Wagner 1 10_14
-
Upload
trec-at-psu -
Category
Documents
-
view
49 -
download
0
description
Transcript of Wagner 1 10_14
![Page 1: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/1.jpg)
A Benefit-‐Cost Evalua1on Method for Transit
Stop Removal
Zef Wagner Robert L. Bertini
Portland State University
Presented at the OTREC Transportation Seminar
Portland, OR January 10, 2014
![Page 2: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/2.jpg)
Overview
§ Introduction § Research § Methodology § Application § Next Steps
![Page 3: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/3.jpg)
Introduc1on
“Ugh! This bus has way too many stops!”
“Hey! Don’t take away my stop!”
![Page 4: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/4.jpg)
Introduc1on
Speed Access
Ridership Coverage
Wider Spacing Closer Spacing
Reliability Proximity
![Page 5: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/5.jpg)
Introduc1on
Close Spacing = Duplicate Coverage
Wide Spacing = Coverage Gaps
![Page 6: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/6.jpg)
Introduc1on
Stop Spacing in Practice = ~1/8-mile (or less!)
Stop Spacing Standards = ~1/4-mile
![Page 7: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/7.jpg)
Research
§ Exis%ng research focused on line-‐level analysis to determine op%mal average stop spacing
§ Needed: a stop-‐level analysis method to determine which specific stops to remove
![Page 8: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/8.jpg)
Methodology
§ Calculate Benefit-Cost Ratio (B/C) for removing each stop
§ B/C > 1 = candidate
for stop removal
![Page 9: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/9.jpg)
Methodology
§ B = (# of through riders on vehicle) x (time saved by not serving stop)
§ C = (# of riders using stop) x (additional time
to access nearest remaining stop) § Passenger-minutes saved vs. passenger-
minutes lost
![Page 10: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/10.jpg)
Methodology
§ Average load and stop-level ridership § Distances between stops § Value of time ratio § Average walking speed § Average time lost per stop (not including
dwells)
![Page 11: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/11.jpg)
Assump1ons
§ Bus serves all stops on every trip § All passengers migrate to nearest stop § Perfect street grid with small blocks
![Page 12: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/12.jpg)
Holland
Morgan
Bryant
Dekum
Rosa Parks
Ainsworth
Jarrett
Killingsworth
Sumner Alberta
Wygant
Prescott
Mason
Beech
Failing
Fremont
Fargo
Morris
Brazee
Knott
Tillamook
Lombard
§ TriMet Line 6 § 20 stops § Outbound § PM Peak § Fall 2011
§ Why chosen? § High ridership § Close spacing § Grid streets § Many stops
Applica1on
![Page 13: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/13.jpg)
Results Holland
Morgan
Bryant
Dekum
Rosa Parks
Ainsworth
Jarrett
Killingsworth
Sumner Alberta
Wygant
Prescott
Mason
Beech
Failing
Fremont
Fargo
Morris
Brazee
Knott
Tillamook
Lombard
Stop Loca1on B/C
Brazee 1.9
KnoA 0.9
Morris 1.4
Fargo 5.3
Fremont 2.0
Beech 4.4
Failing 1.4
Mason 1.9
PrescoA 2.0
Wygant 2.0
Alberta 0.6
Sumner 2.1
Killingsworth 0.5
JarreA 1.3
Ainsworth 0.3
Rosa Parks 1.1
Dekum 1.4
Bryant 3.3
Morgan 7.4
Holland 3.3
Lombard
Cross Street B/C Holland 1.6 Morgan 3.7 Bryant 1.6 Dekum 0.7 Rosa Parks 0.6 Ainsworth 0.2 Jarrett 0.7 Killingsworth 0.2 Sumner 1.0 Alberta 0.3 Wygant 1.0 Prescott 1.0 Mason 0.9 Failing 0.7 Beech 2.2 Fremont 1.0 Fargo 2.7 Morris 0.7 Knott 0.5 Brazee 0.9
§ 5 stops have B/C>1
§ 3 adjacent stops
§ Remove outer stops first
§ Remove stops, then re-evaluate after one year
![Page 14: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/14.jpg)
Sensi1vity Analysis
§ Va = Value of Time § Tr = Time Lost/
Stop § B/C = Benefit/Cost
§ A range of values still support the same conclusion
![Page 15: Wagner 1 10_14](https://reader036.fdocuments.net/reader036/viewer/2022062320/559429561a28abd05a8b45bf/html5/thumbnails/15.jpg)
Next Steps
§ Stop Probability
§ Network Analysis
70% 30% 100% 60% 40%
§ Operational Benefits