At War With The Motorist

Subscribe to At War With The Motorist feed At War With The Motorist
The UK government has declared that they are "ending The War On The Motorist". These are our correspondents' dispatches from the front lines of their policies.
Updated: 6 hours 7 min ago

Shared use and covert cycleways: how cycling on footpaths came to be officially recommended

28 March, 2021 - 19:09

A misunderstanding in the ’00s led cycling organisations to recommend against separating cycling from walking and aided the proliferation of bad shared use paths. We’ve made some progress towards correcting that mistake, but it lingers in Bristol’s covert cycleways.

For anyone who has been missing getting together for Infrastructure Safaris lately, here’s a reminder that I’ve put some on YouTube. The latest is about shared use and covert cycleways:

I won’t try to replicate the video exactly in written form, partly because the point of an Infrastructure Safari is to (virtually) ride along and see everything in context, but mostly because it largely covers topics that I and others have written about before.

Specifically, the infrastructure safari is about the variety of things that fall under the umbrella of “shared use” cycling and walking infrastructure, and how each of them in their own way create and exacerbate conflicts between walking and cycling by persistently bad design.

The early history of shared use, introduced as a quick and cheap way to convert pavements into “cycle tracks” at a time when it was assumed cycling was on the way out, I wrote about in 2012.

But there is one part of the story that I had thought had been written about, if not by me then by another blogger, but which now after some searching I fear perhaps hasn’t, and so maybe deserves a few words here. That is: the story of how design guidance and even cycling organisations came to endorse and encourage mixing walking and cycling rather than trying to keep the two separate.

For decades after their introduction, the norm for shared use paths was, to use the technical jargon, “segregated shared use”. This unhelpful oxymoron means paths on which cycling and walking are separated into their own space. This encompasses a great variety of designs, including those that have a proper cycleway and a proper footway alongside.

But in practice, the overwhelming majority of such paths were the crap facilities we know so well: 3 metre wide pavements and greenways with a white line painted down the middle creating a comically narrow space for each mode.

The design guidance for shared use paths, 1986’s LTN 2/86 and 2004’s LTN 2/04 specified the presumption should be in favour of separating the modes.

And that led to lots of problems. People inevitably can’t always stick to “their” side, but people also get territorial. Cyclists can’t really pass one another in 1.5 metres of space, so will use the walking side. Where sightlines are poor, this can mean last-minute evasive maneuvers. People walking together like to walk side-by-side, and groups especially spill over. People get distracted or lost in thoughts and don’t pay attention to a white line or subtle signage.

Studies started highlighting these problems with segregated shared use — particularly influential amongst them a study for the DfT by Atkins, which looked at a small sample of shared paths, all of them archetypal narrow crap facilities.

There was an obvious lesson to be learned from these studies: build paths wider than 3.0 metres, and with more robust separation than a white line. All of the conflicts were caused not by the modes being separated from one another, but by the fact that the modes inevitably failed to remain separated from one another in such narrow spaces. 

And yet we managed to take away an entirely different lesson: that it is best not to bother trying to segregate shared use, just mix the modes. Cycling Infrastructure Design, the DfT’s 2008 manual for crap facilities, removed the presumption in favour of separation and introduced the idea that it may be better to mix. Sustrans moved towards mixing in the ’00s and adopted a formal position against separating modes in 2010, based on their experience with routes including the Railway Path. Meanwhile, the Atkins study led to the DfT replacing LTN 2/04 in 2012 with LTN 1/12, which emphasised mixing modes. These moves also influenced the decision of the Royal Parks to remove separation on their much wider (but still only paint-separated) paths in 2017.

To be fair, some of the studies and guidance attempted to be a bit more nuanced: suggesting that widening paths is important where usage is high, and that mixing modes is a solution where physical constraints on width are out of the designer’s control. But nuance has never had a place in practice. Every time a design guide has specified technical minimum standards, alongside pages of nuance about best practice, the result is that infrastructure gets built to the technical minimum standards. And so, for more than a decade, rubbish 3.0 metre wide shared paths have proliferated where once rubbish 3.0 metre segregated paths would have been the choice. 

Things are changing. Last year’s LTN 1/20 swept away all the previous guidance, specifying proper cycle tracks in place of shared use pavements, and returning to a presumption in favour of separation of paths, but this time with proper guidance on their widths. Even Sustrans are very slowly and reluctantly accepting the inevitable and will be widening the Railway Path, even if they haven’t yet quite managed to shake off the dogma about mixing modes being a good thing.

But that leads to the other main point of the Infrastructure Safari: that when we build these proper cycle tracks and paths with separation, they need to be clear and legible to all. Covert cycleways will only perpetuate the conflicts that led to the big misunderstanding about separation vs sharing in the first place. Do watch the video if you’re interested in that story.

Categories: Views

Gloucester’s crap cycle facility and the ambition for active travel

22 February, 2021 - 12:42

Gloucestershire lost out on funding in the latest round of DfT active travel grants. Instead of wasting everyone’s time appealing the decision, they should reflect on why.

So the Department for Transport (DfT) have been delivering on their promise to withdraw funding from any council-led active travel projects that don’t meet minimum design standards. Gloucestershire County Council (GCC) are reportedly appealing against one such decision. But the DfT were absolutely right to turn Glos down, and GCC should instead try to learn why. Let’s have a go at that together — there might be something we can all learn from their experience.

The inadequate facility

The proposal in question is for a cycleway along the route of the B4063 between the edge of Gloucester and the outskirts of Cheltenham (Google Map). It’s the old main road between the towns — many decades ago bypassed with a fast dual carriageway, but, as is the British way, always retained as a through route itself.

The plans (PDF) are a mix of cycle tracks — narrower than the design standards require — and shared use pavements, some of them as narrow as 2.0 metres, others with bus shelters in them. There are a few half-hearted attempts at priority crossings of side-roads and property accesses, perhaps an amateurish attempt to fool the DfT into thinking Glos have taken on board the modern design standards, but which instead stand out for their fundamental failure to understand the geometry of cycleways, as they attempt to bend the kerb-hugging cycle track around the vast splays of the side roads. Of course the DfT’s feedback described this as “inadequate”. The DfT’s claim to be turning down councils who waste funds on crap facilities would have zero credibility if they’d given money to this rubbish.

The way to fix this is screaming out from GCC’s own annotations on their designs. They say things like “existing layby to remain”. They mark multiple turning lanes for stacking motor vehicles at signal-controlled junctions. They retain bus stop laybys — deciding that allocating the limited available space to motorists so they can easily overtake buses is a higher priority than allocating it to designing the cycleway well. In a stretch where they deemed there was sadly not enough space to separate cycling and walking, providing only a 3.0m shared pavement, they mark on the adjacent carriageway 3.32 metre wide central hatching!

As the DfT feedback says: the space is there, the council just need to show the political will to reallocate it from motor capacity — on a road, remember, that has already been bypassed by a high capacity parallel route.

Are there any critical fails?

Extensive use of shared use paths in inappropriate areas, poor and indirect provision at junctions. Could seek to maximise pedestrian and cycling space through carriageway narrowing etc. …

Designs as shown are heavily motor centric and could provide higher quality cycle and pedestrian infrastructure in line with LTN 1/20,(e.g. indirect crossings, no side road priority and so on).  Optional aspirations marked on the drawings provide suitable interventions, however the design approach presented could have shown greater ambition to address carriageway space and motor capacity/dominance.

The DfT are saying: be bold. Forget motor capacity, that’s what the bypass is for. Design a fantastic cycling and walking route, and then see what you can fit in for motoring around that. Can’t fit in a proper 3.5 metre cycleway with separate 2 metre footway and fit in 2 motor traffic lanes? Then don’t fit in 2 motor traffic lanes. Have a single lane pinch point with one direction of traffic giving way. Put in a bus gate to remove through private motor traffic if you have to. There’s no shortage of ways you can make this road great for cycling and walking, you just need to accept that the cycleway is a higher priority than the central hatching on a road — I don’t think I can say this enough times — that is already bypassed by a high capacity parallel route.

Sympathy for the councils

It’s hard not to feel just a little bit of sympathy for GCC. For decades they’ve been told to design rubbish, and suddenly they’re being chastised for it. They’ve got a highways department conditioned to expect cash for any old crap, and suddenly they’ve been told they’ve actually got to work for it. The reality of the situation must be terrifying — no wonder they’re in denial. If the DfT really are going to keep up seriously enforcing the new standards, it’s going to be quite the rude awakening for those councils who have found active and sustainable transport funding an easy target for something they can use to keep their highways departments busy, or covertly channel into motor-centric schemes.

You can also perhaps feel a little sympathy for a council that has designed something that, until the introduction of LTN 1/20 just six months ago, was largely what the government’s own design guidance was telling them to do. The designs GCC submitted are classic LTN 1/08 stuff.

Most of all though, I felt genuine sympathy for GCC when I read in the news coverage that the designs were largely drawn up for them by Highways England — an agency of the Department for Transport.

In 2013, Highways England committed to “cycleproofing” the interfaces between their network of motorways/trunk roads and the local road network, and said some genuinely encouraging things about what they do to overcome the severance caused by their roads and junctions. Alas, the sense I get is that policy effectively died long ago. (It’s a fate that awaits many policies: the people who championed it move on in their careers, those who are left to implement it aren’t enthusiastic enough, empowered enough, or qualified enough to make it live up the original vision or to then push it through to its next phase, and so it withers.) The B4063, it seems, is Highways England’s solution to cycleproofing the A40 — that high capacity route which bypasses the B4063 — and its junction with the M5.

From GCC’s perspective, they’ve submitted the DfT’s designs to the DfT for funding and the DfT have come back and said they’re not good enough.

How much initiative would you like us to use?

In their feedback in the designs, the DfT make it clear that GCC should be bold and use their initiative:

B4063 could potentially be more than a local distributor road as the A40 provides a bypass to this corridor and access toward the city centre via roads with more overall highway width available for corridor improvements; there are clear opportunities for network level interventions to help progress more ambitious schemes on the local road network

They’re saying what I said above: if they need to find more room to do the cycle track properly, GCC should take away central hatching, turn lanes and laybys. They should make traffic wait behind the buses at stops if necessary. They should put in single lane pinch points. If it comes to it, filter out through private motor traffic. If there’s too much traffic, enable modal shift and send any remaining excess elsewhere.

The problem is that GCC are at the same time being told not to do that. Mid way along the route is a large industrial estate, which continues to grow. The B4063 therefore has to be a distributor road — it’s the only available route for HGV traffic to service that industrial estate. The council are obliged to accommodate HGV access.

It’s hardly surprising that GCC’s highways department are rather attached to all the turning and stacking lanes at junctions along this road. They’ve spent five decades incrementally producing these motor capacity “improvements”, one by one, as and when developer contributions from housing and industrial estate expansion allow, as central government guidance has advised them to do. As central government still advises them to do. It must be exasperating to be told to fix these mistakes by the same government department that had told you — is telling you — to make them in the first place.

Ideally, perhaps, the industrial estate wouldn’t be where it is. It would be alongside the A40, or the A417, so HGVs could hop straight onto a main road without having to wind through villages on a road shared with cyclists and local buses. But it’s not, and in the past decade central government have been further eroding what little power local councils and local people have to determine where developments like these happen in their areas.

Perhaps, when the DfT say GCC should be ambitious and seek “network level interventions”, they are inviting GCC to bid to the Active Travel Fund for funding to build a whole new HGV access road direct from the industrial estate to the A40 that will enable through traffic to be removed from the B4063?

So I have some sympathy for councils, even the ones who still haven’t got it on active travel. GCC should go back and revise their B4063 plans, they can do much, much better. But they are given all of the responsibility for delivering on a central government vision and none of the power. They must juggle the demands of multiple, contradictory instructions and policies from different branches of central government — even from other teams within the same department.

On balance, at this stage in the Active Travel Fund it’s a good thing that the DfT are micro-managing the bids to prevent the money being wasted. But if this is ever to scale up to something that really delivers for active travel, central government need to fix their own policies that push councils to keep preserving and reinforcing motor dominance, and then empower the local authorities who are expected to do the work.

Categories: Views