How Little New Road Can You Get Away With in Designing Large New Settlements?

1. All of the above statements are true (to a point):

• Building roads induces new traffic
• Taking links away on a traffic network can reduce traffic and adding them can increase it
• Excessive car use is a major cause of air pollution and CO2 emissions
• We face a climate emergency

Does that mean that we should never expand the road network?

That is hard because settlements spawn, extend and grow always have and always will through a road grid, well before the car is invented.

Its all an issue about how roads are used and whether they are used and planned as part of a wider network to minimise the above network effects.

2. Because the following are also true (to a point):

• If people cant afford a house near there place of work and they only have access to a car they will drive ever further out to afford housing;
• If a network is incomplete then unnecessary miles will be driven causing pollution and congestion at its most overloaded points
• Many people living in a new settlement will be locals and be driving on the roads anyway and so wont add net to traffic.

3. Which of these effects predominate will in all cases depend on the topology of the network, the management of the network and the disposition of land uses, in other words its transport and land use planning.  Each case is unique and in large sites can only be discovered through modelling.  Yes its a highly mathematical issue, an issue of graph theory, and one that can only conclusively be concluded through up to date traffic models that fully take into account multimodal trip distribution and assignment.  Its tough stuff – if you want you can look up in various contexts as the Downs-Thomson Paradox, or the Lewis-Mogridge Position.  A more sophisticated variation is known as The Braess Paradox which building new roads in the wrong location can lead to longer travel times for everyone, even without induced demand, because new roads may lead more car drivers to the weakest most congested  links in the network. The reverse may also be true: removing roads may even improve traffic conditions.  Both of these effects occur because each driver chooses the mode (in the first theory) or the route (in the second)  that is quickest without considering the implications his or her choice has on other drivers.  However it all depends on the network and the efficiency of alternative modal choices.  In the US for example traffic growth is highly correlated and elastic to road growth, in the Netherlands many new roads have been built to airports and ports but overall traffic growth for commuting to major city has fallen because of the promotion of alternative modes.

4. This dependence on network conditions and transport policy mean it is unwise to take an absolutist position, no roads ever even if it means building no houses,  Such a position means all of the negative impacts in section 2 apply and could mean that traffic congestion and CO2 emissions get worse.

5. If we want to both build new homes, move towards sustainable transport and reduce CO2 and other emissions we have to take an integrated design led approach to both land use and transport ensuring the network as a whole reinforces sustainable modal choice minimsing the effects of induced traffic.

6. International experience suggest the following are most successful

• Ensuring that major development is based around transit, walking and cycling
• But that isn’t enough, transit, walking and cycling have to be easier, safer and and faster (filtered permeability), which means taking traffic away from residential areas, which in some cases will mean ring roads and bypasses, as well as a dense ‘dutch’ 240m cycling grid and all intersections to safe dutch standards for cyclists.
• Roads should be restricted to essential access, emergency service and PSVs, an major roads only access to major settlements, access to employment premises and transhipment locations (motorways, logistics hubs, ports and airport.

7.  How few major roads can you get away with? You can work this out by calculating a desirable modal split in favour of sustainable modes (say 60=70% low country urban areas split) then working backwards from that (households converted to commuters (% 55.2 working of population of working age x % 77  age dependency ratio – which works out at around 42.5%) .  So a Garden Community of 10,400 homes (one secondary school ) with 23,300 approx population will generate 9,900 workers, assuming 10% working from home equals 8,810  commuters.  Assuming a low 30% car share (assuming a highly sustainable design, infrastructure  and public transport infrastructure) , that’s around 2,643 car commuters a bit less assuming some may have double occupancy.

We can work out how many arterial (that is major access to the settlement) roads this equates to by referring to this official government table (table 9)  on the free flow capacities of roads.

A 40 mph (60 kph approx) carriageway has a free flow capacity of 1,380. So a single carriageway highway with access to two major sources of employment (one in each direction) so we are just fine.  We dont need more that the bare minimum single access road in each direction you would need to service the shops and industries in the settlement anyway.  Any extra connections would add resilience and extra choices to the network, but we don’t need to go overboard.  If we build at higher densities in a sustainable way we don’t need a huge MK like grid, the only occasions where you do need more than one arterial road is where you are building a larger garden community of more than one district where a minimal arterial grid is needed, and most roads accessing neighbourhoods need only be major collector (link) or local collector (feeder) standard.

Once you start pushing modal share by car below 50% you start to take vehicles off the network, the key is is the % below the number of local residents already driving through the area to work?  If it is s sustainable new settlement which means people can reduce their communing distances and potentially not need to drive at all can theoretically reduce not increase traffic.  And we also know from areas of Germany, the Netherlands, Denmark and Sweden that have designed new settlements in this way and seen a reduction in car commuting it can be done, no need to be cynical.

8.  None of this is saying the way we design and plan major housing in the UK is all right.  Strategic growth locations in the wrong place, and /or not designed around sustainable modes from the get go will make matters worse.  Which it is why every agency, especially you Homes England and you MHCLG must see ensuring sustainable infrastructure and design is not just a ‘risk’ but task number 1 in planning.   Whose job is it. mainly county councils and combined authorities, but in most home counties growth is most needed there are one man and a dog outfits unless they have already done growth deals.  If we want to solve out housing crisis then this field is where we currently have the biggest capacity, skills and research gap.