Thursday, 1 November 2012

Traffic on the B1050 (Northstowe Joint Development Committee)

As part of the work of the Northstowe Joint Developement Committee (mainly between South Cambridgeshire District Council, and Cambridgeshire County Council) a local Longstanton District Councillor, Alex Riley, has done some work analysing traffic patterns on the B1050 which joins to the A14 just North of Bar Hill.

While his focus is plainly the other side of the A14 the conclusion he reaches - that without an upgrade the road  just won't cope with the additional traffic caused by Phase 1 of Northstowe - should be a cause for serious concern for Bar Hill residents, especially those that use the A14 to get to work/ school/ college every day!

I've attached the complete report below, as oriignally circulated by Cllr Riley, a more detailed map (from which I've taken the image to the right) is available here;

The use of the B1050 within NorthstoweThe Northstowe Area Action Plan (NAAP) quite clear in its sections on road infrastructure that access to Northstowe will be via two new roads, one coming off “an improved Hattons Road from the A14” and the other being a “new road from the A14…in the vicinity of the Dry Drayton junction into the southern end of Northstowe” (See Policy NS/10).

The NAAP in D6.5 on p47 explains the purpose of the B1050 for Northstowe: “A link to the Longstanton West Bypass / Station Road, Longstanton junction will provide additional access onto the local road network to the north, allowing adjoining villages to access the facilities and services at Northstowe.” The Inspectors clearly thought the B1050 was not up to the job it is now being expected to carry out. The purpose of the access was for other villages, not Northstowe itself.

However, the Phase 1 proposals not only add all Phase 1 traffic to this already busy minor road, but also impose 4 junctions, 2 sets of traffic lights and 3 pedestrian crossings in a few hundred metres. The question is: can it cope?

The need for traffic simulationSimulation uses a computer to generate random traffic, but at a specified average rate. As a former statistician I completely believe that simulations do provide a valid means of checking something like the proposed changes to the B1050, provided that the underlying model and vehicle rates are correct. I am a layman as regards traffic simulation.

The road layout used by the simulation looks accurate and I’m not able to comment on the cycle times used for the traffic lights. Many who have seen the simulation have commented that the rules applied to vehicles for speed, acceleration and deceleration are completely uniform – something which we all know does not happen in real life.

The key reason for locating the Phase 1 Local Centre on the B1050 was so that “passing trade” would allow the shops to be more viable than otherwise. The simulation does not allow for a proportion of traffic in both directions to leave the B1050 and subsequently rejoin it.

So we can assume that even an accurately-specified simulation will tend to give an idealised impression of the operation of this road. By which I mean that it will be shown as working more smoothly than it would in real life. But it can still be a valuable tool.

Observed problems with the current simulationWhen the simulation was shown to the Northstowe Joint Parishes meeting, there was a stunned silence. I believe that this was because all attendees are acquainted with how busy the B1050 is in real life, and no-one could believe how adding Phase 1 traffic, 4 junctions, 2 sets of lights and 3 pedestrian crossings could apparently improve traffic throughput. So there was a fundamental credibility problem with the current simulation.

Also, because it operates speeded up by a factor of 2 it is hard to assess the average vehicle speed and general operation.

In order to give check the validity of the simulation, I said I would check out B1050 traffic flows.

Traffic simulation parametersThe following values need to be input to the simulation for it to work:
  • Through vehicle flow rate north to south 
  • Through vehicle flow rate south to north 
  • Vehicles generated by Phase 1 itself joining the B1050 (in both directions) 
  • Phase 1 Vehicles leaving the B1050 (in both directions) in order to return to Phase 1 
  • Proportion of vehicles in each direction which divert to Phase 1 (e.g. for shops) 
  • Traffic light cycle times 
  • Vehicle acceleration / deceleration rates and average speed 
Items 1 and 2 are relatively easy to verify via a traffic survey. Items 3 and 4 are necessarily subjective and will be based on assumptions of: a) number of dwellings (assumed to be 1500), b) average car ownership/dwelling and c) assumed proportion of vehicles travelling at a given time.

Item 5 is not covered by the simulation, which is a surprise given the importance applied to passing trade as a factor for the location of the Phase 1 Local Centre.

Items 6 and 7 are possibly standard values. In any case it would be easy to adjust them in order to optimise performance of the simulation.

Traffic levels currently used by the simulation[In order to reduce the amount of data, I propose only to deal with southbound data for the morning rush hour. There are also data from all sources for northbound morning traffic and for the evening busy period, but the key points are borne out by the southbound morning figures.]

The morning figures are based on peak flow rates between 8 and 9am. They are 923 vehicles per hour (vph) heading from North to South along the stretch of the B1050 in question, i.e. just north of the “golf course roundabout”. Everyone seems agreed that this is the right piece of road to survey. It means that issues such as rat-running through Longstanton, or traffic generated by Home Farm, whilst both significant issues, can be disregarded for the purposes of this survey.

However, the 923 vehicles consist of 405 generated from within Phase 1, leaving only 518 vph of north-south through traffic. This last figure is based on a survey carried out in 2006.

The figure of 405 vph as a peak rate generated within all parts of Phase 1 (including the Employment Area) and heading south seems reasonable.

Quite why 6-year-old survey data have been used, I cannot explain. The Longstanton bypass was completed in 2008 and traffic levels to have increased dramatically in the past 6 years.

Traffic levels recorded by a County Council survey in November 2011Half-hour commencing - Total vehicles - vph
7.00 - 468 - 936
7.30 - 526 - 1052
8.00 - 479 - 958
8.30 - 448 - 896
9.00 - 347 - 694

The weather that morning is described as: Cloudy, damp, drizzle. This would have tended to reduce inessential travel. So the peak value recorded is more than double that used for the simulation.

Traffic levels recorded by myself on Tuesday 25/9/1215-mins commencing - Total vehicles - vph
7.25 - 257 - 1028
7.40 - 234 - 936
7.55 - 214 - 856
8.10 - 194 - 776
8.25 - 180 - 720

I intended to start at 7.30 but the traffic rate was so great I started 5 minutes earlier. I wish I had started even earlier.

Conclusions from the surveysI deliberately did not try to find out the traffic flow rates used in the simulation before doing my own survey, and I had no idea that the CCC survey even existed.

Recent government figure have shown that in the past few months vehicle fuel consumption has gone down, as a consequence of the recession. This could explain why my own figures are lower that the CCC survey. As the recession lifts, traffic levels will increase.

It is no use saying that we have to look at figures for 8am – 9am. The reason the B1050 is so busy much earlier is because of the state of the A14. No-one in their right mind would try to go in to Cambridge on the A14 from the B1050 between 8 and 9 am.

Both the recent surveys show through traffic rates that are twice the rate used in the simulation.

Simulations shown on 4 OctoberI told WSP that I needed them to rerun the simulation with north-south morning through traffic levels of 1000 vph. We were actually shown it at maximally 700 vph, at which point the proposed Phase 1 road system was not really coping. I can only presume that WSP tried the simulation with higher rates of through traffic but found the results to be catastrophic.

Using CCC’s own professionally-executed traffic survey the road would not have been functioning properly between 6.30am and 9.30am since through traffic rates were around or greater than 700 vph. Similar results apply during the evening.

SummaryEven given the idealised functioning of the simulation, the proposed road arrangements for Phase 1 simply will not cope.

We should not forget that traffic rates in this entire area are steadily increasing. By the time Phase 1 is complete the through traffic levels will be greater than at present.

Alex Riley