Subject: Roundabout Analysis Methodology

As a public agency operating under budget constraints, we have not been

able to acquire the roundabout analysis software (SIDRA, VISSIM)

utilized by design consultants developing multilane roundabout projects

for us. As a check on the designs being submitted I have been utilizing

the methodology outlined in Appendix M of NCHRP Report 572, Roundabouts

in the United States. This "Draft" HCM methodology utilizes a gap

acceptance model based on observed operating characteristics at

roundabouts in the United States.

The documentation for the methodology indicates observed capacities at

US roundabouts are lower than those observed elsewhere. It further

qualifies the results of the methodology by observing, "At high levels

of conflicting flow...a simplified gap acceptance model may not give

reliable results."

We have a multilane roundabout under design. The arterial entry thru

flows are approximately 1400 vph each in the N/S direction during the PM

peak. Side street entry is approximately 250 vph WB & 200 vph EB during

the same period.

The SIDRA analysis indicates good levels of service in 2018 with 6%

growth in traffic over 10 years. A VISSIM simulation appears to support

the SIDRA results. The Draft HCM methodology indicates the same geometry

doesn't work under 2008 volumes.

When doing signalized intersection analyses I have a fair amount of

confidence in the results from the HCM methodology. With the roundabout

methodology, I don't know what to believe. Are the capacity thresholds

in the Draft HCM roundabout methodology too conservative?

Have the users of SIDRA and VISSIM found them to produce reliable models

of US roundabout operations?

Thanks

Ted Coleman

Principal Civil Engineer/Traffic Design

Rhode Island Department of Transportation