Empirical Evidences of Dynamic Speed Limit Impact on a Metropolitan Freeway

The claimed benefits of dynamic speed limit (DSL) strategies (e.g. increase of throughput or speed homogenization) still remain as a controversial topic. The present paper presents an empirical assessment of DSL policies with aggregated traffic flow data.A DSL system installed on the C-32 metropolitan freeway accessing the city of Barcelona (Spain) since February 2009 allows a before-after comparison between non-DSL and DSL scenarios. In order to achieve this objective, a new methodology is proposed. It includes speed-based data stratification for assessing the DSL drivers’ compliance, detecting the causes which motivate drivers’ behavior and the main DSL algorithms inefficiencies. In addition, a computation-friendly approach for searching stationary traffic periods is developed, differing from Cassidy (1998). The whole process permits to obtain a clear fundamental diagram characterization under DSL strategies.The paper proves that when speed limits are lower than free-flow speed, the slope of the free-flow branch of the fundamental diagram is lowered proportionally. This implies that critical occupancies are higher than in their non-DSL analogues. Remarkably, almost any capacity increase is reported in the vicinity of the analyzed bottleneck.