Toronto: Gardiner Expressway Study

Toronto: Gardiner Expressway Study

Paramics 2009 UGM: NewarkOctober 5, 2009

Project Introduction

• Frederick G. Gardiner Expressway– Elevated freeway through downtown Toronto

• Lower Don Lands– Former industrial land redeveloped into

residential and commercial uses• Gardiner is seen as a barrier to the

redeveloped LDL– Investigate solutions

Project Introduction

• Joint Clients– City of Toronto and Waterfront Toronto– City owns the Gardiner– Waterfront Toronto is charged with revitalisation

of the waterfront• Sensitive Project

Project Scope and Goals

• $6M Planning and Design Study– Sub to Dillon Consulting– ~$300k Microsimulation budget

• Investigate four design options for treatments east of Jarvis Street– Do Nothing, Ameliorate, Remove, Submerge– Five construction staging models

• Goal is to investigate ways of reconnecting the redeveloped Lower Don Lands to the City

Microsimulation Model Area

• 5 mi2 (13 km2)• Dense, urban development

SPAD

INA

DUNDAS

WOODB

INE

Methodology

• Create existing conditions– Road selection– Zone system creation– Include transit and pedestrians

• Matrix Estimation• Calibration / Validation• Alternative Testing• Construction Staging

Existing Conditions

• Most roadways in study area– Arterial and up, some collectors, and few locals

• Coding issues with Lake Shore Boulevard

Existing Conditions

• Arup Model as a starting point• Approximately double model area

ARUPJACOBS

Zone System

• Layout zones on paper

Zone System• Simple one zone

per block with access to all roads

• Simplifies creation and coding

• Makes for very difficult matrix estimation

Zone System• One zone per

connector on each block face

• Allows maximum control

• Model becomes unwieldy with so many zones

Existing Conditions

• Transit– Streetcars

• Spadina and Queens Quay – Median-running• Other streets – In mixed traffic

– Buses– Subway

• Not included in our model

• Pedestrians– Taken only from Arup model, no new ped work

Streetcar Coding

• Spadina and Queens Quay– Median-running

• Median Lane Operation– Paramics limitation that similar

movements must be in adjacent lanes– This is an issue where the transit

vehicle through lane is adjacent to a left-turn lane

– Fortunately, coding the through movement in the transit lane as ‘barred’ works perfectly• Transit vehicle moves with normal

traffic ‘through’ green

Streetcar Coding

• Remainder of Streetcar network is in mixed traffic– Typically runs in the left-hand

lane of a two-lane per direction facility

• Congestion issues due to frequent stops– Both lanes in the streetcar travel

direction must stop when loading/unloading passengers

– New feature built into Paramics

STOP

Matrix Estimation

• Counts• Travel Demand Model O/D data• Screenlines

Matrix Estimation

• Issues– Estimation of congested conditions• Counts at congested locations are less useful• Use counts upstream of issues to help estimate

demand– Grid network adds to the challenge• Many parallel routes and relatively short blocks

Calibration / Validation• Counts• Travel Times• Screenlines• Major Queue

Locations– Gardiner and

on/off ramps

Calibration/Validation

• Issues– Major queues on Gardiner• WB in PM, EB in AM• Replication of WB queue may require network

extension or link speed modification at external station

STUDYAREA

CONGESTIONSOURCE

Future Alternatives

• Do Nothing• Ameliorate

JARVIS

EASTERN

• Remove (bring to ground)• Submerge

Construction Staging

• With a preferred alternative selected– Build a series of 5 incremental networks that take

the geometry to the final design– Stages given to us for analysis

• Provide results and recommendations

Conclusions

• Most challenging work will be: – Matrix Estimation– Calibration / Validation

• Questions or Comments?– Adam Lanigan adam.lanigan@jacobs.com