Page Contents
Why is SIDRA INTERSECTION the best software for roundabout capacity analysis? - Comparison of SIDRA, HCM 2010 and TRL Roundabout Capacity Models
Why is SIDRA INTERSECTION the best software for roundabout capacity analysis?
You may have seen a claim from another software developer that the new version of their roundabout capacity software package “is the only commercially available tool to offer full geometric and gap acceptance modelling capability within a single product”.
For at least two decades now, SIDRA INTERSECTION software has used the gap-acceptance methodology for roundabout capacity estimation where gap-acceptance parameters are estimated from the roundabout geometry. In short, SIDRA INTERSECTION has employed a combined (hybrid) geometry and gap-acceptance modelling approach in order to take into account the effect of roundabout geometry on driver behaviour directly through gap-acceptance modelling. This approach accepts the importance of roundabout geometry but states that roundabout geometry alone is not good enough and modelling of driver behaviour is needed for roundabout capacity estimation, just as it is needed for modelling the capacity of any other intersection type (signals, sign control).
There has been a lot of debate about the relative benefits of models based on "gap acceptance theory" vs "empirical models" in the past, with claims that gap-acceptance modelling does not work for roundabouts. Much of this debate has been misleading due to simplistic model categorization based on the suggestion that these modeling approaches are mutually exclusive. Opinion has also been expressed that it does not matter which one of these modelling approaches is used. This has overlooked the fact that SIDRA INTERSECTION has used the gap-acceptance method not only for capacity but also for performance estimation includingunique equations for back of queue and stop rate based on modelling of gap-acceptance cycles.
With the aim to enhance understanding of the fundamental aspects of the different roundabout capacity models available around the world, Dr Rahmi Akçelik presented two papers at the TRB International Roundabout Conference in Carmel, Indiana, USA in May 2011, discussing three well-known analytical models of roundabout capacity, namely the Australian SIDRA INTERSECTION model, the US Highway Capacity Manual 2010 (HCM 2010) model and the UK TRL (linear regression) model. Dr Akçelik pointed out the common features as well as differences of these models, and presented a detailed table comparing the features of these roundabout capacity models. The features compared include methodology, model level of detail considering entry and circulating lanes, parameters used in the model to represent driver behaviour and roundabout geometry, and model calibration methods. A short version of the model comparison table is presented below. Roundabout examples comparing these three models are given in the below papers by Dr Akçelik.
REFERENCES
TRB. Highway Capacity Manual. Transportation Research Board, National Research Council, Washington, DC, USA, 2010. ("HCM 2010")
KIMBER, R.M. (1980). The Traffic Capacity of Roundabouts. TRRL Laboratory Report 942. Transport and Road Research Laboratory, Crowthorne, Berkshire, UK
Comparison of SIDRA, HCM 2010 and TRL Roundabout Capacity Models
|
SIDRA Model |
HCM 2010 Model |
UK TRL model |
|
Methodology |
||
|
Based-on gap-acceptance theory with empirical (regression) equations to model gap-acceptance parameters. |
Empirical (exponential regression) capacity model with clear basis in gap-acceptance theory. |
Empirical (linear regression) capacity model with no stated basis in traffic theory. |
|
Lane-based model: capacity and performance of individual entry lanes modelled. |
Lane-based model: capacity and performance of individual entry lanes modelled. |
Approach-based model: lane capacity and performance not available. |
|
Driver Behaviour |
||
|
Gap-acceptance parameters (Follow-up Headway, Critical Gap), entry lane-use model and circulating stream bunching represent driver behavior. Driver response times determined. |
Gap-acceptance parameters (Follow-up Headway, Critical Gap) and entry lane-use model represent driver behavior. |
No direct representation of driver behavior. Capacity is sensitive to the circulating flow rate only. |
|
Follow-up Headway and Critical Gap depend on roundabout geometry. |
Follow-up Headway, Critical Gap values are constant. |
Not used. |
|
Follow-up Headway and Critical Gap values are reduced (more aggressive driver behavior) with increased circulating flow rate. |
Follow-up Headway, Critical Gap values are constant. |
Not used. |
|
Roundabout Geometry |
||
|
Average entry lane width |
Not used |
Total entry width |
|
Number of entry lanes |
Number of entry lanes |
Not used |
|
Approach lane disciplines and configuration including bypass lanes |
Approach lane disciplines and configuration including bypass lanes |
Not used |
|
Number of circulating lanes |
Number of circulating lanes |
Not used |
|
Inscribed diameter |
Not used |
Inscribed diameter |
|
Entry radius |
Not used |
Entry radius |
|
Entry angle |
Not used |
Entry angle |
|
Approach short lanes: capacity and overflow into adjacent lane modelled using gap-acceptance cycles and back of queue modeling. |
Not used. Extension to the HCM 2010 model in SIDRA INTERSECTION. |
Approach flaring (approach half width and flare length). |
|
Number of exit lanes (can affect upstream approach lane use) |
Not used |
Not used |
|
Exit short lanes (merge lanes): effect on upstream approach lane utilisation modelled. |
Not used |
Not used |
|
Model Level of Detail - Entry and Circulating Lanes |
||
|
Variations in lane disciplines (exclusive and shared lanes, slip and continuous lanes) can be modeled. |
Variations in lane disciplines can be modeled. |
Variations in lane disciplines cannot be modeled. |
|
Dominant and subdominant entry lanes identified. |
Dominant and subdominant entry lanes identified. |
Entry lanes not identified. |
|
Entry lane flow rates are calculated. |
Entry lane flow rates are calculated. |
Lane flows cannot be modelled. |
|
De facto exclusive lanes are identified. |
De facto exclusive lanes are identified. |
De facto exclusive lanes cannot be identified. |
|
Unequal lane use can be modelled by specifying lane utilization ratios. |
Unequal lane use can be modelled by specifying lane volume percentages. |
Unequal lane use cannot be modelled. |
|
Critical lane v/c ratio (degree of saturation) for a multilane approach is determined. |
Critical lane v/c ratio for a multilane approach is determined. |
Critical lane v/c ratio cannot be determined (only the average v/c ratio for the approach is available). |
|
Number of circulating lanes affects capacity. |
Number of circulating lanes affects capacity. |
Number of circulating lanes does not affect capacity. |
|
Circulating lane flow rates used allowing for unbalanced flows. Amount of queuing before entering circulating stream affects capacity. |
Total circulating flow rate used. Circulating lane flows not used. |
Total circulating flow used. Circulating lane flows not used. |
|
Uses a bunched arrival headway model for the circulating stream. |
Uses a random arrival headway model for the circulating stream. |
No explicit assumptions about circulating stream headways. |
|
Extra bunching to model upstream signal effects allowed. |
Not used. Extension to the HCM 2010 model in SIDRA INTERSECTION. |
Not used. |
|
Unbalanced Flows |
||
|
Capacity is sensitive to Origin-Destination demand flow pattern, lane use and level of queuing on entry lanes. Roundabout modeled with high level of interaction between traffic using all intersection approaches |
Not used. Extension to the HCM 2010 model in SIDRA INTERSECTION. |
Not used (roundabout modeled as a series of T- intersections with no sensitivity to Origin-Destination flow patterns). |
|
Adjustment options exist for high Entry Flow / Circulating Flow ratio (increased entry capacity at very low circulating flow rates due to increased driver aggressiveness level). |
Not used. Extension to the HCM 2010 model in SIDRA INTERSECTION. |
Not used. |
|
Model Calibration |
||
|
Intersection-level or approach-level calibration using Environment Factor. Movement-level calibration using Follow-up Headway and Critical Gap parameters. |
Method described to calibrate the model parameters using known Follow-up Headway and Critical Gap values. |
The y-intercept value of the linear regression capacity function can be adjusted. Some issues exist. |
Disclaimer: In order to avoid misleading statements about particular software packages, the third model in the table has been referred to as the UK TRL model to distinguish it as the original published model as opposed to the ARCADY and RODEL software packages which have implemented it since the software may include some differences from the original model.
14-Day Free Trial
