<p>Performance of an intersection affects the total travel time in a traffic network. An intersection with good performance can easily be described by its contribution to lower travel time than in a comparable intersection. The travel time is of importance in the cost calculations when a specific intersection is to be chosen and designed. It is important to be able to make correct and precise performance calculations so that the cost calculations will not be incorrect. For that purpose the software Capcal, which is a traffic model for intersection analysis on performance, is used in Sweden. Capcal is frequently revised but part of it is still based on old theory and data. Therefore is it important to continually validate the model to control its reliability when it comes to traffic conditions of today with increasing traffic. The model can be validated by comparing its results with reality. </p><p>Performance of an intersection without signal control is defined by the so called performance measurements queue length, capacity, degree of saturation, delay and service time. We have studied a few three-way intersections and roundabouts in Norrköping and Linköping to collect these measurement from reality. These field studies were performed by filming and manually measuring of traffic flows. The films were further processed afterwards to collect additional data. The results from the field studies were compared to output from Capcal. </p><p>Generally our study shows a good correspondence between the output of Capcal and the performance measurements from reality. The differences that exist are caused by the reliability of the field studies and what speed that is possible to model in Capcal. The study showed that if the speed limit is modeled it does not always correspond to the real traffic behavior in all of the studied intersections.</p>

Performance of an intersection affects the total travel time in a traffic network. An intersection with good performance can easily be described by its contribution to lower travel time than in a comparable intersection. The travel time is of importance in the cost calculations when a specific intersection is to be chosen and designed. It is important to be able to make correct and precise performance calculations so that the cost calculations will not be incorrect. For that purpose the software Capcal, which is a traffic model for intersection analysis on performance, is used in Sweden. Capcal is frequently revised but part of it is still based on old theory and data. Therefore is it important to continually validate the model to control its reliability when it comes to traffic conditions of today with increasing traffic. The model can be validated by comparing its results with reality. Performance of an intersection without signal control is defined by the so called performance measurements queue length, capacity, degree of saturation, delay and service time. We have studied a few three-way intersections and roundabouts in Norrköping and Linköping to collect these measurement from reality. These field studies were performed by filming and manually measuring of traffic flows. The films were further processed afterwards to collect additional data. The results from the field studies were compared to output from Capcal. Generally our study shows a good correspondence between the output of Capcal and the performance measurements from reality. The differences that exist are caused by the reliability of the field studies and what speed that is possible to model in Capcal. The study showed that if the speed limit is modeled it does not always correspond to the real traffic behavior in all of the studied intersections.