Influence of rainfall on quality of service at multilane roundabouts and its time headway implications.

Abstract

Roundabouts, or traffic circles as they are often called in South Africa, are priority intersections with a unique yield rule. Drivers approaching the roundabout must give way to those that are already circulating the central island. The fixed features and yield rule do not change relative to rainfall; however, vehicular flow rate and driver behaviour are often affected by ambient conditions like rainfall among others. Consequently, in this the study the influence of rainfall on the quality of service delivery at multilane roundabouts and their implications for time headways have been investigated. Based on the hypothesis that rainfall, irrespective of intensity, has adverse effects on the quality of service delivery and time headway at roundabouts, an impact study was carried out in Durban, South Africa. Entry, circulating traffic flow rate and rainfall data were collected at four selected sites in Durban, South Africa. Over one million traffic volume data was collected during the August 2016 to February 2017 rainy season. The key selection criterion is proximity to an active rain gauge. Empirical data were collected continuously for six weeks on each selected roundabout. Rainfall data were collected from surface rain gauge stations with a distance range of 0.75km – 1.18km from the selected sites. Three classes of rain precipitation intensity (i) (light rain, i < 2.5mm; moderate rain, 2.5mm < i ≤10mm; and heavy rain 10mm < i ≤ 50mm) were considered. Very heavy rain, with an intensity greater than 50mm/h, was not considered because of associated drag force and aquaplaning which might be difficult to separate from the rainfall effect. Daylight data were separated into peak and off-peak traffic periods. Peak period data were used to develop a quality of service criteria table and the off-peak data were used to determine traffic flow rate performance. Passenger car equivalent (PCE) values used to convert vehicles per hour to pce per hour was investigated for analytical suitability given rainy conditions. Entry flow rate was used as a function of circulating flow rate to model entry capacity and, hence, determine the reserve capacity. Initially, both linear and exponential models were used, in turn, to test for analytical suitability. Linear model was the preferred after exponential function failed empirical tests. Linear function was used to model the relationships between entry and circulating traffic flow rates. The ensuing entry capacity was also used in conjunction with headway and degree of saturation to estimate entry delay under dry, light, moderate and heavy rainy conditions. The impact study reasons that quality of service is not the same as level of service, hence, the criteria table cannot be the same. This is a clear departure from Highway Capacity Manual (HCM) prescription for roundabout level of service criteria table. The novel quality of service criteria table prescribed in this thesis, has delay and reserve capacity as the xxxi key determinants of service grade. It is also referred to as Functional Quality of service (FQS) in the thesis. FQS criteria table was developed for each study site and used to assess their service delivery. The criteria table was divided into six classes (A to F), where A is the best grade and F is the worst. In any case, traffic performances were analysed and results show that; i) there is no significant difference between South Africa passenger car equivalent values and those estimated in the study; ii) the novel criteria table developed in the study is an effective determinant of FQS delivery at roundabouts; iii) entry traffic flow rate rates decreased because of rainfall and by extension induced a reduction in quality of service delivery at all surveyed sites; iv) entry delay and attendant queue increased during rainfall; v) time headway increased and entry reserve capacity decreased because of rainfall. It has been concluded that rainfall has an adverse effect on the FQS and also, that heavy rainfall has the most significant impact on FQS at roundabouts. It is proposed that in future research, on roundabout entry capacity estimation based on polynomial quadratic function where the single-variable quadratic polynomial would have density as the independent variable and flow rate as the dependent be considered.