It’s long been a contention of this blog that Boris will, eventually, bring back the discredited, costly and destructive 1960s Ringway scheme in some form or other.  The first inkling was the Silvertown Tunnel proposal (which is essentially a Third Blackwall Tunnel by another name) which TfL brought forward once Boris had canned the Thames Gateway Bridge, itself a shadow Ringway remnant.  The other shoe has now dropped and unless London mobilises in the next two months to stop it we’re going to throw away forty years of progress on building a non-car dominates city based on nothing more than a politician’s whim and a scheme cooked up by rank incompetents and greedy spivs.  Hyperbole?  Let’s lay out some facts.

The libertarian Conservative administration at Hammersmith & Fulham have long been far more pro-car than is reasonable in an inner city borough (compare Camden or Hackney, for instance).  This manifested itself in opposition to the congestion charge western extension and subsequently in a pointless widening of the awful Hammersmith Gyratory, sold to TfL as a way of ‘smoothing traffic flow’ on Fulham Palace Road. This was lavishly funded by TfL and so far as I can see is a complete failure – I cycle through the area to work reasonably often and it’s still insanely bad, but with more space for cars to get stuck.  Heartened by this, and with a transport agenda (‘Get h&f Moving’) that concentrates mainly on speeding up the road network they are now pushing the first major urban motorway scheme proposed in London since the Thatcher Government’s post-GLC proposals were binned by Cecil Parkinson in March 1990 with this surprising speech:

The main points are clear. First, there was strong support for improvements to public transport. Secondly, there was widespread opposition to most of the major new road schemes suggested by the consultants. Thirdly, there was support for proposals to slow traffic in residential areas, both to improve safety and to deter rat-running. Fourthly, there was general recognition of the need for better traffic management but concern about the level of traffic and a wish to see higher priority given to buses, cyclists and pedestrians.

A number of the most important public transport schemes identified in the assessment studies are now under active examination. We are evaluating urgently with LRT and BR the proposed Chelsea-Hackney underground line and east-west cross-rail. Subject to the satisfactory outcome of the work, I expect to authorise the introduction of a Bill for one of these lines in November 1990. LRT is appraising the extensions of the Docklands light railway to Lewisham and of the east London line northwards to Dalston and Highbury and southward to east Dulwich. It is taking forward studies of the Croydon light rail system with the borough and BR. Funds to begin upgrading the Northern line are already in LRT’s investment programme

I have asked the chairman of LRT to consider further the case for the extension of the Northern line from Kennington southwards to Streatham and Crystal Palace and for a further extension of the east London line to Balham. I am asking BR to give further consideration to service improvements and new stations suggested by the consultants.

I have decided not to proceed with the major road schemes recommended by the studies. I have rejected the western environmental improvement route and the idea of a tunnel from Chiswick to Wandsworth. I have also ruled out schemes on the south circular such as those at Stanstead road and Brownhill road, and the tunnels under Clapham common, Dulwich park, Tulse Hill and Forest Hill. My Department will press ahead with limited improvements along the south circular, and A3 West hill to improve conditions in Wandsworth. In south London, we shall not pursue new routes across Chipstead valley or along the Wandle valley. We shall take forward proposals for improving the M23-A23 junction, for a Hooley bypass, and for widening and junction improvements between Coulsdon and Thornton Heath. I have ruled out new roads at Norbury and Streatham, but I propose to consult the local authorities on establishing an improvement line so that new development is set back to allow eventual widening.

The roads lobby is never quiet for long in London, however, and has been biding its time and changing its skin yet again and finally re-emerged with Boris’s formation of the Thatcheresquely-named Roads Task Force after May 2012. This had a hugely disparate membership from the usual lorry industry suspects to Sustrans and the cycle blogger Danny Williams.  This produced a report which basically came out in favour of every single idea it had been given from streetscape improvement to huge urban motorways, which is unsurprising in the absence of any serious central direction about how London should tackle road congestion.  The foreword gives the game away:

What unites the Mayor and the RTF is a common belief that London’s roads and streets – which carry 80 per cent of people’s trips, 90 per cent of freight in London and account for 80 per cent of its public space1 – deserve a long-term strategy, commitment to investment and increased ambition in how they are planned, managed and developed.

The aim is to tackle road congestion, and also to ensure London’s streets and roads provide better and safer places for all the activities that go on there, along with transformed conditions for walking and cycling.

The obvious answer, which has been London policy since 1973, is traffic restraint – this reduces congestion (or at least keeps it the same while the city grows, which is the same thing) while allowing roadspace reallocation to sustainable modes. Instead of this the RTF has wound the clock back 50 years and taken as fact the discredited assertion that economic growth will always lead to more congestion. This is doubly odd because despite London’s population growing 12% between 2001 and 2011 road traffic *fell* 9%.  The clear evidence that economic growth can successfully be decoupled from rising road traffic by political action and that modern urban societies can thrive without endless increases in motoring does not appear to have entered into the RTF’s thinking.  There is other evidence for this in the high tech industry that Boris is so keen on – this doesn’t set up near main arterial roads, it sets up in the inner city congestion charge zone where there’s good public transport.  Even out here in west London’s TV Triangle Hogarth Business Park, with a single bus route every 15 minutes, is unlettable and being demolished for housing while the converted warehouses and so on along Chiswick High Road are full of start up video production companies staffed by people who get the bus, tube or cycle.

Ignoring the obvious evidence and choosing to base themselves on a myth the RTF inevitably ended up in an impossible situation – they want to improve the city for sustainable modes everyone will be using while providing more roadspace for relief of congestion caused by the motor vehicles everyone will be using:

[T]he RTF has concluded that if the full aspirations for different streets and places, and for increased use of more sustainable modes of travel, are to be achieved without undermining the ability to get around the city efficiently, this will require the application of more strategic interventions – including both managing demand and investing in improved/new infrastructure

Note the subtle division between ‘sustainable’ modes (walking and cycling) and ‘getting round the city efficiently’ (driving, presumably) – up until now it’s been widely understood by everyone involved in London that walking and cycling are, quite apart from environmental and health benefits, substantially more efficient uses of land than private motoring and should be encouraged on that basis.  Given this level of intellectual bankruptcy it’s not entirely surprising that the ‘toolbox’ they use is prone to grandiose, expensive and unfeasible solutions, which will therefore need to be sold to the public under false pretences. Which brings us back to Hammersmith & Fulham council and the Chiswick-Kensington Urban Motorway.

They don’t call it to CKUM, of course, they called it the ‘Hammersmith Flyunder’, and present it as a straightforward replacement for the 1960 Hammersmith Flyover at the end of its life.  This is actually 10-15 years away – TfL are investing tens of millions of pounds refurbishing it right now.  That’s the first lie, therefore – the timescales being talked about for a replacement are roughly 2017-2020, so it’s not an externally forced decision at all, it’s a choice to invest a large amount of public money in a large road scheme in an inner city borough.  There are therefore alternatives and the time to explore them.

The original ‘flyunder’ idea stems not from engineers or transport planners or anyone with any relevant knowledge or experience but from local architects and the Business Improvement District (in association with construction company Halcrow, significantly), and is naturally an incoherent mess with a bucket of PR sloshed over it.  Cited as rest of the world examples are:

  •  the Big Dig in Boston, which spent $24.3bn burying a major road through the city, now increasingly considered to be a failure in transport terms – the city has a nice park instead of an elevated highway but its roads are still hugely congested.  A comparison to what that sort of eye-watering sum could have bought in the way of 21st century public transport is interesting.
  • Embarcadero, San Francisco – this was a highway *removal*, with no attempt to replace the elevated road damaged in the 1989 Loma Prieta earthquake, either above or below ground.  The result is less traffic and a significantly nicer area, and the roads lobby’s fearmongering about ‘gridlock’ proved entirely illusory
  • New York High Line – this isn’t a road project, but an internationally significant prettification of a redundant rail viaduct.  Its appearance in any UK context is usually a herald for an outbreak of spivvery by someone selling something; the reason we can’t do one in London is that our redundant rail viaducts mostly have new railways on them, such as the DLR in Docklands and the East London Line.
  • Cheonggyecheon, Seoul – like Embarcadero a full highway removal and replacement with landscaping and restoration of the existing river that was buried under the road.
  • Central Waterfront, Seattle – a genuine tunnel-replacing-elevated-highway plan, but a long way from completion and already controversial for diverting huge funds into a hole in the ground. Recently, the world’s largest TBM ‘Bertha’ became stuck amid reports that the geotechnical research wasn’t entirely accurate
  • Madrid’s inner ring motorway, in scope much like the aborted Ringway 1, was buried in 2004 in a glorious example of Spain’s lunatic infrastructure period, funded by borrowing and PFI. The environmental credentials of a scheme designed to attract more traffic are, to say the least, iffy, and the actual lesson to take away is ‘don’t build a ring motorway in your capital city, it’s a traffic and pollution generating money pit’.  Wikipedia’s view: ‘The M30 is the busiest Spanish road, famous for its traffic jams’ hardly inspires one to imitate it.

The West London Link study goes on to recommend not a flyover replacement but a full scale urban motorway tunnel from Hogarth Roundabout in eastern Chiswick to Earls Court in Kensington, completely bypassing the entire borough of Hammersmith and Fulham and with no intermediate junctions.  This would apparently:

  • Eliminate through traffic and the pollution it emits from a 3.7km stretch of West London.
  • Release a strip of land for the profitable development of Chiswick, Hammersmith, Kensington and Chelsea.
  • Create much needed space for parks and recreation.
  • Reroute traffic to recreate Hammersmith as an urban centre.
  • Reconnect Chiswick and Hammersmith to the river.

The reference to Chiswick is a tad misleading.  I live here and know full well that Chiswick technically ends within yards of the start of the tunnel just east of Fuller’s brewery, so the plan spends a large amount of money leaving Chiswick’s A4 section completely unaltered.  This map shows the tunnel in yellow and borough boundary in red.  The total reconnection in Chiswick is therefore two streets.


The bulk of Chiswick’s also already connected to the river by dint of about half of the area being south of the A4 anyway. Our problem is air quality due to being forced to act as the main end point of a large motorway and there’s nothing in this plan that remotely accepts that is an issue.  Other problems with the study are curious things like proposing a tunnel ending at Earls Court and illustrating it with pictures of a six lane tunnel emerging just east of the Ark at the current end point of the flyover. The thought process appears to be this:

  1. A straight burial of the flyover means you need two ugly, noisy, polluting, intrusive tunnel portals and approach ramps, roughly on the sites of the existing flyover ramps (although probably further out on the east side due to the need to go under the District and Piccadilly line
  2. Subways and surface crossings are incompatible with these ramps, so any existing subways would have to close, limiting the possible sites*
  3. Not only does this mean no flyover during the construction phase, with 80,000 vehicles with nowhere to go, but the council will come under pressure from residents west and east of the tunnel portals to move it 100 yards further out
  4. Repeat until you get to the edge of borough
  5. Crayon in the tunnel entrances there

[*The subway point is interesting; between Hogarth Roundabout and Hammersmith there are five subways and two surface crossings, which essentially limits the downwards ramp sites to the central section between Black Lion Lane and the Town Hall.  Building the ramp outside this means you have to remove an existing crossing and create *more* severance.  Notably the current western flyover ramp has a surface pedestrian and cycle crossing directly underneath it and a moment’s thought will illustrate that as people need less headroom than HGVs and buses  a flyover will be able to clear a surface crossing well before a tunnel can.  The other alternative is replacing surface or subway crossings with pedestrian bridges, but again the headroom aspect means that you need much longer ramps and more steps to get over the road than you do to get under it, and those structure are somewhat intrusive.]

Nevertheless, on the basis of this the council decided to include the scheme in…

…the borough’s response to the road task force suggestion to explore ‘alternative tunnelled routes’

This refers to a brief case study on the A4 in the RTF’s Annexes which states:

However, to exploit the potential new surface developments and make greater use of Hammersmith’s established transport hub, longer-term proposals for providing alternative tunnelled routes for through-traffic should be explored. This would have transformative effects on the town centre, greatly improving the quality of life for its residents, and reducing severance of communities along the corridor.

which sounds suspiciously like a circular argument – the council presumably lobbied for that (the language is a dead giveaway, and it’s 100% focused on Hammersmith) and are now using the RTF’s incorporation of this as a justification to proceed, although they’ve conveniently forgotten the ‘longer term’.  The subsequent feasibility study, which I have in draft form and which is supposed to be handed to Boris this week came up with yet another set of different options:


  1. (green) a 1.6km tunnel 15m below the surface on the line of the current flyover, which I suspect is what most residents will assume is the only plan given the careful branding as ‘flyunder’.  This would be excavated using cut and cover, which given that the roads you’re cutting carry around 120,000 vehicles a day implies some drastic traffic restraint across a wide area for the three years of construction (so why not just make that permanent?).  Added to that you need to dig a trench down to and below the District and Piccadilly Lines – closing them for any period is likely to be politically and economically impossible, so we’re talking some kind of progressive jacking of a structure under the lines while they’re operating.  Not least of the issues there is that the same area has high north-south traffic flows as well, plus there’s a major bus station.  Being essentially the same layout as now there is zero reduction in severance for most of the corridor with just a short 100m or so section south of St. Paul’s Green available for repurposing.
  2. (red) a 3.6km tunnel from just before the Sutton Court Road junction in Chiswick to just before North End Road on the Hammersmith/Kensington border, well south of the gyratory.  This would be 25m down and bored, cut and cover not being particularly easy to do on rivers.  There would be major road junctions constructed on the surface at each end.
  3. (blue) a 4.1km tunnel from just before the Sutton Court Road junction in Chiswick to a point which appears to be between the two A3220 junctions on the Cromwell Road near Earls Court Station.  One odd aspect to this is that drivers heading east couldn’t turn north up the A3220 and drivers heading north up the A3220 couldn’t turn west.  Eastbound traffic would have no exits between Sutton Court Road and Earls Court.  This latter point is crucial to what follows.

‘From the project engagement four principal concerns were identified: traffic redistribution, cost, traffic disruption and construction traffic.’

When I heard about what I call the ‘long tunnel’ options I immediately suspected that there was a serious flaw in them.  The A4 east to Hammersmith has three lanes, the flyover has two lanes, but is rarely congested at peak times.  What is congested is the slip road from the A4 onto the gyratory, which suggests a significant proportion of A4 traffic east of Hogarth is trying to get to areas well west of North End Road, let alone Earls Court. If you bypass the whole area with your long tunnel all this traffic ends up east of where it wanted to be and doubling back on local roads, or alternatively comes off the A4 in Chiswick and uses local roads to bypass your expensive new tunnel.  The DfT publish traffic surveys yearly which, while notoriously inaccurate nevertheless serve as a rudimentary means of assessing broadly how much traffic moves along particular corridors today. As a rough rule I have the following:

  •  A4 Chiswick-Hammersmith – 120,000 vehicles/day
  • A4 Hammersmith Flyover – 80,000 vehicles/day
  • A4 west to Hammersmith Broadway – 40,000 vehicles/day
  • N-S flow through Hammersmith Broadway – 20,000 vehicles/day
  • E-W flow on the Chiswick-Hammersmith local roads – 15,000 vehicles/day
  • Hammersmith Broadway to A4 east – 14,000 vehicles/day

These are all before any of the lunatic traffic forecasts in the RTF and London Plan are applied.  What leaps out immediately is that unless you provide an alternative for the 40,000 vehicles coming off at Hammersmith today they’ll try and squeeze onto Chiswick High Road, which can’t possibly cope with 55,000 vehicles per day, it’s a single lane road while that sort of traffic justifies six lanes or so.  So that’s out. Casting around for an alternative, the only vaguely feasible route is on the surface along the A4, which now takes 80,000 vehicles/day underground and 40,000 overground. This is too many for a single carriageway so we need a local four lane dual carriageway replacing the existing six lane A4.  Assuming four lanes underground the Sutton Court Road junction is now eight lanes wide (2 westbound slip, 2 westbound tunnel, 2 eastbound tunnel, 2 eastbound slip), which is at least 33% more land occupied by roads in the middle of Chiswick while far from ‘reconnecting Hammersmith with the river’ what we’ve ended up with is keeping the A4 with 1/3 of the traffic and two lanes removed.  We retain Hogarth Roundabout, slimmed down a bit.  We don’t release a great deal of land for development along the corridor, either, if you moved the remaining four lanes to one side you’d fit in a strip about 8-10 metres wide, minus any road access or other reconnections you might want to do (remember?).  It’s a mainly residential area, too, so don’t go getting carried away with building height.  Finally you’re a long way from any public transport along there, so land value and commercial desirability may not be all its cracked up to be.

East of Hammersmith things are a bit different as far less traffic leaves the gyratory heading east towards North End Road, so Talgarth Road can probably slim down to a single carriageway road with around 14k vehicles/day. This would release a fair bit of land on the north side, at least as far as the point where you need to squeeze in two slip roads round a four lane tunnel portal.  There’s also a Tube station right alongside at Barons Court, so public transport accessibility is less of a bother.

All of this assumes that the extra road capacity doesn’t induce a large amount of traffic, but since that’s apparently the point of the exercise it’s likely that disproportionately more traffic will try and get down the new roads.

The feasibility study recognises the traffic displacement issue and immediately agrees with a key point of my analysis:

the longer the tunnel the less traffic would be likely to use it…opportunities to remove or reduce the existing surface road network diminish as tunnel length increases

This has two key implications – the first is that the long tunnel options are more expensive and have less benefit in terms of re-routing cars, the second is that far from being a key artery into the heart of the city the A4 is to a significant degree a bypass of Chiswick High Road and King Street, and this function would have to be retained in some form in any reorganisation.  The more exits you remove from your buried A4 the more traffic displacement needs to be accommodated on the surface (particularly at Hammersmith) and the less land there is available for all the supposed benefits of the scheme (particularly at Hammersmith) and the whole scheme basically disappears up its own backside.

Having literally dug themselves into a hole, however, the consultants reach for the only possible solution – more tunnels, specifically underground junctions with roads up to the surface that can be used by the displaced traffic to reach its original destination.  There are a few issues:

  1. They are most urgently required at the two major junctions, Hogarth Roundabout (for the A316 traffic joining the A4 eastbound) and Hammersmith Gyratory (for the 40,000 a day leaving the A4)
  2. You’ll need a six lane tunnel now, not a four lane one, which is a fair bit more expensive
  3. You’ll still need a surface road for local access between the ramps, which is another two lanes
  4. So we’ve still got eight lanes of road, only we’ve added another couple of miles of tunnel plus some land take where we really didn’t want to, such as bang in the middle of our new Hammersmith town centre or in residential Chiswick along the A316.

There is in fact a plan knocking about for this:

There are four problems I can see immediately with the Chiswick end

  1. the south-to-east tunnel goes through the moribund Hogarth Business Park, which as mentioned before has just had planning permission granted for a large luxury housing development smack on the line of the tunnel.
  2. it also goes through William Hogarth’s house, which is one of the most important historical buildings in the area
  3. the distance from the A316 to the A4 where they show their tunnel is only about 125m, which is shorter than their figure of 200m ramps needed to get enough headroom in – we’re presumably going between the surface A4 and the buried tunnels too, in a triple decker sandwich
  4. the west-to-south tunnel appears to emerge from underground in Chiswick’s cemetery, then turns sharp right to go up Corney Road, which is a heavily traffic calmed tree-lined residential street, before terminating at a T-junction with the A316.  This is unlikely to be popular in the area.

The study is actually more pessimistic than me about the displacement effect of a long junction-free tunnel – the percentage of A4 traffic they estimate is displaced by each option is:

  1.  0% (as it’s a direct on-line replacement)
  2. 40% (so around 48,000 vehicles/day compared to my 40,000 assumption)
  3. 50% (so around 60,000 vehicles/day)

Even without the side tunnels the costs of the bored option are significantly higher – they’re assuming twin tunnels, thus doubling the length bored, bringing both options 2 and 3 in at well over a billion pounds for effectively no gain in capacity.  Curiously they don’t state whether the long tunnel options are four or six lane tunnels, although the rejected option of a 20m diameter double stacked tunnel implies six (the architect’s report shows a double decked four lane tunnel which appears to be about 14m diameter).

Given all this, the inescapable conclusion is that the two long tunnel options perversely put more traffic through Hammersmith Gyratory, destroying the entire basis of the scheme.  Flailing around for a solution to this they reach for, yes, more tunnels, in this case a north south one from Fulham Palace Road to Shepherds Bush Road.  This is clearly nonsensical – both roads are inner urban high streets, not arterial roads, it needs to be deep enough to go under the London Underground *and* the new road tunnels – and it’s immediately rejected.  In any case it didn’t provide enough relief to the gyratory.

Bizarrely, having effectively trashed the long tunnel options on traffic grounds and pointed out that they force you to keep substantial surface roads on the A4 corridor, the report blithely goes on to assess how much developable land is available between Hogarth Roundabout and Barons Court.  This is not actually any of the three tunnel options – the surviving Option 1 has the entire current A4 on that land, for instance, barring a tiny bit in the middle where there are no existing surface roads.  Instead this appears to be based on GLA work which in turn appears to be based on the original architectural sketches rather than any feasible scheme and an FoI is going in for this.  Even assuming this magical fourth option it’s not clear that the released land is going to cover the costs – the developer has to make a profit and someone has to pay to build the blocks of flats, too, the earliest feasible data for getting a return would be some years after you’ve started pouring money into the hole in the ground.  There’s also no indication of any increase in public transport provision such as new bus routes.  Finally the report cheekily makes a bid to push further road tunnel research (i.e. side tunnels and underground junctions) into TfL’s proposed rework of the Hammersmith Gyratory for cyclists.  Alex Ingram has been working on this at hfcyclists, and his latest post  is here, covering the strange Gilligan/RBKC-linked death of Cycle Superhighway 9, which was supposed to run from Hyde Park Corner to Hounslow via Hammersmith and Chiswick and would actually, if done properly, provide an alternative to driving.  It’ll be interesting to see if RBKC’s transport people are less fastidious about having the end point of Hammersmith and Fulham’s Option 3 dumped on them.

Where are we now?  A lot depends on TfL’s response, which as usual will be coloured by Boris jumping the gun on the radio and declaring himself in favour of a scheme costing £1.7bn, which can only be one of the long tunnel options.  It’s clear from the feasibility study though that the entire scheme as currently understood and which Hammersmith & Fulham council have lobbied for, stirred up the residents to support, got Bill Bailey to back and stuck a massive sign on the town hall to advertise to motorists on the A4, isn’t actually workable – either you replace the flyover like for like, free up almost no land and make sod all difference to the severance along the A4 or you spend well over a billion pounds burying the A4 in a manner that forces half the traffic off it onto a surface network that can’t cope without taking up most of the land that’s supposed to be redeveloped to pay for it.  As a pilot scheme for the resurgent roads lobby it’s hard to think of one that could have more internal inconsistencies and logical paradoxes.  That’s what’s always killed London’s road maniacs and we will be making damn sure it kills them again this time.

Set your Twitter account name in your settings to use the TwitterBar Section.