Critique of M4 East tunnel EIS Human Risk Assessment

(Editor: These comments have been prepared to assist other citizens and groups who are responding to the EIS. They are published anonymously. The author has spent several months investigating these health related issues. We invite discussion, using the reply button below.)

This is a critique of Appendix J The Human Risk Assessment which was prepared for Westconnex by Environmental Risk Sciences Pty Ltd (enRiskS).

Executive Summary

  • This Human Health Risk Assessment is not accepted as a reliable study of the health risks of the WestConnex project

  • This assessment has several serious flaws, but the key one is that it is not independent. It relies completely on the findings of the Air Quality Impact Assessment (Volume 2B, Appendix H), which in turn relies on the WRTM traffic forecast model. It does not question, in any way, the findings of the Air Quality Impact Assessment, thus does not assess the health risks of the very real potential for induced demand on Parramatta Road.

  • A truly independent study would have noted the “hot spots” already existing in the corridor and suggested that NSW Health should be monitoring, via a longitudinal study, the health of residents living, schooling and working near those places a truly independent health risk assessment would have concluded that peoples’ health was already being compromised in this corridor, and that further “hot spots” are going to be created by the project and that the relevant government authorities should be taking action about it, including by urgently implementing the National Clean Air Agreement

  • Both NO2 and PM2.5 levels are already elevated and above guidelines in many of the localities within the project area, so any additional traffic due to induced demand (in tunnels or on surface roads) will add to this situation.

  • This risk assessment does not point out that the air quality standards or guidelines used in the EIS may be superseded by stricter ones at the end of 2015 when the new National Environment Protection (Ambient Air Quality) Measure (NEPM) standards are passed

  • Other cleaner forms of transport are required in this corridor.

  • Diesel fuel vehicles need to be phased out quickly, as is happening in France and proposed in other parts of Europe.

The Westconnex/AECOM Human Health Risk Assessment reaches the following overall conclusions

  • The tunnel doesn’t need filtration as there will be negligible pollution at the portals and the ventilation stack will disperse pollutants across the air-shed with negligible effect on local residents

  • A detailed assessment of two key pollutants (nitrogen dioxide – NO2 – and particulate matter measured in microns – PM10, PM2.5) is required, given their known associated health effects and the impacts of their co-exposure (common in urban environments) and the known main source of PM2.5 (and finer) in the urban environment being vehicle diesel exhaust

  • Potential health impacts associated with changes in air quality (specifically NO2 and PM) within the local community was assessed as low and essentially negligible

  • There are some residents who will have an increased exposure to PM as a result of the redistribution of emissions from surface roads

  • There are some residents who will be exposed to less PM2.5 as a result of the project

  • Summary of mortality data for three key factors (COPD, lung cancer, cardiovascular disease) indicates that Sydney Area Health Service has rates higher than the average NSW rate for all three indicators; summary of hospitalisation data for key factors (asthma, cardiovascular disease, COPD) shows asthma rates similar to NSW and cardiovascular disease and COPD less than NSW rates (possibly accounted for by high GP service level in Inner Sydney)

  • Increases in population (and its distribution) of the suburbs within the project footprint will have no impact on the health outcomes

  • The time spent exposed to pollutants within the tunnel is very short (minutes) and in the absence of published guidelines for NO2 and PM within tunnels, advises keeping vehicle windows up and air con on recirculation setting

  • Potential for noise and vibration impacts during construction could result in adverse health effects so management and mitigation plan required for health effects

  • After construction, some properties will have elevated road noise so mitigation measures are required

  • In such a complex project, there are inherent uncertainties in each of the methods used to estimate emissions and concentrations, and there are limits to how accurately any impacts in future years can be predicted

Summary of key air quality issues accepted by the Health Risk Assessment

  • By 2021, without the project, the maximum (residential and commercial) 1-hour concentration of NO2 estimated to be (in micrograms per cubic metre) 375ug/m3 and 360ug/m3 (respectively) – which is well above the guideline (of 246ug/m3); with the project completion, the levels estimated as 307ug/m3 and 286ug/m3 (respectively) – still well above the guideline.

  • By 2021, with the project, the cumulative (maximum annual average, rather than acute) NO2 concentrations estimated to be below the guideline of 62ug/m3

  • By 2031, with the project, cumulative NO2 concentrations estimated below the guideline of 62 ug/m3

  • By 2021, without the project, the maximum (residential and commercial) 24-hr average concentration of PM2.5 estimated to be 29.3 ug/m3 and 30.5ug/m3 (respectively) – which is significantly above the guideline (25ug/m3); with the project completion, the levels estimated as 28.2ug/m3 and 26.6ug/m3 (respectively) – still above the guideline

  • By 2031,with the project, PM2.5 levels estimated as above the guideline

  • By 2021, without the project, the maximum (residential and commercial) 24-hr average PM10 concentration estimated as 54ug/m3 and 55.4ug/m3 (respectively) – above the current guideline (50ug/m3) and well above the recommended (from the 2014/15 review) of 40-50ug/m3

  • By 2021, with the project, the maximum (residential and commercial) 24-hr average PM10 concentration estimated as 52 ug/m3 and 50ug/m3 (respectively) – above both the current guideline and recommended (review) guideline

  • By 2031, both without and with the project, PM10 will be above both guidelines

  • Community locations most at risk by 2031 of mortality, cardiovascular and respiratory illness from increases in PM2.5 concentrations are: Homebush Boys High School, Ella Community Child Care Centre.St Joan Of Arc Catholic School, Dobroyd Point Public School, Woodfield Aged Care Centre

Methodology of the Health Risk Assessment

The Risk Assessment relies on key aspects of the Air Quality Impact Assessment including:

  • The tunnel design claim that there will be negligible impact at sites of portals and ventilation stacks
  • The findings of the air quality report within this EIS, which relies heavily on the air quality data from the 3 OEH monitoring stations at Rozelle, Chullora and Earlwood, one of which, Rozelle, does not monitor PM2.5
  • Traffic estimates within this EIS (which in turn rely on the population estimates).

For the health indicators, it relies on:

  • data that is five (5) years old for the key health indicators
  • hospitalisation data only for respiratory indicators,and has not included any GP data relating to consultations.

It does not include the demolition of buildings (houses and commercial) in the risks associated with the project, even though a high proportion of pre-1980’s buildings could have some asbestos on the sites.

Flaws in the Health Risk Assessment Methodology

  • It accepts the claim that the tunnel design is “state of the art”, and doesn’t canvass options to filter using actual “state of the art” like the Hong Kong Wanchai By-Pass tunnel and Madrid (Spain) M30 Ring-Road.

  • It accepts a tunnel design that infers there will be no pollution at tunnel entry or exit ramps, with the speed into entrance ramps relying on the “piston effect” to pull vehicle emissions into the tunnel and into the ventilation stack. A close questioning of the “experts” (at a number of the WDA consultations) about tunnel exit ramps where they hit Parramatta Rd and West City Link, found they agreed that there would be pollution in the vicinity of ramps if traffic hits congestion on exiting, and the same would be true for queuing traffic to enter tunnels. Thus their definition of “portals” is strictly that entry/exit area within the tunnel, not leading into or out-of the tunnels. An independent risk assessment would have noted the possibility of build-up of pollution around these portal sites and the consequent health impacts.

  • More real-time local air quality data should have also been used, particularly from near the sensitive receivers, given that the nearest two OEH monitors are some distance from these sensitive receivers. The OEH monitor at Chullora is closer to the Homebush/Nth Strathfield receivers, whilst the OEH Rozelle monitor is closer to the Haberfield/Ashfield receivers. However, both of the monitors are more than five (5) kilometres from these receivers and are not measuring real-time air quality in a 6kmx8km square area that has five of the heaviest-trafficked roads in Sydney. A very high proportion of residents in this 48 square km area live within 300metres of one of these main roads, so are exposed to much higher levels of NO2 and PM than are claimed in this report and based on the data from OEH Chullora and Rozelle monitoring stations. In fact, the OEH Rozelle station (located in the leafy grounds of the old psychiatric hospital near the river) does not collect data on PM2.5. ** An AMA submission into air quality states that the current air quality monitoring system in Australia does not adequately capture data relating to the exposure of vulnerable groups; and information on the levels that specific communities are exposed to, and the subsequent risk, is often unknown; the AMA believes the current monitoring system is only capturing data that represents exposure of the broader population, not people at “hot spots” – these “hot spots” would include the large number of residential and commercial properties within 300 metres of major roads (Australian Medical Association (March 2013), Submission to the Senate Standing Committee on Community Affairs, Inquiry into the impacts on health of air quality in Australia, p7). The Human Health Risk Assessment itself indicates there are 10,000 residences and commercial units (which includes multiple apartment blocks) within the affected zone. A conservative estimation of this population would be 40,000 people who may already be living in “hot spots”.

  • the data collected at the four (4) roadside air quality monitoring stations can be supplemented by similar data from the Lane Cove study (Cowie et al, 2012) collected at Parramatta Road Camperdown in 2006-2008, which indicated highly elevated levels of NO2 and PM2.5. The traffic count (2005) at the Camperdown (control) site in that study was 65,000 vehicles/day, which is a little fewer than the City-West Link count of 69,000 vehicles/day and higher than Parramatta Road (at Dalhousie St Haberfield) of 54,000 vehicles/day (NSW RMS Average Daily Traffic 2012). All this data shows that levels are already higher than the guidelines, which themselves have been questioned by the AMA and several other groups in their submissions to the Senate Standing Committee On Community Affairs Inquiry Into The impacts On Health Of Air Quality In Australia.

  • It should have noted that the new national air quality standards to be endorsed at the end of 2015, well before the project starts, may include higher standards for PM10 than are used in this EIS. The outcome could be many more “hot spots” than acknowledged in the Air Quality Impact Assessment.

  • The study should have noted that several international studies have shown the adverse health effects of living less than 500 metres from major roads. In one longitudinal study of 2,300 children in Los Angeles, Guaderman et al found that improvements in the air quality (particularly PM2.5 and NO2) over several years had dramatically improved the lung function of children living less than 500 metres from major roads (Gauderman WJ, Urman R, Avol E, Berhane K, McConnell R, Rappaport E, Chang R, Lurmann F, Gilliland F.2015. Association of improved air quality with lung development in children. N Engl J Med. 372(10):905-913). A paper that calculates the medical costs of air pollution indicates that even the NEPM standards give a false sense of “safe” levels of pollution. This report states that as average levels of air pollution increase so do the average adverse health effects and that in fact there is no safe threshold. The paper claims that proper cost benefit analyses should be undertaken to accurately quantify adverse health effects due to both local and general increases in air pollution (Barnett, A, 2014 “It’s safe to say there is no safe level of air pollution” Australian and New Zealand Journal of Public Health 2014 vol. 38 no. 5). The health risk assessment does not acknowledge any critique of what is considered a safe threshold because it completely accepts the assumptions of the Air Quality Impact Assessment. Available evidence doesn’t indicate there is an exposure threshold for PM below which health effects do not occur; thus this risk assessment should have erred on the side of caution.

  • The study should have noted the large number of people living, working and schooling less than 500 metres from Parramatta Road and thus at an already elevated risk of health damage; it also should have noted the increased risk if this corridor does not have the reduced traffic volume estimated by the WRTM, given the additional sources from the tunnel traffic.

  • The study should have noted that the times when children and parents are walking to/from school in this polluted corridor coincide with the peak readings of pollution from commuter traffic; the same is true of park usage (the main large parks near portals – Cintra Park, Ashfield Park, Reg Coady Reserve and Robson Park all have multiple forms of use day and early evening) – there is no acknowledgement in the study’s methodology of this type of real-time pollution health risk

  • the traffic estimates are problematic for the surface use of Parramatta Road with project completion, as they rely on population forecast data calculated prior to the Parramatta Road Renewal Plan release. The population of the six (6) LGAs within the footprint of the project is forecast to grow by 40% between 2011 and 2031 (higher than the Sydney forecast of 33%); this is a volume of 132,844 people (Vol 2A, sect 5.2.1,p5-6). The Parramatta Road Renewal Plan allows for an additional 40,000 units in eight (8) “precincts” to be build between Auburn and Camperdown. These additional units could add an average of 100,000 people to the forecast figure in 2031 of 132,844 to give a truer forecast total of 232,844. A very high proportion of these new unit dwellers will own/purchase vehicles, thus adding to the pressures on the Parramatta Road corridor.

  • A scenario including all risk factors for 2031 should have been discussed – that of increased traffic on the surface of Parramatta Road the full length of the project. The concept of ‘induced traffic demand’ is a real one and has not been taken into account by this health risk assessment. A meta-analysis of induced demand (Cervero, R, 2001, Induced Demand: An Urban and Metropolitan Perspective, Paper prepared for Policy Forum, US Environmental Protection Agency) concluded that, whilst there were a range of elasticities associated with increased demand with new road infrastructure, the phenomenon should not be trivialised

  • The study ignores the possible contribution of the project to the estimated health costs of the Sydney region. In 2005 in Sydney motor vehicle pollution alone accounted for $1.5b in health costs (Australian Bureau of Transport and Regional Economics, 2005, cited in Senate Standing Committee On Community Affairs Inquiry Into The impacts On Health Of Air Quality In Australia, NSW Environment Protection Authority Submission, September 2013, p15). As this report states … “particle pollution is the driver for the high public health costs of air pollution” (ibid).

  • The mortality data for three key factors (COPD, lung cancer, cardiovascular disease) shows that Sydney Area Health Service has rates higher than the average NSW rate for all three indicators, yet this risk assessment does not question whether exposure to already elevated levels of NO2 and PM2.5 may be a cause of these elevated rates.

  • The reliance on five years old data for the key health indicators is problematic, given background sources of key pollutants (NO2 and PM2.5 and PM10) have increased in that five year period, as evidenced by vehicle fleet data. According to the Sydney Morning Herald (March 15-16 2014) the number of diesel vehicles on the road in Australia has more than doubled since 2005. ABS data shows that in 2015 there are 3.6 million diesel powered vehicles, accounting for 19.7% of the fleet; and over the five year period from 2010, the number of Passenger vehicles and Light Commercial vehicles registered with diesel fuel increased by 96.4% and 62.9% respectively (ABS, Motor Vehicle Census, Australia, cat no 9309.0). This increase in diesel fuel use would have had both short term and longer term impacts on health in the period since this health data was published and has not been captured by it, nor has this health risk assessment mentioned increased risks from the growing diesel fleet. The other concern with growth of the diesel fleet relates to car manufacturers using software called “defeat devices” that allows the masking of exhaust emissions in pollution control tests. Recent revelations (eg SMH, Sep 23 2015) of such corruption by what was thought of as a “good corporate citizen” – the Volkswagen company – indicates the distinct possibility that other manufacturers may be doing the same. If the environmental protection laws are being flouted by many diesel fuel vehicle makers, then the claim by this EIS that tougher laws are making the Australian vehicle fleet (and therefore the air) cleaner is questionable. The best option for any government keen to improve air quality in polluted parts of Sydney would be to quickly phase out diesel vehicles.

  • Hospitalisation data for respiratory indicators should have been supplemented by GP survey data (given Medicare does not collect this level of data), collected by the relevant Local Health Districts.

  • Sydney Area Health Service collects data on health risk factors, published for each Local Health District. The risk factor model is one that can be called a “personal behaviour” model. It includes risky alcohol drinking, smoking, consumption of fruit & vegetables, being overweight or obese, and adequate physical activity). It does not include external risk factors, like living within a few hundred metres of a busy road. The fact that the Area Health Services in NSW do not collect data for external risk factors, although they are known (e.g. exposure to coal dust (mainly PM10) and road pollution (notably NO2 and PM2.5) means that more meaningful data is not available to these types of Human Health Risk Assessments.

  • The health impacts from noise and vibration seem to be underestimated by this health risk assessment. The writer should have conducted an independent analysis of that assessment, rather than relying on its findings. A independent review of the noise and vibration section of the EIS ( which will be published by the People’s EIS soon) raises uncertainty in a range of areas of the noise and vibration assessment. It states that, given the number of potentially impacted properties and people within the project area, further work needs to be done to ensure the local community and other affected stakeholders are provided with a fully informed assessment. This needs to occur as part of the EIS consultation process where further comment can be sought from the community, and not simply resolved through the Submissions Report which does provide for any further input from the community.

  • The health risk assessment should have included a risk assessment for asbestos and a management plan for mitigating risks, given serious community concern has been expressed over issues with asbestos removal and treatment at Granville (M4 widening stage) and at Alexandria (M5E stage).

  • It is common practice in health research (including health risk assessments) to identify and analyse any related health studies to understand their the implications for the project being assessed-in this case the WestConnex project.Two such studies have been conducted on the Lane Cove Tunnel and on the M5East tunnel. The risk assessment ignored the ‘fine print’ findings of both of these studies. The Lane Cove study was of the short-term respiratory (asthma-like symptoms) health effects of exposure to emissions from the tunnel stack. The study was conducted 2006-2008, with thirty six (36) participants originally, falling to twenty (20) by 2008. This is a very small cohort from which to draw any strong conclusions. The “control site” was Parramatta Road, Camperdown. The main conclusion was that residents near the eastern tunnel stack reported increased symptoms (even though there were no increases in pollutants) and this could be accounted for by other sources of pollution not measured in the study (e.g. particles smaller than PM2.5 and/or volatile compounds) or other unknown sources. The study also found significant adverse effects of increased airway inflammation, and chest and eye symptoms associated with exposure along Parramatta Road Camperdown after only 2-hour exposure periods. The authors recommended a follow-up study to better understand the effects of the ventilation stack. (Christine T Cowie et al, 2012, ‘A randomised cross-over cohort study of exposure to emissions from a road tunnel ventilation stack’, BMJ Open 2012;2). No further studies have been conducted.

  • In the study of the pollution effects of the M5 East tunnel (NSW Health, 2012), it was stated that the ventilation stack was an important source of air pollution in the area within a 2 km radius, contributing 23% of NOx and 17% of PM10. The study’s aim was to consider cancer incidence in the vicinity of the tunnel stack. Whilst it found there was a significantly higher incidence of lung cancer in postcodes immediately around the stack, the study concluded that it was unlikely there was a causal link. However, the study did qualify its overall conclusions with the comment that the descriptive epidemiological method used in the study was a relatively weak tool to resolve the type of issues under investigation. The study methodology was unable to discern the types of lung cancers screened in the area (smokers vs non-smokers). The study’s argument was that, given lung cancers can take a number of years to become apparent, the higher incidence recorded after the opening of the tunnel may not have been connected to the stack emissions. The study concluded that the higher incidence of lung cancer probably pre-dated the locating of the stack in that area. This is an alarming finding, given that the cancer registry data would have been available to NSW Health, Department of Main Roads and the other bodies responsible for the EIS prior to the building of the M5East. So it would appear that an area of Sydney which already had an elevated incidence of lung cancer was chosen as the site of the M5East ventilation stack, when health research was already indicating the causal link between particulate matter and lung cancer. Given that mortality data for three key factors (COPD, lung cancer, cardiovascular disease) indicates that Sydney Area Health Service has rates higher than the average NSW rate for all three, is it possible that this project is again ignoring the potential health impacts of a project that independent experts predict will increase total traffic pollution over time.

Flaws in the assumptions on which EIS health findings are based

  • The health risk assessment does not take into account the possibility of total higher traffic along Parramatta Road than the WRTM forecasts. An Australian Government Department of Infrastructure and Transport Research (2012) report into traffic growth in Australia shows that traffic per person in Australia has grown steadily between 1965 and 2011. The measure of traffic volume (number of vehicles x distance travelled = vkt) in all states and major cities for this period indicates that the pattern of increases has been consistent, with only minor changes in response to petrol price rises, unemployment and the global financial crisis. Forecasts of future growth in traffic volumes indicate that traffic in Australia will rise from 55 billion vkt per quarter in 2011 to more than 65 billion vkt per quarter in 2020.

  • There is no reason to believe that traffic volumes on the surface of Parramatta Road will decrease after 2021, especially with the anticipated population growth in Sydney, including in Inner Western Sydney. The traffic studies on which the air quality data assessments are based does not take into account Urban Growth plans to build more than 40,000 apartments in Parramatta Rd. Given that this EIS shows that levels of PM2.5 and NO2 are already elevated in parts of this corridor, it is obvious the project is designed to take current volumes of traffic off this road so that high-density residential development can be enabled even in the face of currently known health risk factors. This is a huge gamble with people’s health.

(Ed:This post was slightly updated on October 31, 2015)

One thought on “Critique of M4 East tunnel EIS Human Risk Assessment

  1. I raised at the Kingsgrove (no) information night the tunnel filtration used on the Hong Kong tunnel. I was advised by the Tunnel Filtration rep that the reason for filtration in Hong Kong was due to the stack located alongside apartment buildings 25 stories high. Looking at some of the Urban Growth designs, there is apparently planned buildings of similar size along Parramatta Road.

    When looking at the Kingsgrove unfiltered stack (in a valley), its proximity to Moorefields Rd Primary School and the many homes (upwind on a hill) the effect of this unfiltered stack will basically give these people a lungful of unfiltered toxins 24/7.

    A complete disregard to the health of the community and the potential medical bills that the state will have to pay on the long term.

    Liked by 1 person

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s