Further to our response to the public consultation, we have prepared seventeen key questions for the City Deal Executive Board to answer in relation to its decision on 13 October 2016 to progress a proposal to build a busway from Cambourne to Cambridge.
The Benefit-to-Cost (BCR) ratio of Option 3 is 0.21 (pp 40-42 of Appendix 1: The Options Assessment Report), which represents ‘very poor value for money’ according to government guidance. Option 3a would cost less than Option 3, but realistically not by more than a third. That still requires the benefits to be six times greater than currently estimated in order to bring the BCR up to 2, the usual threshold for public money to be spent on infrastructure.
By way of comparison, reinstating the railway to Haverhill was rejected by the City Deal on the basis that its estimated BCR is only 0.99.
Q1. Why is an estimated BCR of 0.21 deemed sufficient to progress this scheme?
The City Deal Executive Board claimed in a Cambridge News article on 18 October that the planned Cambourne-Cambridge busway would likely attract as many, if not more, passenger journeys as the existing Guided Busway. Further it claimed that the investment cost for the new busway would work out to be £1.26 per passenger journey. The reasoning that led to such a precise figure is fatally flawed.
Existing Guided Busway as a comparator
It was claimed that “the settlements along [the Cambourne-Cambridge] route will have more residents than currently live along the existing busway.” The population of settlements between St Ives and Orchard Park is 35,750. That is 18% greater than the currently-anticipated 2031 population of settlements along the route of the proposed Cambourne–Cambridge busway (see Appendix A).
This comparison was used to justify an assertion that the Cambourne–Cambridge busway “may well get more” than 3.7 million passenger journeys a year. But that figure covers all services that use any part of the Guided Busway: it includes passenger journeys between Peterborough, Huntingdon, Godmanchester and St Ives, and between the Trumpington Park & Ride (1,340 spaces), Addenbrooke’s, the Cambridge rail station and city centre. Therefore the figure of 3.7 million is irrelevant to the Cambourne–Cambridge busway, and referring to it is misleading.
A figure for the number of passenger-journeys between St Ives and Orchard Park has not been published to date because the bus operators deem it to be commercially sensitive. But that figure (reduced by 15% to account for the difference in populations served – see above) would be the only meaningful comparator for the proposed Cambourne–Cambridge busway.
The monetary figure used to calculate the cost per passenger-journey was the estimated capital construction costs; it omitted the maintenance costs and any service subsidies. The consultants’ report indicates that maintenance adds another £66m, and subsidies a further £13m (though this figure excludes vehicle depreciation/leasing costs see Service Subsidy below). That brings the 60-year cost to £221m.
In order to calculate the impact of the busway, it is necessary to know what the level of bus patronage would be if the busway were not built. Busway services would likely replace two existing services: the Citi 4 (three buses/hour) and 18 (one bus/hour). Patronage of these, extrapolated in proportion to population growth, provides the baseline patronage from which we can measure any increase attributable to the busway.
The busway would provide benefit only during peak times. At other times, the A428, St Neots Road and A1303 provide a direct and only marginally slower route. We’ll assume (pessimistically) that peak-time congestion in 2031 spans two hours at each end of the working day. (Currently it is about 40 minutes, according to analysis by Dr Ian Lewis’s team of Citi 4 bus journey data.)
Extrapolating to 2031
From 2016 to 2031, population is expected to grow 75% in settlements that would be served by the busway (see Appendix A). All other things being equal, we would therefore expect demand for the existing bus services to increase by 75%. If we assume that the current service frequency of 4 buses/hour optimally matches demand, then we would expect 7 buses/hour to run in 2031 if no busway were built.
Modelling for the proposed busway anticipates demand for 9 buses per hour, meaning that only 2 buses/hour can reasonably be attributed to additional demand created by the busway.
If we assume that each bus is occupied by 68 passengers (i.e. 90% of a Guided Busway double-decker’s capacity). That equates to 544 additional peak-time passenger journeys per day attributable to the busway.
The choice of 30 years for depreciating the capital costs is reasonable, though somewhat arbitrary. The report considered 60 year costs, which is a fair estimate for the useful life of the asset. As both time spans are justifiable, both should be used for calculating a per-passenger-journey cost.
If we assume that shoppers and leisure travelers replace commuters during the weekend travel peaks, 544 extra passenger journeys daily equates to just under 200,000 extra passenger journeys per year. That is 5.5% of the 3.7 million claimed.
The 60-year cost of £221m works out at £18.40 per passenger journey, or £36.80 per return journey, or £13,400 per passenger per year.
The 30-year cost of £181.5m works out at £30.25 per passenger journey, or £60.50 per return journey, or £22,000 per passenger per year.
Those figures reinforce the ‘very poor value for money’ 0.21 BCR (benefit to cost ratio) already calculated by the City Deal’s consultants.
We’ve also examined commuter numbers to give the above figures a reality check. The 2011 Census commuting data shows that 303 people commuted by bus to Cambridge from Cambourne, Highfields Caldecote, Hardwick, Toft, Comberton and the Eversdens (see Appendix B). That represents an 11.3% modal share of all commuters to Cambridge. That’s already well above the national average of 7% (DfT Modal Comparisons 2015).
All other things being equal, population growth between 2011 and 2031 should see the number of bus commuters grow by 85% to 560.
If we assume that commuter figures are a close proxy to peak-time travellers, then the 272 extra passengers per day attributable to the busway (dividing by two the 544 extra passenger journeys calculated above) would equate to a 48.5% increase in bus patronage, taking the modal share from 11.3% to 16.8%.
That seems reasonable, and therefore provides confidence that the investment cost calculations above are also reasonably accurate.
Q2. Why did the Board go public with a severely underestimated per-passenger-journey investment cost?
Q3. Does the Board accept our calculations? If not, will the board publish how it calculates the investment cost per new passenger journey attributable to the proposed busway?
Q4. Will the Board obtain a figure for passenger journeys between St Ives and Orchard Park so as to have a meaningful comparator to the proposed Cambourne–Cambridge busway?
Even though the City Deal cannot commission work on the Girton Interchange, the business case for adding connections between the A428, M11 and A1307/Huntingdon Rd should still be examined. A BCR analysis of this would set the other proposals in context.
The Oxford–Cambridge Expressway project is currently looking at the A428, and will most likely recommend the addition of connections between the A428 and M11. This would have a very material impact on the business case for Option 3/3a.
If the Oxford-Cambridge Expressway receives funding from the second phase of the Road Investment Strategy (RIS 2), construction could start soon after 2020, well within the timescale needed to support developments at Cambourne West and Bourn Airfield.
A428–M11 connections would facilitate movements from the west (Oxford, Milton Keynes, Bedford, St Neots, Cambourne, etc.) to south Cambridge, including the Biomedical Campus, science parks along the A505, and Stansted Airport.
Currently, traffic aiming for the M11 has to queue on Madingley Hill (accounting for about 20% of the traffic there) or ‘rat-run’ through Toft, Comberton and Barton to reach the Barton Interchange. This situation is set to worsen as new housing is built around Cambourne, blighting the lives of residents of those villages, and delaying people who have no choice but to drive.
Further benefits would include:
- A Park & Ride could be located at the Girton Interchange, replacing the one on Madingley Rd (see next section).
- There would be no immediate need to build a Western Orbital busway to Trumpington P&R, saving an estimated £30m.
With traffic bound for the M11 and for the Madingley Rd P&R all displaced to the A428, traffic levels on the A1303 would fall dramatically and therefore flow freely.
There is plenty of spare capacity on the section of A428 between Madingley Mulch Roundabout and the Girton Interchange, where the AADF (Average Annual Daily Flow) is 17,454 vehicles, compared with 30,609 west of the Cambourne interchange.
Park & Ride
A P&R at the recommended site (Crome Lea Farm) will undermine rural bus services (including along the new busway – see section on Service subsidy). It will also concentrate traffic at the Madingley Mulch roundabout, adding to congestion there. These are well-documented side-effects of park-and-ride.
A P&R at the Girton Interchange would catch longer-distance traffic from the A428 and A14, without being so attractive to village residents who, we all hope, will be provided with a high-quality express bus service to Cambridge, complemented by a mix of rural and community services.
The Girton Interchange location is also less environmentally sensitive than any of the potential sites identified close to Madingley Mulch roundabout. There will be at least 12 hectares of land enclosed by the road network, ideally suited for a P&R site (see Figure 1).
Officers claim that the benefits of ‘fixing’ the Girton Interchange would be quickly lost, with other traffic filling the A1303, replacing the traffic displaced to the A428–M11.
It is true that demand to travel on the A1303 is suppressed by consistent congestion on Madingley Hill at peak times. This is proven by the large number of drivers currently using alternatives routes through Toft, Comberton and Barton to avoid the A1303.
However suppressed demand would be channelled onto the A428, not the A1303, if:
- The right-turn onto the M11 were also eliminated. This would force drivers arriving from the west to stay on the A428 to the Girton Interchange (which should generally be a quicker route anyway).
- The existing P&R at Madingley Road were closed in favour of a better-connected site at the Girton Interchange. This would require drivers wishing to use Park & Ride to continue on the A428 to the Girton Interchange.
A right-turn ban onto the M11 would also allow for the removal of the right-filter lane on the bridge, making space to either improve cycling provision or add a bus lane, obviating the need to build an expensive new bridge.
To ensure that this effect is not temporary, Smarter Cambridge Transport has proposed using Inbound Flow Control to queue traffic at specific locations at times when the volume of traffic would create congestion. This would entail holding traffic temporarily in a queue before 800 Wood and on the M11 exit ramp (with localised widening of the road). It would incentivise use of Park & Ride and minimise the length of bus lane required to bypass any congestion.
Q5. Will the Board commission an expert evaluation and BCR analysis of adding connections at the Girton Interchange, along with a P&R there (replacing the one on Madingley Rd), and using Inbound Flow Control on the A1303? If not, will be Board please provide justification?
The consultants’ report indicates that the net ‘nominal modelled operational costs’ of the bus service using the new busway will be £13.3m over 60 years, excluding fleet costs. That works out as an annual subsidy of £220,000 per year. Add fleet leasing or depreciation costs, and that figure would most likely double.
Does this figure also take into account the competition created by having a P&R close to the Madingley Mulch roundabout? People (with access to a car) living in Cambourne, Bourn Airfield, Highfields Caldecote and Hardwick will have a choice between using the busway or driving to the P&R.
If the P&R service is cheaper, more frequent and running for longer hours, then people will choose that over the busway. To counteract this, the Council would need to undertake to ensure the bus service from Cambourne is cost-comparable to using P&R (for individuals and families), and runs as frequently and for as many hours as the P&R service. That will almost certainly require even higher subsidy than indicated above.
Only if and when the A428 becomes heavily congested will the balance start to shift back in favour of the busway, but that may not be for many years.
Q6. Can the project team confirm that the effects of competition between P&R and busway services have been taken into account?
Q7. Will the project team publish their analysis of demand for P&R and busway services?
Cycling and walking
There is already conflict and minor collisions between people cycling and walking along Burrell’s Walk, Garret Hostel Lane and Senate House Passage. Has there been any consideration of the capacity of this route to carry a large increase in cycling via the Coton cycle path to/from the new busway and intensified West Cambridge site?
There is also no longer a proposal to improve the very poor provision for cycling and walking along Madingley Road. This omission undermines a key objective of the NW Cambridge site, set out in its Sustainability Statement: “Create an accessible, pedestrian friendly site, with good connectivity to surrounding areas.”
Q8. What options are being considered to address the limited capacity of the Garret Hostel Lane cycle route?
Q9. Why is there no longer any consideration given to the need for a high quality and capacity cycling and walking route along Madingley Road?
Bus route from Grange Road
The busway ends at, or just short of, Grange Road. The quoted 28-minute journey time is for Cambourne–Grange Rd–Cambourne. But Grange Rd is not a destination and buses will not be terminating there, though officers have admitted not knowing where in the city people will want to travel on to.
The shortest route to Drummer St is severely constrained and unsuited to standard buses: Silver St is narrow, and the route includes sharp turns at Pembroke St and St Andrew’s St, creating a dangerous conflict with the many people who cycle and walk along this route. Pembroke St is frequently congested with cars queuing for the Grand Arcade car park, which unavoidably holds up buses.
The enclosed nature of these streets also means that vehicle emissions frequently concentrate to dangerous levels. Average levels of nitrogen dioxide on Pembroke Street have exceeded the EU-mandated annual limit of 40 microgrammes per cubic metre in eleven out of fourteen years to 2014. Routing more diesel-powered buses through these streets will further exacerbate the problem, at a time when governments and the WHO are recognising the serious dangers to health that exposure to pollution poses.
The alternative route (which is the only route available for buses returning to Grange Rd) is via The Fen Causeway, Lensfield Road and Regent St. That is likely also to be heavily congested at peak times (with little scope for diverting traffic elsewhere, or even adding a bus lane).
Option 3/3a is incomplete without analysis of bus movements between the busway and the city centre.
Q10. What is the plan for bus journeys between Grange Road and destinations in and around the city?
Q11. When will this plan be integrated into the business case analysis for the project?
Evaluation of options
Feedback from the public consultation accounted for a maximum of just 5 points out of a total score of 125, i.e. 4% (pp 47-49 of Appendix 1: The Options Assessment Report). Had public opinion been given a 25% weighting, Option 3/3a would have come last. Given the huge amount of time and effort made by many people to contribute to the consultation, and by officers to collate and report on it, why was so little consideration given to the wishes of the public?
This project is intended to underpin the delivery of new housing, as set out in the draft Local Plan for South Cambridgeshire. So why does ‘Accessibility to more housing’ also count for just 4% of the total score? No option scores more than 3/5: isn’t this a concern, and doesn’t it conflict with claims that Option 3/3a performs ‘highly’ for delivering on the Local Plan strategy?
Other aspects of the methodology lack clear justification: e.g. why does Option 3/3a rank as ‘medium performing’ for ‘Capacity of the City Deal to deliver’ when it is ranked as having the highest ‘Constructability risk (complexity of delivery)’ and scores just 1/5 in the overall evaluation?
Q12. Why was so little consideration given to public feedback in the consultations?
Q13. What is the formal justification for the weightings given in the Multi Criteria Analysis Framework?
Q14. Will be Board please explain apparent inconsistencies in the evaluation of criteria such as risk, deliverability, and accessibility to more housing?
The most relevant objective in the original Project Brief was “to have congestion free [public transport] serving the corridor including the new developments, to avoid an increase in congestion and journey times.” To be ‘congestion free’, the solution could only ever be a segregated bus or rail route. This was overly prescriptive and appears to render the brief non-compliant with the Department for Transport’s TAG Transport Appraisal Process, from which the following extract (with our emphases) is pertinent:
2.5.9 It is crucial that the causes of the problems are investigated before solutions are generated. Focusing on problems (rather than underlying causes) as the stimulus for option development may result in solutions which ‘patch up’ the symptoms without addressing the real underlying causes.
2.8.2 It is important that as wide a range of options as possible should be considered, including all modes, infrastructure, regulation, pricing and other ways of influencing behaviour. Options should include measures that reduce or influence the need to travel, as well as those that involve capital spend. Revenue options are likely to be of particular relevance in bringing about behavioural change and meeting the Government’s climate change goal.
2.8.3 Studies should not start from an assertion about a preferred modal solution, or indeed that infrastructure provision is the only answer. Following the Eddington Transport Study, Sponsoring Organisations will be looking to encourage the better use of existing infrastructure and avoiding “solutions in search of problems”. In this context, it is recognised that small schemes can represent high value for money.
The brief could instead have set targets for modal shift and perhaps stated a maximum peak-time delay tolerable under normal circumstances. That would have opened the way for consideration of:
- Diverting traffic from the A1303 to the A428 by connecting the A428 directly to the M11.
- Building a Park & Ride site at the Girton Interchange, accessible from the A428, to replace the Madingley Rd P&R.
- Building a bus station in Cambourne and high quality bus stops along the A428 corridor, with convenient and safe access by foot and cycle.
- Using new, high quality stops to support express bus services (with many fewer than the 40 stops along the Citi 4 bus route).
- Providing more interchange opportunities in and around the city, in particular to Cambridge central train station and Addenbrooke’s.
- Increasing the peak-time frequency from 3 buses/hour (every 20 minutes) to 5 (every 12 minutes).
- Using Inbound Flow Control to reduce the impact on buses of congestion on Madingley Hill.
It should also be borne in mind that not just buses need to access Cambridge city centre: segregating buses ignores the many vehicles involved in delivering goods and services, including health and personal care to an ageing population. A strategy focused entirely on bus priority condemns those vehicles to ever-worsening congestion. A joined-up, long-term transport plan would seek to reduce overall congestion in and around the city to a tolerable and sustainable level, enabling all vehicles that need to access the city, including buses, to do so with minimal delay. Groningen has achieved this; so can Cambridge.
Arguments produced to date against Inbound Flow Control (IFC) are flawed in that they model the effect as if the IFC ‘gate’ created an extra flow restriction, when in fact it would automatically regulate flow to match the rate at which traffic is able to disperse freely. More efficient traffic flow will reduce, not increase, queuing times.
Q15. Why did the project brief limit consideration to only those options providing ‘congestion free’ public transport, when some congestion is both tolerable and compatible with much more cost-effective solutions?
Q16. Why did the project brief limit consideration to bus journeys when other vehicles have a need to access the city (and M11)?
Smarter Cambridge Transport urges the City Deal to commission a business-case analysis of the following by appropriately qualified professionals:
- Adding connections between the A428 and M11.
- Adding connections between the A428 and A1307/Huntingdon Road.
- Building a Park & Ride at the Girton Interchange, linked to the A1307/Huntingdon Road.
- The effect on the business case for the busway of having a competing Park & Ride site close to Madingley Mulch roundabout.
- The effect of peak-time congestion control points and more on-street parking controls in the city on traffic levels and bus journey times between Cambourne and Cambridge.
- Improving bus infrastructure along the route, with a new bus station in Cambourne.
- Optimising the existing Citi 4 bus service to reduce journey times (cashless ticketing, fewer stops, and better information provision).
- Creating a bus gate on Cambridge Road to enable some peak-time buses to be routed via Madingley village, bypassing queues on Madingley Hill.
- Using Smart Traffic Management and Inbound Flow Control on the A1303 instead of a bus lane/way from Madingley Mulch roundabout to Queen’s Road (both with and without the Girton Interchange upgrade).
- Addressing the lack of capacity of Burrell’s Walk, Garret Hostel Lane and Senate House Passage for more walking and cycling journeys.
For more detail on how these features fit together, please refer to our original submission to the consultation.
Q17. If the Board does not accept our recommendations, could it please explain why?
Appendix A: Population figures
Population figures have been sourced from Cambridgeshire Insight.
Population of settlements from Cambourne to Cambridge
* Assumed average occupancy of 2.5 people per new home
† New homes included set out in masterplans and planning applications:
- Cambourne: 950 homes built since 2011
- Cambourne West: 1,200 in local plan (though developer proposal is for 2,350 new homes)
- Bourn: 16 new homes being built
- Bourn Airfield: 1,700 new homes in local plan; developer proposal is for 3,500
- Caldecote: 140 new homes were refused consent, but new applications are likely
- Hardwick: 235 new homes planned
Population of settlements along existing Guided Busway
|Oakington & Westwick||1,550|
Appendix B: Commuter data
Commuter data has been sourced from 2011 Census, analysed according to MSOA (Middle Layer Super Output Area) by UCL at Datashine Commute. The MSOA boundaries are shown on the Government ONS website. Figures are for commuting to any of the thirteen MSOAs in Cambridge city (which includes the Biomedical Campus).
|South Cambridgeshire 020||Cambourne||1,096||135||12.3%|
|South Cambridgeshire 010||Highfields Caldecote, Hardwick, Toft, Comberton, Eversdens||1,597||163||10.2%|
These figures do not include Bourn or Coton, as these are part of MSOAs that would mostly not be served by a Cambourne–Cambridge busway.