WATTLEUP ROAD DEVELOPMENT COMPANY PTY LTD and WESTERN AUSTRALIAN PLANNING COMMISSION
[2014] WASAT 159
•21 NOVEMBER 2014
WATTLEUP ROAD DEVELOPMENT COMPANY PTY LTD and WESTERN AUSTRALIAN PLANNING COMMISSION [2014] WASAT 159
| STATE ADMINISTRATIVE TRIBUNAL | Citation No: | [2014] WASAT 159 | |
| PLANNING AND DEVELOPMENT ACT 2005 (WA) | |||
| Case No: | DR:362/2013 | 7 8 MAY 2014, 1 5 AND 16 SEPTEMBER 2014 | |
| Coram: | JUDGE D R PARRY (DEPUTY PRESIDENT) MR P CURRY (SENIOR SESSIONAL MEMBER) MR B HUNT (SENIOR SESSIONAL MEMBER) | 21/11/14 | |
| 52 | Judgment Part: | 1 of 1 | |
| Result: | Tribunal recommendation to Hon Minister for Planning to refuse subdivision approval for both proposed subdivisions | ||
| B | |||
| PDF Version |
| Parties: | WATTLEUP ROAD DEVELOPMENT COMPANY PTY LTD WESTERN AUSTRALIAN PLANNING COMMISSION ALCOA OF AUSTRALIA LIMITED PRIMEWEST (WATTLEUP) PTY LTD |
Catchwords: | Town planning Residential subdivision 'Call in' by Minister for Planning Tribunal report and recommendations to Minister Site 1.2 kilometres to 1.7 kilometres from land used for drying/disposing of bauxite residue resulting from alumina production Site proximate to sand quarry Air quality Dust Health and amenity impacts of dust Whether proposed subdivisions are acceptable in relation to health and amenity impacts of dust Air quality monitoring Adequacy of air quality monitoring in relation to amenity Averaging of air quality monitoring results for amenity over 24 hours Potential shorter term amenity impacts Whether air quality monitoring and other evidence demonstrates that proposed subdivisions are acceptable in relation to health and amenity impacts of dust in future years Whether conditions during monitoring year representative of future Climatic conditions relevant to dust generation in monitoring year in comparison to other recent years Climate change Increase in height of residue disposal area where most proximate to site Consistency with planning framework Affect on industrial amenity of residue disposal area Planning precedent |
Legislation: | City of Cockburn Town Planning Scheme No 3 Metropolitan Region Scheme Planning and Development Act 2005 (WA), s 246(1), s 246(2), s 246(2)(b), s 251(1) State Administrative Tribunal Act 2004 (WA), s 27(2), s 37(3) |
Case References: | Macri v Western Australian Planning Commission [2014] WASC 153 Nicholls and Western Australian Planning Commission [2005] WASAT 40; (2005) 149 LGERA 117 Wattleup Road Development Company Pty Ltd and Western Australian Planning Commission [2011] WASAT 160 Wattleup Road Development Company Pty Ltd and Western Australian Planning Commission [2014] WASAT 29 |
Summary | Two applicants sought review by the Tribunal of the deemed refusals of residential subdivision applications of land at Wattleup Road, Hammond Park. The Hon Minister for Planning formed the opinion that the applications for review raise issues of such State or regional importance that it would be appropriate for them to be determined by the Minister and therefore 'called in' the applications. The Minister directed the Tribunal to hear the applications but, without determining them, refer them with recommendations to the Minister for determination.,The site is located approximately 1.2 kilometres to 1.7 kilometres northeast of an area used for drying/disposing of bauxite residue resulting from alumina production and close to a sand quarry. A previous subdivision application of part of the site was refused by the Tribunal in an earlier review proceeding having regard to potential dust impacts from the bauxite residue disposal area and the sand quarry. The applicants subsequently commissioned air quality monitoring on the site in order to demonstrate that the proposed subdivisions are acceptable in relation to the health and amenity impacts of dust.,The Tribunal heard concurrent expert evidence from a panel of eight expert witnesses in relation to air quality. Having heard this evidence, the Tribunal was not satisfied that the proposed subdivisions are acceptable in relation to the amenity impacts of dust, because the air quality monitoring for amenity purposes was only undertaken on days adding up to about onefifth of a year and the monitoring and evidence presented by the applicants did not address shorter term than 24 hour adverse amenity impacts from dust, even though such impacts on residents of the proposed subdivisions are likely to be real and material. ,The Tribunal also considered that the applicants' air quality monitoring cannot reasonably be relied upon to demonstrate that the subdivisions are acceptable in relation to the health and amenity impacts of dust in the future, because it is unlikely that relevant conditions during the monitoring year are representative of the future. In particular, relevant climatic conditions during the monitoring year were average and not worst case in comparison to recent years; climate change evidence suggests that relevant climatic conditions in the future will be different; and there will be a significant increase in the height of the bauxite residue disposal area where it is closest to the site (and a consequent altitudinal increase in wind speed) between 2014 and 2030.,Furthermore, for these reasons, the Tribunal considered that it is likely that relevant conditions affecting dust generation and impacts on the site will be different in the future and consequently that dust impacts on the site are likely to be greater from time to time, at least until operations in the most proximate part of the bauxite residue disposal area to the site and at the sand quarry cease and those areas are rehabilitated. The evidence indicates that this will not occur for at least 15 to 20 years.,The Tribunal therefore reported to the Minister that it is not satisfied that the proposed subdivisions are acceptable in relation to the health and amenity impacts of dust. The Tribunal also, consequently, reported to the Minister that it is of the view that proposed subdivisions are inconsistent with the applicable planning framework, would likely adversely affect the industrial amenity of the bauxite residue disposal area, and would set an adverse planning precedent.,The Tribunal therefore recommended to the Minister that the correct and preferable decision is that the applications for review should be dismissed and subdivision approval should be refused for the proposed subdivisions. |
JURISDICTION : STATE ADMINISTRATIVE TRIBUNAL ACT : PLANNING AND DEVELOPMENT ACT 2005 (WA) CITATION : WATTLEUP ROAD DEVELOPMENT COMPANY PTY LTD and WESTERN AUSTRALIAN PLANNING COMMISSION [2014] WASAT 159 MEMBER : JUDGE D R PARRY (DEPUTY PRESIDENT)
- MR P CURRY (SENIOR SESSIONAL MEMBER)
MR B HUNT (SENIOR SESSIONAL MEMBER)
- Applicant
AND
WESTERN AUSTRALIAN PLANNING COMMISSION
Respondent
ALCOA OF AUSTRALIA LIMITED
Intervener
- Applicant
AND
WESTERN AUSTRALIAN PLANNING COMMISSION
Respondent
ALCOA OF AUSTRALIA LIMITED
Intervener
Catchwords:
Town planning Residential subdivision 'Call in' by Minister for Planning Tribunal report and recommendations to Minister Site 1.2 kilometres to 1.7 kilometres from land used for drying/disposing of bauxite residue resulting from alumina production Site proximate to sand quarry Air quality Dust Health and amenity impacts of dust Whether proposed subdivisions are acceptable in relation to health and amenity impacts of dust Air quality monitoring Adequacy of air quality monitoring in relation to amenity Averaging of air quality monitoring results for amenity over 24 hours Potential shorter term amenity impacts Whether air quality monitoring and other evidence demonstrates that proposed subdivisions are acceptable in relation to health and amenity impacts of dust in future years Whether conditions during monitoring year representative of future Climatic conditions relevant to dust generation in monitoring year in comparison to other recent years Climate change Increase in height of residue disposal area where most proximate to site Consistency with planning framework Affect on industrial amenity of residue disposal area Planning precedent
Legislation:
City of Cockburn Town Planning Scheme No 3
Metropolitan Region Scheme
Planning and Development Act 2005 (WA), s 246(1), s 246(2), s 246(2)(b), s 251(1)
State Administrative Tribunal Act 2004 (WA), s 27(2), s 37(3)
Result:
Tribunal recommendation to Hon Minister for Planning to refuse subdivision approval for both proposed subdivisions
Summary of Tribunal's report:
Two applicants sought review by the Tribunal of the deemed refusals of residential subdivision applications of land at Wattleup Road, Hammond Park. The Hon Minister for Planning formed the opinion that the applications for review raise issues of such State or regional importance that it would be appropriate for them to be determined by the Minister and therefore 'called in' the applications. The Minister directed the Tribunal to hear the applications but, without determining them, refer them with recommendations to the Minister for determination.
The site is located approximately 1.2 kilometres to 1.7 kilometres northeast of an area used for drying/disposing of bauxite residue resulting from alumina production and close to a sand quarry. A previous subdivision application of part of the site was refused by the Tribunal in an earlier review proceeding having regard to potential dust impacts from the bauxite residue disposal area and the sand quarry. The applicants subsequently commissioned air quality monitoring on the site in order to demonstrate that the proposed subdivisions are acceptable in relation to the health and amenity impacts of dust.
The Tribunal heard concurrent expert evidence from a panel of eight expert witnesses in relation to air quality. Having heard this evidence, the Tribunal was not satisfied that the proposed subdivisions are acceptable in relation to the amenity impacts of dust, because the air quality monitoring for amenity purposes was only undertaken on days adding up to about onefifth of a year and the monitoring and evidence presented by the applicants did not address shorter term than 24 hour adverse amenity impacts from dust, even though such impacts on residents of the proposed subdivisions are likely to be real and material.
The Tribunal also considered that the applicants' air quality monitoring cannot reasonably be relied upon to demonstrate that the subdivisions are acceptable in relation to the health and amenity impacts of dust in the future, because it is unlikely that relevant conditions during the monitoring year are representative of the future. In particular, relevant climatic conditions during the monitoring year were average and not worst case in comparison to recent years; climate change evidence suggests that relevant climatic conditions in the future will be different; and there will be a significant increase in the height of the bauxite residue disposal area where it is closest to the site (and a consequent altitudinal increase in wind speed) between 2014 and 2030.
Furthermore, for these reasons, the Tribunal considered that it is likely that relevant conditions affecting dust generation and impacts on the site will be different in the future and consequently that dust impacts on the site are likely to be greater from time to time, at least until operations in the most proximate part of the bauxite residue disposal area to the site and at the sand quarry cease and those areas are rehabilitated. The evidence indicates that this will not occur for at least 15 to 20 years.
The Tribunal therefore reported to the Minister that it is not satisfied that the proposed subdivisions are acceptable in relation to the health and amenity impacts of dust. The Tribunal also, consequently, reported to the Minister that it is of the view that proposed subdivisions are inconsistent with the applicable planning framework, would likely adversely affect the industrial amenity of the bauxite residue disposal area, and would set an adverse planning precedent.
The Tribunal therefore recommended to the Minister that the correct and preferable decision is that the applications for review should be dismissed and subdivision approval should be refused for the proposed subdivisions.
Category: B
Representation:
DR 362 of 2013
Counsel:
Applicant : Mr PJ McQueen
Respondent : Ms CA Ide
Intervener : Mr PG McGowan
Solicitors:
Applicant : Lavan Legal
Respondent : State Solicitor for Western Australia
Intervener : Clayton Utz
DR 444 of 2013
Counsel:
Applicant : Mr MC Hotchkin (7 and 8 May 2014)/ Mr PJ McQueen (1 to 5 and 16 September 2014)
Respondent : Ms CA Ide
Intervener : Mr PG McGowan
Solicitors:
Applicant : Hotchkin Hanly (7 and 8 May 2014)/ Lavan Legal (1 to 5 and 16 September 2014)
Respondent : State Solicitor for Western Australia
Intervener : Clayton Utz
Cases referred to in report:
Macri v Western Australian Planning Commission [2014] WASC 153
Nicholls and Western Australian Planning Commission [2005] WASAT 40; (2005) 149 LGERA 117
Wattleup Road Development Company Pty Ltd and Western Australian Planning Commission [2011] WASAT 160
Wattleup Road Development Company Pty Ltd and Western Australian Planning Commission [2014] WASAT 29
Introduction
1 Wattleup Road Development Company Pty Ltd (Wattleup) and Primewest (Wattleup) Pty Ltd (Primewest) each seeks review, under s 251(1) of the Planning and Development Act 2005 (WA) (PD Act), of the deemed refusal by the Western Australian Planning Commission (Commission) of a residential subdivision application that it made to the Commission in respect of land at Wattleup Road, Hammond Park. Wattleup proposes the subdivision of Lots 809, 9002 and 9003 Wattleup Road into 147 residential lots and two public open space reserves. Primewest proposes the subdivision of Lots 117 and 805 Wattleup Road into 93 residential lots, two balance lots and two public open space reserves. The five lots which are the subject of the subdivision applications are located on the southern side of Wattleup Road, either adjoin or are closely proximate to one another and comprise five of the six lots between Lot 9002 to the east and Lot 9003 to the west. The five lots which are the subject of the subdivision applications are collectively referred to in this report as 'the site'.
2 The Minister for Planning, the Hon John Day MLA (Minister), has formed the opinion, under s 246(1) of the PD Act, that each of the applications for review raises issues of such State or regional importance that it would be appropriate for it to be determined by the Minister. The Minister, therefore, 'called in' both applications for review under s 246(2) of the PD Act and, in particular, directed the Tribunal, under s 246(2)(b) of the PD Act, to hear each application but, without determining it, refer it with recommendations to the Minister for determination.
3 The Commission subsequently resolved to advise the Tribunal that it supports approval of the subdivision applications subject to conditions. Wattleup and Primewest each accept the conditions proposed by the Commission. The applicants and the Commission each therefore contend that subdivision approval should be granted for the proposed subdivisions subject to the agreed conditions.
4 However, the Tribunal granted leave, under s 37(3) of the State Administrative Tribunal Act 2004 (WA) (SAT Act), to Alcoa of Australia Limited (Alcoa) to intervene in the proceedings in relation to whether the proposed subdivisions are acceptable in relation to health and amenity impacts of dust: see Wattleup Road Development Company Pty Ltd and Western Australian Planning Commission [2014] WASAT 29. Alcoa contends that subdivision approval should be refused for each of the proposed subdivisions.
5 The site is located in the district of the City of Cockburn (City). The City advised the Commission that it supports the proposed subdivisions. However, Mr Andrew Trosic, a town planner and the City's Manager of Strategic Planning, gave evidence that the City's support of the subdivisions is 'premised on the Tribunal accepting that the subdivisions are acceptable from a dust amenity and health point of view[.]' (T:99.5; 08.05.14)
6 The review applications were heard together over eight days in May and September 2014. The Tribunal comprised a judicial member, a senior sessional member who is an environmental scientist and a senior sessional member who is a town planner. The Tribunal received 51 exhibits and heard evidence from 12 witnesses, including eight expert witnesses in various disciplines who gave concurrent evidence in relation to air quality in a panel session which occupied four days (not all of the eight witnesses remained for the whole of the panel session, as some of the topics did not fall within the expertise of some witnesses). On the first day of the hearing, the Tribunal conducted a view of the site and of Alcoa's Residue Drying/Disposal/Storage Area (RDA) accompanied by the parties' representatives and various expert witnesses. After the evidence was concluded, the Tribunal adjourned for 10 days to enable the parties to file and exchange written closing submissions and then reconvened to hear oral closing submissions.
7 This document comprises the Tribunal's report to the Minister under s 246(2)(b) of the PD Act. For reasons set out in this report, the Tribunal considers that the 'correct and preferable decision' under s 27(2) of the SAT Act is that the applications for review should be dismissed and subdivision approval should be refused for both of the proposed subdivisions, because the Tribunal is not satisfied that the proposed subdivisions are acceptable in relation to the health and amenity impacts of dust and also because, in consequence, the proposed subdivisions are inconsistent with the applicable planning framework, are likely to adversely affect the industrial amenity of Alcoa's RDA and would set an adverse planning precedent for the approval of relevantly indistinguishable subdivision applications on other land on the southern side of Wattleup Road. Under s 246(2)(b) of the PD Act, the Tribunal therefore recommends to the Minister that the Minister should dismiss the applications for review and refuse to grant subdivision approval for the proposed subdivisions.
RDA and sand quarry
8 The site is located approximately 1.2 kilometres to 1.7 kilometres northeast of the boundary of the Alcoa RDA which Alcoa uses for drying/disposing of bauxite residue resulting from its alumina production facility at Kwinana. The RDA occupies approximately 625 hectares of land of which approximately 170 hectares is active residue storage area. The RDA receives approximately 4.4 million tonnes of bauxite residue each year.
9 Bauxite residue consists of a coarse sand fraction (termed 'red sand') and a fine silt fraction (termed 'red mud'). The drying/disposing of the bauxite residue takes place in large open areas on the RDA commonly known as 'red mud lakes'. The lakes are progressively built up with bauxite residue to a height of 80 metres AHD with a typical embankment slope of 1:6.
10 As Dr David Honey, Alcoa's Western Australian Operations Residue Manager, explained in his evidence to the Tribunal, in order to minimise dust emissions from the RDA, 'the embankments which are not under active work are vegetated or covered with aggregate, mulch or bitumen' (Exhibit 8 paragraph 6.12), and when an area is no longer 'operational', that is 'once we've closed the area[,] it will be [planted] with native vegetation to the top' (T:46.746.8; 01.09.14). However, until that occurs, there is potential for dust generation from the RDA when wind speed reaches 5 metres per second or greater. The prevailing wind direction on summer afternoons is from the south west, that is from the general direction of the RDA to the site. The applicants' chemical composition evidence shows that bauxite residue dust from the RDA reaches the site.
11 Dr Honey gave evidence that Alcoa anticipates operating the RDA for at least the next 20 years. However, the current State Agreement allows the RDA to operate for another 30 years, until 2045.
12 There is also a large sand quarry operated by WA Limestone to the south and southeast of the site which is only approximately 50 metres from the closest part of the site. The WA Limestone sand quarry has operated since 1998. Although the current development approval for the sand quarry is for five years, until January 2016, the most recent development application in June 2010 sought approval 'for the whole site for a period of 20 years to permit resources to be taken' (Exhibit 47). It is therefore likely that the sand quarry will operate for at least another 16 years, until at least 2031.
2011 Wattleup decision
13 In October 2011, the Tribunal refused subdivision approval for an essentially identical subdivision to the subdivision now proposed by Wattleup, having regard to potential dust impacts from the RDA and the sand quarry: see Wattleup Road Development Company Pty Ltd and Western Australian Planning Commission [2011] WASAT 160 (2011 Wattleup decision). In the 2011 Wattleup decision, the Tribunal determined that although it found that 'the proposed subdivision is consistent with the longterm strategic and statutory planning framework for the site' and that this finding 'would usually be a powerful and compelling consideration in favour of approval of the proposal', in the particular circumstances of the case, '[t]he precautionary principle … is an overwhelming consideration in favour of refusal of the proposed subdivision': 2011 Wattleup decision at [67]. The Tribunal balanced these considerations in the exercise of planning discretion at [66] of the 2011 Wattleup decision as follows:
As noted earlier, a finding, such as in this case, that a proposed subdivision is consistent with the long term strategic and statutory planning for a site would usually be a powerful and compelling consideration in favour of approval. However, in the circumstances of this case, the precautionary principle warrants refusal, unless and until adequate air quality monitoring is undertaken and reviewed in relation to the site demonstrating that the proposed subdivision would be acceptable in relation to dust health and amenity impacts. The evidence shows that dust can cause significant adverse health and amenity impacts. Furthermore, the Urban Deferred classification was 'lifted' in the face of advice from the EPA that there could not be confidence in predictions about air quality in the vicinity of the site. As, under the precautionary principle, it must be assumed that the threat of serious or irreversible environmental damage is no longer uncertain but is a reality, precluding subdivisions until adequate monitoring of air quality on the site demonstrates the acceptability of the proposal is a proportionate, appropriate and costeffective measure to ensure that the incoming population will not be exposed to unacceptable health or amenity impacts. (Emphasis added)
14 The Tribunal also said the following in its conclusion at [68] of the 2011 Wattleup decision:
There is a threat of serious or irreversible environmental damage for residents of the proposed subdivision in relation to dust from the RDA and the sand quarry to the south and south-east of the site. Furthermore, there is scientific uncertainty as to the environmental damage. The conditions precedent to the satisfaction of the precautionary principle are therefore established. Consequently, a precautionary measure may be taken to avert the anticipated threat of environmental damage, provided that it is proportionate to the threat. Precluding subdivision until adequate air quality monitoring at the site demonstrates the acceptability of the proposal is proportionate to the threat, appropriate and cost-effective. Adequate air quality monitoring requires prior consultation with the DEC, monitoring for a period of at least 12 months and assessment and reporting in relation to relevant standards for PM10 and TSP and chemical composition of dust. (Emphasis added)
15 As Alcoa correctly submits, the present proceedings are not a 'continuation' of the earlier Wattleup proceeding. These are fresh applications for review. Furthermore, as Alcoa submits, 'nothing in the [2011 Wattleup decision] did or could constitute a direction by the Tribunal, satisfaction of which, precludes the full exercise of discretion by the Tribunal in these proceedings.' In the 2011 Wattleup proceeding, the Tribunal also did not have the benefit of the detailed air quality evidence given by five of the eight air quality expert witnesses called in the present proceedings and did not have the benefit of submissions from Alcoa. Nevertheless, consistently with the Tribunal's determination in the 2011 Wattleup decision, as a matter of orderly and proper planning, in order for the consent authority to approve the proposed subdivisions, it must be demonstrated to its satisfaction that the proposed subdivisions are acceptable in relation to the health and amenity impacts of dust.
Applicants' monitoring
16 Following the 2011 Wattleup decision, Dr Peter Forster, an environmental consultant and chemist who gave evidence on behalf of Wattleup in the 2011 Wattleup proceeding (and who also gave evidence on behalf of Wattleup and Primewest in these proceedings) was, as he agreed in crossexamination in these proceedings, 'charged with the responsibility of setting up as a project what [Wattleup] wanted in relation to what [Wattleup] understood was a requirement of the Tribunal [in the 2011 Wattleup decision]' (T:50.4; 03.09.14). Dr Forster agreed in crossexamination that he was 'the person … engaged by [Wattleup] to develop a methodology to carry out what was understood to be a requirement for 12 month's monitoring' on the site (T:52.2; 03.09.14). Dr Forster said that he 'wrote the proposal' (T:52.1; 03.09.14) and agreed that he 'determined what methodology' he would recommend to the Department of Health (DoH) and the Department of Environment and Conservation (DEC) (now the Department of Environment Regulation (DER)). Dr Forster gave evidence that:
The pathway was provided to the agencies and there were modifications made to the program based on their requirements. The modifications were made, the proposal was finalised, submitted to the the client.
(T:52.8; 03.09.14)
17 Dr Forster gave evidence that consultation with DEC included an assessment by DEC of the location of the monitoring station on the site, a critique of the methods proposed for the monitoring, a direction from DEC as to the frequency of sampling for Total Suspended Particulates and agreement on reporting frequency as the monitoring program progressed. Dr Forster also said that the proposed monitoring program was presented to officers of DoH who provided recommendations for health and amenity based air quality standards to be used to assess the outcomes from the monitoring. On 24 February 2012, Mr Eric Lumsden, the Director General of the Department of Planning, wrote to Wattleup confirming his advice that Wattleup 'has now obtained the support (approval) of DEC and DoH to the scope and content of the required monitoring program and subsequent reporting of the results.'
18 Strategen Environmental Consultants Pty Ltd (Strategen) was commissioned by Wattleup to conduct air quality and meteorological monitoring in accordance with the DEC/DoH approved monitoring program on the site during the period 1 July 2012 to 31 July 2013. Strategen engaged Dr Forster to manage the monitoring program. This engagement included 'authoring the Strategen reports on the scope of work and the results of the monitoring program' (Exhibit 29 paragraph 38). The monitoring carried out by Strategen is referred to in this report as 'the applicants' monitoring'.
19 Mr Martin Matisons, the Principal Toxicologist in the Environmental Health Directorate of DoH, confirmed in his evidence that both his Department and the Air Quality Management Branch of DEC/DER 'worked closely with the applicant regarding the dust monitoring program to be established' and that '[t]oxicologists from DoH, myself included, attended all briefings (approximately 3-monthly) given by the consultants regarding progress of the dust monitoring, including update on results' (Exhibit 19 paragraphs 12 and 14). Mr Matisons also gave evidence that DoH 'considers that the applicant engaged a suitably qualified and experienced consultant to undertake the dust monitoring' and that DoH 'is satisfied with the scope of the monitoring program' (Exhibit 19 paragraphs 13 and 15). Ms Constance Dewan, an environmental scientist and chemist, and the Senior Environmental Officer in the Air Quality Management Branch of DER, also gave evidence that 'we were happy with the monitoring' (T:116.2; 03.09.14).
Issues for determination
20 The parties each filed statements of issues, facts and contentions raising various issues. The Tribunal distilled the issues identified by the parties and suggested to the parties during the hearing that there are four principal issues for determination. The parties agreed with the Tribunal's identification of the issues. The following four principal issues arise for determination in these proceedings:
1) Whether the proposed subdivisions are acceptable in relation to the health and amenity impacts of dust.
2) Whether the proposed subdivisions are consistent with the applicable planning framework.
3) Whether the proposed subdivisions would adversely affect the industrial amenity of the RDA.
4) Whether approval of the proposed subdivisions would set an adverse planning precedent.
21 The Tribunal will address each of these issues in turn.
Are the proposed subdivisions acceptable in relation to the health and amenity impacts of dust?
22 The Tribunal heard expert evidence in relation to air quality from the following eight witnesses:
Witness | Expertise | Called by |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
23 The air quality expert witnesses agree that the currently accepted regulatory standard in Western Australia for health impacts of dust on residential populations is the National Environment Protection (Ambient Air Quality) Measure (NEPM) standard of PM10, meaning 50 micrograms per cubic metre of dust particles with a maximum diameter of 10 micrometres, averaged over 24 hours. NEPM allows a maximum of five exceedances of this standard per year.
24 The applicants' monitoring included monitoring for PM10 every day during the period from 1 July 2012 to 30 June 2013 and found two exceedances of that standard. Although Ms Dewan expressed some concerns in relation to these exceedances, the expert witnesses in relation to air quality all agreed that the applicants' monitoring was appropriately carried out and demonstrates compliance with the PM10 NEPM standard.
25 The Tribunal is, therefore, satisfied that the health impacts of dust were acceptable for the proposed subdivisions during the year of monitoring. However, the Tribunal does not accept the Commission's submission that the 'evidence before the Tribunal is that the subject land is not presently subject to dust levels which would pose an unacceptable health impact' (emphasis added), because, for reasons set out below, the applicants' monitoring results cannot reasonably be relied upon as predictive of the health impacts of dust on the site at present or in the future.
26 The air quality expert witnesses agree that the currently accepted regulatory standard in the area of the site for amenity impacts of dust on residential populations is the Kwinana Environmental Protection Policy (Kwinana EPP) standard of Total Suspended Particulates (TSP) of 90 micrograms per cubic metre of dust particles of any size, averaged over 24 hours. The applicants' monitoring did not detect any exceedances of this standard. However, the applicants' monitoring for TSP was undertaken for a total of only 72 or 74 days (there is a discrepancy in the evidence as to whether it was 72 or 74 days, although this difference is inconsequential) or days adding up to a total of only approximately onefifth of the monitoring year.
27 Dr Forster gave evidence that 'the methodology for the program was developed in conjunction with the [D]epartments for [E]nvironment [R]egulation and [H]ealth' and 'the Department of Environment Regulation actually had requirements for the frequency of TSP monitoring' (T:160.7; 02.09.14). Ms Dewan and Mr Matisons both agreed in their evidence with Dr Forster that DEC/DER and DoH had been consulted in developing the methodology for the applicants' monitoring and, in particular, that their Departments recommended that monitoring for TSP did not have to take place every day, but only one in three days during the socalled 'dust season' from September to April and one in six days during the rest of the monitoring year.
28 There is a discrepancy in the evidence as to whether DoH and DEC/DER required monitoring for TSP for 80 days, 90 days or 103 days during the year. As noted earlier, there is also a discrepancy as to whether 72 days or 74 days of TSP monitoring was in fact undertaken. (The difference between the number of days of monitoring required by DoH and DEC/DER and the number of days of monitoring actually provided was in part explained by Dr Forster in that some 24 hour samples had insufficient dust deposited in the filer to allow chemical analysis and were therefore aggregated over two or more days). Nothing turns on which of these numbers is in fact correct. The Tribunal considers that any of these number of days was wholly inadequate to assess whether the proposed subdivisions are acceptable in relation to the amenity impacts of dust.
29 Ms Dewan gave the following evidence in relation to the frequency of TSP sampling:
Just to say that the one in six days for TSP sampling is generally recommended to ensure that and what Peter Forster said to ensure that every day of the week is sampled and that you don't only have one particular day, so to cater for the weekends and the weekdays. And we also take into consideration the fact that monitoring can be a very expensive exercise and, your Honour, is especially for high volume samples using for TSP analysis.
So we usually recommend one in six days, unless the situation requires if the risks are very high then we may recommend a greater degree of monitoring. For example, in this monitoring we did recommend one in three days during the summer period, because we expected higher dust emissions during that time.
…
Ideally we would like to have continuing monitoring for everything. That will give you a very good assessment like what Mr Ormerod has pointed out. Ideally you would want to have even for the TSP samples you would want to have daily samples. But like I indicated, it's because of the costs involved, which we consider that as well, and the method you have to change the filter every 24 hours.
It's a very onerous process. So because of those difficulties, the cost and the onerousness of the process of sampling, that's why we make it one in six days. And the one in six days is also to make sure that you're monitoring emissions from a particular facility on different days of the week, not just every Sunday or every Wednesday. And so that one in six days covers everything. It helps address the cost issue, it helps address the onerous nature of the sampling and it also ensures that you are sampling in different days of the week.
(Emphasis added) (T:195.5195.7 and 199.5199.8; 02.09.14)
30 Ms Dewan's evidence does not provide a satisfactory explanation as to why DEC/DER and DoH required sampling for amenity purposes for days adding up to only approximately one-quarter of the year and accepted sampling for amenity purposes for days adding up to only approximately one-fifth of the year. Sampling for a full year may well be 'costly' and 'onerous'. However, the proposed subdivisions would facilitate the introduction of large, permanent, residential populations on the site in circumstances where '[t]here is a threat of serious or irreversible environmental damage for residents of the proposed subdivision in relation to dust from the RDA and the sand quarry …' (2011 Wattleup decision at [68]). The 'cost' and 'onerousness' of undertaking a full year of monitoring for amenity purposes, rather than days adding up to one-fifth or one-quarter of a year, is necessary given the proximity of the site to the RDA and the WA Limestone sand quarry and the nature of the proposal. In the circumstances of this case, daily monitoring for TSP is not only 'ideal', but necessary for an adequate monitoring program.
31 On the first day of the air quality concurrent evidence, Mr Matisons gave the following evidence in relation to the TSP monitoring:
I will make a comment that the Department of Health did advise the consultants undertaking the monitoring on the program that they should undertake. Our advice was consistent with or similar programs in terms of what needs to be measured, what standards you're looking at and how the measurements are undertaken. And that advice was provided to the consultant, along with advice from the Department of Environment [R]egulation.
In terms of the adequacy with standards that we used and the nature of the elements that we asked to be measured, we consider that to be adequate and, in our experience, that is we can make satisfactory decisions based on those results. It does depend sometimes on the results you get as to whether monitoring needs to be continued if there are issues there that need to be followed up. In this case there weren't any and we were happy with the end result and to look at the end result and make a decision based on that. (T:195.1195.4; 02.09.14)
32 On the second day of concurrent evidence, Mr Matisons returned to the topic of TSP sampling and gave the following evidence:
… I just want to this is just commenting on the six monitoring every sixth day. There is an Australian standard for sampling TSP and it does discuss monitoring every seventh or every sixth day. And that's the national standard that is actually used. So it's ASNZS3580 dated 2003.
(T:41.5; 03.09.14)
33 Mr Matisons said that he had 'only found that last night to confirm where it was and I apologise that it hasn't been done earlier' (T:41.6; 03.09.14). The 'national standard' referred to by Mr Matisons in his evidence is the Australian/New Zealand Standard Methods for sampling and analysis of ambient air Method 9.3: Determination of suspended particulate matter Total suspended particulate matter (TSP)High volume sampler gravimetric method (Australian Standard). Mr Matisons' reference to the Australian Standard took the Tribunal and counsel (even Ms CA Ide who called him: 'I didn't know he was going to raise that today' (T:44.9; 03.09.14)) by surprise. The Australian Standard was tendered (Exhibit 50) and was reviewed by the Tribunal and the parties during an adjournment.
34 However, as was discussed in the evidence following the adjournment, the Australian Standard does not require or even support DoH's and DEC/DER's decision to require sampling for days adding up to only about onequarter of the year. The Preface states that the 'objective' of the Australian Standard is 'to provide regulatory and testing bodies interested in the sampling and monitoring of ambient air with a standard method for the determination of total suspended particulate matter in ambient air'. Clause 7.7 of the Australian Standard sets out a procedure for sampling, including the following in paragraph (e):
Set the high volume sampler time clock to commence running at a predetermined time and for a predetermined period.
NOTE: For routine monitoring, sampling is often carried out every sixth day for 24 h, from midnight to midnight. Other sampling periods and sampling frequencies may be instituted, if required. For sampling periods of less than 24 h, it is preferable to sample on the hour.
(Emphasis added)
35 As Mr Matisons agreed when these provisions were brought to his attention, the Australian Standard does not require any particular sampling period and only contemplates sampling every sixth day 'for routine monitoring' (T:106107; 03.09.14). The note also states that other sampling periods and sampling frequencies may be instituted, if required.
36 Mr Ormerod and Mr Gwynne also gave the following evidence which is entirely consistent with the Preface and cl 7.7 of the Australian Standard and which the Tribunal accepts:
ORMEROD, MR: Just very briefly, you know, your Honour. Dr Forster, I agree that his comment that there's no prescription, as such, in the in the Standard as to for how long, or how often to measure. It is a more a technique guidance than anything else. It's really a minimum a minimum Standard for monitoring with that particular type of device. So that, yes, the and I think the routine really refers to the type of monitoring that is conducted year in, year out, by regulatory agencies at fixed locations and monitoring stations, and also long-term monitoring as part of licence compliance programs through industry.
(T:113.3113.4; 03.09.14)
GWYNNE, MR: I support Dr Forster's and Mr Ormerod's comments, and simply say that the Standard makes it's not a Standard with regard to with regard to sampling design. It's purely to do with how you do the sampling. So the frequency is not an aspect of the Standard at all. So that's part of your sampling program design. And that's varied, depending on what you want to get out of your sampling program.
(T:113.7; 03.09.14)
37 As noted earlier, in his evidence given on the first day of concurrent evidence on this topic, Mr Matisons said that '[o]ur advice was consistent with or similar programs in terms of what needs to be measured, what standards you're looking at and how the measurements are undertaken' (T:195.2; 02.09.14). However, when asked by the Tribunal on the second day of concurrent evidence as to whether DoH has previously been involved in nominating the frequency of monitoring for a subdivision or development approval purpose, rather than for, for example, an ongoing assessment of a plant, Mr Matisons could recall only one instance of a proposed rather than an existing development. In that case, DoH required daily TSP monitoring. Ms Dewan, who was also familiar with that case, confirmed that daily monitoring was required for TSP in that instance (T:109110; 03.09.14).
38 Alcoa submits that the monitoring proposal that DEC/DER and DoH agreed to 'is not determinative of these proceedings' and that '[t]he proper question for the Tribunal is whether the actual monitoring undertaken provides a proper basis to conclude that there is no unacceptable amenity or health impact for future residents of the proposed subdivisions.'
39 The planning consent authority (in this case the Minister, following the report and recommendations by the Tribunal) is certainly not bound by the advice given by DEC/DER and DoH to Dr Forster and Strategen. Nevertheless, usually, the planning consent authority would place considerable weight upon the advice of specialist environmental and health regulatory authorities. However, in this case, no meaningful basis has been put forward by either authority to support the soundness of the nomination of less than a full year of daily monitoring for TSP amenity purposes.
40 Mr Ormerod is an experienced environmental scientist with 35 years' specialist experience in air quality including air quality associated with the alumina industry. He has written or coauthored over 50 papers on air quality issues. He was also a most impressive witness throughout a four day concurrent evidence process and the Tribunal found his evidence to be measured, balanced, objective and thoughtful. Mr Ormerod volunteered the following evidence in relation to the Strategen monitoring:
… the monitoring was conducted in accordance with, obviously, an agreed protocol or agreed arrangement with with the department, and, you know, to me it appears to have been conducted well and it and it provides some very good analyses and good information. (T:175.6; 02.09.14)
41 However, Mr Ormerod also gave the following evidence:
24-hour TSP, for example. So so we're dealing with a you know, 365 days in a year a sample of 80 days, so we we conceivably have, on some of those other 280odd days in the year, higher concentrations than have been measured on those 80 days. So we don't know the answer to that; therefore, it's not possible to say strictly whether conditions do comply with with the standard or not.
(T:176.9177.2; 02.09.14)
42 Similarly, Mr Gwynne, who also has considerable relevant experience and gave balanced, objective and helpful expert evidence, said:
GWYNNE, MR: So that means that we've actually really only got 72 24hour samples which is, I don't believe, sufficient to draw too many conclusions from. …
And, I mean, by contrast, I Alcoa's monitoring at at Sayer Road there's 360odd daily TSP samples from just down the road and my my my interpretation that that the 72 are not necessarily representative is based on comparison in some in part, and based on comparison with those where they did, on other days, measure higher concentrations.
THE D. PRESIDENT: So Alcoa does it every day?
GWYNNE, MR: Every day at Sayer Road, which is a short distance to the west. So I guess in terms of sufficiency, I don't deem that to be a sufficient monitoring program to characterise the now. The the origin of the, you know, three-day and six-day monitoring programs is to assess longer term I guess, to assess against longer term criteria. So the idea of taking a sample every three days or a sample every six days is in order for you to collect a databank which you can then assess against, say, an annual average. So you take all your 70 or 80 samples you collect over the year, you average them out, and you assess against an average criterion, and in that basis you should if you take, you know, a sample every three days or a sample every six days, it should average out to the same number as if you had taken a sample every day. That's the the statistical theory behind behind not doing every day. But if you are assessing against a 24 24hour criterion, if you sample anything less than every 24 hours, you are running the risk of missing a a compliance or you can't state that you're you're always in compliance. You can state that you were in compliance for 72 days, but but no more.
(T:179.9, 180.5181.2; 02.09.14)
43 Dr Forster gave evidence that '[n]o exceedances of the ambient air quality standard for total suspended particulates (TSP) were observed' (Exhibit 29 paragraph 66.1) by the applicants' monitoring. While this is, strictly speaking, correct, the Tribunal accepts Mr Ormerod's and Mr Gwynne's evidence that it cannot be said that the Kwinana EPP TSP standard has been complied with on the basis of only 72 or 74 days of monitoring; much less can it be said on the basis of the applicants' monitoring that the proposed subdivisions are acceptable in relation to the amenity impacts of dust, whether during the year of monitoring, at present or in the future.
44 Furthermore, as Mr Gwynne said, Alcoa was able to monitor for TSP in accordance with the Kwinana EPP at its Sayer Road monitoring station to the west of the site on approximately 360 of the 365 days in the 12 months from July 2012 to June 2013. Alcoa's monitoring during that period revealed 12 exceedances of the Kwinana EPP TSP standard at its monitoring station at Sayer Road.
45 Dr Forster responded to this evidence by referring to a 'correlation' assessment that he carried out between the results of sampling at Sayer Road on the days on which amenity sampling was also undertaken as part of the applicants' monitoring on the site, and suggesting that because he found a poor 'correlation coefficient' between the two sets of samples, 'the use of Sayer Road data to infer or to predict concentrations at the applicant's site in Wattleup Road [has] serious shortcomings, just as the consequence of the poor correlation' (T:203.8; 02.09.14). However, as Mr Gwynne said in response to Dr Forster's correlation assessment:
… it would only be a valid comparison is if you had two completely comparable data sets which means that you would need to have two data sets where the sampling started. …
However, we don't have that in our case so I don't believe that there is a lot of weight that you can place on a correlation, a simple correlation, in terms of time paired correlation in this particular instance because it's not statistically valid. You are to use Dr Lyons' expression comparing apples and oranges and to try and correlate them is mathematically incorrect.
(T:208.8208.9; 02.09.14)
46 As Mr Gwynne said, not only are the data sets from the Sayer Road monitoring station and the site not validly comparable because of the different number of days of monitoring at each location during the monitoring year, but also because 'we've got different start times and end times even [where] we're reporting against the same day … .' (T:209.9; 02.09.14)
47 Mr Coffey, an environmental engineer with over 40 years' experience, agreed with Mr Gwynne's criticism of Dr Forster's 'correlation' assessment as follows:
… I just make the point that that was for 365 days or 350-odd days of monitoring. So we're not, again, comparing apples with apples in the sense that at Wattleup Road there was only 83 days of TSP. …
(T:211.5; 02.09.14)
48 In light of Mr Gwynne's and Mr Coffey's sound criticism, we do not accept the correctness or utility of Dr Forster's 'correlation' assessment in relation to the results of sampling at Sayer Road. Although the Sayer Road monitoring station is located north of the RDA, whereas the site is located northeast of the RDA, and although the Sayer Road monitoring station may well be affected by nearfield dust sources which may not affect the site (just as the site may be affected by nearfield dust sources, such as the sand quarry, which may not affect the Sayer Road monitoring station), the Sayer Road monitoring station is located only 300 metres closer to the RDA than the site. The fact that TSP monitoring was undertaken at the Sayer Road monitoring station for approximately 360 days between July 2012 and June 2013 and that 12 exceedances of the Kwinana EPP TSP amenity standard were observed in that location during that period, whereas only 72 or 74 days of TSP monitoring was undertaken during the same period at the site, strongly suggests that had TSP monitoring been undertaken at the site for the full year or close to it, exceedances of the amenity standard would have been observed. (Interestingly, in its submissions, Alcoa notes that the two exceedances of PM10 that were recorded in the applicants' monitoring at the site (26 February 2013 and 10 April 2013) occurred on days when no TSP sampling was undertaken at the site).
49 However, even if the Tribunal were to ignore the results of TSP monitoring at Sayer Road during the period of the applicants' monitoring, the Tribunal considers that the applications warrant refusal because of the absence of evidence that would be necessary to satisfy it that the proposed subdivisions are acceptable in relation to the amenity impacts of dust even in the year of monitoring.
50 Finally, on this aspect, the TSP amenity monitoring undertaken on behalf of the applicants plainly does not satisfy the Tribunal's contemplation in the 2011 Wattleup decision at [68] of 'monitoring for a period of at least 12 months and assessment and reporting in relation to relevant standards for PM10 and TSP and chemical composition of dust' (emphasis added).
51 The Tribunal is not, therefore, satisfied that the proposed subdivisions are acceptable in relation to the amenity impacts of dust, whether in relation to the year of dust monitoring between July 2012 and June 2013, currently or in the future.
52 Furthermore, the Tribunal is not satisfied that the proposed subdivisions are acceptable in relation to the amenity impacts of dust during the year of monitoring, currently or in the future, because of the applicants' failure to address what the air quality expert witnesses each recognised is 'a gap in the ability to assess the acceptability of air quality from an amenity perspective' (Exhibit 11 paragraph 11) as the 24 hour averaging under the Kwinana EPP TSP standard is likely to 'mask' the true adverse impacts of dust, both generally and from the RDA in particular, upon residents of the proposed subdivisions.
53 Mr Coffey gave the following evidence in relation to this aspect:
In my opinion an assessment of whether air quality is acceptable must include consideration of the fluctuating levels of dust over shorter time intervals than the 24hour averages. This is because 24hour averages can mask variations in dust levels that would be noticed by a human recipient.
Dust emissions result from wind driven processes, as well as arising from other causes. Wind by nature is highly variable, with shorter term fluctuations in wind speed and direction continuously varying around their mean values. It is the instantaneous wind gusts, as well as the average winds over short periods of time, that will typically drive the processes of dust generation. Sudden changes in wind direction and wind speed can occur, as for example the onset of the afternoon sea breeze on summer afternoons. Winds may strengthen and subside over periods of less than 15 minutes to several hours.
For all of these reasons, episodes of dust generation and its movement away from a source site will usually occur over time spans substantially shorter than a day that is minutes to hours. In some extreme circumstances these conditions may persist over a day or more, though this is the exception rather than the norm. In my experience dust emission events, when they occur, typically have durations from less than an hour up to several hours.
(Exhibit 4 paragraphs 9.1 9.3)
54 Similarly, Ms Dewan gave the following evidence:
One of the key problems we face is the average in times. 24 hours versus short term standards. Sometimes we find there could be impacts on a short term basis which are not captured in the standards because of the long averaging period. I would say that's the major limitation.
…
It's mainly due to amenity issues. We find that people experience dust events but when but looking at the monitoring data you find that the 24 hour standards have been have not been exceeded. But that doesn't mean that people are not being impacted.
(T:141.8 and 142.3; 02.09.14)
55 Mr Ormerod gave evidence to the same effect:
I'm very aware of from various projects and discussions over the years is the deposition or amenity guideline is quite a poor indicator of the effects on communities.
… in fact, the events that caused the most consternation occur over periods of hours generally speaking. So a significant dust event from one to four hours in duration, let's say, for argument's sake, could be enough to leave deposits of dust lying around the household and on cars and whatever.
People may have to wash things and clean things which they find a bit of a nuisance. All that sort of thing. These are the actual effects. They can occur quickly but they won't be picked up in a month long average deposition sample unless it keeps on happening on a very frequent basis, of course. So those are sorts of issues that are quite a limitation around trying to interpret amenity impacts of dust.
I've been to communities where there have been monitoring programs in place where companies have tried their best to deal with these things and people are told that the levels are within the guidelines, and they believe that they've been got at basically because their experience doesn't match the objective standards.
THE D.PRESIDENT: So the impact is masked by the strict application of the guidelines. Is that what you're saying?
ORMEROD, MR: Correct. That's right.
(T:142.5-143.5; 02.09.14)
56 Similarly, Mr Gwynne gave the following evidence:
I reiterate that point which is, put another way, strict adherence to the guidance to those guidelines is by no means protective of amenity and that has been borne out many times, and it's not just dust amenity, that this is a common issue with all sorts of amenity criteria. Odour is another one that falls into the same category. The significant limitation to do with (indistinct) on amenity, just to extrapolate a little bit, is mostly to do with the averaging period.
I would reiterate that. The dust events that we have observed and monitored are typically between one and probably eight or nine hours, and there's a very good reason for that. That's pretty much the meteorological time it takes for a front or a wind system to materialise and that's a sort of meteorological effect, and that's what drives the dust in this particular environment. So we are talking about dust events which have their basis in meteorology and that time scale is not captured by the current guidelines. (T:143.6-143.9; 02.09.14)
57 The applicants did not provide any meaningful response to this evidence. Indeed, Dr Forster and Dr Kabay both recognised in the experts' joint statement (Exhibit 11 paragraph 11) that there is a 'gap' in the ability to assess the acceptability of air quality from an amenity perspective.
58 In light of the evidence of the expert witnesses set out earlier and the prevailing southeasterly winds on summer afternoons, during the socalled 'dust season', it is highly likely that residents of the proposed subdivisions would experience adverse amenity impacts of dust during shorter, concentrated periods than the 24 hours averaging periods used for TSP monitoring under the Kwinana EPP. The periods of adverse impact are likely to correspond with the 'meteorological time it takes for a front or a wind system to materialise', as Mr Gwynne said, of 'typically between one and probably eight or nine hours' (T:143.8; 02.09.14). Whilst lasting less than 24 hours, such adverse amenity dust impacts on residents of the proposed subdivisions are likely to be real and material.
59 It is also to be noted that, as Dr Forster conceded under crossexamination, the monitoring program was developed by him at a time when he was unaware of the 'gap', as he only became aware of it during the conferral between expert witnesses required by the Tribunal in these proceedings (T:53; 03.09.14). Dr Forster made the following significant concession under crossexamination:
McGOWAN, MR: You said that you believe that the methodology, the process, the package, the proposal that you were putting up in May 2012 was sufficient to enable short-term dust events, so far as they related to amenity and so far as they were related to a shorter period of 15 minutes, to be captured.
131 Dr Kabay's fourth criticism, namely that the GHD report's dispersion model 'grossly overpredicts' PM10 levels from the RDA on the site, was his principal concern expressed in his evidence in this proceeding. Dr Kabay demonstrated to the Tribunal, in a careful and detailed analysis, that in a handful of what the model understands to be very high wind events, it significantly overpredicts PM10 health criterion exceedances when compared to actual monitoring data at the site. Dr Kabay said that, while this occurs only in a small number of cases, 'a regulator or an impact assessor would look at the top concentration' (T:151.5; 03.09.14). Dr Kabay, therefore, considers that the GHD report is not 'fit for purpose' (T:149.9; 03.09.14).
132 However, having heard Mr Gwynne's and Mr Ormerod's evidence on this point, the Tribunal considers that Dr Kabay's criticism is misplaced, because the GHD report's model was developed and is intended to be a 'statistical' model, not a 'deterministic' model. The GHD report is not intended to necessarily accurately predict the PM10 at a definite location at a definite time. Rather, as Mr Gwynne explained:
The other thing that the evidence that Dr Kabay's evidence shows quite clearly is that there is huge well, there's not there is variability significant variability in the model's ability to predict from day to day. And this is not surprising, because I do not have what I would call a deterministic model. I do not have a model that is designed, nor is it appropriate for it to consider it, to predict accurately every single day.
The sort of model that I've developed here is a statistical model. It's designed to be assessed statistically, as most of these models are. You would have to use a different totally different modelling approach to try and come up with a fully deterministic model. The variables in emissions of the [RDA] to start with are such that I can't possibly parameterise every single management activity that goes on on those [RDAs]. I've got to average them out.
I don't know exactly what was going on on that particular day on the [RDA], but I do know that what was going on on the [RDA] can have a huge effect on how much dust is emitted. Whether or not they're pouring; whether or not they have not poured; whether or not the sprinklers are working that day, or they're down for maintenance; whether or not their dust management is above or below, you know, what is their normal characteristic are just some of the influences that can happen on a day to day basis.
So anything that we do here that assesses individual days data, monitored versus predicted, as Dr Kabay has done, in my view, is inappropriate, okay. He has illustrated that that is a deficiency in the model and I well, it's not so much a deficiency as it's a limitation and I totally accept that this model does not accurately predict day by day numbers all that well. You need to look at it statistically. However, something that it does do is point out the, I guess, effect of that variability.
(T:12.3-13.1; 04.09.14)
133 Mr Ormerod strongly endorsed Mr Gwynne's evidence that the GHD report's model is, and is appropriately, a statistical, rather than a deterministic, model. Mr Ormerod gave the following evidence:
… Firstly, I guess, about the purpose and scope of dispersion modelling of this nature. It is true that the only valid way in which a model of this type, used in this way, should be looked at is in a statistical statistical modelling, not event modelling. Certainly, the event analysis can inform as to how it's constructed.
But Mr Gwynne was perfectly correct in saying that to expect a model like AUSPLUME, with a more or less hands-off approach or a simple approach to the inputs, it will not be capable of looking at specific events in time matched in time. But it will produce statistics that are highly useful for assessment. And that's the way most air dispersion modelling is done for regulatory purposes.
And Mr Gwynne's analysis, particular the last graphs that he showed you, the scatter plot and the quartile-quartile plot, those data certainly are the sort of thing you expect to find in a reasonably well-performing model. So I would agree with what he said on that. So we shouldn't expect too much out of out of any model. …
The common phrase that's used about models is that they have inherent limitations, which is true. But I think it can also be used as a bit of a copout to avoid understanding where they are valid and how they should be used correctly. Without models of this sort of type then, you know, we would be really guessing as to all sorts of impacts of air quality.
So we need them. And it's just important to drill down and understand what those limitations are. And, having understood that, move on and accept, you know, within reason, what those models are telling us. So that's, again, a fairly standard approach to living with imperfect models. But they have a high degree of utility in in certain applications. Now, Mr Gwynne mentioned the sort of the fairly narrow range of application of this particular model. It is a special purpose model that has been constructed. So, again, we shouldn't expect it to do anything outside that range of application. (T24.9-26.1;04.09.14)
134 Mr Ormerod was crossexamined at length about this topic. He gave the following evidence under crossexamination which further satisfies the Tribunal that Dr Kabay's criticism that the GHD report's model significantly overpredicts PM10 concentrations at the site is misplaced, given what is in fact the role and purpose of a statistical model, as opposed to a deterministic model:
McQUEEN, MR: And, certainly, it's accepted, I think, by me and by all the witnesses I heard yesterday, that when you're looking at the [middle] figures, or if you're looking at the cluster of numbers around the middle, there's not a lot of difference between modelling and monitoring. But I'm just putting it to you that for the outliers, at the more extreme end and you can see quite a big difference between what is modelled and what is monitored. And that the factor of two takes that into consideration.
WITNESS, ORMEROD: Yes. And that's again where the statistical nature of using [models] becomes really important. So if you're saying, at 1 o'clock yesterday, it was 100 and, in fact, it was measured as 50, that's a factor of two, that's not it's bad, but it's not as bad to be completely written off. But what you would be doing what you wouldn't be doing with the model, if you were using it properly, is to actually try to say what it was at 1 o'clock yesterday. You would say that, at some time under relatively similar circumstances, you could get a concentration like that. So there is a difference in how you look at it as far as interpretation.
…
McQUEEN, MR: And then you talk about a reasonably good level of performance.
WITNESS, ORMEROD: Yes. That's right. I and reasonably used in the in the sort of, you know, context that I think it is reasonable what has what has what has come out of it. Since the writing of that review, again, I've had the opportunity to understand more about, you know, having a chance to talk to Mr Gwynne and and discuss some of these things.
It it's quite clear to me now that I probably would have done the same sort of thing that he has done, in terms of we need not get too worried about some of the lower and higher percentile data. We're concerned about the the emissions under near worst case conditions, which then feed through into the rest of the the modelling exercise. So I'm not concerned at all about how that all ended up, to be perfectly honest.
…
McQUEEN, MR: All right. So but what I was talking to you about was prior to taking to that ultimate issue, the potential for the model to overstate at extreme events, and you have accepted that.
WITNESS, ORMEROD: Let let's let me put it this way: you can't be you can never be certain how future events are going to occur in relation to a model prediction. We know that the way models are created and applied. They are what we call ensemble average predictions.
…
And this is important to understand, because what it tells us is when we're predicting a certain concentration to come out of a dispersion model, we're really only predicting the most likely value in a range of outcomes which are all equally part of the same answer. And that's simply to do with the way that the atmosphere itself is turbulent and it can throw off all sorts of answers to the same question.
McQUEEN, MR: Yes.
WITNESS, ORMEROD: Right? So when we get an answer, we have to take into account that if those conditions were actually occurring, the the real measured value could be less than or higher than by a certain amount. So when we say a model could be over-predicting or under-predicting, we actually have to add in that extra consideration of the uncertainty. So which is why you can't use it as a necessarily a firm black line on the on the on the thing, because there's other things going on that we can't predict. (T:23.9-24.3; 27.9-28.2; 31.2-31.8; 05.09.14)
135 Having heard Mr Gwynne's and Mr Ormerod's evidence, the Tribunal considers that the criticisms of the GHD report and model made by Dr Kabay in the 2011 Wattleup proceeding and in this proceeding have been appropriately answered by the evidence given in this proceeding or have been shown by the evidence given in this proceeding to be misplaced.
Conclusion and recommendations
136 For the reasons set out in this report, the Tribunal considers that the 'correct and preferable decision' under s 27(2) of the SAT Act is that the applications for review should be dismissed and subdivision approval should be refused for the proposed subdivisions.
137 The Tribunal therefore recommends that the Minister should determine the review applications as follows:
1) The applications for review are dismissed.
2) The deemed refusals of the subdivision applications by the respondent are set aside and in their place decisions are substituted that subdivision approval is refused for each of the proposed subdivisions.
I certify that this and the preceding [137] paragraphs comprise the report and recommendations of the State Administrative Tribunal.
___________________________________
JUDGE D R PARRY, DEPUTY PRESIDENT
3
4
4