Implementing an LOS Replacement Metric in San Francisco
Corresponding Author: Drew Cooper, San Francisco County Transportation Authority
Presented By: Drew Cooper, San Francisco County Transportation Authority
Abstract
Level-of-Service (LOS), the measure of vehicle delay at an intersection or on a roadway, has long been used as an indicator of network performance, focusing on the roadway users’ experience. In California, under the California Environmental Quality Act (CEQA) worsening LOS as a result of a land use development or transportation project is considered environmental impact. This arrangement is problematic for several reasons. First, LOS measures users’ experience or inconvenience, not harm to the environment. Furthermore, and more importantly, it incentives development with worse environmental outcomes by encouraging sprawl: building in low-density or undeveloped areas where there is little traffic and a low likelihood of impacting LOS, over urban infill: building in high-density, developed areas where an LOS impact is likely. Finally, LOS also prioritizes drivers and driving over other users of the transportation network and other modes.
In 2012 the California state legislature passed Senate Bill 743, which initiated a process to remove LOS from the environmental review process and replace it with a measure that more accurately reflects harm to the environment. The Mayor’s Office of Planning and Research, in their draft report, recommend a set of metrics based on vehicle miles traveled, or VMT. Based on this guidance, City and County staff adapted a set of VMT metrics and thresholds for San Francisco and, In Winter of 2016, the City of San Francisco adopted those metrics to replace LOS.
This work combined planners’ insight into how new projects interact within local and regional land use context with quantitative methods using output from the SF-CHAMP travel demand model. This talk will describe, first, the process of multi-agency collaboration to determine a set of metrics for quantifying the environmental traffic impacts of land use development, and second, the use of an activity-based travel demand model to develop thresholds of significance and estimate likely VMT by geography.