A New Three Part Test for the Disclosure Branch of Anticipation

Wenzel Downhole Tools Ltd v National-Oilwell Canada Ltd 2011 FC 1323, 98 CPR(4th) 155, Snider J

The patent in issue (‘630) in Wenzel Downhole Tools concerned a bearing assembly for use in downhole drilling motors in oil and gas wells. Snider J held that the patent was invalid on the basis of both anticipation and obviousness. Her decision ultimately turned on the facts, but it raises a number of interesting points of law.  

The most striking development concerned the test for anticipation. Snider J began her anticipation analysis by citing the two part analysis set out in Sanofi 2008 SCC 61, which asks (1) was there disclosure to the public; and (2) was that disclosure enabling? [82]. She then summarized the law by stating a new three part test for the disclosure branch:

[94] The first branch of the test is disclosure. The question of disclosure raises three distinct sub-issues:

1. Does the prior art contain the subject matter defined by the ‘630 claims?

2. Was the prior art disclosed before the ‘630 Patent claim date?

3. Was the disclosure of the subject matter of the ‘630 Patent available to the public?

[95] Only if the answer to all of these questions is affirmative, will the Defendants be able to establish the first branch of the test set out in Sanofi-Synthelabo, above.

This does not change the law, but it is helpful in clarifying the distinct issues which must be addressed.

The first branch of Snider J’s test in effect asks whether the patent in issue reads on the prior art. As she put it in the section heading at [96], “Does the [prior art] contain the subject matter of the ‘630 Patent?” In this section, she asks the traditional question of whether the prior art “would infringe the ‘630 Patent had it been constructed at a later date” [102]. Of course, because of the Sanofi two-part test, this can no longer be considered a complete test for anticipation, but it is useful to have this traditional test situated clearly in the new framework.

The second branch asks whether the prior art pre-dates the claim date of the patent in issue [105]. This is essentially a factual inquiry. As an interesting practice point, the prior art (also a bearing assembly), was designed using AutoCAD software, and the defendants introduced expert evidence from a digital forensic specialist to establish the creation date of the design [111]. Snider J allowed this evidence and also awarded costs in respect of this witness, saying “[his] expertise in digital media was useful in establishing the genesis of the [prior art], a question that was, to a large degree, essential to my findings in this matter” [218].

The third branch asks whether the prior art was available to the public. In this case, the prior art bearing assembly had been built and rented to an end-user which had used it in the field in a drilling operation, and then returned it to owner. The key question was whether this use was disclosure to the public. The patentee argued that it was not, because the bearings could not be dismantled at the job site, and in fact were not dismantled at the job site. Snider J held that these points were simply irrelevant as a matter of law: “proof of anticipation does not require demonstrating that a member of the public actually analyzed the [prior art] assembly” [118] (and also [121]). The question is simply whether the prior art was available to be inspected. This, in my view, is consistent with the authorities cited by Snider J (Aldous J’s analogy of a book on a dusty shelf of a public library is particularly helpful [118]), and it is also sound as a matter of policy. The overarching policy of the Patent Act is to provide incentives for the creation of useful inventions, but because of the chilling effect of patents, the goal is to provide those incentives only when they are necessary for call forth the innovation. If the inventor of the prior art was willing to deliver it into the hands of a third party, without concern for whether it might be reverse engineered, this establishes that the inventor of the prior art did not find the incentive of a patent or secrecy necessary; and if the prior art anticipates, it follows that the patent incentive is not necessary for the invention generally.

Snider J also convincingly distinguished Bauer v Easton 2010 FC 361. In that case the invention was a design for a hockey skate which had been used in a Test League “composed of employees and other persons all covered by a confidentiality agreement but which took place in a public arena” [Bauer 212]. The public could see the skates, but Gauthier J held that there was no evidence that a visual inspection available to a spectator would allow that person to discern the essential elements of the invention [Bauer 218-19]. In Bauer, the invention was only available to the public for visual inspection; in Wenzel the prior art bearing assembly was available to the public to be dismantled, even if it was not in fact dismantled [121]. As Snider J put it: “What is made available to the public will depend on the type of examination possible in the circumstances” [88].

After this three part analysis of the disclosure branch of the Sanofi test for anticipation, Snider J turned to the enablement branch. Here, Snider J stated that

[144] In this case, enablement is not an issue. This is because the clear evidence before me is that the 3103 bearing assembly has been used in a downhole drilling motor in the same manner that the ‘630 bearing assembly is used. Accordingly, the enablement branch of the test for anticipation has been met. On the facts before me, the disclosure branch of the test is determinative.

With respect, I am not sure this follows. The enablement requirement, if not exactly the same as the sufficiency requirement, is very similar: Sanofi [26]. The question is whether the disclosure would allow the person to make and use the invention so that the public will have the full benefit of the invention at the end of the term. This is not the same as the question of whether the public can use the invention. So, if the invention is a new drug that cures cancer, and the drug is distributed publically and actually used to cure cancer in many patients, it will be disclosed, but the disclosure will not be enabling if a skilled person in possession of the drug cannot determine how to synthesize it without undue effort. This was essentially the point of Synthon, though there the disclosure was solely in a patent, and on the facts a skilled person would have been able to make the compound from the disclosure without undue effort [Synthon 18]. As another example, if in Bauer members of the public had been permitted to skate around the arena in the skates, but they had not been permitted to leave with the skates, or take them apart, so that they got no more information from skating around than they had from visual inspection, the fact that they had actually used the skates, rather than merely seen them, would not make the disclosure enabling.

Put another way, the enablement question asks whether the invention could be reverse engineered from the disclosure. Snider J’s discussion of enablement does not answer this question. However, her discussion of whether the invention was“available to the public” makes it perfectly clear that a skilled person would indeed have been able to reverse engineer the bearings from the disclosure: see [121] and especially [122], where she says that an analysis of the disclosure “enabled the skilled person to perform the invention.”

Thus I would suggest that Snider J’s “available to the public” analysis slightly confuses the disclosure requirement and the enablement requirement. The question with respect to disclosure is whether the invention was available to the public. That raises issues such as whether the disclosure was made in confidence (which was raised but decided against the patentee on the facts [125-138]), and whether the disclosure would have allowed a member of the public to fully inspect the invention (as in this case) or only partially inspect it (as in Bauer). The enablement question is addressed to what the skilled person could have learned from that disclosure, without undue effort.

In the end, it doesn’t matter to this case whether the reverse engineering question was dealt with under Part 1(c), or Part 2, so to speak, since Snider J clearly did deal with it. It does raise an interesting point about three-part tests of any stripe; while such tests help focus the analysis of the relevant questions, they risk creating additional complexity as to what each part means. But so long as the test is used as a guide rather than a straightjacket, the overall effect should be helpful.

PS: This decision was apparently released in November of last year, but I’ve only just become aware of it with its publication in the latest issue of the CPR. I’m not sure how it escaped me when it was released, as I am diligent in checking “Decisions recently posted” on both the FC and FCA websites, but I note that the CPR, contrary to its usual practice, published this decision with only a docket number, and not a neutral citation. This makes me think that it was never posted on the FC website in the weekly releases. I’ve finally signed up for Alan Macek’s e-mail alert service, so I hope to avoid this problem in the future