Visy Packaging Pty Ltd v Siegwerk Australia Pty Ltd

Place: Melbourne

Division: GENERAL DIVISION

Category: Catchwords

Number of paragraphs: 103

Counsel for the respondent/ cross-claimant Siegwerk Australia Pty Ltd: Mr A Schlich

Solicitor for the respondent/ cross-claimant Siegwerk Australia Pty Ltd: Wotton & Kearney

Counsel for the cross-respondent/cross-claimant Nuplex Industries (Aust) Pty Ltd: Mr S Donaldson SC with Mr T Cordiner and Ms SL Ross

Solicitor for the cross-respondent/cross-claimant Nuplex Industries (Aust) Pty Ltd: Sparke Helmore

Counsel for the cross-respondent QBE Insurance (Australia) Ltd: Mr RA Cavanagh SC with Mr AJ Bowen

Solicitor for the cross-respondent QBE Insurance (Australia) Ltd: Gillis Delaney Lawyers

IN THE FEDERAL COURT OF AUSTRALIA VICTORIA DISTRICT REGISTRY GENERAL DIVISION VID 339 of 2009

JUDGE: GRAY J DATE OF ORDER: 19 March 2013 WHERE MADE: MELBOURNE

THE COURT DECLARES THAT, pursuant to the broadform liability policy entered into between the cross-claimant Nuplex Industries (Aust) Pty Ltd ("Nuplex") as insured and the cross-respondent QBE Insurance (Australia) Ltd ("QBE") as insurer, the cross-respondent QBE was and is obliged to defend in the cross-claimant Nuplex's name and on the cross-claimant Nuplex's behalf the cross-claim brought by the cross-claimant Siegwerk Australia Pty Ltd (in liquidation) ("Siegwerk") in this proceeding.

THE COURT ORDERS THAT:

1. The cross-claim of the cross-claimant Siegwerk against the cross-respondent Nuplex be dismissed.
2. The cross-claimant Siegwerk pay the cross-respondent Nuplex's costs of the cross-claim of the cross-claimant Siegwerk.
3. In accordance with its undertaking to the Court by way of security for costs, by letter dated 24 May 2012, Zurich Australian Insurance Limited ("Zurich") pay the amount agreed or assessed as the costs pursuant to the order in paragraph 2 up to the limit of $750,000.
4. Subject to the deduction of: (a) the sum of $500,000; and (b) the amount of the costs recoverable from Siegwerk pursuant to the order in paragraph 2 (whether by the enforcement of the undertaking referred to in paragraph 3 or otherwise) the cross-respondent QBE pay the cross-respondent Nuplex's costs of the cross-claim of the cross-claimant Siegwerk.
5. The costs payable by the cross-respondent QBE pursuant to the order in paragraph 4 be agreed or assessed on an indemnity basis.
6. The cross-respondent QBE pay the cross-claimant Nuplex's costs of the cross-claim of the cross-claimant Nuplex.
7. The costs payable by the cross-respondent QBE pursuant to the order in paragraph 6 be agreed or assessed on an indemnity basis.
8. For the purposes of the order in paragraph 4: (a) the cross-respondent QBE be responsible for agreeing or assessing, and for collecting, the costs payable by the cross-claimant Siegwerk to the cross-respondent Nuplex, pursuant to the order in paragraph 2; (b) payment by or on behalf of the cross-claimant Siegwerk to the cross-respondent QBE of the amount of the costs agreed or assessed pursuant to the order in paragraph 2 be a sufficient discharge of the obligation of the cross-claimant Siegwerk to pay those costs; (c) the cross-respondent QBE be entitled to the benefit of the undertaking by way of security for costs referred to in the order in paragraph 3; and (d) payment by Zurich to the cross-respondent QBE of the amount of the costs agreed or assessed pursuant to the order in paragraph 3 be a sufficient discharge of the obligation of Zurich to pay those costs in accordance with the undertaking referred to in the order in paragraph 3.
9. Liberty to apply be reserved. Note: Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011. IN THE FEDERAL COURT OF AUSTRALIA VICTORIA DISTRICT REGISTRY GENERAL DIVISION VID 339 of 2009

JUDGE: GRAY J DATE: 19 MARCH 2013 PLACE: MELBOURNE

The nature and history of the proceeding 1 There are two issues to be determined in this case. The first is an issue of causation. The question is whether the substitution of one epoxy resin for another in the manufacture of an epoxy-phenolic resin, which was used as a component in the manufacture of a lacquer for coating the insides of can ends, caused the breakdown of that lacquer, so that the food sealed in some of the cans came in contact with the can ends and caused their corrosion. The second issue is whether a claim for loss resulting from that corrosion falls within the terms of a broadform liability policy, by which the manufacturer of the epoxy-phenolic resin is insured. The latter issue involves the construction of the insuring clause of the policy and, if the claim falls within that clause, of four exclusion clauses found in the policy. 2 The proceeding began on 5 May 2009, when Visy Packaging Pty Ltd ("Visy") filed an application against Siegwerk Australia Pty Ltd ("Siegwerk") (sometimes also referred to as SICPA, or Sicpa, from its former name), seeking to recover damages. Visy manufactured cans and ends for those cans, described as "easy open ends", capable of being opened by means of a ring-pull device attached to the end (and therefore not requiring the use of a can opener). Visy claimed to have supplied cans and easy open ends to canners of tuna and to have settled claims by those canners, following the discovery that some cans had corroded as a result of their contents coming into contact with the cans. Visy claimed that the reason for the corrosion was that lacquer supplied by Siegwerk, and applied by Visy to the easy open ends in the course of the manufacture of the cans had failed to seal the can ends so as to prevent their contents coming into contact with the can ends. In addition to the statement of claim that accompanied the application, Visy filed three amended versions of the statement of claim on 14 July 2009, 19 August 2009 and 7 September 2009. Siegwerk filed its defence on 28 September 2009. 3 Also on 28 September 2009, Siegwerk filed a cross-claim against Nuplex Industries (Aust) Pty Ltd ("Nuplex"), the supplier of the resin used by Siegwerk as an ingredient of the lacquer it supplied to Visy. 4 On 17 December 2009, I referred the proceeding to a registrar for mediation. The mediation continued for an unusually long time, because the parties wished it to continue. From time to time, the parties returned for directions hearings, seeking directions that they believed would aid them in the conduct of the mediation. These included directions for the discovery of specific categories of documents and for expert witnesses who had been engaged by the parties to confer and produce for the Court joint documents, setting out the matters agreed, matters of disagreement and reasons for the disagreement. In the course of the mediation, on 17 August 2010, Visy and Siegwerk reached agreement. The claims by Visy against Siegwerk were settled by an agreement that Siegwerk pay a lump sum of $2,250,000, inclusive of interest and costs. The mediation continued between Siegwerk and Nuplex. On 6 April 2011, I gave leave to Nuplex to file and serve a cross-claim directed to its relevant insurer. On 6 May 2011, Nuplex filed a cross-claim against QBE Insurance (Australia) Ltd ("QBE"). 5 In the meantime, Siegwerk went into liquidation, having sold its business. The cross-claim against Nuplex was prosecuted by Siegwerk's insurer, using its right of subrogation. Because of the liquidation, on 16 April 2012, I ordered that Siegwerk's cross-claim be stayed unless, on or before 21 May 2012, Siegwerk provided security in a form satisfactory to the Registrar for Nuplex's costs of the proceeding in the sum of $750,000. By letter dated 21 May 2012, the District Registrar, Victoria District Registry, advised the solicitors for Siegwerk and the solicitors for Nuplex that she was satisfied that a form of undertaking by an insurer was sufficient security for Nuplex's costs of the proceeding. The actual undertaking is in a letter dated 24 May 2012, from Zurich Australian Insurance Limited ("Zurich"). The letter is signed by its Claims Account Manager. It is described as an irrevocable undertaking. It is an undertaking: to pay Nuplex without deduction any and all costs which the Court in this proceeding orders or in the future orders Siegwerk to pay to Nuplex in the event that these costs are unpaid by Siegwerk for a period of 30 days. The amount which it is agreed in writing to pay will be the amount taxed or otherwise agreed. Payment of the said amount is to be within 60 days of the failure by Siegwerk to pay any taxing order or written agreement between the parties as to the amount of costs. This Undertaking shall be subject to a financial limit of not more than $750,000. 6 The trial of the proceeding was therefore a trial of Siegwerk's cross-claim against Nuplex, and of Nuplex's cross-claim against QBE. Although the pleadings raised almost every conceivable issue that might be relevant to those cross-claims, on the first day of the trial the parties were able to reach agreement about the facts, so as to enable the issues that were really in dispute to be determined. The evidence of two expert witnesses, one called on behalf of Siegwerk and the other on behalf of Nuplex, was heard jointly on the second day of the trial. There was then a further day and a half devoted to submissions.

The agreed facts 7 A significant part of the agreed statements of facts was concerned with the history of the entitlement to use the formula for producing the epoxy-phenolic resin and the formula for producing the lacquer using that resin, which were the subject of the proceeding. In 1989, ICI in the United Kingdom developed the formula for an epoxy-phenolic resin, given the code DEP216, for use in coating deep drawn cans. In the same year, ICI UK granted to ICI Australia/Dulux a licence to manufacture a lacquer. When ICI was sold to Orica in 1997, the licence was transferred to Orica. In April 1998, Orica made a lacquer designated as 788-R-3171, using DEP216 and adding waxes and aluminium paste. Orica conducted trials of this lacquer on full panel easy open ends for cans. 8 In September 1998, Orica sold its industrial coatings business to PPG and transferred the relevant licence. In August 1999, PPG supplied its first commercial quantities of 788-R-3344 lacquer to Visy, using a code 788-52161. 9 In 2000, Les Chambers, who had previously worked successively for ICI, Orica and PPG, moved to Sicpa to become Sicpa's ICI Business Manager. Subsequently, ICI transferred the licence for the lacquer to Sicpa. On 11 April 2000, Siegwerk (as Sicpa had by then become) provided Nuplex with confidential process instructions to make an epoxy-phenolic resin, designated as 3490 EP. Among the ingredients these instructions required were a product called ML98 and Dow Chemical's DER669E epoxy resin. Almost a month later, on 10 May 2000, Siegwerk provided Nuplex with a second version of the confidential process instructions, stating that the DER669E epoxy resin must be sourced from Dow Chemicals in Germany. 10 On 8 June 2000, Nuplex made a batch of 3490 EP epoxy-phenolic resin, for the purpose of a laboratory trial. Siegwerk approved the laboratory trial by approving a laboratory sample on 14 June 2000. On 10 July 2000, Nuplex manufactured its first trial batches of 3490 EP. It did so pursuant to a toll manufacturing agreement, entered into between Sicpa and Nuplex on 30 June 1999, the terms of which covered the manufacture of products ordered by Sicpa from Nuplex. A further batch or batches of 3490 EP was or were made by Nuplex on 24 August 2000. 11 In September 2000, ICI licensed PPG to make DEP216 epoxy-phenolic resin and 788-52161 lacquer. In October 2000, ICI licensed Siegwerk to make epoxy-phenolic lacquer with a product code 788-50821, as a substitute to PPG's 788-52161. The 788-50821 was found by Visy to be unsuitable for use with full panel easy open ends for cans, with the result that ICI provided an alternative product lacquer formulation code, being 788-52161. On 2 April 2001, Nuplex made a commercial production "scale up" batch of 3490 EP, under the supervision of Steve Browning and Ron Laherty. 12 In April 2001, Siegwerk and ICI both received a letter of demand from PPG regarding the use of 788-52161. This was the start of a dispute between PPG and Siegwerk about the entitlement to use the formula for 788-52161 lacquer. The licence agreement between ICI and PPG was the subject of a series of three-month extensions beginning in September 2001, December 2001 and March 2002. In the meantime, Siegwerk was proceeding to develop its own lacquer with the product code 650772. On 23 May 2001, Nuplex made its first batch of 3490 EP for use in Visy customer trials. Siegwerk approved that batch for that use on 29 May 2001. In July 2001 Visy gave interim customer approval for Siegwerk's 650772, following three-month test pack trials. In January 2002, Siegwerk provided 600 litres of 650772 to Visy for extended trials on one million can ends, by way of a final trial before full customer approval of the lacquer. On 20 February 2002, Nuplex finalised its permanent production process for 3490 EP. 13 In April 2002, ICI gave Siegwerk a new formulation for its lacquer, with a new formula number 650814. The change in the formula was a change of the aluminium component from Metana Superfine to Silveral 7NL. Siegwerk provided 400 litres of 650814 to Visy and Visy commenced the approval process. 14 In May 2002, PPG demanded from Siegwerk royalties in respect of the production of 650772 lacquer. On the advice of ICI, Siegwerk withdrew its 650772 lacquer. At this stage, PPG was still supplying 788-R-3171 lacquer to Visy for commercial production. 15 Between August 2001 and July 2002, Visy carried out tests. The result was that, in September 2002, Siegwerk gained full customer approval for its 650814 lacquer. In May 2003, Visy commenced commercial production using 650814 lacquer on can bodies. In October 2003, the viscosity specification for 650814 lacquer was changed from 80-100 FORD 4 to 65-95 FORD 4. 16 In October 2003, Visy began commercial production of cans using 650814 lacquer on can ends and bodies and stopped using PPG's lacquer. The cans so produced were supplied to PLTP, pursuant to a supply agreement dated 12 December 2001. In January 2004, PLTP commenced filling cans supplied by Visy, which were coated with Siegwerk's 650814 lacquer. The product specification for the lacquer was contained in a specification manual dated 4 August 2000. The cans were filled with seafood, including tuna supplied by Simplot. Simplot sold these cans of tuna to retailers for purchase by consumers. 17 On 25 July 2004, the first report was received of corrosion in cans coated with Siegwerk's 650814 lacquer. The cans were filled with tuna, with a dressing called "zesty vinaigrette", marketed under the John West brand. Corrosion in cans filled with this product occurred after a minimum of four to six weeks from the filling of the cans. On 4 August 2005, Simplot initiated a product recall of cans filled with its zesty vinaigrette tuna. On the following day, Simplot notified PLTP that it would be seeking to recover all of its costs associated with the recall of that product. Between August 2004 and January 2005, Simplot, Visy and PLTP investigated the cause of the corrosion in the cans. Simplot and PLTP determined initially that they would continue to use the cans provided by Visy. In January 2005, further cans were observed to be leaking as a result of corrosion. On 28 January 2005, Simplot initiated a further product recall, this time in respect of a larger range of its products. 18 Between April 2003 and August 2004, Siegwerk ordered 3490 EP epoxy-phenolic resin from Nuplex from time to time. In response to those orders, Nuplex manufactured batches of that epoxy-phenolic resin and supplied them to Siegwerk. Each such supply took place no earlier than two weeks after the order was placed. Once supplied, the resin was used by Siegwerk within one or two months of the order being placed with Nuplex. It was used to make the lacquer supplied to Visy for the coating of cans. All batches of the resin supplied to Siegwerk and used by Siegwerk to make lacquer complied with the specifications agreed between Siegwerk and Nuplex, according to the results of tests required to be performed by agreed process instructions. As was revealed subsequently, in a number of batches, Nuplex substituted an epoxy resin known as Epikote 1009 for the epoxy resin DER669E specified under the toll manufacturing agreement. It was agreed between Siegwerk and Nuplex that this substitution constituted a breach of an implied term of the toll manufacturing agreement that the composition of the epoxy-phenolic resin supplied would correspond with the instructions given by Siegwerk. 19 The table in the schedule to these reasons for judgment shows the available information about the batches of 3490EP resin supplied by Nuplex to Siegwerk during the relevant period. The table shows the batch number given to each batch, the epoxy resin used (DER669E or Epikote 1009), the approximate latest likely order date, the date of manufacture by Nuplex, the date on the pre-shipment sample testing document, the date of the Nuplex tax invoice and the date of the Nuplex delivery document, these two dates being relevant to receipt of the batch in question by Siegwerk. In some cases, the information that would be contained in one or more of the columns in the table is not available. Where that is the case, the table so indicates. 20 Some batches of Nuplex's 3490 EP epoxy-phenolic resin were tested by Visy, or by experts retained for the purpose, with respect to issues of corrosion in cans. There were four batches identified by Nuplex's batch numbers as having been tested, three of which appear in the table. They are batch numbers 309024 (item 9), 311010 (item 14) and 311034 (item 16). In addition, batch number 207010 is identified as having been the subject of testing, but it is not shown in the table. 21 There were also tests conducted on several batches of Siegwerk's lacquer. These tests were identified by reference to Siegwerk's batch numbers and the testing dates as follows: Batch number Test Date 590111 25 November 2003 590112 28 November 2003 593405 12 December 2003 594406 20 December 2003 597954 20 January 2004 598891 4 February 2004

Descriptions and the results of the tests conducted are described in various expert reports. The parties are agreed that those expert reports record those tests accurately. 22 It was also an agreed fact that at all times, the ML98 component of the Siegwerk lacquer was supplied by Siegwerk and not by Nuplex. 23 There was a separate agreed statement of facts as between Nuplex and QBE, presumably in order to reflect the facts pleaded in Nuplex's cross-claim against QBE and QBE's amended defence to that cross-claim, filed on 24 November 2011. It is unnecessary to set out or summarise the facts contained in this agreed statement, because they are included in the facts already set out, except to the extent that the agreed statement refers to the contents of Visy's further amended statement of claim, which are summarised in [25]-[26] below.

Absence of direct connection 24 Counsel for Nuplex attempted to expand the information in the table that is now in the schedule to these reasons for judgment. Nuplex claimed that, at some time, Visy had kept batch records, linking particular batches of cans with particular batches of lacquer. Nuplex had been unable to obtain those records. Counsel for Nuplex said that it was possible to identify those batches of lacquer which Visy had said at the time that it associated with can failures. Using this information in conjunction with the information in the table in the schedule to these reasons for judgment, Nuplex claimed to be able to show that some of the suspect lacquer batches were made with lacquer containing DER669E epoxy resin and others were made with Epikote 1009. By this means, Nuplex attempted to rule out substitution of epoxy resin as a cause of can failure. There was no agreement between the parties about the basis of this argument. The argument fails, because of the absence of sufficient data about batches of epoxy-phenolic resin actually used in the manufacture of particular batches of lacquer. The assumption is that a batch of lacquer tested on a date after a delivery of epoxy-phenolic resin was manufactured using that batch (or one of several batches) delivered prior to that date. No such assumption can be made. It is not possible to connect any specific batch of epoxy-phenolic resin with any specific batch of lacquer. Nor is it possible to connect any specific batch of lacquer with any specific can. Any direct connection between the substitution of Epikote 1009 for DER669E and the corrosion of cans cannot be established by evidence, unless the expert evidence of Dr John Scheirs is accepted.

The pleading of Visy's claim 25 For completeness in relation to the facts, before examining the expert evidence, it is necessary to refer to the composition of the claim brought by Visy against Siegwerk. The further amended statement of claim filed by Visy on 7 September 2009 pleaded causes of action in misleading and deceptive conduct, in contravention of s 52 of the Trade Practices Act 1974 (Cth) (paras 41(d) and 48-57), and breach of contract, including breach of the terms implied by ss 18 and 20 of the Goods Act 1958 (Vic) (paras 58-61.2). Paragraph 68 of the further amended statement of claim referred to the claims made against Visy by PLTP and Simplot for indemnity "in respect of costs incurred by them as a result of the need to withdraw the product packaged in the cans with Defective Ends from supply". The particulars to para 68 described the claims as follows: By letter dated 5 August 2004, Simplot informed PLTP of its intention to seek recovery of all costs associated with the failure of the product referred to in paragraph 60. By letter dated 12 August 2004, PLTP informed Visy in turn of its intention to seek from Visy recovery of all such costs. Copies of these letters are available for inspection at the offices of Visy's solicitors. Between those dates and July 2006, Simplot and PLTP respectively claimed the following losses from Visy: PLTP claimed a total of approximately $1,778,000, comprising: (a) loss of stock on hand $1,466,000 (b) labour costs in replacing stock $ 109,000 (c) pallet hire costs wasted $ 22,000 (d) costs of disposal of condemned stock $ 63,000 (e) finance costs on condemned stock $ 101,000 and (f) miscellaneous expenses $ 17,000.

Simplot claimed a total in the approximate range $7,482,000-$9,865,000, comprising: (a) loss of stock on hand $4,887,000 (b) costs of recalling stock $ 293,000 (c) sums paid to its customers $1,319,000 (d) loss of profits $ 813,000-3,196,000 and (e) claim preparation costs $ 170,000.

26 Paragraph 69 of the further amended statement of claim alleges that there was a substantial risk to Visy that, as a result of its supply of defective ends to PLTP, Visy would suffer permanent loss of its ongoing custom with PLTP. Paragraph 70 alleges that, in order to avoid that substantial risk of the loss of business, and to mitigate that loss, Visy reasonably paid amounts totalling $7,735,786.00 to PLTP and Simplot. The particulars to para 70 list the payments made as follows: Visy paid PLTP $1,630,000 and Simplot $6,105,786. The sum paid to PLTP comprised payment on account of: (a) loss of stock on hand $1,317,266 (b) labour costs in replacing stock $ 109,464 (c) pallet hire costs wasted $ 21,857 (d) costs of disposal of condemned stock $ 63,429 (e) finance costs on condemned stock $ 101,054 and (f) miscellaneous expenses $ 17,072.

The sum paid to Simplot comprised payment on account of: (a) loss of stock on hand $4,084,665 (b) costs of recalling stock $ 208,501 (c) sums paid to its customers $1,187,998 (d) loss of profits $ 514,622 and (e) claim preparation costs $ 100,000.

The Expert Reports 27 In order to understand the following summary of the expert reports generated by the corrosion of the cans filled with tuna, it is necessary to understand one aspect of the manufacture of the can ends. It is common ground among the parties that, before each easy open can end is attached to its can, it is scored, so as to facilitate the removal of the end by pulling on its ring attachment. The score line is close to the circumference of the end, so that when the end is attached to the can, the score line will be adjacent to the side of the can. The scoring process is completed after the application of the lacquer that will form the internal seal of the can end. It is performed by means of a machine that applies a scoring tool to the can end whilst the end is in contact with an anvil. From an early stage, the scored area was identified as a key area in the search for the cause of corrosion. 28 In July 2005, Visy produced an internal technical report, consisting of a synopsis accompanied by eight documents covering Visy's investigations, analyses and conclusions. This report was not tendered in evidence, but its findings are summarised in subsequent expert reports. In particular, Dr John Scheirs was invited to comment on the Visy report, which he did by means of his first report, dated 17 March 2006. Dr Scheirs gave evidence on behalf of Siegwerk. He is an analytical chemist, specialising in, among other things, polymer processing and problem solving, failure analysis of polymers, and polymer degradation and stabilisation. His capacity to give opinion evidence, based on his specialist knowledge and experience, was not questioned. 29 In his report dated 17 March 2006, Dr Scheirs summarised the conclusions of the Visy report of July 2005 as follows: (i) Practical filling trials support the fact that there are differences between the Sicpa coating and the PPG coating. The Sicpa coating performs worse in the Tempters product range than the PPG supplied coating. The Sicpa material has not provided the required shelf life warranty for commercial reliability of the container performance. (ii) An overseas expert validated that Visy had correctly applied the Sicpa coating and that it remained intact during the can production process. (iii) Analysis of the coating using infrared spectrum and GPC found a difference in the coating between Sicpa supply and PPG supply. (iv) Differences between the coatings were found by analysis of dried applied pigmented samples. Visy also claim that FTIR infrared analysis detected differences between the PPG and Sicpa dried film coatings in August 2004. 30 Dr Scheirs examined over 400 corroded cans, some with PPG lacquer and some with Sicpa lacquer, taken from unopened boxes of product recalled from South Australia and conducted a range of tests. He regarded high-performance liquid chromatography ("HPLC") on wet coatings as a superior technique to fourier transform infrared spectroscopy ("FTIR"). 31 Dr Scheirs's observations of cans led him to conclude that the corrosion had occurred from the outside in, rather than from the inside out. He found no internal corrosion or perforation in the cans he opened, except where gross corrosion and spoilage was already present. He found signs of corrosion on the external score lines of the easy open ends in dozens of cans. 32 The findings of Dr Scheirs in his report dated 17 March 2006 can be summarised as follows: (a) microscopic examination of Sicpa-coated ends showed intermittent cracking on the inside score region. PPG coating in the same region also showed intermittent cracking, but to a lesser extent. Gold Organosol coatings showed no evidence of cracking in the score region. The presence of "ridge" marks in the Sicpa-coated lids indicated that the scoring process had stressed the coating, making it more susceptible to cracking in the internal score region. (b) HPLC testing showed that there were no significant compositional differences between the Sicpa and PPG coating samples that were tested. Nor did reflectance IR scans show any significant differences between the Sicpa and PPG coatings. (c) Corrosion could be attributed to external corrosion initiated at the bare metal on the external score line of the easy open end, which advanced inwards through the score residual and ruptured the internal lacquer, leading to escape of the liquid contents. Corrosion from the outside in was confirmed by photo micrographs of the internal lacquer blistering. The external corrosion could be related to the score repair coating for the easy open ends, which appeared to be inadequate or absent. Visy applied only an oil to the external score to prevent corrosion, but this was removed during the filled-can sterilisation process. An external repair lacquer to protect the score line and the cut edges would have provided better protection. (d) Visy applied a relatively thin coating of the Sicpa lacquer (typically 7-8 microns), whereas the thickness of PPG lacquer coating was typically 10-12 microns and Gold Organosol coatings were approximately 20 microns. Visy also applied two coats, whereas three coats would have been preferable to resist damage caused by the scoring operation. A minimum internal coating thickness had not been established, but there would be a critical coating thickness below which the scoring operation would cause intermittent damage to the coating in the score region. A thicker coating was more likely to withstand deformation during the scoring process. (e) "When the abrasion resistance measurements conducted on the coatings are corrected for thickness it can be determined that the abrasion resistance (i.e. indicative of mechanical performance and crosslink density) of the PPG and Sicpa coatings are comparable." (f) The internal score profile of cans with PPG and Sicpa coatings were not the same. Many Sicpa-coated tins showed a distinct "ridge line" in parts of the internal score regions where the coating had been locally compressed and deformed. This contributed to cracking in the internal score channel, in some instances along the apex of the "ridge line". The PPG score profile was a smoother channel without any ridges. (g) The thinness of the Sicpa-coated ends, compared with the PPG and Gold Organosol coated ends, was confirmed by energy dispersive x-ray spectroscopy ("EDAX"), which detected distinct signals for iron and tin in the Sicpa-coated ends compared with barely perceptible signals for those elements from the other two products. (h) The conclusion that corrosion was from the outside in was not affected by the provision to Dr Scheirs, during the course of his work which led to the report, of 15 lids from pre-opened cans, showing corrosion and pitting initiated on the inside. Because Dr Scheirs had no knowledge of the storage of the lids, the effect of any washing, or of any other post treatment and of any storage, he restricted his findings to the unopened cans that had been supplied to him. 33 In November 2009, Visy received a report of an expert named Robert Steele, dated 18 November 2009. This report was a review of Visy's documents and of the report of Dr Scheirs dated 17 March 2006. The Steele report is not in evidence, but is summarised in a brief report of Dr Scheirs dated 9 September 2010. The conclusions of the Steele report were: • The corrosion of the failed cans was initiated from the inside. • The cause (of corrosion) was the failure of the Sicpa lacquer to properly protect the can from the corrosive contents of the can. • There are significant differences between the production lacquer from Sicpa and the production lacquer from PPG. • He [Steele] was unable to form an opinion on the significance of differences between the production lacquer from Sicpa and the sample lacquer from Sicpa. • He [Steele] was unable to identify the resin component, ML98 or DER 669E, which may account for the lower molecular weight of the lacquer but felt it was the DER 669E as it was the dominant constituent present. 34 The Steele report appears to have been the first occasion on which reference was made to the molecular weight of Sicpa lacquer, and to compositional differences between it and PPG lacquer. In his summary report dated 9 September 2010, Dr Scheirs criticised Dr Steele's own reliance on a comparison of HPLC peak heights to conclude that significant compositional differences existed. He said: Steele correctly notes large differences in molecular weights of the epoxy resin precursor used in Sicpa resin (component of the lacquer) compared to the PPG resin; however, I noted that the calculation of the molecular weights from the peaks failed to integrate (i.e. include) all the peaks so the results cannot be relied upon as presented. Dr Scheirs expressed agreement with: the adhesion theory and with the technical argument presented by Steele that the flexibility of the cured lacquer and its ability to withstand the stresses of the score process is affected by the low starting molecular weight of the epoxy resin He said that Steele presented insufficient evidence that this was actually the case in practice. 35 Dr Scheirs then referred to reports of two independent materials laboratories, which he had commissioned to perform tests on PPG, Sicpa and Gold Organosol coating thicknesses, score depth and coating integrity. He summarised the results of these tests as follows: • The EO [ie easy open] end was made of tinplate steel with inner tin weight of 2.8 g/m2 and the body made from TFS (tin free steel). Lacquer adhesion to tinplate is not as good as adhesion to TFS and it is also recommended that for more acidic fill products such as shellfish in vinegar, that tinplate with 11.2 gsm tin coat weight be used with 2 coats of EP lacquer for adequate protection from corrosion (ref: Canmaking -the [sic] technology of metal protection and decoration, T.A. Turner, 1998, pg 67,68). • Organosol coating thickness was twice that of the PPG and Sicpa coatings. • There is evidence of thinning of all three coating types to half their original thickness in the region under the score. • The score depth was between 54 to 62% of the overall end thickness across all three coated ends. Dr Scheirs then said: From these results it can be concluded that the thin initial coat weight of the EP lacquers combined with deep scores results in further deformation and thinning of the coating within the score region. This reduces the effective 'safe operational window' of the lacquer system. The use of low tin coated ends and highly aggressive fill products places further stress on the ability of the EP system to perform adequately within normal conversion parameters and the rigors of thermal sterilization treatment (retort) which is usually 15 minutes at 120 deg.C. It is reported by Visy that Visy had experienced no problem using the PPG product. 36 It appears from Dr Scheirs's report of 9 September 2010 that, during April and May of that year, documents from Nuplex had come to the attention of Dr Scheirs. These documents disclosed for the first time the substitution of Epikote 1009 for DER669E in the manufacture of Nuplex's 3490 EP resin. This information caused Dr Scheirs to review his views as to the cause of the lacquer failure. His new views were expressed as follows: My views have changed as follows. The use of the unauthorized lower MW Epikote 1009 (molecular weight 4000) in place of the stipulated DER669 (molecular weight 5000-8000) reduced the resistance of the coating to cracking under scoring stresses and reduced the level of adhesion of the lacquer to metal under autoclave sterilization. The low MW epoxy gave a coating that was more brittle and with less adhesion to the metal substrate than the PPG resin. The Epikote epoxy was used by Nuplex until March 2004 based on the information that I have seen. Use of Epikote 1009 in place of DER 669E would explain the variability in performance both of the Sicpa product as well as between the Sicpa product and the PPG product. 37 Dr Scheirs performed further work, which resulted in his report dated 3 December 2010. This report confirmed that Dr Scheirs had come to the view that differences between the Sicpa lacquer and the PPG lacquer, as well as variability in performance of the Sicpa product, were explained by the differences in molecular weight between DER669E and Epikote 1009. The report was based on gel permeation chromatography ("GPC") testing of DER669E and Epikote 1009 samples by a consultant. The testing showed differences in molecular weight when measured according to three molecular weight conventions, with DER669E being higher in molecular weight in every case. These results led Dr Scheirs to conclude that the substitution of Epikote 1009 was the root cause of the lacquer failure. He explained his reasoning as follows: The lower molecular weight of the Epikote 1009 epoxy as compared with the prescribed Dow DER 669E has reduced the window of operability of the final lacquer by giving rise to a less flexible final lacquer that was less able to resist cracking under scoring stresses and which exhibited reduced adhesion during autoclave sterilization. The use of a lower molecular weight product in the Sicpa lacquer explains the problem of cracking encountered and the variable performance between PPG and Sicpa products. 38 Mr Jim Haig gave evidence on behalf of Nuplex. He is a materials scientist and analytical chemist. It was accepted that his qualifications and experience entitled him to give evidence of expert opinion for the purposes of this proceeding. Mr Haig's report is dated 8 April 2011. Mr Haig did not do, or commission, any testing of any materials. He reviewed the available documents extensively. In his report, he raised for the first time the possibility that the damage to the can ends that led to the corrosion problem occurred as a result of the failure of Visy to replace worn scoring tools and anvils often enough, so that the scoring became rough on certain can ends. 39 In his critique of the Scheirs report of 3 December 2010, Mr Haig expressed the opinion that the testing of a single sample of each resin to determine molecular weight had no statistical credibility, when the manufacturer of DER669E expressed its molecular weight as a range from 5,000-8,000 Mn, which gave rise to a possible variability of plus or minus 25%. He referred to other literature suggesting figures for the molecular weight of Epikote 1009 that differed from those found in the test commissioned by Dr Scheirs. He referred to research in Japan on Epikote 1009 suggesting that: even a plain blend of epoxy and phenolic resin does not follow the rule of thumb quoted by Scheirs that lower molecular weight reduces epoxy's adhesion, increases brittleness, etc. 40 Mr Haig referred to papers by Kojima and by Kojima and Watanabe ("the Kojima papers"), which he annexed to his report. 41 The conclusions of Mr Haig were expressed as follows: Conclusion 1 144 I have not seen any test document or the like to suggest to me any difference between: 144.1 pre-production sample; and 144.2 production product, of the SICPA lacquer ultimately used by Visy to line the Tuna Cans. Conclusion 2 145 Scheirs' contention in the Scheirs Second Report and the Scheirs Third Report that substitution of Epikote 1009 for DER 669E is substitution with a lower molecular weight resin cannot be and is not substantiated and his consequential inference, that such substitution would (on the basis of a rule of thumb for pure epoxy molecular weight) and did produce an inferior lacquer, therefore fails. Conclusion 3 146 In my experience, if the SICPA lacquer was at fault that the incidence of failure would have been particularly high over a batch of product, which it was not. However, if the source of the problem with the Tuna Cans was another variable (such as Visy production operation in which variables can creep out of specification) I would expect the failure rate to be sporadic (which is what I have observed from the Visy Documents actually happened). Conclusion 4 147 In my opinion, based on my experience I am of the view that the SICPA lacquer: 147.1 was not applied by Visy with sufficient thickness to enable the coating to withstand the rigors of Visy's production line processes which in turn caused the Defective Ends of Tuna Cans; and 147.2 was cracked by the Visy scoring process in the manufacture of the Easy Open Ends and its corrosion barrier properties thus compromised. Conclusion 5 148 It is apparent to me that, while Visy appeared to have a comprehensive QC and monitoring procedure in place on its Easy Open End production line, Visy failed to incorporate sufficient back-up capability into the design of its product in that the coating suffered unacceptable damage when process variables such a [sic] scoring tool quality began to degrade. Conclusion 6 149 Specifically, I am of the opinion that Visy applied the absolute minimum thickness of protective lacquer and that a thicker lacquer would have been more likely to have survived the rigors of the production line. Indeed, the 25% lower thickness seen in samples of lacquers applied during the problem period, and which corresponds to the period in which the Nuplex resin was used by SICPA, possibly accounted for the difference between maintaining a corrosion barrier and not maintaining that barrier. 150 Consequently, my view is that: 150.1 Visy did not have perfect control of its processes for the Tuna Cans; 150.2 the most probable cause of the failure of the Tuna Cans is that: 150.2.1 Visy had a design of its Easy Open Ends that was not robust insofar as that design needed to establish a lacquer thickness able to cope with all expected conditions on the production line; 150.2.2 Visy applied the lacquer to the Tuna Cans at a lower thickness than that was required for its intended use in the Tuna Cans, being an aggressively corrosive product; 150.2.3 Visy cracked the coating during the scoring process, probably during a production period in which the scoring tool and anvil were reaching the end of their useful life. 151 I say this notwithstanding the possibility that at least some failures may have occurred due to corrosion of the external scoring mark (as found by Dr Scheirs in the Conclusion of his first report). Despite Steele's contention that the samples Scheirs examined were old and therefore more susceptible to external score corrosion, Scheirs argues persuasively that the score repair oil would have been removed by the retort processing making the score line once again vulnerable to corrosion, especially when exposed to salt residues as cans frequently are.

The joint expert report 42 In accordance with a direction of the Court, Dr Scheirs and Mr Haig conferred and prepared a joint report dated 22 September 2011. This report was tendered in evidence and formed the basis of the discussion between those two expert witnesses when they gave their evidence concurrently in the course of the trial. The joint report identified nine issues for discussion. There was no agreement between the two experts as to how any of these nine issues should be resolved, although there were concessions made in the discussion of aspects of the issues. The following is intended to be a summary of the way in which the nine issues were discussed in the joint report. 43 The first issue was the general one of establishing a connection between the substitution of Epikote 1009 for DER669E and the corrosion of cans. Mr Haig maintained that there is no technical evidence to link the two, such as any capacity to trace coatings on failed cans back to resin involving the substitution. He pointed out that the only link was the coincidence of the occurrence of can failures during the production run of Sicpa lacquers that involved the substitution. Dr Scheirs certainly relied on the correlation between the time period of the substitution and the recall of cans. He focused on five particular batches of Sicpa lacquer that Visy had identified for investigation, and the link between those batches and the substitution. He also said that he had relied on independent testing, technical literature and Visy's actual experience and usage. 44 The second issue concerned quality control testing. Mr Haig said that all of the Nuplex EP3490 resin passed the quality control criteria specified in the manufacturing instructions. He pointed out that quality control was designed to set a minimum standard. If a product is well above the standard most of the time, then its quality drops but remains above the standard, the product is still compliant. The specification is set by the designer of the product, who determines what is needed for the product to perform satisfactorily through conversion processes and end use. Dr Scheirs responded that it is not the function of quality control tests to determine whether two materials are equivalents. Material equivalency was assumed in such tests. Whilst both materials might meet quality control criteria, that did not mean that they would meet end use requirements, as quality control tests were not conducted after scoring of can ends, filling of cans and retorting. The tests were not designed to assess the ultimate performance of the EP3490 resin. 45 In the third issue, the experts began to focus on the question of molecular weight differences. Mr Haig did not consider the analytical determination of molecular weights relied on by Dr Scheirs as robust, because the determination was only in relation to a single sample of resin. It did not provide information as to variability in molecular weights that might be expected over many batches. He referred to the large range quoted in technical data sheets for epoxy equivalent weights, which he said was indicative of a similarly wide range in molecular weights. Dr Scheirs said that the molecular weights were determined analytically in accordance with accepted methodologies and showed large differences between the epoxies. He conceded there would be some variability between batches, but said that the difference in the measured molecular weights between DER669E and Epikote 1009 was significant and larger than the difference measured between different grades of Epikote (1009 and 1007). Dr Scheirs pointed out that Mr Haig had not undertaken any testing. 46 The next issue was a factual one concerning the number of batches of EP3490 resin in which Nuplex had used Epikote 1009 instead of DER669E, or had used a mixture of the two. Mr Haig's understanding was that there were only 10 batches out of 47 manufactured involving total substitution, and an additional three in which a mixture was used. Dr Scheirs relied on Nuplex records to say that some 24 batches out of 46 involved partial or full substitution. 47 The fifth issue was the crucial one, as to whether variations in molecular weight were linked to coating failure. Mr Haig was "[n]ot convinced" that such a link existed. He believed that the molecular weights of Epikote 1009 and DER669E were sufficiently similar not to have made a difference in performance. He said that final performance (ie the ability to provide a protective barrier) had to take into account many factors, including the final degree of cure, conversion stresses on the coating and coat weight. Mr Haig agreed that the broadness of the molecular weight distribution curve of Epikote 1009 was tighter than that of DER669E. He agreed that as molecular weight of epoxy resin increases, the coating tensile strength of the final product increases. He believed that literature references to coating flexibility increasing with the use of higher molecular weight epoxy resin is a "rule of thumb" based on general usage. In his view, failure is more affected by tensile strength than flexibility and at lower coat weights the tensile strain is sufficiently high to cause cracking failure. Crucially, Mr Haig believed the Kojima papers showed coating performance to increase with increasing epoxy molecular weight, but to reach a plateau of relatively insensitive response and then start to decrease. He said that, unless Epikote 1009 and DER669E were evaluated side by side, it could not be known whether they would perform differently ultimately. 48 Dr Scheirs maintained that lacquer failure was caused by the substitution of a lower molecular weight epoxy resin, giving rise to lower flexibility coating unable to withstand the rigors of scoring and retorting combined with aggressive fill product, when converted at conditions near the edge of the window of operability of the Visy process. The same conversion conditions were being tolerated by lacquer containing DER669E. In the view of Dr Scheirs, the nature of the failure mode (microcracking and microdelaminations leading to lacquer cracking and blisters after retort) were unable to be detected by normal quality control procedures used by Visy. Dr Scheirs said: One would expect the observed sporadic failure rather than wholesale failure of an entire batch of faulty lacquer due to the fact that the coating and scoring processes occur over several production sequences and a single batch of lacquer can be subjected to a wide range of conversion variables and the converted ends used in cans containing a range of fill products. Dr Scheirs agreed with Mr Haig that process factors can affect the final coating performance but believed that those factors were able to be tolerated by DER669E but not by Epikote 1009. In his view, in coping with the window of operability, inherent lacquer properties were more important than process factors. Specifically, higher epoxy resin molecular weight and polydispersity produces lacquers with increased flexibility. The substitution therefore produced a lower flexibility coating which narrowed the window of operability. Dr Scheirs said that his measurement of the molecular weights of DER669E and Epikote 1009 resulted in differences commensurate with the differences published in technical data sheets for different Epikote grades. The same technical data sheets stated the higher molecular weight of starting epoxy resins in epoxy-phenolic coating systems, the better the film flexibility. The statement was unqualified in the literature. Dr Scheirs criticised Mr Haig's reference to the Kojima papers, saying that Mr Haig had misunderstood the Kojima experimental parameters and misinterpreted the results, ignoring Kojima's conclusions relating coating flexibility to epoxy resin molecular weights. Dr Scheirs said that Kojima found that coating films with more flexibility were achieved by the use of epoxy resin with higher molecular weight or lower content of lower molecular weight components or by the control of cure. He accused Mr Haig of misinterpreting technical detail relating to molecular weight between crosslinks within cured epoxy-phenolic coatings achieved by formulating the coatings with different ratios of epoxy to phenolic resins, different epoxy resins and different cure temperatures and times. Dr Scheirs said that, if flexibility dropped as molecular weight increased, formulators would choose cheaper, lower molecular weight epoxy resin alternatives and epoxy grades having higher molecular weights would not be used in can coatings. 49 The sixth issue again concerned the adequacy of testing. Mr Haig said that sufficient numbers of can ends had been tested to include some using lacquer manufactured with Epikote 1009 resin. The test results indicated equivalence with PPG lacquer when tested with challenges equivalent to those found in service, namely light lacquer application rates and highly stressed scoring regions. Taber abrasion testing (reported by Dr Scheirs in his report of March 2006) also showed this equivalence. Mr Haig said that the same report had showed damage in the score region when Sicpa lacquer had been used, but not when PPG lacquer had been used, indicating "poor optimisation of scoring parameters for the Sicpa lacquer in general." Dr Scheirs described Mr Haig's view on this issue as "speculative at best". He said there was no evidence that the substitute epoxy resin had been evaluated for performance as an easy open end coating within Visy's operational parameters, or in any controlled tests for comparison to PPG lacquer or to Sicpa lacquer incorporating DER669E. Dr Scheirs referred to tests that had been conducted to evaluate PPG and Sicpa lacquers in January 2003, with no evidence that substitution had occurred in relation to the latter. The tests determined that Sicpa lacquer was considered acceptable to replace PPG lacquer as a coating on easy open end cans. Dr Scheirs also said that the Taber abrasion testing did not identify whether the Sicpa lacquer tested contained Epikote 1009. He rejected the assumption that the testing could be said to have involved Sicpa lacquer containing Epikote 1009. 50 The seventh issue concerned the sporadic nature of the can failures. Mr Haig said that if resin containing Epikote 1009 suffered an inherent defect resulting in poor performance, it would be expected that most, if not all, production runs of easy open ends using the batches containing Epikote 1009 would have failed in service, rather than the sporadic failures that occurred. He said there was no suggestion by any experts or any independent laboratories of chemical, thermal or other degradation of any of the Sicpa lacquer in service. Dr Scheirs repeated that sporadic failure was what would have been expected, because coating and scoring processes occur over several production sequences and a single batch of lacquer can be subjected to a wide range of conversion variables. The converted ends were then used in cans containing a range of fill products. He said: Coating and conversion process variables could be well within the lacquer's window of operability whilst others are right on the edge and when combined with the impact of inherent reduction of the lacquers' flexibility pushes the coating outside of the designed performance parameters required to withstand the rigors of retort with an aggressive fill product. 51 The eighth issue related to the thickness of the lacquer coats. Mr Haig referred to the reduction in lack of thickness from 10-12 microns to 7-8 microns as applied by Visy when they changed from PPG to Sicpa lacquer. He also said there was independent evidence that all coatings (PPG, Sicpa and Gold Organosol) suffered thinning to half their original thickness under pressure from the scoring anvil. Dr Scheirs agreed that the coat weight was applied to a lower control limit with Sicpa lacquer. PPG coating was thicker than Sicpa coating and the latter underwent more thinning in the scoring process. Dr Scheirs said this was a controlled mode of Visy's operation and passed quality control tests. He rejected the suggestion that thickness reduction for the Sicpa coatings was a cause of the failures. 52 The ninth and final issue focused on the scoring process. Mr Haig said that Visy had chosen the coating thickness for Sicpa lacquer, which did not allow for variations in the scoring process, which had been identified in a number of reports. He said that Visy was operating on the edge and fell over at times, because they allowed scoring tool quality to slip, which overstressed the Sicpa lacquer and caused it to crack. Dr Scheirs described this as "a simplistic conclusion". He said it was based on assumptions that inherent lacquer flexibility was unchanged and that Visy operated a process with tooling that progressively increased the stress on the coating without adequate process control measures to monitor when the safe operational window was about to be breached. Dr Scheirs disagreed with both those assumptions, referring again to his view about lower molecular weight resulting in a coating of lower flexibility, reducing the safe operational processing window of the coating. According to Visy's process control documentation seen by Dr Scheirs, Visy's use of daily statistical process monitoring and control helped them to see when they were about to "fall over" and they halted production pending further quality control inspection. In the opinion of Dr Scheirs, from October 2003, Visy fell over because they did not allow for sporadic epoxy resin substitution in the Sicpa lacquer, which led to reduction in the ability of the coating to resist the formation of microcracks and microdelaminations under the score lines, which were undetectable by quality control tests but became apparent as corrosion initiators after the filled cans were retorted.

The oral evidence of the experts 53 Dr Scheirs and Mr Haig gave their oral evidence concurrently. With few interruptions, they debated in sequence the nine issues the subject of their joint report. When that process had been exhausted, and an opportunity had been provided for each of the witnesses to make any final statement, counsel were able to ask questions of both of the witnesses. The questioning by counsel was directed to the issues. There was no attempt to engage in questioning of the kind designed to discredit or to damage the reputation of either witness. The process of taking expert evidence was a great deal more satisfactory than the traditional method of having the expert witnesses called separately and be examined, cross-examined and re-examined. As well as savings in time and money, the process ensured that the focus was on the substance of the evidence and the resolution of the issues, rather than on tactical considerations. 54 The discussion about the first issue, whether the only connection between epoxy resin substitution and can failure was a coincidental temporal connection, ranged over the nature of testing that had been done by Visy and by Dr Scheirs in the preparation of his earlier reports. There was agreement about a number of issues in relation to the testing history and the significance of the tests that were carried out. The discussion did give Dr Scheirs an opportunity to articulate his case about temporal coincidence very clearly. He said: So we know that the SICPA lacquer was used around October 2003, and we know that some, say, nine months later the perforations were observed in the market place. And based on the pack trials that I've seen, the accelerated ageing trials, that sort of time window is consistent with the aggressiveness of the vinaigrette if the lacquer is compromised. Because it does take time. And it also takes time between when the lacquer is applied to the sheets to when the ends are actually punched out to when the ends and the cans are actually filled. So there's a number of months. So that nine months is consistent, in my mind, that - that timeframe of a relationship between the onset or the introduction of the SICPA lacquer and the emergence of the failures in the market place. But we don't have any stronger links in terms of batch traceability. That's for sure. In the end, the difference between the two expert witnesses on the first issue remained much as it had been. Mr Haig pointed to evidence of failure of both PPG and Sicpa lacquers and to the lack of any opportunity to trace failed cans to batches of resin, and was unable to accept that there had been a sudden change from acceptable performance to unacceptable performance. Dr Scheirs conceded that Visy had applied a thinner coating of Sicpa lacquer than they had used of PPG lacquer, but maintained that the coating was shown to be suitable during testing. He said that there had been some factor that caused the process to fall out of the window of operation, and that factor had later been discovered to be the substitution of an epoxy resin of a different molecular weight. 55 On the second issue, the significance of quality control testing, the witnesses remained at odds. Mr Haig emphasised the consistency of the resins and the lacquer as demonstrated by the quality control tests. Dr Scheirs pointed to the lack of any tests that would predict failure of coated can ends. 56 The third issue concerned the manner in which differences in molecular weight between Epikote 1009 and DER669E had been determined. Mr Haig relied on the technical data sheet for Epikote 1009 and on information he had sourced from the manufacturer of DER669E as to the ranges of molecular weights quoted for the two products, to say that the two products could be regarded as equivalent in terms of molecular weight. He said that the single sample analysis procured by Dr Scheirs was statistically very weak. Dr Scheirs said that the range of molecular weights for DER669E was broad because that resin was manufactured in a number of places. The specification given by Siegwerk to Nuplex required that the epoxy resin must be purchased from Dow Chemicals in Germany, because the German resin was known to have a high molecular weight. Dr Scheirs defended his testing on the basis that two different types of tests on the samples gave numbers consistent with DER669E being of higher molecular weight than Epikote 1009. Even if there were an 80% overlap in the ranges of molecular weights as between the two products, Dr Scheirs maintained that such a degree of overlap could not lead to the products being regarded as equivalents. Dr Scheirs gave a technical explanation as to the reason for the importance of molecular weight. The explanation depends upon the existence in each molecule of the resin of pendant groups called hydroxy groups, which are responsible for crosslinking when the resin is combined with a phenolic resin. The number of pendant hydroxy groups remaining unreacted in the crosslinking process increases the molecular weight of the final product. This improves adhesion of the final product. Further, with higher molecular weights, the distance between crosslinks is greater, which contributes to enhanced flexibility. In the view of Dr Scheirs, even a 20% difference in molecular weight could lead to a 20% difference in adhesion, tensile strength and flexibility. If Visy were operating at the edge of a window, such a 20% change could cause the result to fall out of that window. 57 Mr Haig responded that molecular weight was an important issue, but both DER669E and Epikote 1009 were type 9 resins, which meant they were both of very high quality. A single sample sent for analysis would only produce a single result. A molecular weight figure was not the same for each molecule, but was an average. The 80% overlap of which he spoke was an 80% overlap of probability that the distributions within two samples were going to overlap. Mr Haig would prefer to have seen analyses of five or 10 individual samples from different batches, to see if a pattern emerged. Mr Haig repeated that, on the technical data available, there was reason to believe that there was equivalence and a single result was difficult to accept at face value. 58 Dr Scheirs pointed out that he had also tested a sample of Epikote 1007 and found the difference between Epikote 1007 and Epikote 1009 to be similar to the difference between Epikote 1009 and DER669E. Mr Haig said that such a result would be quite disturbing and questioned the accuracy of it. He also said that one of the results of testing DER669E was outside the published range of molecular weights for that product. Dr Scheirs said the test data within one particular sample was reliable. 59 There was no additional discussion about the fourth issue, the extent to which substitution of epoxy resins occurred. 60 The fifth issue again related to molecular weight. Some time was spent dealing with the standard that Mr Haig had applied when saying that he was "[n]ot convinced" that differences in molecular weight affected performance. 61 It was in the discussion of the fifth issue that differences occurred between Mr Haig and Dr Scheirs about how the Kojima papers should be interpreted in one respect. To some extent, the differences were over the meanings of technical terms used in the Kojima papers. In one respect, however, there remains a dispute between Dr Scheirs and Mr Haig as to the results of the experiments reported in the Kojima papers. Dr Scheirs maintained that those results demonstrated that flexibility increased with higher molecular weights of the epoxy resin used to create an epoxy-phenolic resin. Mr Haig maintained that, in the range of molecular weights of type 9 epoxy resins, the results showed that there was a lack of sensitivity to variations in molecular weights. 62 The sixth issue concerned the testing that was in fact conducted in relation to can ends. Mr Haig referred to testing that had revealed that both PPG and Sicpa lacquer (both using DER669E epoxy resin) were cracking. He also referred to Taber abrasion testing that Dr Scheirs had done. In the course of discussing this issue, Dr Scheirs described in some detail the coating process used by Visy in relation to can ends. Each end was coated twice, with curing by means of heating in a stove taking place after each coating. His point was that different batches of lacquer might well be used for the two coats, making it extra difficult to do any tracing of particular batches of lacquer into particular can ends. Dr Scheirs dismissed the significance of Taber abrasion tests to the present case, on the basis that they were purely designed to test hardness and abrasion resistance, not flexibility. 63 In relation to cracking, Dr Scheirs said that electron microscope photos revealed that both PPG lacquer and Sicpa lacquer cracked underneath the score, but the fissures in the Sicpa lacquer cracks were deeper than those in the PPG lacquer. He also said that cracking did not necessarily mean corrosion. There was substantial discussion about the efficacy of quality control, particularly in relation to the wearing of scoring equipment. Dr Scheirs placed weight on the report of an expert who was brought from the USA at an early stage of investigating the corrosion of cans, specifically for the purpose of investigating the scoring process. 64 The seventh issue concerned the sporadic nature of the can end failures. Mr Haig would have expected a greater number of failures from an inherent problem with the lacquer, whereas Dr Scheirs argued that breakdown of lacquer using epoxy resin of a lower molecular weight as a result of the scoring process was only more likely but not inevitable. 65 The eighth issue concerns the difference in thickness of coatings between PPG lacquer and Sicpa lacquer. Mr Haig saw this as significant. Dr Scheirs said it was still within the control limits adopted by Visy. The difference between the two experts seem to be largely as to the efficacy of Visy's quality control in detecting wearing of the scoring equipment and consequent imperfections in the scored can ends. 66 Both experts thought that the ninth issue, which again raised the effect of the quality of scoring tools, had been covered effectively in earlier discussion. 67 Nothing of great significance came from the questioning of the expert witnesses by counsel.

Causation 68 In approaching the task of determining whether there is a causal link between Nuplex's substitution of Epikote 1009 for DER669E in the epoxy-phenolic resin it supplied to Siegwerk and the corrosion of can ends manufactured by Visy using lacquer containing the substitute resin, it is necessary to bear in mind that the usual rule of civil litigation applies. It is for the party making the claim (in this case, Siegwerk as cross-claimant against Nuplex) to establish on the balance of probabilities that the cause is as alleged. In other words, the evidence must demonstrate that it is more probable than not that the substitution, whether or not in conjunction with other causative elements, led to the damage. 69 The issue of causation in the present case turns upon the evidence of the two experts, Dr Scheirs and Mr Haig. The thesis of Dr Scheirs is that Epikote 1009 had a lower molecular weight than DER669E; the use of an epoxy resin with a lower molecular weight resulted in a lacquer that was less flexible when applied to can ends and cured; less flexibility resulted in a greater likelihood of the lacquer coating being damaged in the scoring process applied to the can ends before they were affixed to the filled cans; and, as a result, the contents of the occasional can were able to penetrate the coating in a damaged area and corrode the metal of the can end. Mr Haig was critical of this reasoning in a number of respects, but pointed also to the alternative possibilities that the thinness of the lacquer coating on the can ends (thinner than had been the case with the earlier use of PPG lacquer), coupled with the propensity of the lacquer to crack under the scoring process and the changes in the results of that process when the scoring equipment became worn, could have produced the breakdown in the lacquer coating in the occasional can. 70 The task of the Court is "not concerned with a choice among rival conjectures." See Jones v Dunkel (1959) 101 CLR 298 at 304 per Dixon CJ. His Honour was dissenting as to the result of the case, but there can be no doubt as to the correctness of the principle that his Honour stated. The same principle was discussed at some length in the speech of Lord Brandon of Oakbrook, with whom Lords Fraser of Tullybelton, Diplock, Roskill and Templeman concurred in Rhesa Shipping Co SA v Edmunds [1985] 1 WLR 948. In that case, the issue was as to the cause of the sudden sinking of a ship. In particular, the question was whether the cause was "perils of the sea", for the purposes of an insurance claim. The ship sank while sailing in deep water in a well-used shipping lane, in good weather, when an aperture suddenly appeared in its hull, through which water entered. On the basis of expert evidence, the primary judge had found that the sinking was due to a collision with a submerged submarine, travelling in the same direction as the ship and about the same speed. The primary judge thought such an event in itself highly improbable, but found on the balance of probabilities that it was the cause, because he regarded himself as having to choose between the expert opinion as to the submarine and the alternative theory that the aperture was due to deterioration of the ship's hull as a result of poor maintenance. The House of Lords held that it was wrong for the primary judge to have ignored the third possible solution, that the cause of the aperture was simply unknown and the ship owners' claims should be dismissed on the basis that they had not discharged the burden of proof that lay on them. Early in his speech in Rhesa Shipping, at 951, Lord Brandon of Oakbrook stated two principles: In approaching this question it is important that two matters should be borne constantly in mind. The first matter is that the burden of proving, on a balance of probabilities, that the ship was lost by perils of the sea, is and remains throughout on the shipowners. Although it is open to underwriters to suggest and seek to prove some other cause of loss, against which the ship was not insured, there is no obligation on them to do so. Moreover, if they chose to do so, there is no obligation on them to prove, even on a balance of probabilities, the truth of their alternative case. The second matter is that it is always open to a court, even after the kind of prolonged inquiry with a mass of expert evidence which took place in this case, to conclude, at the end of the day, that the proximate cause of the ship's loss, even on a balance of probabilities, remains in doubt, with the consequence that the shipowners have failed to discharge the burden of proof which lay upon them. 71 In the present case, there can be no doubt that the thesis of Dr Scheirs provides a plausible theory as to the cause of the damage. It would be easy enough to say that it is at least as plausible as the theories offered by Mr Haig. The choice is not between competing theories, however. In circumstances in which the substitution of one epoxy resin for the other is the only possible ground on which Nuplex could be liable to Siegwerk, the question is whether there is sufficient evidence to support the theory of Dr Scheirs and to make that theory an explanation that is more probable than not to be the explanation of the cause. 72 The fact that there is a competing hypothesis, advanced by Mr Haig, should not be permitted to convert the case into a competition between two hypotheses as to which has more plausibility. Such an approach would be even more dangerous in a case such as the present, in which far more attention has been given to investigating the theory of Dr Scheirs than to any alternative. The impact of a competing hypothesis is merely to make it plain that the one advanced by Dr Scheirs does not have the advantage of being the only theory advanced. 73 The first assumption on which Dr Schier's theory is based is that there was in fact a difference in molecular weight between the Epikote 1009 used by Nuplex in substitution for the DER669E that Nuplex was bound by its contract with Siegwerk to use. The validity of this assumption is very difficult to establish. In the first place, the use by the respective manufacturers of the two epoxy resins of ranges in the technical data they publish about molecular weight has the result that the only comparison possible on the basis of that technical data is a comparison of the ranges, rather than a direct comparison of the molecular weights of the particular inputs into resin manufactured by Nuplex. Dr Scheirs said that this difficulty was overcome by the analysis he had carried out of a sample of each of the two epoxy resins, which revealed significant differences in molecular weight. I accept Mr Haig's criticism of this proposition. Given that molecular weights are expressed in ranges, analysis of a single sample of each epoxy resin is unlikely to be of much significance in determining whether there were actual differences in molecular weights. This is not to say that I agree with Mr Haig's statement that he would need to be "convinced" of the difference in molecular weight. It is clear from his evidence that Mr Haig was applying a standard well above that of the balance of probabilities in saying that he was "[n]ot convinced". It is true, however, that separate comparative analyses of a series of samples from different batches of the two epoxy resins could have provided a firmer basis on which it could have been determined with any confidence whether there was a difference in molecular weights between the two resins used by Nuplex. The evidence as it is falls short of validating the first assumption. 74 The second assumption made by Dr Scheirs is that a lower molecular weight of the epoxy resin used results in a less flexible coating on a can end, after the epoxy resin has been mixed with a phenolic resin, the resulting epoxy-phenolic resin has been mixed with other ingredients to make a lacquer, the lacquer has been applied to can ends in two coats and each of those coats has been separately cured by heating. Dr Scheirs says that this assumption is supported by the Kojima papers, which show the results of experimental work demonstrating the correlation between higher molecular weight of epoxy resin and greater flexibility of final coating. There can be no doubt that the contents of the Kojima papers do support the existence of such a correlation up to a point. Mr Haig's criticism is that Grade 9 epoxy resins (which include Epikote 1009 and DER669E) appear to be beyond that point. In other words, the higher the grade of epoxy resin, the more likely it is that there will be a plateau effect or (Mr Haig says) even a downturn in the flexibility of the finished product. Grade 9 epoxy resins are very high grade products. I do not accept that the assumption made by Dr Scheirs that any lowering of molecular weight involved in the substitution of Epikote 1009 for DER669E resulted in less flexible lacquer coating on can ends is borne out. The assumption might well be correct, but further evidence would be needed to validate it. 75 Without any validation of these two assumptions, the theory of Dr Scheirs remains a plausible theory, but amounts to nothing more. It cannot establish on the balance of probabilities that the substitution of one epoxy resin for the other was a causal factor in the damaged cans of tuna. There remains the possibility that the damage was caused by Visy's decision to use a thinner coating of Sicpa lacquer than had been used when PPG lacquer was applied. There remains the possibility that occasional excessive wear of the scoring tools, whether in conjunction with the thinner coat of lacquer or by itself, was the cause. There even remains a possibility, which has never been investigated, that the composition of the lacquer was affected by the change of the aluminium-based pigment added to the lacquer, in substitution for the pigment used in PPG lacquer, in order to avoid disputes about infringement of intellectual property rights (see [13] above). So far as the evidence discloses, there has been no investigation at all of any effect that this change might have had in the composition of the lacquer. Even if the theory of Dr Scheirs remains the most plausible of all the theories, Siegwerk is not entitled to succeed against Nuplex on that basis. 76 The result is that there is insufficient evidence to establish any cause of the damage to the cans of tuna, out of which Visy's claim against Siegwerk arose. More importantly, the theory of Dr Scheirs that the causal chain can be traced to a difference in molecular weights between Epikote 1009 and DER669E has not been established to be a probable cause, in the sense of being more probable than not. The cross-claim of Siegwerk against Nuplex must be dismissed.

The insurance claim 77 The failure of Siegwerk's claim against Nuplex does not bring to an end the issues in the proceeding. Nuplex's cross-claim against QBE seeks not only indemnity under the broadform liability policy in respect of any liability Nuplex may have to Siegwerk (including interest and costs), but also indemnity in respect of all costs and other expenses incurred by Nuplex in the conduct of its defence of the allegations contained in the amended cross-claim. This claim is based on a term in the policy that would require QBE to defend in Nuplex's name and on Nuplex's behalf any claim, even a failed claim. As a consequence, it is necessary to go to the terms of the policy to see if Siegwerk's claim against Nuplex falls within those terms, even though the claim has failed.

The terms of the policy 78 The broadform liability policy is in a standard form. It is of the kind often described as written in "plain English". This means that it is necessary for the reader to overcome the annoying practice of using second person pronouns to refer to the insured (Nuplex). The insurer is described as "QBE". The "plain English" nature of the policy leaves the meaning of many of its terms obscure, so that there was substantial argument about each of them in the course of the trial. 79 Clause 2 of the policy, headed "COVERAGE" specifies the circumstances in which QBE will indemnify Nuplex. For present purposes, the relevant clauses are as follows: 2.1 Liability QBE will pay in respect of Personal Injury or Property Damage first happening during the Period of Insurance and caused by an Occurrence within the Territorial Limits in connection with Your Business: 2.1.1 all sums which You become legally liable to pay by way of compensation; 2.1.2 all costs awarded against You. 2.2 Defence of Claims With respect to the indemnity provided by this Policy QBE will: 2.2.1 defend in Your name and on Your behalf any claim or legal action against You seeking damages on account of Personal Injury or Property Damage even if the action is groundless, false or fraudulent and QBE will investigate, negotiate and settle any claim or legal action as QBE sees fit; 2.2.2 pay all legal costs and expenses incurred by QBE and all interest accruing after entry of judgment until QBE have paid, tendered or deposited in court such part of the judgment as does not exceed the Limit of Liability; 2.2.3 reimburse You for all reasonable expenses, other than loss of earnings, incurred with QBE's consent, in the defence of a claim or legal action against You seeking damages on account of Personal Injury or Property Damage; Clause 3 of the policy is headed "EXCLUSIONS" and provides relevantly: This Policy does not cover liability in respect of: ... 3.5 Contractual Liability any obligation assumed by You under any agreement or contract except to the extent that: 3.5.1 the liability would have been implied by law; 3.5.2 the liability arises from a provision in a contract for lease of real or personal property other than a provision which obliges You to effect insurance or provide indemnity in respect of the subject matter of that contract; 3.5.3 the liability is assumed by You under Your warranty of fitness or quality as regards to Your Products; ... 3.13 Loss of Use loss of use of tangible property which has not been physically damaged, or physically lost or physically destroyed resulting from: 3.13.1 a delay in or lack of performance by or on Your behalf of any agreement; 3.13.2 the failure of Your Products to meet the level of performance, quality, fitness or durability expressed or implied, warranted or represented by You, but this Exclusion does not apply to the loss of use of other tangible property resulting from the sudden and accidental physical damage to, or physical loss or physical destruction of Your Products after they have been put to use by any person or organisation other than You. ... 3.17 Product Recall Claims arising out of or resulting from any loss, cost or expense incurred by You for the loss of use, withdrawal, recall, inspection, repair, replacement, adjustment, removal or disposal of Your Products or of any property of which they form a part, if such Products or property are withdrawn from the market or from use because of any known or suspected defect, deficiency, inadequacy or dangerous condition in them. 80 The policy also contains a schedule, specifying a number of matters, including the period of insurance, the limit of liability, the deductibles ($500,000 for each and every Occurrence (inclusive of costs and expenses)) and endorsements to the policy. Endorsement 3 adds to the policy an additional exclusion, in the following terms: This Policy does not cover liability in respect of Personal Injury or Property Damage resulting from the failure of Your Products, or work completed by You or for You, to perform the function or serve the purpose intended by You but this Exclusion does not apply to Personal Injury or Property Damage resulting from the Active Malfunctioning of such products or work. For the purposes of this Exclusion, "Active Malfunctioning" means the failure to function in its normal manner (for which it was designed) and such failure was active. 81 It is also necessary to have regard to some definitions in cl 1 of the policy. So far as relevant, cl 1 provides: Any word or expression which this Policy defines as having a particular meaning will have the meaning everywhere it appears. ... 1.4 "Deductible" means the amount You first bear in relation to each Occurrence. The Deductible will apply to any costs and expenses incurred in the defence or investigation of claims under Coverage Clause 2.2 first, then to amounts (if any), which are subject to indemnity under the Coverage Clause 2.1. ... 1.11 "Occurrence" means an event which results in Personal Injury or Property Damage, neither of which is expected nor intended from any person's standpoint. All Personal Injury or Property Damage arising out of continuous or repeated exposure to substantially the same general conditions shall be construed as arising out of one Occurrence. ... 1.17 "Property Damage" means: 1.17.1 physical damage to or physical loss or physical destruction of tangible property including any resulting loss of use of that property; or 1.17.2 loss of use of tangible property which has not been physically damaged, lost or destroyed provided such loss of use is caused by physical damage to or physical loss or physical destruction of other tangible property. ... 1.24 "Your Products" means any goods, products and property (after they have ceased to be in Your possession or under Your control), which are or is deemed to have been manufactured, grown, extracted, produced, processed, assembled, constructed, erected, installed, repaired, serviced, treated, sold, supplied or distributed by You (including any container thereof other than a Vehicle).

The coverage clause 82 Counsel for QBE raised two issues in respect of the coverage clause in the broadform liability policy. The first was whether there had been an "Occurrence", as defined that had caused the loss claimed. The second issue was whether the loss claimed by Siegwerk against Nuplex was entirely economic loss, and not compensation for "Property Damage" as defined. 83 Authorities dealing with insurance policies containing the same or similar wording make it clear that the event constituting the occurrence is something different from the suffering of the property damage to which the coverage clause attaches. The occurrence is the mishap that causes the damage. See Distillers Company Bio-Chemicals (Australia) Pty Ltd v Ajax Insurance Co Ltd (1974) 130 CLR 1 at 18-19 per Stephen J, GIO General Ltd v Newcastle City Council (1996) 38 NSWLR 558 at 567 per Kirby P (with whom Powell JA agreed), Windsurf Pty Ltd v HIH Casualty & General Insurance Ltd [1999] QCA 360 at [7]-[13] per de Jersey CJ (with whom White and Muir JJ agreed) and Selected Seeds Pty Ltd v QBEMM Pty Ltd [2009] QSC 70 at [20], upheld by the Queensland Court of Appeal in Selected Seeds Pty Ltd v QBEMM Pty Ltd [2009] QCA 286 at [16] per Fraser JA (with whom Holmes JA and White J agreed) (the issue was not dealt with by the High Court in Selected Seeds Pty Ltd v QBEMM Pty Ltd [2010] HCA 37). It is also clear that, in a case in which there is a supply of a product, the relevant occurrence is not that supply itself. For instance, in Windsurf, the relevant occurrence was not the negligent laying of carpet on stairs, it was the subsequent shifting of that carpet, which caused a person walking on it to fall and suffer injury. 84 In the present case, the property damage from which the loss was claimed to follow was the corrosion of cans, resulting from the contents coming into contact with the metal of the can ends, with the consequent corrosion of the cans and deterioration of the products within them. Between Nuplex supplying to Siegwerk resin containing Epikote 1009 and the corrosion of cans there lay any number of events, each of which could amount to a relevant "Occurrence", for the purposes of the coverage clause of the broadform liability policy. They included Siegwerk mixing that resin with other ingredients to make lacquer, Siegwerk supplying that lacquer to Visy, Visy applying that lacquer to can ends, Visy scoring the can ends to which the lacquer had been applied, Visy supplying the can ends with cans to fish canners, and fish canners using those can ends to seal cans in which they had placed their fish products. Each of these events lies along a chain of causation, alleged to lead from the supply by Nuplex of resin containing Epikote 1009 to the corrosion of some cans of fish. If Siegwerk's claim against Nuplex had been successful, it would have been because of that chain of causation and the events that were part of it. It follows that any of those events can constitute an "Occurrence" as defined in cl 1.11 of the policy. Of course, each of those events was itself more than one event. Siegwerk used resin from a number of different batches containing Epikote 1009 to make lacquer. It sold a number of different batches of lacquer to Visy. Visy applied those different batches of lacquer to many different can ends, which it then scored and supplied to the fish canners. The fish canners filled a number of different cans with fish and sealed them, using the can ends. In each case, however, these events amounted to "continuous or repeated exposure to substantially the same general conditions", within the meaning of the second sentence of cl 1.11, so they are to be construed as a single occurrence. 85 It is of no consequence that many of these events in the chain of causation did, or might have, fallen outside the period of insurance covered by the broadform liability policy (which began on 1 April 2004). It is not the occurrence that must happen within that period, in order to give rise to liability under the policy. By cl 2.1, it is the property damage that must first happen during that period. By reason of the chain of causation, it is also apparent that whatever is chosen as the relevant occurrence was in connection with "Your Business", ie the business of Nuplex. The connection is supplied by the allegation that it was Nuplex's supply of resin with the substituted Epikote 1009 that began the chain of causation. 86 Counsel for QBE also contended that there had been no "Property Damage", but only pure economic loss resulting from the recall and withdrawal from sale of large numbers of cans of fish. Clearly, this argument cannot be accepted. The definition of "Property Damage" in cl 1.17 of the policy is in two parts. The first refers to physical damage to tangible property. The corrosion of some cans containing fish, and the consequent deterioration of the contents of those cans clearly amount to physical damage to tangible property. The second part of the definition concerns loss of use of tangible property which has not been physically damaged, lost or destroyed, where the loss of use is caused by physical damage to other tangible property. The undamaged cans of fish recalled and withdrawn from sale (the use of which was thereby lost) were caused to be recalled and withdrawn from sale by the physical damage to the cans that were damaged and their contents. The claim by Siegwerk against Nuplex is therefore one "in respect of" property damage, within the meaning of cl 2.1 of the policy, whichever of the two limbs of the definition of "Property Damage" is applied. There was physical damage to some property, being the damage to cans and their contents. There was loss of use of other property, consequent upon that physical damage. In either case, any liability Nuplex might have incurred to pay compensation to Siegwerk, if the claim against Nuplex had succeeded, would have been a liability "in respect of…Property Damage" within clause 2.1 of the policy. Either the whole claim was "in respect of", ie causally related to the physical damage, or the claim was partly in respect of loss of use of undamaged property consequent upon damage to other property. 87 The property damage unquestionably first occurred during the period of insurance. It follows that, if Siegwerk had succeeded on its cross-claim against Nuplex, QBE would have been liable (subject to its right to deduct $500,000) to indemnify Nuplex for the amount claimed by Siegwerk and all costs awarded against Nuplex. In those circumstances, cl 2.2.1 of the policy required QBE to defend the proceeding in Nuplex's name and on Nuplex's behalf. Subject to the application of any of the exclusion clauses upon which QBE relied, QBE was liable to defend Siegwerk's claim against Nuplex, even though that claim was unsuccessful. 88 Although, in its defence to Nuplex's cross-claim, QBE pleaded reliance on the contractual liability exception in cl 3.5 of the policy, at the trial QBE conceded that the proviso to that clause precluded the application of the exclusion. Either any liability of Nuplex to Siegwerk would have been implied by law, within cl 3.5.1 of the policy, or the liability would have been assumed by Nuplex under a warranty of fitness or quality as regards Nuplex's product. 89 The loss of use exclusion in cl 3.13 of the policy requires closer examination. There is no question that the claim made by Visy against Siegwerk was predominantly (although not entirely) a claim in respect of loss of use of tangible property which had not been physically damaged, lost or destroyed. By far the largest element of Visy's claim was in respect of amounts it had paid to PLTP and Simplot in respect of "loss of stock on hand". The figures are shown in [26] above. Because the settlement between Visy and Siegwerk was for a lump sum, there has been no specification of the form or forms of loss for which Siegwerk was compensating Visy. The total of the amounts claimed by Visy other than "loss of stock on hand" would be more than enough to cover the entire sum of $2,250,000 paid by Siegwerk to Visy in settlement of the claim, which is the amount of the claim against Nuplex. This fact places QBE in some difficulty when it attempts to bring the claim within the opening words of the exclusion in cl 3.13. 90 Even if it can be assumed that the claim against Nuplex is a claim in respect of loss of use of tangible property not itself damaged, lost or destroyed, the claim would not fall within cl 3.13.1. It was not a claim that resulted from any delay in or lack of performance by Nuplex of any agreement. There is no issue at all of delay. There was performance of the toll manufacturing agreement between Nuplex and Siegwerk, in the sense that there was a supply of epoxy-phenolic resin each time an order for a further batch was placed. In respect of some batches, the performance involved the substitution of one epoxy resin for another, so that the batch did not conform to the agreed specifications, but this was not a lack of performance. 91 Any loss of use of tangible property which had not been physically damaged, lost or destroyed did result from the failure of Nuplex's products to meet the level of performance, quality fitness or durability expressed or implied, warranted or represented by Nuplex, within the meaning of cl 3.13.2 of the policy. It is therefore necessary to look at the proviso to the exclusion. The question is whether the loss of use of other tangible property resulted from "the sudden and accidental physical damage" of Nuplex's resin, after it had been put to use by any person or organisation other than Nuplex. Plainly, the resin had been put to use by Siegwerk, by being mixed with other ingredients to make a lacquer. It had also been put to use by Visy, which applied two coats of the lacquer to can ends, cured each coat, and then subjected the can ends to the scoring process. Nuplex's resin was also put to use by the fish canners to whom Visy supplied the can ends, when they used those can ends to seal the cans after they had been filled with fish and other products. The question is whether, as the claim has been put, there was "sudden and accidental physical damage" to the resin. The claim is based on the proposition that the lower molecular weight resulting from the substitution of epoxy resin in Nuplex's epoxy-phenolic resin gave rise to a lacquer which had less flexibility than that required, and was therefore less able to withstand the rigors of the scoring process carried out by Visy. It was the damage to the lacquer, which included Nuplex's product, that caused the contents of some cans to come into contact with the metal of the cans. The question is whether that damage can properly be described as "sudden and accidental". 92 It is clear that these words are intended to have their ordinary meanings when used in the policy. They are not technical terms. The Macquarie Dictionary provides two alternative meanings of the word "sudden", being "happening, coming, made, or done quickly, without warning or unexpectedly" or "sharp; abrupt". In construing the meaning of the word "sudden" in an insurance policy in Sun Alliance & London Insurance Group v North West Iron Co Ltd (1974) 2 NSWLR 625 at 631-633, Sheppard J adopted the "unforeseen and unexpected" meaning of the word. The Macquarie Dictionary defines "accidental" as "happening by chance or accident, or unexpectedly". Again, in an insurance context, the High Court in Australian Casualty Co Ltd v Federico (1986) 160 CLR 513 adopted the meaning of "accident" involving its unintended and unforeseen elements. At 527, Wilson, Deane and Dawson JJ said: As a matter of ordinary language in this country, an "accident" (from the Latin accidens) means very much what the etymologist would expect. It is something which happens without intention or design. When used with reference to something which causes injury, it means an unexpected and unintended mishap. 93 The focus on the unexpected element of both "sudden" and "accidental" means that there is considerable overlap between the two words. It is not surprising in a document like an insurance policy to find an element of tautology in expressions chosen. "[S]udden and accidental" is a composite phrase, and must be interpreted as such. In the present case, the damage to Nuplex's resin, after it had been used by others, was undoubtedly sudden and accidental. It was unintended and unexpected. As the claim is put, it was the sudden and accidental physical damage to the resin, as a component of the lacquer, that caused the loss of use of the tangible property being the undamaged cans that were recalled and withdrawn from sale. The claim therefore falls within the proviso in cl 3.13.2 of the policy, and consequently is not excluded by cl 3.13. 94 Counsel for QBE argued that the product recall exclusion in cl 3.17 of the policy operated to exclude the whole of the claim, because the vast majority of the cans of fish of which there was a loss of use had been withdrawn from the market because of a suspected defect. In my view, this argument is based on a false assumption. Clause 3.17 does not exclude claims for loss of use of Nuplex's products or any property of which they form a part. The clause excludes claims "arising out of or resulting from any loss, cost or expense incurred by [Nuplex] for the loss of use" etc of the products. The exclusion is directed to the expenses of the process of withdrawal from the market or from use. Thus, a claim for loss, cost or expense of acquiring a temporary substitute for a product, during a period of loss of use would be excluded. Likewise, it is the loss, cost or expense involved in withdrawing from the market, recalling, inspecting, repairing, replacing, adjusting, removing or disposing of property that is covered, not the loss of the value of the property itself. Reading the exclusion in cl 3.17 in this way makes sense of the clause itself, as the phrase "loss of use" is used in conjunction with the word "withdrawal" and the other words that follow, indicating that the target of the exclusion is not the loss of use of the property itself, but the cost and expense associated with the loss of use by reason of withdrawal from the market or from use of the item in question. Further, reading the clause in this way makes it consistent with cl 2.1 of the policy and the definition of "Property Damage", particularly that part of it in cl 1.17.2. To regard cl 3.17 as excluding any claim for loss of use of products withdrawn from the market because of a suspected defect, suspected because of physical damage to other property, would amount to a substantial negation of the intention of the policy. 95 In any event, the attempt to rely on cl 3.17 encounters the same difficulty as beset QBE in relation to the attempt to rely on cl 3.13 in relation to loss of use. Parts of Visy's original claim related to loss of stock. Other parts could be said to fall within cl 3.17, particularly items like "labour costs in replacing stock", "pallet hire costs wasted", "costs of disposal of condemned stock", "finance costs on condemned stock" and "costs of recalling stock". Because of the nature of the settlement between Visy and Siegwerk, QBE cannot identify to which items the sum claimed by Siegwerk against Nuplex relates. QBE cannot therefore bring the claim within the exclusion in cl 3.17. It cannot apply the principle derived from Wayne Tank & Pump Co Ltd v Employers Liability Assurance Corporation Ltd [1974] 1 QB 57. That principle relates to insurance claims in which the relevant insurance policy excludes indemnity in respect of loss from a particular cause. If there are two causes, one of which falls within the exclusion and the other outside it, the principle is that the exclusion applies. In the present case, Siegwerk's claim is brought in respect of a single cause and that cause is not the subject of the exclusion in cl 3.17. The original claim by Visy involved more than one head of damage. The principle in Wayne Tank does not apply to exclude an entire claim because it involves more than one head of damage, some falling within and some falling outside an exclusion clause in the relevant insurance policy. The undifferentiated nature of the settlement sum claimed by Siegwerk deprives QBE of the opportunity of arguing that the claim falls within the exclusion in cl 3.17. 96 The final exclusion relied on is found in Endorsement 3 to the policy. It is a difficult provision to construe. There is no doubt that the claim in the present case is in respect of the alleged failure of Nuplex's products to perform the function or serve the purpose intended by Nuplex. The intended function or purpose of Nuplex's resin was to form part of a lacquer that would prevent the contents of the cans coming into contact with the metal of the can ends. The claim is that the lacquer containing the resin failed to perform this function. The difficulty is in construing the proviso to Endorsement 3, that the exclusion does not apply to property damage resulting from "the Active Malfunctioning" of the products. The phrase "Active Malfunctioning" would be difficult to construe by itself. It is made even more difficult by the definition of that phrase as meaning "the failure to function in its normal manner (for which it was designed) and such failure was active." The ordinary meaning of "failure" connotes a passive event. The Macquarie Dictionary refers to "proving unsuccessful", "lack of success", "non-performance of something due or required", "running short", or "insufficiency". To couple the word "failure" with the word "active" invites a question as to the intended meaning of the word "active". Of the collection of meanings of "active" in the Macquarie Dictionary, the only one that appears to give sense to the notion of an active failure is "causing change; capable of exerting influence". Thus, a failure that brings about change appears to be regarded as an active failure, and therefore as "Active Malfunctioning". 97 Applying that meaning to the facts of the present case, it is clear that Siegwerk's claim depends upon the allegation that Nuplex's resin failed to function in the normal manner, for which it was designed, and that, in doing so, it brought about change in the form of corrosion of the cans. It follows that the proviso to Endorsement 3 of the policy prevents the exclusion in that endorsement from applying. 98 For the foregoing reasons, Siegwerk's claim against Nuplex falls within the coverage clause of Nuplex's broadform liability policy with QBE and is not excluded from that coverage by any of the exclusions on which QBE relied. Even though Siegwerk's claim has failed, cl 2.2 of the policy required QBE to defend in Nuplex's name the proceeding. QBE is obliged to bear the costs of that defence.

The appropriate orders 99 The cross-claim of Siegwerk against Nuplex must be dismissed. Applying the usual principle, that costs follow the event, there should be an order that Siegwerk pay Nuplex's costs of that cross-claim. In accordance with the undertaking by way of security for costs referred to in [5] above, Zurich should be ordered to pay the amount agreed or assessed to be the costs ordered against Siegwerk, up to the limit of $750,000. 100 There should also be a declaration, substantially to the effect of cl 2.2.1 of the broadform liability policy, as to the obligation of QBE to defend the proceeding in Nuplex's name and on Nuplex's behalf. It follows from this obligation that, subject to two deductions, QBE should pay all Nuplex's legal costs and expenses of defending Siegwerk's cross-claim. The first exception is the sum of $500,000, described in the schedule to the policy as a "Deductible" in respect of "each and every Occurrence (inclusive of costs and expenses)". The second exception is the amount of any costs recovered by Nuplex from Siegwerk, whether by calling upon the security referred to in [99] above or otherwise. The obligation to defend should be satisfied by an order that, subject to the two exceptions, QBE pay Nuplex's costs of Siegwerk's cross-claim on an indemnity basis. 101 To enforce QBE's obligation to defend Siegwerk's cross-claim in Nuplex's name and on Nuplex's behalf, it was necessary for Nuplex to cross-claim against QBE. Nuplex has succeeded on its cross-claim and the costs of that cross-claim should also follow the event. Further, QBE should be ordered to pay Nuplex's costs of its cross-claim on an indemnity basis. Such an order, together with the orders referred to in [100] above, is necessary to place Nuplex in the position it would have been in, if QBE had honoured its obligation under the policy. By reason of r 40.01 of the Federal Court Rules 2011, an order simply for costs would result in those costs being assessed on a party and party basis, which would result in Nuplex suffering a loss it would not have suffered if QBE had honoured its obligation. 102 In view of QBE's obligation pursuant to cl 2.2 of the policy, it appears pointless to cast upon Nuplex and its legal representatives the burden of agreeing upon, or undergoing the process of assessment of, the costs that Siegwerk is obliged to pay to Nuplex. QBE and its legal advisers should undertake that burden. In conjunction with that obligation, QBE should be entitled to the benefit of the security. Orders to this effect will also be made. 103 Liberty to apply will be reserved, in case any difficulty should arise in relation to these matters. I certify that the preceding one hundred and three (103) numbered paragraphs are a true copy of the reasons for judgment herein of the Honourable Justice Gray.

Associate: Dated: 19 March 2013 Schedule 1 No. SICPA 3490 EP resin batch number DER669E or Epikote 1009 Approximate latest likely order date Date of manufacture by Nuplex Date on pre-shipment sample testing document Date SICPA 3490 EP resin batch received by SICPA Date SICPA 3490 EP resin batch received by SICPA (two weeks before date of manufacture) (Summary 1 - created by Steve Browning) (Nuplex Batch & Activity Testing records) (SICPA pre shipment sample test records) (Nuplex Tax invoice) (Nuplex Delivery docket) 304003 DER669E 2 April 2003 Not dated 30 April 2003 15 April 2003 15 April 2003 304004 DER669E 2 April 2003 Not dated 30 April 2003 15 April 2003 15 April 2003 304036 DER669E 15 April 2003 Not dated 13 May 2003 6 May 2003 6 May 2003 304037 DER669E 16 April 2003 Not dated 12 May 2003 9 May 2003 9 May 2003 305043 Epikote 1009 29 May 2003 Approximately 17 June 2003 17 June 2003 17 June 2003 11 June 2003 (received 23 June 2003) 305044 Epikote 1009 2 June 2003 Approximately 19 June 2003 17 June 2003 17 June 2003 16 June 2003 (received 23 June 2003) 308007 Epikote 1009 22 July 2003 6 August 2003 2 September 2003 29 August 2003 29 August 2003 (received 1 September 2003) 308008 Epikote 1009 24 July 2003 8 August 2003 2 September 2003 No tax invoice No delivery docket 309024* Epikote 1009 27 August 2003 10 September 2003 26 September 2003 1 October 2003 1 October 2003 (received 2 October 2003) 309039 Epikote 1009 4 September 2003 18 September 2003 30 September 2003 No tax invoice No delivery docket 310024 DER669E 14 October 2003 28 October 2003 5 November 2003 6 November 2003 6 November 2003 (received 10 November 2003) 310060 DER669E 10 October 2003 Not dated Circa 31 October 2003 6 November 2003 6 November 2003 (received 10 November 2003) 310080 DER669E 31 October 2003 14 November 2003 19 November 2003 No tax invoice No delivery docket 311010* DER669E 31 October 2003 No Batch Activity & Testing Record 21 November 2003 No tax invoice No delivery docket 311033 Epikote 1009 31 October 2003 14 November 2003 21 November 2003 16 December 2003 16 December 2003 (received 19 December 2003) 311034* Epikote 1009 31 October 2003 14 November 2003 26 November 2003 16 December 2003 16 December 2003 (received 19 December 2003) 401011 Epikote 1009 24 December 2003 8 January 2004 Approval probably by email or telephone 19 January 2004 19 January 2004 (received 27 January 2004) 401019 Epikote 1009 31 December 2003 14 January 2004 Approval probably by email or telephone 21 January 2004 21 January 2004 (received 27 January 2004) 401022 DER669E 6 January 2004 Not dated Approval probably by email or telephone 27 January 2004 No delivery docket 401023 Epikote 1009 6 January 2004 14 January 2004 Approval probably by email or telephone 27 January 2004 No delivery docket 401053 DER669E 23 January 2004 Not dated 10 February 2004 9 February 2004 No delivery docket (received 11 February 2004) 402005 DER669E 24 January 2004 6 February 2004 11 February 2004 10 February 2004 10 February 2004 (received 11 February 2004) 402006 Epikote 1009 25 February 2004 9 February 2004 13 February 2004 12 February 2004 No delivery docket 402007 Epikote 1009 23 January 2004 5 February 2004 13 and 16 February 2004 12 February 2004 12 February 2004 (received 16 February 2004) 402036 DER669E 24 February 2004 1 March 2004 13 February 2004 1 March 2004 No delivery docket 402037 DER669E 24 February 2004 1 March 2004 2 March 2004 4 March 2004 4 March 2004 402038 DER669E 24 February 2004 1 March 2004 4 March 2004 5 March 2004 5 March 2004 402039 DER669E 20 February 2004 4 March 2004 9 March 2004 10 March 2004 10 March 2004 403040 DER669E 9 March 2004 23 March 2004 29 March 2004 30 March 2004 30 March 2004 403041 DER669E 10 March 2004 24 March 2004 29 March 2004 30 March 2004 30 March 2004 403042 DER669E 11 March 2004 25 March 2004 29 March 2004 30 March 2004 30 March 2004 403043 DER669E 12 March 2004 26 March 2004 31 March 2004 30 March 2004 30 March 2004 404002 DER669E 24 March 2004 7 April 2004 13 April 2004 14 April 2004 14 April 2004 404003 DER669E 24 March 2004 8 April 2004 14 April 2004 19 April 2004 19 April 2004 404004 DER669E 31 March 2004 13 April 2004 16 April 2004 19 April 2004 19 April 2004 404010 DER669E 6 April 2004 20 April 2004 23 April 2004 27 April 2004 No delivery docket 404031 DER669E 30 April 2004 13 May 2004 17 May 2004 18 May 2004 18 May 2004 404032 DER669E 29 April 2004 12 May 2004 18 May 2004 19 May 2004 19 May 2004 405036 DER669E 13 May 2004 27 May 2004 1 June 2004 3 June 2004 3 June 2004 405037 DER669E 14 May 2004 28 May 2004 4 June 2004 4 June 2004 4 June 2004 406054 DER669E 22 June 2004 1 July 2004 6 July 2004 7 July 2004 No delivery docket 406055 DER669E 23 June 2004 2 July 2004 8 July 2004 9 July 2004 9 July 2004 407043 DER669E 20 July 2004 2 August 2004 10 August 2004 12 August 2004 12 August 2004 407044 DER669E 22 July 2004 4 August 2004 10 August 2004 12 August 2004 12 August 2004 407056 DER669E 30 July 2004 Not dated 12 August 2004 13 August 2004 13 August 2004 408008 DER669E 5 August 2004 19 August 2004 8 September 2004 5 November 2004 5 November 2004 408017 DER669E 7 August 2004 21 August 2004 8 September 2004 5 November 2004 No delivery docket 408018 DER669E 11 August 2004 24 August 2004 8 September 2004 5 November 2004 No delivery docket 408019 DER669E 25 August 2004 Not dated 8 September 2004 5 November 2004 5 November 2004