RECYCLING OUR CITIES
Richard Mann
Oxford Friends of the Earth
1.
BACKGROUND
This paper has grown out of our experience of providing for cyclists in Oxford. Oxford
is a flee-standing city of approximately 120,000 people. It is the major employment
centre of the'county, with a large number commuting in every day, primarily by bus.
Twenty-five years ago, the proposal to build an inner relief road through Christchurch
Meadow was abandoned, and the "balanced transport policy" adopted instead. The
primary feature of this has been that no new roads have been built in the city centre,
nor are ever likely to be. The main effect is that traffic levels have been stable. A range
of measuras have tried to promote cycling and the bus, but with limited effect.
The political situation has been confused by there being two local authorities: the C i t y
Council and the County Council. The County Council is the Highways Authority, and
makes decisions about the use of the main roads. The City Council has an agency
agreement to look after all the other roads within the city boundary. The County places
more emphasis on the needs of those from outside Oxford who need to travel in. The
City places more emphasis on the poor environment that its citizens have to put up
with. There is inevitably some tension between the two authorities.
The City Council has for many years sought the views of local cycling groups through
co-opting them onto a sub-committee of its Highways Committee. This has allowed an
ongoing dialogue between councillors, officers and local cyclists as to how best to
provide for cycling. The objective can be described quite simply: to have Dutch levels
of cycling, but without having to spend large amounts of money. The result is a
strategy for cycling that is largely as described in this paper.
This paper goes on to describe how such a cycling strategy can be the viable basis for a
sustainable transport policy, for Oxford and for other cities.
2.
AN IDEAL CYCLE NETWORK
2.1
The Politician's View
From the polificianls point of view, the ideal cycle network is one that is cheap,
politically uncontroversial, and effective (probably in that order). Despite the change of
government and the likelihood that more money ~ be available for cycling, there Hill
always be competing demands for funds. The amount of money spent on cycling is not
likely to increase dramatically above our historical spending levels in Oxford.
The principal sources of political opposition come from the other modes of transport,
as space is reallocated to cycling. Particularly vociferous are pedestrians, who are
extremely hostile to the recognition of cyclists using the pavement. Elderly people and
aduk cyclists don't mix at all well, and teenage cyclists are probably worse. We
therefore have to avoid the promotion of cycling on the pavement. The alternative,
cycling on the road, has the potential to upset cardrivers. Providing for cycling may
well require reducing road capacity (for cars) in places. This needs to be done with
care if it is not to generate congestion, and potentially, driver hostility.
2.2
The Cyclist's View
Cyclists have lots of competing demands for cycle facilities. They variously want them
to be fast, direct, safe, pleasant and personally secure. They also need them to link
together to make complete routes.
Adult commuter cyclists will not use routes that are slow or circuitous. There is some
debate whether potential new adult cyclists are looking for speed or safety in the first
instance. However, it is clear that when he or she has achieved a degree of confidence
on a bike, and if there are no major hazards, the route of choice will be determined by
speed. Since our road system was originally designed around the economic activities of
aduks, the route of choice will generally be the main traffic streets.
Many adults are not quite so in a hurry. They will often seek out routes through
pleasant surroundings, preferably away from cars. Children will often be obliged to use
routes away from the traffic by their parents. The absence of a route that is perceived
as safe will often be suppressing a significant demand among children to be allowed to
cycle to school.
At night, attitudes change. Few women are happy using back streets, let alone cycle
tracks through parks. There need to be routes available that are busy enough to avoid
the fear of being attacked.
2.3
Types of Facilities
A range of facilities have been developed over the years, many of which will be
familiar. It is worth quickly reviewing them in the light of the previous paragraphs.
Cycle lanes have been used in many places in Oxford. They have a habit o f stopping
when they are most needed, and still leave you to negotiate traffic at junctions. They
are certainly not regarded as safe for children to use. They are however very cheap.
Cycle tracks are often superb. However, they rarely link up to provide for real
journeys. Cycle tracks are inevitably built away from houses, making them little use for
short journeys. In personal security terms, especially when unlit, they can be unusable
at night. They are also very expensive to build for any distance.
Back streets form the informal cycle network that is used on a dally basis. A few have
been signposted where they form through routes. However, back streets tend to link
rather better to main roads (for cars) than across main roads (for cyclists). Using back
streets for cycle routes usually presents few problems, though some traffic calming is
sometimes needed. The main expense is in providing safe and convenient routes across
main roads.
Pavements o f main roads are sometimes used where the back streets do not link up.
Using them tends to be cheap but highly distressing to elderly residents. There are
often difficulties crossing side roads, which can make them more dangerous than using
the main road itself.
All of these facilities are used in Oxford. None of them are ideal. The best (cycle
tracks) can't be used without incurring ridiculous expense and rebuilding most of the
city.
3.
RETHINKING ~
IDEAL
It is impossible to provide cycle facilities that work for all cyclists at reasonable cost
(and probably not even then). The approach we have developed in Oxford is to provide
two separate cycle networks, one aimed at adult commuters, the other aimed at
schoolchildren. This is called the Twin Network Approach.
Adult commuters are primarily interested in the time it takes them to get to work. We
can therefore reduce the emphasis placed on absolute safety, and provide simple
facilities on the main roads. A small number of other road junctions will need
treatment, where there are traffic lights or roundabouts. This is especially likely to be
required around large traffic generators, such as hospitals and campus universities.
Schoolchildren need complete routes at a reasonably safe standard. Directness is not
quite so important. We can therefore pick out a network Of back streets and cycle
tracks that can be made to link up. In Oxford we sought routes every half-mile (800m)
or so. Our efforts are then concentrated on creating main road crossings.
Both networks are supported by other back streets. These form the conventional
access routes to the main roads for adults, but also access to the designated back street
network for schoolchildren. Some of these other back streets will require traffic
calming, but few will need any cycle facilities as such.
4.
~
MAIN ROAD NETWORK
4.1
Cycle Lanes
The main things that cyclists benefit from on main roads is a bit o f space of their own,
protection at junctions, and slower traffic. So the basic requirement is for oncarriageway cycle lanes on all urban main roads. These need to be as continuous as
possible. They need to be marked across side road junctions, and alongside pedestrian
crossing markings (zigzags).
The minimum dimensions we regard as acceptable are 1.0m for the cycle lane and 2.5m
for the general traffic lane. General traffic lanes of this width need to be separated by a
median strip o f some kind. We have main roads laid out with these dimensions that
appear to operate quite satisfactorily, though a little uncomfortably. The minimum
width required for this median strip is still being debated locally, but is probably at least
0.3m.
Where width is available, and there is no difficulty with illegal parking, cycle lane
widths can increase to 1.5m, and general traffic lane widths to 3.0m. Wider traffe lanes
will probably lead to dangerous speeding; wider cycle lanes are probably less useful
than additional pavement or space in the middle of the road for crossing and turning.
4.2
Median Strips and Refuges
A wide median strip between the general traffic lanes provides a continuous informal
crossing opportunity for pedestrians. This can take the form of a long traffic island, or
more simply (and cheaply) p~allel painted dashed lines. This can be enhanced by the
us.e of pedestrian refuges (small traffic islands with illuminated bollards) at key crossing
points on street comers. These refuges also create a protected central area at junctions,
which can considerably assist cyclists turning right across the traffic coming in the
opposite direction.
4.3
Parking
Where parking is required, it needs to be provided for in bays, with the cycle lane and
general t r a ~ c lane marked around it. Residential parking bays can be a minimum of
1.7m wide. Short-stay parking bays should be 2.0m wide, to provide some protection
against car doors opening.
4.4
Bus Stops and Bus Lanes
Cycle lanes are not compatible with bus stop markings, so the bus stop markin.gs will
need to be removed. If there is a problem with illegal parking, narrowing the
carriage.way is more effective than markings. Except perhaps at major stops, the
overtaldng of stationary buses should be discouraged, in that it disadvantages bus
users, and is dangerous for cyclists.
Bus lanes (minimum 3.0m) can safely be shared with cyclists, if the bus drivers can be
persuaded to drive non-agressively. Stopping buses travel at about the same net speed
as cyclists, so overtaking is usually unnecessary. Where there are express buses
operating, the bus lane should ideally be 4.0m wide, and have a 1.0m cycle lane marKeo
within it. Bus lanes should operate at all times, not just in the peak.
4.5
Junctions
Traffic-light controlled junctions should feature advanced stop lines for cyclists. If at
all possible, the use of multiple queuing lanes should be avoided. Especially where
there are separate filters for the lanes, this creates a very dangerous situation for
cyclists turning right. Cyclists have to be able to pull across flowing traffic without fear
of being cut up. This can only happen safely if the tratfic is proceeding slowly.
Widening at junctions can literally be fatal. The extra lane at trame lights often adds
only marginally to the capacity of the junction, and should be dispensed with unless the
congestion this would cause is genuinely intolerable.
The radii of the junctions should be kept to the bare minimum. The reduces the
distance across the junction for cyclists, and makes it more reasonable for them to stay
in front of the tratfic until aider the junction is passed. Overtaking cyclists at junctions
is dangerous, and should be discouraged by engineering measures as much as possible.
For lett-tuming cyclists, the cycle lane markings can usually recommence immediately
beyond the stop line, and run across the junction.
4.6
Roundabouts
Preventing overtaking on roundabouts requires them to be designed to operate at
about 20mph (30kph) maximum. This is m sharp contrast to conventional British
designs!
The most direct method is to make the carriageway narrow, the roundabout quite small
and the radii tight. This will obviously cause difficulties for large vehicles. The best
solution will vary considerably with circumstances. When adapting existing
roundabouts, the answer may well lie in making the central island bigger or widening
pavements using granite setts, to allow them to be over-run.
5.
THE BACK STREET NETWORK
The bulk of this network can be formed of existing back streets. These need linking
together with cycle tracks and crossings of main roads. We sought routes between
each of the main radial roads, and crossings of the main roads roughly every half a mile
(800m). Wherever possible, we tried to run close to schools and other major traffic
generators (colleges, hospitals, shopping eentres). A key design tool was choosing
names for the routes; if there wasn't a sensible name, the route probably didn't go
anywhere useful.
5.1
Traffic Calming
Some back streets will need traffic calming, especially if they are parallel to main roads
and are likely to be heavily used by schoolchildren. Traffic calming can take the form
of humps, though if these are very severe they cause problems for cyclists just as much
as cars. Traffic cushions may be an alternative, but are ineffective against motorbikes.
Narrowings and chicanes can be problematical; in that drivers often speed through
them to get out of the way of an oncoming vehicle. If cycle bypasses are provided, they
are usually traps for litter, and often obstructed by parked ears; a slightly wider
narrowing (4.0m) is probably better than bypasses.
The preferred forms of traftie calming are shallow humps, pedestrian refuges
(effectively parallel narrowings), and junction-tightening (usually in association with
parking bays).
5.2
Cycle Tracks
Where tracks are viable (usually across green land of some kind), they should be
designed tO be shared with pedestrians. This means they should be at least 2.0m wide,
depending on the volume of pedestrians. If pedestrians are to use them as well, they
should not be thought of as fast cycle routes. Tracks should not be segregated; cyclists
should expect pedestrians to get in their way, and behave accordingly. Some form of
cycle-calming may b e required to enforce this. These could be: access barriers that
allow the passage of cycles but not motor cycles; cattle-grids or rumble strips; swing
gates that push open from both directions. Signs indicating that priority ought to be
given to pedestrians may also be appropriat e .
In general cycle tracks ought to be sealed. We have had some success using a stonecoloured rolled pea-gravel finish. After a short while this blends in quite well into a
semi-rural environment. Ideally there ought to be a grass verge, for dogs. Vegetation
will need trimming back at intervals in the spring and summer. A hard surface next to
the traek that is permeable enough to allow a smattering of grass to grow, but not
weeds, reduces the need for trimming.
Ideally cycle tracks should be lit at night. If street lights are not provided, the edge of
the path and obstructions need to b e picked out with small amounts of reflective paint.
5.3
Main Road Crossings
The simplest crossing of a main road is a large pedestrian refuge. This ought to be
2.0m wide, so that a cycle can comfortably fit inside. This type of crossing will
generally be sufficient where a cycle track crosses a main road.
An adaptation 0f this is where a back street crosses a main road from one side road to
another. In this instance a refuge before and after the junction provides a protected
5
central area, so the cyclist can cross each direction of traffic separately. The presence
of the refuge also means that the traffic lanes will have been narrowed, which improves
the safety of the crossing by slowing the traffic down.
Where a cycle track crosses a particularly busy main road, or where there isn't enough
room to provide a refuge, a Toucan crossing (a light controlled crossing for cyclists
and pedestrians) can be used. However, these cost the equivalent of at least ten
refuges, so will need to be used only when necessary.
Toucan crossings can be provided for crossing between side roads. However, the only
arrangement that is deemed safe is if left turns out of the side roads through the
crossing are banned. This may not always be possible. Simply closing the side roads, or
installing full traffic lights at the junction might prove easier.
5.4
Staggered Crossings
Where the crossing is slightly staggered, the crossing methods described in the
previous section can generally still be used. Judicious narrowing of the side roads can
help to reduce or increase the stagger, as appropriate. Increasing the stagger slightly,
for instance, may make a refuge crossing feasible.
The protected central area crossing can be used with quite substantial staggers. If the
second turn is across oncoming traffic (a left then a fight in Britain), the protection
afforded by the refuges is usually suttleient for cyclists to quite happily cycle between
the two junctions in the middle of the road. If the stagger is any distance, the section
between the two turning areas can be made into a short cycle track.
If the against-direction crossing has to be made first (a right then a left in Britain), the
best option will be to get the cyclist across the road at the first junction, then use the
cycle lane to the second junction. The cycle lane could usefully be wider than normal,
and be segregated from the general traffic slightly, by paint if by no other means.
5.5
Links Along Main Roads
When there is no other route available, it wiU be necessary to have short sections of the
back street network follow a main road. Rather than use the pavement, this can be
done by building a cycle track using the edge of the carriageway. This might consist of
a two-way track on one side of the road, or two one-way tracks either side. Which
depends on the feasibility of crossings at either end, and on whether cycle traffic is
predominantly on one side of the road or the other. These edge-of-carriageway cycle
tracks usually have to incorporate one or both of the cycle lanes as well.
The cycle track can be cheaply separated from the main road by using small concrete
blocks (0.4m x 0.4m) and paint. If there is room, and the required length is not
prohibitively expensive, a narrow traffic island can be used.
This type of cycle track has all the usual problems that face routes on pavements.
There are complicated interactions at bus stops, pedestrian crossings and side roads.
However, they don't take space away from pedestrians, which gets past one major
problem. Having the cycle track as part of the road also clarifies the approach at bus
stops, pedestrian crossings and side roads. In general, however, it is better to only use
such cycle tracks where there is no back street alternative, and for as short a distance
as possible. In this way, most of the problem situations can be avoided.
5.6
Pavements
r
The pavements of main roads should not be used unless there is absolutely no back
street alternative, and there is insufficient room to build a cycle track in the
carriageway. In the whole of Oxford, there were only two situations where the
pavements o f main roads had to be used. In both situations the pavement is over 4.0m
wide, so we are able to mark out a cycle track without having to narrow the pavement
unreasonably.
6.
OTHt2R BACK STREETS
Little treatment is generally needed for roads that don't form part of either network.
The main exception is traffic-light controlled junctions and roundabouts, which are
occasionally used on minor roads. These need approach cycle lanes, advanced stop
lines and calming, as appropriate. The aim should be for all these roads to operate at
around 20mph (30kph).
Other minor facilities will be useful for local access. These include exemptions to road
closures and turning bans, use of one-way streets in both directions, and short cycle
tracks. However, these facilities are unlikely to link together to form a coherent
network, and shouldn't be signposted in the same way as the main back street network.
• 7.
SUSTAINABLE TRANSPORT
So far, this paper has concentrated on providing for the bicycle. Some of you may well
regard the bicycle as a bit of a minority interest, only suitable for the few traditional
cycling cities, such as Oxford.
In reality, cycling is going to have to play a key part in developing a sustainable
transport system. Buses are simply incapable of providing for all the short journeys in
towns that are made by car and cause so much congestion. Similarly, walking has
limited growth potential,, given the suburban densities at which we live, and the
increasing distances to local facilities.
A strategy that uses cycling as its basis has the potential for achieving much of the
sustainable transport agenda. This is because providing comprehensively for the bicycle
not only promotes an alternative to the car. It also requires space for cars to be
reduced, sometimes slowing traffic, sometimes reducing capacity. This is a neat
combination o f carrots and sticks, making a clear link between restrictions on cars
(sticks) and provision of alternatives (carrots).
It's worth now turning to have a quick look at the inter-relationship between providing
for cycling and other "green" modes of transport.
7.1
Walking
Reducing roadspace for cars often translates directly into providing more space for
pedestrians. This can either be for wider pavements, or for traffic islands in the middle
of the road that can be used for crossing. More informally, the space may simply be
used f o r a painted median strip. Although this gives pedestrians little physical
protection, it generally increases their ability to cross the road when they want to, with
less fear o f being run over. One o f the key advantages of walking is that you can
generally take the shortest route between two points, such as between two shops on
opposite sides o f the road. Making it easier to cross the road where and when you
want considerably enhances the experience of being a pedestrian.
7
Reducing roadspace for cars is also designed to reduce traffic speed. This reduces road
noise, making the walking environment more pleasant. It also makes it easier and safer
to cross the road. Ultimately, it will also reduce the quantity of traffic, as car drivers
shift to other modes, leading to an even greater improvement.
Pedestrians also benefit from the back street cycle network, especially where new links
are created. Providing alternatives to cycling on main roads creates more pleasant
alternatives to walking along the main roads. A typical cycle track will have as many
pedestrian users as cyclists. The surfacing of the track will also benefit pedestrians,
especially those with pushchairs or wheelchairs to look after. It also is a better surface
to walk on in wet weather.
7.2
Buses
The interaction with buses has already been discussed in relation to the need to allow
cyclists to share bus lanes. However, unlike pedestrians, buses gain little directly from
the provision of cycle networks as described. Buses need to have provision made for
them too.
In order for bus services to operate efficiently, they usually need bus-only space in city
centres, and reasonably clear approaches on the main radials. It will often be possible
to provide bus lanes from the edge of the city, since the roads are quite wide. However
these usually end before the main bottleneck. In this situation, it will be useful to use a
"bus gate" to create a false bottleneck, where the traffic is held back to let the bus get
into the restricted section first.
There will sometimes be competition between bicycles and buses for precious
roadspaee. We have shown that compromises can be made, with general traffic lanes as
narrow as 2.5m. Bus gates can also be used to give buses advantage without requiring
the use of road width for bus lanes.
Buses and bicycles can also share bus lanes reasonably safely, as long as bus-drivers
avoid aggressive driving. The situation where there is most likely to be problems is
where there are express bus services operating. This can be solved by widening the bus
lane to 4.0m, and painting a 1.0m cycle lane within it. Alternatively, it may be.possible
to provide an alternative route on the pavement for use by cyclists in the mormng peak
(when the bulk of the problems occur). This will only be appropriate if the pavement is
wide, and the number of pedestrians low. Ultimately, the best option for express buses
will be to confine them to routes where there is space, or to provide separate
guideways for them on alternative alignments. In Oxford we are investigating the use
o f the railway corridor for just this purpose.
8.
O T I ~ R CITIES
In Oxford, we have a major advantage over many other cities: we have not spent the
last thirty years building roads. The level of traffic we have now in the city is little
changed from the situation then. The current Transport Strategy goes one further, by
removing virtually all the remaining sections of dual carriageway in the city centre
(generally so that one carriageway can be used exclusively for buses).
So how does the Twin Network Approach work in other cities? Ideally, you should
dispense with your dual carriageways and multi-lane roads. They are basically hostile
environments, attuned to the needs of the car, and have no real place in a sustainable
transport system.
8
Short o f that, the answer lies in the nature of the road. If the road concerned is an old
road that has been widened, it will be valuable to cyclists. The best treatment is to
provide a cycle lane (preferably quite wide), and advanced stop lines at traffic lit
junctions. On particularly busy roads, cyclists may choose to loop round to the left,
rather than turn directly right. This should be facilitated if required.
I f the road forms part o f a gyratory, the important thing is to ensure that each junction
has lights, giving the cyclist the opportunity to turn right. In some situations, a cycle
lane on the inside o f the gyratory may be useful. However, in general gyratory systems
are dangerous for cyclists, and abandoning them in favour of the conventional use of
two-way streets is much to be preferred.
New roads have usually been built on top of the original street network. They usually
have little frontage, and generally don't need to be used by cyclists. However, they may
have severed the old street network, breaking a link that cyclists would otherwise use.
Restoring that link will usually be the best way of providing for cyclists, and undoing
the damage caused by the building of the road.
At the interface between old and new roads, such as an old radial route meeting the
inner relief road, providing for cyclists will be particularly difficult. Wherever possible,
sucla junctions should be light controlled, with cyclists given access to the old road
beyond the junction. This may require a short dedicated phase in the traffic lights.
The alternative approach is to provide shared pedestrian and cycle facilities, and to
take the opportunity to improve matters for both groups. Even if this uses a subway,
the ideal is to use the line of the original road as closely as possible, and to recreate the
type of frontage (eg small shops) that was destroyed. When you build a pedestrian
subway, you are building a new street, and as closely as possible, it should have all the
features o f a street.
9.
CONCLUSION
In this paper I have explained the cycle strategy that has developed over the years in
Oxford. The Twin Network Approach has enabled us to concentrate resources
separately on the problems of providing for cycling that is both fast and safe.
Such a cycling strategy has also proved to be an excellent basis for a sustainable
transport policy, combining as it does incentives to use alternatives to the ear (carrots)
with restrictions on the use of the car (sticks). The main addition needed to the basic
cycling strategy is a parallel development o f facilities for buses.
The strategy as described, while most readily applicable in a city such as Oxford, can
be extended to be applied in cities which have more history of roadbuilding. The
strategy shows that by planning for cycling, we can envisage how the road network
must change to allow the sustainable modes to flourish. This requires a fundamental
reallocation o f roadspaee on our main roads that can truly be described as recycling
our cities.
I0