Esso Australia Pty Ltd v The Australian Workers� Union [2016] FCAFC 72 (25 May 2016) Last Updated: 25 May 2016 FEDERAL COURT OF AUSTRALIA Esso Australia Pty Ltd v The Australian Workers’ Union [2016] FCAFC 72 Appeal from: Application for leave to appeal and appeal: Esso Australia Pty Ltd v The Australian Workers’ Union (Federal Court of Australia, VID 115 of 2015, Orders dated 13 August 2015) Esso Australia Pty Ltd v The Australian Workers’ Union (2015) 253 IR 304 ; [2015] FCA 758 File number(s): VID 435 of 2015 VID 470 of 2015 Judge(s): SIOPIS, BUCHANAN AND BROMBERG JJ Date of judgment: 25 May 2016 Catchwords: INDUSTRIAL LAW – construction of s 418 of the Fair Work Act 2009 (Cth) – powers of the Fair Work Commission under s 418 – such orders are limited to the scope of identified industrial action – primary judge held that orders of the FWC lacking in specificity were invalid – no error in the approach of the primary judge INDUSTRIAL LAW – Fair Work Commission ordered that “de-isolation of equipment” was protected industrial action – dispute between parties as to the practical meaning of “de-isolation of equipment” – primary judge found that the construction to be given to what is protected industrial action is that which a reasonable person in the position of the employer would understand it to mean – no error in conclusion by the primary judge INDUSTRIAL LAW – whether the union was in breach of coercion provisions under the Fair Work Act 2009 (Cth); ss 343 , 346 and 348 – two elements to be satisfied to find coercion ( Seven Network (Operations) Ltd v Communications, Electrical, Electronic, Energy, Information, Postal, Plumbing and Allied Services Union of Australia [2001] FCA 456 ; (2001) 109 FCR 378) – not necessary to establish the person intended to act unlawfully – not a defence that a person believed their action to be lawful – intent to coerce requires showing negation of choice Legislation: Fair Work Act 2009 (Cth), ss 13 , 14 , 14 (1), 19 , 186 , 186 (5), 272 , 272 (2), 340 , 341 , 341 (1)(b), 341 (2)(e), 343 , 343 (2), 346 , 347 , 347 (b)(iv), 348 , 360 , 361 , 407 , 408 , 409 , 410 , 411 , 413 , 413 (4), 413 (5), 414 , 414 (6), 415 , 417 , 417 (1), 417 (2), 418 , 418 (1), 418 (3), 419 , 421 , 421 (2), 556 , 570 , 675 Workplace Relations Act 1996 (Cth), ss 4 , 127 , 127 (1), 170NC, 170NC(2), 420 , 496 , 496 (1), 496 (9) Cases cited: Anthony Hordern and Sons Limited v The Amalgamated Clothing and Allied Trades Union of Australia [1932] HCA 9 ; (1932) 47 CLR 1 Australian Mines and Metals Association Inc v The Maritime Union of Australia [2015] FCA 677 ; (2015) 251 IR 75 Australian Mines and Metals Association Inc v The Maritime Union of Australia [2016] FCAFC 71 Construction, Forestry, Mining and Energy Union v Boral Resources (Vic) Pty Ltd [2015] HCA 21 ; (2015) 89 ALJR 622 ; 320 ALR 448 Construction, Forestry, Mining and Energy Union v Endeavour Coal Pty Ltd [2015] FCAFC 76 ; (2015) 231 FCR 150 Fair Work Ombudsman v National Jet Systems Pty Ltd [2012] FCA 243 ; (2012) 218 IR 436 Finance Sector Union of Australia v Commonwealth Bank of Australia [2000] FCA 1468 ; (2000) 106 FCR 16 Johns v Australian Securities Commission [1993] HCA 56 ; (1993) 178 CLR 408 Keen Mar Corporation Pty Ltd v Labrador Park Shopping Centre Pty Ltd & Anor [1989] FCA 54 ; (1989) 67 LGRA 238 Metal Trades Industry Association of Australia v Automotive, Food, Metals, Engineering, Printing and Kindred Industries Union [1997] FCA 1355 ; (1997) 77 IR 87 Pacific Carriers Ltd v BNP Paribas [2004] HCA 35 ; (2004) 218 CLR 451 R v Wallis [1949] HCA 30 ; (1949) 78 CLR 529 Re Cram; Ex parte The Newcastle Wallsend Coal Company Proprietary Limited [1987] HCA 29 ; (1987) 163 CLR 140 Seven Network (Operations) Ltd v Communications, Electrical, Electronic, Energy, Information, Postal, Plumbing and Allied Services Union of Australia [2001] FCA 456 ; (2001) 109 FCR 378 Transport Workers’ Union of New South Wales v Australian Industrial Relations Commission [2008] FCAFC 26 ; (2008) 166 FCR 108 Victoria v Construction, Forestry, Mining and Energy Union [2013] FCAFC 160 ; (2013) 218 FCR 172 Esso Australia Pty Ltd v The Australian Workers’ Union PR561701 Order (6 March 2015) Esso Australia Pty Ltd v The Australian Workers’ Union and another PR561655 Order (5 March 2015) Esso Australia Pty Ltd v The Australian Workers’ Union PR561142 Order (17 February 2015) Esso Gippsland (Barry Beach Marine Terminal) Enterprise Agreement 2011 Esso Gippsland (Longford and Long Island Point) Enterprise Agreement 2011 Esso Offshore Enterprise Agreement 2011 Date of hearing: 26, 27 November 2015 and 14, 15 December 2015 Registry: Victoria Division: Fair Work Division National Practice Area: Employment & Industrial Relations Category: Catchwords Number of paragraphs: 382 Counsel for the Appellant in VID435/2015 and the Respondent in VID470/2015: Mr F Parry QC with Mr M J Follett Solicitor for the Appellant in VID435/2015 and the Respondent in VID470/2015: Clayton Utz Counsel for the Respondent in VID435/2015 and the Appellant in VID470/2015: Mr H Borenstein QC with Mr P Rozen (26 and 27 November 2015) and with Mr J Fetter (14 and 15 December 2015) Solicitor for the Respondent in VID435/2015 and the Appellant in VID470/2015: The Australian Workers’ Union ORDERS VID 435 of 2015 BETWEEN: ESSO AUSTRALIA PTY LTD (ABN 49 000 018 566) Appellant AND: THE AUSTRALIAN WORKERS’ UNION Respondent JUDGES: SIOPIS, BUCHANAN AND BROMBERG JJ DATE OF ORDER: 25 MAY 2016 THE COURT ORDERS THAT: Leave to appeal be granted. Declaration 4 made on 13 August 2015 be set aside and in lieu thereof it be declared: “4. By operation of s 413(5) of the FW Act, from 6.01 pm on 6 March 2015 until 6.00 pm on 20 March 2015, all industrial action organised by the respondent and taken by the Esso employees in support of claims in relation to bargaining for a replacement enterprise agreement or enterprise agreements for the Esso Gippsland (Barry Beach Marine Terminal) Enterprise Agreement 2011 , the Esso Gippsland (Longford and Long Island Point) Enterprise Agreement 2011 and the Esso Offshore Enterprise Agreement 2011 was unprotected industrial action.” The appeal be otherwise dismissed. Note: Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011 . ORDERS VID 470 of 2015 BETWEEN: THE AUSTRALIAN WORKERS’ UNION Appellant AND: ESSO AUSTRALIA PTY LTD (ABN 49 000 018 566) Respondent JUDGE: SIOPIS, BUCHANAN AND BROMBERG JJ DATE OF ORDER: 25 MAY 2016 THE COURT ORDERS THAT: Leave to appeal be granted. Declaration 3 made on 13 August 2015 be set aside. The appeal be otherwise dismissed. Note: Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011 . REASONS FOR JUDGMENT SIOPIS J: I have had the considerable benefit of reading the reasons for judgment of Buchanan J. I agree with those reasons and with the orders proposed by Buchanan J. I certify that the preceding one (1) numbered paragraph is a true copy of the Reasons for Judgment herein of the Honourable Justice Siopis . Associate: Dated: 25 May 2016 REASONS FOR JUDGMENT BUCHANAN J: Introduction These two appeals, respectively by Esso Australia Pty Ltd (“Esso”) and The Australian Workers’ Union (“AWU”), a federally registered organisation of employees, arise from declarations and orders made by Jessup J on 13 August 2015 for reasons which were published on 24 July 2015 ( Esso Australia Pty Ltd v The Australian Workers’ Union [2015] FCA 758 ; (2015) 253 IR 304). In the proceedings before the primary judge Esso alleged that the AWU contravened a number of provisions of the Fair Work Act 2009 (Cth) (“FW Act”) when it organised industrial action at Esso’s plant at Longford, Victoria. The primary judge recorded the general nature of the activities in which the industrial action was organised as follows (at [2]-[7]): The applicant [Esso] is in the business of the exploration for and the production of oil and gas, the refining of petroleum and the supply of fuels, including natural gas. Relevantly to the present proceeding, the applicant operates three onshore facilities in Gippsland, and a number (presently 23) of offshore platforms, and associated infrastructure, in Bass Strait. The onshore facilities are at Longford, Long Island Point and Barry Beach. The Longford Plant is the onshore receiving point for all of the crude oil and gas produced by the applicant’s offshore platforms in Bass Strait. It consists of three separate gas plants, a crude oil stabilisation plant and an ethylene glycol plant. Two pipelines run 220 km from Longford to Long Island Point, carrying crude oil for storage and distribution and gas liquids for final processing and distribution. These pipelines are managed and maintained by Longford-based personnel. Hydrocarbons recovered from the seafloor are initially separated, at the offshore platforms, into two streams consisting of unstabilised crude oil and raw gas. These streams are piped separately to the plant at Longford, where impurities are removed and the two streams processed into stabilised products. These products are methane gas, which is sold directly to customers as natural gas, heavier hydrocarbon liquids (ethane and LPG), and stabilised crude oil. The LPG and the crude oil are sent for further processing and storage at Long Island Point. Operations (ie as distinct from maintenance) personnel employed by the applicant at the facilities referred to are represented industrially by the respondent. They, and other personnel, are covered by industrial agreements approved under the FW Act, namely – the Esso Gippsland (Longford and Long Island Point) Enterprise Agreement 2011, which covers employees at two of the applicant’s onshore processing operations (Longford and Long Island Point); the Esso Offshore Enterprise Agreement 2011, which covers employees at the applicant’s offshore oil and gas platforms; and the Esso Gippsland (Barry Beach Marine Terminal) Enterprise Agreement 2011, which covers employees at the applicant’s Barry Beach Marine Terminal. The nominal expiry date of each of these agreements (see FW Act, s 186(5)) was 1 October 2014. Since about June 2014, the applicant and the unions representing its employees, including the respondent, have been engaged in bargaining for the making of a new enterprise agreement, or agreements, to take the place of those referred to in the previous paragraph. The respondent is a bargaining representative for the applicant’s operations employees. It is clear that much has happened in that bargaining, including, at times, proceedings in the Fair Work Commission (“the Commission”). Save to the extent mentioned below, it is not necessary to refer further to those proceedings. I shall further address the applicant’s operations in some detail in the next section of these reasons, but, by way of broad introduction, I indicate now that the setting for the controversy which has led the parties to court is the return to operational service of items of plant or equipment which have earlier been taken out of service to have some work, such as repair, maintenance or upgrading, carried out on them. That work will normally have been carried out either by the applicant’s own maintenance personnel or by specialised contractors. The present case is not directly concerned with this work. The employees whose work is directly relevant, rather, are the operational personnel employed by the applicant. They control and monitor the plant and equipment in its normal operating state. They also have responsibilities at the point of removing plant and equipment from service for work to be done on it, and at the point of returning plant and equipment to service after the work has been completed. The industrial action organised at Longford which became the subject of this case was directed to the latter area of activity. (Emphasis added.) As to the character of the operation at Longford, the primary judge said (at [8]): The applicant’s operations at Longford involve the processing of highly toxic, volatile, pressurised and flammable products. Plant and equipment required to process these products involves heat, flame and pressure. Potential ignition sources are adjacent to highly flammable hydrocarbons. The risk of fire or explosion is ever-present ... In the period after the nominal expiry date of the three agreements had passed, the AWU gave a number of notices of intention to take “protected industrial action” (see FW Act s 413(4) and s 414) to assist its claims for new enterprise agreements, or for a single enterprise agreement. The course of those notifications, and some consequent proceedings before the Fair Work Commission (“FWC”) are described in the primary judgment. It is not necessary to repeat all those details in this judgment. On Esso’s application, the FWC made a number of orders under s 418 of the FW Act that industrial action stop, or not occur. When some forms of industrial action continued, Esso commenced proceedings in this Court alleging various contraventions of the orders made by the FWC, and of the FW Act. Esso’s case was partially successful. The orders made on 13 August 2015 reflect that partial success in the form of declarations that the AWU contravened the provisions of the FW Act in particular respects. There were to be further proceedings to consider questions of penalty and compensation. The orders made on 13 August 2015 were interlocutory, rather than final, orders. Leave to appeal is required in order to challenge them. Esso sought leave to appeal about particular respects in which its application did not succeed. The AWU sought leave to appeal about particular respects in which Esso’s application did succeed. At the hearing, the parties were informed that leave to appeal would be granted. Arguments on the appeals were then taken. The primary judgment is detailed and thorough. It will not be possible to adequately extract, or summarise, it in all respects. However, it is possible to trace the legal issues presented by the two appeals by commencing with questions concerning the validity and operation of certain orders made by the FWC under s 418 of the FW Act, and then moving progressively to the other legal issues which are presented for decision by the two appeals. For the time being, the legal issues may be described and discussed without being diverted to any particular extent by which appeal most directly concerns them. Section 418 The first issue in the appeals concerns the proper construction of s 418(1) of the FW Act. Resolution of that question of construction does not depend upon the particular facts of the case. Once that task has been undertaken, other matters concerning the application of further provisions of the FW Act to the facts of the case may be considered. Although some historical matters will need attention in connection with the initial issue of construction, an attempt should first be made to construe s 418(1) of the FW Act in its present, overall, statutory context. The FW Act establishes a scheme for the negotiation and making of enterprise agreements (see generally Part 2-4 of “Chapter 2—Terms and conditions of employment”) and (in some circumstances) workplace determinations (see generally Part 2-5 of Chapter 2). Enterprise agreements and workplace determinations, when made, must specify a term of not more than four years (s 186(5), s 272(2)). Section 417(1) and (2) of the FW Act provide: Industrial action must not be organised or engaged in before nominal expiry date of enterprise agreement etc. No industrial action (1) A person referred to in subsection (2) must not organise or engage in industrial action from the day on which: (a) an enterprise agreement is approved by the FWC until its nominal expiry date has passed; or (b) a workplace determination comes into operation until its nominal expiry date has passed; whether or not the industrial action relates to a matter dealt with in the agreement or determination. Note: This subsection is a civil remedy provision (see Part 4-1). (2) The persons are: (a) an employer, employee, or employee organisation, who is covered by the agreement or determination; or (b) an officer of an employee organisation that is covered by the agreement or determination, acting in that capacity. Industrial action, where that term is used in the FW Act, is given the meaning assigned by s 19 of the FW Act: Meaning of industrial action (1) Industrial action means action of any of the following kinds: (a) the performance of work by an employee in a manner different from that in which it is customarily performed, or the adoption of a practice in relation to work by an employee, the result of which is a restriction or limitation on, or a delay in, the performance of the work; (b) a ban, limitation or restriction on the performance of work by an employee or on the acceptance of or offering for work by an employee; (c) a failure or refusal by employees to attend for work or a failure or refusal to perform any work at all by employees who attend for work; (d) the lockout of employees from their employment by the employer of the employees. Note: In Automotive, Food, Metals, Engineering, Printing and Kindred Industries Union v The Age Company Limited , PR946290, the Full Bench of the Australian Industrial Relations Commission considered the nature of industrial action and noted that action will not be industrial in character if it stands completely outside the area of disputation and bargaining. (2) However, industrial action does not include the following: (a) action by employees that is authorised or agreed to by the employer of the employees; (b) action by an employer that is authorised or agreed to by, or on behalf of, employees of the employer; (c) action by an employee if: (i) the action was based on a reasonable concern of the employee about an imminent risk to his or her health or safety; and (ii) the employee did not unreasonably fail to comply with a direction of his or her employer to perform other available work, whether at the same or another workplace, that was safe and appropriate for the employee to perform. (3) An employer locks out employees from their employment if the employer prevents the employees from performing work under their contracts of employment without terminating those contracts. Note: In this section, employee and employer have their ordinary meanings (see section 11). (Emphasis in original.) Thus, a wide range of industrial action is prohibited during the nominal term of an enterprise agreement. Outside the nominal term of an enterprise agreement or workplace determination, industrial action is not, as such, prohibited by the FW Act. Some forms of industrial action, if stated conditions are met, may in addition be or become “protected industrial action”. Section 418 finds its place in the statutory arrangements which revolve around the notion of protected industrial action. I shall return to the text of s 418 shortly. Three forms of protected industrial action are identified by s 408 of the FW Act. They relate to industrial action which is organised or engaged in for supporting or advancing claims in relation to an enterprise agreement, or in response to such action, or in response to the response. The three forms of protected industrial action are employee claim action for the agreement (see more particularly s 409), employer response action (see more particularly s 411) and employee response action (see more particularly s 410). Each form of industrial action must, in order to be protected industrial action, meet the “common requirements” stated in s 413. Section 413 provides: Common requirements that apply for industrial action to be protected industrial action Common requirements (1) This section sets out the common requirements for industrial action to be protected industrial action for a proposed enterprise agreement. Type of proposed enterprise agreement (2) The industrial action must not relate to a proposed enterprise agreement that is a greenfields agreement or multi-enterprise agreement. Genuinely trying to reach an agreement (3) The following persons must be genuinely trying to reach an agreement: (a) if the person organising or engaging in the industrial action is a bargaining representative for the agreement—the bargaining representative; (b) if the person organising or engaging in the industrial action is an employee who will be covered by the agreement—the bargaining representative of the employee. Notice requirements (4) The notice requirements set out in section 414 must have been met in relation to the industrial action. Compliance with orders (5) The following persons must not have contravened any orders that apply to them and that relate to, or relate to industrial action relating to, the agreement or a matter that arose during bargaining for the agreement: (a) if the person organising or engaging in the industrial action is a bargaining representative for the agreement—the bargaining representative; (b) if the person organising or engaging in the industrial action is an employee who will be covered by the agreement—the employee and the bargaining representative of the employee. No industrial action before an enterprise agreement etc. passes its nominal expiry date (6) The person organising or engaging in the industrial action must not contravene section 417 (which deals with industrial action before the nominal expiry date of an enterprise agreement etc.) by organising or engaging in the industrial action. No suspension or termination order is in operation etc. (7) None of the following must be in operation: (a) an order under Division 6 of this Part suspending or terminating industrial action in relation to the agreement; (b) a Ministerial declaration under subsection 431(1) terminating industrial action in relation to the agreement; (c) a serious breach declaration in relation to the agreement. (Emphasis added in s 413(4).) Section 414 provides: Notice requirements for industrial action Notice requirements—employee claim action (1) Before a person engages in employee claim action for a proposed enterprise agreement, a bargaining representative of an employee who will be covered by the agreement must give written notice of the action to the employer of the employee. (2) The period of notice must be at least: (a) 3 working days; or (b) if a protected action ballot order for the employee claim action specifies a longer period of notice for the purposes of this paragraph—that period of notice. Notice of employee claim action not to be given until ballot results declared (3) A notice under subsection (1) must not be given until after the results of the protected action ballot for the employee claim action have been declared. Notice requirements—employee response action (4) Before a person engages in employee response action for a proposed enterprise agreement, a bargaining representative of an employee who will be covered by the agreement must give written notice of the action to the employer of the employee. Notice requirements—employer response action (5) Before an employer engages in employer response action for a proposed enterprise agreement, the employer must: (a) give written notice of the action to each bargaining representative of an employee who will be covered by the agreement; and (b) take all reasonable steps to notify the employees who will be covered by the agreement of the action. Notice requirements—content (6) A notice given under this section must specify the nature of the action and the day on which it will start. (Emphasis added in s 414(6).) Thus, at least one requirement to be met, if industrial action is to be protected industrial action, is that notice be given which specifies the nature of the action and when it will commence. One important consequence of the fact that action is protected industrial action is stated by s 415: Immunity provision (1) No action lies under any law (whether written or unwritten) in force in a State or Territory in relation to any industrial action that is protected industrial action unless the industrial action has involved or is likely to involve: (a) personal injury; or (b) wilful or reckless destruction of, or damage to, property; or (c) the unlawful taking, keeping or use of property. (2) However, subsection (1) does not prevent an action for defamation being brought in relation to anything that occurred in the course of industrial action. It is in the context set by the foregoing provisions that s 418 operates. The essential premise for its operation, and the effectiveness of the orders which it permits, is that the requirements for protected industrial action have not been met. Section 418 provides: FWC must order that industrial action by employees or employers stop etc. (1) If it appears to the FWC that industrial action by one or more employees or employers that is not, or would not be, protected industrial action: (a) is happening; or (b) is threatened, impending or probable; or (c) is being organised; the FWC must make an order that the industrial action stop, not occur or not be organised (as the case may be) for a period (the stop period ) specified in the order. Note: For interim orders, see section 420. (2) The FWC may make the order: (a) on its own initiative; or (b) on application by either of the following: (i) a person who is affected (whether directly or indirectly), or who is likely to be affected (whether directly or indirectly), by the industrial action; (ii) an organisation of which a person referred to in subparagraph (i) is a member. (3) In making the order, the FWC does not have to specify the particular industrial action. (4) If the FWC is required to make an order under subsection (1) in relation to industrial action and a protected action ballot authorised the industrial action: (a) some or all of which has not been taken before the beginning of the stop period specified in the order; or (b) which has not ended before the beginning of that stop period; or (c) beyond that stop period; the FWC may state in the order whether or not the industrial action may be engaged in after the end of that stop period without another protected action ballot. (Emphasis in original.) Section 418 calls for an assessment by the FWC about whether industrial action (i.e. as defined in s 19) that is happening, threatened etc or being organised is, or would be, protected industrial action. That assessment by the FWC could not be legally conclusive, but it is an important preliminary step nevertheless (see Re Cram; Ex parte The Newcastle Wallsend Coal Company Proprietary Limited [1987] HCA 29 ; (1987) 163 CLR 140 at 149). The assessment to be made by the FWC requires identification of the existing or potential industrial action as a matter of fact, or probable fact. The requirement on which s 418 operates is that the FWC may conclude that the industrial action is not protected or would not be protected if it took place although it may not be possible, in some cases of industrial action which is being organised or threatened, to say with any certainty what action is proposed because it cannot be assumed that notice of the kind referred to in s 414(6) has been provided. In fact, no steps at all may have been taken to attempt to meet the “common requirements” for protected industrial action. Nevertheless, the power of the FWC under s 418(1) is to order that the industrial action stop or not occur or not be organised, even though the FWC does not need to specify the particular industrial action (s 418(3)). Is it open to the FWC under s 418 simply to order that no industrial action (i.e. as defined by s 19), which is not protected industrial action, occur for a specified period (the stop period)? In my view, it is not. First of all, I see no reason to doubt that the FWC must, for the purpose of its own assessment, make some attempt to identify the existing or potential industrial action. Then it must form a view whether that existing or potential industrial action is or would be protected industrial action. Perhaps in a case where the industrial action was occurring, threatened or being organised during the term of an enterprise agreement or workplace determination, little would turn on the precise character of the industrial action but, on the other hand, s 417 needs nothing by way of assistance from s 418. The more critical case (perhaps the only practical case) for the invocation of s 418 will arise from concern about industrial action (actual or potential) outside the nominal term of an enterprise agreement or workplace determination. In such a case, as I have said, some effort must be made by the FWC, before issuing an order, to establish that the statutory foundation for the order is present. Similarly, the scope of the order cannot simply be at large because it must be directed at the industrial action (existing or potential) which has been identified. In my view, it is an abdication of the responsibilities of the FWC to make an order which simply states that industrial action must not occur which is not protected industrial action. If industrial action is not protected industrial action then the immunity from suit given by s 415 will not apply. An order under s 418 does not adjust the operation of s 415. An order under s 418 need not be complied with if industrial action is, or would be, protected industrial action (s 421(2)). However, otherwise, breach of an order under s 418 is an offence (see s 675) whether or not supported by an injunction under s 421. My view is reinforced by consideration of the operation of s 419, which forms part of the scheme of the same Division of Part 3-3 of Chapter 3 of the FW Act. Section 407 identifies the employees and employers, to which Part 3-3 applies, to be national system employees and national system employers, terms which are defined by ss 13 and 14(1) (relevantly) as follows: Meaning of national system employee A national system employee is an individual so far as he or she is employed, or usually employed, as described in the definition of national system employer in section 14, by a national system employer, except on a vocational placement. Note: Sections 30C and 30M extend the meaning of national system employee in relation to a referring State. (Emphasis in original.) Meaning of national system employer (1) A national system employer is: (a) a constitutional corporation, so far as it employs, or usually employs, an individual; or (b) the Commonwealth, so far as it employs, or usually employs, an individual; or (c) a Commonwealth authority, so far as it employs, or usually employs, an individual; or (d) a person so far as the person, in connection with constitutional trade or commerce, employs, or usually employs, an individual as: (i) a flight crew officer; or (ii) a maritime employee; or (iii) a waterside worker; or (e) a body corporate incorporated in a Territory, so far as the body employs, or usually employs, an individual; or (f) a person who carries on an activity (whether of a commercial, governmental or other nature) in a Territory in Australia, so far as the person employs, or usually employs, an individual in connection with the activity carried on in the Territory. Note 1: In this context, Australia includes the Territory of Christmas Island and the Territory of Cocos (Keeling) Islands (see paragraph 17(a) of the Acts Interpretation Act 1901 ). Note 2: Sections 30D and 30N extend the meaning of national system employer in relation to a referring State. (Emphasis in original.) Section 419 operates with a very similar mechanism to s 418. It provides: FWC must order that industrial action by non-national system employees or non-national system employers stop etc. Stop orders etc. (1) If it appears to the FWC that industrial action by one or more non-national system employees or non-national system employers: (a) is: (i) happening; or (ii) threatened, impending or probable; or (iii) being organised; and (b) will, or would, be likely to have the effect of causing substantial loss or damage to the business of a constitutional corporation; the FWC must make an order that the industrial action stop, not occur or not be organised (as the case may be) for a period specified in the order. Note: For interim orders, see section 420. (2) The FWC may make the order: (a) on its own initiative; or (b) on application by either of the following: (i) a person who is affected (whether directly or indirectly), or who is likely to be affected (whether directly or indirectly), by the industrial action; (ii) an organisation of which a person referred to in subparagraph (i) is a member. (3) In making the order, the FWC does not have to specify the particular industrial action. Section 419, therefore, is concerned with conduct (industrial action) by persons normally outside the operation of the FW Act which potentially causes “substantial loss or damage to the business of a constitutional corporation”. Like the basis upon which s 418 operates, s 419 states the satisfaction of the FWC about that matter as the foundation for its obligation to make an order. In the case of s 419, it is quite clear that the FWC could not simply make an order in general terms stopping all industrial action regardless of its effect. That would be to use the power for a purpose which was not its intended one. Such an order would be ultra vires in my view. Similarly, orders made under s 418 are confined within the statutory limits for which the power is granted. The order must relate to the industrial action which triggers the statutory obligation. Without resort to history or authority, therefore, I would construe s 418(1) as limited to orders where, so far as the circumstances permit, the order operates in relation to, and is confined to, the industrial action (existing or potential) which the FWC has decided is not, or would not be, protected industrial action, having regard to the reasons for that conclusion which may be quite particular. What, then, of history and authority? In the Workplace Relations Act 1996 (Cth) (“WR Act”) when it was enacted, s 127(1) provided a power to the Australian Industrial Relations Commission (“AIRC”), similar to that in s 418 of the FW Act, but referring only to industrial action, without reference to protected industrial action. In Metal Trades Industry Association of Australia v Automotive, Food, Metals, Engineering, Printing and Kindred Industries Union [1997] FCA 1355 ; (1997) 77 IR 87 (“ MTIA ”), Marshall J considered an order which was made simply prohibiting industrial action as defined by s 4 of the WR Act. Marshall J refused to grant an interim injunction restraining breach of the order, in part because (at 92): ... the order is void because it fails to adequately specify the particular conduct which it seeks to prohibit. ... By later amendments to the WR Act in 2006, the effect of s 127 was replaced by s 496 of the WR Act. Section 496(1) and (9) are a sufficient parallel with s 418(1) and (3) of the FW Act. In Transport Workers’ Union of New South Wales v Australian Industrial Relations Commission [2008] FCAFC 26 ; (2008) 166 FCR 108 (“ TWU ”), Gray and North JJ considered the effect of s 496(1) of the WR Act. An order had been made which simply prohibited industrial action by reference to the definition of that term in s 420 of the WR Act, although excluding protected industrial action. The AIRC had not found that industrial action was threatened etc; but that it was occurring. Gray and North JJ said (at [21] and [24]): In the present case, it appeared to Senior Deputy President Hamberger that industrial action, not being protected action, was happening. So much is evident from [6] of the Senior Deputy President’s reasons. There was no finding that it appeared to the Commission that industrial action was threatened, impending or probable, or was being organised. On that basis, the Commission was limited to making an order that the industrial action stop. ... ... For present purposes, it is enough to say that, in the absence of any finding other than that industrial action, not being protected action, by employees was happening, the Commission had no power to go beyond the making of orders that the industrial action stop. Without it appearing to the Commission that industrial action was threatened, impending or probable, the Commission was under no duty, and had no power, to make any order that the industrial action not occur. Similarly, in the absence of a finding that the industrial action was being organised, the Commission had no duty, and no power, to make an order that the industrial action not be organised. Later, their Honours said (at [39]): It is also necessary to bear in mind that the duty of the Commission to make orders is confined by s 496(1) of the WR Act to orders that “the industrial action stop, not occur and not be organised”. The reference to “the” industrial action is a reference to industrial action that appears to the Commission to be happening, to be threatened, impending or probable, or to be in the process of being organised. It is necessary for the Commission to identify the industrial action that appears to it to be happening, threatened, impending or probable, or being organised, and to make orders that that industrial action stop, not occur or not be organised, as the case may be. Section 496(1) contains neither a duty nor a power to make orders that any act or omission that might possibly fall within the definition of “industrial action” in s 420 of the WR Act stop, not occur or not be organised. The Commission’s duty, and power, is limited to the industrial action that is the subject of the application before it. (Emphasis added.) One difficulty with their Honours’ analysis and conclusion (which was to a similar effect to Marshall J in MTIA ) is that no reference was made to s 496(9). Section 496(9) appeared to have been introduced to address, and overcome, the conclusion of Marshall J in MTIA . The absence of any discussion of its operation in TWU makes that latter case an unsafe authority on this issue, as is the judgment in MTIA . However, those difficulties do not attend the judgment under appeal, where the primary judge dealt with all of them. Before I turn to that, I should mention that reference was made in argument to what was said about s 418(3) of the FW Act (the counterpart of s 496(9) of the WR Act), in the Explanatory Memorandum: 1689. In making an order to stop or prevent industrial action, FWA does not have to specify the particular industrial action (subclause 418(3)). This is intended to allow FWA to make effective orders that do not require the separate identification of each particular instance of industrial action. I regret to say that, like so many paraphrases of this kind, this statement does not assist much although it does to my mind suggest that some specificity was contemplated – e.g. sufficient to disclose the legal operation of the order and provide sufficient certainty to allow compliance with it. The FWC orders considered by the primary judge, to some extent or other, prohibited industrial action by reference to the statutory definition (here s 19 of the FW Act). Relevant parts of the primary judge’s reasoning about the construction of s 418(1) and (3) included the following (at [107]-[111], [114]-[116]): ... the obligation to make orders, expressed in the main clause of the sentence which constitutes subs (1), is to be read distributively. That is to say, for example, unless the Commission had found that unprotected industrial action was being organised, the Commission would have no power to order that it not be organised. This is one aspect of the view about s 496(1) taken by Gray and North JJ in Transport Workers’ , notwithstanding that this dimension of the subsection was then expressed less clearly than it is in s 418(1) of the FW Act: see (166 FCR at 120-121 [17]). ... subs (3) permits the Commission to frame its order in a way that does not “specify the particular industrial action”. That is to say, it is permissible for the industrial action to be identified without specification of whether it is, or would be, a work stoppage, a ban, or something else. But that does not mean that the Commission can go beyond the findings made under subs (1). Nor, in my view, does it mean that the Commission can frame its order by reference to “industrial action”, without more. The order which it is required to make may not extend beyond “the” industrial action which has been found to be happening, to be threatened, etc. If these observations are sound ones, the question will inevitably arise: if the Commission is limited to the industrial action which was the subject of its findings under s 418(1), but is not required to specify the form that the industrial action being prohibited by its order might take, how is it to be expected to identify the subject-matter of its prohibition? A ready, but rather unsatisfying, answer to that question would be to say that the operation of s 418 in the way I have expressed it is sufficiently clear to make recourse to practical issues such as this both unnecessary and impermissible as on matters of construction. A more satisfying answer would be to recognise that the section contemplates that the Commission must, or at least will normally, identify the industrial action in some way. This may involve specifying the particular industrial action: the existence of subs (3) does not mean that the Commission may not so proceed. Or it may make use of some other identifier which makes sense to the parties in the facts of the case, such as the purpose of the action, the place in which it is to occur, the timing of the action, or something else. The point here is that the existence of subs (3) does not, as a matter of construction, involve the proposition that the Commission no longer need identify the industrial action which is being prohibited by its order, or the conclusion that, in making its order, the Commission may travel beyond the scope of “the” industrial action, the subject of its findings under subs (1). ... Although the issue does not arise for resolution in the present case and I have not been addressed upon it, I would offer the tentative view that the mere inclusion of a term in an order that the order did not apply to protected industrial action, made in circumstances where the Commission had not made a positive finding that the industrial action which it had found to be happening etc was not, or would not be, protected industrial action, would not be within power under the section. Applying the foregoing legal analysis to the facts of the case, I commence with the order made by the Commission on 17 February 2015 (see para 38 above). Reading the definition of “industrial action” (cl 3.1) into the operative provision (cl 4.1), the order prohibited the respondent from organising, and the employees from engaging in, any industrial action within the definition in s 19 of the FW Act, including action of the kinds referred to in paras (a)-(d) of cl 3.1, but excluding action of the kinds referred to in paras (e)-(g) of that clause. For the reasons I have given above, I consider that, if the definition had referred only to the overtime ban the subject of para (d) – the subject, and the only subject of the Commission’s findings under subs (1) – the order would have been within power under s 418. On the other hand, if it had referred only to the definition in s 19, or to that definition together with the inclusions set out in paras (a)-(c), the order would have been beyond power. Either way, in my view, the exclusionary provisions of paras (e)-(g) would not affect the result. ... I would add that the specific inclusion of industrial action of the kind referred to in cl 3.1(d) bespeaks an intention on the part of the Commission that, whatever else might be conveyed by the terms of the order, there should be no doubt but that that industrial action was caught thereby. It is as though the Commission contemplated that there might be persons bound by the order who were unaware of the terms of s 19 of the FW Act, or to whom the legislative jargon used in paras (a)-(c) might not be familiar. The reference to the overtime ban at Longford in para (d) had the purpose, I infer, of putting up in lights the specific matter which had brought the parties to the Commission. There is every reason to suppose that the Commission intended that this aspect of the order, if no other, should operate. It will be clear from my reasons above that I take the view that the Commission’s order of 5 March 2015 had, and has, no valid operation. With respect to the Commission’s order of 6 March 2015, the position is, mutatis mutandis , the same as that reached above in relation to the order of 17 February 2015. The Commission had found that the respondent was organising, and that its members were implementing, bans on equipment testing, air freeing and leak testing. Paragraphs (a) and (b) of cl 3.1 of the order referred specifically to industrial action of that kind and, by reason of the operation of s 46(2) of the AI Act, the order was within power to that extent. The order did, however, have no wider valid operation. (Bold emphasis added.) (Italic emphasis in original.) The primary judge thus applied a “blue pencil” test confining two orders within the statutory limits which he had identified, and preserving their operation with respect to the industrial action which had been adequately identified by the FWC, namely (order made on 17 February 2015): DEFINITIONS 3.1 For the purposes of this Order, ‘ Industrial Action ’ has the meaning prescribed by section 19 of the Fair Work Act 2009 ( Act ) and includes: ... (d) ... a ban on the performance of overtime contrary to the Esso Gippsland (Longford and Long Island Point) Enterprise Agreement 2011 and contrary to custom and practice regarding availability for and the performance of overtime; ... (Emphasis in original.) and (order made on 6 March 2015): DEFINITIONS 3.1 Subject to 3.2, for the purposes of this Order, ‘Industrial Action’ has the meaning prescribed by section 19 of the Fair Work Act 2009 (Act) and includes: (a) the performance of work in a manner different from that in which it is customarily performed, or the adoption of a practice in relation to work the result of which is a restriction or limitation on, or a delay in, the performance of equipment testing, air freeing, and leak testing; (b) a ban, limitation or restriction on the performance of work on the performance of equipment testing, air freeing, and leak testing. ... Orders which lacked that kind of specificity were found to be invalid. The orders set out above, which the primary judge found to be valid, are accepted to have been validly made, at least to that extent. Esso argued that s 418(3) required that they not be so confined but in my view they provide a very good illustration of the practical operation of s 418(3) which supports the approach taken by the primary judge. The orders which survived identify industrial action by its nature and character. There is no doubt that the identification of that industrial action in that way is meaningful for the parties. At the same time, it may fairly be said that the orders do not (in this instance) identify the “particular industrial action” which will be prevented by their terms. It should be emphasised that orders of this kind may descend to that level of particularity if the FWC thinks it appropriate, but they need not do so. What they may not do is move beyond the industrial action identified as the foundation for making the order as required by s 418(1). In my respectful view, the approach taken by the primary judge was the correct one. The orders made by the FWC were valid only to the extent identified. The primary judge found that the AWU had breached the s 418 order made on 6 March 2015 in three respects. Those findings were reflected in the following declarations made on 13 August 2015: THE COURT DECLARES THAT: The respondent contravened section 421(1) of the FW Act by contravening clause 4.1 of the Third s 418 Order, by its conduct in organising unprotected industrial action to be taken by the Longford members in the form of a ban on air freeing and leak testing from 6:00 pm on 6 March 2015 until 9:30 am on 7 March 2015. The respondent contravened section 421(1) of the FW Act by contravening clause 4.1 of the Third s 418 Order, by its conduct in organising unprotected industrial action to be taken by the Longford members in the form of a ban on the manipulation of bleeder valves from 9:30 am on 7 March 2015 until the making of the Court’s interim order on 17 March 2015. The respondent contravened section 421(1) of the FW Act by contravening clause 5.1(a) of the Third s 418 Order, by its conduct in failing to prepare a Written Notice as soon as practicable following 6:00 pm on 6 March 2015, as required by clause 5.1(a) of the Third s 418 Order. In the next two sections of the judgment I will deal first with aspects of the subject matter of declarations 1 and 2 and then with declaration 3. De-isolation The next issue concerns the contention by the AWU that those parts of the order made on 6 March 2015, which had been found not to fail for non-compliance with s 418 of the FW Act, nevertheless failed to identify industrial action which was not, or would not be, protected industrial action. Specifically, “a ban, limitation or restriction on the performance of work on the performance of equipment testing, air freeing, and leak testing” (cl 3.1(b) of the order made on 6 March 2015) was said by the AWU to be protected industrial action because it was embraced by a notice of intended industrial action given on 3 February 2015, as follows: ... Employee claim action: ... An indefinite ban on the de-isolation of equipment by employees covered by the Agreements commencing at 12.01 a.m. on Thursday, 12 February 2015. ... In an application for an order under s 418 relating to this issue on 5 March 2015, Esso said: 4.4 Following the issuance of the Notice, the Employees engaged in a ban on the performance of de-isolation of equipment at the Longford Plant. 4.5 Esso responded by arranging for its managerial and supervisory employees to carry out that task. 4.6 Since 2 March 2015, the Employees have expanded their ban to testing procedures associated with the recommissioning of plant and equipment. In particular, the Employees have refused to perform equipment testing, air freeing, and leak testing (the Action). 4.7 In recent days, Esso managers have pointed out to the AWU that the Action is not protected industrial action because it is not covered by the Notice. The AWU, through the local delegate Rob Steed and the Victorian Branch Secretary Ben Davis, confirmed that: (i) the Action was in place; and (ii) the Action would continue because the AWU regarded the Action as part of the notified ban on de-isolation of equipment. The respective positions of the parties may be seen clearly enough from this extract. Esso was saying that the notified ban on de-isolation of equipment did not include equipment testing, air freeing and leak testing, but had as a matter of practice been extended to those activities subsequently. The AWU said that the notified ban had always, in terms, so extended. The FWC agreed with the position put by Esso, and made the order on 6 March 2015 which contained the part of the order on that day which the primary judge found survived the challenge to validity based on s 418. The primary judge also found that the work of equipment testing, air freeing and leak testing did not fall within the notified ban on de-isolation of equipment. The result was that the AWU breached s 421 of the FW Act by contravening an order made under s 418. The specific factual allegations of contravention relating to this issue were stated as follows (at [118]): The contraventions of s 418 orders alleged against the respondent are the following: (a) organisation of a ban on air freeing and leak testing imposed by its members at Longford on and from 2 March 2015; ... (d) organisation of a ban on the performance of bleeder valve manipulation “as part of air freeing and leak testing” imposed by its members at Longford on and from 7 March 2015. The primary judge’s ultimate findings were as follows (at [119]-[120], [126]): With respect to allegation (a) on this list, the only valid order under s 418 which applied to the ban on air freeing and leak testing was that made on 6 March 2015. That order came into effect at 6 pm that day. It was not until about 9.30 am on 7 March 2015 that Mr Jones, conformably with the advice which he had received from Mr Tschugguel, informed Mr Lloyd that he would perform leak testing. I infer that members of the respondent would also have performed air freeing if required (subject, of course, to the ban covered by allegation (d) on the list). There was, therefore, a period of about 15 hours during which members of the respondent at Longford maintained a ban on air freeing and leak testing in contravention of the Commission’s order of 6 March 2015. It is uncontroversial that the respondent organised that ban. With respect to allegation (d) on the list above, there was a ban on the performance of bleeder valve manipulation, although the respondent would regard the applicant’s qualifier “as part of air freeing and leak testing” as tendentious. The ban, of course, was imposed only when valves had to be manipulated in preparation for – or, as the respondent would have it, as an element of – the “de-isolation of equipment”, the term used in the respondent’s s 414 notice. In the context of the Commission’s order of 6 March 2015, the question is not whether the manipulation of bleeder valves was part of air freeing and leak testing, but whether the ban on the manipulation of bleeder valves amounted to the adoption of a practice in relation to work the result of which was a restriction or limitation on, or a delay in, the performance of air freeing or leak testing within the meaning of cl 3.1(a) of that order. The question needs only to be stated in those terms for an affirmative answer to be self-evident. I would hold that, by organising this ban, the respondent contravened the order of 6 March 2015 from about 9.30 am on 7 March 2015 until the making of the court’s interim order of 17 March 2015. ... Because they were not covered by any notice under s 414, the respondent’s bans on air freeing and leak testing, and later on bleeder valve manipulation associated with those functions, were not protected industrial action within the meaning of the FW Act. Those findings followed an extensive discussion of the positions put by Esso and the AWU about the meaning conveyed by the notice of 3 February 2015 and, in particular, what was meant and should be understood by a reference to a “ban on the de-isolation of equipment”. Important aspects of that discussion, and the evidence relied upon by the parties, concerned the significance of parts of Esso’s Work Management System (“WMS”) manual. When the primary judge introduced his discussion of what was conveyed by the notice he said (at [66], [69]-[70]): As mentioned above, on 3 February 2015 the respondent notified the applicant of a ban on the “de-isolation of equipment”. An important question is whether that notification covered the respondent’s members’ refusal to carry out air freeing and leak testing and, from 7 March 2015, their refusal to manipulate bleeder valves. ... The applicant’s complaint is not that the respondent’s notice of 3 February 2015 was bad for want of sufficient specificity. Indeed, the applicant says that it well understood what was conveyed by the notice: the de-isolation of equipment in the defined sense under the WMS manual, which involved, it is said, both positive and negative aspects. Positively, the WMS manual provided a definition which referred to unlocking and moving isolation valves, reconnecting systems, removing blinds, and unlocking electrical switches to their normal operating state. Negatively, the WMS manual provided separate definitions of “air freeing” and “leak testing”, thereby suggesting that these operations were not the same as de-isolations. Either way, it is said, the respondent’s notice should be understood as a reference to de-isolation of equipment as such, and as not including air freeing, leak testing or the manipulation of bleeder valves preparatory to, or associated with, those tasks. By contrast, it was submitted on behalf of the respondent that the term “deisolation of equipment” had an accepted, and well-understood, meaning at Longford. It was by reference to that meaning that the applicant’s management would have understood the respondent’s notice of 3 February 2015. In considering this submission, it is necessary to commence with the purely factual question whether there was such an accepted and wellunderstood meaning, both on the part of the operators employed by the applicant and on the part of the managers whose function it was to consider what was conveyed by the notice. His Honour then referred to aspects of the evidentiary case for both parties which he had earlier set out in detail (at [8]-[28]). Further detail was given (at [71]-[83]). Then, the primary judge said (at [84]): All things considered, I am not persuaded that, in a normal operational setting at Longford, the term “de-isolation of equipment” had an accepted, and well-understood, meaning as proposed by the respondent [the AWU]. To the contrary, at least in a practical context involving the identification of work and tasks, the term related to de-isolation as such. When air freeing and pressure testing were required to be carried out, they were referred to in terms – either those terms or, in the case of the former, “purging”, and, in the case of the latter “leak testing”. This conclusion was seen by the primary judge to be consistent with the operational practices and directions contained in the WMS manual, which set out in detail the procedures and processes relevant in relation to the “work permit” systems and isolation procedures. The primary judge had earlier observed at [11]: The detailed processes in relation to work permits and isolation procedures are prescribed in a separate procedures manual, called the Work Management System (“WMS”) manual. All operations personnel employed by the applicant have been trained in the provisions of this manual, and refresher training is also undertaken. The WMS manual is available at the relevant workplaces, in both electronic and paper forms. Some of the relevant definitions contained in the WMS manual are the following: ... Some definitions in the WMS manual were as follows (there arranged alphabetically): De-isolations Unlocking and moving isolation valves, reconnecting systems, removal of blinds, and unlocking electrical switches to their normal operating state . ... ... Equipment testing The process of temporarily de-isolating equipment and energizing to a live-state for operational testing or fault finding. ... ... Air freeing Removing oxygen from process equipment to prevent flammable mixtures occurring when hydrocarbons are reintroduced. ... ... Leak testing Introducing pressure to the system to confirm that integrity has been restored and there are no leaks. ... ... (Italics in original.) The introduction to Section 4.6 of the WMS manual, “Reinstating Facility Systems and Equipment” included: Introduction This section describes the procedures and precautions to be followed when reinstating equipment and facilities and defines personnel responsibilities. The general sequence of events, when reinstating equipment and facilities, is as follows: Recommissioning Note: Recommissioning is an activity that takes place throughout the reinstatement. Mechanical completion checks must be made before and after equipment testing, air freeing, leak testing, and de-isolating. Equipment testing Air freeing Leak testing Removing energy isolations (mechanical, electrical, instrument) Removing temporary defeats Acceptance testing ... The WMS goes on, in pages of detailed instructions, to deal separately with “Recommissioning”, “Equipment Testing”, “Leak Testing” and “Air Freeing”, before addressing procedures for “Removing Energy Isolations”. A general procedure is then stated as follows: Energy isolation removal procedure The following general procedure regarding energy isolation removal should be followed: Upon completion of all recommissioning, leak testing, and air freeing, the Permit Holder brings the Isolation Control Certificate to the Area Operator or CCR and advises that the system is ready to be de-isolated. The Isolating Authority and Area Authority must confirm that it is safe to perform the requested de-isolations. The Area Operator must confirm that the isolated system has been recommissioned, leak tested, and air freed, if required. − The Area Operator must confirm that no other planned or ongoing work requires any of the existing isolation points to remain isolated. The Area Operator collects the original and all copies of the Isolation Control Certificate and requests the Area Authority to approve for reinstatement. The Area Authority approves the deisolation activities to commence. For remote locations where the Area Authority is not present, the Area Operator or Permit Holder acts as the Area Authority and approves the de-isolation activities to commence. Once the Area Authority or designate has approved the de-isolation: − The Permit Holders remove their Functional lock(s) from the isolation control point (ICP). − The Area Operator removes his or her Functional lock from the ICP. − The Isolating Authority obtains the key(s) from the ICP and removes the locks and tags the isolation points. The Area Operator is responsible for coordinating the deisolating of the equipment. − The Isolation Control Certificate (ICC) is closed out. (Emphasis added.) Before the primary judge, Esso emphasised the separate identification of different stages in the overall recommissioning process, whereby equipment testing, air freeing and leak testing were not subsumed within the notion or process of de-isolation. Rather, before energy isolations are removed, any necessary equipment testing, air freeing and leak testing occurs. The case for the AWU focussed on the part in the overall process which is played by a safety feature known as the Isolation Control Certificate (“ICC”). The AWU sought, at first instance and on appeal, to emphasise the importance of this feature and the alleged understanding of its purpose and function possessed by operators. The primary judge (at [19]) referred to use of the ICC as follows: Section 4.5 of the WMS manual refers to an electronic artefact known as the Isolation Control Certificate (“ICC”). There are designated persons responsible for preparing and issuing the ICC to record the fact that the isolation has been requested and approved. The ICC becomes part of the work permit documentation. According to the WMS manual, the ICC “must be used to document the isolation and approval of equipment or systems and will be used to track the status of all isolations”. ... and then went on to describe those arrangements in greater detail. The AWU placed particular significance on a specific stage of that process, identified by an ICC heading in a drop down electronic version which was “De-isolation in Progress”. The written submissions by the AWU in the present appeal accepted: The purpose of a s. 414 notice is to enable an employer which is to become affected by protected action “to take appropriate defensive action”: Davids Distribution v NUW [1999] FCA 1108 ; (1999) 91 FCR 463 ( Davids ) at 495 [87]. The description of the “nature of the action” should be in “ordinary industrial English”: Davids at [88]. Because such notices are often drawn by nonlawyers acting without legal advice, an approach to their interpretation that places a “premium on legalism” is to be eschewed: Davids at [86]; CFMEU v Yallourn Energy Pty Ltd (2000) 100 IR 52 at [21]; Tidewater Marine Australia Pty Ltd v MUA [2014] FCA 172 at [19] . An assessment of the adequacy of the description of the industrial action “must take account of all the circumstances and examine expressions used in the context of whether the concepts embodied in the expressions are well recognised in workplace relations”: CEPU v Pinnacle Career Development Pty Ltd [2010] FCA 1350 ; (2010) 190 FCR 581 at [58] . Applying these authorities to the facts, the task of this Court is therefore to ascertain “the meaning which the [notice] would convey to a reasonable person having all the background knowledge which would reasonably have been available to the parties in the situation in which they were at the time [the notice was served]”: Maggbury P/L v Hafele Aust P/L . (Footnote omitted.) However, the AWU’s written submissions went on to pursue the issue as follows: It was the AWU’s case that these various steps are referred to at Longford collectively as “de-isolation” because they are performed while the computerised work method known as the Isolation Control Certificate (ICC) is designated ‘de-isolation in progress’. The ICC forms part of Esso’s Work Management System manual (WMS). The role of an ICC is explained on page 4-43 of the WMS. ICC’s are described in s.4.5.1 of WMS as the document by which energy isolations are managed. Section 4.5 (page 4-36) of the WMS describes the process of isolation, listing a series of actions that need to be taken. Isolation is a process. De-isolation is the re-instatement of the equipment to service, and it is submitted that, as a matter of consistency of interpretation, it too should be seen as a process. This proposition is confirmed in WMS at 4-38 which sets out the isolation process flow at Figure 4-7. It will be seen that that figure identifies a stage of “isolation in progress” and a later stage of “de-isolation in progress”. (Emphasis added.) (Footnote omitted.) By contrast, Esso submitted: Turning to the AWU’s specific criticisms of the primary judge’s approach to this issue: (a) ... The AWU places almost total reliance on the ICC. In its terms, nothing in the ICC advances the AWU’s case. Paragraph 12 advances a flipside type of argument, but the terms of the WMS are directly inconsistent with the submission. The flipside “de-isolation process” as the AWU would have it, is actually called “Reinstating Facility Systems and Equipment” in the WMS, and within this process, equipment testing, air freeing, leak testing and de-isolating are clearly separate activities; ... (c) ... The ICC does not describe any whole process as “de-isolation”. There is an electronic status on a computer called “de-isolation in progress”, but that is as far as the ICC goes. The AWU’s submission was that this defined or captured all reinstatement activities, but the evidence did not reflect the submission. ... (Emphasis added.) (Footnotes omitted.) That part of the WMS manual which explained the ICC recorded: Isolation Control Certificate The Isolation Control Certificate (ICC) must be used to document the isolation and approval of equipment or systems and will be used to track the status of all isolations. ... ... The ICC performs the following functions: Lists the location of the isolation points and their normal status. It is a record of any fittings (such as analyzer points, sample points and plugs) that have been removed or moved from their normal state to ensure that they are reinstated. It is a record of all drain, vent, and bleed points for valve integrity tests. (Operation of these valves does not need to be recorded as an amendment on the ICC but would be checked when returning equipment to service.) Authorizes any temporary de-isolation for equipment testing (Sanction to Test). Authorizes and records each de-isolated point. As necessary, includes attachments of marked-up P&IDs, electrical diagrams, and isolation plans/procedures. (Emphasis added.) The primary judge rejected the contention by the AWU of an “accepted, and well understood meaning” that extended the notion of de-isolation to equipment testing, air freeing or leak testing and found that the prevalent understanding about what de-isolation involved was governed by the WMS. In his conclusions about this part of the case, the primary judge rejected the idea that the ICC had significance as any form of work instruction in its own right and he rejected the suggestion that de-isolation extended either generally to equipment testing, or to manipulation of particular valves in connection with that work: Returning to the respondent’s reliance on the ICC, of the two presently contentious senses in which the term “de-isolation of equipment” might have been understood by the applicant as recipient of the notice of 3 February 2015, that referred in the WMS manual is, in my view, the more natural one. It refers to the de-isolation of equipment as such, and is, therefore, more closely aligned with the specific tasks which would, in the normal course, be carried out by operators, and which the applicant would understand to be the subject of the ban. By contrast, the ICC is concerned not with tasks or functions but with recording the positions of points at particular stages in the process of taking some equipment out of, and of returning it to, service. Insofar as it deals with the points that must be changed from one state to another, it records what has been done rather than, for example, instructing what should be done. Nowhere is this clearer than in the evidence of Mr Jackson. In short, of the two documents, the WMS manual is the more closely related to the work as such, and is the more directly concerned with marking out de-isolation as an activity of work. For the above reasons, I would reject the proposition that the expression “deisolation of equipment” in the respondent’s notice of 3 February 2015 would reasonably have been understood by the applicant as referring to every aspect of operators’ work that would be performed during the period that the ICC was headed “[De-isolation] in Progress”. To the contrary, in my view it would have been so understood as referring to the specific function of deisolation as such. It would not have been so understood as encompassing equipment testing, air freeing or leak testing. Nor did it refer to the manipulation of valves associated with those activities, notwithstanding that such manipulations were mentioned on the ICC. It follows that the respondent’s ban on work of that kind was not protected industrial action within the meaning of the FW Act. (Emphasis added.) In my view, those conclusions reflected the evidence to which the primary judge referred. I will refer to that evidence in more detail shortly. It is convenient, however, to make the observation at this point that, stripped of the attempt to colour the meaning of the WMS instructions by some subjective interpretation of them (which seems, if I might say so, to be foreign to the use to be made of instructions of this kind, having regard to safety implications) the AWU’s case could not ultimately depart from the terms of the WMS manual. Its argument about the role and significance of the ICC was one which was necessarily confined by the use of the ICC within the procedures directed by the WMS manual. No error was shown in the understanding of the primary judge which was recorded by him at [89]. He had a particular advantage as the trier of the facts. Furthermore, as I shall return to mention, I have considerable reservations about the notion that an enterprise in Esso’s position can be told what its safety procedures signify as a matter of purely textual debate or subjective assertion. First, however, I shall endeavour to trace, in a little more detail, some of the other evidentiary strands with which it was necessary for the primary judge to deal in his comprehensive discussion of this issue. The course of the industrial dispute which was revealed in evidence before the primary judge, and referred to in his findings, was also consistent with his understanding of the process of deisolation. It is apparent that the AWU modified its position about the extent of the bans it had imposed, in response to Esso’s success in having members of staff perform some of the necessary work. Initially, after the advice given on 3 February 2015, which took effect on 12 February 2015, AWU members banned de-isolation work in the sense referred to in the WMS. The effect of that action, and Esso’s response, was referred to as follows by Mr James Kristeff, Maintenance Superintendent at Longford: The de-isolation ban commenced on 12 February 2015 as notified, and has not ceased. The ban had a paralysing effect on all work that governed by the WMS. Consequently, Esso was forced to decide whether it was able to continue to operate the Longford Plant safely. On 25 February 2015, Esso decided to make the following alternative arrangements to cope with the de-isolation ban: (a) instruct appropriately qualified managerial staff to perform deisolations; and (b) triage and prioritise certain work so that managerial staff could perform critical de-isolations. Esso consulted with supervisors prior to making a decision as to whether to implement this proposal. Following that consultation, on 26 February 2015, Esso implemented this arrangement. Managerial staff began to perform the de-isolation in lieu of employees. That is: (a) operators performed recommissioning work, equipment testing, air freeing, and leak testing; (b) supervisors then stepped in to perform the de-isolation; and (c) the operator would then step in and complete the work by removing any temporary defeats and perform acceptance testing, and thereby move to close out the work permit. This enabled Esso to move through the critical work activities. It significantly reduced the impact of the de-isolation ban. The primary judge recorded (at [44]): ... in response to the respondent’s ban on the “de-isolation of equipment”, the applicant had instructed its supervisors to perform de-isolations. It seems that there were at least two, and possibly more, de-isolations performed by supervisors in the period which followed the respondent’s notice of 3 February 2015. Ross Dunbar, the Operations Superintendent – Gas Asset of the applicant (whose normal responsibilities lie in the area of the applicant’s offshore facilities but who was temporarily working at Longford in the co-ordination of de-isolation activities at this time) said in his affidavit that the first de-isolation by a supervisor was done on 19 February 2015. Robert Steed, an operations technician and a delegate of the respondent, said in his affidavit that de-isolations were done by supervisors on 26 February and 3 March 2015. It is sufficient to find that, by the latter date at the latest, it would have been apparent to the respondent and its members at Longford that the applicant had developed a modus operandi by which equipment deisolations, banned by the respondent since 12 February 2015, could be done by supervisors. Mr Kristeff said: On 2 March 2015, operators began to advise their supervisors that they would not perform air freeing or leak testing required to be performed before deisolations. I am informed that on 4 March 2015 a meeting occurred between Esso managerial staff and AWU representatives to discuss this issue. At that meeting were Rob Mackie (Longford Plant Operations Superintendent), Ross Dunbar (Operations Superintendent – Gas Asset), Rob Steed (AWU delegate), and Kain Jackson (an AWU member who refused to perform air freeing/leak testing), amongst others. I am advised and believe that Rob Steed: (a) advised that it was the AWU’s position that the de-isolation ban included air freeing/leak testing; and (b) confirmed that all 81 members of the AWU were aligned to that position. I am informed that there was a 30 minute break between the meeting to allow Esso and the AWU to reconsider their positions. When the meeting was recommenced, I am informed that the AWU representatives: (a) confirmed their position; and (b) advised that their members would refuse to perform air freeing/leak testing if requested to do so. On the basis of this information provided to me, I decided to call Ben Davis, AWU Victorian Branch Secretary, that evening. Melinda Fairbanks, Human Resources, was present when I telephoned Mr Davis. In that conversation: (a) I said that air freeing, purging and pressure testing (i.e. leak testing) are not classified as de-isolations; (b) I invited Mr Davis to consult with his members and delegates at Esso about the position that has been taken; and (c) Mr Davis advised that he would consult his delegates at Esso. Mr Davis called me and Ms Fairbanks shortly afterwards and said that it was the AWU’s view that the ban on air freeing/leak testing was part of the deisolation ban. The primary judge recorded the following (at [45]-[46]): Over the period 28 February to 2 March 2015, preparations began for the deisolation of the Gas Plant 1 rich oil fractionator tower. The scoping of these works included operations supervisors reviewing the ICC and drawings, and walking the process lines in the field. On 3 March 2015, as part of these preparations, supervisors were involved in de-isolating the rich oil fractionator tower level bridle, which was required to allow the tower deisolation to commence the following day. At about 4:45 pm on 4 March 2015, Messrs Dunbar and Mackie met with Messrs Steed and Jackson. Mr Steed told Messrs Dunbar and Mackie that it was the respondent’s position that the de-isolation ban included air freeing and leak testing, and that all members of the respondent were aligned to that position. After a 30 minute break in this meeting, Mr Steed reiterated that this was the respondent’s position, and that its members would refuse to perform air freeing or leak testing if required to do so. James Kristeff, the Maintenance Superintendent at Longford telephoned Mr Davis, who confirmed what Mr Steed had said. However, after proceedings in the FWC on 6 March 2015, the AWU’s position was further refined. The FWC recorded the respective parties’ positions on 6 March 2015 as follows: ... In this present matter, the AWU considers that the bans in (e) referred to on airtesting, equipment testing, air-freeing and leak testing as included within that paragraph. Esso disagrees. ... Various phrases were used by the AWU to describe their understanding, including “custom and practice” and “the ordinary usage in the workplace.” A number of phrases were used. The employer on the other hand relies on the definitions used in safety manuals produced by them in accordance with their extremely important obligations to provide a safe workplace. These obligations are in any sense critical. Their ability to function depends on implementation of proper safety procedures. ... The FWC concluded that equipment testing, air freeing and leak testing was not notified in the notice on 3 February 2015, and the s 418 order made on that day prohibited bans on that work. Then the AWU’s position changed. The AWU then commenced to contend that although such work was not banned as such, manipulation of any valve listed on an ICC for the purpose of that work fell within the earlier notification. The primary judge traced the development of this new position. On 7 March 2015, particular work (air freeing and leak testing) was scheduled and three employees were assigned to perform it. The primary judge said (at [55]-[64]): One of those employees was Gary Jones, an operator and a member of the respondent. On the morning of 7 March 2015, the members of his shift held a meeting. They were addressed by their delegate, Karl Tschugguel. He informed them of the Commission’s order, and what it required. He said that they were not to ban air freeing and leak testing work. There followed a discussion about what could be done without breaching the order. The operators decided that, if points were listed on the ICC, they were deisolation work and were covered by the ban. They decided that, under the ban, it was open to them to refuse to manipulate the points or valves. At approximately 9:00 am, Mr Lloyd met with Mr Jones. They discussed the leak testing to be performed on the propane header, the pressure rating on the vessels, and the scope of the work generally. Mr Lloyd came away from those discussions with the understanding that Mr Jones was going to perform the leak testing. Mr Jones then spoke to Mr Tschugguel, and informed him that Mr Lloyd had requested that he perform leak testing work on the propane header. He sought clarification as to what he could and could not do as part of the protected industrial action. He told Mr Tschugguel that a bleeder valve on the propane header would have to be manipulated before he could conduct pressure testing. Mr Tschugguel then asked Mr Jones to access a computer to check the electronic ICC. Having done so, Mr Jones said that the ICC was “in place”, and that it also listed the bleeder valve as a tagged valve. The valve was also tagged “in the field”. On the basis of this information, Mr Tschugguel advised Mr Jones that, if a supervisor manipulated the bleeder valve and recorded “de-isolation in progress” on the ICC, both on the computer and in the field, he should comply with the Commission’s order and perform the pressure test. At about 9:30 am, Mr Jones returned to Mr Lloyd’s office and informed him that he would not shut the bleeder valves to facilitate the leak testing. Mr Jones said that he was refusing to do this because the open/shut status of bleeders was recorded on the ICC, and, accordingly, work involving the manipulation of bleeder valves formed part of the de-isolation process. It was, therefore, as Mr Jones informed Mr Lloyd, covered by the ban imposed upon de-isolations of equipment notified on 12 February 2015. Mr Jones said that he was willing to perform the leak testing if a supervisor manipulated the bleeder valves. Mr Lloyd’s response, as stated in his evidence-in-chief, was as follows: I questioned that. To me it wasn’t part of a ban. I was under the impression we were okay to purge and pressure test, and there was no protected action around that. I felt I understood the ban on deisolation of equipment. I certainly wasn’t asking him to de-isolate any equipment. I was asking him to perform a leak test. So I suggested he think about that, because I was just asking him to manipulate some bleeders so we could start the pressure test. According to the evidence of Mr Lloyd, Mr Jones also said that he had made this decision after consulting his delegate, Mr Tschugguel. In his evidence-in-chief, Mr Jones denied that, adding that Mr Tschugguel was present at the time, as was another of the respondent’s delegates, Brian Rawnson. Neither Mr Lloyd nor Mr Jones was cross-examined about this point of disagreement. However, what I have said in para 60 of these reasons is taken from the unchallenged evidence of Mr Tschugguel, from which it appears that Mr Jones had indeed had the consultation referred to in Mr Lloyd’s evidence. Mr Lloyd then asked Mr Tschugguel to come to his office, which he did. He (Tschugguel) confirmed that he had instructed Mr Jones not to shut any bleeder valves to progress the leak testing. He said that he had told Mr Jones to undertake pressure testing only after a supervisor had put the “de-isolation in progress” on the computer and manipulated the bleeder valve listed on the ICC. According to Mr Tschugguel’s evidence, the respondent’s members were not refusing to perform air freeing or leak testing as such. This meeting was followed by another at about 10:45 am, also in Mr Lloyd’s office. In addition to Mr Lloyd himself and Mr Tschugguel, Mr Jones and Rob Mahon, Operations Supervisor at Longford, were present. Mr Lloyd stated the applicant’s position that the manipulation of bleeder valves was work that formed part of air freeing and leak testing. Mr Tschugguel then asked Mr Rawnson, to join the meeting. Having done so, Mr Rawnson confirmed, on behalf of Mr Tschugguel and Mr Jones, that it was the respondent’s position that bleeder valve manipulation formed part of the ban on the de-isolation of equipment. He confirmed that the respondent’s members would perform air freeing and leak testing once all bleeder valve manipulations had been completed by supervisory staff. At about 3:30 pm, there was yet another meeting in Mr Lloyd’s office, convened by Mr Kristeff. In addition to Messrs Lloyd and Kristeff, Messrs Rawnson and Tschugguel were present, as was Rob Mackie, Operations Superintendent at Longford. At the meeting, Mr Kristeff asked Messrs Rawnson and Tschugguel whether they had been provided with the Commission’s order made the previous day. They said that they had. Mr Kristeff said that the Commission had made a ruling about what a deisolation was, which confirmed the applicant’s view that leak testing and air freeing were not part of the ban. He said that de-isolations were defined in the WMS, and did not include leak testing and air freeing. He directed Messrs Rawnson and Tschugguel to perform leak testing and air freeing when that work became available later in the day. Messrs Rawnson and Tschugguel did not agree with Mr Kristeff. They held to the respondent’s position as explained to Mr Lloyd earlier that day. They said that they would not perform the manipulation of bleeder valves because that was identified on the ICC. Mr Kristeff asked them, and they agreed, to confirm their position after speaking to Mr Steed. The meeting concluded on that note. At about 4:15 pm on 7 March 2015, Messrs Rawnson and Tschugguel met with Messrs Kristeff and Mackie in Mr Kristeff’s office. They telephoned Mr Steed, who joined the conversation on speaker phone. He told Mr Kristeff that the respondent’s position was that its members would not shut bleeder valves as part of air freeing or leak testing, because bleeder valves were identified on the ICC. He said that supervisors would need to shut bleeders, and to direct operators what to do subsequently, if air freeing or leak testing were to proceed. (Emphasis added.) It is noteworthy, in my view, that the reference point adopted by the AWU and its members at that point was that particular valves were “listed” or “identified” on the ICC. It was no longer asserted, as it had been in the FWC, that equipment testing, air freeing and leak testing were, as a matter of custom and practice, regarded as part of de-isolation. Before the primary judge this broader contention was re-introduced, but the evidence of a number of witnesses (from both the AWU and Esso) was consistently to the effect that the ICC was a list of the status of particular valves (i.e. a record of what had been done), rather than any form of work instruction. The evidence before the primary judge was that the WMS was the source of instructions about duties and tasks. For example, the primary judge recorded (at [74]): Under cross-examination, Mr Steed agreed that he had given evidence in support of the respondent’s position in the Commission, where he had produced, and relied on, Sections 4.5 and 4.6 of the WMS manual. He accepted that he saw the manual as crucial in setting out the duties and tasks at Longford, that the manual was a critical document on site, that there were copies of it in the operators’ workplaces, that it was consulted by both operators and management, that there were electronic copies of it available to operators, and that there had been training, including refresher training, provided with regard to the manual. He was taken to the evidence he had given in the Commission, from which it became clear that he was well aware of the various definitions in the WMS manual, including the definition of “de-isolations”. After referring to the evidence of further AWU witnesses, the primary judge said (at [78]): That was the evidence led by the respondent in support of its contention that the term “de-isolation of equipment” had an accepted and well-understood meaning amongst operators, that is to say, a meaning that would comprehend, without ambiguity, the whole process of returning the equipment to service, including air freeing and leak testing. The evidence was, however, ultimately equivocal, and not really helpful in the present context. For one thing, the acceptance by the witnesses, specifically Mr Steed, of the central importance of the WMS manual – and, one might add, the reference to that manual in the evidence-in-chief of Mr Jackson – makes it inevitable that the court would find, as I do, that the term “deisolations” has an official, or authorised, meaning. If it is the respondent’s evidentiary case that the term also has an everyday meaning which differs from that in the manual, that would produce the result that there were two meanings, something which the respondent’s contention does not, and could not, contemplate. Then the primary judge referred to evidence of typical daily work instructions which also, in a highly practical sense, distinguished between purging (air freeing), pressure testing (leak testing) and de-isolation (or removing the ICC). The primary judge said (at [82]): Mr Lloyd’s evidence as to the language used in the daily notes, and as to the use of those notes in the allocation of tasks each shift at the workplace, was both more specific and more reliable (ie in the sense that it was based on documentary materials) than the high-level references by the respondent’s witnesses to what they do in the process of de-isolation. Evidence of the latter kind did not have the focus sufficient to establish the terminological discriminations that the respondent’s case involved. By contrast, the daily notes, and the evidence given in connection with them, dealt with actual situations which had arisen in relevant contexts, and did so at a very low level, where the operators themselves were intimately involved. Based largely on those notes, I would find that, as a rule, operators tend to be instructed to purge, to pressure test, or to de-isolate (or “remove the ICC”) a piece of equipment, as Mr Jones was on 7 March 2015. There was nothing in the daily notes that would provide support for the respondent’s case that, in the normal course of their duties, operators tended to be instructed merely to “de-isolate” some piece of equipment which required all the relevant stages to be carried out. The primary judge then said (at [84]) (in a passage which I set out earlier): All things considered, I am not persuaded that, in a normal operational setting at Longford, the term “de-isolation of equipment” had an accepted, and well-understood, meaning as proposed by the respondent. To the contrary, at least in a practical context involving the identification of work and tasks, the term related to de-isolation as such. When air freeing and pressure testing were required to be carried out, they were referred to in terms – either those terms or, in the case of the former, “purging”, and, in the case of the latter “leak testing”. The understanding which the primary judge said was crucial was Esso’s, for the following reasons (at [86]-[88]): The task for the court is not the conventional one of the construction of a document, with a view to understanding what the author intended. The document with which we are concerned here was a notice: its purpose was to convey information. Thus the question is not what the author intended, but what the addressee would reasonably have understood from the terms used in the notice. Two things follow from this. First, if the notice might reasonably carry more than one denotation, I see no reason to err on the side of the giver of the notice, thereby permitting him or her to take advantage of his or her own ambiguity. The FW Act leaves it entirely to the giver of the notice to identify the “nature” of the action intended to be taken, and it should not be open to him or her to complain if the terms chosen leave scope for the addressee to see things differently from what the giver, subjectively, might have had in mind. Secondly, it would not be sufficient for the respondent to establish what its own members understood by the presently contentious expression. Even if they knew what they meant by “de-isolation of equipment”, the question is what the applicant’s management would reasonably have understood by that expression. Although I have noted above that the judgment of Wilcox and Cooper JJ in Davids Distribution does not directly provide the answer to the present question, one aspect of that judgment which is here valuable is the identification of the purpose of a notice of industrial action given under the predecessor to s 414 of the FW Act. Their Honours saw the purpose as enabling the party who would be adversely affected by the intended action to take appropriate defensive action. Their Honours recognised the importance of a defending employer, for example, having the opportunity to protect sophisticated equipment from damage. In my view, it is no less important for the affected party to know what functions, operations, etc will not be touched by the intended action. To take an example far from the facts of the present case, an employer handling perishable foodstuffs should be entitled to assume, with confidence, that its operations will not be affected beyond those notified to it under s 414 of the FW Act. Understood in this sense, the notification of industrial action has a negative, as well as a positive, dimension, each of which may be perceived as within the broad purpose of the statute. In the present case, the respondent’s notice of 3 February 2015 was given in the knowledge of the applicant’s forthcoming shutdown: indeed, the bans notified were to take effect on the day the shutdown was due to commence. The inference that the bans were intended to affect the progress of the shutdown is irresistible. In this setting, the applicant would reasonably have understood the notice as referring to the respects in which the shutdown would be affected. That the applicant was entitled to use the notice as an indication of what it would be unable to do, and, no less importantly, of what it would be able to do, during the shutdown was, in my view, a critical aspect of the process of collective bargaining mandated by the legislation, in its application to the circumstances of these parties. (Bold emphasis added.) (Italic emphasis in original.) Both parties to the appeal agreed that useful guidance was available from the judgment of the High Court in Pacific Carriers Ltd v BNP Paribas [2004] HCA 35 ; (2004) 218 CLR 451. At [22] , the High Court said: ... The case provides a good example of the reason why the meaning of commercial documents is determined objectively: it was only the documents that spoke to Pacific. The construction of the letters of indemnity is to be determined by what a reasonable person in the position of Pacific would have understood them to mean. That requires consideration, not only of the text of the documents, but also the surrounding circumstances known to Pacific and BNP, and the purpose and object of the transaction. ... (Footnotes omitted.) In the present case, the “transaction” was the notice of intended industrial action which was required to specify the “nature” of the intended industrial action. The construction to be given was that which a reasonable person in the position of Esso would understand the notice to mean, what it would understand the ban to involve and, as the primary judge found, what the ban would not involve. On the appeal, it was argued that the primary judge had failed to discuss, and therefore gave insufficient weight to, evidence by two senior Esso employees, Mr James Kristeff and Mr Ross Dunbar. The evidence was said to involve concessions by those employees about the significance of the ICC. Before I refer to that evidence, I will refer to the evidence of a third Esso senior employee, Mr Mark Lloyd, to whom the primary judge referred in another connection at [82]. Mr Lloyd was Operations Supervisor – Longford Plants. In his cross-examination, the following evidence was given: [MR BORENSTEIN:] All right. Now, you would agree, wouldn’t you, that the ICC sets out the various actions that need to be taken by an operator, both when equipment is being isolated and when equipment is being de-isolated?---It sets out the isolation points that need to be manipulated. And then manipulated back again for de-isolation?---Correct. That evidence was consistent with the findings made by the primary judge. Mr Kristeff was Maintenance Superintendent at Longford. His evidence was less definite, but I would not regard it, read fairly and fully, as involving any concession of the kind suggested in submissions. The evidence in cross-examination included: [MR BORENSTEIN:] ... And the certificate also then goes on to identify all the particular steps that need to be taken to bring the equipment back online. That’s correct, isn’t it?--- Well, it lists the – all your points that you want to account for on that ICC. ... But can you agree with me that when it talks – when you get to the stage where the ICC is marked as de-isolation in progress, what happens then it whatever steps need to be taken to bring the equipment back online are then taken?---Yes. I would agree with that. Yes. And then when those items of work are completed, the certificate is marked as de-isolation confirmed; correct?---Once all the de-isolations had been confirmed, yes. ... Yes. Okay. Now, to your knowledge, the steps that could be required to be taken, once the equipment is put into de-isolation in progress, could include – depending on the equipment, obviously, could include equipment testing?---Yes. Could include air freeing?---Yes. Could include leak testing?---Yes. Could include removing energy isolations?---Well, that’s the final step. Yes. ... Yes. Those things that I mentioned to you - - -?---Yes. - - - that list of things, they are actions or activities that are carried out – depending on the equipment in question, are carried out between the stage when the ICC is marked de-isolation in progress and the stage when it’s marked de-isolation confirmed?---Yes. ... Okay. And you would agree that if a – an item of work or a piece of equipment is noted in the ICC as requiring manipulation to achieve de-isolation, that it’s reasonable to refer to that as part of the de-isolation process?---Say that again. (Emphasis added.) At this point, Mr Kristeff resisted the cross-examiner’s proposition, and did not later accept it. His resistance led to a more extended exchange of questions and answers which included the following: [MR BORENSTEIN:] But whatever it’s for, it’s marked in here. It has to be done, doesn’t it?---It has to be logged and - - - As being done?---As being the status. Yes. Well, that’s just playing with words, isn’t it?---No, no. It’s the status of whether it has – whether it has been manipulated from one stage or- - - That’s right. Well, okay. It either has to be opened or closed, whichever way you’re going?---Correct. Correct?---Yep. And until that’s done you can’t mark off the de-isolation as completed?---Correct. And that means it’s part of the de-isolation that the ICC requires?---It records the – the – it records the steps you’ve taken. Yes. And it records the steps because it requires you to do those steps before you can say that de-isolation is complete. That’s correct, isn’t it?---Well, it records them so you – you don’t put something back into service that you’ve missed. That’s right. ... But if they are on the ICC then they are part of the de-isolation, aren’t they?---Under – not under my definition of - - - ... They are part of the steps in the ICC, are they not?---Yes. Okay. And because they’re in the ICC they are part of the de-isolation process under the ICC, aren’t they? They have to be done, don’t they?---They would have to be, yeah. ... [THE WITNESS:] ---... if you’re going to manipulate your bleeder valves for air freeing or purging or pressure testing then it has nothing to do with the deisolation. [MR BORENSTEIN:] But these – like, these valves are noted in the ICC?---Yep. They have to be done?---Yeah, but you’re going to manipulate them numerous times, right? That’s what I’m – that’s what I’m saying. ... We’re talking about equipment that’s on the ICC. You have to follow the ICC. That’s the rule, isn’t it?---Correct. It’s a safety rule as much as anything else, isn’t it?---Well, it’s a catch-all, yeah. Yes. But it’s a safety rule so people know what has happened before and they don’t get caught out?---Correct. Yes. And the ICC tells you, because somebody has drawn up this plan before they started – it tells you what you have to do step by step. That’s what the ICC is for, isn’t it?---Yep. So you have to follow those steps to follow the ICC?---Yeah, but it’s – it’s not – it’s not in order. No, it’s not in order?---No. I was going to ask you that. It’s not in order?---Yep. But it sets out all the steps that need to be taken?---As a final close, yep. To get to that?---Yep. To get to that, yes?---Yeah. And the system also describes the process of bringing the equipment back online as de-isolation. Now, we’ve been through that and you’ve agreed to that, and that’s right, isn’t it?---Yes. The document says so. The document doesn’t say any other words, does it?---No. And so the operators out in the field who are working with this document are working with a document that describes what they do as de-isolation. That’s correct, isn’t it?---That’s correct. Yes. And so they’re talking about doing a de-isolation of equipment why would you think that they were referring to anything else?---Because that’s the electronic version versus – there’s a lot of – lot of things in the manual that relies on going back to definitions, right, and that’s why we have a definitions page. (Emphasis added.) The following points may be noted from this exchange: air freeing and pressure testing is separate from de-isolation; valves may need to be manipulated a number of times, and not just for ultimate deisolation; the ICC is an overall safety guide; the ICC does not record the order in which valves are manipulated (i.e. it is not a work method but a record); ultimately, the WMS is the guide. Mr Dunbar was Operations Superintendent – Gas Asset. His cross-examination included: [MR BORENSTEIN:] ... The ICCs aren’t generic, are they?---They’re developed for a specific piece of equipment per the manual. That’s right. So they’re not generic, are they?---No. So when a particular piece of equipment has to be worked on you have to follow the ICC, don’t you?---You put the ICC in place to perform work and you - - - And you do what’s in the ICC, don’t you? MR PARRY: Can we just slow down? Let the witness finish his sentences. If my learned friend could do that. THE WITNESS: And then you reinstate the equipment per the work management system. MR BORENSTEIN: Yes. And the steps you have to take are the steps that are set out in the ICC for that piece of equipment, aren’t they?---The steps you need to take are defined here. The air freeing, the – to reinstatement of your air free, your leak test. If I need to do those things it’s not documented in the ICC. The ICC just lists the valves, the energy isolations that we put in place to take that equipment out of service. If you do not attend to each of the items that’s listed in the ICC, the ICC cannot be marked off as “de-isolation confirmed”, can it?---That’s correct. Okay. Now, isn’t the corollary of that that you must do everything that’s in the ICC in order to achieve a de-isolation?--- To remove the energy isolations, yes, you need to de-isolate the points that are listed in the ICC. And when you achieve that the job is marked as “de-isolation confirmed”?---That’s correct. Yes. And the terminology, as you said before, aggregates all those tasks that you need to do that are listed in the ICC and describes them as “de-isolation”, doesn’t it?---I don’t believe that’s correct. But I thought you told us that a moment ago?---No. I believe that the electronic system uses the term “a grouping de-isolation progress” - - - And the electronic system is that, isn’t it?--- The work management manual defines the sub – the tasks that need to be done. (Emphasis added.) I do not see any inconsistency between the evidence by Mr Kristeff or Mr Dunbar and the findings made by the primary judge about the role, function and significance of an ICC. During the appeal, two examples of an ICC were tendered. One was headed “Deisolation in Progress”, and the other “Archive”. The archive example showed a period of three days between commencement and completion of de-isolation. The terms of such certificates were the object of central attention before the primary judge. Notwithstanding the passage of several days, Esso’s operational procedures distinguished between de-isolation, as such, occurring in that period and other procedures such as air freeing, equipment testing or leak testing which might also require manipulation of valves listed on the ICC. In my view, on the evidence before the primary judge, he was correct to conclude that a ban on de-isolation did not extend to a ban on those procedures, nor upon manipulation of valves connected with them. As I understand the evidence to which we were specifically directed on the appeal, the ICC has nothing to say about equipment testing, air freeing or leak testing. In some circumstances, those processes may be carried out in the absence of any ICC at all. In other cases, where an ICC has been put in place as a checklist for the purpose of shutdown or recommissioning, the ICC is not directed to those tasks either. Rather, the ICC is a checklist directed to the final processes of energy restoration. Its record of valve positions and status is directed to that purpose and does not govern (even though it may perhaps record) any manipulation for a different, more immediate, purpose. Thus, manipulation of a bleeder valve for the purpose of leak testing (pressure testing) was not “de-isolation”. Any change in status which was recorded on the ICC was not directed to the process of de-isolation as such, but to a distinct and earlier process. In my view, the primary judge was correct to accept the evidence of both Esso and AWU witnesses to the effect that an ICC is a record, not an instruction. That it might be used to verify valve positions, or as a checklist, does not invest it with the status of a guide to work. The simple fact that it does not record an order of valve manipulation makes that clear enough, I would have thought. In my view, however the matter is examined, no error in the analysis by the primary judge of the issue has been demonstrated. There are two further aspects of the present matter which seem to me to warrant some additional emphasis. The first is, as the primary judge observed, that the requirements of s 414(6) cannot be seen as operating for the benefit of whoever gives a notice which the subsection requires. The correct identification of whether action is protected industrial action serves to mark the limits of the immunity given by s 415, and the limits on the operation of s 421 of the FW Act as well as s 418. The recipient of such a notice (whether employer, union or employee) is entitled to sufficient precision to identify what is, and what is not, immune from legal challenge. The second matter is more specifically related to the facts of the present case. The discussion about “de-isolation” was not a discussion about ordinary English. It was a discussion loaded with technical and operational significance. That technical and operational significance extended to important questions of safety. It assumed a body and application of expert, highly sophisticated, engineering principles reflected in the operation manuals. That body of knowledge and expertise was not the province of individuals; it was an accumulated, synthesised, technical wisdom having far-reaching consequences. It is no disparagement of the knowledge and appreciation of individual employees to say that what was crucially important (and consistent with authority) was what Esso, as operator of the Longford plant, would understand from the notice of industrial action which it received. Whether particular forms of industrial action would be protected (i.e. whether they fell within or outside the notification) could be critical to Esso’s operational response and its evaluation of the prudence of that response, a question which might go well beyond industrial consequences. In my view, there is a limit to the extent to which any form of textual debate about Esso’s manuals and engineering practices in the isolated atmosphere of a courtroom could be used to impose or attribute some meaning about those manuals and engineering practices which was not Esso’s own. In my respectful opinion, the conclusions by the primary judge at [90] identify precisely the problem with the AWU’s contention. That case was, in effect, that a court should find and declare that the operator of the Longford plant had, or should have had, a different understanding of its own manuals and processes than it did in fact have. The present is a case where any such approach might involve unanticipated dangers which a court or industrial tribunal is not equipped to assess. Moreover, the present seems to me to be a case where the primary judge had an undeniable advantage in the assessment of the evidence. Part of the AWU’s argument was that the primary judge gave too little, or too much, weight to parts of the evidence of senior Esso managers. Judgments of that sort, in a case of the present kind, are ones which a judge at first instance is best placed to make, unless the advantage which the judge had was not used, or was squandered. Neither contention is applicable to the present case. The primary judge made a conscientious and detailed evaluation of the evidence. In such a case, an appellant must identify some error, or other reason for intervention, to require an appeal court to attempt some even more remote, and truncated, consideration of the same evidence, while lacking all the natural advantage of the primary judge. No error of that quality (or any error) was able to be shown in the present case despite the diligent and persistent efforts of counsel for the AWU. I would reject this challenge. The first two declarations made by the primary judge should not be disturbed. Breach of requirements about service The next issue concerns allegations that the AWU breached aspects of the orders made on 5 March 2015 and 6 March 2015 concerned with publishing the orders made on those days. The primary judge referred to the allegations as follows (at [118]): The contraventions of s 418 orders alleged against the respondent are the following: ... (b) failure to prepare, distribute, provide or publish the written notice referred to in cl 5.1(a) of the Commission’s order of 5 March 2015; (c) failure to prepare, distribute, provide or publish the written notice referred to in cl 5.1(a) of the Commission’s order of 6 March 2015; ... Because the primary judge had found that the order made on 5 March 2015 (which was in general terms) did not comply with s 418, and was not valid, it was not necessary to give further attention to (b) (see [121]). Attention was therefore confined to (c). This was the subject of declaration 3. The primary judge made an important observation at [122]: With respect to allegation (c) on the list above, it is uncontroversial that the respondent did not prepare, distribute, provide or publish the notice referred to by 6:00 pm on 6 March 2015. I shall deal with its defence to this allegation next, but I should indicate at the outset that the respondent accepted that, in the setting of a valid order made under s 418 of the FW Act, provisions in the terms of cl 5.1 of the order made on 6 March 2015 were within the power of the Commission as ancillary to the substantive provisions with which they were associated, and, moreover, that cl 5.1 answered the description of “an order under section 418” within the meaning of s 421(1). I shall proceed conformably with that concession, but I make it clear that it is not a matter upon which I have been required to rule. I shall return to the significance of those apparent reservations shortly. Apart from identifying the industrial action to which it was directed, the order made by the FWC on 6 March 2015 contained further aspects. In orders 2, 4 and 6 thereof the order provided: APPLICATION and PARTIES BOUND This Order is binding on and applies to: (a) the Australian Workers’ Union (the AWU); (b) the delegates, officers, employees, servants, and agents of the AWU; (c) employees of Esso Australia Pty Ltd who are members of the AWU and who are covered by the Esso Gippsland Longford and Long Island Point Enterprise Agreement 2011 (Employees); (d) Esso Australia Pty Ltd (Esso). ... INDUSTRIAL ACTION MUST STOP AND NOT BE ORGANISED 4.1 On and from the time specified in clause 6.1 of this Order: (a) the AWU; (b) delegates, officers, employees, and agents of the AWU; (c) the Employees; must not organise any Industrial Action. 4.2 On and from the time specified in clause 6.1 of this Order, Employees must not engage in Industrial Action. ... TERM AND DATE OF EFFECT This order shall come into effect at 6.00 pm on Friday 6 March 2015 and shall cease to have effect at 6.00 pm on Friday 20 March 2015. Those clauses have the effect of stating, and limiting, the scope, application and period of operation of the order sustained by s 418. There has been, and can be, no criticism of those aspects, but they stand in contrast to the matters which now require attention. Order 5 provided as follows: SERVICE OF ORDER 5.1 The AWU must: (a) as soon as practicable following the time specified in clause 6.1 of this Order, prepare a written notice (the Written Notice) signed by an authorised official of the AWU in the following terms: “The Fair Work Commission has issued a section 418 order to stop or prevent unprotected industrial action. The unprotected industrial action identified by the Fair Work Commission was the refusal of employees to perform equipment testing, air freeing and leak testing. The order is called the Esso Australia Pty Ltd Industrial Action Order 2015 No 3 (the “Order”). The Order prohibits the AWU and members of the AWU from organising and engaging in any ban, limitation or restriction on the performance of equipment testing, air freeing and leak testing. The Order applies to: (i) the AWU; (ii) delegates, officers, employees, and agents of the AWU; (iii) employees of Esso Australia Pty Ltd who are members of the AWU and who are covered by the Esso Longford and Long Island Point Agreement 2011. AWU members must not refuse to perform, or impose limitations or restrictions on the performance of equipment testing, air freeing and leak testing for the duration of the order.” (b) by 6.00 pm on 6 March 2015, the AWU must: (i) distribute the Written Notice to all Employees who are members of the AWU by email and by SMS message “Esso members are prohibited from taking unprotected industrial action. Please see link:” with a link to the Written Notice; (ii) provide a copy of the Written Notice to Esso; (iii) publish the Written Notice on a prominent place on the AWU website www.awu.net.au; 5.2 As soon as practicable following the time specified in clause 6.1 of this Order, Esso must: (a) email all Employees, which must: (i) attach a copy of this order; and (ii) contain the text of the Written Notice outlined in 5.1(a) but adapted to clarify that the Written Notice is communicated by Esso to the Employees. (b) place a copy of this order on the notice board(s) usually used by Esso for the purposes of communicating with the Employees. Those are directions about service. In my view, they are not readily seen as authorised by s 418 and it is not necessary that they should be. The scope of the order (to the extent found valid by the primary judge) was sufficiently clear, the persons bound were stated and the time at which the order came into effect had been stated. Obviously enough, in any subsequent proceedings under s 421 of the FW Act it would be necessary to prove that persons alleged to have contravened the order had been served with it, or otherwise had actual or constructive knowledge of its existence. But that would be an issue in those proceedings. It need not be addressed, necessarily, by the FWC. Order 5 seems to erect an independent and free-standing obligation on both the AWU and Esso to publish the order. In cl 5.1(a) and cl 5.2 the obligation does not arise until after the order has come into effect. Clause 5.1(b) requires the AWU to distribute material (that it is not yet required to prepare) before the order comes into effect and before the time for preparation. In my respectful view, order 5 was ineffective to impose any sensible obligation to distribute material before the order took effect. The primary judge (in my view, correctly) dismissed the allegation of breach of cl 5.1(b). However, he found a breach of cl 5.1(a). As recorded in [122] set out earlier, the AWU accepted that an order like 5.1 was generally within power. Its defences were that it was impossible to comply (summarised by the primary judge at [123] and [124]) and that it had otherwise effectively informed its members of the order. The primary judge rejected the “impossibility” defence and reserved consideration of the other defence until he dealt with questions of remedy. As the general validity of an order like order 5 has not been put in issue, it should not be dealt with differently on appeal than by the primary judge at [122]. However, I do not wish to let that question pass without further comment, before I return more specifically to the question of cl 5.1(a). The concession by the AWU that an order in the general terms of order 5 is authorised by s 418 was attributed, on the appeal, to observations by Gray and North JJ in TWU at [52]: Clause (iii) of para 4(c) of the order imposes an obligation to take positive action, as distinct from refraining to do something. This does not necessarily cause it to travel beyond the power of the Commission. It may be that an order to take some positive step can be seen as necessary for, incidental to, or consequential upon an order that industrial action stop, not occur or not be organised. In the absence of any finding that there was any direction, advice or authorisation to delegates or members of the TWU NSW in the present case, it is hard to see how an order requiring immediate advice that such direction, advice or authorisation was withdrawn and that any action must cease was justified. (Emphasis added.) Two points may be made about the statements I have emphasised. First, they do not satisfactorily accommodate order 5 in the present case. Orders 2, 3 (to the extent valid), 4 and 6 were fully effective without order 5. Order 5 established a free-standing administrative obligation, separate from the rest of the order. The AWU, as a party bound by the order, could have easily been ordered to take all necessary steps to bring about a cessation of industrial action by employees (for example, by inserting an additional cl 4.3 to that effect). Secondly, in my view their Honours in TWU were simply leaving open a possibility because it was unnecessary to say anything further about it. They struck down the clauses in question (at [47]-[54]) on other grounds. On the appeal, the AWU referred also to observations by Brennan J in Johns v Australian Securities Commission [1993] HCA 56 ; (1993) 178 CLR 408 at 428-429 and by a Full Court of this Court in Keen Mar Corporation Pty Ltd v Labrador Park Shopping Centre Pty Ltd & Anor [1989] FCA 54 at 33; [1989] FCA 46 ; (1989) 67 LGRA 238 at 241. The statements referred to the well-known principle that a grant of power under a statute extends to what is fairly or reasonably incidental to that power. The central question which might have arisen for consideration is whether order 5 was authorised by s 418. I have some doubt about that. In my view, if order 5 had been challenged as being outside the grant of power in s 418 there may have been very good grounds to strike it down. However, as the primary judge observed, the question was not argued and, therefore, no more need be said about that aspect in the present proceedings. There may, nevertheless, remain room to doubt the validity of cl 5.1(a), standing unconnected with cl 5.1(b), because it is impossible to see how such an order could be incidental to the function performed by the FWC under s 418. In other words, I do not see how cl 5.1(a) can survive the complete failure of cl 5.1(b). A reservation of this kind was not raised on the appeal, however, and I shall put it aside also. Having done so, I conclude nevertheless that there could be no breach of cl 5.1(a) for two reasons. Either it was intended in aid of cl 5.1(b) and can have no effective legal operation in light of the complete failure of cl 5.1(b) as a legal obligation; or, it was intended as part of a composite obligation to prepare and distribute material which cannot survive independently. In either case, in my respectful view, the primary judge should not have found that a breach of cl 5.1(a) was established. The appeal against declaration 3 made by the primary judge should therefore be upheld. I would set that declaration aside. The operation of section 413(5) Section 413(5) of the FW Act states one of the common requirements for industrial action to be protected industrial action. It provides: Compliance with orders (5) The following persons must not have contravened any orders that apply to them and that relate to, or relate to industrial action relating to, the agreement or a matter that arose during bargaining for the agreement: (a) if the person organising or engaging in the industrial action is a bargaining representative for the agreement—the bargaining representative; (b) if the person organising or engaging in the industrial action is an employee who will be covered by the agreement—the employee and the bargaining representative of the employee. Having made findings that the AWU had contravened s 418 in the particular respects identified in declarations 1 to 3 set out earlier, the primary judge was then required to assess the consequences of those findings for the argument put by Esso that no further protected industrial action could be taken in support of enterprise bargaining for a new agreement or agreements. It is not necessary, in my view, to further consider declaration 3, which I would set aside. However, I have concluded that declarations 1 and 2 were validly made, and the basic argument remains the same. The primary judge took a view of the operation of s 413(5) which would, if he had given effect to it, have meant that thereafter the AWU could not engage at any future time in any further protected industrial action in support of bargaining for any of the proposed enterprise agreements. However, for reasons of comity, the primary judge thought it appropriate to apply the construction of s 413(5) given a short time earlier by Barker J in Australian Mines and Metals Association Inc v The Maritime Union of Australia [2015] FCA 677 ; (2015) 251 IR 75 (“ AMMA ”). In AMMA , Barker J expressed the view that only orders applying at the time “industrial action is proposed” (see at [171]) fall within the operation of s 413(5). Applying the same approach, the primary judge confined his further declaratory order, arising from the operation of s 413(5), as follows: By operation of s 413(5) of the FW Act, from 6:01 pm on 6 March 2015 until 6:00 pm on 20 March 2015, all industrial action organised by the respondent and taken by the Longford members in support of claims in relation to bargaining for a replacement enterprise agreement or enterprise agreements for the Esso Gippsland (Longford and Long Island Point) Enterprise Agreement 2011 was unprotected industrial action. There are two features of this declaration to be noted. The first is that it was limited to the period of operation of the s 418 order found to be partly valid. As will appear below, in my view the approach taken by the primary judge to that question was the correct one. It might be conveniently noted at this point that the matter being addressed by the declaration concerned the period in which any industrial action would be unprotected. That period was established by the terms of the valid s 418 order, even though declarations 1 and 2 identified two specific periods of particular action in breach of the s 418 order. The second feature is that the fourth declaration was confined in its operation by a reference to AWU members at Longford and to industrial action in support of an enterprise agreement at Longford. Esso challenged the adoption of the approach taken in AMMA , and confinement of the order to industrial action at Longford. It is convenient to deal with that second aspect first. It is clear from the judgment of the primary judge that he considered the various issues before him by reference to the circumstances at Longford. That is clear from the very first paragraph of the judgment (at [1]): In this proceeding, the applicant, Esso Australia Pty Ltd, alleges that the respondent, The Australian Workers’ Union, contravened ss 340(1) , 343 (1), 346 , 348 and 421 (1) of the Fair Work Act 2009 (Cth) (“the FW Act”) when it organised industrial action by its members employed by the applicant at its plant at Longford, Victoria. The fact that industrial action, in the form of bans and stoppages, was taken by those members, and was organised by the respondent, is uncontroversial. The issues in the case are whether the organisation of action amounted, in the circumstances which existed, to contraventions of the provisions referred to. Thereafter, the references to the evidence concern Longford. However, the primary judge also described the operational and enterprise agreement position early in the judgment in more general terms. In particular, he identified the three enterprise agreements which the AWU was seeking to replace. The declaration sought by Esso before the primary judge, based on s 413(5), was not confined to Longford. Esso’s contention was that the AWU and its members were pursuing a single, integrated campaign for either one new enterprise agreement to replace all three existing agreements applying at all the relevant locations, or three replacement agreements. The AWU organised, and the AWU members took, industrial action in support of the totality of the claims. That was not disputed. Accordingly, breach of an order applying to the bargaining process as a whole was, Esso contended, a breach which affected whether future industrial action at any of the facilities would be protected or not. Esso contended that the common requirements in s 413 could not be met once any breach of an order had occurred in the bargaining process. After judgment was delivered, this contention was pursued in further proceedings directed to establishing the form of order to be made to give effect to findings in the judgment. The primary judge expressed the view on transcript that insufficient attention had been given to other facilities to include a reference to them in the declaration. He was concerned, in particular, that there may be issues of statutory construction which should not be passed over by potentially unwarranted assumption. It appears, from the position taken by the AWU on the appeal, that the primary judge need not have been concerned about that issue. Furthermore, it is not in contest that a co-ordinated approach to a replacement agreement or agreements was being adopted by the AWU. On the findings which were made, therefore, it seems to me that Esso was entitled to the wider order it sought, in that respect. I may now return to the larger question of whether declaration 4 should have been confined to the period of the s 418 order, or whether breach of that order had the consequence that there could be no further protected industrial action taken in support of bargaining for any of the agreements, even after the s 418 order had no further application. The question of the proper construction of s 413(5) has been considered by the same Full Court which is dealing with the present appeals, in an appeal in AMMA . Judgment in the AMMA appeal is being handed down with this judgment ( Australian Mines and Metals Association Inc v The Maritime Union of Australia [2016] FCAFC 71). In the AMMA appeal judgment, I express my preference for a construction of s 413(5) whereby an assessment of whether a bargaining representative, such as the AWU, fails to comply with a relevant order which applies to it when it organises or engages in industrial action focusses upon whether there is, at the relevant point of time, an existing or current order with which it is not complying, rather than whether at some time in the past it has failed to comply with an order. That construction is, in substance, the one favoured by Barker J in AMMA . My reasoning and explanation for that construction should be taken as incorporated here. On the findings made by the primary judge in the present case, once the declared breach of cl 5.1(a) of the order of 6 March 2015 is put aside, the declared breaches of orders are those in declarations 1 and 2. Those declared breaches will have caused other industrial action occurring at that time to cease to meet the common requirements in s 413 so that it all became unprotected, during the period of operation of the s 418 order – i.e. 6 to 20 March 2015. Industrial action occurring during that period was referred to by the primary judge at [35], [40] and [42]: At Longford, the industrial actions, organised by the respondent, which were taken on and from 12 February 2015 were the five bans referred to in items (a)-(e) of the respondent’s letter of 3 February 2015 and the stoppages referred to in item (f). The bans referred to in items (b)-(e) remained in place until an order made by the court on 17 March 2015 (see para 65 below). The ban referred to in item (a) remained in place until withdrawn on 13 March 2015. The stoppages referred to in item (f) continued until 12 March 2015. ... On the same day as the Commission made its order, 17 February 2015, the respondent gave a fifth notice of intention to take protected industrial action, namely, an “indefinite ban on the performance of overtime by Longford employees commencing at 12.01 a.m., Thursday, 26 February 2015”. That ban was implemented on 26 February 2015 as notified, and remained in place until 17 March 2015. ... On 2 March 2015, the respondent gave its seventh protected industrial action notice. By this notice, and effective on and from 12 March 2015, the stoppage referred to in item (f) in the respondent’s letter of 3 February 2015 was replaced by a like stoppage of Longford employees, but commencing at 7.30 am each day. Those stoppages continued until 17 March 2015. In my view, therefore, an order of the kind in declaration 4 was appropriate but the order made by the primary judge should be amended in two respects: it should not be confined by reference to AWU members at Longford, but should refer to those members more generally; and, it should identify all three existing enterprise agreements. Coercion The remaining elements of the AWU’s appeal deal with alleged breaches of ss 343, 346 and 348 of the FW Act. They provide: Coercion (1) A person must not organise or take, or threaten to organise or take, any action against another person with intent to coerce the other person, or a third person, to: (a) exercise or not exercise, or propose to exercise or not exercise, a workplace right; or (b) exercise, or propose to exercise, a workplace right in a particular way. Note: This subsection is a civil remedy provision (see Part 4-1). (2) Subsection (1) does not apply to protected industrial action. ... 346 Protection A person must not take adverse action against another person because the other person: (a) is or is not, or was or was not, an officer or member of an industrial association; or (b) engages, or has at any time engaged or proposed to engage, in industrial activity within the meaning of paragraph 347(a) or (b); or (c) does not engage, or has at any time not engaged or proposed to not engage, in industrial activity within the meaning of paragraphs 347(c) to (g). Note: This section is a civil remedy provision (see Part 4-1). ... Coercion A person must not organise or take, or threaten to organise or take, any action against another person with intent to coerce the other person, or a third person, to engage in industrial activity. Note: This section is a civil remedy provision (see Part 4-1). It will be noted that s 348 contains no equivalent of s 343(2), a matter which will require further discussion. The particular “workplace right” and “industrial activity” relied upon in the present case by Esso are identified in s 341 and s 347 respectively. Section 341(1)(b) and (2)(e) provide: Meaning of workplace right Meaning of workplace right (1) A person has a workplace right if the person: ... (b) is able to initiate, or participate in, a process or proceedings under a workplace law or workplace instrument; ... ... Meaning of process or proceedings under a workplace law or workplace instrument (2) Each of the following is a process or proceedings under a workplace law or workplace instrument : ... (e) making, varying or terminating an enterprise agreement; ... (Emphasis in original.) Section 347(b)(iv) provides: Meaning of engages in industrial activity A person engages in industrial activity if the person: ... (b) does, or does not: ... (iv) comply with a lawful request made by, or requirement of, an industrial association; ... ... (Emphasis in original.) I shall return to the allegation of breach of s 346 a little later. As to the allegations of breach of s 343 and s 348, in each case Esso’s allegation was that the AWU took action against it with intent to coerce it into making an enterprise agreement. Esso had the benefit of the operation of s 360 and s 361 of the FW Act, which provide: Multiple reasons for action For the purposes of this Part, a person takes action for a particular reason if the reasons for the action include that reason. Reason for action to be presumed unless proved otherwise (1) If: (a) in an application in relation to a contravention of this Part, it is alleged that a person took, or is taking, action for a particular reason or with a particular intent; and (b) taking that action for that reason or with that intent would constitute a contravention of this Part; it is presumed that the action was, or is being, taken for that reason or with that intent, unless the person proves otherwise. (2) Subsection (1) does not apply in relation to orders for an interim injunction. The AWU, therefore, was required to prove, on the balance of probabilities, that it did not have an intent to coerce Esso to make an enterprise agreement when it took non-protected action. For that purpose, Esso did not rely on protected industrial action to sustain any claim under s 348. Coercion has been held to require the satisfaction of two elements: negation of choice; and, the use of unlawful, or illegitimate or unconscionable means (see the discussion in Finance Sector Union of Australia v Commonwealth Bank of Australia [2000] FCA 1468 ; (2000) 106 FCR 16 ; Seven Network (Operations) Ltd v Communications, Electrical, Electronic, Energy, Information, Postal, Plumbing and Allied Services Union of Australia [2001] FCA 456 ; (2001) 109 FCR 378 (“ Seven Network ”); Fair Work Ombudsman v National Jet Systems Pty Ltd [2012] FCA 243 ; (2012) 218 IR 436 and Victoria v Construction, Forestry, Mining and Energy Union [2013] FCAFC 160 ; (2013) 218 FCR 172). The AWU argued before the primary judge and on appeal, that an intent to coerce could not be established unless it was proved that there was intent with respect to both elements – i.e. an intent to negate choice and an intent to use unlawful, illegitimate or unconscionable means to bring about that result. At the same time, however, the AWU submitted that the second element had also to be objectively established and that it was not sufficient to point to a belief or understanding that action might, if taken, be unlawful, etc. In my view, it should not be accepted that it is necessary to establish that a person intended to act unlawfully etc. Nor would it be a defence to show that a person believed their action would be, or was, lawful. The AWU relied, at first instance and on appeal, on observations made by Merkel J in Seven Network . The primary judge held that the passages in question did not convey the asserted meaning and that, in any event, such an approach should not be accepted. I agree. I propose to examine first how the primary judge approached the matter, and then return to look more closely at Seven Network . The primary judge first dealt with the significance and operation of s 343(2). He concluded that s 343(2) was a complete defence if the relevant action was, or would be, protected industrial action, but that an erroneous belief to that effect would afford no defence. In my respectful view, that is the correct approach. Similarly, it does not matter if a belief is held that conduct is not unlawful, if it is. The nub of s 343 is that it is not necessary to wait until the action takes its toll and has overborne the will of the target. An intention to negate choice will suffice, if the conduct is unlawful, illegitimate or unconscionable. There is no room for debate if the conduct is protected industrial action. However, it may be important in some cases to bear in mind that industrial action is not rendered unlawful, illegitimate or unconscionable only because it is not protected industrial action. The absence of the immunity given by s 415 does not carry that immediate consequence. The specific allegation was that the AWU had “organised” particular bans. Two aspects of the allegations were upheld as being not protected and were reflected in declarations 1 and 2 referred to earlier. The primary judge examined the conduct and intent of the Victorian Branch Secretary of the AWU, Mr Ben Davis, to understand the “corporate” intent of the AWU. Although the primary judge referred in various places to the necessity for an intention to negate choice, at [174]-[176], he said: In the light of the evidence to which I have referred, I find that the intent of Mr Davis, and therefore of the respondent, in organising industrial action conformably with the notification of 3 February 2015 was to apply sufficient direct pressure on the applicant to cause it to act otherwise than in the exercise of its own free choice. It was to cause it to agree to terms in a prospective enterprise agreement to which it would not, as a matter of choice, have agreed in the absence of that pressure. In my view, that application of pressure was illegitimate. In every respect, the bans and stoppages notified on 3 February 2015 involved refusals by the employees concerned to perform some aspects of their required, customary, duties pursuant to their contracts of employment. The obligation to serve lies at the heart of any employment relationship. The conclusion that it is illegitimate for an employee to refuse to serve as a means of extracting beneficial terms from his or her employer is one that will rarely be difficult to draw. On the facts of the present case, when, to use Mr Davis’s concession, the employer was in a vulnerable position, this conclusion is readily to be drawn, and I do so. For the above reasons, I find that the industrial action notified by the respondent on 3 February 2015 was organised with intent to coerce the applicant to make an enterprise agreement, or to make one on terms acceptable to the respondent. Then the primary judge found that those bans and stoppages ceased to be protected when the AWU contravened the s 418 order made by the FWC on 6 March 2015, because of s 413(5). Then the primary judge referred to his earlier findings that the ban on air freeing and leak testing imposed on 4 March 2015 (shortly thereafter the subject of the valid s 418 order from 6 pm on 6 March 2015) was not (after that time) protected industrial action and nor was organising a ban on 7 March 2015 on manipulating bleeder valves which was associated with the ban on air freeing and leak testing. The primary judge found that those bans also amounted to action taken with an intent to coerce for the same reasons. The direct challenge to those findings is twofold. First, it is contended that all of the bans represented protected industrial action so that s 343(2) applied. I have already explained why I would not accept that line of defence. Secondly, it was contended that it was necessary for the AWU to know that its conduct was coercive by understanding and appreciating that the bans would not be protected industrial action, whereas in the present case it was contended that the AWU and its officers believed that their actions were protected. This argument was advanced by reference to Seven Network and what Jessup J and Perram J said in Construction, Forestry, Mining and Energy Union v Endeavour Coal Pty Ltd [2015] FCAFC 76 ; (2015) 231 FCR 150 (“ Endeavour Coal ”) at [30] and [77] respectively about the regard to be paid to the subjective reasons of a decision-maker accused of adverse action. In my respectful view, the line of enquiry is not the same as in Endeavour Coal . I agree with the primary judge, as I said earlier, that if the defence in s 343(2) is not available on the facts, then a misplaced belief (or intention) that industrial action is protected industrial action will afford no defence to a case under s 343(1) if the elements of the construction to which I have referred are otherwise made out. Seven Network requires closer attention. The particular facts must be borne in mind. The case turned entirely on the character of threats of industrial action; not on an assessment of actual industrial action. Merkel J stated (at [1] and [3]): The present matter concerns threats alleged to have been made by the respondents to take industrial action that would interfere with the televising by the applicant, Seven Network (Operations) Limited (Seven Network), of the Australian Football League finals (the AFL finals), the Brownlow Medal count and the Olympic Games in August and September 2000. ... Seven Network seeks the imposition of penalties pursuant to s 170NF of the Act in respect of the alleged contraventions. After setting out the background to the threatened industrial action and the relevant statutory provisions (which were relevantly sufficiently similar to s 343 of the FW Act for present purposes), Merkel J posed the question for decision (at [30]-[31]), as follows: The question in the present case is whether the threats of industrial action were made with intent to coerce Seven Network not to agree to enter into a national enterprise agreement but, rather, to enter into a local enterprise agreement under Div 2 of Pt VIB of the Act. There is no real dispute that the CEPU threatened industrial action with the intent to induce Seven Network to enter into a local, rather than a national, enterprise agreement. However, the respondents contend that there was no intention to “coerce” in breach of s 170NC(1) as at all times their intention was that employees at Seven Network would take industrial action that is protected action under the Act. Mighell and McCallum stated that they believed that, as the CEPU initiated the bargaining period, the action threatened by it was protected action. Although the respondents do not dispute that the action threatened, if taken, would not have been protected action under the Act, they claim that s 170NC(1) was not contravened by them because their intent and belief, albeit wrong, was that the action they threatened was protected action. The respondents’ contention requires consideration of the meaning of the phrase “intent to coerce” in s 170NC(l). (Emphasis in original.) Then, his Honour surveyed both criminal and civil cases, in the High Court and this Court, before stating, in a passage (at [41]) which has been repeatedly adopted or approved in this Court: The ... cases establish that there must be two elements to prove “intent to coerce” under s 170NC(l). First, it needs to be shown that it was intended that pressure be exerted which, in a practical sense, will negate choice. Secondly, the exertion of the pressure must involve conduct that is unlawful, illegitimate or unconscionable. ... Then his Honour said (at [43]): It follows from the foregoing discussion that Seven Network must establish that: (a) the respondents’ threats of industrial action were made with intent to negate Seven Network’s choice by the exertion of pressure that was, in the circumstances, unlawful, illegitimate or unconscionable; and (b) the respondents had actual knowledge of the circumstances that made their conduct coercive in the sense discussed in (a) above. To this point in the analysis, in my view the position is clear. The requirement of intent applies to the purpose of negating choice. The additional element that the means employed be unlawful, etc. involves an objective test. That approach is consistent with the common law origins of the notion of coercion which can be traced back to the tort of economic duress, as explained in the cases to which I referred earlier. In that common law context, the notion of purpose, or intent, applies to the first element but not the second. There is no reason to think that the statutory adoption of the common law concept has altered its nature. I do not read Merkel J, in the passage cited above, to be making any suggestion to the contrary. Indeed, I read his Honour’s statements as confirming the position. At [44], Merkel J went on to deal with the effect of s 170NC(2) of the WR Act (now s 343(2) of the FW Act): Where it is contended that the action threatened or taken is protected action the operation of s 170NC(2) must be considered. Section 170NC(2) does not provide that subsection (1) does not apply to action that is intended to be protected action. Rather, it provides that the subsection does not apply to action that is protected action. Thus, if the action relied upon to establish a breach of s 170NC(1) is protected action then the taking of that action cannot amount to a breach of s 170NC(1). If the conduct relied upon is a threat to take action that is protected action the reason why the threat may not breach the section will be because the making of the threat to take protected action, for example by the giving of notice under s 170MO, does not have the element of unlawfulness, illegitimacy or unconscionability required to establish a breach of s 170NC(1). In that regard it may be relevant that the only threat of protected industrial action that is provided for under the Act is the giving of notice under s 170MO. Accordingly, it does not follow that threats of such action made outside of the statutory scheme (for example, a “softening up” process) can be made with impunity. In determining whether the requisite elements of s 170NC(1) are established all of the circumstances of the case must be considered. (Emphasis in original.) The primary judge analysed this passage, and placed it in the context of the FW Act as follows (at [163]-[167]): With respect, I would regard the first four sentences in this paragraph from his Honour’s reasons as unexceptionable. I do, however, regard the fifth as problematic. I would, with respect, have expressed that sentence as follows: “If the conduct relied upon is a threat to take action that is protected action the reason why the threat will not breach the section will be because the section does not apply to protected action.” By reference to the premise upon which his Honour’s fifth sentence was based, there would be no occasion to consider whether the action would have been unlawful, unconscionable or illegitimate. This point at which I would depart from his Honour is not, however, the most important dimension of his reasons in the context of the present case. That dimension is this. His Honour did not hold that it was exculpatory if, at the point in time when the action was organised or taken, the putative respondent had actual knowledge of circumstances that brought, or would bring, that action within the concept of protected action in the legislation. To the contrary, his Honour held that s 170NC(2) operated by reference to what the action was , as distinct from what the respondent intended it to be . I see nothing in his Honour’s reasons that would provide support for the proposition that, under s 170NC of the WR Act, proof that the respondent either intended or believed that his or her action was, or would be, protected, even if based on facts then known to him or her, would be sufficient to defeat the applicant’s allegation; or, for that matter, would even be relevant. Unlike s 170NC of the WR Act, s 343 of the FW Act deals specifically with the act of “organising” action. In practical situations, that may mean that the presumptively wrongful conduct occurs some time before the actual taking of the action to which it relates. It may be that, at the point in time when the respondent’s intent was formed, it appeared that the action, when taken, would be protected industrial action. In the view I take, however, the prospect of the action being protected industrial action is neither a necessary nor a relevant element in so much of s 343 as is tied to the intent of the respondent. If the elements laid out in subs (1) are established, subs (2) will save the respondent from liability only if the action was – or, in the case of action which has not yet occurred, would be – in fact protected industrial action. The next issue requiring consideration relates to the extent of the “subjective” element in the operation of s 343(1). This issue arises because of the case law which has developed around what is involved in the concept of “coercion”. The two elements, as summarised by Merkel J in Seven Network , are set out in para 161 above. That the actor have actually intended to induce the putative target to agree or not to agree to do certain things, and to apply such pressure as would negate choice in that regard, is uncontroversial. But the question arises: is it also necessary that the actor have intended that his or her action be unlawful, illegitimate or unconscionable? I would hold not. These adjectives reflect legal characterisations of the action organised or taken. From Seven Network it is clear that it will not be exculpatory for the actor to prove that he or she believed that the action was lawful. It follows, in my view, that it will be sufficient for an applicant to establish that the action organised or taken was, or would be, unlawful, illegitimate or unconscionable. It need be no part of an applicant’s case to establish that the actor intended – ie consciously in the sense conveyed by Board of Bendigo Regional Institute of Technical and Further Education v Barclay (No 2) [2012] HCA 42 ; (2012) 248 CLR 549 – that his or her action should be characterised in these terms. The question in the present case, therefore, is whether the actions relied upon by the applicant (see ...) were, or any of them was, organised or taken with the intent of applying pressure to the applicant that would negate its choice in the matter of making an enterprise agreement, either at all or on terms other than those proposed by the respondent; or, as it was put by the parties in this case, pressure that would overbear the applicant’s will in this regard. Unless the respondent proves that this was not the intent of those actions, it is presumed that it was: FW Act, s 361(1). (Emphasis in original.) Then, the primary judge turned to the facts bearing on the question of intent to which I have already referred, and found that the requisite intent to negate choice had been established. In that analysis, the primary judge put aside as irrelevant any suggestion that the industrial action was intended to be, or was believed to be, protected. His Honour found that the application of pressure, intended to negate choice, was also illegitimate. In the Seven Network case, Merkel J found, as a fact, that the relevant union officers did not have a belief that the action threatened by them was protected. The case is, therefore, not an example of the AWU’s contention being considered against the facts of a particular case. However, the actual findings suggest an approach consistent with that taken by the primary judge, rather than the contrary. Merkel J found, for example (at [59], [60]): In the above circumstances there can be little doubt that the conduct of Mighell and McCallum on behalf of the CEPU was coercive conduct. In summary, I am satisfied, on the probabilities, that the following factors constituted coercive conduct of the CEPU for the purposes of s 170NC(1): ... (i) if the action threatened had been taken by Seven Network’s employees (who were not members of the CEPU) that action would not have been protected action under the Act and would therefore have been unlawful; (j) the CEPU did not hold a positive belief that the action threatened was protected industrial action; rather, the CEPU believed that the action threatened may not be protected action but refrained from making any proper inquiry concerning that matter; ... The conduct engaged in by the CEPU constituted a threat to take industrial action with intent to coerce Seven Network not to agree to a national enterprise agreement but, rather, to agree to the local enterprise agreement sought by the CEPU. The conduct was coercive as it was a threat to take action that, in all the circumstances, was unlawful, illegitimate or unconscionable. The CEPU, through its officers Mighell and McCallum, was aware of all the circumstances set out in pars (a)-(h), (j) and (k) above that made the threat unlawful, illegitimate or unconscionable. That approach seems to me to reflect a view, consistent with that of the primary judge, that an examination of intent is relevant to the first element of coercion, and an examination of objective circumstances is relevant to the second element. I agree with the approach taken by the primary judge to the meaning and application of s 343 of the FW Act to the facts of the present case. The case involving s 348 of the FW Act has some additional complexities. At [191]-[192], the primary judge held: Both parties treated the application of s 348 to the facts of the case as relevantly indistinguishable from that of s 343. In fact, s 348 differs from s 343 in one important respect: there is no exclusion in relation to protected industrial action. In the submission of the respondent, that did not present a difficulty, since, it was put, the circumstance that the action taken was protected “... [negated] the element of intent”. Consistently with the construction of s 343 which I have held to be correct, I could not accept that submission. However, counsel for the applicant made it clear that it was no part of their client’s case to make any complaint or allegation, whether under s 348 or otherwise, about the organisation of industrial action that was protected. They ask the court to take into account only actions which the court finds to be unprotected. I propose to proceed in accordance with that indication. In the circumstances, I would hold that there were contraventions of s 348 corresponding to the contraventions of s 343 referred to above. On the appeal the AWU argued that, by reason of the content given to s 343 and s 348 by s 341(1)(b), (2)(e) and s 347(b)(iv), Esso had relied on conflicting provisions of the kind addressed by the maxim generalia specialibus non derogant . It is necessary to observe, at this point, that the same argument was advanced to answer reliance by Esso on s 346 of the FW Act, namely that the AWU had taken “adverse action” against Esso because it had not engaged in the “industrial activity” referred to in s 347(b)(iv). The primary judge rejected the “conflicting provisions” argument in this particular respect, finding the AWU had contravened s 346 by the same conduct which contravened s 343 and s 348. His Honour’s reasoning must apply to the similar contention about s 348. His Honour thought that a degree of overlap did not signify that one provision must yield to the other. I agree. Overlap of this sort is to be accommodated when penalties are fixed, or by the operation of s 556 of the FW Act which provides: Civil double jeopardy If a person is ordered to pay a pecuniary penalty under a civil remedy provision in relation to particular conduct, the person is not liable to be ordered to pay a pecuniary penalty under some other provision of a law of the Commonwealth in relation to that conduct. Note: A court may make other orders, such as an order for compensation, in relation to particular conduct even if the court has made a pecuniary penalty order in relation to that conduct (see subsection 546(5)). Nevertheless, the AWU had wished to make submissions to the primary judge about whether it had effective immunity from any form of “double jeopardy”, even in the form of declarations stating more than one contravention for effectively the same conduct. At the end of the hearing it recorded its concerns about that issue. When reasons for judgment were published, no orders were immediately made. Discussion of appropriate orders occurred later. The reasons for judgment clearly stated the primary judge’s conclusions that a contravention of each of ss 343, 346 and 348 had occurred, arising from the same conduct. The AWU appears not to have pursued the point further, or resisted the terms of the declarations prepared by Esso to reflect the particular findings. The AWU also contended on the appeal that s 556 was not applicable in the case of multiple contraventions of the FW Act, but I cannot see why the language of s 556 of the FW Act would not be applicable to contravention of other provisions of the FW Act. In my view, the challenges made by the AWU should not be accepted. It will be a matter for the primary judge how penalties are fixed having regard to the course of conduct explained in the various findings recorded by his Honour. Moreover, in my respectful view, the AWU’s reliance on the concept of double jeopardy, or the more general canon of construction ( generalia specialibus non derogant ) is misplaced. First, so far as double jeopardy is concerned, these were civil proceedings, not criminal proceedings. The idea of protection of an accused in a criminal case does not apply with the same overriding force in civil cases (see Construction, Forestry, Mining and Energy Union v Boral Resources (Vic) Pty Ltd [2015] HCA 21 ; (2015) 89 ALJR 622 ; 320 ALR 448). Furthermore, the canon of construction relied upon is operative in cases where resort to a general provision would permit avoidance of a particular limitation imposed by a more specific provision (see e.g. R v Wallis [1949] HCA 30 ; (1949) 78 CLR 529 ; Anthony Hordern and Sons Limited v The Amalgamated Clothing and Allied Trades Union of Australia [1932] HCA 9 ; (1932) 47 CLR 1). Such a concern does not arise in the present case. It is open to the legislature to create multiple offences arising out of the same, or similar, conduct. It does not infringe some overriding imperative of our system of civil law to contemplate multiple findings of contravention. When penalties are imposed it has been accepted that the same conduct should not be punished more than once, but that is a different principle, which does not arise for application until penalties are imposed. Finally, the conclusions by the primary judge that the AWU had breached s 346 of the FW Act should be accepted for the reasons which his Honour gave. Summary Each of Esso and the AWU should be formally granted leave to appeal, as they sought. The result of the foregoing conclusions for the two appeals is as follows. Esso Appeal The orders necessary appear to me to be as follows: (1) Leave to appeal be granted. (2) Declaration 4 made on 13 August 2015 be set aside and in lieu thereof it be declared: “4. By operation of s 413(5) of the FW Act, from 6.01 pm on 6 March 2015 until 6.00 pm on 20 March 2015, all industrial action organised by the respondent and taken by the Esso employees in support of claims in relation to bargaining for a replacement enterprise agreement or enterprise agreements for the Esso Gippsland (Barry Beach Marine Terminal) Enterprise Agreement 2011 , the Esso Gippsland (Longford and Long Island Point) Enterprise Agreement 2011 and the Esso Offshore Enterprise Agreement 2011 was unprotected industrial action.” (3) The appeal be otherwise dismissed. AWU Appeal The orders necessary appear to me to be as follows: (1) Leave to appeal be granted. (2) Declaration 3 made on 13 August 2015 be set aside. (3) The appeal be otherwise dismissed. Other procedural matters The course of further proceedings concerning penalties and compensation remains under the control of the primary judge. It does not appear to me to be necessary to make orders directed to those matters. There should be no order as to costs (see FW Act s 570). I certify that the preceding two hundred and twenty (220) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice Buchanan. Associate: Dated: 25 May 2016 REASONS FOR JUDGMENT BROMBERG J: The Australian Workers’ Union ( AWU ) represents operations personnel employed by Esso Australia Pty Ltd ( Esso ) at three on-shore facilities in Gippsland and a number of offshore platforms in Bass Strait. Esso is in the business of the exploration for oil and gas, the refining of petroleum and the supply of fuels, including natural gas. Its operations personnel are covered by three enterprise agreements made under the Fair Work Act 2009 (Cth) ( FW Act ). In about June 2014, and on behalf of its members employed by Esso, the AWU commenced negotiations with Esso for replacement enterprise agreements. The FW Act permits industrial action to be taken as a means of imposing pressure in the context of industrial bargaining. When taken in conformity with the FW Act, industrial action is “protected industrial action” and, broadly speaking, immune from suit. For industrial action to be protected industrial action, notice of the action must be given in accordance with s 414 of the FW Act. In February and March 2015, the AWU gave various notices of the intent of its members employed by Esso to take protected industrial action in support of their claims for replacement agreements. Industrial action organised by the AWU occurred at Esso’s Longford Plant ( Longford ). Longford is the on-shore receiving point for all crude oil and gas produced by Esso’s offshore platforms. The industrial action at Longford took the form of various bans and stoppages including, most relevantly, a ban on work which the AWU had notified as “[a]n indefinite ban on the de-isolation of equipment”. As part of the ban on the “de-isolation of equipment”, AWU members refused to carry out various tasks associated with reinstating isolated equipment back to its operative state, including three particular tasks known as “leak testing”, “air freeing” and the manipulation of bleeder valves. A dispute then ensued as to whether those tasks fell within the description “de-isolation of equipment”. The AWU’s position was that they did because the term “deisolation” had a well-known meaning which encompassed all tasks (including the three particular tasks) necessary to reinstate equipment from an isolated to an operative state. Esso’s position was that “de-isolation” had a narrower meaning which did not encompass the three particular tasks and that, as a consequence, the AWU had failed to give the requisite notice of proposed industrial action and had therefore organised unprotected industrial action. Following various proceedings before the Fair Work Commission ( FWC ), in which a number of s 418 orders were issued against the AWU, Esso commenced proceedings in this Court claiming that the AWU had contravened ss 340(1), 343(1), 346, 348 and 421(1) of the FW Act. Esso sought injunctions, compensation for loss, and the imposition of pecuniary penalties. After an expedited hearing, the primary judge gave reasons for his decision ( Esso Australia Pty Ltd v The Australian Workers’ Union [2015] FCA 758). On 13 August 2015, the primary judge declared that the AWU had contravened ss 421(1), 343, 346, and 348 of the FW Act. A declaration was also made that, by operation of s 413(5), all industrial action organised by the AWU between 6–20 March 2015 was unprotected industrial action. A hearing was scheduled to determine Esso’s claims for compensation and penalties. Both the AWU and Esso have appealed the primary judge’s judgment. In each case, it was indicated during hearing that leave to appeal would be granted, although that has not yet been formalised by the making of an order. The central issue raised on the appeals is whether the AWU organised unprotected industrial action. All of the declarations of contravention made by the primary judge, other than one, depend upon a foundational finding made by the primary judge that the AWU organised unprotected industrial action. That finding, in turn, is based on the primary judge’s holding that there was a refusal by AWU members to perform the three tasks of leak testing, air freeing and the manipulation of bleeder valves. Contrary to the AWU’s case, the primary judge held that those tasks would not have been reasonably understood by Esso as falling within the description “[a]n indefinite ban on the de-isolation of equipment”. That issue raises s 414 of the FW Act and the notification requirements for the taking of protected industrial action. It involves the determination of difficult questions as to how a particular term with multiple denotations would have been reasonably understood. I will turn to deal with that issue first, before considering other important questions including the proper construction of ss 413(5) and 418. DID THE AWU COMPLY WITH THE NOTICE REQUIREMENTS? Legal principles At the heart of the issue that arises for determination on this part of the case is the meaning and effect of s 414(6). Section 414(6) states the content requirements ( content requirements ) for a notice of industrial action. It provides (emphasis added): (6) A notice given under this section must specify the nature of the action and the day on which it will start. Section 414(6) is one of a number of requirements ( notice requirements ) set out in s 414. The other sub-sections of s 414 are these: Notice requirements—employee claim action (1) Before a person engages in employee claim action for a proposed enterprise agreement, a bargaining representative of an employee who will be covered by the agreement must give written notice of the action to the employer of the employee. (2) The period of notice must be at least: (a) 3 working days; or (b) if a protected action ballot order for the employee claim action specifies a longer period of notice for the purposes of this paragraph—that period of notice. Notice of employee claim action not to be given until ballot results declared (3) A notice under subsection (1) must not be given until after the results of the protected action ballot for the employee claim action have been declared. Notice requirements—employee response action (4) Before a person engages in employee response action for a proposed enterprise agreement, a bargaining representative of an employee who will be covered by the agreement must give written notice of the action to the employer of the employee. Notice requirements—employer response action (5) Before an employer engages in employer response action for a proposed enterprise agreement, the employer must: (a) give written notice of the action to each bargaining representative of an employee who will be covered by the agreement; and (b) take all reasonable steps to notify the employees who will be covered by the agreement of the action. A consideration of the meaning of the content requirements should begin with a consideration of the text itself. The meaning of the text may require consideration of the context, which includes the general purpose and policy of the provision and in particular the mischief the provision is seeking to remedy: Alcan (NT) Alumina Pty Ltd v Commissioner of Territory Revenue (Northern Territory) [2009] HCA 41 ; (2009) 239 CLR 27 at [46] (Hayne, Heydon, Crennan and Kiefel JJ). The text suggests that the content requirements are not intended to be onerous. Only two matters need to be addressed: the “nature of the action” and the “day on which it will start”. The phrase “the action” in s 414(6) is a reference to “employee claim action” referred to in s 414(1), “employee response action” referred to in s 414(4), or “employer response action” referred to in s 414(5). In each case, it is apparent from the text of ss 409, 410 or 411 that “action” is a reference to “industrial action”. Leaving aside the qualifications provided for by s 19(2), the term “industrial action” is defined in s 19(1) as follows (emphasis added by unbolded italics): Meaning of industrial action (1) Industrial action means action of any of the following kinds : (a) the performance of work by an employee in a manner different from that in which it is customarily performed, or the adoption of a practice in relation to work by an employee, the result of which is a restriction or limitation on, or a delay in, the performance of the work; (b) a ban, limitation or restriction on the performance of work by an employee or on the acceptance of or offering for work by an employee; (c) a failure or refusal by employees to attend for work or a failure or refusal to perform any work at all by employees who attend for work; (d) the lockout of employees from their employment by the employer of the employees. Note: In Automotive, Food, Metals, Engineering, Printing and Kindred Industries Union v The Age Company Limited , PR946290, the Full Bench of the Australian Industrial Relations Commission considered the nature of industrial action and noted that action will not be industrial in character if it stands completely outside the area of disputation and bargaining. It is necessary to observe that the definition categorises industrial action by “kind”. A “lockout” is the only kind of industrial action identified for employer action and is defined in s 19(3) as action that “prevents the employees from performing work under their contracts of employment without terminating those contracts”. Multiple “kinds” of employee industrial action are identified in s 19(1), but in each case, the action concerns the non-performance or the altered performance of work. The differentiation in the definition between one kind of non-performance and the next is largely driven by whether the non-performance is full or partial. In each case, the characterisation of industrial action by its “kind” is given at a very high level of generality. “Kind” is a synonym for “nature”, as is “character”, or “sort” (see the definition of “nature”, Macquarie Dictionary (5th ed., 2009). They are words which, when used to call for a description of an activity or a thing, evince an intent for generality rather than particularity, in the description to be provided. The text of s 414(6) strongly suggests that, in requiring that “the nature of the action” be specified in the notice, s 414(6) demands no more than a broad or general description of the industrial action proposed to be taken. However, a description can be given at various levels of generality and the context in which the description has been called for may indicate the level of generality required. It is necessary, then, to turn to context before returning to the text. A consideration of the context requires that the provisions surrounding s 414(6) be considered, together with the place of the notice requirements in the scheme of which they form part. I will begin with the latter. Broadly stated, and consistently with the objects set out at ss 3(f) and 171(a), the FW Act provides for enterprise-level collective bargaining for the making of enterprise agreements, as a means of enabling industrial parties to resolve their industrial disputes. Collective bargaining is a process in which employees bargain with their employer as a collective rather than individually. To be successful, bargaining usually involves the making of concessions. Sometimes concessions are freely made, but an inherent feature of a collective bargaining regime is the recognition that concessions may need to be extracted through the application of industrial pressure. Industrial action is an available form of pressure and the capacity to lawfully exert such pressure, including by inflicting loss or damage, is permitted but is subject to certain conditions. As to the last-mentioned characteristic of collective bargaining, the FW Act calls permitted industrial action “protected industrial action”. The notice requirements must be met in order for industrial action to be protected industrial action (s 413(4)). If met, together with each of the other “common requirements” specified by s 413, the industrial action taken will be protected industrial action, meaning that, subject to the qualifications expressed by s 415, the industrial action will be immune from suit. It is evident, then, that compliance with the notice requirements is a condition for the grant of the statutory immunity conferred by s 415 to facilitate the taking of industrial action and the exertion of industrial pressure in the context of collective bargaining. The notice requirements are part of the requirements that those who have engaged in industrial action must have satisfied in order that they are able to take the benefit conferred upon them by the FW Act. The notice requirements differ as between different kinds of protected industrial action. It is necessary to explain those differences. There are three categories of protected industrial action contemplated by Part 3–3 and specified by s 408. “Employee claim action” is industrial action for a proposed enterprise agreement that is organised or engaged in for the purpose of supporting or advancing claims made for or on behalf of employees (s 409(1)(a)). “Employee response action” is industrial action for a proposed enterprise agreement that is organised or engaged in as a response to industrial action taken by an employer (s 410(1)(a)). “Employer response action” is industrial action for a proposed enterprise agreement that is organised or engaged in by an employer as a response to employee claim action (s 411(a)). Employee claim action must be authorised by a “protected action ballot” (s 409(2)). Protected action ballots are the subject of Div 8 of Part 3–3, which provides for the conduct of a secret ballot of employees for the purpose of authorising protected industrial action for a proposed enterprise agreement. Under that process, the application for a protected action ballot order must specify the question or questions put to the employees for ballot, including the specification of the same matter as that required by s 414(6), namely, “the nature of the proposed industrial action”: s 437(3)(b). A “protected action ballot order” made by the FWC must also specify the “nature of the proposed industrial action” (s 443(3)(d)) and the order must be provided by the FWC to the employer of the employees to be balloted (s 445(b)). As the results of a protected action ballot must also be provided to the employer (s 457(1)(b)(ii)), notice of “the nature of the proposed industrial action” that the employees have authorised will be provided to the employer in writing, as soon as practicable after the vote (s 457(1)). Beyond notice to the employer of that kind before employee claim action commences, a bargaining representative of the employees who will engage in the employee claim action must give the employer written notice (s 414(1)). The period of notice must be at least three working days unless the protected action ballot order has specified a longer period (s 414(2)). Pursuant to s 443(5), the FWC may provide for a longer period of notice of up to seven working days if satisfied that the longer period is justified by “exceptional circumstances”. A notice of proposed industrial action must not be given until after the result of the protected action ballot has been declared (s 414(3)). The notice given “must specify the nature of the action and the day on which it will start” (s 414(6)). Unlike employee claim action, there is no requirement that employee response action be authorised by a protected action ballot. No advance warning of the nature of the industrial action that may be taken against an employer will be provided through a balloting process as is the case for employee claim action. Further, whilst written notice is required and “must specify the nature of the action and the day on which it will start”, there is no requirement for any period of notice to expire prior to the industrial action commencing. In other words, the industrial action notified may be taken from the moment that the notification is given. The position is the same in relation to employer response action. Whilst written notice specifying “the nature of the action and the day on which it will start” is required to be given, employer response action may commence as soon as the notice is provided (s 414(5)). For “employer response action”, written notice is required to be provided to each bargaining representative of an employee, and the employer is required to take “all reasonable steps to notify the employees” of the action (s 414(5)). What, then, is the discernible mischief that the content requirements seek to address? It seems to me that the context does provide some assistance. The fact that notice of industrial action must be given to its intended target supports the idea that the notice requirements were intended to provide some utility to the person or persons the subject of the proposed industrial action. The provision of notice foreshadows an event. Notice is commonly given so that the person notified can make some preparation for the foreshadowed event. To my mind, that must have been a purpose of the notification provisions. But the capacity to prepare will usually be dependent upon the extent of the information that the notice gives about the foreshadowed event. Where the event is industrial action, three categories of preparation may be of utility to the target of the proposed industrial action. First , with the forewarning of the notice, proceedings can be brought to stop the proposed industrial action from occurring if the FW Act provides a basis for the FWC or a court to make an order that the proposed industrial action not occur (as it does in various circumstances: ss 418, 419, 424). Second , prior warning may provide the target with time to prepare its own industrial action as a timely response to the claim action. Third , time and knowledge of what is proposed may provide an opportunity for defensive action to be taken to defeat or diminish the impact of the proposed industrial action: Davids Distribution Pty Ltd v National Union of Workers [1999] FCA 1108 ; (1999) 91 FCR 463 at [87] . The opportunity to take defensive action is likely to have utility, particularly for an employer. But there are two reasons why the potential for utility, and thus the priority that may have been given to it by the notification requirements, should not be overstated. First , the efficacy of taking defensive action will often be undermined by a lack of certainty as to whether the proposed action will actually occur. There is nothing in the FW Act which requires notified industrial action to be taken either in whole or in part and it is often the case that industrial action is called off. Knowing what action might be taken may not be particularly useful, especially where covering for some or all of the possibilities involves significant costs. Second , the fact that the notification requirements do not require any effective forewarning for responsive industrial action tends against any suggestion that the notification requirements were enacted with high-order priority given to aiding the taking of defensive action. Whilst the existence of significant likely utility for the person informatively notified of proposed industrial action cannot be denied, the fact of that utility does not necessarily mean that the interests of that class of person were fully recognised and taken up by the notification requirements. That is particularly so when it is recognised, as I seek to do in the observations that follow, that provision of detailed notification involves disutility for the notifier. In the context of competing interests, it is necessary to discern where Parliament sought to strike the balance between the interests of the notifier and those of the notified. The notice requirements were unlikely to have been intended to neutralise or significantly compromise the FW Act’s objective of facilitating the taking of effective industrial action in the limited circumstances in which that is permitted. As Wilcox and Cooper JJ said in Davids at [84] in addressing a legislative predecessor to s 414(6): The question addressed by North J in the lengthy passage just quoted is one of considerable difficulty, about which people may reasonably reach different conclusions. Parliament did not indicate what degree of specificity it intended by the term “nature of the intended action”. To interpret this term, on the one extreme, as requiring no more than an indication of industrial action, as argued by NUW, would be significantly to devalue s 170MO(5); the notice would provide little information. To interpret it, on the other extreme, as requiring precise details of every future act or omission would be to impose on the giver of a notice an obligation almost impossible to fulfil. Industrial disputes are dynamic affairs. Decisions as to future steps often need to be made at short notice, sometimes in response to actions of the opposing party or other people, including governments, and changing circumstances. It would be a major, and unrealistic, constraint on industrial action to require a party to specify, three clear working days in advance, exactly what steps it would take. An unduly demanding interpretation of s 170MO(5) would seriously compromise the scheme of Div 8 of Pt VIB of the Act; it would be difficult for a party to an industrial dispute to obtain the protection contemplated by the Division. The observation that industrial disputes are dynamic affairs and that it would be an unrealistic constraint on industrial action to require industrial parties to specify, three days in advance, exactly what steps they intended to take, was made prior to protected action ballots being a mandatory precondition, in every case, to protected employee claim action. As earlier stated, the “nature” of the proposed industrial action must also be specified in a protected action ballot. Section 459 contemplates that protected action ballots may be held well before the industrial action authorised is actually taken. Authorised industrial action remains protected so long as it is commenced during the 30 day period after the declaration of the ballot. An extended period of a further 30 days may be provided by order of the FWC. Presuming, as I do, that the word “nature”, when used in relation to proposed industrial action, was intended to have consistent meanings in both the protected action ballot provisions and the notice requirements, the “unrealistic” constraint, which Wilcox and Cooper JJ spoke of in Davids , has far more force today than it previously had. There is also obvious disutility to the notifier in providing its target with forewarning of industrial action. The more detailed the warning, the greater is the capacity for the target to neutralise or diminish by defensive action the potential effectiveness of the proposed industrial action. Furthermore, the interests of the notifier favour simplicity: Davids at [85]. The greater the level of detail required, the greater is the danger that an infelicitous description will turn conduct intended to have been lawfully taken into a potentially-crippling claim for compensation. That would be so even where the target of the industrial action had not been misled by the misdescription and suffered no loss or damage by reason of such misdescription. That analysis demonstrates that an informative notice has utility for the target of the proposed industrial action but that there is disutility to the notifier. The extent of utility or disutility is largely a function of the extent of the particularity included in the notice. It is difficult to discern from that context where the intended balance between the interests of the notifier and the notified was intended to be struck. But, evidently, the legislature has struck a balance and, with deliberation, it has chosen the word “nature” to specify the extent of the descriptional task required. Whilst the context identifies the competing considerations, the text which has been chosen to strike the balance seems to me to be of greatest assistance in discerning the statutory purpose of s 414(6). If the interests of the person or persons notified were of priority, the word “nature” would not have been chosen. If that were the case, the “particulars” of the industrial action would have been required or the information required to be included in the notice would have been expressly identified. An example of the latter is s 234A of the Trade Union and Labour Relations (Consolidation) Act 1992 (UK) ( UK Act ). To my mind, the use of the word “nature” in the call for a description of the proposed industrial action suggests that considerable, but not exclusive, weight was given to the interests of the notifier. As I have earlier stated, when used to call for a description of a thing or an action, the word “nature” evinces an intention for generality. There are, however, gradations of generality of description capable of communicating the “nature” of industrial action. The most generalised description of the nature of proposed industrial action would state its essential character. There are probably only three essential characters of industrial action: the non–performance of work, the altered performance of work, and the refusal to allow work to be performed. I do not accept that the content requirements were pitched at that level of generality. A notice from a union stating that the nature of the proposed industrial action was the nonperformance of work would be of little more informative utility than if the notice had simply said employees will take industrial action. It must have been intended that notices be of some practical utility, otherwise no requirement for the giving of a notice would have been imposed at all. If that be the case, then drilling down to the next gradation of general descriptors of industrial action suggests that the kind or sort of non–performance, or altered performance, or refusal to allow performance of work, should be specified by the notice. In relation to the nonperformance of work, a specification of the nature of the non–performance has, to my mind, three dimensions to address. There is a functional dimension: what type of work function or duties will not be performed? There is a quantitative dimension: will the nonperformance be full, partial, ongoing or intermittent? Lastly, a description of the scope of the non–performance: what categories of employees will be involved? Each of those dimensions may be addressed expressly or inferentially. A simple expression will ordinarily suffice including because the operational circumstances in which the relevant work is ordinarily performed will provide context and thus meaning. For instance, “an indefinite strike of employees” identifies that there will be no performance of any kind of work function by any category of employee on an ongoing basis; or, “rolling stoppages by maintenance employees” identifies that there will be intermittent non–performance of maintenance functions by maintenance employees. Similar examples were given by Wilcox and Cooper JJ in Davids at [88] including, “a lockout of all employees employed in the AB fabrication plant”, “a ban on overtime”, and “a ban on the use of MN equipment”. Wilcox and Cooper JJ provided those examples in support of the observation, with which I respectfully agree, that it would be sufficient for industrial parties to describe the intended action in “ordinary industrial English” (at [88]). Additionally, at [86], their Honours emphasised that a construction which places “a premium on legalism” ought to be rejected because it is unlikely that such an approach was intended in circumstances where “notices will often, perhaps ordinarily, be prepared by non–lawyers acting without legal advice”. Relatedly, in Communications, Electrical, Electronic, Energy, Information, Postal, Plumbing and Allied Services Union v Pinnacle Career Development Pty Ltd [2010] FCA 1350 ; (2010) 190 FCR 581 at [58] , Greenwood J said that “[t]he adequacy of the notice must take account of the practical applied circumstances of the workplace ... .” As to the constructional task, observations to similar effect, favouring a “workable construction” free of prescriptions rather than a strict construction, were made in relation to s 234A of the UK Act by Elias LJ (with whom Etherton LJ and Mummery LJ agreed), in Serco Ltd t/a Serco Docklands v National Union of Rail, Maritime and Transport Workers [2011] EWCA Civ 226 ; 3 All ER 913; ICR 848 at [9] (the below quote is from the All England Law Reports, there being non-substantive differences between the various reports): There is one respect, however, in which I think that the recognition of a right to strike does have a bearing on the issues before us. Mr Charles Béar QC, counsel for the employers, submitted that since the unions were seeking to take advantage of an immunity, the legislation should be construed strictly against them. There is undoubtedly some authority to support that submission: see for example Express Newspapers Ltd v McShane [1979] 2 All ER 360 at 364, [1979] 1 WLR 390 at 395 per Lord Denning MR. But I do not think that it is a sustainable argument today. The common law’s focus on the protection of property and contractual rights is necessarily antithetical to any form of industrial action since the purpose of the action is to interfere with the employer’s rights. The statutory immunities are simply the form which the law in this country takes to carve out the ability for unions to take lawful strike action. It is for Parliament to determine how the conflicting interests of employers and unions should be reconciled in the field of industrial relations. But if one starts from the premise that the legislation should be strictly construed against those seeking the benefit of the immunities, the effect is the same as it would be if there were a presumption that Parliament intends that the interests of the employers should hold sway unless the legislation clearly dictates otherwise. I do not think this is now a legitimate approach, if it ever was. In my judgment the legislation should simply be construed in the normal way, without presumptions one way or the other. Indeed, as far as the 1992 Act is concerned, the starting point it that it should be given a ‘likely and workable construction’, as Lord Bingham put it in P v National Association of Schoolmasters/Union of Women Teachers [2003] UKHL 8 at [7] , [2003] UKHL 8 ; [2003] 1 All ER 993 at [7] , [2003] UKHL 8 ; [2003] 2 AC 663. The primary judge’s reasoning The primary judge commenced his consideration of this issue by identifying, as an important question, whether the AWU’s notice covered the refusal by operators to carry out air freeing and leak testing and, from 7 March 2015, a refusal to manipulate bleeder valves. It is clear from the reasons (including at [120] and [126]) that the industrial action of refusing to perform air freeing, leak testing and the manipulation of bleeder valves was action confined to the performance of those tasks when required to be done as part of the reinstatement of isolated equipment. I will explain later what I mean by reinstatement of isolated equipment. At [69] of the reasons, the primary judge identified Esso’s position as being that, when “deisolation of equipment” was used in the AWU’s notice, it conveyed that that term was used as defined by what (at [11]) the primary judge characterised as a “procedures manual” called the “Work Management System” manual ( WMS Manual ). That is, it meant “[u]nlocking and moving isolation valves, reconnecting systems, removal of blinds, and unlocking electrical switches to their normal operating state ” (italicisation as in original). The primary judge then identified the AWU’s submission that the term “deisolation of equipment” had “an accepted, and well-understood, meaning at Longford” and that Esso’s management would have understood the AWU’s notice by reference to that meaning. Although the meaning contended for by the AWU was not identified until later (at [90]), the primary judge described the denotation contended for by the AWU as “referring to every aspect of operators’ work that would be performed during the period that the ICC was headed ‘[De-isolation] in Progress’”. “ICC” is an acronym for “Isolation Control Certificate” ( ICC ). At [70], the primary judge said this (emphasis added): In considering [the AWU’s] submission, it is necessary to commence with the purely factual question whether there was such an accepted and well-understood meaning, both on the part of the operators employed by [Esso] and on the part of the managers whose function it was to consider what was conveyed by the notice. To that end, the primary judge then considered the evidence called by the AWU from Mr Kain Jackson, an operations technician level 2 ( Jackson ) ([71]-[72]); Mr Robert Steed, an operations technician level 2 ( Steed ) ([73]-[74]); Mr Anthony Malady , an operations technician level 2 ( Malady ) ([75]); Mr Mark Vos, an operations technician level 2 ( Vos ) ([76]); and Mr David Ryan, an operations technician level 1 ( Ryan ) ([77]). Much of that evidence was to the effect that when operators are asked to “de-isolate” a particular piece of equipment they carry out all the tasks necessary to reinstate the isolated equipment into its operative state, including equipment testing, leak testing and air freeing. Some of those witnesses gave evidence to the effect that the expression “de-isolation of equipment” meant the opposite of “isolation” and that “de-isolation of equipment” had the same meaning as “reinstating equipment”. At [78], his Honour made these observations (emphasis added): That was the evidence led by the respondent in support of its contention that the term “de-isolation of equipment” had an accepted and well-understood meaning amongst operators, that is to say, a meaning that would comprehend, without ambiguity , the whole process of returning the equipment to service, including air freeing and leak testing. The evidence was, however, ultimately equivocal, and not really helpful in the present context. For one thing, the acceptance by the witnesses, specifically Mr Steed, of the central importance of the WMS manual – and, one might add, the reference to that manual in the evidence-in-chief of Mr Jackson – makes it inevitable that the court would find, as I do, that the term “de-isolations” has an official, or authorised, meaning. If it is the respondent’s evidentiary case that the term also has an everyday meaning which differs from that in the manual, that would produce the result that there were two meanings, something which the respondent’s contention does not, and could not, contemplate . The first and third emphasised passages indicate that the primary judge understood the AWU’s case (or at least its primary case) to be that “de-isolation” would necessarily be understood at Longford to have the broad meaning for which the AWU contended. Indeed the word “necessarily” is used at [79] to make that point. The second emphasised passage is, it seems to me, an important part of the primary judge’s ultimate conclusion, a matter to which I shall return. The primary judge thereafter turned to the evidence called by Esso. The judge said at [80] that, as against the AWU’s evidence to which he had just referred, there was the evidence of Mr Mark Lloyd, Operations Supervisor – Longford Plants ( Lloyd ). The primary judge found that Lloyd understood the term “de-isolation” to mean the narrow denotation contended for by Esso. The primary judge also accepted, by reference to Lloyd’s evidence, that “daily notes” which “form the basis of instructions given by supervisors to the operators at toolbox meetings at the commencement of each shift”, contained instructions that used expressions consistent with the narrow sense of “de-isolation” as contended for by Esso. At [82], the primary judge then said: Mr Lloyd’s evidence as to the language used in the daily notes, and as to the use of those notes in the allocation of tasks each shift at the workplace, was both more specific and more reliable (ie in the sense that it was based on documentary materials) than the high-level references by the respondent’s witnesses to what they do in the process of de-isolation. Evidence of the latter kind did not have the focus sufficient to establish the terminological discriminations that the respondent’s case involved. By contrast, the daily notes, and the evidence given in connection with them, dealt with actual situations which had arisen in relevant contexts, and did so at a very low level, where the operators themselves were intimately involved. Based largely on those notes, I would find that, as a rule, operators tend to be instructed to purge, to pressure test, or to de-isolate (or “remove the ICC”) a piece of equipment, as Mr Jones was on 7 March 2015. There was nothing in the daily notes that would provide support for the respondent’s case that, in the normal course of their duties, operators tended to be instructed merely to “de-isolate” some piece of equipment which required all the relevant stages to be carried out. Finding at [83] that Lloyd’s evidence about the daily notes was not controverted (or effectively controverted), the primary judge at [84] concluded that: All things considered, I am not persuaded that, in a normal operational setting at Longford, the term “de-isolation of equipment” had an accepted, and well-understood, meaning as proposed by the respondent. To the contrary, at least in a practical context involving the identification of work and tasks, the term related to de-isolation as such. When air freeing and pressure testing were required to be carried out, they were referred to in terms – either those terms or, in the case of the former, “purging”, and, in the case of the latter “leak testing”. As indicated above at [266], the primary judge was, to this point, dealing with the proposition that “de-isolation” necessarily had the broad meaning contended for by the AWU. At [84], the primary judge rejected that proposition. In that respect, I think the primary judge was saying that “de-isolation” was not so well understood as a form of instruction or other use in the “normal operational setting” at Longford that, when used, it would be readily and confidently understood as intending the broad sense contended for by the AWU. However, that conclusion was not regarded by the primary judge as determinative of how Esso would have understood the term when used in the notice. If it had been, the primary judge’s analysis would have ended at [84]. It did not. As the commencing words of [85] indicate, there was a further submission to be dealt with. From that point, it seems to me that the primary judge dealt with an alternative basis for determining what Esso should have understood from the AWU’s notice. That basis raised the contest between the parties about the effect of the documents upon Esso’s reasonable understanding of the notice. Esso’s position was that it would have reasonably gained an understanding of what “de-isolation” meant from the WMS Manual. The AWU, in contrast, relied upon the ICCs. Ultimately, as is apparent from [89], the primary judge accepted Esso’s position. At [85], the primary judge considered the AWU’s reliance on the nature and use of the ICCs and said (emphasis added): It was also submitted on behalf of the respondent that its case on the denotation of “de-isolation” derived support from the terminology used in the ICC. It was said that the ICC was to be regarded as the controlling document, and that the respondent’s notice of 3 February 2015 encompassed every function and activity that was required to be carried out subsequent to the ICC being headed “[De-isolation] in Progress”. Further, it was said that I should regard the terminology of the ICC as particularly persuasive, since the ICC was the electronic system which the applicant required its operators to use, and the heading referred to was, during a period when equipment was being returned to service, conspicuous to anyone using that system. At the purely factual level, this submission is well-founded , but, before I place the respondent’s reliance on the ICC against the applicant’s reliance on the terms and definitions used in the WMS manual, I should make some general, but nonetheless important, observations about the context in which the present issue arises for resolution. On my reading of the primary judge’s reasons (including the last sentence of [85]), at [86], and then at [89], the primary judge explained why what he regarded as wellfounded at the purely factual level was not ultimately persuasive. In those paragraphs the primary judge said this: [86] The task for the court is not the conventional one of the construction of a document, with a view to understanding what the author intended. The document with which we are concerned here was a notice: its purpose was to convey information. Thus the question is not what the author intended, but what the addressee would reasonably have understood from the terms used in the notice. Two things follow from this. First, if the notice might reasonably carry more than one denotation, I see no reason to err on the side of the giver of the notice, thereby permitting him or her to take advantage of his or her own ambiguity. The FW Act leaves it entirely to the giver of the notice to identify the “nature” of the action intended to be taken, and it should not be open to him or her to complain if the terms chosen leave scope for the addressee to see things differently from what the giver, subjectively, might have had in mind. Secondly, it would not be sufficient for the respondent to establish what its own members understood by the presently contentious expression. Even if they knew what they meant by “de-isolation of equipment”, the question is what the applicant’s management would reasonably have understood by that expression. ... [89] Returning to the respondent’s reliance on the ICC, of the two presently contentious senses in which the term “de-isolation of equipment” might have been understood by the applicant as recipient of the notice of 3 February 2015, that referred in the WMS manual is, in my view, the more natural one. It refers to the de-isolation of equipment as such, and is, therefore, more closely aligned with the specific tasks which would, in the normal course, be carried out by operators, and which the applicant would understand to be the subject of the ban. By contrast, the ICC is concerned not with tasks or functions but with recording the positions of points at particular stages in the process of taking some equipment out of, and of returning it to, service. Insofar as it deals with the points that must be changed from one state to another, it records what has been done rather than, for example, instructing what should be done. Nowhere is this clearer than in the evidence of Mr Jackson. In short, of the two documents, the WMS manual is the more closely related to the work as such, and is the more directly concerned with marking out de-isolation as an activity of work. The primary judge then concluded at [90]: For the above reasons, I would reject the proposition that the expression “de-isolation of equipment” in the respondent’s notice of 3 February 2015 would reasonably have been understood by the applicant as referring to every aspect of operators’ work that would be performed during the period that the ICC was headed “[De-isolation] in Progress”. To the contrary, in my view it would have been so understood as referring to the specific function of de-isolation as such. It would not have been so understood as encompassing equipment testing, air freeing or leak testing. Nor did it refer to the manipulation of valves associated with those activities, notwithstanding that such manipulations were mentioned on the ICC. It follows that the respondent’s ban on work of that kind was not protected industrial action within the meaning of the FW Act. To my mind, having rejected that “de-isolation” in the broad sense had a “well-understood” meaning, the primary judge was ultimately not persuaded that the broad denotation of “deisolation of equipment” would have reasonably been understood by Esso as the intended meaning. The primary judge regarded the “more natural” meaning in the WMS Manual as being the meaning that Esso would reasonably have understood by the expression “deisolation of equipment”. The reason that the Manual’s denotation was regarded as “more natural” was explained by the primary judge at [89]. In short, and relative to the use of “deisolation” in the ICCs, the denotation used in the WMS Manual was “more closely related to the work as such”. That was because it was “more closely aligned with the specific tasks ... carried out by operators” and “more directly concerned with marking out de-isolation as an activity of work”. The AWU’s contentions Ground 1 of the AWU’s Notice of Appeal stated: His Honour erred in interpreting the notice of proposed protected industrial action dated 3 February 2015 given by the Appellant pursuant to s 414 of the Fair Work Act 2009 (Cth) (FW Act), in that he found that the phrase “deisolation of equipment” would not reasonably have been understood by the Respondent as encompassing equipment testing, air freeing and leak testing or the manipulation of valves associated with those activities. Ground 1 complained that the primary judge erred in interpreting the notice of proposed protected industrial action given by the AWU on 3 February 2015 ( AWU’s notice ). In particular, the meaning which the primary judge gave to the expression “de-isolation of equipment” was said to be erroneous. Particulars of the asserted errors of the primary judge were addressed in the AWU’s submissions. In summary, the AWU complained that in construing the expression “de-isolation of equipment”, the primary judge failed to properly take into account and mischaracterised an important surrounding circumstance. The alleged consequences of the error raised by ground 1 were identified in the AWU’s grounds 2 and 3 as follows: By reason of the error identified in ground 1, his Honour erred in finding that the organising by the Appellant of bans on air freeing and leak testing by the persons employed by the Respondent as production employees at its Longford site who were members of the Appellant ( Longford members ) between 6.00 pm on 6 March 2015 and 9.30 am on 7 March 2015 contravened section 421 of the FW Act by contravening clause 4.1 of the order made by the Fair Work Commission on 6 March 2015 ( the third s. 418 order ). By reason of the error identified in ground 1, his Honour erred in finding that the organising by the Appellant of bans on the manipulation of bleeder valves by the Longford members between 9.30 am on 7 March 2015 and the making of the Court’s interim order on 17 March 2015 contravened section 421 of the FW Act by contravening clause 4.1 of the third s. 418 order. The AWU contended that the proper question, in construing the meaning of the AWU’s notice, was what a reasonable employer in Esso’s position would have understood from the AWU’s notice and, in particular, the expression “deisolation of equipment”. The AWU emphasised that the answer required a consideration not only of the text but also of the surrounding circumstances known to Esso, as well as the purpose and object of the transaction. The AWU relied by analogy on the approach taken by Gleeson CJ, Gummow, Hayne, Callinan and Heydon JJ in Pacific Carriers Ltd v BNP Paribas [2004] HCA 35 ; (2004) 218 CLR 451 , in relation to the construction of a letter of indemnity. At [22], their Honours said this (citations omitted): ... The construction of the letters of indemnity is to be determined by what a reasonable person in the position of Pacific would have understood them to mean. That requires consideration, not only of the text of the documents, but also the surrounding circumstances known to Pacific and BNP, and the purpose and object of the transaction. The object of the transaction, so the AWU contended, was the giving of notice by employees of the withdrawal of their labour. In that context, the AWU contended that the most relevant surrounding circumstance to be taken into account in giving objective meaning to the description of the work to be withdrawn, was the description of that work used in the workplace whilst that work was performed. On the AWU’s case, the evidence established that an ICC was ordinarily generated when isolation/de-isolation work was performed on a particular piece of equipment, and that the ICC used the descriptor “Deisolation in Progress” to describe the performance of all tasks (including equipment testing, air freeing, leak testing, and the manipulation of valves associated with those activities) necessary to reinstate equipment from an isolated state into an operative state. The ICC’s meaning of “deisolation” was the description with which the employees worked. It was the description which referenced the work for the employees. As a surrounding circumstance, it was the most relevant use of the expression “de-isolation of equipment”. That was so because the nature of the transaction (the notification of the withdrawal of work) was of employees communicating their intent not to perform certain parts of their own work. The AWU contended that the primary judge mischaracterised the function of the ICC and thus the closeness of its relationship to the performance of the work. It characterised the primary judge’s approach as “semantic”. It submitted that the ICC had a far closer relationship to the performance of work than did the WMS Manual, because the ICC identified the work to be done. In that respect, the AWU challenged the primary judge’s findings at [89], including the finding that “the ICC is not concerned with tasks or functions” and the finding that the WMS Manual is “the more closely related to the work as such”. The AWU contended that, by wrongly characterising the nature of the ICCs and by failing to recognise their significance and connection to the performance of the work, the primary judge failed to give proper weight to an important surrounding circumstance and thereby misidentified the meaning that a reasonable person in Esso’s position would have given to the expression “de-isolation of equipment”. In short, the AWU contended that the ICCs were more closely connected to the work of operators and the primary judge should have so found. The challenge made by the AWU to the findings at [89] was in part based upon the contention that the findings were directly at odds with evidence given by Esso’s own witnesses and in particular Mr James Kristeff, Maintenance Superintendent at Longford ( Kristeff ) and Mr Ross Dunbar, Operations Superintendent – Gas Asset ( Dunbar ). It was also said that the findings were contrary to the unchallenged evidence given by the AWU’s witnesses. The AWU contended that, as the evidence of Kristeff and Dunbar stood directly against the findings made by the primary judge, as a matter of principle the primary judge was bound to explain why that evidence was put aside. The failure to do so was said to constitute an error in the process of fact finding which undermined the primary judge’s conclusions. The AWU also challenged the reliance placed by the primary judge on the “daily notes”. Finally, the AWU contended that to the extent that there was ambiguity as to the meaning of “de-isolation of equipment”, it was ambiguity created by Esso through its own documents and that the contra proferentem rule should have been applied by the primary judge against Esso, and not against the AWU. Esso’s contentions In broad terms, Esso defended the challenges made to the primary judge’s judgment on the basis of the analysis made by the primary judge. Esso accepted that, in construing the AWU’s notice, an objective test based on the approach taken in Pacific Carriers was correct. Esso contended that the primary judge had posed the right test and come to a conclusion open on the evidence. Esso rejected the AWU’s challenge to the findings made by the primary judge as to the function and significance of the ICC. Esso accepted that the ICC was a surrounding circumstance known to Esso but contended that, as a surrounding circumstance, the use of ICCs was taken into account by the primary judge as part of the mix of relevant circumstances. Esso further contended that the relevant findings of the primary judge were either not challenged or, if challenged, were open on the evidence. Those findings were that the WMS Manual had a definition of “de-isolations” consistent with Esso’s contention as to the meaning of that term. The central importance of the WMS Manual meant that the term “deisolation” had, as the primary judge had correctly found, an “official, or authorised, meaning”. Furthermore, Esso relied on the findings made by the primary judge that instructions given to operators in the “daily notes” used the term “de-isolation” in its narrow sense and consistently with the meaning contended for by Esso. Those findings, so Esso contended, were fundamental and important in the assessment of what a reasonable person in Esso’s position would have taken the notice to have meant. Esso contested the AWU’s characterisation of the evidence given by Kristeff. It contended that, fairly read, the evidence of Kristeff was not fundamentally opposed to other evidence accepted by the primary judge and was not such as to require different findings to have been made by the primary judge. In that respect, Esso emphasised the importance of the evidence given by Lloyd. Discussion The object and purpose of the AWU’s notice was to communicate the intention of Esso’s employees to not perform some of their work. Put another way, its purpose was to tell Esso what part of the work, usually performed by certain employees, those employees intended not to perform. Given the terms of s 414(6), it was only the nature of the work that it was intended would not be performed that Esso needed to be told about. As a recipient of the notice, a reasonable employer in Esso’s position must be taken to have had that understanding of the purpose and object of the notice. The relevant question, then, was what a reasonable person in Esso’s position with knowledge of the surrounding circumstances would have understood to be the intent of its employees to not perform work. Before embarking upon any further analysis, it is important to emphasise that, in this kind of exercise, context will be the touchstone. Context will almost always identify which of two possible meanings of a word or term was likely to have been intended. Context may be provided by the surrounding words in a notice, and that will usually be the first place to look. Context may also be provided by previous communications. That will also be an appropriate place to look. Additionally, context may be provided by the surrounding circumstances. Where the same word has two possible denotations, the occasion in which it is used will often reveal which denotation was intended. Thus, a colloquial or familiar name will commonly be used to identify a person or a thing when the context is informal. An official, authorised or technical name would more likely be used where the context is official or technical. To illustrate, an engineer giving technical instructions to safely shutdown machinery or plant may be expected to use the official or technical terminology found in the manual for that machinery or plant. But an operator of the machinery, when speaking to a supervisor about his or her work upon the machinery, is more likely to use colloquial terminology if that kind of terminology is in everyday use. Meaning may also be given by the particular temporal context in which a word with multiple denotations is used, on a particular occasion. If the same word describes both a process involving many tasks and a particular task within that process, the circumstances in which the word is used will often reveal its intended meaning. Thus, it is unlikely that the word will be used to describe the process, rather than the specific task, where it is used as an instruction to perform work and much of the process has already been performed and all that is left to do is the specific task. In this case, the primary judge at [84] rejected the contention that “de-isolation” necessarily had only one meaning in the sense earlier explained. That finding is not challenged. The evidence was uncontentious that, when used in relation to equipment, “de-isolation” was sometimes used to describe a process involving a series of tasks including tasks known as “air freeing” (also called “purging”) and “leak testing” (also called “pressure testing”). When used in that way, the evidence included that the term was used synonymously with the word “reinstatement” or the terms “reinstatement of equipment” or “returning equipment to service”. For ease of reference, I will refer to that denotation as the broad meaning of “deisolation of equipment”. When used in its narrow sense, “de-isolation of equipment” referred to a particular task required to be done as part of the reinstatement of equipment. That task involved the placement of valves into their normal operating states. That task was also known as “removing energy isolations”. I well appreciate that, in a workplace setting, the meaning ordinarily attributed to a word may be the result of exposure to a document such as a work manual which uses that word with a particular denotation. The existence of a work manual or other work document, and the meaning given to terminology by such a document, may properly be regarded as a surrounding circumstance and there can be no doubt that the primary judge was correct to do so in this case. However, there is a question in my mind as to whether, when considering the surrounding circumstances, it was appropriate for the primary judge to have focussed as much as he did upon the choice to be made between the WMS Manual and the ICCs. It may be thought that the approach adopted of considering, first, the question of whether “deisolation” had a “well-understood” use and, second, the documents and their use, was inappropriately compartmentalised given that the documents also shed light upon the extent of the use of the word “de-isolation”, and given that there were other documents such as the “daily notes” (and Lloyd’s evidence about them), which were also relevant to the ultimate question of Esso’s understanding. But that approach was not the subject of any criticism. That may be because the submissions of the parties did not properly appreciate the compartmentalised approach taken by the primary judge. In any event, on the appeal, the main battle ground was [89] of the primary judge’s reasons and the finding that a reasonable person in Esso’s position would have understood the meaning of “de-isolation” from its use in the WMS Manual rather than its use in the ICCs. The relevant surrounding circumstances were thus substantially confined by the approach taken by the parties at both the trial and on the appeal. That approach essentially narrowed the contest to the dispositive question of which of the two competing documents should have been of greater weight or significance to the understanding that Esso should have reasonably gained from the AWU’s notice. Turning then to the detail, it is not necessary that I comprehensively set out the terms of the WMS Manual. Its provenance was described by the primary judge and many sections of it are set out at [11][20] of the primary judge’s reasons. The primary judge identified the overall context at [8]. Esso’s operations at Longford involve the processing of highly toxic, volatile, pressurised and flammable products. To deal with the risk of fire or explosion, Esso has in place an “Operations Integrity Management System”, a section of which is concerned with “work management” including a process of “energy isolation”. Energy isolation of systems and equipment is required at Longford so that maintenance or other work done pursuant to work permits issued under the work management process may be safely performed whilst the equipment is removed from service. Once isolated, and after maintenance or other work has been completed, the equipment must be reinstated into its safe operating condition. In that context, the provenance of the WMS Manual was described by the primary judge as follows (at [11]): The detailed processes in relation to work permits and isolation procedures are prescribed in a separate procedures manual, called the Work Management System (“WMS”) manual. All operations personnel employed by the applicant have been trained in the provisions of this manual, and refresher training is also undertaken. The WMS manual is available at the relevant workplaces, in both electronic and paper forms. The definition of “de-isolations” contained in the Manual said this (emphasis in original): De-isolations Unlocking and moving isolation valves, reconnecting systems, removal of blinds, and unlocking electrical switches to their normal operating state. The definition section contained separate definitions for “air freeing”, “leak testing”, “equipment testing” and some of the other tasks required to reinstate isolated equipment. In that respect, Esso emphasised that the WMS Manual distinguished between “deisolation”, “leak testing”, and “air freeing”, and identified them as different and separate work tasks. That they were each substantial tasks to which the WMS Manual devoted significant attention was also emphasised by Esso and is uncontroversial. However, the WMS Manual also uses “de-isolation” in its broad sense. To demonstrate that, it is necessary to refer to subsections 4.5 and 4.6 of the Manual. Both subsections sit in a section headed “Equipment Isolation and Reinstatement (Lock-Out/Tag-Out)”. The heading of subsection 4.5 is “Isolating Systems and Equipment”. The heading to subsection 4.6 is “Reinstating Facility Systems and Equipment”. Each subsection sets out in some detail the multiple activities that must be performed in each process. For each activity there is a subsection heading followed by some detail of the procedures required in relation to that activity. At the commencement of subsection 4.5, the “isolation activities” are described to include: Energy isolations (process, mechanical, and electrical) Draining and depressurizing Hydrocarbon freeing Breaking containment Blinding Cleaning A similar introduction, listing the activities for “reinstating equipment and facilities”, is given at the commencement of subsection 4.6 as follows: Recommissioning Note: Recommissioning is an activity that takes place throughout the reinstatement. Mechanical completion checks must be made before and after equipment testing, air freeing, leak testing, and de-isolating. Equipment testing Air freeing Leak testing Removing energy isolations (mechanical, electrical, instrument) Removing temporary defeats Acceptance testing It is clear that “isolation” is a process involving multiple activities. There is no suggestion that the term “isolation” is a single stand-alone activity. Similarly, “reinstatement” is a process involving multiple activities. These processes have an inverse relationship. The purpose of the isolation process is to remove equipment from service and the purpose of the reinstatement process is to return equipment to service. The temptation to describe the process of reinstatement as “de-isolation” is strong and, as I will explain, was not resisted by the author of the WMS Manual. Subsection 4.5 contains a detailed section of some 15 pages headed “Energy Isolation and Isolation Control Certificate”. The introduction to that section says this: Energy isolations are managed by an Isolation Control Certificate (ICC). As the primary judge identified at [19], an ICC is an electronic artefact held on Esso’s computer systems. The evidence demonstrated that ICCs were also printed and carried by operators whilst performing their work. The WMS Manual describes an ICC as follows (italics in original): The Isolation Control Certificate (ICC) must be used to document the isolation and approval of equipment or systems and will be used to track the status of all isolations. The isolation plan must be Verified by the Area Authority (AA) or designate before it can be applied. Isolation points must be signed off (electronically or on the paper certificate for remote fields) by the Isolating Authority (IA) to confirm they have been put in place. Associated work permit(s) can only be issued once the AA or designate confirms that all isolations are In Place . The ICC must remain in force until all work permits associated with the ICC have been completed. The ICC must be cross-referenced to all relevant work permits and is the principal control once isolations are in place. The ICC performs the following functions: Lists the location of the isolation points and their normal status. It is a record of any fittings (such as analyzer points, sample points and plugs) that have been removed or moved from their normal state to ensure that they are reinstated. It is a record of all drain, vent, and bleed points for valve integrity tests. (Operation of these valves does not need to be recorded as an amendment on the ICC but would be checked when returning equipment to service.) Authorizes any temporary de-isolation for equipment testing (Sanction to Test). Authorizes and records each de-isolated point. As necessary, includes attachments of marked-up P&IDs, electrical diagrams, and isolation plans/procedures. The WMS Manual contains a flow chart which the primary judge described (at [20]) as demonstrating schematically what the Manual described as the “Isolation Certificate Process”. The flow chart is as follows: In relation to the ICC flowchart ( flowchart ) the primary judge said this at [21]: It will be noted that this flow chart, as it may be called, has two vertically-arranged streams. That on the left refers to the sequential stages of the functions that are necessary in taking the equipment from, and returning it to, service. That on the right refers to the corresponding headings on the ICC, the result of entries made by supervisors and operators (as the situation requires) to indicate the stage that has been reached. It will be noted that neither stream refers to the tasks or functions that must be performed at any stage: it is a given that those involved in this work will know what is required. The right hand stream in the flowchart contains two descriptors of particular importance; “Isolation in Progress” and “De-isolation in Progress”. I will refer to the evidence shortly, but it is uncontroversial that in the period between when an operator selects “Isolation in Progress” on the ICC as the status of the process, and until “In Place” is selected to show the state of the process, so many of the isolation activities described at [296] above as need to be carried out on the particular piece of equipment or system, are carried out. Similarly, but dealing with the inverse process, whilst an ICC designates the status of the process as “Deisolation in Progress” and until “Completed” is selected, so many of the activities of the reinstatement process described at [297] above as are required to be carried out, are carried out by the operators. In the flowchart in the WMS Manual (and in each ICC created by reference to that flowchart), the designation “De-isolation in Progress”, is used as a synonym for the process elsewhere referred to in the WMS Manual as “reinstatement” or “reinstating equipment”. The word “de-isolation” is undeniably there used in its broad sense to mean the activities necessary to reinstate isolated equipment into service. That use (which constituted part of the text of the WMS Manual) was inconsistent with the definition of “de-isolation” given in the Manual. Whilst not determinative of what the WMS Manual means by “De-isolation in Progress”, it is interesting to observe what Ryan said, as recounted by the primary judge at [77]: David Ryan is employed by the applicant at Longford as an operations technician, level 1. He has been employed there since 2010. He regularly performs de-isolations. He uses the expression “de-isolation of equipment” to refer to “a range of tasks that must be performed to bring a piece of equipment to a stand-by state ready to be returned to service after it has had maintenance or repairs performed on it.” Depending on the equipment, this can involve air freeing, leak testing and equipment testing. He uses the term “de-isolation” to mean the opposite of “isolation”. Noting that the WMS manual uses the expression “Reinstating facility systems and equipment”, Mr Ryan said that this was not an expression that he uses, or that he hears used regularly, at Longford. He considers that reinstating equipment and de-isolating equipment are “one and the same thing”. In the course of his work, other employees and supervisors, when referring to the reinstatement of equipment, also use the term “de-isolation”. Kristeff accepted that “De-isolation in Progress”, as used in the ICC, was synonymous with “return to service”. The inconsistency between that use and the WMS Manual’s definition was also acknowledged in evidence, to which I will shortly refer. There are some further matters to note in relation to the terminology used in the WMS Manual. First , the definition section of the WMS Manual does not contain a definition for “reinstatement” or “reinstating equipment”, although a definition does appear for “Energy isolation”. Second , whilst the word “de-isolation” is used in the WMS in its broad sense as just demonstrated, it is also used in its narrow sense but, interestingly, when it is so used it is used as a secondary descriptor. The primary descriptor used for the tasks identified in the WMS Manual’s definition of “de-isolations” is “removing energy isolations”. That is the term used in the introduction to subsection 4.6 to identify the reinstatement activities. It is also the term used in the heading dealing with that activity (at page 4-106). The terms “deisolate”, or “deisolation”, in their narrow senses, are used within that section. Further, the term “removing energy isolation” is not referred to in the definition of “de-isolations”. The point is that the WMS Manual uses language inconsistently. The primary judge (at [22]–[26]) explained, by reference to the flow chart, how the ICC as an electronic artefact is dealt with through the processes of isolation and then reinstatement. It is sufficient to set out what was said at [25]–[26] in relation to reinstatement of equipment: [25] When the work is completed, the equipment must be returned to service. Once again, the applicant’s operational personnel become involved. The first thing to be done is to alter the heading on the ICC to “[De-isolation] in Progress”. Again, this would be achieved by the selection of that designation from the drop-down menu. Thereafter, anyone who consulted the ICC on the computer would see that a de-isolation was “in progress”. The various physical tasks involved in removing the isolations would be carried out: they are not mentioned on the functional side of the flow chart, but are implied under “Isolation removed”. Correspondingly, on the ICC side of the flow chart, the operator is required to “Update Point Positions” to record the changes made to the state of the various points. This will involve returning the points to their “[De-isolated] State”, be it open or closed. [26] Finally, when all the necessary tasks have been completed, the de-isolation integrity of the equipment must be verified in writing by the area operator and, when this has been done, the heading “Completed” is selected from the drop-down menu on the ICC software. When the ICC itself is closed, it is “Archived”. From the archived records, the state of the ICC at any stage during the previous shutdown of a piece of equipment can be recovered as a report, and printed out. A number of such iterations of ICCs were tendered in evidence, from which it was possible to note, for example, the exact state of each relevant point at a particular stage during “Isolation in Progress” and “[De-isolation] in Progress”. There was uncontroversial evidence before the primary judge as to how an ICC is used whilst operators were carrying out the work required to reinstate isolated equipment. A key witness for the AWU was Jackson, an operations technician who was trained to perform de-isolations of equipment and regularly did so. Jackson also provided training and advice to others including supervisors about “the WMS” and Esso’s electronic permit to work system, as the primary judge stated at [27]. The primary judge recounted that evidence at [28] as follows: Mr Jackson gave evidence about the concrete steps he takes in carrying out the work involved in air freeing and leak testing. He commences by putting the ICC to “De-isolation in Progress”. He prints a hard copy of the ICC which he takes into the “field”, that is, to the physical location of the vessel upon which he proposes to undertake certain functions. He will also have with him a “job safety analysis”, to make sure that he has covered all the hazards and risks associated with those functions. Where the vessel requires air freeing, that will be done first. Depending on the circumstances, there may be a pressure safety valve to be put into service at this point. The air freeing itself involves connecting a nitrogen supply to the vessel and opening a bleeder that will direct the contents of the vessel to a safe location. Once he is satisfied that the vessel has been purged in this way, Mr Jackson will close the bleeder and use nitrogen to increase the pressure to anything up to 400 kPa, to test for leaks. Once he is satisfied that there is no obvious leak, he will introduce another hydrocarbon, with a view to raising the pressure to 80% of the design rating of the vessel, or the “pressure safety valve” rating of that vessel. With the vessel under this pressure, Mr Jackson will check for leaks at various points. During each of these stages, each time he manipulates a point, he will mark, or tick, the hard copy of the ICC that he is carrying with him. He will then return to the computer and transfer on to the ICC itself the changed status of the points concerned. Evidence to similar effect was given by Steed, another operations technician. At [73] the primary judge recounted that evidence: ... In his oral evidence, Mr Steed explained what he did when “called on to perform de-isolations of equipment” (his counsel’s words). He would begin by going to the computer and, from the dropdown box, selecting “perform de-isolation”. The computer would notify him “of all permits and any outstanding jobs on that role”. Assuming that there were none, and after Mr Steed had “got the de-isolation in progress under way”, he would print a hard copy of “all of the points in dot point form, ... [showing him] the state that they are in, in an isolated state”. With this paper in his hand, he would then “walk the entire isolation, confirming all the points, confirming the status of all the points”. Then, generally, but depending on the type of equipment, he would “perform a purging of the equipment”, that is “the [removal] of hydrocarbons and oxygen” using nitrogen or process gas (or, as Mr Steed clarified, air freeing). Pressure testing would follow next. Other operators, and in particular Malady and Vos, gave evidence to the same effect, as the primary judge recounted at [75] and [76]. None of that evidence was relevantly controversial. The primary judge accepted that, at “the purely factual level”, the denotation of “deisolation” used in association with the ICCs supported the AWU’s case that the term, when used in the AWU’s notice, should have been understood as having its broad meaning (at [85]). The primary judge then said that before “I place the respondent’s reliance on the ICC against the applicant’s reliance on the terms and definitions used in the WMS manual, I should make some general, but nonetheless important, observations ...” (at [85]). The primary judge moved to identify relevant principles one of which is recorded at the end of [86] as follows: Even if [the AWU’s members] knew what they meant by “de-isolation of equipment”, the question is what the applicant’s management would reasonably have understood by that expression. At [89] the primary judge returned to the contest about the documents. By this point, the primary judge had moved from the “purely factual”, to the objective question which he had posed as to “what [Esso’s] management would reasonably have understood” by the expression “deisolation of equipment” in the AWU’s notice. That was the basis upon which his Honour dealt with the contest about the impact of the documents. It seems evident from the analysis at [89] that the primary judge must have had in mind that a reasonable employer in Esso’s position would have recognised the existence of two possible meanings of “de-isolation”, and chosen the meaning “in the WMS manual” over any contrary meaning indicated by the ICCs. The primary judge reasoned that it was reasonable for Esso to have done that because the WMS Manual’s use of “de-isolation” was “the more natural one”. That was because it provided a meaning “more closely related to the work as such”. That, in turn, was a conclusion which seems to have been based upon the finding that the definition in the WMS Manual was “more closely aligned with the specific tasks ... carried out by operators” and “more directly concerned with marking out deisolation as an activity of work”. There are two difficulties with that approach. The first is the finding made at [89] that the ICC “is concerned not with tasks or functions” and thus is not concerned with marking out the work. The second is related and more fundamental. Turning first to the finding that the ICC “is concerned not with tasks or functions but with recording the positions of points at particular stages in the process of taking some equipment out of, and of returning it to, service”, in my respectful view, that finding is erroneous. The uncontroversial evidence before the primary judge was that an ICC lists the valves which are to be worked on in reinstating equipment that has been isolated. As is apparent from the examples of ICCs in evidence, an ICC identified the “Action” status of the isolation or reinstatement process according to the labels indicated by the right hand column in the flow chart set out above at [302]. The evidence was also uncontroversial that, on the commencement of the reinstatement process and until that process was completed, the ICC indicated the status of the process as “De-isolation in Progress”. Each of the valves which was required to be worked on was listed in a “Point Summary” on an ICC. In each case, the valve’s “isolated state”, “current state” and “deisolated state” was shown as “open”, or “closed”. Whilst the “Deisolation in Progress” status was in place, a valve the “current state” of which was shown as “closed” but the “deisolated state” of which was shown as “open”, needed to be manipulated into its “deisolated state”. Likewise, where the “current state” of a valve was shown as “open” and its “deisolated state” was shown as “closed”, a manipulation was required. The uncontroversial evidence was that, before that could be done, the tasks of air freeing and pressure testing (which of themselves may involve the manipulation of the valves) needed to be carried out, where necessary. It is clear that an ICC did not expressly list the tasks required of the operator. An ICC did not say, for instance, that valve X must be moved from its current open position into its deisolated closed position. However, the ICCs were, as the evidence showed, documents that experienced operators familiar with the tasks required of them used to reference their work. By identifying the valves to be worked on and the result required, an ICC identified the work to be done and, inferentially, the tasks necessary to perform that work. As a descriptor of work, an ICC described the artefacts (the valves) upon which work was to be performed. To my mind, the ICCs were concerned with work tasks and functions. An ICC may well be characterised as a “list” or a “record”. But, to dismiss the relevance of the ICCs on that basis was to miss the point that, as a practical matter, the ICCs referenced the work required of operators and, whilst the tasks involved in reinstating particular equipment were in progress, the ICCs labelled that work as “deisolation”. It was wrong, in my respectful view, for the primary judge to have put aside the evidence relied upon by the AWU as to the nature of and the use of ICCs on the basis that an ICC was not concerned with tasks or functions. Second, accepting that it was the case that the WMS Manual was more closely aligned with specific tasks, or more directly concerned with marking out deisolation as an activity of work, it does not necessarily follow that the WMS Manual provided “the more natural” descriptor of work, or that, as a descriptor of work, it was more closely related to the work of the operators than were the ICCs. One nature of description of work is not inherently more “natural” than another. It does not follow that a document that marks out the specific tasks associated with particular work (or better marks out those tasks), ought to be regarded as the “natural” point of reference in a search for the likely understanding of what employees intended when describing their work. Work can be described by reference to the tasks it involves. For instance, the work performed in court by a barrister may be identified by task as the examination of witnesses and the making of submissions. But that work can also be described as “court work” or the “work performed in court” and every barrister will know exactly what was intended. A task-based descriptor is not necessarily a more natural means of describing work. Nor is a task-based descriptor of work necessarily a superior descriptor of work. But, more fundamentally, a superior task-based descriptor of work is not necessarily the descriptor that is more closely related to the work it describes. There is an issue as to the discrimen used by the primary judge to identify which sense of “de-isolation” was more closely related to the work of the operators. In my view, the finding that work tasks were marked out or better marked out by the WMS Manual did not justify the finding that the sense in which the Manual used “de-isolation” was more closely related to the work. Before I further explain why it is that I have reached that conclusion, I should first turn to the evidence of Esso’s witnesses Kristeff and Dunbar, which the AWU contended was at odds with the primary judge’s findings at [89]. In his evidence in chief, Kristeff referred to Part 4.5 of the WMS Manual dealing with “Isolating Facility Systems and Equipment”. He described the “isolation process” as involving identifying, and then removing, all energy sources from vessels, pipelines or other equipment, to enable them to be worked on safely. He noted that energy isolations are managed by an ICC. He said that the ICC becomes part of the work permit documentation. Kristeff noted that there is a concept in the “electronic permit to work system” (by which I assume he meant or at least included the ICC), known as “De-isolation in Progress”. Kristeff noted that that term appeared in the flow chart in the WMS Manual. He said the flow chart sought to explain broadly how the process steps of the electronic permit to work system mapped to the field procedures and standards described in the WMS Manual. He then addressed the “purpose” of implementing the “De-isolation in Progress” status. He said the purpose was simply to identify that work was being undertaken by operators in the field in connection with the isolated equipment. He acknowledged that “such work” may include tasks such as air freeing and equipment testing, as well as de-isolations. In cross examination he was taken to that evidence given in chief. He said that the flow chart was the template for the ICC process. He confirmed that the process involved the planning of a particular isolation, including drawing up any necessary plans showing various isolation points in the equipment to be worked on. He confirmed that the ICC set out all the various steps that needed to be taken to isolate the piece of equipment. It was suggested to him that the ICC then went to identify all the particular steps that needed to be taken to bring the equipment “back online”. To that he said “Well, it lists the – all your points that you want to account for on that ICC”. He acknowledged that the ICC “sets out what you need to do for your isolation”, and also for bringing the equipment “back onto line again” or “reinstating it”. Kristeff agreed that the ICC was meant to be “a control mechanism which workers have to pay attention to”. He confirmed that whenever workers “do anything out in the field they have to record it on the ICC so everybody knows where things are up to”. He agreed that “so far as the day-to-day work of isolating and reinstating equipment is concerned people have to know how to operate the ICC”. Kristeff agreed that “they have to work according to it”. Kristeff agreed that when the ICC is marked as “de-isolation in progress”, what happens then is that “whatever steps need to be taken to bring the equipment back online are then taken”. He also agreed that once those items of work are completed the ICC is marked as “deisolation confirmed”. It was suggested to Kristeff that the process starts off with the isolation of the equipment and that when the work on the equipment has been completed “you go into the de-isolation of the equipment”. To that he said, “well, I use the term return to service”, but agreed that the ICC did not use that phrase and that in the ICC “it’s called de-isolation”. Kristeff acknowledged that the steps required to be taken, once equipment is put into “de-isolation in progress”, could include, depending on the equipment, equipment testing, air freeing, leak testing, removing energy isolations and removing temporary defeats. He acknowledged that these were actions or activities that are carried out, depending on the equipment in question, “between the stage when the ICC is marked de-isolation in progress and the stage when it is marked de-isolation confirmed”. It was suggested to Kristeff that it was reasonable to refer to the steps that have to be taken on various valves as part of getting the de-isolation completed, as being part of the de-isolation. To that, Kristeff answered “Well, it’s part of the – part of the process” and then said “Well it could have been part of the air freeing or pressure testing”. To the cross-examiner’s suggestion that “it’s all part of the process of de-isolation”, Kristeff said de-isolation in progress wouldn’t be completed unless “you’ve done all the other steps”. Kristeff was then taken to a statement made in his evidence in chief, that “The fact that the position of any particular bleeder valve might be recorded in [the] ICC ... doesn’t make the manipulation of the bleeder valve part of the de-isolation of equipment”. It was put to Kristeff that that statement was wrong and inconsistent with the evidence he had given under cross examination. To that Kristeff said “No, because you can manipulate bleeder valves without actually being on the ICC”. In the following exchange Kristeff resiled from that statement (but in a later exchange re-asserted it): But if they are on the ICC then they are part of the de-isolation, aren’t they?---Under – not under my definition of --- Well, I’m really not interested ---? ---It’s – it’s – yeah, I know that, and it’s part of the steps. They are part of the steps in the ICC, are they not? ---Yes Okay. And because they’re in the ICC they are part of the de-isolation process under the ICC, aren’t they? They have to be done, don’t they? ---They would have to be, yeah. Kristeff confirmed that it was a rule that the ICC had to be followed. He agreed that it was a safety rule and said “it’s a catch-all”. Kristeff acknowledged that the ICC “tells you what you have to do step by step” and that that is what the “ICC is for”. He noted, however, that the steps set out in the ICC were not “in order”. There was then the following exchange: And the system also describes the process of bringing the equipment back online as de-isolation. Now, we've been through that and you've agreed to that, and that's right, isn't it?---Yes. The document says so. The document doesn't say any other words, does it?---No. And so the operators out in the field who are working with this document are working with a document that describes what they do as de-isolation. That's correct, isn't it?---That's correct. Yes. And so they're talking about doing a de-isolation of equipment why would you think that they were referring to anything else?---Because that's the electronic version versus - there's a lot of - lot of things in the manual that relies on going back to definitions, right, and that's why we have a definitions page. Well, why haven't you changed the definition in the ICC?---Don't know why that hasn't happened. Well - - -?---Or why - why that's different. Well, all right?---I can't answer that. No. The isolation certificate is provided in the WMS, isn't it? It's in that flow chart that we spoke about?---Yes. And that flow chart also uses the word "de-isolation", doesn't it?---Yep. In cross examination, Dunbar agreed that the ICC was “part of the WMS”, but added that the definition of de-isolation was provided “earlier in the document”. When it was put to him that the definition of “de-isolation” in the WMS Manual was not the way in which the expression is used in the ICC, Dunbar said “the way it’s described in the electronic system I think is a overarching term used to define a number of sub activities that need to be performed”. He added, “the electronic system is grouping a number of things”, and that “the definitions of those things [are] provided [in] more detail in the manual”. When asked if he agreed that the ICC “groups all of those activities under the description of De-Isolation”, he agreed that the “electronic software system we use uses that term”. He added that the tasks being performed “are defined more clearly in the overarching manual”. Dunbar agreed that the WMS Manual was “generic” and that, conversely, the ICCs are “developed for a specific piece of equipment per the manual” and were not “generic”. Dunbar rejected the proposition that the steps that needed to be taken were set out in the ICC. He said “[t]he ICC just lists the valves, the energy isolation that we put in place to take that equipment out of service”. Dunbar accepted that when he referred to the “electronic management system” he meant an ICC and that the fact that it’s electronic does not make an ICC less important. There was then the following exchange: No. And it’s electronic and it’s kept in this way because [Esso] does it, isn’t it? It’s Esso’s choice to do it this way, isn't it?---Yes. And it's Esso’s choice to give instructions to its operators in this form, isn’t it?---Yes. And it's Esso’s choice to require the operators to comply with it, isn’t it?---Yes. And it’s Esso's choice to have a document which refers to these tasks as de-isolation, isn’t it?---No. Well, who put that word in there, then?---The - the electronic system refers to - - - But whose electronic system is it?---Yes. It is Esso’s system. Right. And who paid to have it done? Esso?---Yes. So Esso put the word in, didn’t they?---Yes. Neither Kristeff or Dunbar conceded that, when the term “de-isolation” was used, it necessarily meant de-isolation in the broad sense. However, Kristeff’s evidence was generally confirmatory of the evidence given by AWU witnesses in relation to the use and significance of the ICCs to the work of operators when reinstating isolated equipment. His evidence was consistent with the term “de-isolation” being used in the ICCs synonymously with “returning equipment to service” or “bringing the equipment back on line”, and he acknowledged that such a process included the tasks of air freeing and leak testing. He agreed that the operators were working with a document (the ICC) that described that work as “deisolation”. He agreed that the ICCs told operators the work they have to do and that operators have to work according to it. In sum, Kristeff accepted that the ICCs were working documents which referenced the work done by operators to reinstate isolated equipment, and in doing so, called that work “de-isolation”. Dunbar agreed that ICCs used “de-isolation” to refer to the activities required to reinstate equipment but maintained that those activities were more clearly defined in the WMS Manual. He accepted that ICCs gave instructions to the operators and referred to the required tasks as “de-isolation”. Their evidence, and in particular that of Kristeff, demonstrated Esso’s actual knowledge as to how the term “deisolation” was used in association with the ICCs. It was confirmatory of the conclusion that an objective assessment would arrive at in any event. If the extent of the nexus between the competing work descriptions and the actual performance of the work is the determinative discrimen (as I think it must be given the manner in which the case was conducted), then I would conclude that there was a stronger nexus between the ICCs’ description and the work of the operators in reinstating isolated equipment, than there was between the WMS Manual’s description and that work. The evidence of Kristeff, and that of the AWU’s witnesses set out at [310]–[312] above, to the effect that the ICCs were working documents which referenced the work of operators whilst reinstating isolated equipment, was not relevantly contradicted. That evidence established that, whilst that work was performed, it was, in accordance with the ICC process, labelled by the operators themselves as “de-isolation in progress”. The evidence demonstrated the prevalence of the ICCs, the frequency of operators’ exposure to them, and their presence whilst work was performed. To my mind, the evidence provided a sound rationale for the conclusion that the denotation of “de-isolation” used by the ICCs had a close connection to the work of the operators. In contrast, whilst the WMS Manual was an important document which gave a superior taskbased description of the operators’ work to that given inferentially by the ICCs, its use or its prevalence at the coal face was far more limited. As described by the primary judge, the WMS Manual was a “procedures manual” which detailed the processes required for undertaking work permit and isolation procedures. It was used for both initial and refresher training. But for experienced operators repetitively performing isolation and de-isolation functions, the WMS Manual was not an instructional document which guided them through the performance of that work. The operators did not require guidance from a manual whilst performing those tasks and functions. As the primary judge said at [21] “... it is a given that those involved in this work will know what is required”. Whilst I think the same can be said of the ICCs, that is, that the ICCs were not instructional in that they were not used by operators to guide them through their requisite tasks, the ICCs had a function which to my mind was of central importance. The ICCs identified the valves upon which the well-understood tasks had to be performed. In that respect, they referenced the work of the operators, moving the operator from one valve to the next and prompting the operator to perform all of the well-understood tasks necessary to be performed in order for the operator to be able to place the valves into their “de-isolated” states. In practical terms the nexus between the ICCs and the actual performance of the work of reinstating isolated equipment was closer than that between the WMS Manual and that work. A reasonably-based objective assessment made by Esso should have come to that conclusion including because in construing the likely intent of the reference to the work made in the AWU’s notice, substantial weight had to be given to the work descriptor that the operators would likely have regarded as most-closely connected with the work that they did in reinstating isolated equipment. Whilst I would have arrived at that conclusion even were it the case that “de-isolation” was only used in its narrow sense in the WMS Manual, my conclusion is reinforced when it is recognised, as I have earlier explained, that the WMS Manual used “de-isolation” in both of its senses and thus inconsistently with its own definition of that term. The objective test applied by the primary judge relied upon a factual foundation which was erroneous in two respects. First, the primary judge mischaracterised the nature of the ICCs by holding that an ICC was not concerned with tasks or functions. Second and relatedly, the primary judge mischaracterised the closeness of connection between the ICCs and the work of the operators. As trial judge, the primary judge enjoyed no relevant advantage over this Court in characterising, as he did, the uncontroversial facts. In those circumstances an appellate court is obliged to conduct a “real review” and give the judgment which, in its opinion, ought to have been given at first instance: Ashby v Slipper [2014] FCAFC 15 ; (2014) 219 FCR 322 at [57] –[58] (Mansfield and Gilmour JJ), citing Fox v Percy [2003] HCA 22 ; (2003) 214 CLR 118 and CSR Ltd v Della Maddalena [2006] HCA 1 ; (2006) 80 ALJR 458. For the reasons given, in my view, the proper application of the requisite test results in the conclusion that Esso would have understood that the ban on the “de-isolation of equipment” was an intended reference to a ban on all of the tasks involved in reinstating isolated equipment, including leak testing, air freeing and the manipulation of bleeder valves associated with those tasks. There is a further fact that I should mention, although, in the end, it is of no consequence. The protected action ballot which authorised the taking of industrial action approved: An unlimited number of indefinite and/or periodic bans on: ... xi. the isolation or de-isolation of equipment; As earlier outlined, the word “isolation”, as it is used in the WMS Manual, is an unambiguous reference to the process of isolating equipment. The juxtaposition of “deisolation” and “isolation” in the ballot question, strongly suggests that the antonym of “isolation”, that is, the inverse process to the process of isolation, was intended by the reference to “deisolation”. The AWU’s notice did not notify a proposed ban on isolation in conjunction with a ban on deisolation. A notice of intended industrial action, given by the AWU four days earlier on 30 January 2015, did. The ballot and the earlier notice are prior communications which provide a cogent contextual basis for construing “de-isolation” in the AWU’s notice consistently with its prior use, as an antonym for the process of isolation. Those observations would serve to reinforce my view that a reasonable employer in Esso’s position would have understood the AWU’s notice to be giving notice of an intended ban upon all of the tasks involved in reinstating isolated equipment. However, no reliance was placed on the prior communications. It is therefore not appropriate for me to rely upon them. Nor, in the circumstances, is it necessary to resolve the competing contentions as to whether, and if so in whose favour, the contra proferentem rule ought to have been applied. Although the primary judge took a position, in the end the judge did not give application to that rule. Nor has it been necessary to do so in the approach I have taken to re-determining the issue as to the scope of the AWU’s notice. I should add, however, in deference to the submissions made, that in my view a contra proferentem approach to resolving ambiguity in a notice of industrial action is unattractive. That is particularly so in a case such as this where, if ambiguity existed, both the notifier and the notified bore responsibility for its existence. For those reasons, I would uphold the AWU’s first ground of appeal. I turn, then, to the consequence that has for the AWU’s challenge to the declarations made by the primary judge. THE DECLARATIONS MADE BY THE PRIMARY JUDGE For reasons I will now explain, the AWU needed to do more to escape liability than establish that the refusal by AWU members to conduct air freeing, leak testing and bleeder valve manipulation ( primary industrial action ) was protected industrial action. The finding made by the primary judge that the AWU’s notice did not encompass the primary industrial action led to the finding that the primary industrial action was not protected industrial action. It was uncontentious that the primary industrial action was organised by the AWU. It was also uncontentious that the AWU’s members at Longford also took other industrial action, also organised by the AWU, in support of claims for a replacement enterprise agreement ( other industrial action ). That industrial action was extensive. It included various bans on the performance of specific tasks, an overtime ban and work stoppages. The detail is given at [35], [40] and [42] of the reasons of the primary judge. There was no issue at trial that the other industrial action had been validly notified and would have been protected industrial action save for the operation of s 413(5) of the FW Act. The primary judge’s finding that the primary industrial action was unprotected was a foundation for each of the declarations made by the primary judge, other than declaration 3 (to which I will return). The organisation by the AWU of the other industrial action, in so far as that industrial action was taken after 6 pm on 6 March 2015 (being the time from which the primary judge determined that the operation of s 413(5) was enlivened) was also significant for at least some of the declarations made. Broadly speaking, the declarations made by the primary judge were that: (1) By reason of the primary industrial action, the AWU contravened s 421(1) of the FW Act by contravening the FWC’s s 418 order of 6 March 2015 ( declarations 1 and 2 ); (2) By operation of s 413(5) of the FW Act, the primary industrial action and the other industrial action were unprotected industrial action ( declaration 4 ); (3) By organising the primary industrial action with intent to coerce Esso, the AWU contravened ss 343 and 348 of the FW Act ( declarations 5, 7, 8 and 10 ); (4) By organising the other industrial action with intent to coerce Esso, the AWU contravened ss 343 and 348 of the FW Act ( declarations 6 and 9 ); (5) By organising the primary industrial action, the AWU contravened s 346(b) of the FW Act ( declarations 11 and 13 ); and (6) By organising the other industrial action, the AWU contravened s 346(b) of the FW Act ( declaration 12 ). Declaration 3 was that the AWU contravened s 421(1) of the FW Act by contravening cl 5.1(a) of the FWC’s s 418 order made on 6 March 2015, by its conduct in failing to prepare a written notice as soon as practicable following 6 pm on 6 March 2015. It is not in contest that if the AWU’s first ground of appeal succeeded, as I have determined it should, grounds two and three should also succeed. It would follow that declarations 1 and 2 referred to at (i) above should be set aside. With one reservation, it was accepted by Esso that if the AWU succeeded on appeal grounds 1–3, the AWU would avoid all liability under ss 343, 384 and 346(b). If Esso’s reservation was not sustained, it would follow that each of declarations 5, 7, 8 and 10 (referred to in (iii) above), 6 and 9 (referred to at (iv) above), 11 and 13 (referred to at (v) above), and 12 (referred to at (vi) above) should be set aside. It would also follow that declaration 4 (referred to at (ii) above) should be set aside. The only remaining declaration is declaration 3 and that declaration is the basis for Esso’s reservation. Esso contended that, unless the AWU succeeded in showing that it did not contravene cl 5.1(a) of the FWC’s order of 6 March 2015 (which required the preparation of a written notice), s 413(5) would operate to render all industrial action taken after 6 pm on 6 March 2015 unprotected industrial action and that, consequently, the organisation of that action by the AWU would be in contravention of ss 343, 348 and 346(b) of the FW Act. That analysis suggests that I should turn next to consider the AWU’s appeal grounds 4 and 5, which challenge the primary judge’s finding that the AWU contravened cl 5.1(a) by failing to prepare a written notice. Thereafter I will turn to consider the operation of s 413(5) and the challenge to the approach taken by the primary judge which is raised by Esso’s second ground of appeal. It will also be necessary to consider Esso’s first ground of appeal, which raises s 418 and challenges some of the primary judge’s findings as to the validity of parts of the s 418 orders made by the FWC. THE AWU’S FAILURE TO PROVIDE A WRITTEN NOTICE Order 5 of the FWC’s Order of 6 March 2015 was headed “Service of Order”. That Order provided: 5.1 The AWU must: (a) as soon as practicable following the time specified in clause 6.1 of this Order, prepare a written notice (the Written Notice) signed by an authorised official of the AWU in the following terms: “The Fair Work Commission has issued a section 418 order to stop or prevent unprotected industrial action. The unprotected industrial action identified by the Fair Work Commission was the refusal of employees to perform equipment testing, air freeing and leak testing. The order is called the Esso Australia Pty Ltd Industrial Action Order 2015 No 3 (the “Order”). The Order prohibits the AWU and members of the AWU from organising and engaging in any ban, limitation or restriction on the performance of equipment testing, air freeing and leak testing. The Order applies to: (i) the AWU; (ii) delegates, officers, employees, and agents of the AWU; (iii) employees of Esso Australia Pty Ltd who are members of the AWU and who are covered by the Esso Longford and Long Island Point Agreement 2011. AWU members must not refuse to perform, or impose limitations or restrictions on the performance of equipment testing, air freeing and leak testing for the duration of the order.” (b) by 6.00 pm on 6 March 2015, the AWU must: (i) distribute the Written Notice to all Employees who are members of the AWU by email and by SMS message “Esso members are prohibited from taking unprotected industrial action. Please see link:” with a link to the Written Notice; (ii) provide a copy of the Written Notice to Esso; (iii) publish the Written Notice on a prominent place on the AWU website www.awu.net.au ; 5.2 As soon as practicable following the time specified in clause 6.1 of this Order, Esso must: (a) email all Employees, which must: (i) attach a copy of this order; and (ii) contain the text of the Written Notice outlined in 5.1(a) but adapted to clarify that the Written Notice is communicated by Esso to the Employees. (b) place a copy of this order on the notice board(s) usually used by Esso for the purposes of communicating with the Employees. Both paragraph (a) of cl 5.1 and cl 5.2 referred to “clause 6.1 of this Order”. The Order did not have a cl 6.1, but did contain cl 6, which provided: 6. TERM AND DATE OF EFFECT This order shall come into effect at 6.00 pm on Friday 6 March 2015 and shall cease to have effect at 6.00 pm on Friday 20 March 2015. It seems apparent that the reference to “the time specified in clause 6.1”, in cl 5 of the Order, was an intended reference to 6 pm on Friday 6 March 2015. At [123] of his reasons, the primary judge noted that the AWU raised two defences to Esso’s case that it had failed to comply with cl 5.1. The first of those defences corresponds with what is raised by the AWU’s fourth ground of appeal. The second defence corresponds with what is raised by the AWU’s fifth ground of appeal. As to the first defence and the response given to it by Esso, the primary judge identified the competing contentions (which in substance are those now made on the appeal), and gave his answer at [123]–[124] of the reasons: [123] ... First [the AWU] submits that it was impossible to comply with cl 5.1. By para (a) of the clause, it was to “prepare” a notice in the terms there set out “as soon as practicable following” the time specified in cl 6.1 of the order, namely, 6:00 pm on 6 March 2015. However, by para (b) of the clause, it was to distribute, to provide and to publish that very notice “by” 6 pm on 6 March 2015. The respondent’s point is that, if it complied with para (a) to the letter, it would be impossible to comply with (b). On the other hand, if it took the steps necessary to comply with para (b) to the letter, it would have been in contravention of para (a) by preparing the notice before, as distinct from “following”, the time specified in cl 6.1. [124] The applicant’s response to this defence is to submit that, considered separately, each of paras (a) and (b) of cl 5.1 could be complied with, and that it was the respondent’s obligation to do so, notwithstanding the imperfect fit of the two paragraphs when considered jointly. With respect to para (b), that submission cannot be accepted: the “written notice” which had to be distributed etc was the written notice which, by definition, had not been prepared by the time for compliance with this aspect of cl 5.1. With respect to para (a) of the clause, however, the submission is a good one: albeit that it might by then have been in contravention of para (b), there is no reason why the respondent could not, in compliance with para (a), have prepared the notice as soon as practicable following 6 pm on 6 March 2015. Because it did not do so, the artefact by reference to which para (b) would have operated never came into existence. So the conflict problem to which the respondent refers was never an issue. Subject to the respondent’s second defence to this allegation, I would hold that, by the morning of 7 March 2015 at the latest, the respondent was in contravention of cl 5.1(a) of the Commission’s s 418 order made the previous day (to the extent, of course, that that order had a valid substantive operation). In making cl 5.1 of the Order, the FWC must have intended that an enforceable obligation would be imposed on the AWU to prepare and distribute a notice with particular content. The Order failed to achieve that objective. It was not possible for the AWU to distribute the notice at a time prior to the time it was required to prepare its content. Whether the Order is construed as containing one composite obligation or two interdependent obligations does not matter. In either case, the Order failed to achieve its intended purpose and was legally ineffective. Any other conclusion would impute to paragraph (a) of cl 5.1 a purpose for which it was not made. Declaration 3 should therefore be set aside. It is not necessary for me to consider the alternative attack made on declaration 3 by the AWU’s fifth appeal ground. To the reservation expressed by Buchanan J about whether the FWC was empowered by s 418 to make an order such as Order 5, I would add, without deciding, my own reservation. If s 418 does authorise the making of incidental orders requiring the service of an order made under s 418, that is not what cl 5 (other than cl 5.2(a)(i)) did. The clause purported to require that a notice be distributed which characterised (and arguably mischaracterised) the effect of the substantive order made by the FWC. Assuming that the FWC is empowered by s 418 to require that its order be served, I doubt that within that incidental power is a power to require any more than that the terms of the order made be communicated. Lastly, by its first ground of appeal, Esso seeks, inter alia , to overturn an invalidity finding made by the primary judge in relation to cl 5 of an Order made by the FWC on 5 March 2015. If that aspect of that ground of appeal succeeded, Esso sought a declaration that the AWU contravened s 421 of the FW Act by not complying with the requirement to distribute a notice in accordance with clause 5 of that Order. In so far as they are relevant to the question of any contravention of s 421 by the AWU, the terms of cl 5 of the 5 March 2015 Order are relevantly identical to the terms of cl 5 of the 6 March 2015 Order, which I have just addressed. The same issues are raised as those which I have just determined. If I had upheld Esso’s first ground of appeal, I would nevertheless have rejected the declaration of contravention sought by Esso for the same reasons as I have determined that declaration 3 should be set aside. SECTION 413(5): DID THE CONTRAVENTION OF ANY SECTION 418 ORDER RENDER ALL INDUSTRIAL ACTION UNPROTECTED? I next turn to consider s 413(5). The proper construction of that provision is raised by Esso’s second ground of appeal. It is convenient that I also deal with Esso’s third ground of appeal and the AWU’s contention that declaration 4 should be set aside. Subsections 413(1) and (5) of the FW Act provide: Common requirements that apply for industrial action to be protected industrial action Common requirements (1) This section sets out the common requirements for industrial action to be protected industrial action for a proposed enterprise agreement. ... Compliance with orders (5) The following persons must not have contravened any orders that apply to them and that relate to, or relate to industrial action relating to, the agreement or a matter that arose during bargaining for the agreement: (a) if the person organising or engaging in the industrial action is a bargaining representative for the agreement—the bargaining representative; (b) if the person organising or engaging in the industrial action is an employee who will be covered by the agreement—the employee and the bargaining representative of the employee. Esso contended that, by reason of s 413(5), once the AWU contravened an order made by the FWC under s 418, all industrial action thereafter organised by the AWU in support of the making of replacement enterprise agreements with Esso, was thereby rendered unprotected. As earlier indicated, beyond the primary industrial action, the AWU had organised other industrial action which, but for the operation of s 413(5), was otherwise accepted by Esso as protected industrial action. At the trial, Esso alleged multiple breaches of s 418 orders. It succeeded in establishing three contraventions and they are the subject of declarations 1–3. Declarations 1 and 2 were concerned with the taking of the primary industrial action and declaration 3 with the AWU’s failure to prepare the notice which the primary judge found contravened clause 5.1(a) of the 6 March 2015 order. Other contraventions of s 418 orders alleged by Esso were not established because the orders upon which those allegations depended were held by the primary judge to be wholly or partly invalid. If Esso should succeed in the challenge raised by its first appeal ground to those findings of invalidity, then other possible contraventions will become relevant. However, I have rejected that ground of appeal, for reasons I will shortly explain. Furthermore, it follows from my view that declarations 1–3 should be set aside that there are no contraventions of s 418 orders upon which Esso could rely to invoke the operation of s 413(5). It also follows that, as no unprotected industrial action was organised by the AWU, declaration 4 should be set aside. If it had been necessary for me to determine Esso’s challenge to the finding made by the primary judge as to the operation of s 413(5), I would have rejected it. To that end, I refer to and adopt the observations of Buchanan J (with whom Siopis J and I agreed) as to the proper construction of s 413(5) in Australian Mines and Metals Association Inc v The Maritime Union of Australia [2016] FCAFC 71 , at [53]–[110]. It follows that Esso’s second ground of appeal should be dismissed. By its third appeal ground, Esso sought to extend the scope of the unprotected industrial action declared by declaration 4. The approach I have taken makes the determination of the issues raised by that ground academic. If it had been necessary to determine that ground, I would have respectfully agreed with the approach taken by Buchanan J to the issues thereby raised. THE PROPER CONSTRUCTION OF SECTION 418 Subsections 418(1), (2) and (3) provide: (1) If it appears to the FWC that industrial action by one or more employees or employers that is not, or would not be, protected industrial action: (a) is happening; or (b) is threatened, impending or probable; or (c) is being organised; the FWC must make an order that the industrial action stop, not occur or not be organised (as the case may be) for a period (the stop period ) specified in the order. Note: For interim orders, see section 420. (2) The FWC may make the order: (a) on its own initiative; or (b) on application by either of the following: (i) a person who is affected (whether directly or indirectly), or who is likely to be affected (whether directly or indirectly), by the industrial action; (ii) an organisation of which a person referred to in subparagraph (i) is a member. (3) In making the order, the FWC does not have to specify the particular industrial action. The FWC made three orders of relevance. The first was made on 17 February 2015. The second and third are those earlier referred to and were made on 5 and 6 March 2015 respectively. The primary judge held that the 5 March Order was wholly invalid as being beyond the power conferred by s 418. The primary judge also held that parts of the Orders made on 17 February and 6 March 2015 were in excess of the FWC’s power under s 418. By reference to s 46 of the Acts Interpretation Act 1901 (Cth), the primary judge blue-pencilled or read down those two Orders so that only those terms that were within power, were treated as valid. In terms, each of the three s 418 orders enjoined all “industrial action” as defined by s 19 of the FW Act. The 17 February Order also specifically identified “a ban on the performance of overtime” whilst the 6 March Order also specifically identified “a ban, limitation or restriction ... on the performance of equipment testing, air freeing, and leak testing”. The effect of the blue-pencilling exercise engaged in by the primary judge was that only those parts of the orders that enjoined specifically-identified industrial action were held to be valid. Esso contended that the primary judge was wrong to hold that any parts of orders made by the FWC were beyond power. Esso contended that once the FWC’s jurisdiction to make orders was enlivened (for example, by the FWC finding that particular unprotected industrial action was happening), the FWC had the power to enjoin industrial action more broadly. Subsection 418(3) exists, so Esso contended, so that the FWC does not have to limit its orders to the particular industrial action found to be happening or being organised. If Esso is correct, the FWC made a valid s 418 order enjoining all unprotected industrial action by its Order of 17 February 2015. That would have the effect of bringing back by some 2 days (to 4 March 2015) the point at which it could be found that a s 418 order was first contravened by some aspect of the primary industrial action. It would also have the effect that cl 5.1 of the Order made on 5 March 2015 would be prima facie valid. Esso sought revised declarations which would take those additional effects into account. However, even if Esso’s construction of s 418 was accepted, on the basis of my holdings above, there would be no utility to Esso. That is because the primary industrial action was protected action and, for the reasons stated at [364] above, cl 5.1 of the 5 March 2015 Order was invalid in any event. In the circumstances, it is sufficient for me to say that, like Buchanan J, I see no error in the primary judge’s construction of s 418. Esso’s approach would have s 418(3) displace the plain words of s 418(1). The order that the FWC “must make” is an order that “ the industrial action stop”. There is no scope in those words for the order made by the FWC to travel beyond “the” industrial action. Section 418(3) does not expand the scope of an order that must be made pursuant to s 418(1). Nor, as the primary judge said at [109] does s 418(3) suggest that the FWC no longer need identify the industrial action that its order will prohibit. As the FWC must make a finding that identifies “the” industrial action which it has determined is happening (etc.), there is no basis for thinking that the identification made is not to be reflected in the order. There is every reason for thinking that it should be so reflected, given that a contravention of s 418 gives rise to a civil penalty. That all suggests that s 418(3) is not concerned with whether the industrial action must be identified by the order, but instead simply with the form in which the industrial action may be identified in the order made. Section 418(1) marks out the FWC’s jurisdiction. In so far as that jurisdiction can be properly exercised without the order made particularising the industrial action, s 418(3) makes it clear that particularity is not necessary. Or, in other words, s 418(3) makes it clear that particularity is not necessarily a condition of the effective exercise of jurisdiction. Section 418(3) recognises that in the proper exercise of the jurisdiction conferred by s 418(1), it will not always be possible to specify the industrial action with particularity. That is unlikely to be the case where the industrial action “is happening”, and is thus manifest, but is more likely where the industrial action is prospective and its intended form insufficiently apparent to permit it to be described with particularity. In the latter circumstance, the proper exercise of the FWC’s s 418 jurisdiction, including the requisite assessment to be made as to whether the industrial action is protected industrial action, will need to be made by reference to a general rather than a particular identification of “the” industrial action in prospect. CONCLUSION For the reasons given above, I would formalise by order the grants of leave to appeal and make such orders as were necessary concerning the filing of the draft notices of appeal. I would uphold grounds 1–4, 6, and 12 of the AWU’s notice of appeal. Grounds 5 and 7–11 do not arise because of the way in which I have decided the issues. I would allow the AWU’s appeal. Further, I would dismiss grounds 1–2 of Esso’s notice of appeal. Ground 3 does not arise, but if it did I would dismiss it. I would dismiss Esso’s appeal. I would set aside each of the declarations made by the primary judge on 13 August 2015. I would set aside Orders 16–22 of the Orders made by the primary judge on 13 August 2015, which deal with or are incidental to the conduct of a hearing on the issues of compensation or penalties. Order 14 dismissed so much of Esso’s application as was unsuccessful, and Order 15 discharged an earlier restraint, by Order dated 25 March 2015, on the taking of further industrial action by the AWU. I would let Order 15 stand, its appropriateness not being in dispute on the appeal. I would supplement Order 14 by making an order that Esso’s Second Further Amended Originating Application dated 14 May 2015 be dismissed. Esso’s written submissions in each appeal included a submission that there should be no order as to costs. The AWU made no submission in relation to costs. I am not aware of any circumstance that would warrant departure from the usual order, pursuant to s 570 of the FW Act, that there be no order as to costs. But, I would give the parties an opportunity to make submissions (supported by affidavit, if necessary), within 14 days, in support of a different order. I would order that, in the event any party makes such a submission, the other party have 7 days to respond (again supported by affidavit, if necessary). I would order that, if no submission is filed in support of a different costs order, there be no order as to costs. I certify that the preceding one hundred and sixty-one (161) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice Bromberg. Associate: Dated: 25 May 2016