It is inevitable that intersections and crossings will be included in a cycle network. Planners and designers have choices in the type of intersection control and the type of facilities for cycling provided at individual locations, within the constraints of the space available and competing demands of different users.
When planning intersections for cyclist use, the goal is to provide people cycling with a safe environment and a reasonable level of service (LOS), at a reasonable cost and within the available constraints. Intersections are where most urban cycle crashes happen and usually involve a lower LOS for people who cycle, so intersection issues will be most important for developing a cycle friendly network.
Wherever possible, it is advisable to address potential conflicts between users at a distance from intersections, rather than at an intersection where there are more demands on road users. For example, where a kerbside cycle facility is involved between intersections, a transition on the approach to the intersection can be used to ensure that people intending to cycle straight ahead through the intersection are not situated to the left of left turning motor traffic at the intersection. Left turn slip lanes can also achieve this effect.
The key planning principles relate to the type of intersection control and the provision of adequate space; these are effectively design aspects, but need to be considered from the outset of planning to ensure they can be achieved.
Planners should consider the following design aims:
the intersection performs efficiently for cyclists under the traffic conditions expected throughout the planned life of the facility
an acceptable LOS for other users is considered through a network operating plan framework
all normal manoeuvres through the intersection are possible, from all origins to all destinations, especially for cyclists transitioning at the intersection to and from dedicated cycle routes (note that this includes right turns in particular, even if via a hook turn)
the conflict area between through-cyclists and left-turning traffic (especially heavy vehicles) is managed, preferably before the intersection
conflict points are easily identified and highlighted (eg coloured/textured surfacing)
cyclists and drivers know where cyclists are expected to be on the road, and all measures are taken to dissuade drivers from queuing over cycle facilities
the intersection is consistent with mid-block facilities on approach and departure.
As for provisions between intersections, treatments for cycling applied at signalised intersections may suit a particular target audience but be less appropriate for other cyclist types.
The differences in preferences of different target audiences can become more acute at intersections. Interested but concerned cyclists require a higher degree of separation from motor vehicles; at signalised intersections it is generally not possible to achieve spatial separation but temporal separation can be achieved by using specific cycle signals and phasing. However, this will reduce the overall level of service at the intersection and result in cyclists receiving a low proportion of green time compared to that of adjacent through traffic.
Enthused and confident and strong and fearless cyclists in particular are unlikely to accept this additional delay and decrease in LOS in comparison to adjacent motor vehicles, and they may thus choose to not comply with the cycle signals. Full temporal separation may not be required and it may be preferable to simply provide cyclists with a head start through the intersection and then allow filter turning, so as to not significantly increase the delay to cyclists and risk the resulting non-compliance issues. If speeds and volumes of filter-turning vehicles are kept low through planning and design, these vehicles will not pose a great safety risk to cyclists, and may not be a great deterrent, either.
Enthused and confident cyclists are more likely to be comfortable with basic intersection storage facilities such as advanced stop lines, advanced stop boxes and hook turn boxes.
Grade separation may be considered as an alternative, as this gives people who cycle a continued physical separation from motor traffic through the intersection, and an excellent LOS, without compromising the overall intersection LOS.
Roundabouts involve a higher proportion of cycling-related injuries than other intersection types. Multi-lane roundabouts are the main culprits and high-speed is also a critical factor. It should be considered whether such roundabouts on proposed cycle routes can be modified so that they are ‘cycle-friendly’ (ie by reducing number or lanes and vehicle speeds), whether they can be converted to traffic signals or whether grade separated crossing options can be provided. If not, routes should be planned to avoid such roundabouts.
Speeds at urban multi-lane roundabouts can be managed by providing traffic-calming measures prior to the roundabout or platforms on the approaches (which can also act as crossing points).
Small, single-lane roundabouts that are designed to tame traffic speeds have been proven to reduce cycling injuries (Wilke, 2014). These roundabouts slow traffic by using the shape of the islands to deflect traffic onto a curved path, and/or by ensuring visibility to other traffic is restricted. Where operating speeds can be capped at desirably 25 km/h, or at a maximum of 30 km/h, they require no special provision for people cycling.
Sharrows may be added to encourage less confident cyclists to take the lane when approaching roundabouts (and raise awareness of lane sharing among drivers). If traffic volumes reach high levels, cycling on the road will become increasingly uncomfortable, and alternative approaches to negotiating the roundabout may need to be considered.
Some enthused and confident cyclists may feel comfortable mixing with traffic (ie adopting a vehicular cycling approach) at larger roundabouts with higher operating speeds, but few are likely to be comfortable at large multi-lane roundabouts and so external perimeter paths should be considered (Austroads, 1999; Bach and Diepens, 2000). Crossings at roundabout legs will generally result in a poor LOS for cycling owing to crossing delays, and some approaches or departures of busy roundabouts can simply not be crossed during peak times due to a lack of available gaps. Platform crossings where pedestrians and cyclists are given right of way could be considered to address this, but these may not be appropriate on multi-lane arterial roads. Grade separation or conversion to traffic signals is strongly preferred over multi-lane roundabouts.
Priority controls are generally used for intersections of minor roads with major roads. At priority intersections, the side road(s) has either a Give Way or Stop control. An uncontrolled intersection is generally used where two low-order roads meet; no control is implemented and normal give way rules apply.
Priority and uncontrolled intersections are the most common intersections in the transport network, and as a result involve a high proportion of cycle crashes. Of all crashes involving cyclists that occurred from 2003–12 at urban intersections/driveways, 57% occurred at uncontrolled or priority intersections (Leggat et al, 2014(external link)).
Cyclists on the major road travelling through priority/uncontrolled intersections generally travel to the left of the general traffic and are opposed by vehicles emerging from side roads and turning into the side road from the major route.
Where the major route is congested, some drivers will leave gaps in the queue at side roads to allow other vehicles to turn right into the side road. This is a significant hazard to cyclists who may be travelling past the stationary traffic to the left (within a wide general traffic lane, or in a dedicated cycle lane) and who may not be seen by a turning driver.
On-road cyclists emerging from minor roads at priority intersections also need to be considered in design. If there are not enough gaps in traffic on the major road it may be necessary to consider a treatment to help cyclists cross the road. One way of doing this would be to provide a crossing treatment downstream of the intersection on the major road.
Where minor side roads intersect pathways or separated cycleways, the side road should be designed to seem like a driveway so that motorists will give way to path/cycleway users.
Driveways, especially commercial driveways, are similar to priority intersections and should thus be treated carefully in design, especially when a two-way or contra-flow cycle facility is involved. Crashes at driveways make up 14% of all urban cycling crashes (Leggat et al, 2014)
Treatments at driveways can include coloured surfacing, cycle logos (and pedestrian logos if a footpath or shared path is involved) to alert motorists to the potential for cyclists. Locating the cycle facility further from the property boundary can assist safety as cyclists are more visible to drivers exiting driveways.
Where a separated cycleway at carriageway level is involved, it may be suitable to employ a small speed hump across the gap in the separation device (as in the photo below). This gap can also be designed to limit vehicle turning speeds to an appropriate level; this requires consideration of the design vehicle to be used. For example, it is unlikely that a residential driveway will need to accommodate a large truck in general circumstances; therefore the design vehicle should be a 95thpercentile car (note that furniture removal trucks, for example, would still be able to access the property, but these would need to cross the centreline to manoeuvre to and from the driveway). The kerb and channel between the cycleway and footpath/berm generally acts as a mini-speed hump.
Where a separated cycleway at footpath level is involved, the kerb and channel is between the cycleway and the roadway and so it may be more appropriate to employ a speed hump device between the footpath/berm and the cycleway.
Alternatively, the cycleway can be ramped up and down above the driveway level to slow down down crossing over it in either direction.
Bridges, underpasses and overpasses are the main forms of grade separation associated with cycling provision. Providing grade separation at intersections can help cyclists cross the intersection without being delayed by waiting for other users and without being exposed to potential conflict with motor traffic. Grade separation may also be applied between intersections, for example to help cyclists cross a railway line, a river or a high-volume, high-speed road safely, without being delayed or needing to detour.
In some situations there will be the option of whether to build a bridge over the obstacle, or a tunnel under it. Cycle bridges involve greater vertical deviation than cycle underpasses, as bridges must clear trucks using the road below, whereas underpasses need only descend low enough to provide enough clearance for people on bicycles. Cyclists also prefer to speed up going down a ramp to an underpass, and use their momentum to travel up the ramp on the other side rather than having to first climb up a bridge and then speed down the other side. Also, the fact that bridges generally involve a greater vertical deviation means they require more effort to negotiate.
However, bridges have lesser drainage requirements, easier lighting requirements and are more likely to offer a higher degree of natural surveillance. Site topography may favour either a bridge or an underpass.
People biking along bridges and underpasses tend to shy away from the sides, so required facility width must be carefully considered and should ideally be greater than the width of the path it is located on. This is doubly important because it is not so easy to retrofit additional width to a structure later.
Personal security is important. NZ Transport Agency’s Underpass guidance notes (NZ Transport Agency, 2009) provides numerous suggestions to enhance personal security at underpasses.
Because structures are expensive, the needs of people cycling and others must be properly identified, particularly in relation to:
constructing a motorway or highway interchange
planning new residential areas
designing a new bridge or tunnel.
Koorey (2002) identifies several possible methods of providing cycle crossings at a lower cost than building a completely new bridge:
install pre-fabricated structures for small crossings
recycle disused road/rail bridge structures (eg retain the old bridge for cyclists when upgrading)
use abandoned bridge abutments and piers to create a new crossing facility over top
provide a lightweight crossing in conjunction with existing utility (pipe/cable) bridges
make alterations so that cyclists can use an existing footpath on a bridge (however, this should not cause considerable conflicts between cyclists and pedestrians)
re-mark the road to provide narrower traffic lanes with adequate shoulders – it may also have the effect of slowing traffic down
consider a lower speed limit on some long narrow rural bridges
provide warning signage and/or ban overtaking on the bridge
add clip-on lanes to the bridge sides for cyclists and pedestrians
provide alternative routes via sealed fords across generally dry channels
widen the existing bridge carriageway (perhaps in conjunction with strengthening projects).
It is important to consider the connections to a bridge or tunnel facility for cyclists. For example, if a bridge involves a two-way cycle facility on one side, this will require cyclists traveling in a certain direction to cross the road to access and depart from the bridge.
A road interchange is a combination of grade separation and interconnecting roadways at the intersection of two or more roads, at least one of which is a motorway or expressway. The scale of interchanges can vary; they may involve a simpler at-grade, on-ramp and off-ramp arrangement at a rural motorway interchange giving access to and from a collector road, or they may involve several grade-separated roads within a busier urban context.
The Design guidance section outlines the specific possibilities and considerations for providing for cycling at interchanges. The important thing to remember at the planning stage is that carefully considered treatments will be required wherever cycling is to be accommodated at (or near to) an interchange. The ONRC road classification may determine what level of cycling provision is required. For ‘National Roads’ cyclists should be provided with separated facilities, which generally entails grade separation of their own, or protected signalised crossings wherever the cycle route crosses a national road. Otherwise, carefully designed crossing points may be needed where cycle routes cross on/off ramps.
A cycle route crossing a road between intersections effectively creates an intersection itself although with particular features given motor vehicles are present on the road being crossed but not in the direction of the cycle route.
A number of different treatments are available for cycle crossings between intersections:
no specific provision
kerb build outs
Less-confident cyclists are more likely to prefer signalised crossings that offer temporal separation over other treatments where they are required to judge gaps in the motor traffic, especially where traffic volumes and/or speeds are high. Cyclists who are more confident may not appreciate having to wait at signalised crossings if they perceive that there are enough available gaps in traffic and they could otherwise cross at their own judgement. Most cyclists will appreciate grade separation treatments, assuming that the gradients are not too steep, as these enable crossing without being delayed by or exposed to motor traffic.
Priority at the crossing can be either in favour of the road or the cycle route, and should be determined on the basis of relative importance of the cycleway and roadway.
The Australasian Pedestrian Crossing Facility Selection Tool(external link) can be used to identify the suitability of crossing treatments between intersections for cyclists as well. While cyclists travel at faster speeds than pedestrians, for short distances (eg crossing distances) beginning from a standstill, the pedestrian speed used in the guide is a suitable approximation for cyclist speed.