The goal of this research was to study if bicycle streets as applied in The Netherlands are traffic safe. The main motivation for this research was the growing number of bicycle streets in The Netherlands, while so far less is known about the traffic safety of those streets. This research was based on the theory that function, design and behaviour of a street and traffic on a street influence each other. The research was split up in three research steps, each focussing on one of the aspects. Those aspects have been studied for eight bicycle streets in The Netherlands, four single-lane and four two-lane bicycle streets. The selection of two different designs made it possible to compare the two common types of bicycle streets. In the first step, the possible functions of bicycle streets have been determined by a literature review. This resulted in a theoretical framework for finding the functions of the eight bicycle streets. The study of the functions shows that the most important functions of the bicycle street are the flow function for cyclists and the access function for motor vehicles. Those functions are conflicting and therefore a source for conflicts between motor vehicles and cyclists on the bicycle street. In the second step, the driving behaviour on the eight selected bicycle streets was investigated by collecting data via road tube counters and observing the traffic on the bicycle streets. The data collected by the road tube counters provided information about the speed and type of the vehicle passing. The observations gave information about the behaviour, position on the road and possible conflicts of road users on the bicycle street. The results of this step showed that a large part of the cars drive faster than the safe speed limit of 30 km/h on the bicycle street. Further was concluded that the ratio cyclists/motor vehicles does not influence the speed of motor vehicles on the bicycle street, but that the width of the profile has a big influence on the speed. Also has been concluded that vehicles parking or stopping on the bicycle street are a serious obstacle for cyclists. The stronger the access function of the street is present, the more vehicles are parking and stopping on the street. The access function of the street should therefore be relatively unimportant compared with the flow function for cyclists. By the use of an interview, knowledge about Human Factors in traffic has been used to test the recognisability of the designs. Based on those two research steps, the outcomes have been tested to the Sustainable Safety principles functionality, homogeneity, predictability and forgivingness. The most important conclusion that has been drawn from this test is that bicycle streets have so many appearances that they are not recognisable for the different road users. The road course of the bicycle street and the driving behaviour of road users on the bicycle street are therefore difficult to predict. The recognisability and predictability can be improved by applying one standard design with recognisable design elements. The homogeneity of the traffic flow could be improved by reducing the speeds of motor vehicles on the street. When the functionality is considered, there can be concluded that the mix of functions causes conflicts, but is at the same time the main characteristic of a bicycle street. To deal properly with this mix of functions, the street and its users should be forgiving. In the design step, recommendations have been made for the design of new bicycle streets. Important recommendations are the choice for single-lane bicycle streets as standard design, adding bicycle path lining to those type of streets and the application of raised intersections to reduce the speeds of motor vehicles on the bicycle street. Further has been recommended to integrate the gutter with the rabat strip to clarify the function of the rabat strip and it is recommended to apply one standard traffic sign on all the bicycle streets in The Netherlands. Although the sample of eight bicycle street is too small to draw significant conclusions that could lead to a revision of guidelines, the study indicates some aspects that need further investigation. A more extensive research with a larger sample is needed to validate many conclusions that are drawn in this study. (Author/publisher)
Samenvatting