(Ed: This submission is by transport planner Chris Standen, who is currently completing a doctorate in transport economics at the University of Sydney. For more on Chris, see below. Readers will note that Chris has many questions for the M4 East Traffic Modellers. The People’s EIS are keen to hear the answers. We note that Chris Standen’s submission to EIS may have been slightly updated before it was finally submitted and we will publish the final version later)
1) I strongly object to the M4 East project, and to the broader WestConnex scheme.
2) The EIS has failed to model the impacts of implementing the proposed project (M4 East) relative to not implementing the proposed project (the ‘future do minimum’ scenario). The ‘future do something’ scenarios, on which the traffic, air quality, health and greenhouse modelling is based, include the M4 East project plus another uncommitted project to convert kerbside general traffic lanes on Parramatta Road to bus priority. With these additional bus lanes, the capacity of Parramatta Road would be significantly reduced and traffic volumes would fall accordingly, with drivers opting to use the M4 East tunnel instead. As such, the traffic volumes for the M4 East tunnel have been dramatically overestimated, and the traffic volumes for Parramatta Road have been dramatically underestimated in the ‘future do something’ scenarios.
The impacts of the project as proposed by the proponent (and as defined in Section 5 of the EIS), that is, the M4 East Tunnel with no new priority bus lanes on Parramatta Road, has not been presented in the EIS, as required by the SEARs.
(3) The stated objectives for the project were contrived to fit the project after it had already been announced. In a democratic strategic planning process, objectives are set first based on the needs and desires of the community, and then alternative projects/policies are appraised against their ability to meet those objectives.
(4) The EIS has not modelled alternative policy scenarios that could meet the transport/accessibility needs of NSW’s growing population, e.g.
a) Greater investment in public transport;
b) Road pricing reform;
c) Land use planning that places more homes closer to employment and services.
(5) It is no secret that the real purpose of the WestConnex scheme is to increase the road freight accessibility of Port Botany and Sydney Airport, and that private passenger vehicles have been included as a means of paying for it (through tolls). However, there are various policy alternatives for dealing with the growing freight task that do not appear to have been considered, e.g.:
a) Increase the capacity and reliability of rail freight
b) Increase rail freight subsidies to match/surpass those of road freight.
c) Divert container operations to other ports outside the city centre. Very few cities concentrate container operations in the city centre where road access is costly and has significant impacts on highly populated areas.
(6) The M4 East will be used by less than 1% of the NSW population each day. The costs will be borne by the whole population. It can hardly be argued that it is providing for the “greater good”.
(7) The Traffic and Transport Assessment does not stand up to scrutiny. There is not enough information about the methodology, input data or assumptions for the forecasts to be independently verified.
(8) There is no sensitivity analysis in the Traffic and Transport Assessment. The effects of varying key assumptions (e.g., willingness to pay the M4 East toll) have not been disclosed.
(9) The Traffic and Transport Assessment has not modelled the travel time and accessibility impacts for non-motorised modes (walk and bicycle).
(10) The issue of induced demand has not been fully addressed in the Traffic and Transport Assessment.
(11) Given the seriously flawed Traffic and Transport Assessment, there can be no confidence in the accuracy of the other impact analyses in the EIS that are dependent on the traffic forecasts, in particular:
a) Air quality,
b) Noise and vibration,
c) Human health,
d) Greenhouse gas emissions.
(12) The role of motorways in a multimodal urban transport network is to allow traffic to circulate around the edge of a city connecting low density suburbs, where the traffic does not directly impact highly populated areas. For radial transport into and out of employment/activity centres, mass transit (e.g., rail) is more quick/efficient, requires less space, and has fewer impacts on highly populated inner-urban areas.
(13) The EIS does not consider the cumulative costs of adding more urban motorways to those previously built through the heart of Sydney since the 1950s. Although the economic, social and environmental costs of each individual motorway (as reported in an EIS) may be considered by some stakeholders to be acceptable, the cumulative costs are considerable:
(a) Following decades of road expansion and consequential sprawl, Sydney now spends about 13% of its GDP on transport, while the average European or Asian city spends only between 5% and 8%.1 (1)
(b) Serious human health impacts due to petrochemical vehicle emissions/smog, including:
i) Lung cancer,
iii) Heart disease,
iv) Impaired lung development in children living near motorways/exhaust stacks.
(c) Waterways contaminated with road runoff (heavy metals and carcinogens in brake and clutch dust, exhaust particulates etc.).
(d) High traffic crash costs (of deaths/traumatic injuries and material damage).
(e) Urban sprawl and increasing commuting distances.
(f) Social isolation for non-drivers living in car-dependent suburbs.
(g) Noise pollution from traffic and its impacts on sleep.
(h) Impacts on visual amenity (pollution stacks, concrete interchanges, concrete flyovers).
(i) Extreme summer temperatures (urban heat island effect).
(j) Community destruction and severance.
(k) Destruction of heritage.
(l) Less incidental physical activity from walking and cycling (including to/from public transport), resulting in higher rates of obesity, diabetes, cancer and heart disease.
(m) Increased chauffeuring burdens for parents and carers.n)
(n) Less independence for children.
(o) High per-capita greenhouse gas emissions.
2 Issues with the Traffic and Transport Assessment ( Appendix G)
2.1 General Comments
(14) The Traffic and Transport Assessment does not stand up to scrutiny. There is not enough information about the methodology, input data or assumptions for the forecasts to be independently verified.
(15) There is no sensitivity analysis in the Traffic and Transport Assessment. The effects of varying key assumptions (e.g., willingness to pay the M4 East toll) have not been disclosed.
**(16) Travel time and accessibility impacts for non-motorised modes (walk and bicycle) have not been modelled.
(17) Impacts of disruptive technology on future driving demand have not been not considered (e.g.automated vehicles).
(18) Inter-generational changes in vehicle ownership, driver licensing and transport preferences have not been considered.
(19) Changes in aggregate transport measures have not been provided for the various scenarios.For example:
(a) Overall increase in VKT (Vehicle Kilometres Traveeled.)
(b) Change in average trip distance.
2.2 Comments on Specific Sections
(20 One of the stated purposes of the report is to “complete a holistic traffic and transport assessment including crash analysis, travel speeds and travel time analysis and opportunities to enhance public and active transport networks within the project area”. However, the report does not provide any travel time forecasts for active transport.
(21) Another purpose is to “Recommend a suite of measures to mitigate and manage traffic and transport impacts of the project for construction and operational scenarios”. The general consensus among transport experts is that the most effective way to manage traffic demand is through demand management, e.g., road pricing reform. However, the report does not recommend any demand measurement measures.
2.2.2 Section 3 – Strategic Context
(22) The stated justification for the project is based on the discredited ‘predict and provide’ approach to transport planning, whereby it is assumed that transport demand will continue to grow, and that capacity must be increased to accommodate it. In practice, transport demand in cities is limited by capacity: as capacity increases, so does demand (induced demand). It is geometrically impossible to provide enough roadway capacity to accommodate all the latent demand for driving (i.e., where everyone can live and work where they want, and make all the driving trips they want, when they want, to wherever they want in free-flow traffic) in a city of Sydney’s population.
(23) Furthermore, the most efficient way to accommodate the transport and accessibility needs of a growing population is through mass transit and better land use-transport integration. Urban motorways are a very inefficient way of moving people around. A single traffic lane can transport a maximum of only 2000 people per hour (in ideal conditions); a single railway line can transport 20,000 people per hour.
(24) The statement “It is acknowledged that any investment in motorway infrastructure has to be aligned with supporting public and active transport initiatives to achieve an increase in capacity, while aiming to reduce the reliance and demand of private vehicles on the future road network” is contradictory: increasing motorway capacity will only serve to increase private vehicle demand.
2.2.3 Section 4:Assessment Methodology
(25) There is not enough information about the modelling methodology for it to be replicated and the outputs independently verified.
(26) The transport model (WRTM) has not been made available for independent verification.
(27) The model input data and assumptions have not been made available for independent verification.
What toll prices have been assumed?
(28) The model coverage area is too small to capture all the transport impacts of the project. The project will affect transport demand and behaviour across the whole metropolitan area.
(29) More detail on the Value of Travel Time Saving (VTTS)/Willingness to Pay (WTP) model is needed.
(a) The form and parameters of the model have not been given.
(b) If it was based on stated preference surveys, then how has the issue of hypothetical bias been addressed?
(c) Has the model been validated? Previous toll choice models in Australia have overestimated WTP for toll roads.
(d) Does it include the negative utility of the tunnel environment (monotony, no natural light, poor air quality)?
(30) The weekend period has not been modelled, despite current weekend traffic volumes being higher than weekday traffic volumes on many corridors.
(31) Insufficient information about the travel zone structure in the WRTM:
(a) What are the travel zones based on? How big are they?
(b) How are intra-zonal trips modelled?
(c) How are trips to/from external zones modelled?
(32) Induced demand has not been fully addressed:
a) The model ignores the impact of the project on the long-term transport decisions of individuals and firms, including:
(1) Residential location choice – the project will encourage more people to move further from work (sprawl), thereby increasing average travel distances/demand.
(2) Work location choice – the project will encourage more people to work further from home, thereby increasing average travel distances/demand.
(3) Car ownership choice – the project will encourage more car ownership.
(4) Firm location choice – the project will encourage firms to locate in locations further away from their labour supply/customers/suppliers than they otherwise would, thereby increasing travel distances/demand.
(b) To my knowledge, there has been no long-term evaluation/verification of the methodology used to forecast induced demand (New Zealand Transport Agency Economic Evaluation Manual (EEM)). Induced demand by its nature materialises over several years, as people gradually move home/work location etc. Without a long-term evaluation/verification of the methodology, there can be no confidence in the induced demand forecast produced.
(33) Insufficient detail on origin-destination demand matrix generation:
(a) What are the form and parameters of the generalised cost function?
(b) How were shortest paths calculated?
(34) Insufficient detail on trip generation:
(a) What are the form and parameters of the trip production function, and how was it estimated?
(b) What are the form and parameters of the trip attraction function, and how was it estimated?
(c) Were trips were balanced towards attractions, or towards productions?
(35) Insufficient detail on trip distribution/modal split:
(a) What are the form and parameters of the gravity model used?
(b) What are the form and parameters of the deterrence function used?
(c) How has modal split been estimated?
36 Insufficient detail on road traffic assignment:
(a) Is assignment stochastic or deterministic?
(b) What link loading/flow function was used?
(c) Were intersection delays included?
(37) Insufficient detail on public transport assignment:
(a) How were access and egress points determined?
(b) How were route strategies determined?
(38) Non-motorised trips have not been included in the WRTM or LinSig modelling.
(39) Impacts on accessibility have not been modelled. Most transport is not an end in itself – it is a means to access work, education, services etc. How does the project affect population accessibility?
(40) Equity and equality impacts not described.
(a) How many people are better off with the project?
(b) How many people are worse off with the project?
(c) Do benefits/impacts accrue to any population groups more than others, e.g., people with a disability or on low incomes?
(41) Downs-Thomson Paradox not considered.
The project will attract passengers away from public transport to road. As such, public transport patronage will be lower than it would be without the project. This could result in public transport service levels being cut, which will encourage further mode shift from public transport to road.
(42) No sensitivity analysis.
Given the numerous assumptions and approximations in the model, there needs to be some sensitivity analysis, e.g.: How will traffic volumes be affected if (when) the WTP for the toll turns out to be higher than the point estimate used?
2.2.4 Section 7 Assessment of construction impacts
(43) Impacts on walking/bicycle demand and travel times have not been modelled.
(44) Does the LinSig intersection modelling take into account the impact that changes in intersection LOS will have on travel demand?
2.2.5 Section 8 Future year traffic volumes and patterns
(45) Impacts on walking/bicycle demand and travel times have not been modelled.
(46) Does the LinSig intersection modelling take into account the impact that changes in intersection LOS will have on travel demand?
2.2.6 Section 8 Future conditions without the project
(47) Impacts on walking and bicycle demand and travel times have not been modelled.
(48) Does the LinSig intersection modelling take into account the impact that changes in intersection LOS will have on travel demand?
2.2.7 Section 9 Assessment of operational impacts
**(49) Impacts on walking and bicycle demand and travel times have not been provided.
- Newman P, Kenworthy J. Costs of automobile dependence: global survey of cities. Transp. Res. Rec. J. Transp. Res. Board 1999;1670(1):17-26. doi:10.3141/1670-04.
(From The Conversation website: Chris is currently researching for a PhD in transport economics at the University of Sydney. He holds a BSc in Physics from Brunel University, where his honours research was on subatomic particle measurement at the Stanford Linear Accelerator, and a Master of Environmental Management from UNSW. He has previously worked as a transport planner in local government, as a policy advisor in the NSW Parliament, and as a systems architect designing telecommunications networks for operators in Europe and Asia.)