Section 75AK: state of scientific or technical knowledge at the time
74The second defendant submitted that the state of scientific or technical knowledge as at March 2005 was not such as to enable the defect, if there be found to be one in the subject joist, to be discovered.
75While it has not been necessary for me thus far to determine the precise nature of the defect, since it was sufficient for the s 52 and s 53 claims that the wood did not conform to MGP 10, it is necessary for some consideration to be given to the nature of the defect in the subject joist for the purposes of the s 75AK defence.
76A substantial amount of expert evidence was given by two experts: Professor Crews, who was instructed by the second defendant's solicitor and to whom reference has already been made, and Mr Beckett, who was instructed by the plaintiff's solicitor. They gave concurrent evidence in the course of the hearing.
77In a joint statement the experts agreed on two matters of significance:
(1) the physical appearance of the wood in the zone of failure could be described as a "brash" fracture; and
(2) the SEM undertaken by Mr Beckett confirms the presence of damage to the cell walls in a zone approximately 40mm away from the fracture.
78The subject joist was graded MGP 10, in accordance with AS 1748. This process of grading did not detect the defect in the wood which caused it to fail.
79Mr Beckett said that the defect could have been detected by visual inspection. In summary, the second defendant's evidence, based on Professor Crews' opinion, was that the 'state of the art' in 2005 was machine stress grading, that this was what was done, and this did not enable the defect to be discovered.
80For the reasons that appear below, I prefer Professor Crews' evidence to that of Mr Beckett.
81Professor Crews gave evidence that machine stress grading (such as that to which the failed joist was subjected) is a non-destructive process that uses a machine to bend each piece of timber (generally about its minor axis). The machine measures the stiffness of the board and uses a loose correlation between stiffness and strength to assign a strength grade. Professor Crews gave evidence that MGP 10 is a descriptor of a product that has a parcel of statistically derived properties formulated through destructive testing by the CSIRO of some 10,000 pieces of timber. The properties of MGP 10 are published in the Timbers Structures Code AS 1720, Part 1.
82The leading and trailing ends of the timber (700mm from each end) are not graded by the machine because the machine is unable to measure the deflection of the leading and trailing ends. However, they are subject to some testing because the loads are applied to the ends in the course of the grading process. They are also subject to visual inspection.
83The relevant grading standard for MGP 10 is AS 1748, and was developed concurrently with MGP 10. It grades according to the modulus of elasticity ( MOE ), which is concerned with deflection. The MOE is to be distinguished from the modulus of rupture ( MOR ) which is the actual fracture of the material. It also makes provision for visual inspection in s 6, which provides as follows:
" 6 PHYSICAL REQUIREMENTS - STRENGTH CONSIDERATIONS
6.1 General The limitations on visual characteristics listed in this Clause are required in order to meet strength considerations.
6.2 Limitations for softwood species All characteristics passed by the machine shall be permitted in a piece of softwood except the following, to the limits given:
(a) Resin streaks, resin pockets, bark pockets Extending from one surface to the opposite surface, and individually longer than the width of the piece.
(b) Heart shakes Exceeding 3 mm in width or extending from one surface to the opposite surface.
(c) Cross shakes, splits other than end splits Of any length.
(d) End splits Aggregate length at each end exceeding the lesser of twice the width or 200 mm; length individually exceeding half the width."
84Professor Crews gave evidence that machines such as the Metriguard machine in use at the second defendant's mill was the only commercially available technology for grading radiata pine for the purposes of MGP 10 in Australia in 2005 with respect to its structural qualities.
85Mr Beckett gave evidence that there were other techniques available for non-destructive testing of timber in 2005. Professor Crews agreed that there were such techniques available. However, his evidence was:
"What we have to talk about is whether they are available and whether they are actually useful because there were trials already undertaken in Australia on some of these techniques prior to 2005 and they were found to be unreliable and inconsistent. So to answer your question yes a number of these techniques were being investigated overseas and there were some pilot projects being undertaken overseas but in terms of what represented the state of the art the machine grading process was the state of the art at that point of time."
86Whether there were other methods for grading or testing radiata pine in Australia in 2005 is a large question. But the one that I have to decide is a much more confined question: would the particular defect in the subject joist that failed and caused the plaintiff to fall have been detected by any available technology in Australia in 2005?
87It was accepted by both experts that the joist that failed contained compression wood, which had at least the potential to affect its structural strength. It was not, however, accepted that compression wood per se is a defect. Professor Crews said, and Mr Beckett reluctantly agreed, that all radiata pine contains compression wood to some extent. Compression wood is, accordingly, a characteristic of wood, rather than necessarily a defect. It will only amount to a defect where, as here, it materially affects the structural strength of the wood.
88Accordingly, it is necessary to determine whether there was commercially available technology for detecting not only the presence of compression wood in radiata pine in 2005, but also compression wood that would materially affect the structural strength of the wood, since that is the relevant defect in the subject joist.
89In addition to the expert evidence referred to above, there were two substantial documents tendered by the plaintiff on this question. The first was a research paper prepared in 2005 by Mr Duff with funding from J.W. Gottstein Memorial Trust Fund, the National Educational Trust of the Australian Forest Products Industries ( the Duff paper ). The second was a report prepared by Bailleres, Hopewell and Boughton in 2009 for Forest & Wood Products Australia, entitled "MOE and MOR assessment technologies for improving graded recovery of exotic pines in Australia" ( the Bailleres report ).
90The plaintiff relied on these two documents to counter the opinion expressed by Professor Crews that scientific or technological knowledge at the relevant time was not such as to enable the defect to be detected.
91According to Professor Crews, whose evidence on this matter was not challenged, the Duff paper contains four references to "compression wood". The first reference appears in the context of mechanical stress graders, of which the Metriguard machine is one. The words appears in the following context:
"The microfibril angle (MFA) is another characteristic that affects strength and stiffness. The MFA can only be determined at high magnification of the wood fibre walls - and is therefore impossible to directly measure in a production sense. The MFA refers to the 'mean helical (spiral) angle that the fibrils of the S2 layer of the fibre wall make with the longitudinal axis of the fibre' (Jozsa & Middleton, 1994). Mature wood is characterised by a small microfibril angle, while juvenile and compression wood have higher microfibril angles . In the case of juvenile wood, Jozsa & Middleton conclude 'these large fibril angles in juvenile wood have been correlated to lower strength and stiffness in lumber products where these lower values could not be attributed to appreciable differences in density'. As with slope of grain, by measuring stiffness directly, mechanical grading machines are also responding to the effects of MFA." [Emphasis added]
92I do not consider this reference to provide any support for the plaintiff's expert's contention that there were machines commercially available that would have been able to detect the particular defect in the subject joist. Rather, the passage demonstrates that the Metriguard machine measures stiffness which is an effect of the slope of grain to be found in compression wood.
93The first reference establishes, in my view, that the Metriguard machine used by the second defendant was the industry standard. It had the capacity to detect defects, but not all defects. It did not have the capacity to detect compression wood per se, although if the presence of compression wood gave rise to a structural weakness at any part of the length apart from the ends, it would be likely that that would be detected in the mechanical stress grading process.
94The second reference appears in the context of the tracheid effect, which
"... relies on the principle that wood fibres act like optical conductors. The parallel wood fibres in softwood are known as tracheids - the tracheid effect measures the ability of these fibres to scatter laser light. Nystrom (2002) describes this technique: 'when a narrow beam of light strikes a wood surface, a part of the light penetrates into the outermost layers of tracheid fibres where it is scattered and conducted, better in the direction of fibres than across them'.
...
The tracheid effect can be employed to assist in the classification of a number of characteristics, including:
Sound knots - where the wood fibre deviates significantly around the knot
Decay
Juvenile wood and compression wood
Bark"
[Emphasis added]
95Professor Crews explained that this system works by applying a sensor at one end of the timber and an imp act at the other. The signal is timed and gives an average which correlates with the average density of the board. It is to be contrasted with the Metriguard machine, which loads the timber boards continuously with a view to discarding boards by reference to the weakest stiffness in the board.
96The contents of the Duff report do not enable me to conclude that a device that uses this tracheid effect would have been able to detect the defect that caused the joist to fail.
97The third reference appears in the context of a description of a machine called BoardMaster NT. It uses between three and five PCs to manage the image processing and optimising data that can amount to between 10MB and 30MB per board. The more PCs, the greater the production capacity of the system.
" Grading Modules
The BoardMaster NT provides classification of the following characteristics:
Knots
Split
Rot and red stain
Blue stain
Bark and resin pockets
Wane
Warp - bow, spring, twist
FinScan acknowledge that there are a number of characteristics that are still very difficult to detect and classify. Compression wood is o ne of these ." [Emphasis added]
98Professor Crews gave evidence that the obstacle delaying getting machines such as BoardMaster NT into commercial production has been the availability of computing power. He said that it was only in relatively recent years that there has been computing power available at a speed capable of doing these assessments and analyses in real time.
99I accept Professor Crews' evidence that this machine was not commercially available at the relevant time. In any event, even had it been, I could not be satisfied that it could detect the defect that caused the joist to fail, because of the difficulties referred to in the passage extracted above of using this technique to detect and classify compression wood.
100The fourth reference appears in the context of a grading modulus which predicts board strength by working out the density of wood. The resolution of this measurement is around 2.54mm across the board. It is said to allow certain characteristics to be identified due to their effect on density, including compression wood and knots.
101Professor Crews gave evidence, which I accept, that the grading modulus had a heavy dependence on computer power which was not available at the relevant time for commercial applications. Professor Crews said:
"There are eight characteristics identified, one of which is compression wood, and the algorithm works on the basis of identifying density. So the problem that the machine still faces that it may pick up a knot, which may not be any problem at all, it may pick up a small amount of compression wood, and the density may be the same. So it's not distinguishing between compression wood."
102Ms Norton, for the plaintiff, put to Professor Crews certain categories of devices that were addressed in the Duff report. Professor Crews rejected the proposition that any of these machines were commercially available and also opined that they would not, in any event, have picked up the defect in the subject joist. He adhered to his evidence, which I accept, that the Metriguard machine and visual inspection were the 'state of the art' technology in March 2005, and that they were not capable of detecting the subject defect.
103I do not consider there to be anything in the Bailleres Report that undermines Professor Crews' evidence.
104Accordingly, I find that the second defendant has established its defence under s 75AK. It is therefore not necessary for me to determine whether the plaintiff's damages should be reduced by reference to s 75AN of the TP Act. Had I considered that they should be reduced by reference to s 75AN, I would have, for the reasons earlier given, reduced them by an amount of 60 per cent. Although s 75AN does not in terms incorporate the principles of contributory negligence, it has been held to be analogous to such common law concepts: Glendale Chemical Products Pty Ltd v Australian Competition & Consumer Commission [1998] FCA 1571, per Wilcox, Tamberlin and Sackville JJ.