-vehicle time: Includes all walk, boarding, and wait time.

Drive-access time: Represents driving time between a trip end and a transit station parking lot.

Terminal time: Represents the time it takes to travel between a vehicle and the trip origin or destination.

Fare: Represents the average transit fare, in dollars, a transit rider will pay to use the system. This is calculated on a modal basis by dividing the total revenue by ridership; this accounts for differences in “discounted” fare media used throughout the system and in the study area. Also included along with this average fare is one-half of any applicable parking costs (one-half because such costs are calculated on the basis of a one-way trip) at a transit station parking facility.

Auto cost: Represents auto operating and toll costs. Also included is one-half of any applicable non-transit parking costs (one-half because such costs are calculated on the basis of a one-way trip) on the street or in a parking facility. Also, for shared-ride modes, total auto costs are divided by the appropriate auto occupancy.

Household size: Represents the number of persons per household. This estimate is obtained from the trip generation model.

Vehicles/person: Represents the total number of vehicles per person in a household. Vehicles are estimated using the vehicle availability model described earlier.

Population density: Represents total population per acre of dry land in the TAZ.

Percent SOV origins/destinations: Represents the AM peak period single-occupant vehicle share of work trip ends within a TAZ, as computed by the home-based work mode choice model.

The Four Trip Purposes and the Station Choice Model

Home-Based Work Model

Home-based work (HBW) is the only trip purpose for which the mode choice models distinguish between two-person carpools (HOV2) and three-or-more-person carpools (HOV3+). The model specifications are shown in Table 9.

A transit nest is incorporated into the model on the basis that the decision to take transit over the other modes is made before selection of a particular transit mode. The transit coefficients are the same for both walk access (WAT) and drive access (DAT) transit and include coefficients for in-

CTPS 28 May 8, 2017

vehicle, initial wait, transfer wait, and total walk time. Drive-access time and production terminal times are included in DAT as one parameter.

The WAT fare includes the transit fare in dollars. For DAT, costs include the average transit fare and half of any parking cost. Population density by traffic zone, in people per acre, is included in walk-access transit, and it is positively correlated: the greater the density, the more likely a traveler is to choose this mode. The zones with high population densities also have more transit stops. Vehicles-per-worker is a socioeconomic input unique to this trip purpose for DAT. It is also positively correlated, since a higher vehicles per worker ratio increases the likelihood of a vehicle’s being available for a trip to a park-and-ride lot.

The auto times and cost coefficients are the same for the three auto modes. For HOV2 and HOV3+ the auto cost is divided by the average vehicle occupancies to reflect the sharing of costs between vehicle occupants. Household size is included as a positively correlated variable for the shared-ride modes and has a somewhat greater impact for HOV3+ than HOV2.

Home-Based Other Model

The home-based other (HBO) mode choice model combines the home-based shopping and home-based recreational trip tables output from the trip distribution process into a single HBO trip table. The model specifications are shown in Table 10. The model is similar to the HBW mode choice model, except for the following three differences. First, since there is only one shared-ride mode, HOV2+, household size is only a parameter for this one mode. Second, the vehicles-per-person-in-a-household is used, as opposed to vehicles-per-worker. Finally, a distance dummy equal to one if the trip distance is less than a mile and zero otherwise is added to the walk mode. This reflects the fact that people taking short trips for this purpose are more likely to walk than choose another mode.

Non-Home-Based Model

The non-home-based (NHB) model splits work trips and non-work trips. The model specifications are shown in Table 11. There is a work dummy variable in the two auto modes which is equal to one if the trip is a non-home-based work trip and zero otherwise. The coefficient is positive for SOV and negative for HOV, indicating that the SOV mode is more likely on work-related trips than on non-work trips. The percentage of trips attracted to the origin and destination zones that is SOV is a variable in the drive-alone mode. The percentage is taken from the results of the AM peak period HBW mode choice model and is positively correlated. Finally, the distance dummy in the walk mode is equal to one if the distance is less than a mile. It has a positive coefficient.

Home-Based School Model

The home-based school (HBSC) model was re-estimated and restructured in 2004 to allow for compatibility of the HBSC purpose with the Federal Transit Administration’s Summit program. The previous HBSC model had one nest comprising all motorized modes. The revised HBSC model has two nests, transit and highway. The revised HBSC model specifications are shown in Table 12.

CTPS 29 May 8, 2017

Station Choice Model

The final part of the mode choice model is the assignment of drive-access transit trips to transit stations in the station choice model. This model uses estimates of highway travel times and costs from the highway assignment model, estimates of transit impedances from the walk-access transit assignment model, and estimated transit parking lot capacities to distribute drive-access transit trips among the transit stations with parking lots. The model also estimates the impedances associated with the drive-access transit trips between each TAZ pair and, if parking at the transit parking lots is constrained, reassigns demand for full parking lots to other parking lots or to other modes of transportation.

The probability of selecting a station is determined by the combination of utilities for the auto and transit legs of the drive-access transit trip. The utility of the auto leg (Uik) is a combination of the auto travel time between production TAZ i and transit station k (ATTik) and the parking capacity at transit station k (PCk).

Uik = -.125 ATTik + .00025 PCk

The utility of the transit leg (Ukj) is a function of the composite impedance used in transit path selection, which includes transit in-vehicle travel time (ITTkj), boarding time (BTkj), waiting time (WtTkj), fare (Fkj), and walk time (WkTkj) accumulated between station k and attraction TAZ j.

Ukj = -.05 * (ITTkj + BTkj + Fkj + (2 * (WtTkj + WkTkj)))

The auto leg utilities are used to identify the five most likely stations to be used for each production TAZ, the combined utilities are used to estimate the probabilities of selecting each of those stations for each pair of TAZs, and the trips are assigned to transit stations. If transit parking is constrained to capacity, some trips may not be possible since the parking demand exceeds the capacity at the station, For those trips, the auto leg utilities are re-estimated to identify the five most likely stations with available parking capacity, and the trips are assigned to transit stations based upon the combined utilities. Trips which are still not assignable due to inadequate parking capacity are then switched to the walk-access transit, single-occupancy vehicle, or high-occupancy vehicle mode in the same proportion of other trips of the same purpose between the same pair of TAZs.

CTPS 30 May 8, 2017

TABLE 9 Home-Based-Work Mode Choice Model Specifications

Impedance Variable Socioeconomic Variable Nest Terminal Walk Initial Transfer Auto Boarding Fare Auto Population Vehicles/ HH Coeff IVTT Time Time Wait Wait Access Time ($) Cost ($) Density Worker Size Drive-Alone Top Level 1 -0.05466 -0.292 -0.32 Application Level -0.05466 -0.292 -0.32 Ratio to IVTT ($/hr) 1 5.34211 $ 10.25 HOV2 Top Level 1 -0.05466 -0.292 -0.32 0.07322 Application Level -0.05466 -0.292 -0.32 0.07322 Ratio to IVTT ($/hr) 1 5.34211 $ 10.25 -1.33955 HOV3+ Top Level 1 -0.05466 -0.292 -0.32 0.2168 Application Level -0.05466 -0.292 -0.32 0.2168 Ratio to IVTT ($/hr) 1 5.34211 $ 10.25 -3.96634 Walk Top Level 1 -0.1007 Application Level -0.1007 Ratio to IVTT ($/hr) Walk-Access Transit Top Level 0.6791 -0.05466 -0.1007 -0.11292 -0.11292 -0.05466 -0.32 0.01889 Application Level -0.08049 -0.14828 -0.16628 -0.16628 -0.08049 -0.47121 0.02781 Ratio to IVTT ($/hr) 1 1.8423 2.06593 2.06593 1 $ 10.25 -0.34551 Drive-Access Transit Top Level 0.6791 -0.05466 -0.292 -0.1007 -0.11292 -0.11292 -0.13665 -0.05466 -0.32 -0.32 0.2897 Application Level -0.08049 -0.42998 -0.14828 -0.16628 -0.16628 -0.20122 -0.08049 -0.47121 -0.47121 0.4266 Ratio to IVTT ($/hr) 1 5.34211 1.8423 2.06593 2.06593 2.5 1 $ 10.25 $ 10.25 -5.30011

TABLE 10 Home-Based-Other Mode Choice Model Specifications

Impedance Variable Socioeconomic Variable Nest Terminal Walk Initial Transfer Auto Boarding Fare Auto Population Vehicles/ HH Distance Coeff IVTT Time Time Wait Wait Access Time ($) Cost ($) Density Worker Size Dummy Drive-Alone Top Level 1 -0.01965 -0.2308 -0.22378 Application Level -0.01965 -0.2308 -0.22378 Ratio to IVTT ($/hr) 1 11.7463 $ 5.27 HOV2+ Top Level 1 -0.01965 -0.2308 -0.22378 0.1976 Application Level -0.01965 -0.2308 -0.22378 0.1976 Ratio to IVTT ($/hr) 1 11.7463 $ 5.27 -10.0566 Walk Top Level 1 -0.05895 0.9005 Application Level -0.05895 0.9005 Ratio to IVTT ($/hr) -15.2757 Walk-Access Transit Top Level 0.3722 -0.01965 -0.05895 -0.05895 -0.05895 -0.01965 -0.22378 0.00883 Application Level -0.05279 -0.15838 -0.15838 -0.15838 -0.05279 -0.60123 0.02373 Ratio to IVTT ($/hr) 1 3.0002 3.0002 3.0002 1 $ 5.27 -0.44951 Drive-Access Transit Top Level 0.3722 -0.01965 -0.2308 -0.05895 -0.05895 -0.05895 -0.04912 -0.01965 -0.22378 -0.22378 0.71239 Application Level -0.05279 -0.6201 -0.15838 -0.15838 -0.15838 -0.13198 -0.05279 -0.60123 -0.60123 1.914 Ratio to IVTT ($/hr) 1 11.7463 3.0002 3.0002 3.0002 2.5 1 $ 5.27 $ 5.27 -36.2564

TABLE 11 Non-Home-Based-Work Mode Choice Model Specifications

Impedance Variable Socioeconomic Variable Nest Terminal Walk Initial Transfer Auto Boarding Fare Auto Work Distance Percent Coefficient IVTT Time Time Wait Wait Access Time ($) Cost ($) Dummy Dummy SOV Drive-Alone Top Level 1 -0.03022 -0.3197 -0.1817 0.1926 0.00885 Application Level -0.03022 -0.3197 -0.1817 0.1926 0.00885 Ratio to IVTT ($/hr) 1 10.5791 $ 9.98 -6.37326 -0.29295 HOV2+ Top Level 1 -0.03022 -0.3197 -0.1817 -0.7627 Application Level -0.03022 -0.3197 -0.1817 -0.7627 Ratio to IVTT ($/hr) 1 10.5791 $ 9.98 25.2383 Walk Top Level 1 -0.07525 0.493 Application Level -0.07525 0.493 Ratio to IVTT ($/hr) -6.5515 Walk-Access Transit Top Level 1 -0.03022 -0.07525 -0.08333 -0.08333 -0.03022 -0.1817 Application Level -0.03022 -0.07525 -0.08333 -0.08333 -0.03022 -0.1817 Ratio to IVTT ($/hr) 1 2.49007 2.75745 2.75745 1 $ 9.98 Drive-Access Transit Top Level 1 -0.03022 -0.3197 -0.07525 -0.08333 -0.08333 -0.07555 -0.03022 -0.1817 -0.1817 Application Level -0.03022 -0.3197 -0.07525 -0.08333 -0.08333 -0.07555 -0.03022 -0.1817 -0.1817 Ratio to IVTT ($/hr) 1 10.5791 2.49007 2.75745 2.75745 2.5 1 $ 9.98 $ 9.98

TABLE 12 Home-Based-School Mode Choice Model Specifications

Impedance Variable Nest Terminal Walk Wait Drive-Access Fare Auto Population Coefficient IVTT Time Time Time Time ($) Cost ($) Density Drive-Alone Top Level 0.5559 -0.0305 -0.0904 -0.1803 Application Level -0.0548 -0.1626 -0.3244 Ratio to IVTT ($/hr) 1.0000 2.9672 $10.14 HOV2+ Top Level 0.5559 -0.0305 -0.0904 -0.1803 Application Level -0.0548 -0.1626 -0.3244 Ratio to IVTT ($/hr) 1.0000 2.9672 $10.14 Walk Top Level 1 -0.0791 Application Level -0.0791 Ratio to IVTT ($/hr) Walk-Access Transit Top Level 0.5559 -0.0305 -0.0791 -0.0791 -0.1803 0.0150 Application Level -0.0548 -0.1423 -0.1423 -0.3244 0.0270 Ratio to IVTT ($/hr) 1.0000 2.5967 2.5967 $10.14 -0.4927 Drive-Access Transit Top Level 0.5559 -0.0305 -0.0904 -0.0791 -0.0791 -0.0762 -0.1803 -0.1803 0.0150 Application Level -0.0548 -0.1626 -0.1423 -0.1423 -0.1371 -0.3244 -0.3244 0.0270 Ratio to IVTT ($/hr) 1.0000 2.9672 2.5967 2.5967 2.5018 $10.14 $10.14 -0.4927

TRIP ASSIGNMENT

Trip assignment is the fourth step in the travel demand forecasting process and in CTPS’s regional travel demand model. Trip assignment is the process by which each trip in the trip tables resulting from the mode choice model is assigned to a specific submode (for example, bus or rapid transit) and a specific route. The CTPS model uses two distinct assignment procedures, one for the transit trips and one covering the highway modes.

Highway Assignment Routine

The highway assignment implemented in EMME is an equilibrium assignment. The fundamental assumptions underlying such an assignment procedure are that each user of the highway network knows all of the network times, costs, and distances and will choose the route that he or she perceives to be the best. The assignment is an aggregate assignment in that traffic volumes on any given link are an aggregate number, as opposed to being associated with a specific trip. There are several inputs used by the EMME equilibrium assignment procedure. The key inputs are the highway demand matrices, the volume delay function, and the highway network:

 Highway demand matrices

The demand matrices that the highway assignment procedure uses as an input are the demand matrices that result from the mode choice and distribution models and other sources. These are origin-destination matrices of single-occupancy vehicles, trucks, taxis, internal-external/external-internal trips, through trips, and high-occupancy vehicles.

To prepare the mode choice trip tables for use in highway assignments, it is necessary to convert person trips to vehicle trips by applying vehicle occupancy factors for HOV modes. These occupancy factors, presented below, vary by trip purpose and are based upon the latest household travel survey.

Home-based work trips HOV2: 2 persons/vehicle HOV3+: 3.373 persons/vehicle Home-based other trips HOV2+: 2.404 persons/vehicle Home-based school trips HOV2+: 2.788 persons/vehicle Non-home-based trips HOV2+: 2.385 persons/vehicle

In addition to manipulating the output matrices from mode choice, it is necessary to bring in vehicle trip tables produced outside of the mode choice process. These vehicle trip tables are:

o External Through – This matrix consists of trips that pass through the study area without stopping and hence are exogenous to the travel model. The trips were estimated from the latest external travel survey, 2000 Census Journey to Work data, and traffic counts. o Taxi – The taxi vehicle trip table was originally developed from a 1993 survey and has since been revised several times based upon a factoring process.

CTPS 35 May 8, 2017

o Logan Airport SOV and HOV – This trip table is developed from a separate Logan Airport Passenger Mode Choice Model, which was developed based on a 2007 Massachusetts Port Authority survey. o Drive-Access Transit Auto Access – DAT trips are determined through the station choice model, which is a part of the mode choice process. Each DAT trip requires a vehicle access trip. o Interregional SOV and HOV – The interregional vehicle trip tables are generated through the interregional trip distribution model. o Pick-Up/Drop-Off SOV and HOV – The pick-up/drop-off (PUDO) tables, produced by the interregional trip distribution model, cover those trips in which a person is dropped off at his or her destination (not an intermediate parking lot). o Truck – The truck trip tables, which cover commercial truck trips within the region, are produced from the CTPS tour-based truck trip model.

This integrated behavioral tour-based model was developed from the Vehicle Inventory and Use Survey, Massachusetts Registry of Motor Vehicle files, and existing truck trip generation rates. These existing sources were supplemented by specifically-focused telephone and travel-intercept surveys and video data capture which provided information used to quantify key behavioral relationships.

The truck travel forecasting model incorporates several new factors into the truck trip tables used for highway assignments. With these changes, the truck model produces truck travel estimates which are sensitive to demographic changes and which are consistent with the roadway network and observed truck travel patterns and operating characteristics. These innovative features can be characterized into three major elements:

o BEHAVIORAL. The model is based upon functional usage categories which capture relatively homogeneous patterns of truck operation and are tied to regional socioeconomic characteristics.

o TOUR-BASED. The model differentiates between truck trip tour ends and their intermediate starts and stops in order to impose a tour-like form on the pairing of truck trip ends. This, as a result, distributes truck trip ends appropriately between truck garage sites and truck starts and stops along delivery routes.

o INTEGRATED. The model is sensitive to changes in a specific set of interacting variables that are internally consistent and externally constrained. The variables include sector employment and population, truck ownership and operational characteristics, highway network truck restrictions, link truck volume counts, time-of-day, and intra-regional and inter-regional truck travel demand. Truck travel volumes are estimated as a function of the population and employment by type of business in each TAZ.

The underlying premise of the modeling approach is that overall truck travel demand can be divided into nine relatively identifiable and homogenous functional usage categories.

CTPS 36 May 8, 2017

Each of the nine is comprised of relatively similar travel characteristics as measured in such variables as tours per day, trips per tour, and trip length.

These nine distinct categories are:

o Tankers – distinct body type, many carry hazardous materials. o Household Goods - perform distinct service, as they move the belongings (not products) of their clients o Less-than-truckload/Truckload – commercial carriers transporting wide varieties of goods o Food and Warehouse Distribution – distributing goods (sometimes non-durable, so time-sensitive) to retail outlets. o Intermodal – picking up or delivering goods also carried by rail or boat. o Package – distinct service type, many stops per tour. o Heavy – large vehicles, most subject to weight limitations. o Retail – delivering goods to end users. o Pickup/Van – small vehicles, least subject to restrictions.

Within the modeling process, a series of relationships were established among firm employment, firm truck ownership/usage, and truck type and usage category. These relationships are expressed in terms of both FHWA physical vehicle classes and usage categories. This correspondence made it possible to validate and, where necessary, to adjust our travel demand matrices through use of trip table estimation techniques. In this way, we adjusted our initial demand levels for four time periods to observed truck volumes from counts conducted on links of the highway system.

 Volume-delay function

The function used in the highway assignment procedure is a volume-delay function, which, when applied in the context of a highway assignment, changes the speeds users of the network experience based upon the volumes on the network. The volume-delay functions employed in the CTPS regional model are variations on the so-called Bureau of Public Roads (BPR) function. Developed by its now defunct namesake, the BPR function is a widely used and validated volume-delay function that is parabolic in shape and takes the form:

Congested Speed = (Free-Flow Speed)/(1 + 0.83*[Volume/Capacity]b) The form of this equation used for expressways uses 5.5 as the value of b, the exponent for the volume-capacity ratio. The form of the equation used for all other roadways in the network uses 2.7 as the exponent.

The CTPS regional model is segmented by time periods. For each time period, the BPR function is altered to reflect the number of hours in that period.

 Highway network

CTPS 37 May 8, 2017

The highway network is an abstract digital representation of the real highway network in eastern Massachusetts. For future-year scenarios, the highway network depicts roadway links that are planned in addition to the existing highway network. The base-year highway network is a depiction of the eastern Massachusetts highway network as it existed in the year 2012. The highway network in the base and future years includes information about number of lanes, free-flow speeds, and capacity (in vehicles per lane per hour). Freeways typically have a free-flow speed of 60 miles per hour, are three lanes, and have a capacity of 1,950 vehicles per lane per hour. Smaller arterials typically have a free-flow speed of 30 to 45 miles per hour, are coded as having one or two lanes, and have a capacity of 900 to 1,000 vehicles per lane per hour. Such parameters are consistent with widely accepted traffic engineering principles and the Transportation Research Board’s Highway Capacity Manual.

The highway assignment procedure performs a multi-class, generalized cost equilibrium auto assignment. The multi-class assignment runs an assignment for each time period for the demand matrices of five modes, SOV, HOV, pickup trucks and vans, hazardous material trucks, and other commercial trucks, from the total vehicle trip tables for each class. Tolls affect the assignment and are stored on the network.

The highway assignment procedure is iterative in that the assignment is calculated repeatedly, in order to mathematically optimize assignment results. Three criteria are used to determine how many iterations of the assignment procedure are used.

First, the relative gap is an estimate of the difference between the current assignment and a perfect equilibrium assignment, in which all paths used for a given origin-destination pair would have exactly the same time. The default relative gap is 0.5%, but CTPS employs 0.01% so that a more accurate assignment will result.

Another criterion for when to stop the iterations is the normalized gap (or trip time differential), which is the difference between the mean trip time of the current assignment and the mean minimal trip time. The mean trip time is the average trip time on the paths used in the previous iteration; the mean minimal trip time is the average trip time computed using the shortest paths of the current iteration. Again, a minimum level is selected, 0.01 minutes, in order for the designated number of iterations to be carried out.

If neither of these criteria is met, the CTPS regional model highway assignment procedure is set to stop after running through 50 iterations.

Transit Assignment Routine

The transit assignment used in EMME is a multi-path assignment based on the calculation of optimal transit strategies for system users. A transit strategy is roughly analogous to a path in highway assignment. The transit assignment allows for users of the transit system switching within the transit network between various available transit services in order to reach their destination. In basic terms, the transit assignment algorithm identifies the optimal service or services at each node in the transit network for each origin and destination node pair. This

CTPS 38 May 8, 2017

algorithm is repeated for all nodes, starting with the destination node and culminating at the origin node.

Like the highway assignment procedure, the transit assignment procedure utilizes several key inputs to estimate a transit assignment. Three of the key inputs are the transit demand matrices, the transit functions, and the transit network:

 Transit demand matrices

The transit demand matrices are just that, matrices of trips that have been split into the transit mode because the utility of their trip suggests that transit is the most attractive mode choice for their particular origin-destination pair. These trip tables come from three sources: o Walk-access transit trip tables from the mode choice model o Drive-access transit trip tables from the station choice model o Logan transit trip tables from the Logan Airport Passenger Mode Choice Model

 Functions

The function used in the transit assignment procedure depicts the relative levels of attractiveness among the numerous paths available in the eastern Massachusetts transit network for each pair of TAZs. Costs are translated to time assuming a value of time of $12 per hour (using 1991 dollars) and doubling the out-of-vehicle time (walk and wait times) before adding it to in-vehicle time.

 Transit network

The transit network is an abstract digital representation of the real transit network in eastern Massachusetts. For future-year scenarios, the transit network depicts transit links that are planned in addition to the existing transit network. The base-year transit network is a depiction of the eastern Massachusetts transit network as it existed in Spring 2012. The transit network includes every commuter rail line, rapid transit line, bus route, bus rapid transit route, and ferry route and many free shuttle routes in eastern Massachusetts. The bus routes run on the highway network, and their run times are influenced by roadway traffic congestion. Among other things, the transit network in the base and future years includes estimated vehicle headways, wait times, transit run times, and fares for each line. The assignment algorithm takes into consideration all of these elements in calculating a transit assignment.

Additionally, the transit network represents and accounts for park-and-ride facilities. Park-and-ride nodes provide connections between the highway and transit networks via a walk link. As a result, drive-access transit trips use both the highway and the transit networks.

The transit network also includes an extensive set of walk-access and transfer links. All these links assume a walk speed of 3 miles per hour.

CTPS 39 May 8, 2017

Walk-access links are an abstract representation of all of the walking routes transit users utilize in eastern Massachusetts to access the transit system. In other words, they are an aggregate abstraction of the sidewalks, roadways, backyards, driveways, and shortcuts people use to walk to the transit system.

The walk-access estimation process is an automated process that involves three steps. The first step builds paths and distances on a walk network roadway geographic information system (GIS) coverage that is created from the most recent statewide digital line graph (DLG) coverage of the roadway network. The roadways that are unsuitable for walking within the study area are then cut from that coverage. The path building and distance skimming between transit stops and zones is calculated on this coverage. The distances between the transit stops and stations active under the scenario under study and each TAZ are then calculated from this coverage. Up to two walk links are created between each TAZ and the stations and stops on each transit line, with no links over one mile. Transfer links are created to connect all stations and stops within a quarter-mile walk of each other.

 Fare Coding

Average fares are used for coding in the EMME network. Each transit sub-mode (boat, bus rapid transit, rapid transit, bus, commuter rail, and shuttle) is assigned a boarding fare that is placed on the walk access and transfer links serving the nodes that serve the stations and stops for that sub-mode. Additional zone fares are represented as segment fares placed on the transit links crossing the fare zone boundaries. In addition, park-and- ride parking charges are coded onto the walk links that connect the park-and-ride nodes to the transit station and stop nodes.

Fares are translated into time for influencing path selection by assuming a value of time of $12 (in 1991 dollars) per hour. Although fares are expressed in minutes to allow them to be included in the impedances that influence path selection, they are kept separate from travel times for input into the mode choice model.

CTPS 40 May 8, 2017

Attachment B

CTPS CENTRAL TRANSPORTATION PLANNING STAFF CENTRAL TRANSPORTATION Staff to the Boston Region Metropolitan Planning Organization PLANNING STAFF

TECHNICAL MEMORANDUM

DATE: May 6, 2017 TO: Allston I-90 Interchange Task Force FROM: K. Grace King, Central Transportation Planning Staff RE: Allston Interchange Traffic Study: Model Validation Process

This memorandum describes the process used to validate the Boston Region Metropolitan Planning Organization’s regional travel demand model for use in the Allston Interchange Traffic Study. After checking and editing the transportation network of the regional model and making sure it accurately represented the base-year condition in the study area, the Central Transportation Planning Staff (CTPS) performed the standard four steps of the travel demand modeling process: trip generation, trip distribution, mode choice, and assignment. Four model outputs—travel demand, trip flows, highway traffic speeds, and mode shares—were the focus of the validation efforts. The following sections describe each of these elements in detail.

1 TRAVEL DEMAND This section discusses the validation processes for highway traffic volumes and transit ridership in the study area. CTPS staff tried to establish benchmarks for specific empirical data that could be matched to the model results, and succeeded in most instances. These benchmarks focused on the study corridor’s transit and highway measures.

1.1 Highway Traffic Volumes The validation efforts for highway traffic volumes focused on local roadways in the study area and on the Interstate 90 (Massachusetts Turnpike) Allston ramps and nearby mainline. The study team collected local roadway counts in 15- minute intervals on December 11, 2013, during the AM peak period—6:00 AM to 9:00 AM—and the PM peak period—4:00 PM to 7:00 PM—at various intersections in the study area (Figure1). Volumes on all of the approaches to these intersections were recorded. Since the PM peak period in the regional model is defined as 3:00 PM to 6:00 PM, and not 4:00 PM to 7:00 PM, the collected PM counts were adjusted to be aligned with the model’s PM period. At the ramps and the nearby Turnpike mainline near the study area, 24-hour traffic volumes were collected in 15-minute intervals on December 18, 19, and 20, in 2013. Average weekday daily traffic (AWDT) counts were calculated by using the

State Transportation Building • Ten Park Plaza, Suite 2150 • Boston, MA 02116-3968 • (857) 702-3700 • Fax (617) 570-9192 • TTY (617) 570-9193 • [email protected] FIGURE 1 Allston Interchange Traffic Study CTPS Locations of Local Roadway Counts Allston Interchange Traffic Study: Model Validation Process May 6, 2017

mean average from these counts. The final counts used for validation integrated these AWDT counts with CTPS’s established 2010 directional and temporal balanced highway counts for these locations.

The highway validation process focused on matching the assigned volumes on the roadway links to count data, but not to turning movements at the selected intersections. Turning movements were not included as part of this validation process because of the historical difficulties associated with achieving such detailed accuracy using regional travel demand models.

Two validation benchmarks were sought for the roadway volumes. The validation benchmark for the total assigned model volumes required the modeled volumes to be within 10-15 percent of the total highway counts in the study corridor; within 15 percent is considered acceptable, but within 10 percent is ideal. As seen in Tables 1 and 2, the collective differences between the model results and the counts in the study area were 12 percent in the AM peak period, 8 percent in the PM peak period, and 8 percent on a daily basis. Because counts for local streets and collector roads were only available for the peak periods and not on a daily basis, the data for those roadways were not used for the benchmark validation. Another validation benchmark required that the root mean square error (RMSE) between the model volumes and the counts for all locations to be less than or equal to 30 percent. The RMSE for the model results and the counts in the study area was 23 percent for the AM peak period, 21 percent for the PM peak period, and 18 percent for the entire day. Therefore, the highway validation targets were readily met in the study area for both peak periods and on a daily basis, although there was more variability on ramps, local streets, and collector roads than on highways. Tables 1 and 2 display the comparisons mentioned in this paragraph.

TABLE 1 Comparison of Modeled Highway Volumes and Observed Counts for AM and PM Peak Periods AM AM PM PM Count Location Count Model Count Model On-ramps and off-ramps 33,667 37,991 37,857 40,689 Turnpike 110,850 121,644 115,700 127,310 Storrow Drive 63,400 67,060 72,400 73,545 Local streets and collector roads 41,919 52,031 53,617 60,688 Total 249,836 278,725 279,574 302,232 Difference 28,889 22,659 Percentage difference 12% 8% RMSE 23% 21%

Page 3 of 9 Allston Interchange Traffic Study: Model Validation Process May 6, 2017

TABLE 2 Comparison of Daily Modeled Highway Volumes and Observed Counts

Count Location Daily Count Daily Model On-ramps and off-ramps 172,871 193,818 Turnpike 535,500 593,707 Storrow Drive 347,000 352,024 Totals 1,055,371 1,139,548 Difference 84,177 Percentage difference 8% RMSE 18%

1.2 Transit Ridership Transit validation efforts focused on transit services in and nearby the study area. These efforts involved matching modeled daily route ridership data for MBTA bus Routes 64, 66, 70/70A, and 57/57A to empirical data. Other data used in these matching efforts included daily station boardings at Auburndale, West Newton, and Newtonville commuter rail stations and boardings at Harvard and Central Stations on the Red Line. Transit counts were taken from the MBTA’s 2014 Ridership and Service Statistics report and CTPS’s 2012 Commuter Rail Passenger Count Results.

The validation benchmark for the total daily boardings at key stations and the bus routes in the study area entailed that the modeled data for boardings were to be within 15 percent of the observed boardings. Table 3 shows the comparison of the observed counts and the modeled results. While the percentage differences at West Newton Station and Newtonville Station exceed 15 percent, the absolute differences were small (223 and 46).

TABLE 3 Comparison of Modeled and Observed Daily Transit Boardings

Model Daily 2012 Percentage Transit Stations and Routes Counts Results Difference Auburndale 183 183 0% West Newton 173 396 129% Newtonville 289 335 16% Central Square 16,525 18,510 12% Harvard Square 23,199 20,438 -12% MBTA Bus Routes (64, 66, 70/70A, 57/57A) 31,259 28,230 -10%

Page 4 of 9 Allston Interchange Traffic Study: Model Validation Process May 6, 2017

2 TRIP FLOWS Validation of the modeled auto trip flows focused on matching the origins and destinations of auto trips in the study area to observed counts. The origin and destination information of the auto trips on the Allston Interchange ramps in the month of October 2013 were purchased from AirSage, a company that utilizes cellular signals to capture the locations and movements of mobile devices. These data (captured signals on roadways) are used as proxies for vehicle movements on roadways. CTPS identified a 2,500-meter square region in the study area for AirSage to use as the target zone (Figure 2). All signals from active mobile devices (text messages, streaming data, and phone calls) in the target zone were captured. The origins and destinations were identified by capturing the locations where the mobile devices were stationary (a duration of at least 500 seconds at a specific location).

The resulting origin and destination findings from AirSage are shown in Figure 3. As expected, the majority of origins and destinations for ramp users are located in areas close to the Turnpike mainline and the ramps, with the greatest concentrations clustered around the ramps.

The modeled trip origins and destinations of the Allston Interchange ramp users were statistically compared to the data from AirSage using the R2 coefficient of correlation, a statistical measure that provides information about the “goodness of fit” of a model. An R2 of one indicates that a model perfectly fits the observed data. The correlations between the AirSage findings and the model results on trip origins and destinations of ramp users were very good: 0.79 for the AM peak period, 0.78 for the PM peak period, and 0.79 for the entire day.

3 HIGHWAY SPEEDS Turnpike speed data for 2012 were purchased from INRIX, a company that also utilizes mobile devices to track traffic speeds on roadways. Modeled average speeds for the roughly six-mile stretch of the Turnpike between the West Newton Interchange (Exit 16) and Massachusetts Avenue during the AM and PM peak periods were compared with INRIX’s empirical data. Table 4 displays this comparison. The modeled speeds for each peak period are within 30 percent of the observed INRIX period speeds. Modeled PM period speeds for both directions match the observed speeds rather closely. Although the modeled AM eastbound speed diverges more from the observed speed than the other modeled directional speeds, the fact that the AM eastbound modeled speed is less than, and not greater than, the observed speed means that the base-year model has more congestion on the Turnpike in the AM peak direction than presently exists. This conservative estimate for a highway-centered project will

Page 5 of 9 FIGURE 2 Allston Interchange Traffic Study CTPS AirSage Data Target Area FIGURE 3 Allston Interchange Traffic Study CTPS Daily Allston Interchange Ramp Users’ Trip End Locations Allston Interchange Traffic Study: Model Validation Process May 6, 2017

translate into slower speeds and greater congestion in the future, given no capacity increases.

TABLE 4 Comparison of Modeled and Observed Turnpike Speeds (mph)

Model Model INRIX AM AM INRIX PM PM MassPike Average Average Percentage Average Average Percentage Direction Speed Speed Difference Speed Speed Difference Eastbound 51.0 36.8 -27.8% 54.8 53.8 -1.8% Westbound 58.7 51.8 -11.8% 43.8 46.2 5.5%

4 MODE SHARES The validation of model mode shares focused on best representing the existing tripmaking modal characteristics associated with the study area. CTPS used estimates of the mode shares in the study area from the 2010 Census Journey- to-Work data and used them to calibrate the 2012 base-year model. The best modeled analogue for the Journey-to-Work data is Home-Based Work trips. Table 5 illustrates that the modeled modal shares of trips in the study area for this trip purpose are within two percent of the Journey-to-Work data. There are no available observed estimates of the mode shares for the other trip purpose types (i.e. Home-Based Other, Home-Based School, and Non-Home- Based). It is interesting to note that the mode shares for Home-Based Other trips are within 10 percent of the Journey-to-Work data, as are the overall daily modeled mode shares. TABLE 5 Comparison of Mode Shares from Census Data and Base-Year Model

Census Model Model Model Model Journey- Home- Home- Home- Non- to-Work Based Based Based Home Model Data Work Other School Based Total TRANSIT 25% 24% 16% 25% 10% 17% AUTO 59% 61% 63% 24% 71% 57% WALK 16% 15% 21% 50% 19% 26% Total 100% 100% 100% 100% 100% 100%

Page 8 of 9 Allston Interchange Traffic Study: Model Validation Process May 6, 2017

4 CONCLUSION The regional travel demand model, after validated in terms of travel demand, trip flows, highway speeds, and mode shares, was deemed satisfactory for use in the Allston Interchange Traffic Study future-year demand-forecasting work. The results of the model runs were utilized as the inputs for further detailed analyses for the study.

KGK/kgk

Page 9 of 9 Attachment C

CTPS CENTRAL TRANSPORTATION PLANNING STAFF CENTRAL TRANSPORTATION Staff to the Boston Region Metropolitan Planning Organization PLANNING STAFF

TECHNICAL MEMORANDUM

DATE: March 27, 2017 TO: Allston Interchange Project Team FROM: K. Grace King, Central Transportation Planning Staff RE: Allston Interchange Traffic Study: Transit Network Assumptions

This memorandum describes the future-year transit network assumptions that Central Transportation Planning Staff (CTPS) used in the modeling process for the Allston Interchange Traffic Study.

1 TRANSIT NETWORKS FOR NO-BUILD The background transit network assumptions are based on the Boston Region Metropolitan Planning Organization’s Long-Range Transportation Plan (LRTP), Charting Progress to 2040, which was amended in August 2016. CTPS modified the transit network in the model to incorporate the projects listed in the LRTP for No-Build scenarios representing the years 2025 and 2040. The only transit service that is not reflected in the modified transit network for the No-Build scenarios is the Worcester/Framingham commuter rail service. Vanasse Hangen Brustlin (VHB) prepared the service plan for this line; in the No-Build scenarios, the trains do not stop at West Station in Allston.

Table 1 contains the list of transit projects. For more details about the projects programmed in the LRTP, go to http://www.ctps.org/data/pdf/plans/lrtp/charting/2040_LRTP_Chapter5_final.pdf and http://www.ctps.org/data/pdf/plans/lrtp/charting/2040_LRTP_Ammendment_0825 16.pdf.

State Transportation Building • Ten Park Plaza, Suite 2150 • Boston, MA 02116-3968 • (857) 702-3700 • Fax (617) 570-9192 • TTY (617) 570-9193 • [email protected] Allston Interchange Traffic Study March 27, 2017

TABLE 1 Transit Projects in the 2025 and 2040 No-Build Scenario

Urban Ring Crosstown Bus Service Worcester Commuter Rail (partial service) South Station Transportation Center (long distance bus station) Acela Train/Electrification of the Northeast Corridor Newburyport Commuter Rail Service Old Colony Commuter Rail Service (two lines) Additional MBTA Park-and-Ride Spaces Route 128 Amtrak/Commuter Rail Station Grafton Station on the Worcester Line Hingham Ferry Improved service on the Haverhill Commuter Rail Line Salem-Boston Express Bus Service North Station Improvements Blue Line Modernization (Bowden to remain open with six car trains) Additional Park-and-Ride spaces Worcester Commuter Rail (full service, including new stations) Silver Line – Transitway, Phase 2 Silver Line – Washington Street, Phase 1 Mattapan Line Refurbishment Silver Line to Airport Industriplex Center in Woburn New Commuter Rail Station at JFK/UMass Station Greenbush Commuter Rail Service Peabody Express to Logan, and Logan Express from Anderson Fairmount Line Improvements – Phase I (state of good repair) Mishawum Station open for outbound service 7:07 AM, 7:49 AM, and 8:34 AM and inbound 4:36 PM, 5:31 PM, and 6:06 PM Silver Line to South Station Fairmount Line Improvements (four new stations) Parking Expansion – 1,000 New Park-and-Ride Spaces Assembly Square Orange Line Station in Somerville Green Line Extension to College Avenue and Union Square Beverly Garage Wonderland Parking Garage Salem Garage

Page 2 of 4 Allston Interchange Traffic Study March 27, 2017

TABLE 1 (continued) Transit Projects in the 2025 and 2040 No-Build Scenario

Winthrop Ferry (10 months of service) Fitchburg Commuter Rail Line Improvements Automatic Electronic Tolling (AET) New Balance Station Increased service to Yawkey Station Increased service on Framingham Commuter Rail Line Silver Line to Chelsea Tri-Town Community Bike Path in Stoneham, Winchester, Woburn Ferry Service – Charlestown to East Boston to South Boston

2 TRANSIT NETWORKS FOR BUILD SCENARIOS For the design year 2040 Build scenarios, the focus of the new transit services in the study area centers around West Station. West Station is to be located in the former Beacon Park Yard between Malvern Street and Babcock Street in Allston. The station will be on the Worcester/Framingham commuter rail line and will be served by two or three new shuttle bus routes; the routes vary depending on the scenarios. For scenario MUI and MUI-T, the shuttle bus routes are 1) Harvard Shuttle, operating between Harvard Square and West Station; 2) Kendall Shuttle, operating between Kendall Square and West Station via Central Square; and 3) LMA Shuttle, operating between Ruggles Station and West Station via the Longwood Medical Area. For scenarios MUI-C and MUI-CT, the shuttle bus routes include 1) Harvard Shuttle, operating between Harvard Square and West Station, and 2) Kendall-LMA Shuttle, operating between Kendall Square and West Station via Ruggles Station and the Longwood Medical Area. Besides the commuter rail line and shuttles, the study area is to be continuously served by several existing MBTA bus services, including Routes 66, 64, and 70/70A.

In the opening year 2025 Build scenario, it is assumed that West Station is not open, and that one new shuttle bus route is serving the study area. This bus route would operate between Harvard Square and Barry’s Corner in Allston.

The proposed headways for the shuttles are included in Table 2.

Page 3 of 4 Allston Interchange Traffic Study March 27, 2017

TABLE 2 Proposed Shuttle Headways

West Station Shuttles AM Midday PM Night Harvard Shuttle 5 min. 15 min. 5 min. 20 min. Kendall Shuttle 5 min. 15 min. 5 min. 20 min. LMA Shuttle 5 min. 15 min. 5 min. 20 min. Kendall via LMA Shuttle 5 min. 15 min. 5 min. 20 min. Harvard - Barry's Corner Shuttle 10 min. 10 min. 10 min. 20 min.

The proposed Worcester/Framingham commuter rail service plans provided by VHB for the Build scenarios are attached to this memorandum.

KGK/kgk

Page 4 of 4 Project #12529.09 MassDOT West Station Project November, 2016 Future Operations Analysis Proposed Worcester Line Weekday Schedule - DRAFT for Ridership

Northbound 500 502 580 504 582 506 508 584 510 552 586 512 514 516 518 522 588 590 524 526 592 528 530 532 534 536 538 Worcester 4:45 5:15 5:55 6:30 7:00 7:30 8:05 8:40 10:40 12:05 14:15 15:45 17:20 18:05 19:15 20:00 20:30 21:35 23:20 0:20 Grafton 4:58 5:28 6:08 6:43 7:13 7:43 8:53 10:53 12:18 14:28 15:58 17:33 18:18 19:28 20:13 20:43 21:48 23:33 0:33 Westborough 5:02 5:32 6:12 6:47 7:17 7:47 8:57 10:57 12:22 14:32 16:02 17:37 18:22 19:32 20:17 20:47 21:52 23:37 0:37 Southborough 5:11 5:41 6:21 6:56 7:26 7:56 9:06 11:06 12:31 14:41 16:11 17:46 18:31 19:41 20:26 20:56 22:01 23:46 0:46 Ashland 5:15 5:45 6:25 7:00 7:30 8:00 9:10 11:10 12:35 14:45 16:15 17:50 18:35 19:45 20:30 21:00 22:05 23:50 0:50 Framingham 5:26 5:56 6:05 6:36 6:46 7:11 7:42 7:50 8:11 8:40 9:21 11:21 12:46 14:56 16:26 16:47 17:40 18:01 18:46 19:05 19:56 20:41 21:11 22:16 0:01 1:01 W Natick 5:31 6:01 6:10 6:41 6:51 7:16 7:49 7:55 8:16 8:45 9:26 11:26 12:51 15:01 16:52 18:06 19:10 20:01 20:46 21:16 22:21 0:06 1:06 Natick 5:36 6:05 6:15 6:56 7:21 8:00 8:20 8:50 9:31 11:31 12:56 15:06 16:57 18:11 19:15 20:06 20:51 21:21 22:26 0:11 1:11 Wellesley Sq 5:41 6:20 7:01 7:26 8:05 8:25 8:55 9:36 11:35 13:00 15:10 17:01 18:15 19:19 20:10 20:55 21:25 22:30 0:15 1:15 Wellesley Hls 5:45 6:24 7:05 7:30 8:09 8:29 8:59 9:40 11:39 13:04 15:14 17:05 18:19 19:23 20:14 20:59 21:29 22:34 0:19 1:19 Wellesley Fms 5:48 6:27 7:08 7:34 8:12 8:32 9:02 9:43 11:42 13:07 15:17 17:08 18:22 19:26 20:17 21:02 21:32 22:37 0:22 1:22 Auburndale 5:53 6:32 7:13 7:39 8:17 9:07 9:48 11:47 13:12 19:31 21:07 0:27 1:27 W Newton 5:56 6:35 7:16 7:42 8:20 9:10 9:51 11:50 13:15 19:34 21:10 0:30 1:30 Newtonville 5:59 6:38 7:19 7:46 8:23 9:13 9:54 11:53 13:18 19:37 21:13 0:33 1:33 Boston Landing 6:05 6:44 7:52 8:43 9:19 11:59 15:28 18:33 19:43 21:19 22:48 1:39 West Station 6:08 6:47 7:55 8:46 9:22 12:02 15:31 18:36 19:46 21:22 22:51 1:42 Yawkey 6:11 6:22 6:52 7:00 7:27 8:00 8:09 8:33 8:50 8:57 9:27 10:02 12:05 13:26 15:34 16:46 17:21 18:07 18:39 19:08 19:49 20:30 21:25 21:45 22:54 0:41 1:45 Back Bay 6:16 6:27 6:57 7:05 7:32 8:05 8:14 8:38 8:55 9:02 9:32 10:07 12:10 13:31 15:39 16:51 17:26 18:12 18:44 19:13 19:54 20:35 21:30 21:50 22:59 0:46 1:50 S Stn 6:22 6:33 7:03 7:11 7:38 8:11 8:20 8:44 9:01 9:07 9:38 10:12 12:15 13:36 15:44 17:06 17:41 18:17 18:49 19:23 19:59 20:40 21:35 21:55 23:04 0:51 1:55

Southbound 501 581 583 503 585 505 587 507 509 511 515 517 589 519 591 521 593 523 525 527 529 551 531 533 535 537 539 S Stn 4:55 5:00 5:30 5:45 6:48 7:15 7:26 8:55 10:15 11:55 14:00 15:30 15:40 16:25 16:35 17:05 17:15 17:40 17:50 18:20 18:45 19:35 19:45 20:30 21:35 22:30 23:30 Back Bay 5:00 5:05 5:35 5:50 6:53 7:21 7:32 9:01 10:21 12:01 14:06 15:36 15:46 16:31 16:41 17:11 17:21 17:46 17:56 18:26 18:51 19:41 19:51 20:36 21:41 22:36 23:36 Yawkey 5:05 5:10 5:40 5:55 6:58 7:37 9:06 10:26 12:06 14:11 15:41 15:51 16:36 16:46 17:16 17:26 17:51 18:01 18:31 18:56 19:46 19:56 20:41 21:46 22:41 23:41 West Station 5:58 7:40 9:09 12:09 14:14 15:54 16:39 17:29 18:04 18:59 19:59 21:49 23:44 Boston Landing 6:01 7:43 9:12 12:12 14:17 15:57 16:42 17:32 18:07 19:02 20:02 21:52 23:47 Newtonville 5:19 10:35 12:19 14:24 16:04 16:55 17:39 18:15 19:09 20:09 20:50 21:59 22:50 23:54 W Newton 5:23 10:39 12:23 14:28 16:08 16:59 17:43 18:19 19:13 20:13 20:54 22:03 22:54 23:58 Auburndale 5:26 10:42 12:26 14:31 16:11 17:02 17:46 18:22 19:16 20:16 20:57 22:06 22:57 0:01 Wellesley Fms 5:29 5:53 6:12 7:11 7:54 9:23 10:45 12:29 14:34 16:14 17:05 17:49 18:27 18:45 19:19 20:19 21:00 22:09 23:00 0:04 Wellesley Hls 5:32 5:56 6:15 7:14 7:57 9:26 10:48 12:32 14:37 16:17 17:08 17:52 18:30 18:48 19:22 20:22 21:03 22:12 23:03 0:07 Wellesley Sq 5:36 6:00 6:19 7:18 8:01 9:30 10:52 12:36 14:41 16:21 17:12 17:56 18:34 18:52 19:26 20:26 21:07 22:16 23:07 0:11 Natick 5:40 6:04 6:23 7:22 8:05 9:34 10:56 12:40 14:45 16:25 16:59 17:16 18:00 18:38 18:56 19:30 20:30 21:11 22:20 23:11 0:15 W Natick 5:24 5:45 6:09 6:28 7:27 8:09 9:38 11:01 12:45 14:50 16:01 16:30 17:04 17:21 17:37 18:05 18:12 18:43 19:01 19:35 20:35 21:16 22:25 23:16 0:20 Framingham 5:28 5:50 6:14 6:33 7:32 7:46 8:14 9:43 11:06 12:50 14:55 16:06 16:36 17:09 17:27 17:45 18:18 18:20 18:48 19:06 19:40 20:40 21:21 22:30 23:21 0:25 Ashland 5:34 6:39 7:52 9:49 11:12 12:56 15:01 16:12 17:15 17:52 18:27 18:54 19:12 19:46 20:46 21:27 22:36 23:27 0:31 Southborough 5:39 6:44 7:57 9:54 11:17 13:01 15:06 16:17 17:20 17:57 18:32 18:59 19:17 19:51 20:51 21:32 22:41 23:32 0:36 Westborough 5:48 6:53 8:06 10:03 11:26 13:10 15:15 16:26 17:29 18:06 18:41 19:08 19:26 20:00 21:00 21:41 22:50 23:41 0:45 Grafton 5:53 6:58 8:11 10:08 11:31 13:15 15:20 16:31 17:34 18:11 18:46 19:13 19:31 20:05 21:05 21:46 22:55 23:46 0:50 Worcester 6:06 7:11 8:24 10:21 11:45 13:28 15:34 16:49 17:54 18:32 19:10 19:32 19:45 20:18 20:40 21:19 22:00 23:09 0:00 1:04

Blue Indicates L stop (regular stop, but train may leave ahead of schedule) Red Indicates f stop (train will not stop unless passengers notify conductor they wish to get off or passengers are visible waiting on the platform) Indicates new station stop at Boston Landing and West Station Indicates peak hour trains

Note that this draft schedule with proposed station stops at Boston Landing and West Station was developed for the purposes of ridership planning. The service plan has not yet been tested through simulation modeling, and is subject to change.

Based on MBTA Framingham/Worcester Line Schedule Effective November 21, 2016. Based on MBTA Equipment Cycle Effective May 23, 2016 Attachment D

CTPS CENTRAL TRANSPORTATION PLANNING STAFF CENTRAL TRANSPORTATION Staff to the Boston Region Metropolitan Planning Organization PLANNING STAFF

TECHNICAL MEMORANDUM

DATE: March 16, 2017 TO: Allston Interchange Task Force FROM: K. Grace King and Bruce Kaplan Central Transportation Planning Staff RE: Allston Interchange Traffic Study: Land Use Assumptions

The proposed new Massachusetts Turnpike mainline alignment and ramp configurations in Allston stands to free up developable land. For the Allston Interchange Traffic Study, the new development associated with the newly available land will need to be estimated to appropriately assess its impacts on local and regional traffic demand. This memorandum describes how the land use assumptions for the year 2040 (the study’s future horizon year) were estimated and later used in the regional model for this project.

1 BACKGROUND The study area encompasses four transportation analysis zones (TAZs) in the Boston Region Metropolitan Planning Organization’s regional travel demand model: TAZs 238, 244, 245, and 246 (see Figure 1). TAZ 238, between Soldiers Field Road and North Harvard Street, contains Harvard Stadium and other athletic facilities and recreational areas. TAZ 244, a triangular area bordered by North Harvard Street, Western Avenue, and Soldiers Field Road, is mostly composed of Harvard Business School facilities. TAZ 245, located between Soldiers Field Road, Western Avenue, and the area east of North Harvard Street, is home to Genzyme Corporation and contains many undeveloped empty lots. Finally, TAZ 246, a triangular zone between Cambridge Street, Soldiers Field Road, and railroad tracks, is the location of the Allston Interchange ramps, the CSX rail yard, a warehouse, and Doubletree Suites Hotel. TAZ 246 contains developable land, which will provide the additional economic opportunities associated with this project; the potential for new development on this land is reflected in the land use projections used in this study.

The driving force behind the new development in the study area is Harvard University, the major land owner in the study area. Harvard’s Institutional Management Plan (IMP), a ten year plan which was approved in 2013, outlines population and employment projections for the study area. Based on this information, the Harvard Planning and Project Management department provided the project team with employment and population forecasts, projected for the horizon year 2040, for TAZs 238, 244 and 245, but not for TAZ 246.

2 2040 NO-BUILD SCENARIO The original land use forecasts for the 2040 No-Build scenario were based on the population and employment estimates in Boston Region Metropolitan Planning Organization’s Long Range Transportation Plan, Charting Progress to 2040. In order to best reflect Harvard’s updated plan for the study area, in the No-Build scenario TAZs 238, 244, and 245 were assumed to have the IMP’s proposed levels of population and employment because such development is not contingent on the Allston Interchange Turnpike project. The development of TAZ 246, however, is directly tied to the project. The land use assumptions in the No-Build scenario for TAZ 246 will remain unchanged from the base year, 2012.

3 2040 BUILD SCENARIO The Build scenario represents the proposed new Massachusetts Turnpike mainline alignment and ramp configurations. For the Build scenario, the population and employment projections for TAZ 246 were calculated by assuming the same land use density as portions of the study area having similar characteristics, which were determined, with input from Harvard, to be TAZs 244 and 245. Since the developable land area in TAZ 246 is 70 percent of the sum of the developable land areas in TAZs 244 and 245, it is projected to have 70 percent of the employment and population of TAZs 244 and 245 (combined) in the 2040 Build condition. No new development is projected for TAZs 238, 244, and 245 in this scenario.

4 2025 NO-BUILD AND BUILD SCENARIOS For the Allston Interchange Study, the year 2025 was designated as the opening year for the realigned Turnpike and ramps in Allston. Although both No-Build and Build scenarios were developed for the opening year, land use was considered to be identical in each because no new development was assumed to occur in TAZ 246. The 2025 land use for TAZs 238, 244 and 245, was based on Harvard University’s Ten-Year Plan (pre-2025) for the study area. TAZ 245 reflects the new land use Harvard proposed in November 2015 for its Allston science center.

5 RESULTS Table 1 displays the final land use assumptions set as the No-Build and Build conditions for the base year (2012), the opening year (2025), and the future year (2040). Table 2 displays the trip ends produced by the Central Transportation Planning Staff’s trip generation model for the four TAZs in the study area. The number of trip productions and trip attractions within a TAZ are estimated as a function of the household and population characteristics, and employment types. Trip productions and attractions were estimated for eight different trip purposes, each having their own distinct trip-rate functions: home-based work (HBW), home-based work-related (HBWR), home-based

Page 2 of 5 Allston Interchange Traffic Study: Land Use Assumptions March 16, 2017

personal business (HBPB), home-based social-recreational (HBSR), home-based school (HBSC), home-based pick-up/drop-off (HBPD), non-home-based work (NHBW), and non-home-based other (NHBO).

TABLE 1 Land Use Assumptions

Base Year Base Year Base Year Base Year Retail Basic Service Group Base Year Base Year TAZ Employment Employment Employment Quarters Population Households 238 0 0 624 0 0 0 244 1 10 1,072 1,503 3,069 867 245 0 424 50 0 127 54 246 154 115 314 0 0 0 Total 155 549 2,060 1,503 3,196 921

2040 No-Build 2040 No-Build 2040 No-Build 2040 No-Build 2040 Retail Basic Service Group No-Build 2040 No-Build TAZ Employment Employment Employment Quarters Population Households 238 113 0 513 400 862 325 244 106 10 3,331 1,863 2,960 654 245 113 424 4,598 0 589 379 246 154 115 314 0 0 0 Total 486 549 8,756 2,263 4,411 1,358

2040 Build 2040 Build 2040 Build 2040 Build Retail Basic Service Group 2040 Build 2040 Build TAZ Employment Employment Employment Quarters Population Households 238 113 0 513 400 862 325 244 106 10 3,331 1,863 2,960 654 245 113 424 4,598 0 589 379 246 153 304 5,550 0 2,484 1,523 Total 485 738 13,992 2,263 6,895 2,881

2025 Retail 2025 Basic 2025 Service 2025 Group 2025 2025 TAZ Employment Employment Employment Quarters Population Households 238 113 0 513 400 862 325 244 54 5 2,202 1,683 2,780 654 245 57 424 2,434 0 127 54 246 154 115 314 0 0 0 Total 378 544 5,463 2,083 3,769 1,033

Retail: store front, restaurant, shopping center, etc. Basic: manufacturing, laboratory, warehouse, transportation, etc. Service: office, hospital, hotel, library, health club, etc. Group Quarters: school residence hall, nursing facility, prison, etc.

Page 3 of 5 Allston Interchange Traffic Study: Land Use Assumptions March 16, 2017

TABLE 2 Productions and Attractions as Trip Ends

Base Year Base Year Base Year Base Year Base Year Base Year Base Year Productions Productions Productions Productions Productions Productions Productions TAZ NHBO NHBW HBPD HBSC HBPB HBSR HBWK 238 1,516 1,145 0 0 0 0 0 244 2,966 1,977 158 3,583 3,580 2,284 1,496 245 45 250 18 55 124 84 83 246 356 620 0 0 0 0 0 Total 4,883 3,992 176 3,638 3,704 2,368 1,579

Base Year Base Year Base Year Base Year Base Year Base Year Base Year Attractions Attractions Attractions Attractions Attractions Attractions Attractions TAZ NHBO NHBW HBPD HBSC HBPB HBSR HBWK 238 1,238 1,116 19 1,608 0 649 831 244 2,535 1,954 122 2,601 1,090 2,103 1,433 245 43 178 20 0 68 65 618 246 367 637 18 0 643 195 792 Total 4,183 3,885 179 4,209 1,801 3,012 3,674

2025 2025 2025 2025 2025 2025 2025 Productions Productions Productions Productions Productions Productions Productions TAZ NHBO NHBW HBPD HBSC HBPB HBSR HBWK 238 1,715 1,202 40 105 421 287 345 244 3,003 3,001 106 3,744 3,243 2,089 1,438 245 811 2,568 17 55 120 84 84 246 356 620 0 0 0 0 0 Total 5,885 7,391 163 3,904 3,784 2,461 1,867

2025 2025 2025 2025 2025 2025 2025 Attractions Attractions Attractions Attractions Attractions Attractions Attractions TAZ NHBO NHBW HBPD HBSC HBPB HBSR HBWK 238 1,503 1,270 54 1,539 905 1,109 891 244 2,559 2,843 140 2,490 1,064 1,775 2,821 245 685 2,181 99 0 315 137 3,503 246 368 644 19 0 662 195 792 Total 5,115 6,938 312 4,028 2,946 3,216 8,006

Page 4 of 5 Allston Interchange Traffic Study: Land Use Assumptions March 16, 2017

TABLE 2 (continued) Productions and Attractions as Trip Ends

2040 2040 2040 2040 2040 2040 2040 No-Build No-Build No-Build No-Build No-Build No-Build No-Build Productions Productions Productions Productions Productions Productions Productions TAZ NHBO NHBW HBPD HBSC HBPB HBSR HBWK 238 2,002 1,322 40 105 422 288 344 244 3,878 4,342 105 4,098 3,449 2,229 1,548 245 1,700 4,703 54 158 522 384 429 246 356 620 0 0 0 0 0 Total 7,936 10,987 199 4,361 4,393 2,901 2,321

2040 2040 2040 2040 2040 2040 2040 No-Build No-Build No-Build No-Build No-Build No-Build No-Build Attractions Attractions Attractions Attractions Attractions Attractions Attractions TAZ NHBO NHBW HBPD HBSC HBPB HBSR HBWK 238 1,735 1,389 51 1,691 866 1,215 913 244 3,292 4,025 168 2,735 1,243 2,022 4,274 245 1,467 3,999 192 0 933 585 6,129 246 367 636 18 0 634 190 792 Total 6,861 10,049 429 4,426 3,676 4,012 12,108

2040 Build 2040 Build 2040 Build 2040 Build 2040 Build 2040 Build 2040 Build Productions Productions Productions Productions Productions Productions Productions TAZ NHBO NHBW HBPD HBSC HBPB HBSR HBWK 238 2,003 1,323 40 105 423 288 344 244 3,881 4,343 105 4,098 3,449 2,229 1,548 245 1,700 4,703 54 158 523 385 429 246 3,044 6,304 254 680 2,152 1,608 1,951 Total 10,629 16,673 453 5,041 6,546 4,510 4,272

2040 Build 2040 Build 2040 Build 2040 Build 2040 Build 2040 Build 2040 Build Attractions Attractions Attractions Attractions Attractions Attractions Attractions TAZ NHBO NHBW HBPD HBSC HBPB HBSR HBWK 238 1,737 1,390 51 1,692 866 1,216 913 244 3,294 4,027 168 2,738 1,244 2,023 4,274 245 1,467 4,000 192 0 933 585 6,129 246 2,806 5,553 349 0 3,143 2,169 7,945 Total 9,304 14,971 760 4,430 6,186 5,993 19,260

NHBO: Non-Home-Based Other HBWK: Home-Based Work NHBW: Non-Home-Based Work HBSC: Home-Based School HBPD: Home-Based Pickup/Drop-off HBSR: Home-Based Social Recreational HBPB: Home-Based Personal Business

KGK/kgk

Page 5 of 5 893 624 623 552 563 561 507 893 891 555 560 509 919 621 553 565 508 505 917 889 622 620 551 559 890 503 504 918 719 618 619 550 916 728 597 615 596 466 920 603 720 612 598 727 718 617 595 502 465 616 602 722 599 594 206 961 909 726 721 613 601 592 915 910 724723 717 725 716 611 600 591 593 462 705 704 921 614 610 590 206 706 703 912 914 702 606 588 589 607 202 707 205 990 713 699 698 586609 587 195 913 911 701700 605 201 908 695 608 604 937 715 580 204 709 708 694 697 585 200 196 696 678 583 194 936 581 199 931 714 710 584 579 197 693 677 582 203 711 687 691 679 193 930 929 661 646 686 688 680 676 629 198 938 690 662 647 630 625 922 685 681 675 648 632 192 712 684 682 649 631 933 932 939 238 673674 651 633 634 626 928 663 1 4 5 927 683 672 650 635 636627 7 934 244 664 2 8 9 671 653 638 628 21 3 6 1002 923 239 665654 652 637 10 926 670 640 22 29 935 245 666 656 639 23 24 32 11 925 924 669 655 642 37 1004 241 657 27 30 36 39 12 240 242 243 658 641 25 35 999 222 667 659 643 31 41 46 1001 1000 246 28 40 5013 223 53 224 247 668 660 644 26 52 6668 1013 212 248 86 71 6263 14 1005 207 249 645 135 225 250252 85 9197 74 65 15 227 230 236 251 254 257 84 88 72 7975 77 17 136 1021 213 235 253 256 92 101 19 226 237 255 94 80 18 143137 234 87 95104 102 78 8182 1014 208 233 258 259 144 214 228 731 260 108 93 83 229231 730 729 263 106 105 20 216 732 261 265 112 114 145 1022 232 262 268 264 109 110 115 266 107 119 146 148 209 215 218 269 217 733 734 735 272 273 117 120 1020 221 279 270271 116 121 1015 276275 123 147 151 280 219 220 739 738 283 274 277 122 737 736 281 312 125 124 152 284 287 278 210 740 282288 313 127 126 130 156 1016 741 286 129 1017 211 743 744 293 289290 320 314 128 153 1019 1018 742 292 131 154 166 745 291 321 315 749 294 155 1038 746 324 132 322 316 1035 750 747 295 1034 Study area 297 296 325 323 TAZ boundaries 298 133 167 1037 748 301 1039 1040 299 332 Town boundaries 326 317 318 134 1036 751 300 752 328 355 304 327 329 333 319 753 302 303 354 359 1043 1042 1041 754 754 306 305 343 334 335 336 330 331 353 FIGURE 1 Allston Interchange CTPS Transportation Analysis Zones of the Study Area Land Use Attachment E I‐90 Allston Interchange Project Transit Crowding Analysis AM Peak Period (6‐9 AM) Results

2025 OPENING YEAR NO BUILD (NB) AM Peak Period Service Characteristics AM Peak 1 Hour Service Info AM Peak 3‐Hr Service Info

Car Set Set Service Model Volume / Volume / Cars / Seated Car Policy Seated Policy Seated Policy Volume ‐ Seated Policy Max 3 Hr Model Peak Hr Commuter Rail Line Service Peak Load Point Set Headway Capacity Max Load Capacity Max Load Frequency Capacity Max Load 1 Hr Capacity Load Volume Factor Worcester/Framingham Line IB Newtonville ‐ Boston Landing 6 20.00 185 204 1,110 1,221 3 3,330 3,663 2,451 0.74 0.67 3,770 0.65 Worcester/Framingham Line OB South Station ‐ BackBay 6 45.00 185 204 1,110 1,221 1 1,480 1,628 275 0.19 0.17 410 0.67 Green B Branch IB ‐ Aboveground BU West ‐ BU Central 3 6.19 45 100 135 300 10 1,311 2,913 3,002 2.29 1.03 5,560 0.54 Green B Branch OB ‐ Aboveground BU East ‐ BU Central 3 7.20 45 100 135 300 8 1,127 2,504 1,297 1.15 0.52 2,820 0.46 Green B Branch IB ‐ Underground Copley ‐ Arlington 3 1.47 45 100 135 300 41 5,512 12,248 3,694 0.67 0.30 6,840 0.54 Green B Branch OB ‐ Underground Boylston ‐ Arlington 3 1.65 45 100 135 300 36 4,920 10,933 5,212 1.06 0.48 11,330 0.46 Harvard ‐ Barry's Corner N. Harvard St ‐ Harvard Sq 1 10.00 39 55 39 55 6 234 328 29 0.13 0.09 60 0.49 66 IB 1 8.57 57 80 57 80 7 399 559 675 1.69 1.21 1,650 0.41 66 OB 1 5.29 57 80 57 80 11 646 904 473 0.73 0.52 1,180 0.40 64 IB 1 20.00 39 55 39 55 3 117 164 127 1.09 0.77 250 0.51 64 OB 1 22.50 39 55 39 55 3 104 146 113 1.09 0.77 170 0.67 70/70A IB 1 10.59 39 55 39 55 6 221 309 240 1.09 0.78 490 0.49 70/70A OB 1 10.00 39 55 39 55 6 234 328 302 1.29 0.92 700 0.43 57/57A IB 1 5.14 39 55 39 55 12 455 637 427 0.94 0.67 1,130 0.38 57/57A OB 1 7.20 39 55 39 55 8 325 455 388 1.19 0.85 960 0.40

2025 OPENING YEAR BUILD AM Peak Period Service Characteristics AM Peak 1 Hour Service Info AM Peak 3‐Hr Service Info

Car Set Set Service Model Volume / Volume / Cars / Seated Car Policy Seated Policy Seated Policy Volume ‐ Seated Policy Max 3 Hr Model Peak Hr Commuter Rail Line Service Peak Load Point Set Headway Capacity Max Load Capacity Max Load Frequency Capacity Max Load 1 Hr Capacity Load Volume Factor Worcester/Framingham Line IB Newtonville ‐ Boston Landing 6 20.00 185 204 1,110 1,221 3 3,330 3,663 2,470 0.74 0.67 3,800 0.65 Worcester/Framingham Line OB South Sta ‐ Back Bay 6 45.00 185 204 1,110 1,221 1 1,480 1,628 281 0.19 0.17 420 0.67 Green B Branch IB ‐ Aboveground BU West ‐ BU Central 3 6.19 45 100 135 300 10 1,311 2,913 3,008 2.29 1.03 5,570 0.54 Green B Branch OB ‐ Aboveground BU East ‐ BU Central 3 7.20 45 100 135 300 8 1,127 2,504 1,316 1.17 0.53 2,860 0.46 Green B Branch IB ‐ Underground Copley ‐ Arlington 3 1.47 45 100 135 300 41 5,512 12,248 3,710 0.67 0.30 6,870 0.54 Green B Branch OB ‐ Underground Boylston ‐ Arlington 3 1.65 45 100 135 300 36 4,920 10,933 5,226 1.06 0.48 11,360 0.46 Harvard ‐ Barry's Corner N. Harvard St ‐ Harvard Sq 1 10.00 39 55 39 55 6 234 328 29 0.13 0.09 60 0.49 66 IB 1 8.57 57 80 57 80 7 399 559 650 1.63 1.16 1,590 0.41 66 OB 1 5.29 57 80 57 80 11 646 904 396 0.61 0.44 990 0.40 64 IB 1 20.00 39 55 39 55 3 117 164 117 1.00 0.71 230 0.51 64 OB 1 22.50 39 55 39 55 3 104 146 106 1.02 0.73 160 0.67 70/70A IB 1 10.59 39 55 39 55 6 221 309 255 1.15 0.83 520 0.49 70/70A OB 1 10.00 39 55 39 55 6 234 328 298 1.27 0.91 690 0.43 57/57A IB 1 5.14 39 55 39 55 12 455 637 427 0.94 0.67 1,130 0.38 57/57A OB 1 7.20 39 55 39 55 8 325 455 396 1.22 0.87 980 0.40 I‐90 Allston Interchange Project Transit Crowding Analysis AM Peak Period (6‐9 AM) Results

2040 DESIGN YEAR NO BUILD (NB) AM Peak Period Service Characteristics AM Peak 1 Hour Service Info AM Peak 3‐Hr Service Info

2040 DESIGN YEAR BUILD AM Peak Period Service Characteristics AM Peak 1 Hour Service Info AM Peak 3‐Hr Service Info

Car Set Set Service Model Volume / Volume / Cars / Seated Car Policy Seated Policy Seated Policy Volume ‐ Seated Policy Max 3 Hr Model Peak Hr Commuter Rail Line Service Peak Load Point Set Headway Capacity Max Load Capacity Max Load Frequency Capacity Max Load 1 Hr Capacity Load Volume Factor Worcester/Framingham Line IB Newtonville ‐ Boston Landing 6 20.00 185 204 1,110 1,221 3 3,330 3,663 2,763 0.83 0.75 4,250 0.65 Worcester/Framingham Line OB Wellesley Farms ‐ Wellesley Hills 6 45.00 185 204 1,110 1,221 1 1,480 1,628 328 0.22 0.20 490 0.67 Green B Branch IB ‐ Aboveground St Paul ‐ BU West 3 6.19 45 100 135 300 10 1,311 2,913 3,256 2.48 1.12 6,030 0.54 Green B Branch OB ‐ Aboveground BU East ‐ BU Central 3 7.20 45 100 135 300 8 1,127 2,504 1,495 1.33 0.60 3,250 0.46 Green B Branch IB ‐ Underground Copley ‐ Arlington 3 1.47 45 100 135 300 41 5,512 12,248 4,039 0.73 0.33 7,480 0.54 Green B Branch OB ‐ Underground Boylston ‐ Arlington 3 1.65 45 100 135 300 36 4,920 10,933 5,359 1.09 0.49 11,650 0.46 Harvard Shuttle Bus IB Stadium Way/West Station 1 5.00 39 55 39 55 12 468 655 34 0.07 0.05 70 0.49 Harvard Shuttle Bus OB West Station ‐ Stadium Way/Western Ave 1 5.00 39 55 39 55 12 468 655 60 0.13 0.09 150 0.40 Kendall Shuttle Bus IB Central Square ‐ Stadium Way/Western Ave 1 5.00 39 55 39 55 12 468 655 172 0.37 0.26 350 0.49 Kendall Shuttle Bus OB West Station ‐ Cambridge St/East Dr 1 5.00 39 55 39 55 12 468 655 40 0.09 0.06 100 0.40 LMA Shuttle Bus IB Ruggles ‐ LMA 1 5.00 39 55 39 55 12 468 655 461 0.98 0.70 940 0.49 LMA Shuttle Bus OB West Station ‐ LMA 1 5.00 39 55 39 55 12 468 655 92 0.20 0.14 230 0.40 66 IB 1 8.57 57 80 57 80 7 399 559 613 1.54 1.10 1,500 0.41 66 OB 1 5.29 57 80 57 80 11 646 904 460 0.71 0.51 1,150 0.40 64 IB 1 20.00 39 55 39 55 3 117 164 132 1.13 0.81 260 0.51 64 OB 1 22.50 39 55 39 55 3 104 146 126 1.22 0.87 190 0.67 70/70A IB 1 10.59 39 55 39 55 6 221 309 280 1.27 0.91 570 0.49 70/70A OB 1 10.00 39 55 39 55 6 234 328 285 1.22 0.87 660 0.43 57/57A IB 1 5.14 39 55 39 55 12 455 637 495 1.09 0.78 1,310 0.38 57/57A OB 1 7.20 39 55 39 55 8 325 455 424 1.30 0.93 1,050 0.40 I‐90 Allston Interchange Project Transit Crowding Analysis AM Peak Period (6‐9 AM) Results

2040 DESIGN YEAR BUILD W/ COMM AVE CONNECTION AM Peak Period Service Characteristics AM Peak 1 Hour Service Info AM Peak 3‐Hr Service Info

Car Set Set Service Model Volume / Volume / Cars / Seated Car Policy Seated Policy Seated Policy Volume ‐ Seated Policy Max 3 Hr Model Peak Hr Commuter Rail Line Service Peak Load Point Set Headway Capacity Max Load Capacity Max Load Frequency Capacity Max Load 1 Hr Capacity Load Volume Factor Worcester/Framingham Line IB Newtonville ‐ Boston Landing 6 20.00 185 204 1,110 1,221 3 3,330 3,663 2,717 0.82 0.74 4,180 0.65 Worcester/Framingham Line OB Wellesley Farms ‐ Wellesley Hills 6 45.00 185 204 1,110 1,221 1 1,480 1,628 328 0.22 0.20 490 0.67 Green B Branch IB ‐ Aboveground St Paul St ‐ BU West 3 6.19 45 100 135 300 10 1,311 2,913 3,348 2.55 1.15 6,200 0.54 Green B Branch OB ‐ Aboveground BU East ‐ BU Central 3 7.20 45 100 135 300 8 1,127 2,504 1,518 1.35 0.61 3,300 0.46 Green B Branch IB ‐ Underground Copley ‐ Arlington 3 1.47 45 100 135 300 41 5,512 12,248 4,045 0.73 0.33 7,490 0.54 Green B Branch OB ‐ Underground Boylston ‐ Arlington 3 1.65 45 100 135 300 36 4,920 10,933 5,548 1.13 0.51 12,060 0.46 Harvard Shuttle Bus IB Stadium Way/West Ave ‐ Stadium Way/Cambridge St 1 5.00 39 55 39 55 12 468 655 34 0.07 0.05 70 0.49 Harvard Shuttle Bus OB West Sta ‐ Stadium Way/Western Ave 1 5.00 39 55 39 55 12 468 655 60 0.13 0.09 150 0.40 West Station ‐ LMA ‐Kendall West Station ‐ LMA 1 5.00 39 55 39 55 12 468 655 98 0.21 0.15 200 0.49 Kendall ‐ LMA ‐ West Station Ruggles ‐ LMA 1 5.00 39 55 39 55 12 468 655 44 0.09 0.07 110 0.40 66 IB 1 8.57 57 80 57 80 7 399 559 720 1.80 1.29 1,760 0.41 66 OB 1 5.29 57 80 57 80 11 646 904 473 0.73 0.52 1,180 0.40 64 IB 1 20.00 39 55 39 55 3 117 164 137 1.17 0.84 270 0.51 64 OB 1 22.50 39 55 39 55 3 104 146 120 1.15 0.82 180 0.67 70/70A IB 1 10.59 39 55 39 55 6 221 309 270 1.22 0.87 550 0.49 70/70A OB 1 10.00 39 55 39 55 6 234 328 341 1.46 1.04 790 0.43 57/57A IB 1 5.14 39 55 39 55 12 455 637 491 1.08 0.77 1,300 0.38 57/57A OB 1 7.20 39 55 39 55 8 325 455 416 1.28 0.91 1,030 0.40

2040 DESIGN YEAR BUILD W/ STADIUM WAY TRANSIT‐ONLY SOUTHBOUND (SB) AM Peak Period Service Characteristics AM Peak 1 Hour Service Info AM Peak 3‐Hr Service Info

Car Set Set Service Model Volume / Volume / Cars / Seated Car Policy Seated Policy Seated Policy Volume ‐ Seated Policy Max 3 Hr Model Peak Hr Commuter Rail Line Service Peak Load Point Set Headway Capacity Max Load Capacity Max Load Frequency Capacity Max Load 1 Hr Capacity Load Volume Factor Worcester/Framingham Line IB Newtonville ‐ Boston Landing 6 20.00 185 204 1,110 1,221 3 3,330 3,663 2,763 0.83 0.75 4,250 0.65 Worcester/Framingham Line OB Wellesley Farms ‐ Wellesley Hills 6 45.00 185 204 1,110 1,221 1 1,480 1,628 328 0.22 0.20 490 0.67 Green B Branch IB ‐ Aboveground St Paul ‐ BU West 3 6.19 45 100 135 300 10 1,311 2,913 3,256 2.48 1.12 6,030 0.54 Green B Branch OB ‐ Aboveground BU East ‐ BU Central 3 7.20 45 100 135 300 8 1,127 2,504 1,500 1.33 0.60 3,260 0.46 Green B Branch IB ‐ Underground Copley ‐ Arlington 3 1.47 45 100 135 300 41 5,512 12,248 4,039 0.73 0.33 7,480 0.54 Green B Branch OB ‐ Underground Boylston ‐ Arlington 3 1.65 45 100 135 300 36 4,920 10,933 5,543 1.13 0.51 12,050 0.46 Harvard Shuttle Bus IB Stadium Way/West Station 1 5.00 39 55 39 55 12 468 655 59 0.13 0.09 120 0.49 Harvard Shuttle Bus OB West Station ‐ Stadium Way/Western Ave 1 5.00 39 55 39 55 12 468 655 60 0.13 0.09 150 0.40 Kendall Shuttle Bus IB Central Square ‐ Stadium Way/Western Ave 1 5.00 39 55 39 55 12 468 655 172 0.37 0.26 350 0.49 Kendall Shuttle Bus OB West Station ‐ Cambridge St/East Dr 1 5.00 39 55 39 55 12 468 655 16 0.03 0.02 40 0.40 LMA Shuttle Bus IB Ruggles ‐ LMA 1 5.00 39 55 39 55 12 468 655 470 1.01 0.72 960 0.49 LMA Shuttle Bus OB West Station ‐ LMA 1 5.00 39 55 39 55 12 468 655 92 0.20 0.14 230 0.40 66 IB 1 8.57 57 80 57 80 7 399 559 613 1.54 1.10 1,500 0.41 66 OB 1 5.29 57 80 57 80 11 646 904 465 0.72 0.51 1,160 0.40 64 IB 1 20.00 39 55 39 55 3 117 164 132 1.13 0.81 260 0.51 64 OB 1 22.50 39 55 39 55 3 104 146 126 1.22 0.87 190 0.67 70/70A IB 1 10.59 39 55 39 55 6 221 309 280 1.27 0.91 570 0.49 70/70A OB 1 10.00 39 55 39 55 6 234 328 281 1.20 0.86 650 0.43 57/57A IB 1 5.14 39 55 39 55 12 455 637 495 1.09 0.78 1,310 0.38 57/57A OB 1 7.20 39 55 39 55 8 325 455 424 1.30 0.93 1,050 0.40 I‐90 Allston Interchange Project Transit Crowding Analysis PM Peak Period (3‐6 PM) Results

2025 OPENING YEAR NO BUILD (NB) PM Peak Period Service Characteristics PM Peak 1 Hour Service Info PM Peak 3‐Hr Service Info

Car Set Set Service Model Volume / Volume / Cars / Seated Car Policy Seated Policy Seated Policy Volume ‐ Seated Policy Max 3 Hr Model Peak Hr Commuter Rail Line Service Peak Load Point Set Headway Capacity Max Load Capacity Max Load Frequency Capacity Max Load 1 Hr Capacity Load Volume Factor Worcester/Framingham Line IB Wellesley Hills ‐ Wellesley Farms 6 60.00 185 204 1,110 1,221 1 1,110 1,221 286 0.26 0.23 440 0.65 Worcester/Framingham Line OB Boston Landing ‐ Newtonville 6 22.50 185 204 1,110 1,221 3 2,960 3,256 1,682 0.57 0.52 2,510 0.67 Green B Branch IB ‐ Aboveground Blandford ‐ Kenmore 3 6.00 45 100 135 300 10 1,352 3,005 1,595 1.18 0.53 4,090 0.39 Green B Branch OB ‐ Aboveground BU East ‐ BU Central 3 6.70 45 100 135 300 9 1,211 2,691 2,143 1.77 0.80 4,870 0.44 Green B Branch IB ‐ Underground Boylston ‐ Park 3 1.51 45 100 135 300 40 5,385 11,966 3,487 0.65 0.29 8,940 0.39 Green B Branch OB ‐ Underground Arlington ‐ Copley 3 1.54 45 100 135 300 39 5,255 11,677 3,590 0.68 0.31 8,160 0.44 Harvard ‐ Barry's Corner N. Harvard St ‐ Harvard Sq 1 10.00 39 55 39 55 6 234 328 147 0.63 0.45 300 0.49 66 IB 1 9.00 57 80 57 80 7 380 532 319 0.84 0.60 780 0.41 66 OB 1 9.47 57 80 57 80 6 361 506 292 0.81 0.58 730 0.40 64 IB 1 30.00 39 55 39 55 2 78 109 249 3.19 2.28 490 0.51 64 OB 1 25.71 39 55 39 55 2 91 127 140 1.53 1.10 210 0.67 70/70A IB 1 11.25 39 55 39 55 5 208 291 270 1.30 0.93 550 0.49 70/70A OB 1 10.59 39 55 39 55 6 221 309 155 0.70 0.50 360 0.43 57/57A IB 1 7.83 39 55 39 55 8 299 419 189 0.63 0.45 500 0.38 57/57A OB 1 6.92 39 55 39 55 9 338 473 343 1.02 0.73 850 0.40

2025 OPENING YEAR BUILD PM Peak Period Service Characteristics PM Peak 1 Hour Service Info PM Peak 3‐Hr Service Info

Car Set Set Service Model Volume / Volume / Cars / Seated Car Policy Seated Policy Seated Policy Volume ‐ Seated Policy Max 3 Hr Model Peak Hr Commuter Rail Line Service Peak Load Point Set Headway Capacity Max Load Capacity Max Load Frequency Capacity Max Load 1 Hr Capacity Load Volume Factor Worcester/Framingham Line IB Wellesley Hills ‐ Wellesley Farms 6 60.00 185 204 1,110 1,221 1 1,110 1,221 286 0.26 0.23 440 0.65 Worcester/Framingham Line OB Boston Landing ‐ Newtonville 6 22.50 185 204 1,110 1,221 3 2,960 3,256 1,702 0.57 0.52 2,540 0.67 Green B Branch IB ‐ Aboveground Blandford ‐ Kenmore 3 6.00 45 100 135 300 10 1,352 3,005 1,626 1.20 0.54 4,170 0.39 Green B Branch OB ‐ Aboveground BU East ‐ BU Central 3 6.70 45 100 135 300 9 1,211 2,691 2,156 1.78 0.80 4,900 0.44 Green B Branch IB ‐ Underground Boylston ‐ Park 3 1.51 45 100 135 300 40 5,385 11,966 3,487 0.65 0.29 8,940 0.39 Green B Branch OB ‐ Underground Arlington ‐ Copley 3 1.54 45 100 135 300 39 5,255 11,677 3,586 0.68 0.31 8,150 0.44 Harvard ‐ Barry's Corner N. Harvard St ‐ Harvard Sq 1 10.00 39 55 39 55 6 234 328 142 0.61 0.43 290 0.49 66 IB 1 9.00 57 80 57 80 7 380 532 307 0.81 0.58 750 0.41 66 OB 1 9.47 57 80 57 80 6 361 506 272 0.75 0.54 680 0.40 64 IB 1 30.00 39 55 39 55 2 78 109 234 3.00 2.14 460 0.51 64 OB 1 25.71 39 55 39 55 2 91 127 120 1.32 0.94 180 0.67 70/70A IB 1 11.25 39 55 39 55 5 208 291 280 1.34 0.96 570 0.49 70/70A OB 1 10.59 39 55 39 55 6 221 309 155 0.70 0.50 360 0.43 57/57A IB 1 7.83 39 55 39 55 8 299 419 181 0.61 0.43 480 0.38 57/57A OB 1 6.92 39 55 39 55 9 338 473 347 1.03 0.73 860 0.40 I‐90 Allston Interchange Project Transit Crowding Analysis PM Peak Period (3‐6 PM) Results

2040 DESIGN YEAR NO BUILD (NB) PM Peak Period Service Characteristics PM Peak 1 Hour Service Info PM Peak 3‐Hr Service Info

2040 DESIGN YEAR BUILD PM Peak Period Service Characteristics PM Peak 1 Hour Service Info PM Peak 3‐Hr Service Info

Car Set Set Service Model Volume / Volume / Cars / Seated Car Policy Seated Policy Seated Policy Volume ‐ Seated Policy Max 3 Hr Model Peak Hr Commuter Rail Line Service Peak Load Point Set Headway Capacity Max Load Capacity Max Load Frequency Capacity Max Load 1 Hr Capacity Load Volume Factor Worcester/Framingham Line IB Wellesley Hills ‐ Wellesley Farms 6 60.00 185 204 1,110 1,221 1 1,110 1,221 332 0.30 0.27 510 0.65 Worcester/Framingham Line OB Boston Landing ‐ Newtonville 6 22.50 185 204 1,110 1,221 3 2,960 3,256 1,910 0.65 0.59 2,850 0.67 Green B Branch IB ‐ Aboveground Blandford ‐ Kenmore 3 6.00 45 100 135 300 10 1,352 3,005 1,771 1.31 0.59 4,540 0.39 Green B Branch OB ‐ Aboveground BU East ‐ BU Central 3 6.70 45 100 135 300 9 1,211 2,691 2,358 1.95 0.88 5,360 0.44 Green B Branch IB ‐ Underground Boylston ‐ Park 3 1.51 45 100 135 300 40 5,385 11,966 3,748 0.70 0.31 9,610 0.39 Green B Branch OB ‐ Underground Arlington ‐ Copley 3 1.54 45 100 135 300 39 5,255 11,677 3,577 0.68 0.31 8,130 0.44 Harvard Shuttle Bus IB Stadium Way/West Ave ‐ Stadium Way/Cambridge St 1 5.00 39 55 39 55 12 468 655 108 0.23 0.16 220 0.49 Harvard Shuttle Bus OB Stadium Way/Cambridge St ‐Stadium Way/West Ave 1 5.00 39 55 39 55 12 468 655 88 0.19 0.13 220 0.40 Kendall Shuttle Bus IB Stadium Way/West Ave ‐ Stadium Way/Cambridge St 1 5.00 39 55 39 55 12 468 655 103 0.22 0.16 210 0.49 Kendall Shuttle Bus OB West Station ‐ Central Square 1 5.00 39 55 39 55 12 468 655 24 0.05 0.04 60 0.40 LMA Shuttle Bus IB LMA ‐ West Station 1 5.00 39 55 39 55 12 468 655 255 0.54 0.39 520 0.49 LMA Shuttle Bus OB West Station ‐ LMA 1 5.00 39 55 39 55 12 468 655 136 0.29 0.21 340 0.40 66 IB 1 9.00 57 80 57 80 7 380 532 299 0.79 0.56 730 0.41 66 OB 1 9.47 57 80 57 80 6 361 506 252 0.70 0.50 630 0.40 64 IB 1 30.00 39 55 39 55 2 78 109 269 3.45 2.47 530 0.51 64 OB 1 25.71 39 55 39 55 2 91 127 140 1.53 1.10 210 0.67 70/70A IB 1 11.25 39 55 39 55 5 208 291 309 1.49 1.06 630 0.49 70/70A OB 1 10.59 39 55 39 55 6 221 309 168 0.76 0.55 390 0.43 57/57A IB 1 7.83 39 55 39 55 8 299 419 200 0.67 0.48 530 0.38 57/57A OB 1 6.92 39 55 39 55 9 338 473 271 0.80 0.57 670 0.40 I‐90 Allston Interchange Project Transit Crowding Analysis PM Peak Period (3‐6 PM) Results

2040 DESIGN YEAR BUILD W/ COMM AVE CONNECTION PM Peak Period Service Characteristics PM Peak 1 Hour Service Info PM Peak 3‐Hr Service Info

Car Set Set Service Model Volume / Volume / Cars / Seated Car Policy Seated Policy Seated Policy Volume ‐ Seated Policy Max 3 Hr Model Peak Hr Commuter Rail Line Service Peak Load Point Set Headway Capacity Max Load Capacity Max Load Frequency Capacity Max Load 1 Hr Capacity Load Volume Factor Worcester/Framingham Line IB Wellesley Hills‐ Wellesley Farms 6 60.00 185 204 1,110 1,221 1 1,110 1,221 332 0.30 0.27 510 0.65 Worcester/Framingham Line OB Boston Landing ‐ Newtonville 6 22.50 185 204 1,110 1,221 3 2,960 3,256 1,903 0.64 0.58 2,840 0.67 Green B Branch IB ‐ Aboveground Blandford ‐ Kenmore 3 6.00 45 100 135 300 10 1,352 3,005 1,798 1.33 0.60 4,610 0.39 Green B Branch OB ‐ Aboveground BU East ‐ BU Central 3 6.70 45 100 135 300 9 1,211 2,691 2,407 1.99 0.89 5,470 0.44 Green B Branch IB ‐ Underground Boylston ‐ Park 3 1.51 45 100 135 300 40 5,385 11,966 3,748 0.70 0.31 9,610 0.39 Green B Branch OB ‐ Underground Arlington ‐ Copley 3 1.54 45 100 135 300 39 5,255 11,677 3,599 0.68 0.31 8,180 0.44 Harvard Shuttle Bus IB Stadium Way/West Ave ‐ Stadium Way/Cambridge St 1 5.00 39 55 39 55 12 468 655 74 0.16 0.11 150 0.49 Harvard Shuttle Bus OB Stadium Way/Cambridge St ‐Stadium Way/West Ave 1 5.00 39 55 39 55 12 468 655 84 0.18 0.13 210 0.40 West Station ‐ LMA ‐Kendall LMA ‐ Ruggles 1 5.00 39 55 39 55 12 468 655 44 0.09 0.07 110 0.40 Kendall ‐ LMA ‐ West Station LMA ‐ West Station 1 5.00 39 55 39 55 12 468 655 108 0.23 0.16 220 0.49 66 IB 1 9.00 57 80 57 80 7 380 532 331 0.87 0.62 810 0.41 66 OB 1 9.47 57 80 57 80 6 361 506 280 0.78 0.55 700 0.40 64 IB 1 30.00 39 55 39 55 2 78 109 269 3.45 2.47 530 0.51 64 OB 1 25.71 39 55 39 55 2 91 127 146 1.61 1.15 220 0.67 70/70A IB 1 11.25 39 55 39 55 5 208 291 304 1.46 1.05 620 0.49 70/70A OB 1 10.59 39 55 39 55 6 221 309 168 0.76 0.55 390 0.43 57/57A IB 1 7.83 39 55 39 55 8 299 419 204 0.68 0.49 540 0.38 57/57A OB 1 6.92 39 55 39 55 9 338 473 267 0.79 0.56 660 0.40

2040 DESIGN YEAR BUILD W/ STADIUM WAY TRANSIT‐ONLY SOUTHBOUND (SB) PM Peak Period Service Characteristics PM Peak 1 Hour Service Info PM Peak 3‐Hr Service Info

Car Set Set Service Model Volume / Volume / Cars / Seated Car Policy Seated Policy Seated Policy Volume ‐ Seated Policy Max 3 Hr Model Peak Hr Commuter Rail Line Service Peak Load Point Set Headway Capacity Max Load Capacity Max Load Frequency Capacity Max Load 1 Hr Capacity Load Volume Factor Worcester/Framingham Line IB Wellesley Hills ‐ Wellesley Farms 6 60.00 185 204 1,110 1,221 1 1,110 1,221 332 0.30 0.27 510 0.65 Worcester/Framingham Line OB Boston Landing ‐ Newtonville 6 22.50 185 204 1,110 1,221 3 2,960 3,256 1,923 0.65 0.59 2,870 0.67 Green B Branch IB ‐ Aboveground Blandford ‐ Kenmore 3 6.00 45 100 135 300 10 1,352 3,005 1,775 1.31 0.59 4,550 0.39 Green B Branch OB ‐ Aboveground BU East ‐ BU Central 3 6.70 45 100 135 300 9 1,211 2,691 2,354 1.94 0.87 5,350 0.44 Green B Branch IB ‐ Underground Boylston ‐ Park 3 1.51 45 100 135 300 40 5,385 11,966 3,748 0.70 0.31 9,610 0.39 Green B Branch OB ‐ Underground Arlington ‐ Copley 3 1.54 45 100 135 300 39 5,255 11,677 3,595 0.68 0.31 8,170 0.44 Harvard Shuttle Bus IB Stadium Way/West Ave ‐ Stadium Way/Cambridge St 1 5.00 39 55 39 55 12 468 655 152 0.32 0.23 310 0.49 Harvard Shuttle Bus OB Stadium Way/Cambridge St ‐Stadium Way/West Ave 1 5.00 39 55 39 55 12 468 655 92 0.20 0.14 230 0.40 Kendall Shuttle Bus IB Stadium Way/West Ave ‐ Stadium Way/Cambridge St 1 5.00 39 55 39 55 12 468 655 59 0.13 0.09 120 0.49 Kendall Shuttle Bus OB West Station ‐ Central Square 1 5.00 39 55 39 55 12 468 655 20 0.04 0.03 50 0.40 LMA Shuttle Bus IB LMA ‐ West Station 1 5.00 39 55 39 55 12 468 655 98 0.21 0.15 200 0.49 LMA Shuttle Bus OB West Station ‐ LMA 1 5.00 39 55 39 55 12 468 655 60 0.13 0.09 150 0.40 66 IB 1 9.00 57 80 57 80 7 380 532 294 0.77 0.55 720 0.41 66 OB 1 9.47 57 80 57 80 6 361 506 252 0.70 0.50 630 0.40 64 IB 1 30.00 39 55 39 55 2 78 109 264 3.39 2.42 520 0.51 64 OB 1 25.71 39 55 39 55 2 91 127 140 1.53 1.10 210 0.67 70/70A IB 1 11.25 39 55 39 55 5 208 291 309 1.49 1.06 630 0.49 70/70A OB 1 10.59 39 55 39 55 6 221 309 168 0.76 0.55 390 0.43 57/57A IB 1 7.83 39 55 39 55 8 299 419 197 0.66 0.47 520 0.38 57/57A OB 1 6.92 39 55 39 55 9 338 473 267 0.79 0.56 660 0.40