Abstract
An acknowledged challenge with protected bike lanes in the United States is that even though the segment is separated from traffic, bicyclists must merge or interact with turning traffic at intersections unless all movements are signalized. This paper presents a comparison of five designs for protected bike lanes at intersections without bicycle signals. The designs represent ideas for how to mix and interact bicycles and motor vehicles. The designs communicate to road users how this interaction is to occur with lane striping, green pavement markings, shared lane-use markings (sharrows), and vertical flexpost delineators. The designs are divided into “mixing zones” and “turning zones with a through bike lane” for evaluation purposes. The paper compares and contrasts the designs with the use of observed user behavior (from 78 h of video analyzed, in which 6,082 bicyclists and 7,574 turning vehicles were observed) and self-reported behaviors and comprehension (from 1,245 nearby residents and 690 intercepted bicyclists). Overall, the evaluation suggests that there are benefits to clear demarcation of the entry to the merge zone for bicycles and cars and to creating a semiprotected through bicycle lane.