In 2009, a series of traditional intersections near Bellingham, Washington, was converted to double-lane roundabouts. The road also was widened from one to two lanes, and a median with a cable barrier was added. The current study evaluated the impact of roundabout conversions at two intersections, one signalised and one with two-way stop signs, on public attitudes, traffic performance, fuel consumption, vehicle emissions, and safety. Analyses also examined whether older drivers may have avoided the roundabouts by taking an alternative route. Telephone surveys were conducted before and approximately six months and one year after the roundabout conversions. Before/after traffic performance, fuel consumption, and vehicle emissions were analysed at each intersection. Poisson regression models examined before/after changes in crash rates per entering vehicles, relative to changes at comparable traditional intersections. Before/after age group distributions of drivers along the study corridor and a parallel road were examined. The proportion of drivers who favoured the roundabouts increased from 34% before construction to 51% six months after and 70% more than one year after. Drivers who opposed the roundabouts most often said it was because the roundabouts were unsafe (31% before, 33% six months after, 27% more than one year after). More than one year after construction, more than 40% of drivers did not believe signs and pavement markings adequately conveyed information about the appropriate speed at roundabouts, the rules for right-of-way, and the appropriate way to navigate the roundabouts in the presence of large trucks. Widening the roadway improved traffic operations at both intersections, and the conversion to roundabouts achieved further improvements. Substantial declines occurred at both intersections in delays on the minor roads (33%, 90%), the proportion of vehicles that waited in queues (35%, 43%), and lengths of queues on minor roads (50%, 64%). Overall intersection delays increased (12%, 22%), due to slightly longer delays on the major approaches as vehicles slowed to enter the roundabouts. The degree of congestion in the most congested lane (based on the ratio of traffic volume to lane capacity) increased slightly (3%) at one intersection and declined by 32% at the other. The roundabout conversions were associated with reductions in fuel consumption (23%, 34%) and vehicle emissions (0%-45%). Relative to the crash rates that would have been expected without the roundabout conversions, based on crash trends at the comparison traditional intersections, the roundabout conversions were associated with reductions in the rates of injury and fatal crashes combined (34%, 100%) and increases in the rates of non-injury crashes (13%, 502%). Only the increase in the non-injury crash rate (502%) at the roundabout converted from a two-way stop sign was significant. The odds that older drivers (estimated age 70 and older) travelled on the study corridor versus an alternative route after the roundabout conversion was 0.32 times the odds before the conversion. This difference was statistically significant, although the proportion of older drivers on these roads was very small (4% before on both roads; 1% on the study corridor, and 3% on the alternative route after). The conversion of stop sign- and traffic signal-controlled intersections to double-lane roundabouts was associated with improvements in most measures of traffic performance, leading to reductions in fuel consumption and vehicle emissions. Drivers’ acceptance of the roundabouts improved over time, and injury crashes declined, although not significantly. These findings are consistent with prior research finding substantial traffic, environmental, and injury-reduction benefits of single-lane roundabouts. However, there were indications that the greater complexity of multi-lane roundabouts may present challenges. Some confusion about navigating the roundabouts persisted one year after construction. There was evidence that some older drivers may have taken an alternative route to avoid them, and non-injury crash rates increased. (Author/publisher)
Abstract