Analysis of left-turn saturation flow rates and capacity at signalized intersections

Ming Heng Wang, Rahim F. Benekohal

Research output: Contribution to journalArticlepeer-review


This study investigated and developed left-turn saturation flow rate models by considering the operation characteristics of left-turn traffic and the arrival and departure types of the opposing traffic. Methodologies for measuring left-turn saturation flow rates and capacity were proposed for the protected-then-permitted phasing with an exclusive left-turn lane at signalized intersections. The permitted left-turn saturation flow rate was developed on the basis of the opposing nonplatoon flow rate during the unblocked (unsaturated) time instead of using average opposing flow rates during the entire green time. The results of simulation show that the protected left-turn saturation flow rate is about 0.97, 0.96, or 0.95 of the saturated through traffic flow rate when opposed by one, two, or three lanes, respectively. The permitted left-turn saturation flow rate depends not only on the opposing flow rate but also on the number of opposing lanes. For facing multiple opposing lanes, the leftturn saturation flow rate decreased by 8% to 57% compared to facing only a single opposing lane. Even for multiple opposing lanes, additional reductions in the left-turn saturation flow rate, when facing three versus two opposing lanes, ranged from 8% to 18%. The estimation of the left-turn capacity using the proposed models was validated by comparing the capacity from the simulation.

Original languageEnglish (US)
Pages (from-to)96-104
Number of pages9
JournalTransportation Research Record
Issue number2173
StatePublished - Jan 12 2010


  • Highways
  • operations and traffic management
  • pedestrians and bicyclists

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Mechanical Engineering


Dive into the research topics of 'Analysis of left-turn saturation flow rates and capacity at signalized intersections'. Together they form a unique fingerprint.

Cite this