Environment Protection Authority v Unomedical Pty Limited

(2) oxidation. 3 Unomedical pleads not guilty to the charge, however, for the reasons given below I find Unomedical guilty of the offence as charged. Uncontroversial Background Facts 4 The background facts in this matter were largely uncontroversial and were contained, in the main, in a statement of agreed facts. They are as follows.

5 Unomedical is a company specialising in the manufacturing of single use medical instruments and equipment. This includes the operation of a sterilising facility used in the manufacture of instruments and equipment. 6 Unomedical currently occupies the premises of 11-17 Wilmette Place, Mona Vale, New South Wales ("the facility"). They have done so since about 1 January 2002. Prior to occupying the Wilmette Place facility, from approximately 1985, Unomedical (and its predecessor companies) operated a sterilising plant nearby at Bassett Street, Mona Vale. 7 In the late 1990s Unomedical decided to consolidate its operations and move to the Wilmette Place facility. As part of this process Unomedical applied for (DA 0919/00 or "the first DA"), and obtained, development consent from Pittwater Council ("the council"), the consent authority and the regulatory authority under the Protection of the Environment Operations Act 1997 ("the POEOA"). 8 Ethylene oxide ("EtO") is the principal chemical used in the sterilisation process.

9 During the charge period there was no standard or rate prescribed for the emission of EtO under New South Wales legislation and the facility was not classified as a scheduled premises under the POEOA. 10 On 10 November 2006, the council's environmental health officer, Mr Nick Ives, received a telephone call advising that EtO was being used at the facility. As a consequence, Mr Ives undertook an inspection of the facility on 17 November 2006. During the site inspection Mr Ives asked Mr Anders Kolding, then the managing director of Unomedical, to prepare a report on its EtO emissions. Mr Kolding engaged the company HLA Enviro Sciences Pty Limited ("HLA") to prepare the report. 11 In or about May or June 2007, the HLA report was submitted to the council. The council forwarded the report to the New South Wales Department of Health and the New South Wales EPA for advice. The Department of Health formed a 'health expert panel' to advise it on the health impacts of the Unomedical EtO emissions. The health expert panel comprised of: (a) Professor Bruce Armstrong;

(d) Professor Bernard Stewart. 12 These persons were chosen by the Department of Health as having expertise on the impact of carcinogenic materials, and in particular the impact of EtO, on environmental, occupational and population health. 13 On 26 July 2007, the council issued Unomedical with a prevention notice under s 96 of the POEOA ("the First Prevention Notice"). Unomedical ceased operations at the facility in response to the First Prevention Notice. 14 In July and August 2007, Unomedical installed a catalytic abatement system in the steriliser. The catalytic abatement system converted the residual EtO from the sterilisation process to carbon dioxide in water vapour or steam. It reduced the amount of EtO vented to the atmosphere at the end of each sterilisation cycle by approximately 99.9%. The purchase, freight and installation of the catalytic abatement system cost Unomedical approximately $840,000. The catalytic abatement system was sourced from a sterilisation plant operated by Unomedical in Denmark. 15 On 16 August 2007, the council issued Unomedical with a second prevention notice under s 96 of the POEOA ("the Second Prevention Notice"). 16 Prior to the council serving Unomedical with the First and Second Prevention Notices there was extensive consultation between the then Department of Environment and Climate Change ("DECC"), the council and the New South Wales Department of Health in relation to the emission of EtO. The consultation process continued after the Second Prevention Notice was served. During this process all agencies sought information, which Unomedical provided voluntarily, including expert information, which it provided at its own cost. 17 In or about September 2007, following the installation of the catalytic abatement system, Unomedical resumed operation in the facility in compliance with the Second Prevention Notice. 18 Unomedical has fully complied with the First and Second Prevention Notices. By way of letter dated 3 October 2008, the council informed Unomedical that it would not take any further action. 19 On 28 April 2008, the Protection of the Environment Operations Amendment (Scheduled Activities and Waste) Regulation came into effect ("the Regulation"). This Regulation amended Schedule 1 of the POEOA so that the use of more than one tonne of EtO per year in the sterilisation of products or equipment was a scheduled activity. Since the amendment to Schedule 1 of the POEOA, Unomedical has held, and continues to hold, an environmental protection licence for the facility in accordance with s 48 of that Act. 20 In relation to the health and environmental impact of the exposure to EtO from Unomedical's facility in Mona Vale, the health expert panel produced a consensus statement entitled "Past Community Exposures to Ethylene Oxide From Facilities in Mona Vale". The consensus was exhibited to Professor Stewart's affidavit. Professor Stewart concurred with the health expert panel's conclusion that: On balance, although there may have been a low level of cancer risk present, it is unlikely that the historical exposure to ethylene oxide from the operations of the Bassett Street and the Wilmette Place facilities have resulted in as much as one additional case of cancer among local residents. Legislative Framework Giving Rise to the Charge 21 Although the charge is framed in terms of the entirety of s 128 of the POEOA, there was no dispute that it is alleged that Unomedical breached s 128(2) only of that Act. This provision states (emphasis added): Where neither such a standard nor rate has been so prescribed, the occupier of any premises must carry on any activity, or operate any plant, in or on the premises by such practicable means as may be necessary to prevent or minimise air pollution.

22 The offence created by s 128(2) of the POEOA is one of strict liability.

23 This is the first case that deals with an alleged breach of s 128(2) of the POEOA. Its predecessor is located in s 15(2) of the Clean Air Act 1961 for scheduled premises and s 19(2) of that Act for non-scheduled premises. Overview of the Proceedings: Elements of the Offence and Defences 24 The EPA and Unomedical agreed that the elements of the offence created by s 128(2) are as follows: (a) that no standard or rate has been prescribed for the emission of an air impurity;

(b) that Unomedical was the occupier of the premises;

(c) that Unomedical carried on an activity, or operated a plant; and

(d) that the carrying on of the activity or operation of the plant by Unomedical was not by such practicable means as may have been necessary to prevent or minimise air pollution. 25 Elements (a)-(c) were admitted by Unomedical. Accordingly, I find they have been proved beyond reasonable doubt. The proceedings therefore concerned only whether sufficient proof of the last element could be demonstrated by the EPA. 26 In addition to submitting that the EPA had not proven to the requisite standard the last element of the offence, Unomedical also relied on two defences: (a) first, honest and reasonable mistake of fact. That is to say, that Unomedical, through its Managing Director, Mr Anders Kolding, held an honest and reasonable but mistaken belief of fact that it was taking all necessary measures to minimise air pollution in the operation of the facility. This is because at the time Unomedical commenced its operations at the facility, Mr Kolding held the positive but mistaken belief that the EtO which was being discharged was minimal and would not harm human health or the environment, and moreover, that no regulatory controls having been imposed on the discharge of the EtO from the facility, Unomedical was at all times complying with the law in relation to its use and emission of EtO; and

(b) second, officially induced error of law, a defence not yet recognised at common law in Australian law.

EPA's Evidence Ms Janice Davis/Taylor 27 Ms Davis was employed as an Investigator, Special Investigations Unit, at DECC. Ms Davis was the investigator, who together with Mr James Goodwin, investigated the alleged breach of s 128(2) of the POEOA. She gathered documentary information and interviewed various employees of Unomedical. In particular, she gathered information on the installation of the catalytic abatement system.

28 Ms Davis conducted an interview with Mr Kolding on 30 April 2008. Ms Davis stated that at the end of the interview with Mr Kolding he was given an opportunity to state anything further he wished to say about the matter or to make a written statement and he declined to do so. Ms Davis also stated that throughout the entire interview Ms Elizabeth Wild was present as Mr Kolding's legal representative. 29 During the course of this interview Mr Kolding relevantly stated that: (a) between 2002 to July 2007, Unomedical did not undertake any independent testing to measure the emission rates of EtO;

(b) he reported to Mr Henrik Brandt who was the CEO of Unomedical Denmark, Unomedical Australia and Unomedical "everywhere";

(c) in Australia there was an overall policy that Unomedical was to minimise any environmental impact and "always to be within legislation";

(d) he did not know what the requirements were for EtO emissions in New South Wales, but Unomedical had asked for advice from Mr Andrew Doig from the Australian Environmental Business Network ("AEBN") who advised Unomedical that because it was not a licensed company there were no applicable emissions guidelines;

(e) Unomedical got advice from the EPA in 2000 that because it was not a licensed company the amount of EtO utilised was sufficiently small that Unomedical was "of no interest to the EPA" and that it should consult with its local council. Since 2001 and up to 2007 this was the course that Unomedical pursued;

(f) because there was no legislation saying that the EtO emitted from the sterilisation process was required to be tested, no testing was conducted prior to 2007. Mr Kolding stated that "we actually didn't even know how to test it" and that even in 2006 when the council asked HLA to test the emissions, it had been difficult to find a company who could carry out the tests;

(g) prior to July 2007 the preventative measures taken by Unomedical to minimise or prevent the release of EtO into the atmosphere were that it was diluted with nitrogen prior to its release. This resulted in, Mr Kolding believed, only 50% of the EtO utilised in each sterilisation cycle being released into the atmosphere;

(h) the abatement system installed in or about 2007 was sourced from one of Unomedical's factories in Denmark. It had initially been installed in premises in Denmark in 1996 and was later recommissioned and rebuilt in 2003. It had been installed in Denmark to meet the requirements of the Danish legislation with respect to the emission of EtO into the atmosphere. These requirements applied due to the significantly larger size (ten times bigger than the facility at Wilmette Place) of the Danish factory;

(i) Unomedical used approximately six tonnes of EtO per year; (j) the EtO emissions were not referred to in the Environmental and Earth Sciences Proprietary Limited ("EES") audit report commissioned by Unomedical because they were not considered to be an issue when the report was being prepared;

(l) the EPA had provided Unomedical with advice in 2001 and 2006 about permissible EtO emissions from the facility, although he could not recall who at the EPA had provided this advice. In relation to the 2001 provision of advice, Mr Kolding stated that he had received the advice via his Operation Manager, Mr Nerinda Kanda; and

(m) the risks associated with EtO were that it had been proven to be carcinogenic to animals but that there was no proof that human exposure to it was harmful. 30 Ms Davis stated that during her site inspection of the facility on 16 May 2008, she saw sensors in place in various positions throughout the facility monitoring the release of EtO inside the facility. She saw a number of warning signs placed throughout the premises in respect of EtO. At all times Mr Kolding was cooperative with her during the site visit and that he willingly showed her around the facility and gave her any assistance required.

31 Ms Davis stated that her enquiries of EPA staff revealed that there were no records held by the EPA of any contact by Mr Kolding or Mr Kanda to the EPA in or around 2001. Mr James Goodwin 32 Mr James Goodwin was the Unit Head, Sydney Industry Section, at DECC. Mr Goodwin, along with Ms Davis, attended a meeting with representatives of the New South Wales Department of Health. During this meeting, the New South Wales Department of Health advised that it had compiled a preliminary report concerning its investigation into community exposure to untreated EtO from Unomedical's facility at Wilmette Place.

33 Mr Goodwin obtained the following information from the council, namely, that: (a) DA 919/00 was lodged by Maersk Medical Pty Ltd ("Maersk") for a change of use to a medical manufacturing and warehouse facility at Wilmette Place from Bassett Street, to council on 19 September 2000. The description of the proposal was for a: small addition (960 m 2 ) to an existing 5,800 m 2 facility. Additional car parking, landscaping, building entry improvements and refurbishment of all office space. For use as a medical manufacturing site (see Statement of Environment Effects) The application form stated that, "Council's staff will assess your proposal and contact you should any additional information be required". The Statement of Environmental Effects contained a description of the "Proposed Development" which read: sterilisation of the manufactured medical products is carried out using ethylene oxide by fully automated process control two cycles run each day. The process is validated and conforms to EN550/ISO11135 with auditing by the Department of Health for licencing and BVQI for CE marking on a yearly basis. Ethylene oxide is classified under the Dangerous Goods Act, & Maersk is licensed under the Act to carry 500 kg with Lic No 35/007026. No noise is involved and no odours generated or released.

(b) on 13 March 2001, the council granted development consent in respect of the first DA. The Statement of Environmental Effects was attached to the consent;

(c) on 13 August 2001, Maersk applied to modify the consent;

(d) the council approved the modification application on 5 October 2001. An identical Statement of Environmental Effects was attached to the approval, together with documents disclosing the location of the EtO gas tanks enclosure.

(e) on 20 October 2004, a second development application was lodged by BOS Constructions Pty Ltd ("BOS") for a gas bottle enclosure extension at the facility ("DA 776/04" or "the second DA"). A new Statement of Environmental Effects attached to the consent made it clear that the application was for an extension of an existing ethylene oxide gas bottle enclosure at the facility. An email from Mr Mark Blake at BOS to the council on 9 February 2005 attached various documents in support of the application. These included a chemical fact sheet stating that most of the ethylene oxide used in sterilisation entered the environment, but once in the atmosphere it would take approximately one week to degrade or to return to earth in rain. The development control plan ( My Development Control Plan - 11 Wilmette Place, Mona Vale ) attached to the application (at C3.7), however, that the development did not adversely impact upon "health, the environment or other lands". It stated that the development must be designed, constructed, maintained and operated in a proper and efficient manner so as to prevent air pollution. It also stated that the development and business operation "must comply with the" POEOA. At the foot of that section of the plan the word "ok" had been written. It is not known by whom; and

(f) the council granted consent to the second DA on 7 March 2005. 34 Mr Goodwin stated under cross examination that local councils had the primary responsibility as the regulatory authority under the POEOA of enforcing the provisions of that Act in relation to non-scheduled industries such as Unomedical's facility, which included conducting environmental audits of these premises. He agreed that if any local council had any particular concerns in relation to any industries falling within its regulatory jurisdiction it could seek advice from the EPA. 35 Mr Goodwin confirmed that prior to Unomedical becoming a scheduled premises there were no legislative enforceable air emission limits for EtO that applied to it.

Ms Amiette Wakenshaw 36 Ms Wakenshaw was a specialist investigator with DECC. Ms Wakenshaw assisted in the investigation into the reports of air pollution from the facility together with Ms Janice Davis and Mr James Goodwin. 37 In particular, Ms Wakenshaw received information and records from Unomedical pursuant to a notice served on it under s 212A of the POEOA. Amongst the material provided by Unomedical in response to the notice was the following: (a) a document entitled 2007 OHS Hazard Register - Steriliser Plant Post Installation of Air Purification System published by Unomedical ("the 2007 Register"). The 2007 Register noted that EtO was classified as a hazardous and dangerous good 2.3. It stated that the legislative requirements, recommendations and guidelines for risk assessment and control of EtO were set out in the documents, 1992 National Code of Practice for the Safe Use of EtO in Sterilisation/Fumigation Process NOHCS:2008(1992) ("the National Code") and the Guidance Note for the Safe Use of EtO in Sterilisation/Fumigation Process NOHSC:3016(1992). These documents stated that EtO was a category 1 carcinogen according to the International Agency for the Research of Cancer. This meant that EtO was a human carcinogen and human exposure might result in the development of cancer;

(b) a BOC material safety data sheet dated 7 June 2005 for "Fumigas 1000 Fumigant" ("the MSDS"), the trade name of the EtO sterilant used by Unomedical, which stated that the gas used in the sterilisation process comprising 99.7% EtO, may irritate eyes, respiratory system and skin, may cause cancer and may cause heriditary genetic damage. The health hazard summary for the gas stated that it was "moderately to highly toxic". The summary went on to state that the symptoms were usually delayed. It stated that EtO was classified as carcinogenic to humans and that special approval for use was required by the Health Department under the poisons legislation. In terms of the ecological information available it stated that "limited ecotoxicity data was available for this product at the time this report was prepared. Ensure appropriate measures are taken to prevent this product from entering the environment"; and

(c) an application for renewal of licence to keep dangerous goods dated 5 December 2003, signed on behalf of Unomedical by Mr Kanda which listed one of the dangerous goods to be 400 kg of EtO (class 2.3). Professor Bernard Stewart 38 In his affidavit sworn 11 September 2008, Professor Stewart stated that he was a medical scientist employed by South Eastern Sydney and Illawarra Area Health Service.

39 In July 2007 he was invited by the Director of the Environmental Health Branch of the New South Wales Health Department to sit on the health expert panel established to assess the effects of the discharge of EtO from the facility. In particular, the purpose of the expert panel was to assess the impacts on human health of EtO emissions. 40 On 5 September 2007, the expert panel considered a report entitled Pittwater Local Government Area Cancer Incident Study 1996 to 2005 produced by the Northern Sydney Central Coast Public Health Unit. The expert panel concurred with the report's conclusions that there "is no evidence of increased cancer rates in the Mona Vale postcode for the period 1996 to 2005". 41 The expert panel convened in meetings on 11 February and 18 March 2008. At these meetings the expert panel reviewed the report entitled Retrospective Air Quality Impact Assessment Unomedical Sterilisation Process, Mona Vale, New South Wales produced by HLA, and considered that "the annual average levels of Ethylene Oxide estimated in the period 1986 to 2007 were suitable for assessing past human health risk". 42 On 18 March 2008, the expert panel also considered a document entitled Estimate of Additional Cancers due to Ethylene Oxide Exposure at Mona Vale. The document was produced by the Northern Sydney Central Coast Public Health Unit at the request of the expert panel. 43 On 9 April 2008, the expert panel produced a document summarising from the expert panel's work, entitled Past Community Exposure to Ethylene Oxide in facilities in Mona Vale. The document was a consensus statement agreed to by the entire panel. 44 Professor Stewart stated that he considered that the document was "an accurate summary of the expert panel's work. I also concur with the expert panel's conclusion that, 'on balance, although there may have been a low level of cancer risk present, it is unlikely that the historical exposure to Ethylene Oxide from the operations of the Bassett Street and Willmette Place facilities has resulted in as much as one additional case of cancer among local residents'". 45 However, Professor Stewart further stated in his affidavit that: Ethylene oxide is categorized by International Agency for Research on Cancer (IARC: an arm of the World Health Organization) as being a cancer-causing hazard in common with tobacco smoke, asbestos and benzene: all four agents are included in the current listing of IARC Group 1 carcinogens. Pollution of the workplace or general environment by these agents and similarly categorized substances is properly addressed from the perspective of preventing exposed persons being at risk. Scientific data particular to ethylene oxide limit specification of quantitative risk - the actual number of cancer cases anticipated to result from a particular circumstance or level of exposure - by comparison with some other carcinogens. This consideration does not qualify or reduce the imperative to ensure that preventable human exposure to ethylene oxide does not occur by the adoption of any practicable procedure(s) known to eliminate or markedly reduce such exposure. 46 Having said this, in cross examination Professor Stewart agreed that: there is a lack of definitive evidence using the word to mean unequivocal proof of the type associated with say asbestos or tobacco smoke as causing cancer. There is not, however, a complete lack of evidence as far as studies have been done and there is what might reasonably be characterised as inference of increased risk but it is not definitive. 47 Professor Stewart also agreed that all the studies were concerned with the occupational exposure to EtO and centred around employees in industrial premises where EtO was being used primarily in the sterilisation process of medical instruments. Moreover, the studies concerned exposure over a number of years in the workplace at "the highest levels that humans are exposed" to EtO. Professor Stewart noted that the highest levels of exposure were in the workplace and levels of exposure "metaphorically speaking beyond the facility fence are invariably less". 48 During the cross examination of Professor Stewart the following exchange took place: Q. And it would it be fair to say so much so that is that environmental exposure is so markedly less than an industrial occupational situation that no one has ever attempted to demonstrate any environmental causation of cancer? A. By ethylene oxide specifically as distinct--

Q. By ethylene oxide? A. Yes, that's correct. And the fact that, as you have correctly in my view cited the IRC study, there is inference of increased risk of cancer in the workers but the data are not definitive. And granted that scenario it would not be rational to attempt to detect cancer following lower levels of exposure the rational thing is to keep on monitoring where records are available, the relevant exposed workforce and that would be the focus of all studies. It would almost be impossible in my view to get funding for a study involving environmental exposure.

Q. So is it fair to say Professor, where we stand today in relation to studies there's simply no evidence that environmental exposure to ethylene oxide has caused cancers in humans? A. There's no, there's no evidence in the direct sense, ..(not transcribable).. study or there's no study as we just -as your evidence just adduced, no studies have been undertaken therefore there's no evidence in humans of that result.

Q. Even without the criminal standard you can't say, can you Professor, that the emissions of EtO from Wilmette Place caused any increase in the rates of cancer in the northern beaches area? A. No, I have not asserted that at and I can't say that. 49 In re-examination, however, Professor Stewart said: So ethylene oxide, being a material that's known to cause tumours in experimental animals, but despite studies is not definitively known to cause tumours in humans, would only be studied further in humans in high dose situations even though in public health policy it's undesirable for people to be exposed to the material. Q. Why do you add that last qualifier about it being undesirable? A. Because all public health approaches to carcinogen exposure involve the caveat that there should be no preventable exposure to a human carcinogen because there is no evidence, biological evidence of a safe level of exposure to such materials. So even though populations exposed may not provide epidermalogical evidence of increased risk of cancer, public health policy mandates that people be not exposed to jointly preventable exposure of a carcinogen. Mr Nicholas Ives 50 In his affidavit affirmed 4 July 2008, Mr Ives stated that he was the Environmental Health Officer for the council during the charge period. His duties as Environmental Health Officer included conducting environmental assessments of industries to ensure compliance, conducting routine inspections of industrial premises and responding to complaints. As part of his duties he was responsible for exercising the provisions of the, amongst other legislation, POEOA. 51 It was Mr Ives who, on 10 November 2006, received a telephone call from an unidentified male person who stated "Unomedical is using Ethylene Oxide with the gas being pumped into the air. EtO is hazardous and can pose serious health risks". 52 On 17 November 2006, Mr Ives attended the premises at Unomedical for the purpose of conducting a site inspection. He was shown throughout the facility by Mr Anders Kolding. During the inspection he saw EtO monitors mounted on the outside walls of the sterilisation chamber and he observed warning signs that EtO was in use. He was informed by Mr Kolding that the monitors and signage were required by WorkCover. 53 While viewing the sterilisation chambers, Mr Ives had a conversation with Mr Kolding about the sterilisation process. Mr Kolding disclosed to Mr Ives the following information: (a) that the steriliser was run on average about one and a half times each day;

(b) that about 15 to 20 kg of EtO was used in the sterilisation process and released into the air;

(c) that the release of EtO was not monitored outside the facility;

(d) that the discharge was, according to Mr Kolding, "safe" because when Unomedical first moved to the site it engaged an environmental consultant to ensure its methods were satisfactory. This culminated in a report by the consultant which Mr Kolding stated that he would send to Mr Ives; and

(e) that another sterilisation process not requiring the use EtO could be employed but that Unomedical was not set up for it and that it tended to discolour the product.

54 Mr Ives subsequently drafted a Environmental Compliance Inspection report dated 22 November 2006. The report stated: Inspection result: ethylene oxide is used in steralisation (sic) process. There does not appear to be any monitoring. The company have an environmental consultant who advised the meathod (sic) is satisfactory. After discussion with DEC there are levels that should not be exceeded and Council will require Unomedical to show that the use of Ethylene Oxide meets the criteria. 55 On 6 June 2007, Mr Ives attended a meeting at Unomedical, the purpose of the meeting was for Unomedical to present the findings of the emission testing conducted by HLA. 56 As a result of the findings by HLA and information received from DECC which specified the standard EtO when measured in ambient air should not exceed 0.0033 mg/m3 (or 3.3 ug/m3), Mr Ives issued the First Prevention Notice. The First Prevention Notice prevented Unomedical from utilising EtO until they were able to produce documented evidence they had complied with the accepted standard of 0.0033 mg/m3. As a result, Unomedical was required to temporarily cease sterilisation activities at the facility, which it did.

57 In relation to the earlier audit report that Mr Kolding had referred to during the site inspection, Mr Ives discovered that council records indicated that the report was completed by EES for Unomedical in November 2002. A copy of the report was not on council files. Between 14 to 17 August 2007 Mr Ives sent a number of emails to Mr Kolding requesting a copy of the EES report. It was never received. 58 Under cross examination, Mr Ives stated that he had observed monitors at various places inside the facility, which Mr Kolding had explained to him were to ensure that the levels of EtO inside the premises met occupational health and safety standards. Mr Ives stated that he had observed the premises to be spotless and efficiently operated and that it was clear to anyone inspecting the premises that EtO was being used as part of the industrial process carried out at the facility. 59 Mr Ives also stated that Mr Kolding had told him that EtO was emitted through the stack at the top of the roof and that he showed Mr Ives the location of the stack. Mr Ives stated that the stack could easily be seen standing beside the facility, because it projected approximately 6 to 8 m above the roofline. However, Mr Ives said that in his opinion to the untrained eye it would not be known that the stack was a discharge point for EtO.

Mr Andrew Doig 60 In his affidavit sworn 23 June 2009, Mr Doig stated that he was the Director of AEBN. On 31 May and 1 June 2006, AEBN held a workshop entitled Environment Management Practitioners' Workshop. One of the attendees of that seminar was an employee of Unomedical, namely, Ms Annegrete Kolding. 61 After the conference, Mr Doig received a telephone call from Mr Paul Orton of Unomedical. Mr Orton requested that AEBN provide Unomedical with some general environmental information in relation to its sterilisation operation. Mr Doig agreed to the request. 62 Over the course of a number of months Mr Doig and Mr Kolding exchanged various communications by way of emails and telephone calls. The subject matter of these communications was the use of EtO at the facility. 63 In or around November 2006, and at Mr Kolding's invitation, Mr Doig took a tour of Unomedical's operations at the facility. During the tour, Mr Doig had a discussion with Mr Kolding in which Mr Kolding sought further information concerning permissible EtO emission limits. Mr Doig was also asked to provide general information on the environmental impact of EtO. Mr Doig stated that the discussions were informal in nature. 64 After the tour Mr Kolding and Mr Doig had a telephone conversation to the following effect: AD Hello Anders, it's Andrew Doig.

AD I've looked at the Protection of the Environment Operations (Clean Air) Regulation 2002 and there are no limits or references associated with ethylene oxide. But there is a reference to ethylene oxide in a document called the Approved Methods for the Modelling and Assessment of Air Pollutants in NSW. While the document doesn't contain any stack limits for ethylene oxide, it does contain ground level concentrations. However, it is largely used during development assessment for premises that are regulated by the EPA. Based on the facts you told me that Unomedical is already operating under a valid development consent and the site isn't regulated by the EPA because, it's not a type of industry listed in Schedule 1 of the Protection of the Environment Operations Act 1997 and the Approved methods only recently included ethylene oxide, it appears it does not apply directly to the site for setting stack emissions limits. AK Thank you for looking into that for me. It's much appreciated.

AD Bye. 65 Mr Doig subsequently referred Mr Kolding to an air pollution consultant for advice. 66 Under cross examination, Mr Doig stated that he had told Mr Kolding that it was his understanding that because the facility was not scheduled there were no stack limits in relation to the emission of EtO.

Ms Angela Daly 67 In her affidavit sworn 3 July 2009, Ms Daly states that from August 2002 until January 2003, she was employed by the council as an Environmental Health Officer. During this period she carried out a number of environmental audits of industrial and commercial premises in the council local government area. These audit inspections were of a "general nature" in as much as "the primary focus was on whether the goods were properly stored, whether stormwater drains were adequately protected from any discharges of foreign materials, whether any waste materials were being dealt with properly". Ms Daly said that she did not undertake any desktop evaluation of premises before undertaking the inspections, rather she simply relied on what she saw and was told during the course of the inspection. 68 One of the inspections carried out by Ms Daly was of the facility. During the inspection she was accompanied by Mr Kanda from Unomedical. Whilst on the inspection, Ms Daly completed an inspection sheet based on her observations and on what she was told. Ms Daly stated that during the course of the inspection she did not see any visible air emissions such as steam or smoke. Ms Daly stated that she was unfamiliar with the processes carried on at the facility. 69 Ms Daly stated that whilst she saw material safety data sheets ("MSDS") for chemicals used in the maintenance of the cooling tower she did not see an MSDS for the goods stored at the premises, including EtO. 70 Ms Daly stated that she had no recollection of anyone telling her at the time of the inspection, or any time thereafter, that EtO was discharged into the air. Ms Daly stated that whilst she was aware that EtO was stored at the facility and used in the process of sterilising medical devices, she did not know how it was used and was "very surprised when I read in the local press in 2007 that EtO was, according to those reports, being discharged to the atmosphere from the premises".

71 The Pittwater Council Environmental Audit Inspection report completed by Ms Daly stated as follows: Air Quality Management No evidence of air emissions (pollution control devices etc not relevant). … Recommendations No recommendations

72 The report was sent to Maersk in Denmark on 29 November 2002. She was not aware of any response from Maersk or Unomedical. 73 On the same day, Ms Daly sent to Mr Kanda a copy of a summary report of conditions observed by her at the time of inspection of premises on 22 November 2002. The letter attaching the summary report referred to having been allowed to read the report prepared by EES in May 2002. The letter stated that if Mr Kanda had any enquiries regarding the matters raised in the audit that he should contact her. No contact was forthcoming. 74 Ms Daly stated under cross examination that at the time she worked with the council she knew that the council was the appropriate regulatory authority for non-scheduled industrial premises (of which Unomedical was one) within the council's jurisdiction. Ms Daly also agreed that the audits that she conducted for the council were aimed at meeting the objectives set out in s 172 of the POEOA.

75 Ms Daly also agreed that the audits she conducted were primarily concerned with the storage of waste to protect the stormwater system and water quality. Thus they were very general audits of no more than half an hour to 40 minutes duration and could not be considered to be comprehensive. Mr Peter Ellis 76 Mr Peter Ellis, who was not required for cross examination, stated in his affidavit affirmed 18 June 2008, that on 25 May 2007 his company, Peter J Ellis and Associates Pty Ltd, prepared a proposal for Unomedical to increase the overall efficiency and capacity of its sterilisation process based on discussions with Unomedical's plant engineer, Mr Hop Lam. 77 Subsequently, Mr Ellis was asked by Mr Ross Halfacree and Mr Brandon Buhler on behalf of Unomedical if he was able to facilitate the installation of an abatement system to the steriliser. Mr Halfacree advised Mr Ellis that he had ordered the abatement system (consisting of a balancer and catalytic converter) from overseas. The function of the abatement system was to treat the EtO prior to it entering the atmosphere following sterilisation. This process involved the EtO reacting with a catalyst and being converted into carbon dioxide and steam prior to it entering the atmosphere.

78 Mr Ellis stated that prior to the installation of the abatement system, there was no external device which treated the EtO before it was released into the atmosphere. Rather, Unomedical's process for disbursing EtO was to dilute it with nitrogen and at the conclusion of the sterilisation cycle to pump the gas through a pipe which went from the steriliser directly through the ceiling roof and into the atmosphere. 79 The installation of the abatement system, which Unomedical imported from Denmark, took approximately one month to complete, including testing. During the testing phase the emission levels of EtO were measured by HLA to ensure that Unomedical complied with air quality requirements. Mr John Court Use of Overseas Literature and Learning

80 The EPA relied upon expert reports by Mr John Court, an environmental engineering consultant with expertise and experience in the investigation, assessment and control of environmental pollution.

81 Attached to his reports were various overseas learning and academic articles on the adoption of controls for the emission of EtO from commercial sterilisers. The material came principally from Canada, the United States and Denmark.

82 In his oral evidence, Mr Court identified which parts of the literature and learning he had relied upon and had adopted thus forming either directly or indirectly the basis of his expert report.

83 Unomedical objected to the admission of this overseas literature and learning on the basis of relevance, first, because the commission of the offence occurred in Australia, and second, because Unomedical had conceded that it could have (and subsequently did) implement practical means to limit the EtO emissions, but that in its opinion it was not necessary to do so. Second, it objected to the material on the basis that it was hearsay. Third, and in the alternative, Unomedical sought to exclude the material pursuant to either ss 135 or 137 of the Evidence Act 1995 on the basis of its prejudicial nature.

84 In PQ v Australian Red Cross Society [1992] 1 VR 19 the plaintiff claimed damages for a HIV infection received from an infusion of blood products administered in the treatment of his haemophilia. Claims in negligence and under the Trade Practices Act 1974 (Cth) were made. In a trial before a jury, an issue arose as to the evidential use that could be made of information in authoritative scientific publications and the data relied upon by scientific experts in forming opinions and making decisions. McGarvie J held that articles that had been discussed and analysed could not be admitted as exhibits because the publication itself was hearsay. That is to say, the publication itself was not evidence of the truth of statements it made as to the information or data it contained. If, however, the expert adopted the publication referred to or quoted, then it became part of the testimony of the expert witness and therefore admissible (at 35-36).

85 However, his Honour went on to hold that the publications could be admissible as exhibits if the contents were relevant to the issue of what information was available at a particular time and/or how reliable the information was (at 36-37). An important issue in that case was what the defendants knew or ought to have known about the risks of persons becoming HIV+ through the use of blood products and the available means of eliminating such risks.

86 The decision in PQ has been subsequently followed in a number of Australian jurisdictions (see, for example, Fazio v R (1997) 69 SASR 54 at [41]; Woods v The Director of Public Prosecutions (2008) 38 WAR 217 at [53]-[61]; Workcover Authority of NSW (Inspector Simpson) v CNH Australia Pty Ltd (formerly Case Corporation Pty Ltd) [2006] NSWIRComm 220 at [89], [106], [109], [110] and [111] - [116]; R v Jung [2006] NSWSC 658 at [60]; Street v Luna Park Sydney Pty Ltd [2007] NSWSC 1529 at [7]; Alphapharm Pty Ltd v H Lundbeck A/S (2008) 76 IPR 618 at [698]-[783]; Bodney v Bennell (2008) 167 FCR 84 at [92]; J D Heydon, Cross on Evidence, Vol 1, (1996) Butterworths at [29150] and I Freckelton SC and H Selby, Expert Evidence Law, Practice, Procedure and Advocacy (4th ed), (2009) Lawbook Co, p 129).

87 Accordingly, the overseas material identified by Mr Court as having been adopted and relied upon by him in the formulation of his reports was admitted into evidence without limitation on the basis that, first, it was relevant in determining the ambit of the term "practicable means" under s 128(2) of the POEOA, second, it was relevant to determining what knowledge Unomedical ought to have had during the charge period as to what were the practicable means available to prevent or minimise air pollution (leaving aside the question of whether these means were necessary), and third, to assess the accessibility of this material.

88 Unomedical took further objection to the literature and learning identified by Mr Court that had not been relied upon or adopted by him in stating his expert opinion. The EPA submitted that the purpose of this tender was not to rely on the truth of the contents of these publications, rather the material demonstrated the knowledge Unomedical ought to have had during the charge period of the practicable means available to it to prevent or minimise air pollution.

89 Given this stated purpose, this latter category of literature and learning was admitted subject to a limitation pursuant to s 136 of the Evidence Act 1995 that the material was evidence only of the fact of their contents and not of the truth of the facts and opinions contained therein (PQ at 36-37 and R v Karger (2001) 83 SASR 1 at [71]-[72]). As it transpired, in light of the evidence of Mr Kolding, very limited weight was required to be attributed to this latter category of publications.

90 With respect to domestic EtO emissions where that product was used as a sterilant: (a) in the Approved Methods and Guidance for the Modelling and Assessment of Air Pollutants in New South Wales (EPANSW 2001) ("the 2001 Approved Methods document") an acceptable design ground-level concentration of EtO from any source after dispersion was specified to be 0.05 mg/m 3 (based on a 3-minute averaging period);

(b) under the 2005 Approved Methods document the design ground-level concentration could not be greater than 0.0033 mg/m 3 and a requirement was introduced to control emissions of EtO to the "maximum extent achievable";

(c) in Victoria, the design ground-level concentration must be 3.0 mg/m 3 and EtO emissions had to be reduced to the "maximum extent achievable"; and

(d) at a Commonwealth level, the National Code stated that EtO was a category 2 carcinogen which meant that it was a "probable" human carcinogen and should be treated as carcinogenic to humans and "handled with great caution". The Code went on to state that the absence of a relevant exposure standard or compliance with a relevant exposure standard should not preclude further efforts to reduce exposure. Exposure should be controlled to the "lowest practicable level". The Code recommended that exhaust gases from a steriliser should be scrubbed or consideration should be given to a catalytic converter or incinerator.

91 In respect of EtO emissions where the product is used as a commercial sterilant, what the literature and learning disclosed in relation to overseas practice was equivocal. For example, as at 2001 (Environment Canada, "Summary of Technical and Socio-Economic Background Study for Ethylene Oxide Sterilisation Applications", Final Report, January 2001, Cheminfo Services Inc): (a) approximately 71% of the annual potential EtO emissions from its use as a sterilant in Canada were destroyed by reason of the use of an acid-water scrubber, thermal oxidation or a catalytic converter;

(b) there were no harmonised European emission standards for EtO used as a sterilant;

(c) the United Kingdom, France, Spain, Portugal, Italy and Greece had no emission standards in place for EtO;

(d) the Netherlands had an EtO stack concentration standard of 2 mg/m 3 ; and

(e) Germany, Belgium, Ireland, Denmark and the Czech Republic had an EtO stack concentration standard of 5 mg/m 3 .

92 In the United States, a 99% elimination of EtO emissions from a normal steriliser cycle was required. Use of the Approved Methods Documents 93 In giving his opinion, Mr Court relied upon the Approved Methods documents (2001 and 2005 versions).

94 Again, Unomedical submitted that these documents were inadmissible because they were only relevant to determine emission standards for the purpose of scheduled premises under s 128(1). That is to say, there was no reference in the documents to non scheduled premises, and therefore, they had no application to the offence as charged under s 128(2).

95 The EPA pressed for the admission of these documents notwithstanding that the facility was not scheduled and that no prescribed emission limits or standards applied for the purposes of the POEOA. Rather, the EPA relied upon oral evidence given by Mr Court that he regularly had regard to these documents, and used them to assess the emission controls of non-scheduled premises. It was Mr Court's evidence that in his experience, including in his experience dealing with local councils, the documents were commonly used as a reference resource for both scheduled and non-scheduled premises. However, Mr Court conceded that the Approved Methods documents were not widely available and, in particular, that they had not been introduced in 2000 when the first DA was lodged with the council by Unomedical for use of the facility.

96 The Approved Methods documents were admitted because: (a) while clearly most relevant to the offence created under s 128(1), this did not of itself preclude any consideration of them in respect of the offence as charged under s 128(2); and (b) of the oral evidence from Mr Court to the effect that he was aware that the Approved Methods documents did not apply to non-scheduled premises but that in his experience the documents were commonly used to assess not only scheduled premises but non-scheduled premises also. 97 Having said this, the weight to be accorded to the Approved Documents was tempered given the limitations identified above surrounding their direct application to the facility.

98 In his report dated 27 June 2008, Mr Court opined that: (a) first, equipment for effectively controlling EtO emissions had been developed in North America and Europe. Either scrubbing with water or acid, oxidation by direct flame or by using a catalytic converter were all well established techniques for EtO control, and moreover, were available for application in commercial sterilisation as at 2000;

(b) second, regulatory controls limiting the amount of EtO that could be emitted into the atmosphere were introduced in the 1990s in the United States and some European countries. Had Unomedical been operating in these countries it would have been compelled to significantly reduce its EtO emissions. Initial air quality modelling by HLA in May 2007 indicated that the EtO emissions from the facility were well in excess of the ground level concentration of 3.3 ug/m 3 . In fact subsequent modelling by HLA predicted that the ground level concentration of EtO was 8.3 ug/m 3 ;

(c) third, while there were no emission limits in New South Wales that were applicable to the facility, the 2005 Approved Methods document assessment requirements for scheduled premises mandated a 3.3 ug/m 3 ground level emission concentration with the additional requirement that emissions at the source be controlled to the maximum extent achievable using "practical means" before dispersion could be considered. Unomedical's emissions would not have complied with the 2005 requirements; (d) fourth, as a result of its operations in Denmark, Unomedical ought to have been aware of the need to stringently control emissions of EtO resulting from its sterilisation processes;

(e) fifth, at a minimum Unomedical should have undertaken an environmental assessment of air quality impacts from the facility much earlier that it did; and

(f) sixth, the subsequent installation in 2007 of an abatement system effectively eliminating the EtO emissions from the steriliser was a demonstration of the "practical means" that Unomedical could have implemented earlier.

99 In his report dated 4 June 2009, Mr Court described seven options for control of EtO emissions from commercial sterilisers. These ranged from absorption, oxidation, condensation and reaction with solids. Further, after surveying the regulations in Denmark and the United States from 2002 to 2007, Mr Court concluded that the size of the operation of the steriliser at the facility during 2002 to 2006 would have necessitated the installation of emission reduction equipment to reduce EtO discharges had it been operating in either of those two countries. 100 To the extent that it was relevant, it was Mr Court's opinion that the words "practicable means" meant minimising pollution as far as possible by using available technology. This, Mr Court acknowledged, meant striking a balance between what was desirable and what was practically and commercially achievable. It did not mean employing the best available technology irrespective of the commercial implications. In this regard, Mr Court readily agreed that there was no abatement system that could be utilised at the facility that would achieve zero emission levels without significantly changing the method of sterilisation. 101 With respect to the term "as necessary", again to the extent this evidence was relevant, it was Mr Court's opinion that given the toxicity of EtO, and consistent with the approach adopted to air pollution management since the mid 1990s in Europe and the United States, this meant to control emissions to the maximum possible extent.

102 Mr Court did not agree with Unomedical that an adequate method of minimising EtO emissions to the atmosphere was to emit them from a tall stack. This was because stack emission did not minimise EtO emission from the sterilisation process, instead it minimised exposure because the height of the stack ensured a greater dispersion of the chemical.

103 Mr Court stated that the amount of additional beneficial dispersion obtained from increasing the velocity of the emissions of EtO from the stack was, in his view, limited. Mr Court conceded, however, that he gave this opinion absent any testing of disbursement velocities from the facility. 104 In terms of reducing emissions by using only a full load for sterilisation, Mr Court stated that while this reduced the overall amount of EtO in use over time, per load any decrease in the rate of EtO emissions would be minimal. 105 Finally, with respect to flushing, Mr Court stated that, in his opinion, this had no impact on the amount of EtO emitted because the same amount of EtO was emitted only in a more diluted form.

106 Unomedical sought to discredit Mr Court on the basis that he lacked sufficient practical expertise or experience with medical sterilisation processes, and in particular sterilisation processes using EtO. I reject this criticism. While it is true that Mr Court did not have any direct experience with the use of EtO in the course of the manufacture and sterilisation of medical instruments, he nevertheless had nearly four decades of experience and training as a chemical engineer, and in particular, in analysis of emissions resulting from industrial processes. I therefore do not accept the submission by Unomedical that Mr Court was unqualified to give the opinions he did. Having said this, Mr Court's lack of direct experience with sterilisation processes using EtO must nevertheless be taken into account when assessing his evidence, albeit, in my opinion, to a limited degree. Unomedical's Evidence Mr Anders Kolding 107 In light of the ultimate conclusion that I have reached concerning Unomedical's guilt, as a matter of fairness to Unomedical I have set out in considerable detail (more than would otherwise be necessary or appropriate) the oral evidence of Mr Kolding, Unomedical's only witness in the proceedings. 108 As referred to above, Mr Anders Kolding was the managing director of Unomedical during the charge period. Mr Kolding commenced his employment with Unomedical on 1 September 1980. He ceased his employment with the company around 31 March 2009. 109 At the time he commenced employment with Unomedical, Mr Kolding was located in Denmark. At that time Unomedical was known as Pharma Plast. The company manufactured medical devices and was owned by the shipping group Maersk. It later changed its name to Maersk Medical, then Unomedical.

110 Mr Kolding stated that he had worked for Unomedical for approximately 30 years. As a consequence, he had considerable experience with EtO and he knew that it was regarded as a carcinogen. After being promoted to plant manager (from production assistant) Mr Kolding became responsible for a plant outside of Copenhagen, which did not have an on-site steriliser. Later he was promoted to general manager for the same site and he then assumed the position of managing director of Unomedical Australia on or about 25 June 1996. As managing director he was responsible for the entire facility, namely, manufacturing, sales, marketing, distribution and its general operation. 111 Mr Kolding stated that Unomedical typically manufactured single use disposable medical products. This included catheters, urine drainage bags, wound drainage products, custom procedure packs, and pre-prepared packs for all kinds of operations. Most of the products were sterilised on site at Unomedical. The products were supplied to hospitals across Australia and New Zealand. The products were generally supplied weekly or fortnightly. 112 Mr Kolding stated that in the Australia and New Zealand market Unomedical was considered to be a dominant operator with considerable market share. Typically, Unomedical serviced approximately 20 million people in Australia and 5 million in New Zealand. By contrast, Unomedical's European plants serviced 600 to 700 million across Europe, the United States and the Asian region. Consequently the sterilisers used by Unomedical Denmark were considerably larger. Mr Kolding's estimate was that in Denmark, Unomedical sterilised up to 10 times more products than at the facility in Australia. 113 When Mr Kolding commenced as managing director of Unomedical Australia in 1996, the plant operated next door to the current facility at Wilmette Place. In February 2002, the company relocated to Wilmette Place because Unomedical needed larger premises. Thus the two sites merged. 114 Unomedical engaged Mr Ross Halfacree, a consultant, to assist in the preparation of the first DA required for the relocation. Mr Kolding stated that he had seen the development application at the time it was lodged with the council. 115 The first DA clearly stated in its description of the proposed development, that there would be sterilisation of manufactured medical products by a fully automated process control using EtO. The application stated that EtO was classified under the Dangerous Goods Act 1975. Mr Kolding conceded that the first DA did not, however, state that EtO would be emitted into the atmosphere as a consequence of the sterilisation process.

116 But Mr Kolding stated that if such information had been asked for it would have been provided. Mr Kolding said that at the time of the relocation he had assumed that the council knew about the EtO emissions because Unomedical had been operating for the past 15 to 20 years in adjacent premises. However, Mr Kolding accepted that no plan was attached to the first DA that showed an emission stack. This was because, Mr Kolding stated, Unomedical was not required to show a stack on the plans given that the application was for a relocation and continuation of Unomedical's existing operations. 117 Mr Kolding stated that prior to submitting the first DA, Unomedical had consulted with the council and had been advised by Mr Nick Sutton, the council planner, that because Unomedical had been operating a sterilisation plant for the past decade, existing approvals could be moved across to the new site and that nothing other than a Statement of Environmental Effects was required. 118 Mr Kolding also stated that before the fully automated steriliser was installed at the facility, Unomedical had checked with the EPA as to what, if any, approvals were required for the operation of the steriliser. Mr Kolding said that he had asked his operations manager, Mr Kanda, to ensure that all necessary approvals were obtained. Mr Kanda had informed Mr Kolding that he had been advised by the council that Unomedical could continue operating under the same approvals that Unomedical had previously had. That is to say, Unomedical was not required to limit the amount of EtO emissions from the facility at Wilmette Place.

119 Further, Mr Kanda told Mr Kolding that he had spoken with the EPA and had been advised by it to ask the council, as the regulatory authority, what, if any, limits there were on the emission of EtO from the facility. It was as a result of this conversation that Mr Kolding concluded that no catalytic converter needed to be installed at the facility at this point in time. 120 Mr Kolding stated that had the council or the EPA required Unomedical to conduct emissions testing or to install an abatement system to reduce emissions, Unomedical would have done so.

121 Mr Kolding stated that the advice given to Mr Kanda, relayed to him via Mr Kanda, was that "10 tonnes [of EtO] was the critical number, as long as you were below 10 tonnes there was no requirement for you to put any controls in and you will be regulated by the council and Unomedical back at that time only used about 5.5 tonnes per year". Mr Kolding stated that in 2006 the maximum tonnage of EtO used by Unomedical was about 9 tonnes.

122 Under cross examination, Mr Kolding accepted that in his record of interview he had made no mention of any of the conversations between himself and Mr Kanda concerning the advice given to Mr Kanda by the EPA nor had he made any mention of the 10 tonne threshold. Mr Kolding accepted that there was no record of the EPA having had any of the conversations with Mr Kanda as alleged.

123 Mr Kolding stated that his belief during the charge period that the level of EtO that was being emitted from the stack never exceeded a level which could cause harm to persons outside the facility, was based on, first, his experience of EtO emissions in Denmark, and second, on what his operations manager, Mr Kanda, who had been working at Unomedical for a decade had told him. 124 Mr Kolding stated that the May 2002 EES environmental due diligence report had not been received by him, but by the head office in Denmark. It had been prepared by Unomedical answering a questionnaire probably, Mr Kolding said, completed by Mr Kanda, but checked by himself. He recalled that the questionnaire, the completed version of which could not be found, dealt with the issue of the EtO emissions.

125 Under cross examination Mr Kolding accepted that no irregularities were identified by EES in the audit. Furthermore, nobody from Denmark had contacted Mr Kolding to say, having read the EES report, that when it stated that the questions concerning air emissions were not applicable to the facility, that this was incorrect because EtO was being discharged from it.

126 Contradicting his earlier evidence, Mr Kolding stated that the responsibility for checking the correctness of the conclusions regarding air emissions contained in the EES report fell not to himself but to his operations manager, Mr Kanda. The task had been delegated to him. 127 In November 2002 the council conducted its own environmental audit of the facility. Ms Angela Daly of the council was the officer responsible. Mr Kolding was not present when Ms Daly conducted the environmental audit. Mr Kanda, however, was. Mr Kolding stated that if, as a result of the council audit, Unomedical had been asked to conduct air emission tests or to install an abatement system, it would have done so.

128 In relation to Ms Daly's environmental audit report, Mr Kolding stated that he knew, as at the date it was sent, that what was stated in the report (namely, that there was "no evidence of air emission (pollution control devices etc not relevant)") was not correct because there were emissions from the facility. It was Mr Kolding's evidence that Mr Kanda would have told Ms Daly during the audit about the EtO emissions. According to Mr Kolding, if Mr Kanda had not discussed this with Ms Daly "it would be a mistake". But, Mr Kolding agreed that while Ms Daly would have seen the emission stack on her inspection, EtO emissions were odourless and were not visible to the naked eye.

129 In relation to the second DA in 2005 to increase the capacity of the gas tanks from 80 kgs to 750 kgs, Mr Kolding accepted that the fact that EtO was being emitted into the atmosphere as a result of the sterilisation process was not disclosed in the application. Mr Kolding stated, however, that had the council made such inquiries this information would have been provided to it. 130 Mr Kolding gave evidence that prior to the beginning of 2006 the steriliser would have been operating at approximately 80 to 85 % capacity. That is to say, it would have been using 8 to 9 tonnes of EtO per year. Mr Kolding stated that typically the operation of the steriliser was five days a week, but there were times when it was operating "24/7". Mr Kolding stated that by the end of the charge period, Unomedical was very busy and the steriliser was generally operating seven days a week. 131 Mr Kolding stated that from early to mid 2006 Unomedical sought to expand its production capacity. This anticipated expansion would have taken Unomedical over the 10 tonne annual limit for use of EtO in its operations and as a consequence Unomedical gave consideration to installing an abatement system using a catalytic converter. Consequently Mr Kolding engaged in discussions with a company called Lesni to provide Unomedical with a catalytic converter. 132 Mr Kolding gave evidence that about this time he consulted Mr Andrew Doig of AEBN and requested him to inspect the premises and to give advice on whether Unomedical was "right in our assumptions of what was required for us to operate the steriliser". Mr Doig attended the premises and Mr Kolding showed him the steriliser and explained the mechanics of its operation. Mr Kolding stated that Mr Doig subsequently telephoned him and "confirmed that he was on the same belief as I was", namely, that there were no emission limits on the operation of the steriliser.

133 Mr Kolding recalled that he and Mr Doig had a conversation about the 10 tonne Commonwealth National Pollutant Inventory ("NPI") threshold. Mr Kolding also recalled Mr Doig telling him that he should engage an air pollutant consultant for the purpose of properly assessing the EtO emissions from the stack at the facility. That is to say, "to get an idea of what the limit and what our emission was, yes".

134 Mr Kolding stated that when he spoke to Mr Doig about the 2005 Approved Methods document, Mr Doig had told him that it did not apply to Unomedical.

135 However, under cross examination Mr Kolding admitted that he was aware of the existence of the document prior to his conversation with Mr Doig (initially Mr Kolding stated that he had not read the document whereas Mr Kanda had and that Mr Kanda had not communicated the contents of the document to him). He agreed that he had read parts of the 2005 Approved Methods document in 2006. In particular, he had read section 7.2.1 "Impact Assessment Criteria" and table 7.2A. Mr Kolding stated that as a consequence of reading section 7.2.1 and table 7.2A he understood that EtO was considered to be a principal toxic air pollutant; that it was carcinogenic; that it was highly toxic; that it was utogenic; that it was highly persistent in the environment and that it had to be minimised to the maximum extent achievable through the application of best practice process design and/or emission controls.

136 Mr Kolding gave evidence concerning the inspection by Mr Ives on 17 November 2006. He stated that Mr Ives had asked to see the emission point and that he had obliged. In the weeks that followed, Mr Ives had asked Unomedical to assess and test the EtO emissions. Unomedical agreed to do so and engaged HLA to perform the tests. 137 Mr Kolding agreed that it was not in dispute that the testing revealed EtO emission levels of 3.3 ug/m3. It was also not a matter of controversy that the subsequent HLA analysis indicated that 99% of the EtO used in the sterilisation process was discharged to the atmosphere via the stack.

138 As a consequence of the tests and the resulting prevention notices, a catalytic converter abatement system was installed at the facility which was obtained from "an idle system which we got from Denmark, from one of the old sites in Denmark".

139 Mr Kolding accepted that prior to the involvement of HLA, Unomedical had established no system to monitor EtO emissions outside the facility and no analysis had been conducted to determine how much EtO was being emitted from the facility's stack.

140 Inside the facility, by contrast, electrochemical detectors were calibrated against EtO and set to a 1 ppm (1 part per million) detection rate. These detectors were there, Mr Kolding stated, for the purpose of ensuring compliance with occupational health and safety standards.

141 Mr Kolding noted though, that these internal detectors were not appropriate for monitoring the ambient value of EtO outside the facility. To obtain this information, dispersion modelling was required based on the stack emission figures.

142 While Mr Kolding initially stated that he was unaware that one available means of monitoring ambient EtO values was to ascertain the stack concentration or emission of EtO and then to use modelling to determine where and how the EtO was being dispersed, he later conceded that during the charge period Unomedical could have readily commissioned an expert to model and quantify the dispersion of EtO from the facility but that it did not. Mr Kolding stated that he had not previously monitored the EtO emitted because "I have never been asked to do it".

143 Mr Kolding did not know whether or not monitoring was undertaken by Unomedical in Denmark. This was because Mr Kolding had never asked. Nevertheless Mr Kolding accepted that there were persons to whom enquiries could have been made about undertaking such testing and modelling, in particular, Unomedical's EtO specialist, Mr Alan Happel.

144 On 21 July 2007, Mr Kolding sent an email to Mr Henrik Brandt, the CEO of the Unomedical group, in Denmark. The email stated as follows: Hi Henry, the reason I called you Friday evening was to inform you about a critical development in our EtO operation. As you probably know we have over the last few years budgeted for a catalytic converter which is a known process to reduce EtO emission to the atmosphere. According to our legal environmental advisors and the way we interpret the legislation we are not obliged to have installed such a converter and we therefore postponed the investment as long as possible. However late last year the local council asked Unomedical to conduct a voluntary air quality impact assessment for the sterilisation process. We conducted the test in early June and submitted the report which was of a critical character. We had included a plan on how to reduce our future emissions to an acceptable level. 145 Mr Kolding's evidence was that he was convinced that Unomedical had been conducting the sterilisation process at the facility by "such practical means which is necessary to minimise pollution, absolutely, yes". This belief was based on: Among other things, I relied on Pittwater Council. I relied on my EtO specialist in Australia Narinda Kanda and our EtO specialist in Denmark and also obviously myself have been exposed to EtO issues over many years and my early experience with the limited amount of EtO we were using I didn't find it necessary to install any abater at that point in time. … I was convinced that there was no harm ever caused to local residents, the environment itself or our employees inside the factory. 146 Significantly, Mr Kolding agreed that an abatement system could have been installed at the facility in 2001 when Unomedical was seeking development approval for the use of Wilmette Place. He agreed that it was "practicable" for Unomedical to have installed a catalytic converter. He stated, "I think in layman wording, it was practical to put an abatement system in. I'm not disputing that. I didn't find it necessary to put an abatement system in." 147 Mr Kolding conceded that he was not aware of the duty in specific terms placed on Unomedical as the occupier of the facility by s 128(2) of the POEOA. But, he stated, "as normal practice and operating you will always try to minimise and improve things". Mr Kolding said, "I think we tried to minimise to the level we thought was necessary". Mr Kolding stated that whilst a catalytic converter was available as an improvement it was not, in his view, necessary. This was because, "we didn't believe there was a risk for human or the environment that was significant enough to warrant a catalytic converter".

148 In particular, Mr Kolding gave the following evidence: Q. You didn't know what level the emissions were once they were emitted in the ambient atmosphere? A. I didn't know the limits - the measurements but I knew how much EtO we used.

Q. But you can't point to any literature or learning which says the amount of ethylene oxide that was being used and emitted was safe, can you, for the environment? A. Not a number but I can say that my experience told me that the level of EtO we were using was an insignificant number of kilos.

Q. You can't point to any literature and learning which says the amount of EtO that was being emitted during the charge period through the stack was safe environmentally? A. I can say that the tests and the service I've seen there has been no proven harm to people, human being or workforces due - let's take the environment, the community of EtO emissions.

Q. None said it was safe environmentally to emit the amount of EtO Unomedical was during the charge period? A. I didn't think anyone said it was unsafe.

Q. Are you saying you had to wait for somebody to prove a link to cancer to an emission of EtO before Unomedical acted? A. No, I'm not.

149 Later Mr Kolding said: Q. During the charge period Unomedical was not minimising to the maximum extent achievable the emission of ethylene oxide through the stack at Wilmette Place, was it? A. That's correct.

Q. You always understood that minimising to the maximum extent achievable for ethylene oxide could have been done by the use of a catalytic converter before the charge even commenced? A. If it had been required it could have been.

Q. But you could have required it? A. Yes.

Q. As the managing director? A. I could but didn't. I didn't find it necessary. 150 Mr Kolding further admitted that "it was general knowledge for me in Denmark that it [EtO emissions] was controlled by an abater system." He gave the following evidence: Q. What did you understand from your reading and knowledge in the mid 1990s in Denmark was the limit or level for the sterilisation factory for environmental purposes? A. The number which was discussed back then was 99% efficiency of the catalytic converter.

Q. You always knew, going now to Wilmette Place and the charge period, that there was never a 99% efficiency - by that I mean destruction of ethylene oxide before it was emitted? A. Yes.

Q. You knew that was achievable from as early as the mid 1990s from what you had learned about and read about in Denmark? A. Yes, I did.

Q. Are you sure you don't remember the concentration or reading required in Denmark from the mid 1990s? A. I don't. … Q. Did you enquire what was the limit or the reading that Danish requirements imposed for the manufacture of medical products where ethylene oxide was used and emitted? A. I don't know whether I asked that question. It was well known to me and it's well known to everyone.

Q. And what did you understand before the charge period a catalytic converter would do for ethylene oxide? A. It would reduce it up to 99%. I didn't find that necessary for the amount of EtO we were using. 151 The cost of installing the abatement system was approximately $840,000. Mr Kolding candidly admitted that financial constraints were not the reason why an abatement system had not been installed, or why emissions testing had not been performed, earlier: Q. But there's no dispute, Unomedical always had the financial capacity to pay and absorb that cost through an abatement system in the charge period for Wilmette Place? A. Correct, we could have done that.

Q. You could have asked this Danish EtO specialist for the exact measurement or reading outside the sterilisation emission area in the environment that the Danish regulations require? A. I could.

Q. You could have always asked that person about the testing equipment that was used in Denmark to check whether that reading was being satisfied? A. I could.

Q. You never did that, did you? A. No.

Q. And it was your responsibility as you say in your record of interview because you were the chief executive officer in Australia as to whether such an inquiry was made to find out from the EtO specialist in Denmark what testing equipment could be used to check for Wilmette Place what the reading or concentration was of the EtO once it was emitted from the stack into the air? A. I could have….

… Q. I'll just clarify again. Did you as the managing director take any steps, during the charge period before HLA got involved, to establish the percent of ethylene oxide that was absorbed by the product? A. No because we are not required to.

Q. Did you take any steps to check the percentage that was actually being emitted through the stack into the atmosphere? A. No, because we are not required to. 152 Again Mr Kolding stated that, "I was aware that there is a general obligation in general that you try to minimise and you try to improve. That's not only the Australian legislation, that's typical legislation worldwide…that's the way I've always operated and that's the way you operate worldwide."

153 Mr Kolding denied that Unomedical had not done anything to minimise its emissions of EtO into the atmosphere. Mr Kolding explained that the following minimisation methods were used at the faciltiy: (a) first, a new automatic steriliser has been installed in 2002. This steriliser was more efficient and did not utilise any additional EtO than that which was absolutely necessary to ensure that the medical products were properly sterilised;

(b) second, Unomedical only ran full loads in the steriliser;

(c) third, a tall emission stack was employed, namely, 8 m above the roof level, which meant that the emission point was approximately 20 m above the point of exhaust;

(d) fourth, the emission velocity was increased to ensure a better disbursement of the EtO in the atmosphere; and

(e) fifth, an extended number of flushes were used to dilute the EtO upon emission.

154 Under cross examination Mr Kolding conceded that he did not have the expertise to dispute Mr Court's expert evidence with regard to the limited efficacy of the minimisation methods employed by Unomedical at the facility.

155 Mr Kolding agreed that Unomedical had the following policy in place during the charge period: Unomedical has well developed management systems covering regulatory affairs, quality assurance and environmental requirements to ensure that the company complies with international requirements and sets high standards for the manufacturing of medical devices…Unomedical is a business-driven organisation with focus on all quality issues as well as a constant focus on the customer's wishes and demands. Coordination and cooperation and sharing of 'best practice' knowledge across business units and geographical borders are essential parameters…whenever we develop new products we base our selection of materials and technologies on a desire to minimise the environmental impact as well as ensuring health and safety at workplace. 156 Mr Kolding did not deny that it was his responsibility in Australia during the charge period to ensure compliance with this policy in respect of the facility.

157 But when pressed, Mr Kolding stated that, in his view, this meant only complying with local statutory requirements and did not extend to adopting the legislative requirements that would have been applicable had Unomedical been operating in Denmark or the United States.

158 This was so, notwithstanding that Mr Kolding was aware prior to the charge period that a facility of Unomedical's size operating in the United States during the charge period would have been compelled to achieve a 99 % reduction of its EtO emissions.

159 Mr Kolding stated that he was aware of this standard because he was familiar with the literature and learning to this effect. Indeed, Mr Kolding went further and said that this requirement was known by everyone working with EtO at Unomedical, including the head office in Denmark. He said, "we know what requirements there are in countries by countries". This was because, amongst other things, Unomedical employed international experience as a guide to what constituted best practice in Australia.

160 However, Mr Kolding did not believe that the United States standard was appropriate for Unomedical's operation in Australia because: (a) there was no legislative requirement in Australia that Unomedical install a mechanical control measure to achieve a 99% clean emission rate;

(b) the studies he had seen had, in any event, not demonstrated that there was proven harm to human health from EtO emissions at the levels at which Unomedical were discharging;

(c) the United States threshold was an arbitrary limit. It was, in his opinion, a "political measure";

(d) he had assumed that the threshold for the installation of an abatement system in Australia was 10 tonnes; and

(e) the council did not require any abatement system to be installed and it was "very well aware", and had been for some time, of the sterilisation processes undertaken by Unomedical at the facility.

161 Thus Mr Kolding ventured the following opinion:

Q. Why do you think they required for use of 907 kilograms or more - that less than 907 kilograms a 99% emission reduction other than as an attempt to prevent or minimise dangers or risks to public health and welfare, what other reason? A. I think that that's the reason why they put it in but I don't know what the hazard or the danger actually was.

Q. What part of public health or welfare do you think they were trying to protect or minimise by requiring a 99% reduction in the ethylene oxide? A. I can read it and I can say that that's the reason why they have said they are doing it but I don't know what they are actually are conducting that statement out of, I don't know what service they have done for - to get that regulation in place, I don't know.

162 Notwithstanding Mr Kolding's belief as to the benign effects of the atmospheric emissions of EtO by Unomedical from the facility, Mr Kolding accepted that he was aware during the charge period that the National Code stated that EtO was a category 2 carcinogen, meaning that it is a "probable human carcinogen" and that "it should be treated as carcinogenic to humans and should be used and handled with great caution". Mr Kolding was also aware that "exposure to EtO should be kept as low as practicable and under no circumstances shall it exceed the national exposure standard of 1 ppm". Equally, Mr Kolding was aware that Pt 6.4 of the National Code provided that: compliance with this exposure standard should not preclude further efforts to reduce exposure … therefore it is a good general policy to keep exposure to any substance as low as practicable irrespective of whether present information indicates its hazardous or not at a particular concentration partly in case future experience shows that the substance represents a hazard which could have been prevented or limited by the application of simple control measures. 163 Further, Mr Kolding knew that Pt 6.5 of the National Code provided that: The absence of a relevant exposure standard for a mixture containing ethylene oxide shall not be considered an indication that the exposure need not be controlled. Where there is no exposure standard exposure shall be controlled to the lowest practicable level and this shall be expressed during the assessment process. 164 While Mr Kolding was aware that there was a high probability of toxicity to humans exposed to EtO, he did not regard the amount of EtO that was being emitted by Unomedical as capable of causing harm because he believed it was disbursed upon emission from the stack. Mr Kolding said that he had read in the literature and learning that "there was no suspicious issue and harm" in relation to EtO. Accordingly, he said, "I haven't seen anywhere where it was proven that it has had significant impact or an impact which alerted me to do something". Mr Kolding stated that he did not believe "whatsoever" that the impact of the emissions was in any way harmful to the environment or people.

165 It appears this belief was premised on an understanding by him that EtO was rapidly absorbed and diluted by water in the atmosphere. After accepting that it did not rain every day during the charge period thereby washing the EtO, the following evidential exchange took place: Q. Did you know that during the charge period that absent rain washing out the ethylene oxide and relying upon just normal water vapour found in the atmosphere that the ethylene oxide one week later had an estimated half of its original concentration when it was emitted? A. Not that exact number. I knew that it had a lifetime in the atmosphere but how long it was or - but I didn't regard it as important, as any significance I would say.

Q. When you say you didn't regard it as important or significant for what period of time did you believe the ethylene oxide retained half of its concentration-- A. I didn't know.

Q. You never informed yourself? A. No.

Q. You never got any advice about that? A. Not as I can recall in that detail.

Q. Are you able to comment on this suggestion that the reduction one kilometre from the factor with a light wind of 1 metre per second would have only been 0.1%, in other words 99% of the ethylene oxide discharged still existed one kilometre away? A. Okay. I accept that.

Q. Didn't it occur to you as a managing director that the ethylene oxide, instead of getting into the factory, might get into adjoining houses, that is adjoining the premises Unomedical were on? A. Yes.

Q. Or that it might get into other factories in the neighbourhood? A. Yes.

Q. And that could occur, as you understood it, from wind blowing from a direction which didn't take the ethylene oxide towards the monitor but the other way? A. Yes.

Q. And there was nothing in place to prevent that occurring was there? A. No.

Q. Nothing in place to minimise that occurring, the ethylene oxide being blown into a factory or a house or both of neighbouring properties? A. Not above the stack to prevent it, no.

166 The belief that there was a "very minimal" impact on the environment during the charge period by reason of the discharge of EtO into the air was held notwithstanding that, as Mr Kolding accepted, Unomedical had not undertaken any monitoring or modelling of EtO emission rates during the charge period. Rather, because Mr Kolding knew that the amount of EtO that Unomedical used in its sterilisation process was less than 10 tonnes he "assumed that the professional regulators in Australia knew that under 10 tonnes there was no risk in that emission". Later Mr Kolding admitted that nowhere in the NPI was it stated that if the threshold of 10 tonnes per annum of EtO use was met, an occupier of premises was absolved from having to prevent or minimise EtO from being emitted into the atmosphere.

167 Finally, Mr Kolding accepted that prior to the charge period he had been aware of the Victorian legislative requirements limiting the discharging of EtO into the atmosphere consequent upon the operation of a steriliser. Mr Kolding stated that he discussed with Mr Kanda whether or not Unomedical should adopt the Victorian EtO emission levels, but that it was decided by him that it was not necessary to do so.

168 The EPA submitted that Mr Kolding was an unreliable witness and that certain aspects of his evidence were inaccurate based, as they were, on reconstruction. In particular, the EPA submitted that Mr Kolding's evidence of his discussions with Mr Kanda could not be relied upon.

169 My overall impression of Mr Kolding was that of an honest witness who did not seek to mislead the Court in giving his account of the events culminating in the charge. Given the passage of time between the writing this judgment and Mr Kolding's evidence, I have sought to confirm this opinion by extensive recourse to the transcript and to notes taken by me during the course of the hearing. There is force in, and I accept, the EPA's submission that Mr Kolding's recollection of conversations between himself and Mr Kanda was reconstructed and was unreliable. Absent confirmation by Mr Kanda or some other source, I am loath to accept them as accurate.

170 I was troubled that in his record of interview with the EPA Mr Kolding made scant mention of critical conversations between him and Mr Kanda. For example, he did not refer to Mr Kanda telling him about the 10 tonne threshold or to advice reputedly given to Mr Kanda by the EPA concerning the lawfulness of the operation of the steriliser by Unomedical at the facility.

171 I was additionally worried by the fact that Mr Kanda was not called as a witness in the proceedings, notwithstanding the evidence disclosed by Mr Kolding that Mr Kanda's whereabouts in Malaysia were known to Unomedical and that Mr Kolding had recently been in telephone contact with him. While I have not drawn any adverse inferences against Unomedical to the effect that Mr Kanda's evidence would not have assisted Unomedical (Louizos v R (2009) 194 A Crim R 223 at [54]-[57], Mahmood v Western Australia (2008) 232 CLR 397 at [27] and Dyers v The Queen (2002) 210 CLR 285 at [5]-[6], but see Azzopardi v The Queen (2001) 205 CLR 50 at [64]), his absence in conjunction with Mr Kolding's failure to mention critical conversations and events in which Mr Kanda was a participant during his record of interview have meant that I have not found these accounts reliable.

172 The corollary of this concern is that I place no weight on what Mr Kanda told Mr Kolding in the absence of any verification of these conversations.

173 On any view, the drafting of s 128(2) of the POEOA is fraught. It is perhaps for this reason that the provision has remained untested to date. The section makes an offence not the failure by occupiers of premises to observe defined standards, but failure to comply with a general duty which, by the words "as may be necessary to prevent or minimise pollution", deliberately leave undefined the scope of this duty and which render uncertain the steps necessary to avoid the imposition of criminal liability. In my view this is unsatisfactory. However, until such time as Parliament clarifies the provision it is the task of this Court to determine if a breach of the section has occurred.

174 Unsurprisingly, the EPA and Unomedical do not agree on the proper construction of s 128(2), and in particular, on the phrase "such practicable means as may be necessary to prevent or minimise pollution".

175 Unomedical submitted that the proper approach to the statutory construction of the provision was for the Court to: (a) determine if the activity minimises or prevents air pollution; (b) if not, then determine if the occupier implemented practical means; and (c) if the occupier did not, then determine whether it was necessary for the occupier to implement practical means.

176 In so doing, Unomedical urged upon the Court a reading down of the scope of s 128(2) of the Act on the basis that: (a) the provision only applied to non-scheduled premises;

(b) the provision would, if construed broadly, capture a range of everyday activities that were clearly unintended to fall within its ambit; and

(c) the provision would, if construed broadly, have detrimental policy ramifications because it would subject industries to a non-defined standard that varied for different industries and which would shift with even minimal technological advances or operative changes. Thus every business, irrespective of its size, would be required to implement world's best practice in order to avoid prosecution, despite the cost implications of doing so.

177 It is now almost trite law that the contemporary approach to statutory interpretation requires in the first instance a court to have regard to the context in which words appear and their purpose and not only after some ambiguity has been identified (CIC Insurance Ltd v Bankstown Football Club Ltd (1997) 187 CLR 384 at 408, Project Blue Sky Inc v Australian Broadcasting Authority (1998) 194 CLR 355 at [69], Network Ten Pty Ltd v TCN Channel Nine Pty Ltd (2004) 218 CLR 273 at [11] and s 33 of the Interpretation Act 1987). As was wryly, but pointedly, observed by Spigelman CJ in Caterpillar of Australia Pty Ltd v Industrial Court of New South Wales (2009) 255 ALR 131 (at [88]), "judges no longer approach a statute with scissors in one hand and a dictionary in the other."

178 In Alcan (NT) Alumina Pty Ltd v Commissioner of Territory Revenue (2009) 239 CLR 27 (at [47]) the High Court explained the process as follows: 47 This Court has stated on many occasions that the task of statutory construction must begin with a consideration of the text itself (105). Historical considerations and extrinsic materials cannot be relied on to displace the clear meaning of the text (106). The language which has actually been employed in the text of legislation is the surest guide to legislative intention (107). The meaning of the text may require consideration of the context, which includes the general purpose and policy of a provision (108), in particular the mischief (109) it is seeking to remedy.

The Emission of EtO Constituted "Air Pollution" as Defined

179 A consideration of the text of s 128(2) of the POEOA does not reveal whether the emission of EtO into the atmosphere by Unomedical constituted "air pollution" thereby engaging the duty imposed by that provision.

180 The Act defined the following terms in the Dictionary (s 4 of the POEOA states that "expressions used in this Act have the meanings set out in the Dictionary at the end of this Act"): (a) "air pollution" was defined as "the emission into the air of any air impurity";

(b) "air impurity" was defined to include "smoke, dust (including fly ash), cinders, solid particles of any kind, gases, fumes, mists, odours and radioactive substances". The terms "emissions" and "air" were, however, not defined;

(c) "pollution" meant, amongst other things, "(b) air pollution";

(d) "activity" was defined to mean "an industrial, agricultural or commercial activity or an activity of any other nature whatever (including the keeping of a substance or an animal)";

(e) "environment" meant "components of the earth, including: (a) land, air and water";

(f) "harm to the environment includes any direct or indirect alteration of the environment that has the effect of degrading the environment and, without limiting the generality of the above, includes any act or omission that results in pollution"; and

(g) "premises" included "a building or structure, or land or a place (whether enclosed or built on or not, or a mobile plant, vehicle, vessel or aircraft)".

181 The EPA submitted that EtO was an "air impurity" because it was a gas or mist within the Dictionary meaning of that term. I agree. Furthermore, and consistent with the defined terms above, the release of EtO into the atmosphere or air constituted "air pollution" and thus "pollution". Moreover, because the composite term "harm to the environment" expressly "includes any act or omission that results in pollution", the release of EtO by Unomedical was, by definition under the statute, pollution that caused harm to the environment.

182 The language of the legislation is clear and there is no warrant for inserting, as was submitted by Unomedical, into the text of s 128(2) an obligation by the EPA to demonstrate that actual harm to the environment or to human safety is required before the emission constitutes "air pollution", thereby narrowing the scope of the provision. To so construe the section would, in my view, result in error. Thus I find beyond reasonable doubt that the emission of EtO into the atmosphere by Unomedical from the facility unequivocally constituted air pollution and consequently resulted in harm to the environment.

183 Unomedical submitted that s 128(2) of the POEOA, when read conformably with cl 6C(2)(c) of the Protection of the Environment Operations (Clean Air) Regulation 2002 ("the Regulation"), cut down the breadth of the operation of s 128(2) and limited the operation of any obligation it imposed to circumstances where there was a possible risk of actual harm to human health and the environment. That clause provides: 6C General obligation to prevent or minimise air pollution (1) A person who burns anything in the open or in an incinerator must do so by such practicable means are necessary to prevent or minimise air pollution (2) Without limiting subclause (1), the means of preventing or minimising air pollution may include the following: … (c) burning only material that is suitable for disposal by burning, having regard to possible effects on human health and the environment. 184 But in my view, the Regulation does not assist Unomedical because not only did cl 6C(2)(c) not come into effect until 1 September 2006 (Protection of the Environment Operations Legislation Amendment (Control of Burning) Regulation 2006), but the clause, when read in context and, in particular, with sub-clause (1), is plainly concerned with air pollution caused by the burning of material in the open or in an incinerator. Neither processes were involved in the present case.

185 In any event, it is generally accepted in Australia that delegated legislation should not be taken into account in interpreting statutes and cannot be used to construe an enactment in a way that would alter provisions of the legislation (Wallaby Grip Limited v QBE Insurance (Australia) Limited; Irene Stewart (As Legal Personal Representative of the Estate of the Late Angus Clugston Stewart) v QBE Insurance (Australia) Limited (2010) 240 CLR 444 at [21] and Paice v Hill [2009] NSWCA 156 at [2]-[3]). To use the clause to interpret s 128(2) of the POEOA in the manner urged by Unomedical given the conclusion I have reached above about the illegitimacy of reading down the unambiguous text of that provision is, in my view, impermissible. If for no other reason, the effect would be to alter the meaning of the section.

186 Both the EPA and Unomedical sought comfort for their respective interpretative positions from the second reading speech for the POEOA. I did not find it, general in scope as it was, to be of assistance. In any event I am mindful of the recent caution given by the High Court in SAEED v Minister for Immigration and Citizenship (2010) 115 ALD 493 that "the words of the Minister must not be substituted for the text of the law" (at [32] and see [30]-[34]).

187 Having regard to context, an examination of the purpose of the legislation reveals the following relevant statutory objects: 3 Objects of Act The objects of this Act are as follows: (a) to protect, restore and enhance the quality of the environment in New South Wales, having regard to the need to maintain ecologically sustainable development, … (d) to reduce risks to human health and prevent the degradation of the environment by the use of mechanisms that promote the following: (i) pollution prevention and cleaner production, (ii) the reduction to harmless levels of the discharge of substances likely to cause harm to the environment, (iia) the elimination of harmful wastes, (iii) the reduction in the use of materials and the re-use, recovery or recycling of materials, (iv) the making of progressive environmental improvements, including the reduction of pollution at source, (v) the monitoring and reporting of environmental quality on a regular basis,

188 The objects reveal that the central mischief to which the POEOA is directed is to avoid, or at the very least, reduce pollution in order to prevent harm to human safety and the natural environment. The stated aim is not only wide in scope, it is preventative, rather than reactive, in direction. Having regard to the context in which s 128(2) is to be construed and to the objects of the Act as contained in, in particular, s 3(a) and (d), it is apparent that, as a matter of objective intention, the legislature did not seek to narrow or overly circumscribe the scope of the general duty imposed by s 128(2) by insisting that the provision was only engaged once actual harm to the environment or to human safety was demonstrated. Were it otherwise, the vice the legislation was enacted to protect against would be at risk of subversion.

189 Unomedical contends that construed broadly s 128(2) has the potential to produce absurd results. Thus, for example, the release of steam from a kettle, the driving of a motor vehicle or the operation of an air conditioner all amount to "air pollution" which, however transitory, could conceivably result in prosecution under the POEOA. This, Unomedical asserts, could not have been the intention of the legislature in enacting s 128(2).

190 It has been held that definitions which have a wide import that would curtail everyday activities manifestly not within the contemplation of Parliament at the time of enactment ought to be read down (Palos Verdes Estates Pty Ltd v Carbon (1991) 72 LGRA 414 at 427-428, Electricity Commission of NSW v Environment Protection Authority (1992) 28 NSWLR 494 at 498-499 and Chief Commissioner of State Revenue v Qantas Airways Ltd (2009) 74 NSWLR 181 at [33]-[36]). It has also been held that statutes containing penal provisions should be strictly construed where ambiguity is inherent within its language (Beckwith v R (1976) 135 CLR 569 at 576, Waugh v Kippen (1986) 160 CLR 156 at 164 and R v Lavender (2005) 222 CLR 67 at [87]-[93]). 191 In my opinion, when regard is had to the ordinary and natural language of the text of s 128(2), including the defined terms contained within the section, it is abundantly clear that the emission of EtO into the atmosphere was intended by the legislature to constitute "air pollution". That no proven actual harm resulted from the emission is irrelevant and there is no warrant for importing an element of actual harm into the definition. This conclusion is reinforced by the context of the provision concentrating, as it does, on precautionary harm prevention. Put simply, it mandates that an occupier of premises err on the side of caution.

192 Having said this, there is considerable force in Unomedical's submission that taken to its logical conclusion the construction posited by the EPA, namely, that all "air pollution" as that term is defined, could result in frivolous breaches of the POEOA that were unlikely to have been intended by Parliament. It is also worth noting that the definitions of "noise pollution", "land pollution" and even "water pollution" in the POEOA, albeit employing different language, all import the concept of actual or potential harm into their respective definitions.

193 But in my opinion, and in addition to what I would regard as a manifest inconsistency with the objects of the Act, an absence of similar language in s 128(2), including in the text of any of its defined terms, strongly militates against the suggestion that the legislature did not intend to define "air pollution" as widely as it did. I therefore do not accept that any element of proven harm ought to be read into the statutory definition of "air pollution". Accordingly, if the emission of EtO into the atmosphere satisfies the statutory definition of that term, then it amounts to "air pollution" for the purposes of s 128(2) of the Act irrespective of whether it has caused actual harm to the environment.

194 This reasoning includes Unomedical's alternative contentions of likely, possible or potential harm. Unomedical submitted, in support of an interpretation of s 128(2) of the POEOA that it required likely, possible or potential harm to be demonstrated in order to enliven the duty imposed by the section, that the provision must be interpreted in the context of other important sections in the POEOA. First, s 130, which prioritises the various air pollution offences to the effect that all other air pollution offences in Pt 5.4 of the POEOA prevail over s 128 thereby making s 128 the lowest tier of all air pollution offences. And second, ss 96 and 101 which permit a regulator or the Minister respectively, cognisant of a breach of s 128(2) to issue a prevention or prohibition notice. These latter provisions, Unomedical noted, require actual or likely harm to the environment before notices can be issued. Accordingly, Unomedical submitted, given the lowly legislative status accorded to the offence created by s 128(2) and given that at least some of the possible sanctions for a breach of the section require actual or likely harm to be demonstrated, the same requirements "should logically apply in a proper interpretation of section 128(2)."

195 I fail to see, as a matter of logic, why this should be so. Even reading s 128(2) in this much wider context, such a gloss on the clear words (at least in this regard) of the provision cannot be justified pursuant to either a literal or purposive construction of the provision. The fact that the legislature has omitted any specific requirement of actual or likely harm from the text of the offence but has included this element in other provisions which may touch upon s 128(2) strongly indicates, in my view, that the exclusion was deliberate.

196 Furthermore, ss 96 and 101 perform a very different function inasmuch as they are directed to sanctions that may be imposed consequent upon an established contravention of the POEOA. In this way, that regard must be had to whether or not actual or likely harm resulted from the polluting activity is no different to the exercise of taking into account this fact as an aggravating factor in the determination of an appropriate sentence pursuant to the commission of any environmental offence. Therefore, while there is, in my opinion, no warrant for imposing as an element of the offence the need to demonstrate likely or actual harm for the purposes of securing a conviction, ultimately the harm caused by the air pollution will be brought to account in determining penalty.

197 The possible vice about which Unomedical is rightly concerned, viz, the prosecution of persons for the unintended consequences of broad statutory conceptualisation of air pollution, is safeguarded against in several ways. First, the application of the principle that the law is not concerned with trifles (de minimus no curat lex). This principle may be used not only to avoid the imposition of punishment after a finding of guilt where the offence is trivial but also "to avoid hysterical or oppressive law enforcement; cases in which a finding of guilt would tend to bring the law or the judicial system into ridicule or contempt because of triviality" (Williams v The Queen (1978) 140 CLR 591 at 602 per Murphy J and see also 600 per Gibbs and Mason JJ and Electricity Commission of NSW v Environment Protection Authority (1992) 28 NSWLR 494 at 499).

198 Second, and as discussed in greater detail below, whatever be the precise meaning of the composite phrase "such practicable means as may be necessary" to prevent or minimise the polluting activity, in cases where the activity results in air pollution that is transitory or wholly benign it is more than likely that no "practical means" will ever "be necessary to prevent or minimise the air pollution" and that therefore no breach of the section will ever occur. The flexibility inherent in the phrase "such practicable means as may be necessary" imposes a control on the ambit of activities that will be caught by s 128(2) of the POEOA.

199 Third, the Division in which s 128(2) is located has expressly excluded residential premises (s 131).

200 If, however, the interpretation I have given to the term "air pollution" is incorrect and it does include some aspect of harm to the environment to be proved, Unomedical cannot, in my view, escape liability under s 128(1). This is because potential, or even possible, harm to the environment would, in my opinion, suffice. As discussed above, to hold otherwise would be to confine the scope of the term "air pollution" in a manner which was not intended by the legislature given the plain and ordinary meaning of the text of s 128(2) and its defined constituent parts. It would also not be consonant with the objects of the POEOA. Accordingly, if the term "air pollution" is to be interpreted as requiring some degree of harm before it can be demonstrated, then potential or possible harm is sufficient.

201 I am of the opinion that the evidence demonstrates that EtO (unlike water vapour from an air conditioner), was a known and accepted carcinogen and had the very real potential to cause harm to the environment, including harm to human health. This is so notwithstanding that the levels at which EtO can safely be discharged into the atmosphere have not been conclusively proven. This absence of proof does not, in my view, derogate from the potential or possible, as opposed to the actual, harm which could have resulted from the emission by Unomedical of EtO into the air from the facility.

202 Were it necessary, I would therefore find beyond reasonable doubt that the emission of EtO into the atmosphere by Unomedical constituted air pollution as defined for the purposes of s 128(2) of the POEOA because it caused potential or possible harm to the environment.

203 The remaining issue is what were the "practical means", if any, that "may be necessary" to prevent or minimise the air pollution by Unomedical.

204 While the submission of Unomedical is undoubtedly correct, namely, that the words "such practicable means" must be read in the context of the words "as may be necessary", as a matter of convenience I intend to first examine the meaning of "such practicable means" and then interpret the phrase "as may be necessary", before finally turning to the words "to prevent or minimise" air pollution.

205 While there is no composite or individual definition of the expression "such practicable means" in the POEOA, it can be readily appreciated that the duty imposed by s 128(2) of the POEOA is not absolute, modified as it is by the adjectives "practicable" and "necessary".

206 The ordinary everyday meaning of the term "practicable" includes "capable of being put into practice, done or effected, especially with the available means or with reason or prudence; feasible" (Macquarie Dictionary, online edition) and "able to be done or put into practice successfully; feasible; able to be used; useful, practical, effective" (Oxford English Dictionary, online edition).

207 These definitions reveal that the question of what comprises "practicable means" raises issues of common sense and specialist knowledge and will vary in any set of given circumstances. Furthermore, its content will vary over the course of time as a polluter's knowledge concerning the risk of harm to the environment the activities cause and available technological improvements to eliminate or reduce this risk increases (Chugg v Pacific Dunlop Ltd (1990) 170 CLR 249 at 260-261).

208 Both Unomedical and, in particular, the EPA relied on a number of authorities from both the United Kingdom and Australia to support their submissions in respect of the meaning to be attributed to the term "such practicable means". Most of the cases cited concerned breaches of occupational health and safety legislation. While recourse to these authorities as a matter of guidance ought not necessarily be discouraged, the decisions must nevertheless be treated with considerable caution given their different statutory context and given that the legislature has not sought in s 128(2) to amplify or reduce the scope of the term "practicable" by modifiers such as "best" or "reasonable".

209 Be that as it may, it follows that the nature of the risk, the cause of the risk, the state of the scientific and engineering knowledge and learning at the relevant time and the means by which the air pollution could have been prevented or minimised, all form part of the assessment of whether Unomedical took "such practicable means" as may be necessary to prevent or minimise air pollution (Chugg at 261 and Bankstown Foundry Pty Ltd v Braistina (1986) 160 CLR 301 at 308-309). Contrary to the submission put by the EPA, in my view, when combined what underlies these factors is an inherent element of reasonableness. That is not to say that the touchstone of reasonableness is a necessary element to which regard must be had of and by itself in construing the provision - if it were it would have been expressly referred to in the body of the section - but rather, that when examined in their totality, these factors necessitate the implementation of a specific measure that is reasonable in all the circumstances.

210 Unomedical submitted that the word "may" in the context of "as may be necessary" means that the occupier of the premises has a discretion to exercise whatever practical means the occupier believes are necessary to minimise or prevent pollution. Moreover, and relying on the decision in Pelechowski v Registrar (1999) 198 CLR 435 (at [51]), Unomedical submitted that the term "necessary" meant that which is reasonable in all the circumstances and not that which is indispensable or essential. Therefore, when read as a whole, the phrase "as may be necessary" permitted the occupier of premises to determine for itself whether or not it is reasonable to implement "such practicable means". That is to say, the occupier has a discretion.

211 In support of this argument it relied on s 9 of the Interpretation Act. An examination of that provision, however, reveals that Unomedical's reliance on it is misconceived. Section 9 states: 9 Meaning of may and shall

(1) In any Act or instrument, the word "may", if used to confer a power, indicates that the power may be exercised or not, at discretion.

(2) In any Act or instrument, the word "shall", if used to impose a duty, indicates that the duty must be performed.

212 Section 128(2) does not, however, confer a power for the purposes of s 9(1) of the Interpretation Act. It therefore cannot assist Unomedical.

213 Unomedical also relied on cl 6C(1) of the Regulation to argue that when a comparison was made between the wording of that subclause and the language of s 128(2), it was clear that "as may be necessary", rather than "as are necessary", provided greater discretion to the occupier of premises to implement such practical means to prevent or minimise air pollution.

214 The reasons why a legislature chooses a different formulation of words in a later enactment are many and varied. For the reasons discussed above, little assistance, if any, is gained from cl 6C(1) of the Regulation in the construction of s 128(2) of the POEOA.

215 I also do not accept Unomedical's submission that "may" confers a discretion on an occupier of premises to implement practicable means.

216 This construction is entirely inconsistent with the objects of the Act and the objective intent of the provision, which, while ambiguously expressed is nevertheless inherently clear, namely, to impose a duty on those who engage in activities which pollute the air to reduce or eliminate this pollution. To interpret the section as conferring a discretion on the polluter would fundamentally undermine this legislative intent. At its highest, the only choice implicit in the word "may" is, when that term is read in the context of the provision as a whole, a choice as to which practicable means are to be employed to prevent or minimise the air pollution. It does not, where available practical means exist, permit the polluter to simply elect not to implement any of them.

217 Instead, I accept the submission of the EPA that the correct statutory construction of the word "may" in s 128(2) is that it imposes a mandatory duty to carry out an activity or operate a plant by such practical means in circumstances where that activity or operation causes air pollution. As was observed in Ballantyne v WorkCover Authority of New South Wales [2007] NSWCA 239 (at [90]-[92]) quoting from, in particular, Samad v District Court of New South Wales (2002) 209 CLR 140: 90 Further, as noted by Santow JA in Raniere Nominees (No. 2) use of the term "may" does not necessarily identify a broad discretionary power. It may confer a power coupled with an obligation to exercise the power in circumstances where such an obligation is engaged by law. This accords with the principle established in Julius v Bishop of Oxford (1880) 5 App Cas 214. Of course, in present times general law principles must submit to statutory provisions and the Interpretation Act 1987 (NSW) provides that in any Act or instrument, the word "may" "if used to confer a power, indicates that the power may be exercised or not at discretion": s 9(1). That principle must be applied, except in so far as the Act or instrument concerned indicates a contrary intention: Interpretation Act , s 5(2). Further, to speak in terms of "discretion" is not to imply that the discretion is unlimited. The power of this Court in dealing with the present appeal, is conferred by use of the word "may" and includes remittal to the Commission and provides that the Court "may make such other order in relation to the appeal as the Court thinks fit": 1998 Act, s 353(2). Nevertheless, it would generally be a legally erroneous application of the power so expressed if this Court, having identified an error in point of law which was material to the outcome of the decision below, declined to grant relief. Nor should a statutory right to compensation for redress of a statutory duty be treated as subject to some broad discretionary limitation. Thus conferral of power on the Human Rights and Equal Opportunity Commission by the word "may" in the Sex Discrimination Act 1984 (Cth), s 81 did not entitle the Commission to refuse to make a declaration that the respondents pay to each applicant compensation for loss or damage suffered by them and, the power being engaged, the failure of the Commission to make such a declaration constituted an error in law: see Hall v A and A Sheiban Pty Ltd (1989) 20 FCR 217 at 241-243 (Lockhart J); 262 (Wilcox J) and 281-286 (French J).

91 The circumstances in which the conferral of a power by the word "may" should properly be understood as limited by an obligation to make an order in an appropriate case has been considered in a number of authorities including Ward v Williams (1955) 92 CLR 496, Finance Facilities Pty Ltd v Federal Commissioner of Taxation (1971) 127 CLR 106, Commissioner of State Revenue (Vic) v Royal Insurance Australia Ltd (1994) 182 CLR 51 and, more recently, Samad v District Court of New South Wales (2002) 209 CLR 140 at [31]-[36] (Gleeson CJ and McHugh J); [66]-[68] (Gaudron, Gummow and Callinan JJ). However, it should be not be understood that there is a bright line to be drawn between two entirely discrete concepts. As explained by Gleeson CJ and McHugh J in Samad at [35] the distinction is to be understood in the following terms: "Whether a statute, which confers a power, on its true construction also imposes a duty to exercise the power in certain circumstances, may also affect the form of relief that is appropriate where there has been a failure to exercise the power; it may affect the question whether the appropriate order is an order to make a certain decision, or to do a certain act, or merely to consider the matter according to law. And even if a statute does not impose a duty to exercise the power, it may circumscribed the considerations that are relevant to the exercise of a discretion. In every case the task is to construe and apply the statute." 92 Their Honours continued at [36]: "The statutory context may be crucial. Where a statute confers rights or entitlements, as in Finance Facilities , it may be easy to conclude that the legislature did not intend that they could be taken away by the exercise of a discretion reposed in an administrative authority."

218 The only choice or election invoked by the section is in determining which "practicable means" are necessary to implement in order to reduce or prevent air pollution. And as stated above, this in turn will depend on a range of factors, for example, the risk (actual, potential or possible) of harm the pollution poses to the environment, the cost of implementing the "means" and the state of scientific and industry knowledge at the time of the pollution. Otherwise the duty imposed on an occupier of premises is mandatory ("the occupier of any premises must", emphasis added) where air pollution results from the activity or operation occurring at the premises.

219 As to the term "necessary", Unomedical argued consistently with its earlier submissions, that before minimisation or prevention of air pollution was "necessary", the EPA was required to prove beyond reasonable doubt that either actual or likely harm was being caused by the unabated atmospheric discharges. Put another way, Unomedical contented that a remote risk of harm or even the potential for harm was insufficient to mandate the reduction or elimination of air pollution. This construction, Unomedical asserted, was consistent with the recognised definition of "likely" in Pace Farm Egg Products Pty Ltd v Newcastle City Council (2006) 151 LGERA 260 (at [54]). Thus the EPA was required to prove that harm to the environment "was a really or not remote chance or possibility regardless of whether it is less or more than a fifty per cent change before minimisation becomes necessary".

220 While in my opinion the risk of harm, either actual, likely, possible or potential, is a significant factor to which the Court ought to have regard in determining whether it was objectively necessary to implement practicable means in the operation of a plant or the carrying out of an activity, it is not the only, or even the determinative, factor to be considered. Furthermore, and for the reasons already given earlier, to interpret the phrase "as may be necessary" to mean, in effect, 'as only necessary where harm exists or is likely to exist', would denude the provision of much of its intended operation and would represent an unwarranted and unjustified gloss on the language of its text.

221 Equally, there is no basis for any construction of the term "to prevent or minimise" air pollution that deviates from the plain and natural meaning of those words. The word "minimise" has been defined as, "to reduce to the smallest possible amount or degree" (Macquarie Dictionary, on-line edition. The Oxford English Dictionary has a similar definition).

222 As a result of Mr Court's evidence that there was no control equipment which could totally prevent the emission of EtO from the facility, the EPA accepted that it was not possible for Unomedical to prevent air pollution. But, the EPA submits, it was possible for Unomedical to "reduce to the smallest amount or degree" its EtO emissions by the adoption of available "practicable means", specifically, by installing a catalytic converter.

223 Construing the section as a whole, in the circumstances of this case it follows from the principles articulated above, that s 128(2) of the POEOA imposed a mandatory duty on Unomedical, as an air polluter, to carry out its sterilisation operation using such measures as could reasonably be implemented in light of, amongst other things, the risk of harm (including actual, likely, potential or possible) to the environment, the scientific and industrial knowledge and practice at the time and the cost, in order to reduce to the smallest possible degree the EtO emissions from the facility.

Unomedical Did Not Carry on Its Sterilisation Operations By Such Practicable Means As Were Necessary to Minimise the Emission of EtO Into the Air

Nature of the Risk of the EtO Emissions and Unomedical's Knowledge of the Risk

224 Turning first to the nature of the risk created by the EtO emissions through the stack at the facility, there can be no doubt that Unomedical, through Mr Kolding, knew that EtO was a hazardous substance which posed a danger to human health due to its carcinogenic properties.

225 For example, the June 2005 MSDS unequivocally stated the dangers of exposure to EtO, which included that it was classified as a carcinogen. So too did the 2007 Register, the Approved Methods documents and the NPI. In particular, the MSDS noted that because limited ecotoxicity data was available at the time the Sheet was prepared, users of the product should "ensure appropriate measures are taken to prevent this product from entering the environment".

226 Likewise, the 2005 Approved Methods document, which Mr Kolding accepted that he had read in 2006, stated that EtO was a principal toxic air pollutant that had to be minimised to the maximum extent achievable through the application of best practice process design and/or emission controls. Unomedical submitted that the 2005 Approved Methods document was irrelevant because there was no emission rate or standard prescribed for the emission of EtO at non-scheduled premises. While plainly correct, in my view, the submission is besides the point. The evidence of Mr Court was not, contrary to the assertions of Unomedical, that the 2005 Approved Methods document applied to the facility, rather it was that the document could, and regularly was, used as a guide to regulate the emission of toxic chemicals such as EtO in the absence of any such prescription. The document was therefore relevant, as was Mr Kolding's awareness of its contents.

227 It was because Unomedical knew that EtO was harmful that one of the warning signs directly outside the sterilisation area at the facility contained the words in relation to EtO: CANCER HAZARD AND REPRODUCTION HAZARD FIRE AND EXPLOSION HAZARD

228 It was, without doubt, why within the factory building Unomedical strictly monitored the presence of EtO by electrochemical detectors set to a 1 ppm detection rate.

229 The risks to human health associated with exposure to EtO were confirmed by Professor Stewart's evidence. While Professor Stewart acknowledged that there was no evidence that indirect environmental exposure to EtO caused cancer in humans, he went on to state that there should never be any preventable exposure to a human carcinogen "because there is no evidence, biological evidence of a safe level of exposure to such materials." As a result, "public health policy mandates that people be not exposed to jointly preventable exposure of a carcinogen."

230 Having said this, the evidence reveals that no actual harm, in the sense of an additional case of cancer among the local population, occurred as a result of an exposure to EtO from the operation of the facility.

231 While initially reticent to accept that EtO was dangerous to humans, under cross examination Mr Kolding admitted that EtO was a highly toxic substance, was carcinogenic (he understood that "it can be cancer, cause cancers"), and therefore, that its presence in the environment ought to be minimised.

232 However, it was Unomedical's stated belief, expressed through MrKolding, that outside the facility the emission of EtO via the stack into the atmosphere was not harmful to either the environment or to humans because, according to Mr Kolding, "it is actually absorbed very quickly into the atmosphere so when you dilute it into air, it actually reacts very quickly with water and it becomes ethylene glycol", and therefore, "the impact on the environment is very, very minimum."

233 It was this reason, together with an absence of any defined regulatory limitations on the facility's EtO emissions, that was the foundation for Mr Kolding's belief that no measures whatsoever were necessary for Unomedical to implement during the charge period in order to minimise its air pollution.

234 But this belief was not credible, particularly in light of the evidence that: (a) although he was not aware of the precise amounts of EtO being emitted from the stack at the facility, Mr Kolding knew that there "was never a 99% efficiency - by that I mean destruction of ethylene oxide before it was emitted"; (b) between 2002 and July 2007, Unomedical did not undertake testing or modelling, independent or otherwise, to measure the emission rates of the EtO or to predict its dispersion once it entered into the atmosphere;

(c) there were no limits at all applied to the volume of EtO discharged from the stack into the atmosphere;

(d) there was no monitoring of EtO emissions outside the facility;

(e) Mr Kolding could have sought assistance and advice about testing to measure EtO emissions from the facility from Unomedical's EtO specialist in Denmark, but did not;

(f) while Mr Kolding told Mr Ives that the discharge of EtO was safe because EES had expressed this opinion in a report when Unomedical had initially moved to the facility, the report was never received by Mr Kolding (it was received by the head office in Denmark while Mr Kolding was away) and it was never provided to Mr Ives as promised by Mr Kolding;

(g) Mr Kolding agreed that none of the literature and learning that he was aware of during the charge period stated that it was environmentally safe to emit the amount of EtO Unomedical was discharging into the atmosphere from the facility. Mr Kolding's attitude in this regard was best summarised by his evidence that "I didn't think anyone said it was unsafe";

(h) Mr Kolding accepted that he was aware during the charge period of Parts 6.4 and 6.5 of the National Code that provided that it was good general policy to keep exposure to EtO as low "as practicable" and that "the absence of a relevant exposure standard" for EtO should "not be considered an indication that the exposure need not be controlled". While Unomedical submitted that the National Code was irrelevant because it was only an occupational health and safety guideline, the fact remained that Unomedical was, by reason of its awareness of the National Code's content, not only cognisant of the dangers of exposure to EtO, but was also aware that exposure to EtO ought to be regulated and monitored in all circumstances;

(i) while Mr Kolding knew that EtO had a half-life in the atmosphere, he did not know that absent rain washing, and relying just on normal water vapour found in the atmosphere, that one week later EtO was still half its original concentration. And moreover, he had not informed himself of this fact;

(j) it was possible that 99% of the EtO discharged still existed 1 km away with a light wind of 1 m/s and that there was nothing to prevent the EtO from being blown into a nearby factory or residential dwelling; and

(k) Mr Kolding had not seen or read s 128(2) of the POEOA until after the charge period and that he was not aware of the duty imposed on Unomedical pursuant to that provision.

235 I also did not find credible Mr Kolding's stated belief that because Unomedical used less than 10 tonnes of EtO per annum no measures needed to be implemented to reduce the facility's stack emissions. This is because, as discussed above, I have not accepted as reliable the conversations Mr Kolding states transpired between him and Mr Kanda, including for present purposes, the conversation where it is asserted that Mr Kanda told Mr Kolding that provided 10 tonnes or less of EtO per annum was used by the facility, no emission controls were necessary.

236 Unomedical submitted that the EPA, by allowing Unomedical to continue operating without installing an abatement system prior to the prevention notices being issued, had effectively conceded that there was no harm, actual or otherwise, to humans from environmental exposure to EtO from the facility's stack emissions. While superficially compelling, I do not find that any perceived or actual delay on the part of the EPA in this regard constitutes an admission by it as to the benign nature of the risks associated with the discharge of the EtO. No concession to this effect was made by any of the EPA's witnesses and no cross examination to this effect was undertaken by Unomedical during the proceedings.

237 Based on the evidence, there is no doubt in my mind, and I find accordingly, that: (a) Unomedical was aware of the fact that EtO was a dangerous substance that was toxic and was classified as a carcinogen with attendant risks to the health and safety of humans who came into contact with it, both inside and outside the facility; and

(b) while no actual harm was caused by the discharge of the EtO from the facility, given its toxicity and in the absence of any knowledge or information as to its possible dispersion once it was emitted into the atmosphere from the stack, there was a very real potential for environmental harm occasioned by its release.

The Amount of EtO Emitted 238 According to the historical modelling undertaken by HLA it is clear that almost 99%, and not the 50% as initially estimated by Mr Kolding, of the EtO that was used in the process of sterilisation was emitted into the air through Unomedical's stack.

239 The most comprehensive assessment of EtO usage and emissions for the facility was found in the 2008 HLA report. Table 9 summarised the emission details as follows:

Annual Ethylene oxide emission at Wilmette Place (HLA2008 p 16) Year Annual Ethylene oxide emitted ( ≈ used), kg Total No. Steriliser Runs Average Ethylene oxide Load Size (kg) 2002 5120 253 20.2 2003 5760 285 20.1 2004 5593 324 17.3 2005 6894 349 19.8 2006 9246 417 22.2 2007 6607 (YTD) 308 21.5

240 These figures accorded with Mr Kolding's recollection of the amount of EtO used by Unomedical in its sterilisation process and with what Mr Kolding told Mr Ives during the inspection by Mr Ives of the facility in November 2006 ("probably about 15-20 kg is used and then released into the air").

Such Practicable Means Existed at the Time to Minimise the Air Pollution and Were Necessary

242 I also accept and find to the requisite standard that equipment for effectively controlling and minimising the emission of EtO into the atmosphere, developed in the United States and Europe, was in existence during the charge period.

243 In particular, the catalytic abatement system installed by Unomedical in about July or August 2007 could have been installed. This system reduced the amount of EtO vented into the atmosphere at the end of each sterilisation cycle by over 99%. It was Mr Kolding's evidence that cost played no part in the decision by Unomedical not to install this type of abatement system any earlier. Rather, it was the belief held by him that it was not necessary to do so because it was not required by law, that delayed its implementation.

244 In addition to the installation of an abatement system, it was Mr Court's evidence that Unomedical could have implemented the following minimising methodologies in the sterilisation process: (a) absorption (mixing the EtO with liquid, usually water, to dissolve the constituents of the gas), which could have resulted in a 99% removal of the gas;

(b) absorption with chemical reaction (scrubbing the EtO with aqueous acid or alkali to achieve hydrolysis to ethylene glycol), which could also achieve a 99% removal rate;

(c) absorption with oxidation, which could achieve a 99% removal of EtO;

(d) direct flame oxidation (controlling of EtO by thermal oxidation or combustion), which would result in atmospheric emissions of carbon dioxide and water vapour resulting in a 99% removal of EtO upon discharge;

(e) condensation (using refrigeration or chilling equipment), which had a removal rate of between 50-85% of EtO; and

(f) reaction with solids (the removal of EtO by reaction with chemical solids held in a porous bed, usually used in conjunction with acid scrubbing), which claimed efficiencies of up to 99% in the removal of EtO.

245 Of the available technologies, it was Mr Kolding's evidence that a catalytic abatement system was the most efficient and preferable means of reducing the discharge of EtO produced by the sterilisation process.

246 While the literature and learning relied upon by Mr Court disclosed that varying standards were imposed overseas for the removal or minimisation of EtO where the product was used as a commercial sterilant, it was not in dispute that because of the size of its operation had Unomedical been operating the facility in the United States, Canada or, more importantly, where its head office was located, namely, Denmark, during the charge period, the installation of air pollution control equipment to reduce EtO emissions would have been mandatory.

247 Furthermore, had the premises been scheduled premises in New South Wales, Unomedical would have been compelled to reduce its ground level EtO emission concentrations from a predicted concentration of 8.3 ug/m3 (based on the HLA modelling) to 3.3 ug/m3. In Victoria, pursuant to the 2005 Approved Methods document, the design ground level concentration of EtO emissions could not exceed 3.3 ug/m3, and moreover, the emission of EtO had to be controlled to the "maximum extent achievable". At a Commonwealth level, the National Code recommended that all exhaust gases from a steriliser, including EtO, should, albeit in the context of workplace health and safety regulations, be scrubbed or that consideration should be given to the installation of a catalytic converter or incinerator.

248 Ultimately, it was conceded by Mr Kolding that practicable means, more specifically an abatement system, could have been implemented during the charge period to minimise the amount of EtO discharged into the atmosphere by the stack emissions. But, according to the evidence of Mr Kolding, it was Unomedical's belief that the installation of an abatement system or other "such practical means" supplementing those already in existence at the facility were not "necessary" because there were no prescribed discharge limits for EtO and there was no proven harm to human health. This belief was maintained notwithstanding that, according to Mr Kolding, Unomedical employed international experience and knowledge as a guide to what was "best practice" in Australia and its published policies referred to the essential "sharing of 'best practice' knowledge across business units and geographical borders".

249 From the evidence it is plain that Mr Kolding knew the following: (a) that he was not in a position to dispute Mr Court's expert evidence concerning the inefficiencies of the minimisation methods used by Unomedical at the facility in respect of the EtO;

(b) that he reported to the CEO of Unomedical Denmark, Unomedical Australia and Unomedical "everywhere";

(c) that he was aware of the overseas literature and learning and "what the requirements there are in countries by countries" for EtO abatement;

(d) that in Denmark EtO emissions were controlled by an abatement system;

(e) that had Unomedical been operating in Denmark or the United States it would have been required to implement further measures to reduce its EtO emissions;

(f) that had Unomedical been required to reduce the EtO emission levels to achieve a 99% efficiency of EtO use, then he "would accept that is the best practice", namely, the installation of a catalytic converter;

(g) that it was practicable for Unomedical to have installed a catalytic converter in 2001 when it was establishing the facility at Wilmette Place; and

(h) that in 2007 an abatement system was installed at the facility that had been sourced from one of Unomedical's factories in Denmark.

250 I therefore find beyond reasonable doubt that Mr Kolding was aware that there were practicable means, in particular the installation of a catalytic converter, that could have been implemented to reduce Unomedical's EtO emissions into the air, and moreover, I similarly find that had Unomedical been operating in the United States or Denmark, some form of abatement system would have been compulsory and that this was known to Mr Kolding during the charge period. 251 By making these findings no burden has been imposed on Unomedical to implement 'best' practicable means. To do so would be, as Unomedical correctly submitted, to impose a requirement beyond the scriptures contained in s 128(2). A duty to conduct activities or operations by employing the best practicable means to minimise air pollution would impose a much higher obligation on the occupiers of premises. It suggests a narrower operation of any concept of reasonableness, for example, in the cost of implementing the means.

252 Accordingly, when regard is had to the evidence adduced in this case, I find beyond reasonable doubt that such practicable means as were necessary ought to have been implemented by Unomedical to minimise the emissions of EtO from the stack and which were not in breach of s 128(2) of the POEOA. This finding is made after a consideration of all the evidence and submissions put before the Court, but particular regard was had to the fact that: (a) Mr Kolding was aware of the toxicity of EtO and the carcinogenic risk it posed to human health;

(b) Mr Kolding's belief that the emissions from the stack were not harmful was not sustainable absent any testing to this effect, or any measurement of the amount of EtO discharged by Unomedical into the atmosphere;

(c) while no actual harm was proven to the residents surrounding the facility, Professor Stewart nevertheless advised that environmental exposure to EtO should be minimised;

(d) albeit not uniform, but of which Mr Kolding was largely aware, the literature and learning, from both overseas and domestically, indicated that a number of countries, including Denmark, where the head office and the CEO of Unomedical to which Mr Kolding reported were located, required some form of abatement system to be installed to limit EtO emissions;

(e) there were no practical or financial impediments to the installation of an abatement system, such as the catalytic converter installed at the facility in 2007, during the charge period; and

(f) Mr Kolding's belief that Unomedical was at all times operating within the regulatory framework applicable to the facility cannot be maintained given his stated ignorance of s 128(2) of the POEOA. 253 In short, Unomedical submitted that absent any standards or limits prescribed for the environmental emission of EtO and absent any evidence that the emissions were "not unsafe", it was not "necessary" to implement any measures to minimise the emission. As is apparent from the discussion and findings made above, these 'absences' have not precluded the imposition of criminal liability upon Unomedical pursuant to its breach of s 128(2) of the POEOA.

Defences Honest and Reasonable Mistake of Fact 254 In reliance on this defence, Unomedical argued that at the time it commenced its operations at Wilmette Place, Mr Kolding held the positive but mistaken belief that the amount of EtO which was being emitted from the stack was minimal and would not harm human health or the environment and that there were no emission controls on the discharge of EtO into the atmosphere from the facility.

Principles and Elements of the Defence 255 In Proudman v Dayman (1941) 67 CLR 536 (at 540) Dixon J described the defence of honest and reasonable mistake of fact in these terms: As a general rule an honest and reasonable belief in a state of facts which, if they existed, would make the defendant's act innocent affords an excuse for doing what would otherwise be an offence. 256 Mere ignorance of a relevant fact is insufficient to establish the defence. So too is a lack of knowledge or the lack of a positive belief in certain facts.

257 In Proudman Dixon J explained the burden of proof as (at 541): The burden of establishing honest and reasonable mistake is in the first place on the defendant and he must make it appear that he had reasonable grounds for believing in the existence of a state of facts, which, if true, would take his act outside the operation of the enactment and that on those grounds he did so believe.

258 Once raised, the prosecution must then prove the absence of that belief, beyond reasonable doubt (He Kaw Teh v R (1985) 157 CLR 523 at 582 and CMT v R (2008) 236 CLR 440 at [8] and [35]).

259 The defence does not apply to mistakes of law (Thomas v The King (1937) 59 CLR 279). This is because, as the age old maxim states, ignorance of the law is no excuse (Martindale v Falkner (1846) 2 CB 706 at 719-720; 135 ER 1124 at 1129-1130 and Ostrowski v Palmer (2004) 218 CLR 493 at [1]). Moreover, "a mistake as to the existence of a compound event consisting of law and fact is in general one of fact and not a mistake of law" (Thomas at 306).

260 But determining whether a mistake is one of law or fact is notoriously difficult (Iannella v French (1968) 119 CLR 84 at 114 and Australian Fisheries Management Authority v Su (2009) 176 FCR 95 at [23]).

261 The defence has been raised in this Court with mixed success (for example, State Rail Authority of New South Wales v Hunter Water Board (1992) 28 NSWLR 721; Environment Protection Authority v Milpharma [1994] NSWLEC 189; Valsane Pty Ltd v Environment Protection Authority [1996] NSWLEC 265; Morrison v ANL Ltd (1996) 91 LGERA 437; Environmental Earth Sciences Pty Ltd v Environment Protection Authority (1999) 103 LGERA 434; Environment Protection Authority v Brazel (2002) 119 LGERA 72 and Hawkesbury City Council v Johnson; Hawkesbury City Council v Johnson Property Group Pty Limited [2008] NSWLEC 138).

262 From the above principles and after a survey of the authorities in this Court, the following elements were identified by Unomedical, which I adopt, as necessary to successfully establish the defence of honest and reasonable mistake of fact. They are: (a) a positive belief in a state of facts which, if they existed, would make the act innocent and take the act outside the operation of the statute ( Proudman at 540 and State Rail Authority at 723);

(b) the act must be based on a reasonable belief ( State Rail Authority at 725);

(c) the mistake must not be one of law ( Ostrowski at [28]). This is because ignorance of the law is no excuse ( CTM at [7]);

(d) a mixed mistake of fact and law will generally be a mistake of law ( Ostrowski at [87]);

(e) the defendant has the evidentiary burden of raising the defence ( He Kaw The at 535 and 551-555 and CTM at [8] and [35]);

(f) the prosecutor must then negative the defence beyond reasonable doubt ( He Kaw The at 539 and CTM at 725-726);

(g) a lack of knowledge or mere ignorance is insufficient to establish the defence ( Proudman at [57] and State Rail Authority at 725-726);

(h) there must be more than a general understanding that everything is in order ( State Rail Authority at 725-726 and Environmental Earth Sciences at [57]);

(i) the mistaken belief must be sufficiently specific to relate to elements of the offence in question ( State Rail Authority at 725-726);

(j) an absence of fault or an absence of negligence is not an element of the defence ( Brazel at [68]); and

(k) the ground of exculpation is a narrow one ( Von Lieven v Stewart (1990) 21 NSWLR 52 at 65).

Any Mistake Made by Unomedical Was Neither One of Fact Nor Was it Reasonable

263 As referred to in part above, Unomedical submitted that the evidence demonstrated that Mr Kolding, as Unomedical's managing director, held three positive but mistaken beliefs, namely, that: (a) the EtO being emitted from the stack was minimal and would not harm human health or the environment;

(b) there were no emission limits or controls on the amount of EtO that could be discharged from the stack; and

(c) Unomedical was making all necessary attempts to minimise the emissions of the EtO. 264 Mr Kolding stated that these beliefs had been formed, first, as a result of conversations that Mr Kanda had had with the EPA in or about 2001 that unless Unomedical's use of EtO exceeded 10 tonnes per annum there was no need for an abatement system to be installed. Second, his own knowledge, research and experience in the medical devices manufacturing industry. Third, the fact that the council had approved Unomedical's development application to operate a medical devices manufacturing plant at Wilmette Place without imposing any conditions to minimise or prevent the emission of EtO into the air, and fourth, his communications with Mr Doig.

265 In my opinion, Unomedical cannot avail itself of the defence. The reasons are three-fold. First, in essence Unomedical's mistake is, when properly analysed, one of law insofar as it mistakenly believed that there being no prescribed limit on the amount of EtO it could emit into the atmosphere, it was not under any duty to mitigate its emissions by the installation of an abatement system. It was Mr Kolding's repeated evidence that because there was no mandated limit on the amount of EtO the facility could discharge into the air, and because the installation of a catalytic converter, or some other form of abatement system, was not required by law, that at all times the operation of the facility by Unomedical was lawful.

266 I accept the EPA's submission that ignorance of the law, and in particular, ignorance of s 128(2) of the POEOA, impermissibly forms the basis of Unomedical's attempt to avail itself of this defence. As Mr Kolding conceded he was not aware of the general obligation imposed by s 128(2), or indeed the provision itself, either before, or during the charge period. He could not be mistaken about that which he was not aware.

267 Second, the facts about which Unomedical contended it was mistaken, again properly characterised, amounted to no more than opinions, beliefs or assumptions which were not sufficient to exonerate Unomedical from criminal liability.

268 In R v Lavender (2005) 222 CLR 67 Gleeson CJ, McHugh, Gummow and Hayne JJ stated (at [57] and [59], with whom Kirby J agreed at [83], likewise Callinan J at [139]-[141] and Heydon J at [148]): [57] The first issue concerns a point that was taken at trial by counsel for the respondent. In considering the issue, it is necessary to note the precise terms of counsel's submission to the trial judge. Reference has earlier been made to the five elements of manslaughter identified by the judge. Counsel said: "I would invite your Honour to add, in relation to the manslaughter, a sixth element, being that the accused did not hold an honest and reasonable belief that it was safe to proceed." The invitation was declined. … [59] The second reason is that the principle on which counsel based his argument, which applies in other contexts, is a principle relating to honest and reasonable mistake of fact. The principle was recently discussed in this Court in Ostrowski v Palmer (50). As the decision in that case illustrates, the principle concerns mistakes of fact. The belief concerning which counsel sought a direction was a (supposed) "belief that it was safe to proceed". Such a state of mind involves an opinion. It might be based upon certain factual inferences or hypotheses (the respondent did not give evidence, so the jury were not told by him exactly what facts or circumstances were operating in his mind), but it necessarily involves an element of judgment. Indeed, it involves a conclusion by the respondent that his conduct was reasonable. The direction sought would be inconsistent with what has been described as the objectivity of the test for involuntary manslaughter. The respondent's opinion that it was safe to act as he did was not a relevant matter. If there had been some particular fact or circumstance which the respondent knew, or thought he knew and which contributed to that opinion, and the jury had been informed of that, and counsel had asked for a direction about it, then it may have been appropriate to invite the jury to take that into account (51).

269 Not dissimilarly, Mr Kolding believed that it was lawful to operate the facility without an abatement system minimising the emission of EtO. This was because he, first, believed that there was no harm being caused to local inhabitants by the discharge of the EtO. Second, he assumed that the consent granted by the council in relation to the first DA without any conditions limiting the discharge of EtO meant that uncontrolled emission of the substance from the stack was lawful. Third, he believed that provided the amount of EtO utilised by the sterilisation process was less than 10 tonnes a year, no controls were necessary. And fourth, he believed that there were no prescriptive standards limiting the amount of EtO that could be discharged via the stack. These were mistaken beliefs, however, and not mistaken facts.

270 It follows that none of the so-called facts raised by Unomedical to ground the defence are capable of satisfying Unomedical's evidentiary onus and there is nothing for the EPA to negative.

271 However, if this is incorrect and these beliefs and assumptions did not constitute mere opinions but were facts about which Unomedical was mistaken, then, third, they were not, in any event, reasonable.

272 In my opinion, it was not reasonable for Unomedical to be mistaken about the level of harm caused to the environment (as this concept has been discussed above) absent any knowledge or testing to establish the level of EtO emitted into the atmosphere during the charge period, or any modelling to establish its dispersion. It was not reasonable for Mr Kolding to assume that EtO was not harmful to humans absent a proven study demonstrating this fact given his recognition and acknowledgement of the carcinogenicity and toxicity of EtO.

273 Moreover, to the extent that Mr Kolding relied on the May 2002 EES report as justification of his belief in the 'safety' of the EtO emissions from the facility, this report was neither able to be produced to the council despite numerous requests nor was it received by Mr Kolding during the charge period (it was received in Denmark). The report was therefore irrelevant to any belief formed by Mr Kolding for the purpose of Unomedical's exculpation.

274 It was also not reasonable, in my view, to rely on a consent given by the council when the council was not made aware of the amount of EtO being discharged into the atmosphere, or more critically, the fact that EtO was being emitted into the air at all. This ignorance was confirmed by the council's Environmental Audit Inspection Report prepared by Ms Daly which noted, on the basis of what she had been told and what she had observed during her inspection of the facility, that there was "no evidence of air emissions". Although the council was the regulatory authority for Unomedical as non-scheduled premises under the POEOA, I nevertheless do not accept as reasonable any misapprehension by Unomedical as to whether or not it was required by law to implement measures to minimise its EtO emissions as a consequence of any of the development consents granted by the council for use of the facility.

275 I also do not find reasonable any mistake as to the fact that because Unomedical was utilising less than 10 tonnes of EtO per annum in its sterilisation process that it was not required to abate any unconsumed discharge of the substance into the atmosphere or report this fact to the council. The evidence from Mr Kolding was that this belief was based on: …the advice from EPA which I was told by Mr Narinder [Kanda] that the 10 tonnes was the critical number, as long as you were below 10 tonnes there was no requirement for you to put any controls in and you will be regulated by the council and Unomedical back at that time used only about 5.5 tons per year.

276 For the reasons already given, I do not accept as reliable the evidence Mr Kolding gave of this conversation. In addition, it was information that was not volunteered to Ms Daly when she undertook her inspection and audit of the facility, it was information that was not mentioned during Mr Kolding's interview with the EPA and it was information that was contradicted in cross examination by Mr Kolding when he stated in evidence: And I haven't said that the 10 tonnes from the EPA was mentioned from the EPA. I said our assumptions were, when we heard the 10 tonnes, that that would come from that report. 277 At its highest, the 10 tonne threshold was no more than an unverified assumption on the part of Unomedical.

278 The EPA correctly, in my opinion, accepted that Mr Kolding's beliefs could be equated to that of Unomedical given Mr Kolding's position of managing director (State Rail Authority at 723-724 and Tesco Supermarkets Ltd v Nattrass [1972] AC 153). However, the EPA submitted that any belief attributed to Mr Kanda and relied upon by Mr Kolding could not be taken into account for the purpose of establishing this defence. Given my adverse finding about the reliability of the communications between Mr Kanda and Mr Kolding, it is strictly not necessary for me to decide this issue. Having said this, I do not consider a managing director's reliance on the information provided to him or her by an operations manager with the requisite expertise and skill and whose duties include knowledge of this type of information, to be wholly unreasonable.

279 Finally, I do not accept as reasonable any mistake Mr Kolding contended that he had made that no abatement system was required to lawfully operate the facility based on his knowledge that the premises were not scheduled and that no emission standards applied to the external discharge of EtO. While the fact that no defined discharge limits applied to the facility was true, this did not, in my opinion, render reasonable any belief by Mr Kolding that no further steps to minimise air pollution from the facility were necessary absent any inquiries by him to this effect. To the extent that Mr Kolding made inquiries of Mr Doig, these were informal in character and were not verified by, as Mr Doig suggested, an independent air pollution consultant (such as HLA) engaged by Unomedical to provide this advice.

280 Finally, I do not accept as reasonable the stated belief by Mr Kolding that he was taking all, or such practicable, means necessary to minimise the emission of EtO. This is because, first, he did not know how much EtO was being emitted, and second, he knew that an abatement system could easily have been installed that would have been considerably more efficacious at reducing the discharge of EtO.

281 It follows that I find that Unomedical cannot avail itself of the defence of honest and reasonable mistake of fact because, first, not only has Unomedical failed to raise any mistakes of fact capable of satisfying its evidentiary onus in raising the defence, but second, even if it had, I find that none were based on reasonable grounds.

Officially Induced Error of Law 282 Unomedical further relied on the defence or excuse of officially induced error of law. It did so notwithstanding that, as it readily accepted, to date the defence does not form part of the common law of Australia.

283 In the Canadian Supreme Court case of R v Jorgensen [1995] 4 SCR 55, the accused were convicted of an offence of knowingly selling obscene material. Lamer CJ held that the accused were to be excused from conviction on the basis of an officially induced error of law by virtue of the Ontario Film Review Board's approval of the films in question (at [2]). While this did not negative the mens rea of the offence, his Honour stated that reasonable reliance on this type of official advice was a sufficient basis for a judicial stay of proceedings to be entered. In so concluding Lamer CJ emphasised that a stay could only be entered in the clearest of cases and that the excuse did not offend the maxim that ignorance of the law was no excuse, rather, it provided an exception to this provision "which ensures that the morally blameless are not made criminally responsible for their actions" (at [2]). 284 After detailing the development of the defence in the United States and Canada (at [12]-[24]), Lamer CJ went on to describe its elements in these terms (at [28]-[36]): 28 The first step in raising an officially induced error of law argument will be to determine that the error was in fact one of law or of mixed law and fact. Of course, if the error is purely one of fact, this argument will be unnecessary. Unlike Professor Barton, I do not agree that officially induced error should be used to eradicate the distinction between mistakes of fact and mistakes of law. This distinction is important for all the reasons that I believe the principle that ignorance of the law does not excuse must stand firm. Distinguishing between mistakes of fact and those of law remains conceptually important. Mistakes of law will only be exculpatory in narrowly defined circumstances. 29 Once it is determined that the error was one of law, the next step is to demonstrate that the accused considered the legal consequences of her actions. By requiring that an accused must have considered whether her conduct might be illegal and sought advice as a consequence, we ensure that the incentive for a responsible and informed citizenry is not undermined. It is insufficient for an accused who wishes to benefit from this excuse to simply have assumed that her conduct was permissible. 30 The next step in arguing for this excuse will be to demonstrate that the advice obtained came from an appropriate official. One primary objective of this doctrine is to prevent the obvious injustice which O Hearn Co. Ct. J. noted -- the state approving conduct with one hand and seeking to bring criminal sanction for that conduct with the other. In general, therefore, government officials who are involved in the administration of the law in question will be considered appropriate officials. I do not wish to establish a closed list of officials whose erroneous advice may be considered exculpatory. The measure proposed by O Hearn Co. Ct. J. is persuasive. That is, the official must be one whom a reasonable individual in the position of the accused would normally consider responsible for advice about the particular law in question. Therefore, the Motor Vehicle Registrar will be an appropriate person to give advice about driving offences, both federal and provincial. The determination of whether the official was an appropriate one to seek advice from is to be determined in the circumstances of each case. 31 My colleague Sopinka J. argues that, in this case, allowing OFRB approval to constitute an excuse is an impermissible delegation of power from one level of government to another. In my view, this argument does not bar officially induced error from constituting an excuse which is considered after culpability has been proven. There is no issue of the action of a provincial board precluding criminal prosecutions. Indeed, we must recall that it is the provincial Attorney General who makes the decision to prosecute offences, even when the offence charged is in the Criminal Code . The advice of officials of any level of government may induce an error of law and trigger this provision provided that a reasonable person would consider that particular government organ to be responsible for the law in question. The determination relies on common sense rather than constitutional permutations. 32 Several other types of advice have been considered in the cases dealing with this excuse, for example advice from private lawyers and reliance on judicial pronouncements. As these examples are beyond the scope of this case, I make no comment at this time on whether these types of advice may provide the basis for officially induced error of law. 33 Once an accused has established that he sought advice from an appropriate official, he must demonstrate that the advice was reasonable in the circumstances. In most instances, this criterion will not be difficult to meet. As an individual relying on advice has less knowledge of the law than the official in question, the individual must not be required to assess reasonableness at a high threshold. It is sufficient, therefore, to say that if an appropriate official is consulted, the advice obtained will be presumed to be reasonable unless it appears on its face to be utterly unreasonable. 34 The advice obtained must also have been erroneous. This fact, however, does not need to be demonstrated by the accused. In proving the elements of the offence, the Crown will have already established what the correct law is, from which the existence of error can be deduced. Nonetheless, it is important to note that when no erroneous advice has been given, as in MacDougall , supra , this excuse cannot operate. 35 Finally, to benefit from this excuse, the accused must demonstrate reliance on the official advice. This can be shown, for example, by proving that the advice was obtained before the actions in question were commenced and by showing that the questions posed to the official were specifically tailored to the accused's situation. 36 In summary, officially induced error of law functions as an excuse rather than a full defence. It can only be raised after the Crown has proven all elements of the offence. In order for an accused to rely on this excuse, she must show, after establishing she made an error of law, that she considered her legal position, consulted an appropriate official, obtained reasonable advice and relied on that advice in her actions. Accordingly, none of the four justifications for the rule that ignorance of the law does not excuse which Stuart outlined is undermined by this defence. There is no evidentiary problem. The accused, who is the only one capable of bringing this evidence, is solely responsible for it. Ignorance of the law is not encouraged because informing oneself about the law is a necessary element of the excuse. Each person is not a law unto himself because this excuse does not affect culpability. Ignorance of the law remains blameworthy in and of itself. In these specific instances, however, the blame is, in a sense, shared with the state official who gave the erroneous advice.

285 Thus to summarise, the constituent elements of the defence of officially induced error of law are that: (a) an error of law or of mixed fact or law was made;

(b) the defendant who committed the act giving rise to the commission of the offence considered the legal consequences of his or her actions prior to committing the act;

(c) advice was sought by the defendant and was obtained from an appropriate official;

(f) the defendant relied on the advice in committing the act.

286 Applying these elements to the facts of the case, Lamer CJ held that if the Crown had proven that the accused possessed the requisite mens rea, which it had not, thereby necessitating an acquittal, then their conduct would have nevertheless been exculpated on the basis of the officially induced error of law that the films were not legally obscene by reason of their approval by the Board.

287 The remainder of the Court (La Forest, L'Heureux-Dube, Sopinka, Gonthier, Cory, McLachlin, Iacobucci and Major JJ) acquitted the accused on the basis that the Crown had failed to prove that they held the necessary mens rea. In relation to the issue of the availability of officially induced error of law as a defence in the appeal, Sopinka J (who delivered judgment for the remainder of the Court) stated that because the matter was not raised either before that Court or the court below, it ought not be considered. But his Honour was at pains to state that, "nothing in these reasons should be taken as agreeing or disagreeing with the reasons of the Chief Justice" (at [125]).

288 In Maitland Valley Conservation Authority v Cranbrook Swine Inc (2002) 225 DLR (4th) 455, the issue before the Ontario Court of Appeal was whether persons who had been granted a building permit could rely on the defence when the resulting construction violated the provisions of another enactment protecting wetlands. In that case, the defendants were granted a building permit upon application to construct two barns and a liquid manure storage tank. They hired a consultant to handle the building application. The Maitland Valley Conservation Authority prosecuted the defendants on the basis that the construction was on swamp or wetland, which was unlawful pursuant to another statute. There was no evidence that the building permit was invalid.

289 Applying Jorgensen, the majority held that there was insufficient evidence to determine if the defence had been established because no reliance could be demonstrated. The defendants neither considered the legal consequences of building the barns and the liquid manure tank in a swamp nor did they obtain advice from an appropriate official pursuant to that consideration. That is to say, that the construction complied with the statute they had in fact infringed (at [60] and [71]). In short, they demonstrated ignorance, and not mistake occasioned by inducement, of the law.

290 A new trial was ordered and when the matter was remitted to the Ontario Court of Justice ([2004] OJ No 5724) the Court held that while the issuing of the permit was not a complete defence to the charges before the Court (at [74]), nevertheless the defendants were able to establish the defence of officially induced error of law. This was because the evidence of the chief building official was that when he had received the application for the building permit he had attended the site and in his mind the proposed location of the manure storage tank had not raised any concerns under the statute in question and therefore he did not direct the defendants' attention to it. The Court held that this evidence clearly established that the defendants had relied upon the permit as a permission to build the manure tank in the particular location upon which it had been built and that such reliance was reasonable (at [84]). To hold to the contrary would, said the Court, "throw the entire mandatory scheme for the issue of building permits under the Building Code Act into disarray….It would be virtually impossible for construction to this province to proceed if one could not rely on a valid building permit" (at [84]).

291 Unomedical submitted that Cranbrook Swine was analogous to, if not on all fours with, the facts of the present case.

292 Lamer CJ's reasons in Jorgensen were subsequently endorsed by the Supreme Court of Canada in Lévis (City) v Tétreault; Lévis (City) v 2629-4470 Québec Inc [2006] 1 SCR 420 (Lebel J, McLachlin CJ and Bastaraceh, Binnie, Fish, Abella and Charron JJ concurring).

293 In Lévis the Court referred to the defence as discussed in Jorgensen by Lamer CJ and observed that (at [22] and [27]): 22 This Court has firmly and consistently applied the principle that ignorance of the law is no defence. It has given effect to this principle not only in the context of the criminal law itself, but also in cases involving regulatory offences ( Molis v. The Queen , [1980] 2 S.C.R. 356; Pontes ). However, the inflexibility of this rule is cause for concern where the error in law of the accused arises out of an error of an authorized representative of the state and the state then demands, through other officials, that the criminal law be applied strictly to punish the conduct of the accused. In such a case, regardless of whether it involves strict liability or absolute liability offences, the fundamental fairness of the criminal process would appear to be compromised. Although the Court has not ruled on this point, Lamer C.J. responded to these concerns, in concurring reasons in R. v. Jorgensen , [1995] 4 S.C.R. 55, by proposing to recognize the defence of officially induced error and attempting to define the conditions under which the defence would be allowed. … 27 Although the Court did not rule on this issue in Jorgensen , I believe that this analytical framework has become established. Provincial appellate courts have followed this approach to consider and apply the defence of officially induced error ( R. v. Larivière (2000), 38 C.R. (5th) 130 (Que. C.A.); Maitland Valley Conservation Authority v. Cranbrook Swine Inc. (2003), 64 O.R. (3d) 417 (C.A.)). I would also note that, in this appeal, neither the prosecution nor the intervener, the Attorney General of Canada, has questioned the existence of this defence in Canadian criminal law as it presently stands. At most, the Attorney General of Canada has suggested another condition in addition to those enumerated by Lamer C.J., namely that the act was committed contemporaneously with the reception of the information. I do not think this addition is necessary. The Attorney General of Canada's concerns relate more to the need to demonstrate that the advice was reasonable and that the accused relied on it. It should be noted, as the Ontario Court of Appeal has done, that it is necessary to establish the objective reasonableness not only of the advice, but also of the reliance on the advice ( R. v. Cancoil Thermal Corp. (1986), 27 C.C.C. (3d) 295; Cranbrook Swine ). Various factors will be taken into consideration in the course of this assessment, including the efforts made by the accused to obtain information, the clarity or obscurity of the law, the position and role of the official who gave the information or opinion, and the clarity, definitiveness and reasonableness of the information or opinion ( Cancoil Thermal , at p. 303). It is not sufficient in such cases to conduct a purely subjective analysis of the reasonableness of the information. This aspect of the question must be considered from the perspective of a reasonable person in a situation similar to that of the accused.

294 The facts of Lévis were as follows. The defendant company was charged with operating a motor vehicle for which its registration fees had not been paid. In response, the defendant raised the defence of officially induced error of law and the defence of due diligence, submitting that a representative of the Société de l'assurance automobile du Québec had earlier made the company pay registration fees for a 15 month period and had told the company that a renewal notice would be sent to it before the period of expiry. Due to an administrative error on the part of the Société (the address to the defendant was incorrect), the renewal notice sent by the Société to the defendant was returned.

295 The Crown's appeal was allowed. Relevantly, the defence of officially induced error of law, although available, could not be established on the facts because the error by the Société related to, at the most, an administrative practice and not to any legal obligation to pay the fees by the prescribed date. Moreover, the company could not have considered the legal consequences of its conduct on the basis of advice from the official in question nor could it have acted in reliance on that opinion since no information regarding the nature and effects of the relevant legal obligations had been requested or obtained by it (at [30]-[34]).

296 Unomedical was not able to point to any English authorities where the defence, as accepted by the Canadian decisions, has been applied. In R v Dodsworth (1837) 8 Car and P 218; 173 ER 467 the accused was charged with falsely stating electoral qualifications. He had acted on the advice of an election committee. Lord Denman stated that the accused should not be convicted if it could be proved that he had acted bona fide and had "been guided in his conduct in a matter of law by persons who are conversant with the law". However, properly analysed the decision stands for no more than, in my opinion, the suggestion that evidence of erroneous legal advice may displace the necessary mens rea that a prosecutor must prove where the commission of an offence requires proof of a particular state of mind.

297 While it has been stated that the doctrine might apply in New Zealand, the courts there have yet to unequivocally endorse the defence and have tended to find that on the evidence it could not be established in any event (Tipple v Police [1994] 2 NZLR 362, MacRae v Buller District Council [2005] NZHC 378 at [39]-[41] and Wilson v Auckland City Council [2007] NZHC 287 at [30]-[47]).

298 It is certainly the case that in Ostrowski the High Court did not dismiss the availability of the defence of officially induced error of law. This may have been because, as Unomedical submitted, at the time that Ostrowski was decided the decision of the Supreme Court of Canada in Lévis had not been handed down and the defence had not been entrenched in that country.

299 Unomedical relied on comments by Callinan and Heydon JJ in Ostrowski (at [61]) to the effect that the facts of this prosecution were an example of how "mandatory penalties can operate harshly and unfairly" and that "in the imposition of harsh mandatory penalties, has the appearance of an act of mindless oppression" (at [70]).

300 Unomedical also submitted that an acceptance of the availability of the defence was consistent with the presumption of regularity that official acts are presumed to have been done rightly and regularly (Morris v Kanssen [1946] AC 459 at 475, McLean Bros and Rigg Ltd v Grice (1906) 4 CLR 835 at 849-851, Minister for Natural Resources v New South Wales Aboriginal Land Council (1987) 9 NSWLR 154 at 163-164, Selby v Pennings (1998) 102 LGERA 253 at 261-262, Hill v Woolhara Municipal Council (2003) 127 LGERA 7 at [51]-[52], Lane Cove Council v Minister for Urban Affairs and Planning (2005) 140 LGERA 185 at [78] and GPT RE Ltd v Belmorgan Property Development Pty Ltd (2008) 72 NSWLR 647 at [78]-[84]).

301 The presumption has been enshrined in legislation, particularly with respect to environmental planning instruments (s 45 of the Interpretation Act). And while the decisions have been divergent, the better view on the authorities is that the presumption can apply to criminal cases, albeit on a more restrictive basis (see the review of the authorities in Selby at 262-266 per Ipp J).

302 In my opinion, however, the presumption is of little assistance to Unomedical. This is because, first, the presumption generally only applies in circumstances where there is a challenge to the performance of a condition essential to the validity of an administrative act or decision upon which an offence depends. No such challenge is mounted here. There is no suggestion that any of the development consents granted to Unomedical were not valid. Second, the presumption has been said not to apply to the substance of development consents (Belmorgan at [82]).

303 Third, the presumption is no more than a rebuttable presumption of fact (Northside Developments Pty Ltd v Registrar-General (1990) 170 CLR 146 at [177]) and as the evidence presently demonstrated, it had been rebutted. For example, and as discussed further below, Unomedical never made a development application that contained an adequate Statement of Environmental Effects detailing the discharge of EtO into the atmosphere, no plan was provided to the council showing the emission stack, no process diagram was provided to the council and the council was never told of the frequency and quantity of the EtO emissions from the facility.

304 Fourth, to the extent that Unomedical sought to draw a comparison with, or make an analogy to, the existence of the presumption in support of the application of the defence of officially induced error of law in Australia, it ignored the fact that Australian courts are generally not supportive of the use of estoppel in administrative law (see the discussion and authorities referred to in Casa v City of Ryde Council (2009) 172 LGERA 348 at [109]-[111] and Page v Manningham City Council (2010) 174 LGERA 48 at [61] and the authorities referred to at footnote 31 of that decision). To permit a defence to be available in Australia, which if established would operate as a stay on the proceedings, would be to, in effect, entrench a species of estoppel against administrative decision-makers.

305 While it is correct to say, as Unomedical did, that Ostrowski does not preclude the operation of the defence in Australia, a close examination of the reasons of that decision cannot give Unomedical the comfort it desires. For example in Ostrowski McHugh J stated (at [53]-[54]): 53 It is irrelevant that Mr Palmer's mistake was induced by the conduct of an employee of Fisheries WA. That conduct cannot convert a mistake as to the applicable law into a mistake of fact. If a defendant knows all the relevant facts that constitute the offence and acts on erroneous advice as to the legal effect of those facts, the defendant, like the adviser, has been mistaken as to the law, not the facts.

54 Four cases which address this issue are Olsen v Grain Sorghum Marketing Board ; Ex parte Olsen (89) , Loch v Hunter; Ex parte Loch (90), Cambridgeshire and Isle of Ely County Council v Rust (91) and Power v Huffa (92). In the first two cases each defendant claimed that he or she was acting on the erroneous advice of a third party - either a legal adviser or a government official - that the acts in question were legal and that this mistake amounted to a mistake of fact. In the last two cases each defendant claimed that he or she had been given lawful authority to act as charged and that this mistake likewise amounted to a mistake of fact. In each case the defendant's argument failed, the court finding that each defendant was acting under a mistake of law. Accordingly, the bare fact that the adviser or official may have been mistaken as to the state of the law does not convert the defendant's mistake into one of fact. Both the adviser or the official and the defendant operate under a mistake of law.

306 Callinan and Heydon JJ, while sympathetic to the plight of the defendant in Ostrowski, expressed a similar sentiment (at [84]-[85], emphasis added): 84 It is impossible not to sympathise with the respondent. On any fair and objective view he was not culpable in any way. To the contrary he was most diligent. He went to the office of the administering authority twice in order to ascertain what his obligations were. Entirely openly and strictly in accordance with his licence he sought to comply with his understanding of what he could do based on official information personally provided by officials.

85 Be that as it may, it is the task of this Court to apply the law by answering the question whether the respondent should be regarded merely as having been ignorant of the law, an excuse which s 22 of the Code would deny him, or whether he had an honest and reasonable, but mistaken, belief in the existence of a state of things which if they had in fact existed would have meant that he was not criminally responsible. The question is an important one. A mockery would be made of the criminal law if accused persons could rely on, for example, erroneous legal advice, or their own often self-serving understanding of the law as an excuse for breaking it, however relevant such matters might be to penalty when a discretion, unlike here, in relation to it may be exercised . 307 The defence of officially induced error of law has been raised on several occasions in this Court. First, in the context of a sentencing decision where in Environment Protection Authority v Fletcher (2001) 114 LGERA 187 (at [116]) Sheahan J went no further than to state that officially induced error may be a mitigating factor in determining an appropriate penalty, and second, in Environment Protection Authority v Goulburn Wool Scour Pty Limited [2003] NSWLEC 200 (at 118]) where pursuant to a finding of not guilty of the offence charged, Talbot J noted that "the concept has not, so far, been imported into Australian law, so far as this Court is aware".

308 Finally, the entreaty by Unomedical to recognise the defence of officially induced error of law in Australia bears more than a passing resemblance to the earlier invitation to the Court to accept the defence of due diligence as available in this country. Again, that defence had been recognised in Canada (R v City of Sault Ste Marie (1978) 85 DLR (3d) 161). In this Court, however, Cripps J rejected any application of the defence of due diligence at common law (State Pollution Control Commission v Australian Iron & Steel Ltd (1992) 74 LGERA 387). On appeal his Honour's decision was unanimously upheld in Australian Iron & Steel Pty Ltd v Environment Protection Authority [No 2] (1992) 79 LGERA 158 at 170-174. In my opinion, a similar result must follow with respect to the defence of officially induced error of law.

309 For these reasons I do not accept that the defence of officially induced error of law is a defence available to Unomedical at common law in Australia.

310 However, assuming for present purposes that this conclusion is incorrect, I have nevertheless considered, first, who bears the onus of proof in respect of the defence, and second, whether or not it can be established by Unomedical on the evidence before the Court.

311 Unomedical submitted that similar to the defence of honest and reasonable mistake of fact, once a defendant adduces sufficient evidence in support of the defence of officially induced error of law on the balance of probabilities, then it is for the prosecutor to negative the defence beyond reasonable doubt. The EPA rejected this contention, noting that the Canadian position with respect to onus in relation to this defence was that it was for the defendant to establish the elements of the offence on the balance of probabilities (see, for example, Cranbrook Swine at [58]).

312 In my opinion, the defence of officially induced error of law is sufficiently analogous in form and substance to the defence of honest and reasonable mistake of fact that to impose any different evidential and legal onus on the defendant could not be rationally justified. I therefore accept Unomedical's submission in this regard.

Unomedical Cannot Avail Itself of the Defence on the Evidence 313 In my opinion, the evidence underlying this prosecution renders incapable any reliance by Unomedical upon the defence of officially induced error of law and there is nothing for the EPA to negative. Even if this conclusion is wrong, then I am of the view that the EPA has successfully negatived the defence on the facts as found.

314 This is so notwithstanding that, on balance, I am prepared to accept (contrary to the arguments of the EPA) that Unomedical made an operative error of law and that the advice, if any, to the extent that it came from the council and not the EPA, came from an appropriate official. I make the latter finding conscious of the contrary position put by the EPA. However, the EPA's position cannot be correct in light of the fact that it was the council that was the regulatory authority under the POEOA in respect of the activities carried out by Unomedical at the facility.

315 In my opinion, Unomedical cannot exculpate itself by recourse to the defence of officially induced error of law because, first, Mr Kolding, and thus Unomedical, was, as he readily conceded, completely ignorant of s 128(2). There could, therefore, be no consideration by Unomedical of the legal consequences of its actions in operating the steriliser and emitting the EtO. That is to say, there could be no consideration of the legal consequences of not implementing such practicable measures as may have been necessary to reduce the emission of the EtO. Although Unomedical submitted that Mr Kolding had considered the use of EtO at the facility more generally within the regulatory framework, as the terms of the first and second development applications, including the Statement of Environmental Effects, evidenced, this was not specific enough.

316 I should add for the sake of completeness that because that I do not accept as reliable the conversation between Mr Kanda and Mr Kolding concerning the NPI 10 tonne threshold, it follows that I do not accept that this was an operative consideration by Mr Kolding of the consequences of Unomedical's activities in emitting the EtO.

317 Second, and although the council was an appropriate official for the purposes of the defence, the evidence discloses that Unomedical did not consult the council for advice. Unomedical consulted Mr Doig and EES but never the council as to whether it was lawful for it to discharge 99%, or indeed any, of the EtO it utilised in the course of the sterilisation processes into the atmosphere.

318 It is undeniable that, as Unomedical submitted, the council was aware of Unomedical's use of EtO in the sterilisation process and that it had several opportunities to regulate the emission of EtO by Unomedical. But not only does this presuppose knowledge by the council that EtO was being vented into the air from the stack, a fact which I find the council was not aware of, it misconceives the nature of the consultation.

319 In respect of the first DA to change the use of the premises at Wilmette Place to a medical manufacturing site, it is undeniable that, as Unomedical submitted, the council officers (Mr Nick Sutton and Ms Robina Warburton) assessing the first DA must have been aware of their responsibilities and the council's obligations under the POEOA when the DA was considered. However, this awareness did not transform the development application into a request for legal advice about the lawfulness of the sterilisation operation.

320 The same may be said of the modification application to the first consent submitted on 13 August 2001. The modifications related only to alterations to the office and factory layout, provisions of a shade structure and new entry to the outdoor recreational area, provision of a new gas tank enclosure, provision of a water storage tank, pump room, sprinkler system and changes to the carpark layout. It was silent in respect of the EtO emissions.

321 A similar conclusion may be drawn in respect of the development application to expand the capacity of the gas bottle enclosure area by BOS in October 2004 (the second DA). While it is accurate to note that the council, in approving the increased storage capacity, stated in a memorandum written by Ms Sarah Winnacotta dated 14 February 2005 that, "there is no evidence to suggest that the quality of the existing environment has been degraded as a result of the existing system…it is unlikely that environment (sic) will experience any change in environmental impacts", this assessment only concerned the expansion of gas tank storage facilities. Viewed accordingly, the opinions expressed in the memorandum were entirely unsurprising and conventional. The development application disclosed neither the frequency nor the amount of EtO likely to be vented into the air as a consequence of the amplification of the storage capacity.

322 Likewise, the environmental audit performed by Ms Daly for the council in November 2002. Although undertaken for the purpose of, amongst other things, providing information to the persons managing the activity "on compliance with legal requirements" (s 172 of the POEOA), the audit was conducted without the benefit of full disclosure by Unomedical of critical information concerning the use and emission by it of EtO. As a consequence, Unomedical cannot rely on the audit to constitute 'consultation' or 'advice' for the purpose of the defence.

323 Furthermore, to the extent that the audit required Ms Daly to form an opinion as to whether the business had any air emissions in the form of smoke, dust and other particulates, it is true that Ms Daly indicated that it did not and her summary of the audit disclosed no problems in this regard. But again, at no point was Ms Daly told that EtO was being vented into the air. The presence of the stack was not explicably solely by reference to a discharge point for the EtO (it could have been, as the EPA submitted, an emergency safety valve). Ms Daly's evidence was that because she was unfamiliar with the processes carried out at the facility and because did not know how the EtO stored at the facility was used, she relied upon what she was told and what she observed in expressing her opinions at the conclusion of the audit. In addition, Ms Daly did not see any visible air emissions, steam or smoke emanating from the facility. This is unremarkable given that the discharged EtO was both colourless and odourless. The evidence of Ms Daly was not challenged in this regard and I accept it.

324 Thus had the council been told of critical details concerning the use and discharge of EtO from the facility, then the absence of any prescribed conditions of consent regulating the EtO emissions may have constituted obtaining advice from the council that it was lawful to operate the facility without the implementation of an abatement system or the installation of other such practicable means to reduce the EtO vented into the air. But at the risk of repetition, neither the fact that EtO was being emitted through the stack into the air nor the amount of EtO being discharged was adequately conveyed to the council. As referred to above, no process diagram demonstrating the steriliser emissions from the stack was ever provided to the council. Indeed, the first DA did not even show that a stack was proposed to be used. The Statement of Environmental Effects accompanying it did not specify that 99% of the EtO used would be discharged from the facility into the atmosphere. Instead it merely stated that "no odours [are] generated or released" during the sterilisation process. In my view, this was wholly inadequate for the purpose of the defence.

325 It was not to the point to submit, as Unomedical did, that the council, as both the regulatory and consent authority, was under an obligation to assess the development applications against the criteria contained in s 79C(1) of the Environmental Planning and Assessment Act 1979 ("the EPAA"). In particular, to assess the applications against s 79C(1)(b) (the "impact of that development") and (1)(c) of that Act (the "suitability of the site for development"), and to modify through the imposition of appropriate conditions of consent, the air emissions from the facility. The absence of any such conditions of consent does not amount to a legal representation by the council that the activities carried out by Unomedical at the facility were lawful.

326 Equally irrelevant was the submission by Unomedical that a development application was not invalid merely because it was not accompanied by an adequate statement of environmental effects or a site plan. The validity of the development consents granted by the council or their compliance with the EPAA or the Environmental Planning and Assessment Regulation 2000 was not at issue.

327 To the extent that Unomedical placed reliance on the assessment and approval process by Fairfield City Council in January 1998 for the use of an EtO steriliser to be operated by Steritech - which resulted in that council imposing conditions of consent concerning the filtration of the EtO used and emitted during the sterilisation process - I did not find this comparison to be of assistance. Each prosecution turns on its own evidence and facts.

328 Ultimately, what was at issue was the quality of the information given to the council which formed the basis of any purported advice, even by omission, the council is contended to have given Unomedical. Absent full, or event adequate, disclosure of the EtO emissions, the council's silence cannot be relied upon by Unomedical as advice properly obtained by it as to the lawfulness of its activities at the facility.

329 I accept, as the evidence discloses (see the council document My Development Control Plan - 11 Wilmette Place, Mona Vale dated 29 October 2004), that the council was aware that it had an obligation to assess the impact of any proposed development against the POEOA. I also accept, as the evidence demonstrates (see, for example, the communications to and from Ms Winnacotta of the council), that the council was aware that Unomedical used EtO during the sterilisation operation, and moreover, that EtO was a dangerous chemical. But I do not accept that the council was aware of the extent to which the EtO was not totally consumed during the sterilisation process, was aware of where the excess EtO was being discharged and was aware of what amount. Had the council been appraised this information at the time it grant the approval to Unomedical to operate the steriliser at Wilmette Place, then a more solid foundation for the defence would have been in existence.

330 To the extent that there is a perceived inconsistency with my findings and the result on retrial in Cranbrook Swine, then I must respectfully cavil with the correctness of that case. In my view, however, the facts of Cranbrook can be distinguished on the basis that, unlike here, in that decision the location of the manure tank, namely, on swampland, was presumably plain for the relevant authority to see on the development permit application or when the relevant officer visited the site. That is not the present case.

331 Accordingly, not only was a greater degree of specificity in the consultation by Unomedical of the council necessary, so too a greater specificity in the disclosure of information to the council concerning the EtO emissions from the facility was required in order to enable Unomedical to avail itself of the defence.

332 Third, it follows that to the extent that the omission by the council to comment or to impose conditions of consent limiting the discharge of EtO (even when not specifically asked to do so) constitutes operative advice for the purpose of the defence, in the absence of sufficient disclosure by Unomedical as to the nature and extent of the EtO emissions, I find that any reliance on the advice obtained was not reasonable.

333 Fourth, the advice was not erroneous. It was correct. The council did not know that Unomedical was polluting the air in the manner in which it was. The corollary of this was that to the extent that it gave Unomedical any 'advice', this 'advice' must have been confined in scope to the facts known to the council at the time. Accordingly, the advice was not wrong because on the information the council had, not only was Unomedical a non scheduled premises with no statutorily prescribed limits on its use of EtO, but there was no reason for the council to impose any conditions of consent requiring an abatement system to be installed in respect of the EtO when the council was not aware that the EtO was being vented to the atmosphere.

334 Fifth, given that evidence discloses that Unomedical was not aware of the duty imposed by s 128(2) of the POEOA, discloses that Unomedical did not seek any advice in respect of the duty imposed by that section and discloses that no such advice was ever given by the council, it logically follows that Unomedical could not have relied on any advice obtained from the council in committing the act of operating the facility without implementing such practicable means as were necessary to minimise its air pollution.

335 For the reasons given above I find that the defence of officially induced error or law is not available to Unomedical to exculpate its liability.

336 I therefore do not need to consider what the appropriate relief should be had Unomedical successfully demonstrated the application of the defence, namely, a permanent stay of proceedings as ordered in Canada, or an acquittal, as suggested by Unomedical.

337 Having found the elements of the offence contained in s 128(2) proven beyond reasonable doubt by the EPA and having found that Unomedical has not been successful in establishing any of the defence relied upon by it, I find the defendant guilty of the offence with which it has been charged.

338 The order of the Court is: (1) stand the proceeding over to the list judge on 22 October 2010 for the purpose of making directions for a hearing on sentence. **********

DISCLAIMER - Every effort has been made to comply with suppression orders or statutory provisions prohibiting publication that may apply to this judgment or decision. The onus remains on any person using material in the judgment or decision to ensure that the intended use of that material does not breach any such order or provision. Further enquiries may be directed to the Registry of the Court or Tribunal in which it was generated.