Subject: Re: Using SIDRA in Roundabout design

GDOT does the same.   The opening year is the year when the  project is expected to be open to traffic, and design year is opening year + 20 years.  We also require use of a second analysis method; the below is from Section 8.2.2 of our Design Policy Manual<http://www.dot.ga.gov/doingbusiness/PoliciesManuals/roads/DesignPolicy/GDOT-DPM-Chap08.pdf>.

*    Section 4, Operational Analyses:  include operational analyses using peak hour traffic volumes for each design alternate, for opening and design years. The results of each analysis should be presented by lane group in terms of volume-to-capacity ratio, average<http://www.dot.state.ga.us/travelingingeorgia/roundabouts/Pages/AnalysisTools.aspx> control delay, level of service, and 95th percentile queue<http://www.dot.state.ga.us/travelingingeorgia/roundabouts/Pages/AnalysisTools.aspx>.  Based on the results of these analyses the performance of each alternate should be evaluated, and intersection types providing adequate performance identified.  Further guidance on evaluating the operational performance of roundabouts can be found in  HCM 2010<http://www.trb.org/Main/Blurbs/Highway_Capacity_Manual_2010_HCM2010_164718.aspx>.   See also Chapter 4 of  NCHRP<http://www.trb.org/Design/Blurbs/Roundabouts_An_Informational_Guide_Second_Edition_164470.aspx>672<http://www.trb.org/Design/Blurbs/Roundabouts_An_Informational_Guide_Second_Edition_164470.aspx>.

Analyses should be performed using more than one analysis methodology to identify a range of expected performance during the design period (i.e., opening to design years).  For example, analyses can be performed using the GDOT Roundabout Analysis Tool <http://www.dot.state.ga.us/travelingingeorgia/roundabouts/Pages/AnalysisTools.aspx> to implement the HCM 2010 <http://www.trb.org/Main/Blurbs/Highway_Capacity_Manual_2010_HCM2010_164718.aspx> method and the "SIDRA Standard" method (suggest environmental factor of 1.1 for the design year) using the software package SIDRA Intersection or the empirical method using the software package ARCADY.  An environmental factor of 1.2 is suggested for opening year analyses.

A simulation software package, such as VISSIM, should be used when modeling of a network of closely spaced intersections is necessary.

Daniel G. Pass, P.E.

Marcus,

I was hoping someone else would take an initial shot at this one but I haven't seen anything so far so here is my response.  Let me state that my opinions are not likely the same as others on this listserv relating to HCM 2010.

To say that I am not a fan of the HCM 2010 roundabout model would be putting it mildly.  My biggest issue is that some organizations are going to use it as is - even though it refers to calibrating it not many organizations have the data to do so.  Using the HCM 2010 model as is could be perceived as being "safe" - by not pushing the gap parameters the organization is assured that the roundabout will work adequately from a LOS perspective.  To quite a few people this makes sense - my concern is that in order to get good LOS with the default model many roundabouts are going to have lanes added that may or not be necessary in the future.  Or, they might not even progress the roundabout alternative because of the worse operations predicted by a model that was based on drivers from 8 years or so ago.  This is being addressed by another FHWA research project.  I am hoping that this recent update will include some attempt at predicting decent gap acceptance values for 10 and 20 year projections.  I never understood the whole thought process of the HCM 2010 roundabout model - what good is it to me when looking 20 years into the future to use gap acceptance values from 8 years ago without some adjustment for driver familiarity....

Personally, I would rather have a single lane roundabout that pushes or even slightly exceeds the capacity limit for an hour or two 16 years from now than a 2 laner with LOS A to C for 20 years after it was built.  Many recent studies have shown capacities quite a bit higher than HCM 2010 predictions.

So, to answer your question:

1st Choice - Run SIDRA standard model with EF set to 1.2 for today and 1.1 for future traffic projections.  At least this still takes geometry in account.

Or

2nd Choice - Run HCM 2010 with at least the most aggressive settings you can find today and then still consider adjusting them a bit more for further capacity improvements

Hopefully this will get a few others with differing opinions out there to respond since it would help all of us to hear other opinions or how others are using the software.

Regards,

Howard