Evaluating the Effect of Road Geometry and Posted Speed Limit on 85 Percentile Speed of Addis Ababa Urban Roads

No Thumbnail Available



Journal Title

Journal ISSN

Volume Title


Addis Ababa University


The main objective of the study was evaluating the effect of road geometry and posted speed limit on 85 percentile speed on urban road of Addis-Ababa. The output of this study is used for transport planner, geometric designer and traffic management agency for evaluation of existing urban road and identifying factors which are used for calming speed of vehicles. The data was collected by field survey, manual pedestrian and side parking vehicle counting and video graphic survey for extracting operating speed calculation data for half hour duration on 23 selected road section. For all selected section, a total of 3779 free flow vehicle were obtained for 85 percentile speed determination in which on the average 160 free flow vehicles per section. The recorded video was played using VLC media player by play and pause in order to obtain entry and exit time of each vehicle for calculation of spot speed. After extracting all data and computing operating speed, the 85 percentile speed was determined using speed frequency distribution graph. After all necessary field data are collected and manipulated, multiple linear regression model was developed using STATA-14 software for predicting 85 percentile speed to use as initial guide line for planning and designing of urban roads. For this study road environment (number of lane, longitudinal grade width and height of median, walking width, road side object density and land use) and traffic factors (pedestrian volume, vehicle hourly volume and parking density) were considered. Based on the result of analysis, the number of lane and road side object density were found to be increase the 85th speed by 3.43km/hr. and 0.425 km/hr. respectively whereas side parking density and hourly traffic volume are found to be reducing 85th speed by 1.27 and 0.018 km/hr. respectively. The degree of determination and root mean square error of the model were also 83.9% and 3.83km/hr. respectively. In order to check the validation of multiple linear regression model parameter effect, driver attitudinal survey was conducted. Based on the collected questionnaire data, factor analysis, was conducted. Factor analysis was done to form few factors from 15 original speeding reason based on correlation between each reason. Based on the result of analysis, 15 original speeding reason was grouped in to four few factors which are understandable and easy for interpretation. These are roadway and variable control condition, lack of knowledge and illegal driver behavior, speed limit condition and over load task of drivers.



85 percentile speed, posted speed limit, speeding frequency, factor analysis