Wednesday, 27 January 2016

Understanding traffic flow (in Cambridge)

Cambridge has serious traffic problems. There is congestion every day, not just when there is an accident (all too frequent on the A14). There is funding available now to ease the problem but how to spend it is controversial. Also it would be good to promote sustainable travel choices at the same time. In practice there are many variables that affect why people travel, where, when and how, and how to make these trips easier.

The Greater Cambridge City Deal has specific proposals for different routes into the city - the one that affects me most directly is Milton Road but I have been also looking at the larger picture. Here are some of my findings which I find surprising.

How and why people travel
34,000 people live and work in Cambridge
16,000 people live in Cambridge and commute out
51,000 people live outside Cambridge and commute in

To minimise traffic coming into the city it would make sense for people who work in the city to live in or near the city as far as possible. One third of working people living in Cambridge actually work outside the city but far more people come into Cambridge than go out. If it were possible for the 16,000 people going out to swap places with 51,000 people coming in it would reduce the inwards pressure by nearly a third. However, there are lots of reasons why people choose to live where they do and that is obviously not possible.

Some of the people who live in Cambridge and work outside commute to London - but only about 2000 per day, judging by the numbers travelling by train. It feels a lot more than that when I am fighting for a seat at 7.45am.

I would have thought that no-one would choose to travel at peak times unless they had to for work or other business. However 14% of peak hour trips are for shopping. An even larger proportion (17%) is health related - which I suppose relates to the fact that Addenbrookes hospital generates an awful lot of traffic.

Sustainable travel choices means public transport, cycling and walking because these use less energy, generate less carbon emissions and less pollution. For people commuting within Cambridge more people cycle (43%) than use a car (26%). This is not surprising considering that Cambridge is well known as a cycle friendly city However, 14% of people coming in from S. Cambs (which completely surrounds the city) also cycle - slightly more than the number who use a bus. The number of people coming in by car is still very high, as shown in the chart.
Travel mode for commuters working in Cambridge. Data from [1]

Park and ride services are under-utilised especially by commuters
Cambridge has five park and ride car parks intended to discourage cars coming into the city. However, more than half of their uses on week days are less than 8 hours and the park and ride buses are most used between 10am and 4pm which suggests that commuters are not their main customers. The Trumpington park and ride is mainly used for very short stays - one third of them are less than one hour even on week days. This is probably because it is close to Waitrose, and also quite close to Addenbrookes hospital.

All the park and ride car parks are under-utilised. Even the busiest one (Madingley Road) is rarely more than 60% full. Price is clearly part of the problem because usage decreased significantly when charges for parking (as well as for using the buses) were introduced. However, I suspect another reason is that the park and ride buses have very few stops so unless you want to travel right into town they are not very convenient.

Term time traffic is only a little heavier than at other times
The chart below shows traffic counts through the day for four arterial routes (Hills Road, Huntingdon Road, Milton Road and Trumpington Road) that are often congested at peak times. There is more traffic in the morning during term time but only 17% more even at the peak. Also there is a great deal of traffic all day. The morning and evening peaks are only 15% more than mid day.
Traffic flow at four automatic traffic counter sites [1]

This is in stark contrast to the pattern of cycle traffic which is mainly at peak times. Probably most of the cyclists are commuting for work or study.
Cycling flow at selected automatic counting sites [1]

Small increases in traffic cause massive increases in delays
The small increase in traffic during term time causes a large increase in congestion as measured by the proportion of buses running to time (defined as less than 6 minutes late). During the morning peak on weekdays Nov/Dec 2014, 10% were late during half term but on normal days it was 29%. As the day goes on the lateness increases - they never seem to catch up.

The traffic in the north of Cambridge is expected to increase because of new residential developments and general growth in the area. Traffic modelling has been used to estimate the effect. For the Milton Road route, (from Mitcham's Corner to the A14, 3.5 miles) a 12% increase in morning peak time traffic is expected to bring a doubling of travel time (southbound traffic, from 10 minutes to 21 minutes) as shown in this table.

Impact on non-bus travel time for morning peak traffic on Milton Road (Mitcham's Corner to the A14. From [2]
ContextJourney time (seconds)Mean mph
Base (2014)
2031 Do nothing (12.0% increase in traffic)
2031 Do everything (10.1% increase in traffic)

Proposed measures for Milton Road are aimed at buses and cyclists but will improve car travel times too.
The 'Do everything' scenario includes a whole package of measures designed to encourage people to cycle or use the bus rather than drive, so overall traffic does not increase quite so much. Included are:
  • Bus and cycle lanes segregated from normal traffic (currently southbound cyclists mostly share with buses while northbound the cycle path is sometimes shared with pedestrians)
  • Traffic signals giving priority to buses.
  • Restrictions on turning at major junctions - partly to reduce the time for traffic lights to cycle round and partly to allow bus lanes to continue right up to the junctions without having to merge with traffic turning left.
I do not use Milton Road regularly but I did once come close to a nasty accident - when I was cycling on a shared pedestrian-cycle path and a car pulled across in front of me to get into a car park. Having a raised segregated cycle path would encourage drivers to take more care at those points, I hope.

The impact on bus travel time will be dramatic (see table below) but apparently even non-bus traffic will travel almost as fast as before. This is mainly due to the changes to traffic signals.

Impact on bus travel time for a 12% increase in morning peak traffic on Milton Road (Mitcham's Corner to the A14. From [2]
ContextJourney time (seconds)Mean mph
Base (2014)
2031 Do nothing
2031 Do everything

No direct route for cyclists using the Milton Park and Ride
I suspect giving cyclists a segregated lane on Milton Road may not be enough to encourage more cycling on this route. It would help if there was a safe direct cycle route from the Milton Park and Ride. However cyclists either have to negotiate the A10 and then the roundabout with the A14 to reach Milton Road or go 0.6 miles out of their way to the cycle bridge over the A14. This may be the reason why the Milton park and ride is the least used. It is rarely more than 30% full.

There are downsides: trees and verges will go.
Unfortunately there are downsides to these plans. Apart from the cost, the main issue is that the extra lanes take extra space so trees and verges will have to go. Also, some of the turning restrictions will bring longer journeys for residents on Milton Road and smaller roads off it.

Integrated ticketing would encourage more people to take the bus.
The intention is that more people will cycle or take the bus because it is quicker. However, travel time is not the only reason people use a car rather than a bus. One issue is the complexity of journeys. For example people are six times more likely to use public transport if there is a direct bus than if they have to take two buses. Also, ticketing in Cambridge is very complicated because there are different companies involved. The charges are different for the normal Stagecoach buses, the park and ride buses and the guided busway.

Not only is traffic a complex problem, any solution has to involve action by a variety of stakeholders, including public transport providers, preferably all working together. The council by themselves are very restricted in what they can do, however much money they spend.

[1] Cambridge Access Study (Mott MacDonald for Cambridgeshire County Council) July 2015


  1. Where traffic lights are staggered they take that much longer to cycle and I question whether this is necessary. I remember when a junction in Ely was changed, what seemed to have been unnecessarily, and did query it but Highways won't listen! It wasn't a danger zone and now causes very long delays as it has the addition of pedestrian controlled add-on now also. This is similar to the Milton Rd, Union Lane junction. When they were timed there's only a 4 second spot for cars exiting UL with a long wait in between. What was wrong with the original layout there with a right turn lane into Arbury? This was lost by adding the cycle lane. Surely safer and more efficient to utilise verge and footpath at that point for cyclists. There aren't many cars traversing UL in off peak. Better traffic light mngt is called for.

  2. All commercial car parks that exit via a shared footpath/cycle way should be clearly signed. A pedestrian isn't likely to walk across if a car is coming in or out, cyclists must observe similarly. Drivers will see a pedestrian as long as view isn't obstructed, will they see a fast moving cyclist when view is restricted? Signage for both and care is the answer.

    When bus and car lanes merge at traffic lights there'd be a maximum of what, two or three cars, not exactly a big hold up.

    Is it really that much worse to use a safe route that's a bit longer? If anyone's keen enough to cycle instead of using bus from Milton park and ride, do so safely.

    Longer journeys will affect all Cambridge residents if turn offs from Milton Rd are blocked, not just those living along it, at least doubling everyone's carbon footprint.

    As it's well known that Park and Ride usage dropped when machines and fees were added, why did the council do it? If they needed additional revenue couldn't they agree a deal with the buses? Why did they not anticipate any of this when it was set up. Something needs to change there. Also why don't all buses use the bus stops? This needs to be discussed fully along with other bus issues.

    No-one wants Milton Rd turned into an A14 lookalike.

    1. As I recall from my highway code, cars are supposed to avoid cutting across pedestrians on footpaths. Unless there is a give way mark on the road, cars aren't supposed to cross in front of you. However I agree it can be hard for drivers to see when cycles move so much faster than pedestrians. My near accident is an example of how sharing with pedestrians is not risk free and not ideal.

      Although the park and ride usage did drop when parking charges were added this was only about 10-15% - they were still under utilised even before that.

  3. This comment has been removed by a blog administrator.

  4. This comment has been removed by a blog administrator.