Best Practices For Installation Of Rectangular Rapid Flashing Beacons With And Without Median Refuge Islands

Over the last decade, Oregon jurisdictions have systematically installed pedestrian crossing enhancements (PCEs) at crosswalks including rectangular rapid flashing beacons (RRFBs), which have been widely used in many jurisdictions, typically at mid-block locations. However, the design details vary widely. The objectives of this research are to develop guidance that practitioners need about the placement of RRFB beacons in combination with median refuges on three-lane roadways by exploring the effect of refuge medians mounted RRFB displays on driver yielding behavior. This study also developed methods for pedestrian volume estimation at midblock locations and reanalyzed the SPR 778 RRFB data to produce more robust estimates of the safety effectiveness of RRFBs. This study explored driver and pedestrian behavior at 23 RRFB sites with and without median islands and beacons. Generally, high yielding rates were observed at all sites and these rates provide evidence that the RRFB is a useful tool alerting drivers to the presence of pedestrians at crosswalks.The data and analysis do generally indicate that the yielding rates increase with the addition of the median beacons. However, the difference is not a large increase (<5%) and is also not statistically significant. The findings suggest that a median refuge beacon could be considered optional on 3-lane roadways with volumes < 12,000 average daily traffic (ADT). Similarly, the data generally show that for the > 12,000 ADT groups, the addition of the median refuge increases yielding. In addition to driver yielding behavior, there are other reasons to install median islands including pedestrian comfort which should be considered. A linear regression model for estimating pedestrian demand was developed and significant predictors included the percent of low wage workers at the home and work locations, intersection density in terms of multi-modal intersections having four or more legs per square mile, intersection density in terms of auto-oriented intersections per square mile, gross population density, and proportion of census block group employment within ¼ mile of transit stop. Crash modification factor's (CMF’s) were developed for pedestrian and rear-end crashes. With the simple-before analysis, CMF’s of 0.84 (standard error=0.25) and 1.42 (standard error=0.12) were obtained for pedestrian and rear-end crashes respectively. A CMF of 0.71 (standard error=0.20) and a CMF of 1.11 (standard error of 0.06) were obtained for pedestrian and rear-end crashes using the empirical Bayes method.


  • English

Media Info

  • Media Type: Digital/other
  • Edition: Final Report
  • Features: Figures; Maps; Photos; References; Tables;
  • Pagination: 93p

Subject/Index Terms

Filing Info

  • Accession Number: 01741566
  • Record Type: Publication
  • Report/Paper Numbers: FHWA-OR-RD-20-06
  • Contract Numbers: SPR 814
  • Created Date: May 31 2020 5:55PM