Present study is an extension of earlier work carried out on two-lane two way roads in the two provinces of Pakistan i.e. N-25,

N-55 and N-5 regarding the measure of operating speed and development of operating speed prediction models. Curved sections

of two-lane rural highways are the main location of run-off road accidents. In addition to that the road alignment having

combination of geometric elements may be more harmful to the drivers than the successive features with adequate separation.

This study is carried out on two-lane two- way road along N-65 (from Sibi to Quetta). Three sections are selected for study with

thirty three horizontal curves. Continuous speed profile data was recorded with the help of VBox (GPS based device) which was

attached with a vehicle to detect vehicle position through satellite signals. VBox is new equipment with modern technology in this

field and it helps in recording continuous speed profile and saving of this information on the computer as a permanent record.

Through the regression analysis, models were developed for estimation of operating speed on horizontal curves and on tangent,

and estimation of maximum speed reduction from tangent to curve. The validation of developed model shows compatibility with

the experimental data.

The present study demonstrates the influence of operating speed on capacity of a midblock section of urban road. Speed – flow data collected at 12 midblock sections of 6-lane and 4-lane divided urban arterials in four metropolitan cities of India are analyzed to determine their capacity. Lane capacity was found to vary from 1482 pcu/hr to 2105 pcu/hr. This variation is explained on the basis of city size and driving behavior, which would influence the free flow speed on the road. Free flow speed was also measured at each section and these speed data were used to determine operating speed (85th percentile of free flow speed of standard car) on the road. Lane capacity was found to be strongly related with operating speed on a road and a second degree polynomial model is developed between the lane capacity and operating speed. This model is further validated by collecting speed flow data at two new sections and their capacity was estimated from field data and from the model developed in the study. The predicted capacity was found to be matching with field capacity and the maximum error was 0.10 percent. Operating speed on a road can vary due to road surface condition, side friction or similar other factors. All these will have influence on capacity of the road. The capacity model suggested in the present study can be a useful tool to determine capacity of an urban road from its operating speed data.