The new Royal College Street cycle tracks

Context map: the completed cycle tracks on Royal College street (and the bit on St Pancras Way) are shown in solid red. The proposed extension of the tracks to Kentish Town Road is shown dashed red. The route I advocate developing as another pair of tracks on Delancey Street and Pratt Street, to link to Gloucester Avenue, is shown dashed blue.

I wrote quite a bit on how the two-way cycle track built in 2000 on Royal College Street, Camden Town, London NW1, came to be, in my post "Understanding Walking and Cycling", "deja vue" and the history of Camden's cycle tracks. I also covered proposals to change the street again in While Boris so far fails to "Go Dutch", Camden quietly gets on with it.

It is now possible to cycle the new pair of one-way cycle tracks on Royal College Street. The scheme is not yet complete. It has been delayed by various utility works, and ongoing work by the National Grid by the Pratt Street junction means it cannot be completed for another 12 months. From this junction there is a temporary northbound cycle lane in place. In addition, though the southbound track now starts at Baynes Street, a section of it is constructed with temporary plastic blocks to allow for lorry movements associated with the electricity works. The commitment from Camden is to build the tracks through the Camden Road junction, to provide the full, direct two-way cycling linkage between Kentish Town Road, Somers Town and Euston Road that has long been called for by Camden Cycling Campaign.

Temporary arrangements near Pratt Street

The scheme is therefore still being built, and in a continual state of flux. However, there has been a lot of comment on it on the cycling blogs, so it deserves some discussion here.

The old Royal College Street track, present from 2000 to 2013

As I've mentioned before, I was never particularly keen on the idea of ripping out the existing 3m wide green two-way kerb-segregated cycle track on the west side of Royal College Street in favour of a new scheme, as the old scheme seemed to be working well, to me, and I believed that if the money was available it would be better spent building a wholly new facility elsewhere, or, alternatively, on radically expanding the Bloomsbury Seven Stations link track, or replacing it with a no-through traffic cycle street, as it is now so clearly over capacity. The Royal College Street track was not over-capacity. Because the flows on it were mainly tidal, the 3m total width was used efficiently.

The over-capacity Seven Stations Link track at Tavistock Place,  photo by Rob Hayles

However, it needs to be understood that it is not a high priority of councils to construct additional cycle infrastructure; not even in one of the most cycle-friendly local authorities in the country. It is not really even a high priority for them to increase cycling, or, indeed, walking. It should be, but it is not going to be, until a very clear steer, and funding to match, is provided by the Department for Transport. Local campaigning can do nothing to change this: we need a massive national movement to make it so. Groups as diverse as the Cycling Embassy of Great Britain, the British Medical Association, the Institute of Civil Engineering and The Times newspaper, are campaigning for this change, but it is still very, very far away. As Gabriel Scally wrote in the British Medical Journal this week:

If we really want to make a substantial difference to the physical activity profile of our population, and to help combat obesity, we need to concentrate on dramatically increasing the proportion of overall personal journeys that are made by walking or cycling. This will simply not happen, or at least not within our lifetime, unless there is a substantial political commitment that follows through into a restructuring of spending plans. The type of pathetically small investments we are continuing to see will make a difference, but given the scale of the challenge that we face, that difference will be insignificant.

... 

The politician that takes the initiative and changes our entire approach to walking and cycling will be revered and remembered.

He put is well. How far away the change is is shown by the DfT's own map of cycle funding in England, excluding London:

They are hoping, with this publicity release, clearly, to make some people who don't analyse very far think they are doing something for cycling. But transport investments are measured in billions of pounds. Estimates for the cost of the High Speed Two line from London to Birmingham are between £40 and £80 billion. Cycling is cheaper than other kinds of transport investment, for the benefits obtained, but the funding needs to be serious. As I pointed out in another post, it is easy to show, with a bit of rough and ready costing, that to solve even the most fundamental cycle network problems of the Borough of Brent it would take more than Boris Johnson's entire allocation of £100 million that is earmarked for selected areas of Outer London. If the sums that the Mayor of London is projecting for cycling are two to four times too low, as I hold it is easy to show they are, if we are serious about making a difference, then the sums the DfT are projecting nationally are two orders of magnitude too low. The point is nicely made in Dave Atkinson's reworked version of the map:

So the point I a making, in a lengthy digression, is that we don't yet have a serious, prioritised policy of building cycle infrastructure (or indeed improving the walking infrastructure) yet in London or in England, and this cannot be laid at the door of local councils, nor probably at the door of the Mayor of London, and neither is there much that local campaigning can do about it, given the legal constraints on local authority spending.

What then is the priority placed on local authority transport departments by the DfT? There are two main priorities: maintaining the flow of "traffic" (whatever that means), and "reducing accidents". "Reducing accidents", of course, is not the same thing as reducing danger, or reducing the barriers to walking and cycling. In many cases the results of the (highly successful, in international terms, it has to be said) policy of "reducing accidents" is to make it harder for people to walk and cycle, and to reduce the share of these modes in the transport mix. We all know how this occurs: through the imposition of indirect and inconvenient "sheep-pen" main road crossings on pedestrians (very safe), and of guardrailing to keep, again, pedestrians out of mischief and maintain the dominance of cars on the streets, and through the imposition of barriers to convenient cycling and the creation of cycle routes that take absurd routings through a city (a terrible example in Durham here), all that stuff.

The point I am getting round to is that the changes to Royal College Street must not be regarded as part of any policy to extend cycle infrastructure, or increase cycling. They had nothing to do with The Mayor's Vision for Cycling in London, predating that, though the concept drawing did appear in that document. They came out of the current prioritised policy for local authority transport departments, imposed on them by government, of "reducing accidents". Officers cannot easily go against this policy; they regard themselves, in my experience, more or less contractually bound by it. For good or ill, this is the position.

The old cycle track on Royal College Street did not have a very good safety record. The problem was the Pratt Street junction. Drivers would emerge from this and only look right, for a gap in the two lanes of northbound traffic, and make their turn, often failing to observe the southbound cyclists on the track, who had marked priority. Everything had been tried over the years: installation of a severe speed table, "Stop" signs and flashing lights like at a railway crossing, a sign saying "Two-way cycle track". But there were still enough crashes that "something had to be done". Officers felt obliged to do something. Clearly we did not want to make the cyclists give way at a minor road, that would be calamitous for the whole principal of segregated cycle tracks (though many officers had wanted to do this in the past and had to be resisted by Camden Cycling Campaign), so the solution needed to be a redesign that placed both directions of cycle traffic on the expected sides of the street.

I continued to lobby against changing Royal College Street for a while, arguing that other locations in Camden, where there had been cycling fatalities, should be addressed first: Kings Cross, and the St Pancras Way – Camden Road junction. But these were both junctions with TfL roads, where TfL and the buses had a big, obfuscating interest. They were locations with a history of policy conflict between Camden, local groups, and TfL, which seemed deadlocked, and where nothing was likely to happen quickly. At Royal College Street, Camden could make changes quite quickly. Moreover, there was a synergy with the Camden Road danger problem. Moving cycle flows onto the expected sides of Royal College Street would make it much easier to construct a safe two-way cycle crossing of Camden Road, creating a junction cyclists could more safely use than the St Pancras Way one, and which they certainly would use preferentially if it became possible. Looking further north, mono directional-tracks would almost certainly be more straightforward to merge into St Pancras Way and then Kentish Town road.

I was won over to the idea that Royal College Street should be rebuilt in the way proposed by these factors:

1. The length of segregated cycle route would be ultimately more than doubled by the scheme
2. The minor junctions would become safer
3. The total width allocated to cycling would increase by 1m
4. The surfaces would be renewed
5. The whole route would become far more useful to far more people with the extension to Kentish Town Road
6. A better crossing of Camden Road than exists currently should be created
7. The proposed scheme would not be very expensive, due to the style of "lighweight segregation" proposed, making it a good template for elsewhere in London, if successful
8. The space dedicated to motor traffic would be halved from two lanes to one and speeds should be reduced. (It is hard to remember now the character of the three-lane motorway that existed here before 2000).
9. It would be demonstrated how a cycle track can co-esist with residents' parking
10. The "lightweight segregation" could easily be changed if the results were not good

My concerns were largely around the effectiveness of "lightweight segregation". Would it work, and be durable, and respected by motorists?

I no longer cycle this route regularly, since I moved from Camden to Brent. I made a special trip to observe the operation of the tracks between 6 and 7pm this week. So I've not seen it under an assortment of conditions, and my conclusions have to be highly provisional.

Northbound track near the south end

The scheme, so far as it has got, seems to be working. It is respected by motorists (and I've not heard reports to the contrary from those who use the route regularly). The segregation consists of an alternation of planters, filled with plants, and "armadillos", about 80cm long and 10cm high (these are officially called "zebras" by manufacturer Zicla and they cost €26 each, if anyone is interested). The tracks have been built at 2m wide. I did have doubts if this would be wide enough for two-abreast, social cycling, but I observed more than one pair of cyclists succeeding in cycling in parallel, so it turns out it is. There is car parking outside the cycle track on the east (southbound) side. This means that the drivers' doors will open on to the cycle track. However, drivers will be facing the oncoming flow of cyclists, so a classic "dooring" accident is unlikely. If it does occur, it will be a glancing blow, and the cyclist will not be thrown into passing traffic, as he or she would be in the usual UK roads arrangement, but deflected towards the pavement.

Southbound track near the north end

One thing that did impress itself upon me, using the Brompton, with its low tolerance of uneven surfaces, is the quantity of ironwork in the track surfaces. The road laying people have tried hard to get the finish as good as possible, but the presence of so much metalwork clearly reduces the quality of this new-build below that of its Dutch equivalent. The Dutch remove and reinstall all utilities when they redesign a street, as a matter of routine, so they achieve absolutely perfect cycling surfaces. Clearly that is not part of our maintenance culture or system, and to have moved all these access covers and drains would have enormously increased the cost of the scheme.

The northern end of the southbound track can only be accessed from Baynes Street at the moment, and most cyclists will probably continue to access it in the traditional way, from Georgiana Street, further south, as this is a slightly shorter route, so this part is slightly out on a limb at the moment. But it will become integrated when the northern extension is built.

Southbound track at Baynes Street junction

The junction of the track with Crowndale road, at its southern end, had to be completely changed. The priority afforded by the signals seems to be about the same as it was before, though I never measured timings. Cyclists positioned at the southbound stop-line have reported they have had some conflicts, or near-conflicts, with the flow of traffic northbound from Midland Road, as two queuing lanes try to merge into one to get into Royal College Street, but I did not experience this, being there at a less-busy time. Some adjustment might prove necessary here.

Looking at the Crowndale road crossing from the south

A lot of attention has focused on the execution of the bus stops. There are two of these, the one opposite the Royal Veterinary College, and the more northerly one.

The southerly bus stop

The northerly bus stop

With these stops, the passengers board from, and disembark on to, a table on the cycle track: there is not a separate boarding island (or "floating bus stop") as you would generally find in Dutch designs, or, indeed, that is being built on the Cycle Superhighway 2 extension, or that has been built on Lewes Road, Brighton. The cycle track is not interrupted by stopping buses; on the other hand, cyclists on the track have a big onus to look out for, and give way to, passengers. Considering that, in the southern case, there clearly was the space for a boarding island, this does not seem to be an optimal solution. There seems to have been an error of measurement in the plans of the road with at this point, as they show insufficient space for moving motor traffic to pass between the bus stop and the car parking, whereas, in fact, there is space. I am not interested in gratuitously holding the cars up behind buses, if there is space for them to get past without causing the buses problems. However, the potential for cycle-pedestrain conflict at the southern bus stop, and possibly also at the northern one, could have been reduced, with boarding islands (which were provided, in a limited way, by the narrow segregating strip in the old design). Certainly this design would not be suitable for busy bus stops.

In practice, on Royal College Street (and people who use the route regularly are welcome to correct me on this), there does not seem to be a significant problem with the bus stops. There are five buses an hour here, on a single service, all northbound. I don't know what the passengers think of these stops, and whether they are finding this arrangement better or worse than the old one. It would be interesting to find out, to apply the findings to future designs. But the number of potential interactions has clearly been reduced, because the southbound flow of cyclists has been transferred to the other side of the street, with no interaction with buses. That being said, the main northbound flow of cyclists will be in the evening peak, when presumably most passengers are getting off northbound buses, so I don't think we can go so far as to say the number of potential interactions has been halved. But at least now the passengers only have to look one way. Similar bus stop designs can be found in Denmark, in comparably busy places, or busier ones, and they do not appear to cause great problems there.

Another criticism that has been levelled is that the tracks could have been made wider. Seeing the street now, that does seem to be the conclusion. On the other hand, doing that would have placed cyclists closer to the motor traffic, travelling in one stream in the centre of its lane. The buffer would have been less, and so would the buffer between the parked cars and the moving motor vehicles (not relevant to the cyclists, but to car occupants). I'm not sure the balance of space achieved is optimal, but one advantage of this design is that it could cheaply be altered.

Another touted advantage is that cyclists can easily get into and out of the tracks, between the "armadilloes". Is this an advantage at all, it has been questioned? Should not the track be wide enough to accommodate all the cyclists, only needing to let them in and out at junctions? If cyclists are entering and leaving the track otherwise, is this not evidence of problems with the design, for example, obstruction at the bus stops? Well, I'm not sure it is all as simple as that. We are starting from where we are, which is a highly fragmentary (indeed, almost non-existent) dedicated cycle network, and an established culture of vehicular cycling. We need to develop the dedicated network, and interface with the established norms of London cycling as well.

Elsewhere in London, I sometimes find myself cycling past bits of cycle facility, ignoring them. And occasionally I think, "If that was in the Netherlands, I would have used it. It is not, in itself, a bad bit of infrastructure." Why is that the case? It is because the more continuous and pervasive cycle infrastructure becomes, the easier and simpler it becomes to stick to it as much as possible. We may, and should, aim for the best possible infrastructure, that will accomodate all cycling styles simultaneously, with enough capacity, but we simply will not get 100% take up from existing cyclists, however good we make isolated sections, because the network is fragmentary. With a fragmentary network, fast cyclists who are used to London traffic will often not find it worth their while to join it, even if sections are excellent.

I observed this effect at the Crowndale Road junction, when I saw that on every phase of the lights, the fast cadre of northbound cyclists coming from Kings Cross on Midland Road (the people who do not use the more intricate, quieter "Somers Town Route" past the British Library) join Royal College Street. Some of them would do what the design allows them to do, and nip in to the track between the planters. But a few would carry on resolutely up the carriageway of Royal College Street, rejecting the cycle track option.

Cyclists joining, or not joining, the northbound track from Midland Road

I can understand why the the rejectors do this. It's a switch of mindset to go between fast, pressurised vehicular cycling, and using cycle facilities, whether the facilities are good, bad, or indifferent. If this is the only bit of cycle track on your journey of 5–10 miles, which you aim to accomplish at an average speed of 17–20mph, and you think you might have to stop for bus passengers, or slow down for a couple of girls cycling side-by-side, in a place where you can't easily get out of the track, would you join it? There would be advantages and disadvantages either way, and it would be a marginal decision, so you might just stick to doing what you have always done. But such decisions, and the need for them, reduce, the more extensive the segregated cycle network becomes. The almost 100% takeup of the segregated networks by Dutch and Danish cyclists is not entirely on account of their point-to-point quality, it is also on account of their ubiquity. We'll have to work with an evolving situation, for a long time, on this street, and elsewhere, where the balance of attractiveness of the tracks versus the carriageway continuously shifts for various groups of cyclists. We're not interested in forcing cyclists to use dedicated infrasteructure if they do not want to. But all the same, it's a good idea to give them a lot of opportunities to get into it. From that point of view, this design is rather good. It's not what the Dutch do, but our whole situation is different.

So, the overall verdict on the scheme so far: is the result attractive? Yes, in my view it is. Is the cycling experience pleasant? In my opinion it is. Is this an improvement on what we had before? It's marginal, I think, at the moment. But the extension northwards will be the game-changer. Is it safer? Probably, we'll have to see. Could the design be improved? Probably. Will it be? It might well be. It could easily be adjusted. Would I be happy to see more streets rebuilt like this? Suitable streets, yes, unsuitable ones, no. As always, we need to proceed case-by-case.

Is it "Dutch"? Hmm. This last question has caused a a surprising amount of heat in some circles. On one level, what does it matter if it's Dutch, Danish, American, Spanish or a uniquely British product? What matters is the cycling experience, and of course, safety. On another level, well, in important senses it is "Dutch". It is a rare example on UK roads of decent-quality physically separated cycling space giving cyclists a high level of convenience and subjective safety. That's what we were always talking about when we campaigned under the slogan "Love London, Go Dutch" – clearly. I know that no street in the Netherlands actually looks quite like this, but let's apply a little common sense here. We need to take into account context.

One point of context is that we have little history in the UK, and few native models, in effective engineering to separate bikes from motor vehicles on the roads. We need to develop these, and, as I say, we need to make these interface well with the cycling culture that we already have. A specific point is that where there is a demand for both cycle space and car parking space, we have traditionally had the idea that bikes should be outside the car parking space on busy roads, sandwiched between the parked cars – with doors opening primarily on the drivers' side, right into cyclists' paths, and fast-moving, heavy traffic, in direct contradiction to the model used in the Netherlands, Denmark and other countries that have developed effective cycle networks. These countries have usually placed parking between the cycle flow and the motor flow, to protect it.  In breaking with UK tradition, and finding a way to combine the cycle flow with parking the correct way round, this can reasonably be described as a "continental" design, if not really all that 'Dutch". The same argument applies to the bus stops. These look like some of the Copenhagen bus stops (except those that I saw where passengers disembarked directly on to the cycle track did not even have a table on the track, so were even less kind to bus users).

Is the result at Royal College Street one that Camden Council should basically be congratulated on? As I say, it's very early days, but I would say "yes', and I would go on to point out what needs to be done next. Camden is very obviously a council that "gets" cycling, in contrast to some other London boroughs, where council policies are working directly against the "space for cycling" agenda. In Southwark a set of guidelines for cycling policy has been drafted that set the council hard against giving cyclists their own space on the roads. In Hackney, the councillor responsible, Vincent Stops, has declared clearly that "Hackney is a 'Share the Road' borough", and implied, in my hearing, that there will be no place on Hackney's roads for cyclists who refuse to cycle with motor traffic.

Camden's councillors and planners, or at least some of them, on the other hand, seem to have understood that to extend the cycling demographic and make cycling objectively safe, as well as convenient, requires that prioritised and protected space to be given over to cycling on significant through-roads. Other boroughs have yet to make that leap. If any London authority is close to "Going Dutch", it is Camden. They need sensible and balanced support. Not uncritical support, but, as I say, sensible support that will tease out new issues as they arise and find good solutions for the borough and for London. We only have to look at the farce of the segregated cycle lane that Tower Hamlets has just completed on Bethnal Green Road to see how far ahead Camden is with Royal College Street.

I look forward to hearing others' opinions, and not just from the usual bloggers, on the new Royal College Street design. Those cyclists I have from on Twitter, who have actually tried it out, appear generally to be pleased with it. It cost, it is reported, £50,000. I presume this includes the resurfacing and also the removal of the old cycle track. If this is the case, then the cost of the same length of track on a road which did not have an existing cycle track  to be removed would be even less. For the results, the money has been pretty well-spent. The original track, for information, cost £1 million in 2000, for less than 1km. Most of this cost came about because the drains had to be rebuilt. That having been done reduced the cost of the recent works. So, to an extent, one thing has been built on another, and earlier investments have not been wasted.

What I'd like to see next, apart from the northwards extension of the Royal College Street tracks to connect with Kentish Town Road, is for the connection of this route westwards through to Primrose Hill and Hampstead to be tacked. This currently goes westbound via Pratt Street, which has reasonable conditions, with a recently-improved mode-filtered crossing of Camden Street, and then via the appalling Delancey Street, with its nasty pinch point at the corner with Camden High Street, that the roaring vehicles exiting Pratt Street always try to get through before the cyclist can reach it, which then widens, pointlessly, into a three-lane one-way motorway on a curve as it approaches the Parkway junction, only for cyclists going straight-on, towards Gloucester Avenue, to be forced into a narrow centre lane between islands, competing with motor traffic to get through that junction, while just beyond lies the low-traffic oasis of relative tranquility that is the Primrose Hill area. Eastbound, there is no good route. The best is via the pedestrian/cycle bridge on Regent's Park Road, and via the congested and traffic-light strewn Chalk Farm Road.

The nonsensical three-lane one way racetrack of Delancey Street

The tight centre lane in which cyclists must jostle with motorists in order to get from Delancey Street to Gloucester Avenue. The short pink cycle contraflow track does not allow the other direction of travel, it goes towards Regent's Park

The relatively quiet haven of Gloucester Avenue, past one of the mode filters that effectively takes the whole district of Primrose Hill out of the through-traffic system

The solution to all this is clearly to replicate something like the Royal College Street design on Delancey Street. Royal College Street was also a three-lane, roaring, one-way motorway before 2000. Delancey Street could be civilised similarly, reduced to one consistent lane, westbound only for motor traffic, with cycle tracks going in both directions. This would effect a high-quality cycling connection in both directions between Kings Cross and Bloomsbury and the residential areas of north-west Camden: Primrose Hill, Belsize Park, Hampstead and Kilburn, and on to Brent, with minimal change to the existing traffic systems.

Context map: the completed cycle tracks on Royal College street (and the bit on St Pancras Way) are shown in solid red. The proposed extension of the tracks to Kentish Town Road is shown dashed red. The route I advocate developing as another pair of tracks on Delancey Street and Pratt Street, to link to Gloucester Avenue, is shown dashed blue.

I believe the new Royal College Street design provides, overall, a good model, that, with tweaks, would be applicable to other streets in Camden and London generally. If it were applied widely, we would see significant results in terms of increased cycling. I commend those involved with it, both in Camden Cycling Campaign and Camden Council.