UNCATEGORIZED - other posts

T 1076/21 – The difference between the "burden to substantiate a case" and the "burden of proof on the substance" in matters of sufficiency – No automatic reversal of the onus of proof when entering appeal

chat_bubble 6 comments access_time 8 minutes

The patent relates to a grain-oriented electrical steel sheet having excellent magnetic characteristics and coating adhesion.

Claim 1 as granted comprised a coating layer A on the steel side and a coating layer B formed on a surface side mainly composed of glass.

Dependent claim 4 mentioned the following parameter: “a coating weight of the coating layer A on the steel sheet side is 1.0-3.0 g/m2 (both sides) as converted to oxygen”

Brief outline of the file

The patent was revoked for lack of sufficiency by the OD and the proprietor appealed.

The board held that the invention was sufficiently disclosed, set the decision aside, and remitted the case to the OD for further prosecution.

The case is interesting for the considerations of the board about the difference between the “burden to substantiate a case” from the “burden of proof on the substance“when it comes to sufficiency.

The opponent’s position

It was not possible to reproduce the invention reliably or with any certainty from the parameters disclosed in the examples. Although there were no differences between the manufacturing processes, some samples fell within the scope of the invention while others did not. For example, samples 1, 6 and 11 of Example 1 were carried out with the same manufacturing process but the achieved product properties were significantly different, even to the extent that sample 11 fell outside the claimed range.

The skilled person was thus confronted with an undue burden to experiment with various coating weights, without any guidance on how to do so.

Due to the unusual parameter in claim 4, there was a weak presumption of sufficiency of disclosure. It was not clear how the conversion to oxygen could be established or how it could be measured.

The OD had already been convinced by the reasoned arguments provided by the opponent. There was no need to provide evidence. Since the OD revoked the patent, in accordance with T 585/92, the burden to provide proof that the patent was sufficiently disclosed shifted to the proprietor.

The proprietor’s position

The examples did not show inconsistent results. Paragraph [0048] of the description stated that the coating weight of layer B, the glass layer, was varied. The variation in the results originated from the varying of the coating weight of layer B.

The qualitative causal relationship between the coating weight and the resulting tensile stress formed part of the skilled person’s common general knowledge, and finding suitable values for adjusting the coating weight was an ordinary task that could be conducted without undue burden.

In opposition, the opponent did not present verifiable facts able to refute sufficiency of disclosure.

T 585/92, referred to by the opponent, was an isolated decision which had to be assessed in view of the underlying case. In the case at issue, the burden of proof did not shift to the proprietor since the patent had been erroneously revoked.

The board’s position

On the basis of an in-depth analysis of the relevant case law, the board has concluded that the burden of proof regarding the facts, arguments and evidence on the substance (which initially lies with the opponent) does not shift to the proprietor just because the patent has been revoked due to an alleged insufficient disclosure.

Only if there are serious doubts, substantiated by verifiable facts, may an application or a patent be objected to for insufficient disclosure, cf. T 19/90, Reasons 3.3.

If a material fact is not or cannot be proven, a decision is taken on the basis of who bears the relevant burden of proof. The fact that the real position cannot be established is to the detriment of the party which bears the burden of proof for this fact, among other things because the other party is given the benefit of the doubt.

After the grant of the patent, i.e. after the end of the examination proceedings, a legal presumption exists that the patent meets the requirements of the EPC. In an opposition based on the ground of Art 100(b), it is the opponent who initially bears the burden to prove that the patent does not disclose the invention in a manner sufficiently clear and complete for it to be carried out by a skilled person.

However, this presumption can be rebutted. The weight of the submissions required to rebut the presumption depends on its strength. A strong presumption requires more substantial submissions than a weak one.

In the context of the ground of opposition of lack of sufficiency of disclosure, the strength depends on the way the invention is disclosed in the patent, cf. T 63/06, Reasons 3.2.1 and 3.3.1; T 55/18, Reasons 1.2.

Often both aspects are together referred to as the burden of proof. This might have led to the misconception that any decision finding insufficiency would lead to the burden of proof being shifted to the proprietor.

However, when distinguishing the “burden to substantiate a case” from the “burden of proof on the substance” it can be clearly deduced from the case law that after a decision by the OD allowing an objection of insufficiency, it is down to the proprietor to substantiate on appeal why that decision is wrong.

The “burden of proof on the substance“, and in consequence the benefit of the doubt, however, is only shifted when the OD’s assessment that the presented facts, arguments and evidence were sufficient to discharge the opponent’s burden of proof turns out to be correct.

An appeal case need not necessarily be substantiated by submitting facts and evidence that counter the arguments and evidence presented by the opponent.

It may also be substantiated by submitting why the OD’s reasoning on insufficiency was flawed and why the burden of proof on the substance still lies with the opponent.

In this respect, the board referred to T 1911/17, Reasons 24, T 942/18, Reasons 2.2 to 3.6, T 55/18, Reasons 1.2 and 2.2 and T 1596/16, Reasons 2.3.

This is also in line with decision T 585/92 referred to by the opponent, T 585/92 merely states that the appellant bears the burden to substantiate the appeal. It does not imply that the decision under appeal reversed the burden of proof on the substance.

In all reviewed cases, the respective boards assessed whether or not the opponents had discharged their “burden of proof on the substanceindependently of the “burden to substantiate the appeal“.

The board distinguished the following situations

  1. Successful discharge of the burden of proof on the substance by submitting experimental data

The board referred to T 480/11, Reasons 3.5, T 1529/17, Reasons 8.10 and T 518/10, Reasons 7.10.1.

2. Successful discharge of the burden of proof on the substance without experimental data, but with reasoned arguments

The board referred to T 188/18, Reasons 3.21, T66/07, Reasons 7, T 428/21, Reasons 1.4; T 262/11, Reasons 3.4; T 347/15, Reasons 2.2. In T 325/13, Reasons 2.4.8 and 2.5.8, the board even considered that there was no need for the opponent to provide further evidence because the deficiency was plausible from the cited prior art and the simple physiochemical context. Under this heading, the board cited further decisions dealing with the topic.

3. Reasons other than a misjudgement on the substance

An appeal against a revocation of a patent for insufficient disclosure need not be based on the submissions of counter-evidence by the proprietor, as can also be seen in cases where the OD’s decision was set aside.

In many cases the board based its decision not on new evidence but on other arguments why the impugned decision was erroneous.

The board referred to T 1596/16, Reasons 2.2 and 2.3, T 1002/20, Reasons 2.4; T 55/18, Reasons 1.2 and 2.2; T 720/17, Reasons 2.1; T 430/15, Reasons 3; T 355/15, Reasons 6.7.2 to 6.7.4; T 1052/13, Reasons 8.10; T 1746/12, Reasons 6.3.2 to 6.5; T 2006/08, Reasons 4; T 499/00, Reasons 1.10.

In T 942/18. Reasons 2.2 to 3.6, the proprietor successfully showed that the OD’s decision had not been sufficiently reasoned within the meaning of R 111(2).

Substantiation of the appeal in the case in hand

The case at issue is similar to T 1596/16. It is disputed whether or not the guidance given in the patent is sufficient for producing the clad material and whether or not the parameters characterising layer A according to claim 4 can be established. Simple physiochemical considerations, like in case T 325/13 referred to above, cannot resolve this dispute.

For the board, the description of the patent does provide guidance. There is no reason to disregard or doubt it from the outset. The opponent did not submit evidence to prove that the guidance disclosed in the patent was insufficient for carrying out the invention, cf. T 1596/16, Reasons 2.2 and 2.3.

The OD thus could not base its decision on verifiable facts as required by established case law. Sufficiency of disclosure therefore cannot be denied. The board further referred to Case Law Book, 10th ed., Chapter II.C.9; T 499/00, Reasons 1.10; T 2006/08, Reasons 4, T 1002/20, Reasons 2.4.

The opponent argued that the coating weight for layer A as defined in claim 4 was an unusual parameter or was not able to be measured.

There is no indication that these aspects were already raised in the opposition proceedings. The decision under appeal consequently does not cover them.

There is no evidence on file showing that the guidance provided in the patent is insufficient, e.g. because identical coatings in terms of manufacture and thickness imparted substantially different tensions on the steel sheet.

Irrespective of the admission of these arguments under Art 12(6) RPBA21, the opponent did not show that the parameter for defining layer A in claim 4 was an unusual parameter or not able to be measured.

Since the opponent did not provide the evidence required for substantiating serious doubts; it did not discharge its burden of proof on the substance.

Since the parties make contradictory but unsubstantiated assertions concerning facts, the proprietor is given the benefit of the doubt, cf. T 72/04, Reasons 3.


It is not so frequent that a board takes the bother to have such wide look at the case law on a specific topic, here sufficiency of disclosure, and the various ways it can be dealt with.

This is a service to the users of the EPO going way further than a bare minimum of reasoning in order to dispose of the case as quickly as possible. This point needs to be emphasised.

As a summary there are two ways to deal with sufficiency

  • With experimental data, which is the most common way. Arguing a lack of sufficiency without any experimental data boils in general down to a mere allegation, or even a hidden clarity objection.
  • Without experimental data, when the opponent makes it plausible that common general knowledge cannot supplement the missing information. In other words, the argumentation has to be solid,cf. T 100/15 or T 275/16. When a parameter cannot be measured, even then experimental data might be required.

An unsubstantiated allegation is insufficient to cast doubt on the presumption that the invention is sufficiently disclosed, see T 1163/12, Reasons 3.3.

The breadth of an expression is more a matter concerning the clarity requirement of Art 84 rather than a ground for opposition under Art 100(b). The mere fact that a term is broad does not prevent a skilled person from carrying out the invention, cf. T 2182/11, Reasons 2.

When appealing a decision on lack of sufficiency, it has to be distinguished between the “burden to substantiate a case” from the “burden of proof on the substance“. The mere revocation of the patent by the OD is, as shown by the present case, not always sufficient for a reversal of the onus of proof.

The board also reminded that in case of contradictory assertions by the parties, whereby the Office is not in a position to decide, then the benefit of the doubt goes to the proprietor. Further to T 72/04, it is also possible to refer to T 808/05 or T 1598/15.

The legal presumption of validity of a granted patent

It can be agreed with the board that there exists a legal presumption that a patent meets the requirements of the EPC once it has been granted.

When looking at published decisions after appeal in opposition, only 18,7% of the oppositions were rejected, but the patent was revoked in 48,1% of the cases or the patent maintained in amended form in 33,2% of the cases.

Serious doubts about the presumption of validity of a granted might thus be raised. It is clear that oppositions are only filed when a patent is disturbing an opponent, be it for the risk of already infringing or infringing in the future. It remains nevertheless that it is just in case of possible litigation that a patent needs to be strong and valid. If a patent crumbles at the first attack in opposition, one wonders why it has at all been granted.

The present case is also to be seen among the 50%+ of decisions of ODs which have been set aside by a board.


Share this post


6 replies on “T 1076/21 – The difference between the “burden to substantiate a case” and the “burden of proof on the substance” in matters of sufficiency – No automatic reversal of the onus of proof when entering appeal”


Dear DXT
I am not sure if your numbers are accurate, because they neglect the number of inter partes appeals terminated without any decision.
In 2022, for example, according to page 31 the BoA’s annual report for 2022, only 60% (1375) of the 2305 inter partes cases were terminated with a decison on the merits. In 40 % of these the 1375 cases, the appeal was dismissed, i.e. the decision confirmed. In around 60% of said 1375 cases, the appeal was at least partly successfull. This percentage remained stable since 2016. This means that in around 36% (0,6×0,6) of the inter partes appeal, the decison of the opposition division is not confirmed. In around 64 % the opposition division’s decision becomes final.
The same number of around 65 % of opposition division’s decisions confirmed can be found in figure 43 of the EPO’s 2022 Quality report

In my opinion, the number of 50% you mentioned is overstimated, as it does not take into account the appeals terminated without any decision.

Avatar photoDaniel X. Thomassays:

As your comment is a copy/paste of your comment in T 558/20, my reply will be as well a copy/paste of that given in T 558/20, with some additions.


I beg to disagree with you. I did not neglect anything. I have made clear what the figure actually represents.

I fully agree that a certain number of inter partes appeals are terminated without any decision.
The partial reimbursement of the appeal fee when an appeal is withdrawn is there to encourage parties to withdraw their appeals.

The % I have indicated is relative to PUBLISHED decisions of the boards of appeal.
Indeed, I did not take into account the absolute number of appeals ever filed.

Even if the absolute number is lower, which I can fully agree with, the trend revealed in PUBLISHED decisions is notable and does not necessarily bode well for the quality of the work delivered by ODs, in spite of what is said in various publications of the EPO.

When a board had a look at the decision of an OD, the boards set aside more than 50% of OD’s decisions.
Thus in cases in which a decision has been taken by a board, I maintain my figure.

Hiding behind decisions which have not been taken is a good trick to try to hide or belittle the problem or even claim that there are problems. The problems are however real and all presentation tricks will not make them less acute.

When looking at the divergences the following are examples of different categories:

Maintenance-Maintenance in a different form
Maintenance-Remittal for further prosecution

Rejection opposition-Maintenance
Rejection opposition-Revocation
Rejection opposition-Rejection for different reasons

Revocation-Remittal for further prosecution
Revocation-Revocation for different reasons

I invite you to look at PUBLISHED decisions of the boards. This source is open and available to any interested person.
When no decision has been taken, it is easy to claim that there has been no problem.

I will however add further comments:

It is only in a few number of cases, i.e. in less than 10% of the PUBLISHED decisions of the boards that opponents come up with documentary evidence which was truly not available when the original search was carried out. Under this heading one can think of public prior uses, oral disclosures in conferences, industry catalogues and PhD dissertations.

In the rest of the cases, opponents came up with documents, especially patent literature, which were present in the search files when the original search was carried out. Under this heading one can think of highly relevant documents, often classified in the same classes as those having been searched, documents belonging to the prior art under Art 54(3), or documents published in the priority interval when the priority was not validly claimed. When opponents come up with such documents in spite of a search report mentioning a lot of A, Y or X documents, and that the patent is revoked or limited, it is difficult to deny that there are no problems.

I can agree that oppositions should not be overestimated as I am fully aware that only a few patents are opposed. In some technical areas there are no oppositions at all or hardly any. In some technical areas, oppositions raise for a while and then decay, e.g. pace makers or lithotripters. Some technical areas are replaced with other ones. Absorbable devices was such an area. It seems to have been replaced to a large extent by treatment of keratinous fibres (=hair).

Oppositions can thus not be considered, as such, as a direct corrective of examination. However, patents opposed have a commercial importance for opponents as other wise they would not be opposed. It is thus important for the proprietors not to see their patents revoked or limited in opposition. When a patent is held by a SME, its revocation can have deleterious effects.

When it is apparent that for important patents, i.e. those opposed, they do not really resist scrutiny, it can legitimately be considered that there is a problem, as the form in which they have been granted is manifestly not correct.

When an appeal after opposition is withdrawn, it means that no party to the opposition procedure has any interest in putting more money in the box. However this does not mean, that those cases can be considered without any problem. The problems, if any, simply do not appear.

Problems can only appear when a board takes a decision which is then published. Neglecting the information found in published decisions might be easy, but this freely available information can easily be retrieved and taken into account by users of the system.


Dear DTX
I completely understand your point.
But it is a matter of fact that only around 36 % of the opposition division’s decisions are set aside. This number is still too high, but remained stable between 2016 and now.
The sentence “ More than 50% of decisions taken by ODs have been set aside since the beginning of 2023.” that you sometimes use is at least misleading.

Avatar photoDaniel X. Thomassays:


If you really claim to understand my point, you would not insist so heavily on the figure of only 36% of the opposition division’s decisions being set aside.

If you consider that my statement that “more than 50% of the opposition division’s decisions being set aside” to be at least misleading, I can easily return the compliment.

I would even allow myself to claim that by taking into account opposition procedures which ended without a decision of a board, in the absence or the withdrawal of an appeal, you are plainly trying to embellish reality.

As there is no decision of a board, due to the absence or the withdrawal of an appeal, it is actually a fallacy to consider that in those cases the decision of the opposition division has not been set aside.

You are trying to mix apples and pears as it gives a much nicer end figure. I compare what is comparable, i.e. decisions in opposition which have been revised by a board of appeal. And this figure amounts to more than 50%.

Taking into account cases in which no decision of a board has occurred, also leads to much nicer figures with respect of revocation, maintenance in amended form and rejection of the opposition.

A proper assessment of the work of an opposition division, and indirectly of search and examination divisions, can only be made when comparing what is comparable, that is before and after an appeal.

I hope that you can at least agree that the result of the appeal is what should have been the result of examination and opposition when considering patent literature in the search files of the EPO. The absence of a decision of a board cannot be equated with the absence of any problem.

Should you not already be aware of it, I take the opportunity to draw your attention to the following post in a Swiss blog:



Well, in the absence of a board decision, an opposition rejected by an opposition division remains rejected.
If you say that the result of the appeal is what should have been the result of examination and opposition, does it mean that one should accept the outcome of T1989/18 with respect to the adaption of the description ?

Avatar photoDaniel X. Thomassays:


As you realise that you are biting on stone, you react in two ways.

The Tibetan mill

Indeed, you repeat your stance like a Tibetan prayer mill without going into a substantial answer to my position.
At a pinch, you can maintain your figure of 36%, but then you should put this figure into proportion by stating the contribution to this figure of the number of cases in which there was no appeal or the appeal has been withdrawn.

What matters for the users is not the figure of 36%, but the fact that, when a board takes a look at published decisions of the boards, 50+% of decisions of ODs are set aside. This is an undeniable matter of fact.

Change the subject

In your last comment, you actually play Monty Pythons: and now for something completely different!! You come up with one of a series of decisions which consider that adaptation of the description is not a requirement of the EPC.

For a start the question raised in this decision has nothing to do with oppositions, as Art 84 is not a ground for opposition.

Next to T 1989/18, there are T 1440/20, with the same legal member, and T 2194/19 with another legal member. In case EP3094648 (=T 56/21) involving the same applicant and the same legal member, the legal member has, in a communication under Art 15(1) RPBA21, invited the applicant to state whether he wishes a referral to the EBA on this topic. It looks a bit odd that a board actually requires the agreement of a party to file a referral to the EBA. If the board was so convinced of its position, it should decide for a referral and not wait for what the applicant has to say. It is clear that the applicant is potentially interested to get read of this problem in a way favourable to him. It also has the large approval of mainly British patent attorneys.

In various contributions, especially on IPKat, my position has been clearly expressed, not under a pseudo, but under my name. You can easily verify.

My position is abundantly clear: once the wording of the claims has been agreed with the applicant, the description has to be adapted to the claims. As long as Art 84 is at it is, there is no way out of this requirement.

Final comment

Last but not least, I would like to draw your attention that the topic of my blog entry was about Art 83 and the reversal of the onus of proof. The divergence between the decision of the BA and of the OD was only a marginal aspect you are trying to blow up.

If you do not have any further contribution on the actual topic of the blog, it appears preferable not continue your stream of comments. It is a matter of fact that we have diverging opinions in matters of setting aside the decision of the OD by a board. For the interest of the audience, we should leave it at the present exchange as the mutual points of view are abundantly clear.

Leave a Reply

Your email address will not be published. Required fields are marked *