Graduation Year

2020

Document Type

Thesis

Degree

M.S.C.E.

Degree Name

MS in Civil Engineering (M.S.C.E.)

Degree Granting Department

Engineering

Major Professor

Robert L. Bertini, Ph.D.

Co-Major Professor

Xiaopeng Li, Ph.D.

Committee Member

Michael Maness, Ph.D.

Keywords

Freeway Capacity, Probe Detector, Vehicle Delay

Abstract

Prevailing traffic conditions were examined for a bottleneck that occurred on northbound SR-91 (Florida’s Mainline Turnpike) during the mass evacuation for Hurricane Irma in September 2017. Radar detector count, occupancy, and speed data polling at 1-minute intervals were collected from the Regional Integrated Transportation Information System (RITIS) and employed to identify the distinct periods one particular bottleneck was active at a service plaza along SR-91. With this rich data, three distinct periods were identified in which the bottleneck was active at a service plaza off-ramp, lasting a total of 27.5 hours during this colossal evacuation. To identify the activation time of the bottleneck, its duration, and the unique stationary traffic features present, curves of cumulative vehicle count and occupancy were utilized. The curves can reveal periods where excess vehicle accumulation and delay persist between successive radar detectors. Results demonstrate that queued traffic propagated from an off-ramp leading to the service plaza. The discharge traffic features, just downstream of the off-ramp, presented unqueued, uniform conditions that exhibited much lower flows than what is usually anticipated for a facility with limited access and high operational speeds. The findings from this thesis present an untouched area for bottleneck analysis during evacuations, which will provide crucial considerations for evacuations planning, traffic management and operations, design, and modeling. The bottleneck studied in this thesis only became active during an evacuation and not during normal recurrent operations, which sheds light on the implications of not including historical events such as this for future planning purposes.

Share

COinS