BASF Coatings Aust Pty Ltd v Akzo Nobel Pty Ltd

IN THE SUPREME COURT OF VICTORIA

Not Restricted

AT MELBOURNE

COMMERCIAL AND EQUITY DIVISION

COMMERCIAL COURT
ARBITRATION LIST

S CI 2012 2310

BASF COATINGS AUSTRALIA PTY LTD (ACN 092 127 501)	Plaintiff
v
AKZO NOBEL PTY LTD (ACN 000 119 424)	Defendant

---

JUDGE:	ROBSON J
WHERE HELD:	Melbourne
DATE OF HEARING:	30 and 31 January 2013
DATE OF JUDGMENT:	18 February 2013
CASE MAY BE CITED AS:	BASF Coatings Aust Pty Ltd v Akzo Nobel Pty Ltd
MEDIUM NEUTRAL CITATION:	[2013] VSC 31

---

COMMERCIAL ARBITRATION – Application for leave to appeal against arbitrators’ interim award – Whether or not manifest error of law on the face of the award or strong evidence that arbitrators made error of law – Construction of contractual terms – No manifest error – Application dismissed – Commercial Arbitration Act 1984, s 38.

---

APPEARANCES:	Counsel	Solicitors
For the Plaintiff	RM Garratt QC with TJ Margetts SC	Ligeti Partners
For the Defendant	AJ Kelly SC with R Andrew	Wotton + Kearney

HIS HONOUR:

Introduction

BASF Coatings Australia Pty Ltd (BASF) seeks leave to appeal against an interim award made in an arbitration between itself and Akzo Nobel Pty Limited (Akzo) under s 38 of the Commercial Arbitration Act 1984 (the 1984 Act).

A convenient summary of the background facts appear in the arbitrators’ Interim Award of 30 March 2012 (the Interim Award). Akzo is an Australian company that carries on business, in part, manufacturing and supplying coatings for the use of the automotive industry. BASF entered into an agreement with Akzo for Akzo to manufacture a priming paint to be sold by BASF to General Motors Holden (GMH) entitled “Toll Manufacturing and Service Agreement” (Toll Agreement) commencing 1 May 2007. The Toll Agreement contained the relevant submission to arbitration.

In essence, Akzo agreed with BASF to manufacture and supply to GMH a white primer known as CC19 White Primer. Under the terms of the Toll Agreement, Akzo agreed that the primer would comply with specifications for the product. Azko manufactured the CC19 White Primer and delivered it to GMH. Unfortunately the CC19 White Primer was defective when used in the manufacture of some 400 cars by GMH. The primary coat applied to motor vehicles did not properly adhere to the primer that was painted over metal surfaces. GMH claimed from BASF damages that it suffered in using the defective primer and BASF agreed to compensate GMH in the sum of $4,098,279.

BASF claimed that Akzo was responsible for manufacturing primer that was defective and sought to be indemnified by Akzo for the damages BASF agreed to pay GMH. Akzo denied the claim made by BASF. Akzo alleged that it made the primer according to the specification provided by BASF. This dispute went to arbitration. The arbitrators rejected the claim of BASF that Akzo was responsible for the defect in the primer that it had manufactured. The arbitrators found that the defect in the primer was the responsibility of BASF in the specifications that it laid down for its manufacture.

The arbitration ran for some nine days. After evidence closed, BASF was given leave to amend its claim against Akzo to allege that Akzo breached the Toll Agreement by not carrying out tests it was required to carry out under the specifications agreed with BASF. BASF did not allege that if the tests had been carried out in accordance with the specifications, the tests would have identified the defect in the primer. Rather, BASF contended that Akzo would should not have delivered the paint to GMH as the primer had not been tested in accordance with the specifications and/or if Akzo had informed GMH that the primer had not been tested in accordance with the specifications, then GMH would have rejected the primer because the specified tests had not been conducted. In either case, BASF would have avoided liability for supplying defective paint to GMH as the defective paint would not have been delivered to and/or accepted by GMH and caused damage to GMH. Accordingly, BASF alleged that by reason of the alleged breach of the Toll Agreement by Akzo in not carrying out the tests as prescribed, BASF suffered loss and damage, being the amount it had agreed to compensate GMH. As discussed, below the Tribunal did not, nor did it need to, address this causation issue. Azko did not address the causation issue at the hearing. Accordingly, my failure to deal with the possible defects in the causation issue should not be taken as an acceptance of BASF’s causation argument.

Issues before the arbitrator

The allegation that Akzo had not properly carried out the required testing (even though the required testing would not have exposed the defect in the primer) was pleaded in the third amended points of claim that was delivered on 2 March 2011.

The relevant amended claim was:

25In breach of the Agreement the CC19 White Primer was not manufactured and/or supplied in accordance with the Specifications in that:
(a)the respondent failed to correctly carry out viscosity test T285.6 as the initial test recorded on page 17 of manufacturing batch card CC19 returned a viscosity of 24” and the results of the test were not referred to a QC supervisor for correction action:
(b)the respondent failed to correctly carry out the following tests in that the CC19 White Primer used in each test had a viscosity of 24”;
iadhesion test T702.5;
iithe application test referred to on page 18 of manufacturing batch card CC19;
iiispecific gravity test T289.0
ivsolids test T253.1;
vopacity test T730;
viresistivity test T269.1;
viiall other tests recorded on pages 17 and 18 of manufacturing batch card CC19 (other than grind test T255.0 and viscosity test T285.6).

Relevant legislation and applicable legal principles

It is convenient to set out in some detail the submissions of BASF on the alleged testing breaches and the findings of the arbitrators on these issues, before I turn to the grounds of appeal and the responses of Akzo.

The Commercial Arbitration Act 2011 (Vic) (the 2011 Act) does not apply to the Arbitration, as the Arbitration was commenced before the commencement of the 2011 Act on 17 November 2011.

The circumstances in which leave to appeal may be granted are set out in s 38(5) of the 1984 Act:

The Supreme Court shall not grant leave under subsection (4)(b) unless it considers that-
(a)having regard to all the circumstances, the determination of the question of law concerned could substantially affect the rights of one or more parties to the arbitration agreement; and
(b)there is-
(i)a manifest error of law on the face of the award; or
(ii)strong evidence that the arbitrator or umpire made an error of law and that the determination of the question may add, or may be likely to add, substantially to the certainty of commercial law.

There was no dispute between the parties as to the meaning “manifest error” – both cited the High Court of Australia’s recent decision in Westport Insurance Corporation v Gordian Runoff Ltd, in which the plurality (French CJ, Gummow, Crennan and Bell JJ) stated:[1]

… the words “a manifest error of law on the face of the award” comprise a phrase which is to be read and understood as expressing the one idea. An error of law either exists or does not exist; there is no twilight zone between the two possibilities. But what is required here is that the existence of error be manifest on the face of the award, including the reasons given by the arbitrator, in the sense of apparent to that understanding by the reader of the award.

BASF’s submissions on contractual compliance[2]

[1](2011) 244 CLR 239, [42]. Justice Keifel agreed on this point (see [163]).

[2]I have gratefully used much of BASF’s written submissions.

Clause 10 of the Toll Agreement between the parties stipulated that the paint manufactured by Akzo must comply with its specifications:

10.     Product. Quality; Returns; Claims; Liabilities
10.1   The Supplier warrants that the Products manufactured hereunder comply with the Specifications for the Products. The Supplier warrants that it will comply with any legislation or regulation having regard to its obligations under this Agreement. All other warranties are expressly excluded.
10.2   For the purpose of determining that a Product is within the Specification for such Product the Supplier shall continue to apply the test methods and use the test equipment as set out in the Specifications for each Product.

The Tribunal found:

[49] The Toll Agreement required that the coating constituted by the CC19 White Primer would comply with the BASF Specification as stipulated in Cl.10.1 of the Toll Agreement.
[50] Relevantly, whether a Product, of which the coating called the CC19 White Primer was one, complied with the “Specification”, as defined in Cl.1.1 of the Toll Agreement, was agreed in Clause 10.2 of the Toll Agreement to be determined by applying “...the test methods and use of the test equipment as set out in the Specifications for each Product.”

The documentation to record compliance

The Tribunal found that the “Specifications for the White Primer, CC19 were constituted by the Batch Card and the specified “test methods” and “test equipment” incorporated by reference” in the definition of “Specifications” in clause 1.1 of the Toll Agreement.[3] This finding was not disputed.
[3]Interim Award, [63](b).

The Tribunal found:

... that the Specification contained in the Batch Card for CC19 ... specified each step of the manufacturing process and also specified the test methods to be required to be undertaken including equipment to be used in testing to a very prescriptive degree;[4]
On all the evidence, and the above interpretation of the Agreement, the Tribunal concludes that the Batch Card is the primary component of the Specification for CC19 ... [5]

[4]Interim Award, [63](g).

[5]Interim Award, [63](l).

On the Batch Card, the quality control requirements were set out.

The Tribunal refers to the Quality related testing required of Akzo by the Toll Agreement and the Toll Agreement Specification (see Batch Card E001 to E012). [6]

[6]Interim Award, [58](i).

BASF contends that Mr Black, Akzo’s operations manager, was called by Akzo to establish that the paint had satisfied all the quality control testing. BASF relies on evidence in his cross-examination:

The content of your witness statement, if one was to summarise in a whole, is in substance to establish that CC19 was manufactured properly and tested properly in accordance with the precise requirements of the specification?
That’s correct.
You agree with the proposition that you’ve gone about that task by examining in fine detail the CC19 batch card?
That’s correct.
As I said to you earlier, it’s fair to say that from your experience, you’re the best witness that Akzo Nobel can produce to interpret the batch card?
Yes.
Just tell the arbitrators why you believe that form of evidence is very important in this arbitration?
In terms of understanding the content of the batch card?
Yes. Why is it very important in this dispute?
I guess the batch itself is key to this arbitration.
But why is it key to Akzo Nobel’s position? How do you understand it’s important?
Well, we made can code 19 and tested it to the specification and it passed all those specifications.
So what does that mean?
What does that mean?
Yes?
It means it was approved.
So it’s absolutely critical in your evidence that the paint was approved by all the appropriate QC processes?
Yes.
Because, in essence, it would be correct, wouldn’t it, if it wasn’t and didn’t go through all those processes, we probably wouldn’t be here because the paint would never have been approved to be put into the filling?
Sorry, can you repeat that?
Is the correct term “filling”? If you go to the last part of the batch card, it’s called a filling batch card?
Yes, correct.
So, if you haven’t followed all these QC tests and processes, the paint would never have been approved for filling?
That’s correct.
It must then follow on, if it had never been approved for filling, the paint would never have gone to General Motors Holden?
Yes.

BASF relies on this last answer to assert that if the batch (the subject of the arbitration) had not complied with the Specification the paint would not have gone to GMH, and (according to BASF) it follows in that event that the damage sustained by GMH would not have occurred.

The Batch Card consisted of several pages. Pages 17, 18 and 19 of the Batch Card comprise the Quality Control section of the Batch Card and the “QC Tests Results” were recorded thereon in handwriting by the QC controller.

The Production Paint Test Report was a typed document prepared by Akzo based on the results of the QC tests as set out on the Quality Control Batch Card.

The test methods required by the Specification

The test methods were designated on the Quality Control section of the Batch Card in the column under the heading “Specification” (pages 17 and 18 of the Batch Card). The test methods were prefaced by the letter “T”. They were as follows:

Grind  T225.0
Viscosity                   T285.6
Specific Gravity       T289.0
Solids (SC)*              T253.1
Opacity  T730.0
Resistivity                T269.1
Colour (SC)*
Gloss  T725.2
Adhesion Test         T702.5
Cure-Xylene Rubs   T890.0
MAR Resistance      T775.0
Large Scale App.     T004.3
Overspray                T557.1

The text of the test methods was contained in the respective Product Test Methods.

BASF says that contractually the tests had to be carried out sequentially. BASF contends that only when the first or prior test had been completed and passed, was the next test to be performed. In support of this construction of the Toll Agreement, BASF relies on evidence on the cross examination of Mr Black as follows:
If the second viscosity test returned a result which was within specification, this test result would be recorded and the QC laboratory chemists would proceed to complete the more than 12 remaining tests specified on the batch card?
Correct.
Your evidence is that these two tests are carried out sequentially after the viscosity test?
Yes.
BASF also relies on the supplementary witness statement of Mr Black, where he said:
After further stirring for about half an hour, a further sample would be taken by the operator to the QC laboratory and the batch would be re-tested a second time. If the second viscosity test returned a result which was within specification, this test result would be recorded and the QC laboratory chemists would proceed to complete the more than 12 remaining tests specified on the Batch Card.
BASF’s written submissions also relies on transcript at 746, which was not tendered at the hearing before me.

The Tribunal divided the QC tests into two types:

There are two types of QC tests which are required, those on the applied primer (T780, Line 27 et seq) are colour, gloss, adhesion, solvent rub, mar resistance, opacity and those on the wet primer, specific gravity, solids, resistivity and viscosity.
For the first set of required tests the primer needs to be reduced to spray application viscosity. For the second, set of required testing is done on the wet primer.[7]

[7]Interim Award, [66](v) and [66](vi).

The first of the wet primer tests was the viscosity test for which the specified standard tolerance limits were 28/32 seconds.[8] Primer within these tolerance limits was designated as at ‘package viscosity’.
[8]Ibid.

BASF asserts that, under the Toll Agreement, this viscosity test had to be passed before any further test was undertaken. Viscosity is measured by reference to the time the paint sample moves through a flow cup.

The test results were initially recorded on the Batch Card and then written up by Akzo on a sheet under BASF’s name which went to GMH, called the Production Paint Test Report.

BASF assert that, under the Toll Agreement, to carry out the seventh test on the Production Paint Test Report (the Adhesion test) a sample panel was required which had been made up starting with paint that satisfied the viscosity requirements specified at page 17 of the Batch Card (that is, “package viscosity”).

BASF rely on evidence about the tests given by Mr Black to establish this contractual term:

A representative sample of the paint I suggest to you is the finished paint?
Correct.
Manufactured. To be representative, it’s got to be of package viscosity?
Yes
It’s not representative if it’s not at package viscosity. Clear as day.
That’s correct.
Then the next test – let’s look at another test. Let’s go to p. 36, Resistivity. It says “sample” – now, this is on p. 36?
Yes, on that.
Point 3.2?
Yes.
“Sample under test at package viscosity”?
Correct.
If one goes to, for example, the adhesion test which is at p.53, remember we got you to change the word under Item 4.2?
Yes.
So again it’s got to be “at application viscosity”?
Yes.
Over the page at 5.1 you have a primed panel?
Yes
Again, I assume that’s a panel that’s primed with primer at package viscosity?
Application viscosity.
Yes, but first of all you’ve got to have a piece of primer at package viscosity which you then adjust?
Fully adjusted, sure.

It is important to note that the Resistivity test that Mr Black was referred to in this cross-examination was a wet sample, whereas the Adhesion test was to be done at application viscosity. The parties agreed this was at 22 seconds viscosity.

BASF alleges the tests did not comply with contractual requirements

The QC controller who tested the CC19 batch for compliance with Specification was Mr Van Nguyen. He did not give evidence, and no longer worked for Akzo. BASF contends that evidence was not given to explain his absence as a witness.

Mr Black gave evidence about the testing undertaken. BASF says that as Mr Black was not involved in the testing, his evidence was necessarily interpretative against the background of Akzo’s practices and procedures.

BASF alleged that Mr Van Nguyen followed Akzo’s practice of working through (a print out of) the Quality Control Batch Card and annotating it with the results of his tests. The specified tests were to be performed over a period of 3 or 4 hours. Two viscosity entries were recorded, the first at a score of 24” which was less than the required range of 28/32. An entry also appears as Test 3 (with Test 2 left blank) at a figure of 28”, and so within range (if it reflected a test result, as put by BASF).

At the foot of the sheet which records the viscosity results (page 17 of the Batch Card) is a hand annotation, which records the results of a paint sample adjustment to prepare paint for spraying for the purpose of carrying out the adhesion test (towards the end of the list of QC tests). The starting viscosity of the paint used is recorded as 24”. BASF points out that this is the viscosity recorded on the first test result recorded in the first column on page 17 of the Batch Card.

As regards the figure of 28” entered under the Test 3 column on that page, BASF asserts that there was no explanation for an entry being made in the third column in the circumstances. BASF claims that there was no evidence – apart from the Batch Card entry – of the manufactured paint achieving a viscosity test result of 28”.

BASF contends that in cross examination Mr Black did not think that the paint viscosity could have gone down from 28” to 24” in the period of 2 hours or so which could have elapsed before the preparation of paint in order to carry out the adhesion test.[9] Mr Black accepted that the paint from which the spray paint sample had been prepared for carrying out the adhesion test was not a package sample, namely at the viscosity of 28/32.[10]
[9]T776/19-21.
[10]T777 to T781.

The Tribunal found that the QC controller “used CC19 at a lower viscosity than required for packaging, namely 24” … in order to carry out the product tests that required application of the primer to electrocuted steel and oven baking before testing” and referred to Mr Black’s evidence in chief for this, and the Quality Control Batch Card. BASF says that the Tribunal did not refer expressly to Mr Black’s evidence in cross-examination.[11] I do not consider anything flows from the failure to refer to this evidence.
[11]Interim Award para [60(i)].

BASF asserts that it was to be inferred that the manufactured paint never achieved the minimum viscosity tolerance limit of 28”. BASF says that there is no other explanation for using manufactured paint at 24” to prepare the spraying sample for the purpose of carrying out the adhesion test.

Contrary to the assertion of BASF that the 28” figure did not constitute the results of a viscosity test, the Tribunal found that the 28” figure represented the result of a viscosity test, and concluded that:

(a) as a result of the short grinding and stirring time, the CC19 contained Bentone 38 which was less well incorporated than other batches and had not reached a stable viscosity by the time it was quality checked. Viscosity was still increasing rapidly under stirring when it was quality checked.[12]

and found that a jump of 24” to 28”:

represents a 16.7% viscosity increase in 30 minutes. The viscosity was also found to have doubled (D223) when two drums were tested after one hour’s further stirring at GMH which indicates a 100% viscosity increase by that stage.[13]

[12]Interim Award, [57](viii).

[13]Interim Award, [57](ix) and [57](x).

BASF did not allege these findings involved any error of law.

Viscosity test T285.6

BASF asserts that under the Toll Agreement that the test method required that a viscosity test reading which was too low was required to be referred to the QC supervisor for correction action.

BASF contends that the evidence established (and the Tribunal should have found) that Akzo did not follow the viscosity test method required under the Toll Agreement in not referring the 24” viscosity test reading to the QC supervisor for correction action.

The Tribunal stated:

Mr Black in his supplementary Witness Statement (see B361A and B362A) stated that he was unable to say whether the first test result at 24” was or was not referred to the supervisor, however, based on his experience, he would say that it was not, because the usual practice at Akzo Nobel was that if initial viscosity was low, then the QC chemist who conducted the initial test would instruct the operator to stir the batch for about half an hour and then provide the laboratory with a further sample so that another test could be conducted. (See T758 Lines 15 to 30)[14]

[14]Interim Award, [66](xi).

BASF says that the Tribunal appears to have accepted the proposition that Akzo was entitled to depart from the contractually required test method if its “normal practice” differed and did not require the aberrant result to be referred to the QC supervisor.

BASF raised two contentions on this point. First, that there was no contractual warrant permitting Akzo to depart from the contractual test method in this way. Secondly, that the evidence did not support a conclusion that Akzo had a normal practice of the kind in relation to paint manufacture.

The Tribunal stated:

The Tribunal accepts the explanation by Mr Black and regards this as evidence of reasonable practice in this regard. Further, the Tribunal also considers that BASF has not established on the balance of probabilities that such failures, namely to carry out viscosity test T 285.6 and/or refer a test result of a viscosity of 24” to a QC supervisor for correction, separately or in combination, caused the adhesion loss complained of by BASF, or if undertaken, as BASF asserts should have occurred, would have prevented the adhesion loss complained of by BASF. The Tribunal refers in this regard to the Tribunal’s above reasons as to the likely actual mode of causation.[15]

[15]Interim Award, [66](xii).

BASF says that this passage reveals a fundamental error on the part of the Tribunal. BASF says that the Tribunal proceeds on the basis that a departure from the Specification is significant only if BASF establishes that the departure explains the later adhesion loss.

BASF contends that the Tribunal does not address the fact that a departure from the Specification entitled BASF and GMH to reject the paint as non-compliant, and that GMH would have done so, regardless of whether the departure carried the risk or certainty that material damage would ensue from the use of the paint.

BASF says that the Toll Agreement required the paint supplied to conform with Specification as a result of the application of the specified tests administered in accordance with the test methods using the test equipment.

BASF asserts that this did not happen in key respects: the first (and probably only) viscosity test result lay outside the specified tolerance limit; there was no referral of the first test result to the QC supervisor as required; and the application tests were conducted on the basis of spray paint that was not constituted from paint at package viscosity. BASF submits that it does not avail Akzo that some (but not all) of these conclusions rest on inferences, as the inferences are readily to be drawn, and BASF says that Akzo did not call Mr Van Nguyen or adduce evidence explaining his absence from the witness box.

Unless the tests required by the Specification were carried out in accordance with the Specification as required by clause 10.2 of the Agreement, BASF submits that the paint would have been rejected by Akzo[16] and GMH.
[16]As Mr Black acknowledged.

The test results were entered up on the Batch Card and then transposed (as many of the test methods indeed contemplated) to a Production Paint Test Report on BASF heading, but prepared by Akzo and sent with the paint to GMH. BASF says that this is the context of the finding of the arbitrators that “the controlling document in terms of whether the properties of the primer complied with the Specification, for toll manufacture, was the BASF Coatings Australia Pty Ltd Production Paint Test Report”.[17] BASF submits that the finding is wrong in that the Toll Agreement does not accord the Production Paint Test Report any status as a “controlling document”.
[17]Interim Award, [51].

The wet primer tests (specific gravity, solids, resistivity and viscosity)

These tests were to be performed on wet primer within the package viscosity range (28/32 secs). In relation to the wet primer tests other than viscosity namely:

the specific gravity test …, solids test … and resistivity test …, the Tribunal observes that there was no evidence adduced to establish that these tests were not undertaken by Akzo within the package viscosity specification range. Accordingly, the tribunal does not consider that BASF has proved these matters on the balance of probabilities and in that regard made out its required positive case in relation to these tests.[18]

[18]Interim Award,[66](ix).

BASF submits that contrary to this finding of “no evidence”, there was evidence of the viscosity of the paint sample at the time of these tests. BASF contends that it is the evidence of Mr Black, already referred to, that the viscosity of the manufactured paint would not drop from 28” to 24” in the course of the testing, coupled with the evidence that Mr Van Nguyen began preparing the spray paint for the adhesion test using manufactured paint with a viscosity of 24”.

BASF relies on the evidence of Mr Black in cross-examination:

Well, you do, because you’ve got to have – if you go back to p.36. Before you do your resistivity testing, you’ve got to have the sample at packaged viscosity?
Correct. What I’m saying is I believe it would have been packaged viscosity.
What do you base that on?
Well, because the viscosity was 28 seconds.
Let me go back then: how did it get to 24?
That I can’t answer.
On your evidence, I thought you told the arbitrators it was unlikely to have gone 24, 28, back to 24?
Correct.
Well, it must then follow that he’s done his tests at the rate of 24 straight down the page?
Yes, I don’t believe that.
Mr Digby: But doesn’t the record seem to indicate that that’s the position and what Mr Margetts is asking and what we’re interested in too is if you don’t believe that’s the case, what do you base your view that it wasn’t tested at 24 and 24 at all material times that Mr Margetts has been asking about?
I base that on the fact that the batch tested to the appropriate viscosity at 28 and I believe then the QC test would have been carried out.

BASF submits that Mr Black had no personal knowledge of the matter, and the annotation of Mr Van Nguyen on the Batch Card confirmed the use of paint at a viscosity of 24” to undertake the preparation for the adhesion test (and other tests). BASF says that it must be that the earlier tests were carried out on the paint with a viscosity of 24” consistent with the evidence of the Test 1 result (being a viscosity of 24”). BASF contends that all test results were listed under “Test 1” on the Quality Control Batch Card, which (BASF claims) is again consistent with the use of paint with a viscosity outside the Specification.

BASF submits that the proper interpretation of this evidence was that all tests, including resistivity, were undertaken with a paint sample outside package viscosity. BASF says that there was evidence on this issue. The Tribunal’s finding that there was “no evidence” is wrong and shows that the Tribunal did not properly consider the evidence from the Quality Batch Card and that of Mr Black, both in chief and under cross examination.

The applied primer tests

The Tribunal found that:

The Quality Control Operator used CC19 at a lower viscosity than required for packaging, namely 24” (see P. Black witness statement B345, para 195), in order to carry out the product tests that required application of the primer to electrocoated steel and oven baking before testing. Eg gloss, adhesion, colour, solvent resistance. This is shown on the Batch Card, (see: equation annotated at the bottom of E011. The viscosity required for application of the primer for testing primed steel was 20 to 22 seconds. (See: E012).[19]

[19]Interim Award, [60](i).

The Tribunal subsequently found:

Tribunal has earlier in the Interim Award noted that it accepts Dr Scheirs’ opinion (see T922 to T923 and T923.24 to 924.4), which was also supported by Professor Lamb (see T703 Lines 6 to 15), that it is not necessary for the wet primer to be at package viscosity prior to the reduction to application viscosity and the preparation of primed panels for testing of the applied primer.
Therefore, the tribunal rejects BASF’s assertion that Akzo failed to correctly carry out the tests referred to in the said BASF Pleading [25(b)i, ii, v and vii] because the Tribunal finds that Akzo was not required to carry out those tests at package viscosity.[20]

[20]Interim Award, [66](vii)-(viii).

BASF submits that these findings are wrong. BASF concedes that the test procedures for the “applied primer tests” (colour, gloss, adhesion, solvent rub, mar resistance, and opacity) do not expressly say that the starting paint for the preparation of the application (spray) paint for the sample panel must be at package viscosity before it is reduced to application viscosity. BASF submits, however, that Mr Black gave evidence that the starting viscosity of the paint used to prepare the paint to be applied (sprayed) for the purpose of the applied primer tests must be package viscosity. BASF contends that Mr Black gave evidence as “a person who is deeply involved in these processes” rather than as an independent expert.[21] BASF submits that the effect of Mr Black’s evidence as to industry practice was that the starting paint had to be at package viscosity, and that the appropriate test result entry, if this did not occur, was that the test “was not done”. BASF asserts that Mr Black’s evidence is supported by the fact that the tests were required to be done sequentially over a few hours, and that several of the test methods refer to ‘reducing’ a sample of the paint down to application viscosity.
[21]As arbitrator Mr Digby QC observed.

BASF submits that the reasoning of the Tribunal based on the opinion evidence of Dr Scheirs and Professor Lamb is legally and factually wrong. BASF says that it is legally wrong because Dr Scheirs and Professor Lamb could not give evidence as to what the Toll Agreement required (that is, what it meant) in relation to the specified applied primer tests; that is not the role of an organic chemist. BASF contends that the reasoning is factually wrong because neither expert was expressing an opinion about industry practice, that is, giving evidence about a matter that might be taken into account as the context for the proper construction of the Toll Agreement. BASF contends that the evidence of Dr Scheirs was no more than an assumption on his part, and the passage from the evidence of Professor Lamb does not stand for the proposition for which it is cited.

BASF concludes that the consequence of the correct finding at [60(i)] – and the evidence of Mr Black – is that the CC19 batch was not properly tested prior to despatch, and that if the testing which had been done had been correctly reported, the paint would have been rejected.

BASF submits that despatching the paint as having complied with the stipulated tests, when the tests had not been properly performed, was a breach of the warranty in clause 10.1 of the Toll Agreement, and the Tribunal ought to have so found.

The arbitrators’ findings on the testing issues

The arbitrators’ findings the subject of the application are essentially in paragraph 66 of the Interim Award (emphasis original).

[66]    BASF allegations in relation to Akzo failing to manufacture and supply in accordance with the Specifications and failing to correctly carry out Viscosity test 285.6 and other tests
(i)      BASF, in its third amended statement of claim [25], alleges that the CC19 White Primer was not manufactured and/or supplied in accordance with the Specifications in that:
(a)       the respondent failed to correctly carry out viscosity test T285.6 as the initial test recorded on page 17 of manufacturing batch card CC19 returned a viscosity of 24” and the results of the test were not referred to a QC supervisor for correction action;
(b)       the respondent failed to correctly carry out the following tests in that the CC19 White Primer used in each test had a viscosity of 24”:
i.        adhesion test T702.5;
ii.        the application test referred to on page 18 of manufacturing batch card CC19;
iii.       specific gravity test T289.0;
iv.       solids test T263.1;
v.        opacity test T730.0;
vi.       resistivity test T269.1;
vii.      all other tests recorded on pages 17 and 18 of manufacturing batch card CC19 (other than grind test T225.0 and viscosity test T285.6).
Akzo denies these allegations.

(ii)     Quality control checks are commenced at the end of the tint 1130, 1140, 1150, adjust resistivity 1170 and adjust viscosity 1190, 1200 stags (see E009, 010) to see if the tinting has produced in-specification colour, resistivity adjustment has produced in-specification viscosity. Following these checks further CA1216 was added to adjust resistivity and further tinters to adjust colour. No adjustments to viscosity were made (see Table on E011) and annotated comments on RHS on E011.
(iii)     Mr Black could not say whether these resistivity and colour adjustments were done before or after the second viscosity check (T755, L 21-25 and T 757, L 19-2) although Mr Black did refer to the test performed resulting in a “pass” (T 781, L22).
(iv)     Context for the evidence given by Mr Black on this issue is to be found in Mr Black’s Witness Statement B (p 338 para 134 to p 346 para 198 and Mr Black’s Supplementary Witness Statement B (p 361A to p 361C) and the cross-examination of Mr Black T754 to T784 by Senior Counsel for BASF.
(v)     There are two types of QC tests which are required, those on the applied primer (T780, Line 27 et seq) are colour, gloss adhesion, solvent rub, mar resistance, opacity and those on the wet primer, specific gravity, solids, resistivity and viscosity.
(vi)     For the first set of required tests the primer needs to be reduced to spray application viscosity. For the second, set of required testing is done on the wet primer.
(vii)    Tribunal has earlier in the Interim Award noted that it accepts Dr Scheirs’ opinion (see T922 to T923 and T923.24 to 924.4), which was also supported by Professor Lamb (see T703 Lines 8 to 15), that it is not necessary for the wet primer to be at package viscosity prior to the reduction to application viscosity and the preparation of primed panels for testing of the applied primer.
(viii)   Therefore, the tribunal rejects BASF’s assertion that Akzo failed to correctly carry out the tests referred to in the said BASF Pleading [26(b)i, ii, v and vii] because Akzo was not required to carry out those tests at package viscosity.
(ix)     In relation to the remaining tests alleged by BSASF at [25(b) iii], solids test [25(b) iv] and resistivity test at [25 (b) vi], the Tribunal observes that there was no evidence adduced to establish that these tests were not undertaken by Akzo within the package viscosity specification range. Accordingly, the tribunal does not consider that BASF has proved these matters on the balance of probabilities and in that regard made out its required positive case in relation to these tests.
(x)     The Tribunal also finds that BASF has not established on the balance of probabilities that, in relation to the CC19 White Primer, a failure to undertake the tests referred to in [25 (b)] of the said BASF Pleading, either individually, or in combination, would have resulted in the adhesion loss complained of by BASF, or that if one or all of the said Tests had been performed differently, the adhesion loss of which BASF complains, would have been prevented. The Tribunal refers in this regard to the Tribunal’s above reasons as to the likely actual mode of causation.
(xi)     BASF also alleges that Akzo failed to correctly carry out viscosity test T 285.6 and further that the results of the test (which BASF asserts were in essence anomalous) were not referred to the QC supervisor for correction (BASF Third Amended Statement of Claim [25(a)]). This claimed breach against relates to the viscosity test method T285.6 (see F011 Section 6 “Interpretation of Results”), which states, inter alia, “If the viscosity is too low, refer to the QC supervisor for correction action”.
Mr Black, in his Supplementary Witness Statement (see B361A and B362A) stated that he was unable to say whether the first test result at 24” was or was not referred to the supervisor, however, based on his experience, he would say that it was not, because the unusual practice at Akzo Nobel was that if initial viscosity was low, then the QC chemist who conducted the initial test would instruct the operator to stir the batch for about half an hour and then provide the laboratory with a further sample so that another test could be conducted. (See T758 Lines 15 to 30).
(xii)    Mr Black explains the basis for taking this corrective action in Sections 5 and 6 on B361B. The normal practice was to continue to stir the batch to make it fully homogeneous. The Tribunal accepts the explanation by Mr Black and regards this as evidence of reasonable practice in this regard. Further, the Tribunal also considers that BASF has not established on the balance of probabilities that such failures, namely to carry out viscosity test T 285.6 and/or refer a test result of viscosity of 24” to a QC supervisor for correction, separately or in combination, caused the adhesion loss complained of by BASF, or if undertaken, as BASF asserts should have occurred, would have prevented the adhesion loss complained of by BASF. The Tribunal refers in this regard to the Tribunal’s above reasons as to the likely actual mode of causation.

(xiii)   Furthermore, the Tribunal does not consider that Akzo was in breach of the terms of the Toll Agreement in that the CC19 White Primer was not supplied in accordance with the Specifications for the sole reason that it “caused adhesion loss”, because as addressed elsewhere in these reasons in detail the Tribunal find that Akzo’s obligation to manufacture the CC19 White Primer, as required by Clause 10.1 and 10.2 of the Toll Agreement, was fulfilled and that, BASF has not established to the Tribunal’s satisfaction that Akzo has breached the Toll Agreement.
(xiv)   Further, the Tribunal consider that BASF has not established, on the balance of probabilities, that Akzo manufactured the CC19 White Primer defectively, and/or from defective raw materials, including Gellant Bentone 38. Accordingly, the Tribunal find that Akzo was not in breach in this regard.
(xv)    Further, the Tribunal also considers that BASF has not established on the balance of probabilities that, in relation to the CC19 White Primer, defective manufacture or defective raw materials or defective Gellant Bentone 38 were, separately or in combination, the cause of the adhesion loss complained of by BASF. Again the Tribunal refers to the Tribunal’s above reasons, on the balance of probabilities, as to the actual mode of causation of the relevant adhesion loss.
(xvi)   Finally, as addressed above the Tribunal also finds that Akzo has not breached the alleged testing requirements detailed on the said BASF Pleading [25 (a) and (b)] and the Tribunal finds further that BASF has failed to establish that any of the alleged testing failures were causative of the relevant adhesion failures or that if testing and/or referral had occurred as BASF assert should have happened that would, on the balance of probabilities, have prevented the relevant adhesion failures which have occurred. The Tribunal again refers to its reasons above as to the probable cause of the subject adhesion failures.

Grounds of appeal

BASF filed grounds of appeal. Subsequently, in submissions it elaborated on these grounds. Both parties submitted that I should address the grounds referred to in the submissions, which are set out below (rather than the grounds set out in the proposed notice of appeal.) BASF relies on ten grounds in the form of questions (I will omit their citations and footnotes).

Ground 1

Whether on the proper construction of the Toll Agreement between the parties: “the controlling document in terms of whether the properties of the primer complied with the Specification, for toll manufacture, was the BASF Coatings Australia Pty Ltd Production Paint Test Report”? …

BASF submits that the Toll Agreement (TA) contains no such provision and does not admit of any such interpretation. The finding of the Tribunal reveals manifest error on the part of the Tribunal.

The Tribunal did state that the controlling document in terms of whether the properties of the primer complied with the Specification, for toll manufacture, was the BASF Coatings Australia Pty Ltd Production Paint Test Report. It is clear, however, that in describing the Production Paint Test Report as controlling the Tribunal did not consider it contained any terms of the Toll Agreement (which they made clear in the previous two paragraphs of their report: [49] and [50].) The Tribunal set out the relevant terms of the Toll Agreement, including clause 10.2 (which they emphasised) that said “[f]or the purpose of determining that a Product is within Specification for such Product the Supplier shall continue to apply the test methods and use the test equipment as set out in the specifications for each Product.” The Tribunal accepted that the criteria for determining whether a product complied with the specification was agreed in clause 10.2.

In paragraph 63 of the interim award, the Tribunal again confirmed their acceptance that the “Specifications” for the primer were constituted by the Batch Card and the “test methods” and “test equipment” incorporated by reference and pursuant to clause 1.1 of the Toll Agreement, when they rejected BASF’s contention that the Specification included materials comprised in the submission by BASF to GMH for approval of the primer pursuant to GMH’s Production Part Approval Process.

The Tribunal’s analysis does not show that they treated the Test Report as having any contractual force but rather the Test Report was merely a repository for the test results. The Tribunal do say that the they found that the primer “complied with the Paint Test Report, so far as the required tests specified by BASF under the Toll Agreement to be carried out by Akzo prior to [filling], in order for Akzo to fulfil its toll manufacture obligations in the Paint Test Report for CC19 (see A57).”

In my opinion, in so saying, the Tribunal are treating the Paint Test Report as referring to and recording the tests otherwise required to be carried out by the Specifications. The Tribunal do not treat the report as having some contractual controlling significance as contended by BASF.

Accordingly, I do not consider that there is a manifest error on the face of the award or there is strong evidence that the arbitrators made an error of law as contended by BASF in Ground 1.

Ground 2

Whether on the proper construction of the Toll agreement, compliance of the manufactured primer with specification was to be determined according to the actual results of testing against specification, or only according to how those results were recorded?

BASF submits that it is only the actual test results that mattered, and not how they were (incorrectly) written up. The finding of the Tribunal that the Production Paint Test Report was the “controlling document” in terms of compliance with specification is manifestly erroneous on this ground too: it was the actual test results that governed compliance with specification.

BASF contends that that on the proper construction of the Toll Agreement, compliance with the manufactured primer with specification was to be determined according to the actual results of testing against specification rather than how those results were recorded.

Test method T285.6, for example, makes clear that there is a distinction between the results of the relevant test and the recording of the results. In my opinion, the Tribunal understood this distinction and made findings as to the results of the tests as distinct from the recording of the results (see paragraphs [57]-[66]). A reference to those paragraphs makes apparent that the Tribunal made findings as to the test results and did not merely make findings as to how they were recorded.

In oral submissions, BASF contends that the Tribunal did not clearly distinguish between the two alleged causes of failure: one, the delivery of product which BASF alleges didn’t comply with the specifications because tests had not been done, and the other, the supply of product which because of its inherent qualities caused the loss. BASF submits that the Tribunal did not look at the first level of contractual compliance because of the reasoning in paragraphs [50] to [53].

I do not accept this submission. The Tribunal did address the first level of alleged contravention in [66] as set out above. The Tribunal did not get to causation as it did not find a breach of the Toll Agreement in testing and there was accordingly no need to address the issue of causation.

Accordingly, I do not consider that there is a manifest error on the face of the award or there is strong evidence that the arbitrators made an error of law as contended by BASF in Ground 2.

Ground 3

Whether on the proper construction of the TA, incorporating by reference the viscosity test method (T285.6), Akzo was obliged to refer its first viscosity test result, which at 24 seconds was too low, “to the QC [quality control] supervisor for correction action” as provided in paragraph 6 of the test method?

Akzo did not take this step. It was a manifest error of the Tribunal not to hold that in so proceeding Akzo breached the TA.

Ground 3 raises the issue of whether Akzo was obliged under paragraph 6 of the test method to refer its first viscosity test of 24” to the Quality Control supervisor for correction action.

Akzo contends that the Batch Card provides for four possible tests to be conducted in relation to each specification requirement. For example, under the card, provision is made for up to four tests to be carried out on viscosity specification . Akzo says that the testing procedure is governed by both the Batch card and the relevant “product test methods” document. Akzo says that the when read together the obligation to refer to the QC supervisor only arises if after the four tests provided for in the Batch card the results of the required testing are that the “viscosity is too low”.

BASF, however, contends that the test method is something to be followed through each time a test is performed even though the Batch Card provides for four tests.

The Tribunal dealt with this ground in paragraphs 66(xi)-(xii), set out above. The Tribunal did not rule on whether or not under the Toll Agreement reference to the QC supervisor should have been made after Test 1 and before any further tests were conducted as provided for in the Batch Card. Rather, the Tribunal referred to the evidence of Mr Black that the usual practice was that if the initial viscosity was low, then the QC chemist who conducted the initial test would instruct the operator to stir the batch for about half an hour and then provide the laboratory with a further sample so that another test could be conducted. The Tribunal referred to the evidence of Mr Black that normal practice was to continue to stir the batch to make it fully homogeneous to achieve the required viscosity.

The Tribunal accepted the explanation of Mr Black and found that “this as reasonable practice in this regard.” “The Tribunal also considered that BASF had not established on the balance of probabilities that such failures, namely to carry viscosity test T285.6 and/or refer a result of viscosity of 24” to a QC supervisor for correction, separately or in combination, caused the adhesion loss complained of by BASF.”

BASF submits that the Tribunal did not address the real issue: that is whether or not Akzo breached the Specifications imposed by the Toll Agreement. BASF says that, relevantly, the Tribunal accepted the evidence of Mr Black that the failure of the first test was not referred to the QC supervisor.

The only evidence that Mr Black gave was in relation to whether the first test of the four possible tests was referred to the QC supervisor. In my opinion, Mr Black was referring to whether or not the first test (of the four permissible tests) was referred to the QC supervisor before any of the other three tests were carried out as provided for in the Batch Card.

Azko submits that the Batch Card provides for the possibility of four tests to be carried out to determine whether or not the specification for a particular characteristic of the paint is satisfied. Akzo’s case is that the contractual obligation to refer a low viscosity to a QC supervisor only arises if after the four opportunities to meet the required viscosity, the viscosity of the batch is found to be too low. In the this case, within the four tests provided for, the batch of paint was found to achieve the specified 28” viscosity. Akzo say that accordingly there was no obligation to refer the first test to the QC supervisor when the third test satisfied the required specification of 28”. In my opinion, the Akzo construction of the Toll Agreement on this aspect is correct.

The Tribunal properly took into account how paint is made. The Tribunal found that if the viscosity of a batch is too low then the viscosity may be increased by further stirring of that batch.[22] The batch is not altered or its ingredients changed in carrying out this normal manufacturing process. No evidence was led as to why a QC supervisor would need to be consulted when under normal manufacturing procedures further stirring, without more, is likely to increase the viscosity. It would appear there would only be a need to refer to the QC supervisor if after further stirring and the completion of the four prescribed tests the viscosity was too low.
[22]See Interim Award, [57], [58], and [66].

BASF submits that the construction that it suggests makes sense because someone with superior expertise comes in immediately and investigates what has been done if the first of the four test fails. On the findings of the Tribunal, however, it would make little sense as there was a well-established procedure in making paint that merely involving further stirring which was used to raise the viscosity. Further, the procedure laid down by the Batch Card implicitly contemplated that up to four tests might be made as part of the prescribed method of testing the batch for compliance with the viscosity specification (and thus some tests failing).

In my view, on its proper construction, the Toll Agreement did not oblige Akzo to refer to the QC supervisor if on the first of the four tests the viscosity of the batch was too low.

Although the Tribunal made no express ruling on the construction of the Toll Agreement on the issue of reference to the QC supervisor, I am not satisfied that the result of the arbitration has been adversely affected by the absence of the express ruling. The arbitration proceeded on the footing that the claim about the supervisor did not lead to a failure of the paint to comply with the relevant Specifications.

Accordingly, I do not consider that there is a manifest error on the face of the award or there is strong evidence that the arbitrators made an error of law as contended by BASF in Ground 3. Alternatively, if there was error of law on the part of the Tribunal, in my discretion I would not grant leave to appeal on this ground as I find that the correct finding on the issue raised by this ground would not have altered the result of the arbitration.

Ground 4

Whether on the proper construction of the Toll agreement, incorporating by reference the viscosity test method (T285.6), Akzo was obliged, in relation its first viscosity test result on the Batch Card, to write “FAIL” near the viscosity result, as provided in paragraph 7 of the test method?

Akzo did not take this step. It was a manifest error of the Tribunal not to hold that in so proceeding Akzo breached the Toll agreement.

Under this ground, BASF asserts that the it was a term of the Specification that in relation to the first viscosity test, Akzo was obliged to write “Fail” near the viscosity result .

Under the third amended statement of claim, BASF alleges that the primer was not manufactured and/or supplied in accordance with the Specifications including in that: (a) Akzo failed to correctly carry out viscosity test T285.6 as the initial test recorded on page 17 of the manufacturing Batch Card CC19 returned a viscosity of 24” and the results were not referred to a QC supervisor for corrective action.

Akzo contends that this plea repeats the issues raised in Grounds 1, 2 and 3, but the third amended statement of claim did not raise the issue raised in Ground 4. Akzo concedes that the issue was raised in the cross-examination of Mr Black. Nevertheless, BASF tendered written final submissions of some 37 pages that did not raise this issue. The written submissions expressly submitted that “BASF did not correctly follow viscosity test T 285.6.” Some seven particulars of this allegation are given. There is no reference in any of these seven particulars, however, to any alleged failure to write the word pass or fail on the test results.

I accept Akzo’s submission that the Tribunal were not asked to resolve this issue, nor did the Tribunal address this issue in their Interim Award. In those circumstances, BASF is not entitled to allege the Tribunal erred in law, when BASF did not press the point at the hearing and ask the Tribunal to rule on it.

Accordingly, I do not consider that there is a manifest error on the face of the award, nor is there strong evidence that the arbitrators made an error of law as contended by BASF in Ground 4.

Ground 5

Whether on the proper construction of the Toll agreement, incorporating by reference the viscosity test method (T285.6), Akzo was permitted not to refer the first viscosity test result to the Quality Controller, and not to record the failure on the Batch Card, and instead to follow some other practice of its own?

This is the necessary effect of the findings of the tribunal at IA [66(xi) & (xii)]. BASF submits that the requirements of the Toll agreement with respect to the proper application of the viscosity test method are clear, and required reference to the Quality Controller, and the annotation of the Batch Card.

This ground again raises the issue of whether or not the alleged failure of those conducting the tests to refer to the QC supervisor the alleged failure of Akzo to record “failure” on the Batch card (before the four prescribed tests had been completed) constituted a breach of the Specifications.

The Batch Card has “OK” written on it with initials and date at bottom of card.

In response to Ground 5, Akzo adopt their submissions on Grounds 3 and 4. Akzo also call in aid Cohen v Ockerby,[23] where Isaacs J said that in construing a commercial contract “the expressions, and particularly any elliptical expressions, in a mercantile contract are to be read in no narrow spirit of construction, but as the Court would suppose two honest business men would understand the words they have actually used with reference to their subject matter and the surrounding circumstances.”
[23](1917) 24 CLR 288, 299-300.

Applying this principle, Akzo submits that the Specifications ought to be construed as the court would suppose two honest people engaged in the business of manufacturing thousands of litres of paint would require things to happen.

Accordingly, I do not consider that there is a manifest error on the face of the award or there is strong evidence that the arbitrators made an error of law as contended by BASF in Ground 5.

Ground 6

Ground 6 was not pressed.

Ground 7

Whether, having:
(a)       held (correctly) that for the specified application tests, “the primer needs to be reduced to spray application viscosity”; and
(b)       having accepted the evidence of Akzo’s Operations Manager (Black) that for the purpose of performing the application tests the manufactured paint had to be reduced from packaged viscosity (28/32 seconds) to application viscosity (20/22 seconds) (which Mr Black affirmed in cross-examination); and
(c)       having found that Akzo had used paint at less than packaged viscosity in order to go about the application tests,
the Tribunal erred in law in not holding that there had been a breach of the TA in the way in which the application tests had been conducted, and in not holding that Akzo should have reported its non-adherence to the test methods?
The Tribunal implicitly acknowledged that these matters exposed Akzo to findings of breach. The Tribunal avoided this result by concluding, by reference to evidence in the case of Dr Scheirs and Professor Lamb, that the application tests did not need to be conducted starting with paint at package viscosity.
Here the Tribunal further erred in law because the evidence referred to did not, and could not, displace the requirement of the Toll Agreement correctly construed. Neither expert could give evidence as to what the Toll Agreement required – and did not purport to do so – or as to Akzo’s known practices or industry practice as regards application testing, as at the date of the Toll Agreement, and neither purported to do so. The evidence of Dr Scheirs and Professor Lamb was not to the effect contended by the Tribunal or relevant to the issue of contractual breach.

In substance, this ground raises the contention that the Tribunal should have found that it was contractual obligation on Akzo not to have conducted the spray tests from the batch that initially tested at 24”, even though that batch later tested at 28” viscosity.

This ground was pleaded in paragraph 25(b) of the third amended statement of claim. The Tribunal rejected “BASF’s assertion that Akzo failed to correctly carry out the tests referred to in 25(b)i, ii, v and vii because the Tribunal finds that Akzo was not required to carry out those tests at package viscosity.”[24]
[24]Interim Award, [66](viii).

In my opinion, BASF have not established any error of law by the Tribunal on this issue. The express terms of the specification did not require the tests to be carried out sequentially. The express terms did not require the primer used in the application tests to have satisfied the 28’’ viscosity.

The scientific evidence did not suggest that the test would be in any way compromised by being carried out on primer that only achieved a viscosity of 24” but later achieved a viscosity of 28”. The Tribunal accepted evidence of Dr Scheirs (also supported by Professor Lamb) that it was not necessary for the wet primer to be at package viscosity prior to the reduction to application viscosity and the preparation of primed panels for testing of the applied primer.[25] The evidence of Dr Scheirs and Professor Lamb was not challenged. The Tribunal had found that all the ingredients had been added to the batch at the stage the sample was taken for the spray tests. Akzo accepts, however, that the evidence of the two scientists was not relevant to the construction of the obligations of Akzo.
[25]Interim Award, [66](vii).

The suggestion that industry practice required that the tests must be carried out sequentially and only after each test had been successful was not supported by any evidence from BASF or Akzo to that effect. The proposition that the Toll Agreement mandated that several specification tests be carried out in the sequence in which they are listed on the Batch Card is not consistent with the Batch Card that made provision for up to four tests for each the specification requirements.

In any event, the construction submitted by BASF does not fit well with the sequence of the tests, as the opacity test (which is an application test) is listed before the resistivity test (which is a wet test).

BASF submits that the obligation imposed by clause 10.2 of the Toll Agreement in relation to the test methods should be construed in light of what is known to both parties, BASF and Akzo, acting honestly and reasonably. BASF contends that Dr Scheirs and Professor Lamb are organic chemists, and it is not proper for them to give expert evidence as to industry practice for the purpose of construing the agreement. BASF contends that what is known to the parties is the knowledge of Mr Black about how in the paint industry the testing procedures are done, namely that the tests should be conducted sequentially after each test has been done and passed. There was no evidence that BASF did have this knowledge save that BASF said that Mr Black said it was standard practice. It was not suggested that this argument was put to the Tribunal.

BASF relies, however, on the evidence of Mr Black. BASF says that he gave evidence that the starting viscosity of the paint used to prepare the paint to be used in spray tests must be package viscosity. I have been referred to his evidence at page 760 of the transcript. I think it is fair to say that he agreed in cross-examination that in the normal course of testing that the spray tests will be conducted with primer that had satisfied the initial viscosity test and achieved the package viscosity. He did not support the contention that it was industry practice that it was mandatory not to conduct spray tests with a batch that on its first test gave a viscosity of 24”.

No authorities were cited to support the proposition that in the absence of an express or an implied term, the obligation could be construed by evidence of the usual method of proceeding with the tests.

Even if the evidence of Dr Scheirs and Professor Lamb is not relevant to the construction of Akzo’s obligations – as BASF contends – I am not convinced there was relevant evidence before the Tribunal as to commonly understood industry practice in relation to the whether the application test would be compromised unless the test batch had already achieved package viscosity. In the absence of such evidence, I can see no reason why the construction of the terms would carry with it an obligation to perform the test beginning at package viscosity; the agreement would merely be silent on the issue.

In my opinion, the Tribunal correctly held that Akzo was not contractually bound to carry out the spray tests with a sample from the batch that had initially tested at package viscosity. I do not consider that there is a manifest error on the face of the award or there is strong evidence that the arbitrators made an error of law as contended by BASF in Ground 7.

Ground 8

Given the findings of breach which the Tribunal made or ought to have made, the Tribunal was obliged to address the question of causation, but did not do so or do so correctly. This was a further manifest error of law.

At IA [66(xii)] the Tribunal reasoned that it was incumbent on BASF to show that a departure from specification which gave rise to a breach had also to be the cause of the intercoat adhesion failure experienced by GMH, in order for BASF to succeed in its claim. This is incorrect in law.
Breach and causation are separate issues to be considered separately. The obligation of Akzo was to supply paint that had been manufactured and tested according to specification. Akzo breached the TA in not doing so and in concurrently providing incorrect testing results. The paint would not have been supplied, and would have been rejected, if supplied, but was instead applied by GMH causing a loss which could not otherwise have been happened. BASF suffered loss because Akzo supplied paint that should never have been supplied, whether or not Akzo’s departure from the specified tests explains the loss which the paint caused, and whether or not the specified tests included a specific intercoat adhesion test, as the Tribunal seems to have thought relevant.

BASF contends that the Tribunal ought to have addressed the question of causation. The Tribunal did not address causation. It found no breach of the Toll Agreement. I do not consider it an error of law for the Tribunal not to consider causation, on the grounds that it may have been wrong on breach.

Accordingly, I do not consider that there is a manifest error on the face of the award or there is strong evidence that the arbitrators made an error of law as contended by BASF in Ground 8.

Ground 9

Was there any evidence to sustain a finding that the manufactured paint passed the viscosity test? Such a finding was wrapped up in the finding that the primer complied with all test requirements, at IA [63(j)]. A tribunal is required as a matter of law to have some evidence in order to make a finding, and may not just make findings up.

The only person who could have given direct evidence of this was Mr Van Nguyen, the Akzo tester at the time. He did not give evidence and his absence from the witness box was not explained. He made entries on the Batch Card. The first viscosity test result of 24 seconds was entered on the Batch Card. It was under specification. There was no second viscosity test result entry. The Tribunal found (correctly) that tests following the viscosity test were performed starting with paint with a viscosity of 24 seconds, and not the specified viscosity of 28 to 32 seconds. Akzo’s Operations Manager gave evidence, not referred to in the IA, that the viscosity of the manufactured paint could not have risen to 28 seconds over the testing period. There was an entry of 28 seconds as a purported “third” viscosity test on the Batch Card. It is submitted that, in the absence of Mr Van Nguyen as a witness, it is mere speculation to treat the “third” result entry as the observed result of a viscosity test actually performed.

BASF contends that there was no evidence to sustain the finding that the manufactured paint passed the viscosity test.

I reject that contention. The Batch Card was tendered. It recorded a test result of 28”. I do not consider that Akzo was obliged to call Mr Van Nguyen. There was no inference open for BASF to contend the Batch Card was forged or that a false entry had been made. Jones v Dunkel might have come into play if there was an inference sought to be drawn by BASF that Mr Van Nguyen might have been able to rebut.

There was no inference open to be drawn that 28” was a false entry. In fact, the evidence of Mr Black, the Batch Card itself, and the evidence of the Test Report Card all infer the opposite.

Accordingly, I do not consider that there is a manifest error on the face of the award or there is strong evidence that the arbitrators made an error of law as contended by BASF in Ground 9.

Ground 10

Did the tribunal fall into further legal error in holding that there was no evidence adduced to establish that the application tests were not undertaken by Akzo within the package viscosity range? …

There was ample evidence sustaining this conclusion, mostly on the face of the Interim Award. The evidence comprises pages 36 and 37 of Mr Black’s witness statement (referred to at IA [53]), the entry on the Batch Card showing that the paint for application testing had been prepared from a starting viscosity of 24 seconds, and evidence in cross-examination of Mr Black that manufactured paint would not drop in viscosity from 28 seconds to 24 seconds in the course of the testing.

In paragraph [66(ix)] the Tribunal were referring to the specific gravity test, the solids test, and the resistivity tests. All these were found to be wet tests and not application tests.[26] The Tribunal observed that there was no evidence adduced to establish that these tests were not undertaken by Akzo with the package viscosity range. BASF have not referred to any evidence adduced to establish that these tests were not undertaking by Akzo with the package viscosity range.
[26]Interim Award, [66](v).

Accordingly, I do not consider that there is a manifest error on the face of the award or there is strong evidence that the arbitrators made an error of law as contended by BASF in Ground 10.

Conclusion

BASF has not established any ground upon which leave to appeal may be given. It is not necessary for me to turn to any discretionary considerations.

The application for leave to appeal should be dismissed, with costs reserved.

I direct that the remaining applications in this matter and the applications in the related matters (S CI 4063 of 2012 and S CI 3789 of 2012) be listed for hearing and determination at a date to be fixed.

[2013] VSC 31

18 February 2013