T 1791/19 16-03-2022

1. The appeal is admissible.

2. Background.

The invention relates to a free-flow pump. In a free-flow pump the impeller is spaced from the pump inlet and from a part of the housing of the pump in such a manner that a free passage for solids contained in the pumped liquid is provided between pump inlet and outlet, see figures 1, 4 and 7. Free-flow pumps of this kind are often used in wastewater that is contaminated with solid matter, see specification paragraphs [0001]-[0002]. In practice materials made of fibres or yarn or similar two-dimensional materials tend to accumulate at the impeller front and clog or partly obstruct the pump, specification paragraph [0004].

An object of the invention is to reduce clogging or obstruction while maintaining a high pump efficiency. To this end claim 1 defines an impeller geometry aimed at reducing suction force in its central area and at increasing suction force at its periphery. This enlarges the liquid flow path around the pump central area and and so reduces accretion of two-dimensional materials over the entire impeller front surface, see paragraphs [0004]-[0009]. The claim in particular defines the impeller base surface geometry, that is the geometry of the surface from which the vanes project, in a combination of features 1e, 1f, 1g and 1j.

3. Main request - Inventive step.

3.1 The claim is directed at free flow pump with an impeller having a base with hub body and disk surface from which vanes project. The features of the claim are concerned primarily with the shape of the impeller base. Considered in the radial direction inner portions of the impeller base should not lie deeper with respect to the inner end of the vane front side than respective maximum depths expressed relative to the height difference H between the vane inner end and the maximum depth of the disk surface (features 1e, 1g): in the inner third no more than a sixth, in the inner half no more than two thirds of H. Furthermore, the disk surface should drop continuously in the outward direction (away from the hub) over at least half of the impeller radius (feature 1f). Finally, in a middle portion along the radius the height difference of the disk surface should exceed half of H.

3.2 In qualitative terms these requirements define a general shape that is shallow or even projects in the hub area and then drops away continuously over the greater part of the disk surface in a way that in the first half the shape can be quite steep up to a maximum drop, but in the overlapping middle third it may be less steep (with a minimum drop lower than the maximum in the first half). The patent associates each of the features with different effects: the shallow or projecting hub (1e) with reduced suction and thus reduction of accretion (paragraph [008]), the limited drop (1g) in the inner half with further reduction of accretion (paragraph [0010]), the sufficiently large drop (1j) with pump efficiency (paragraph [0011]), and the continuous drop (1f) with effective flow through the impeller (paragraph [0012]), so that overall the pump is optimized (paragraph [0009]). The above general shape is illustrated in figures 4 and 7 of the patent, with the disk surface sloping downwardly in a convex curve away from the raised hub portion. Indeed, these figures, which only differ in the shape of the lower vane but otherwise appear identical, provide the sole example in the patent of an impeller shape that meets the requirements of claim, even if in a generalized manner.

3.3 Looking towards the prior art, D17, see figure 16a, indisputably discloses a free flow pump having features 1b', 1c and 1d'. It moreover shows a disk surface that declines continuously over its entire radial length and thus the greater part of the impeller radius so that this known impeller also meets the requirements of feature 1f.

3.4 The general shape of the impeller in figure 16a (and indeed in other figures 1, 12 and 13) is very similar to that of figures 4 and 7 of the patent, with a disk surface that slopes away in a similar convex curve from the raised central hub portion. Figure 16a and the figures of the patent are technical drawings that are of a very similar character and quality. Just as figures 4 and 7 are meridional section views so also is figure 16a as is evident from its symmetry, where any inaccuracies are minor and in the Board's view not intentional but the likely result of reproduction distortion (all pages are skewed for example). Just as figures 4 and 7 are meant to illustrate the general shape and form of the components of the pump and their arrangement with respect to each other, so does figure 16a. As the central feature of the pump is its impeller, it can be safely assumed that also in D17 the impeller has been drawn with the intent to have some semblance to reality; the resemblance to the impeller shapes of the figures of the patent is thus not accidental. As the general shapes are very similar, also the effects must at a qualitative level and to some degree be the same. If there are any differences these must therefore reside in the specific values for radial extent (inner third, inner half, middle third) and limits for relative depth and height difference (in terms of H) given in features 1e, 1g and 1j.

As regards features 1g and 1j, close study of the figure, see for example the figure on page 17 of the Appellant's statement of grounds, reveals that within the inner half of the radius the impeller base lies at about half of H, that is well above the lower limit of 2/3 H (feature 1g) and that in the middle third the drop is roughly 2/3 or 3/4 of H, again easily more than the lower limit of 1/2 H (feature 1j).

3.5 The Respondent argues that figure 16a, because of its schematic nature, cannot be used to derive any measurements, citing inter alia decisions T204/83 and T451/88.

3.5.1 The Board notes that T748/91 had already held that under certain circumstances, relative dimensions may be inferred, for example because a feature (profile depth of an intermediate layer relative to that of a superposed surface layer) was an essential aspect of the invention described. Thus, it held that the findings of T204/83 and T451/88 were not absolute. For its approach to rely on drawings as a means of disclosure, the Board further finds support in decision T 1200/05, point 2.2. of the reasons, where on the basis of a drawing alone a certain double-axle arrangement of a truck was held to be disclosed:

2.2 It is established jurisprudence of the Boards of Appeal and undisputed by the parties that the drawings in a published patent or patent application are part of the disclosure of that document.

In the present case although in the description of D2 it is indicated that in the drawings the invention is represented in a schematic way it appears at a first glance that these drawings are quite precise. As a matter of fact they show for instance a number of bolts including the bolts for fixing the air spring elements to the frame 6 of the lorry, they show a precise shape of the support members 3 with the cross sectional area increasing towards the axle, they show the turned back ends of the spring leafs 13 and their fixation bolt 14 and 15, to name only some of the details shown."

3.5.2 In the case at issue, the Board sees special circumstances that allow some inference of relative dimensions. As stated, it sees figure 16a as illustrating a general shape in the very same manner as the figures of the patent and which in a like manner must have some semblance to reality. Furthermore, the claimed ranges for both features are very broad - feature 1j covering 2/3 of all possible depths in the inner half of the radius, feature 1g covering half of the possible height differences in the middle third, so broad in fact that they cover a very wide range of possible shapes. Rather than laying down strict dimensional or geometric requirements that could aid in clearly delimiting the claimed invention from the prior art, these features are of a broad, qualitative character. Finally, the values that are derived from the general shape of figure 16a lie well within these very broad ranges. Thus, even if the general shape shown is not an exact shape and any values inferred thus cannot be true values, the Board sees them as very rough estimates thereof. Considering reasonable margins arising from the general nature of these figures, these roughest of estimates are still seen to lie well within the broad ranges defined in the claim.

3.5.3 The claim features relate to relative dimensions in the height (axial) direction along relative radial portions of the impeller base. This means that any relative heights or depths that might be inferred from figure 16a will not change for any scaling up (stretching) or down (shrinking) of the impeller shape in axial or radial directions or both. As pointed out by the Respondent in relation to feature 1e, the critical factor is the relative position of the inner end of the vane front side in the radial direction along the (convexly curved) impeller base as this determines the height difference against which the various parameters of features 1e, 1g and 1j are measured. Figure 16 shows that position well within the inner half, but not too close to the hub. The actual position would need to be much closer than that shown, maybe as much as 1/10th of the radial length of the impeller base, before the two criteria were no longer met and this aspect were no longer generally represented in the figure.

3.5.4 The Respondent also argues that the axial and radial scaling of figures in D17 appear to be different. This would be suggested by the axial and radial dimensions recited in column 6, lines 58-62 of D17 with reference to the embodiment of figures 1 and 2. This argument is moot. As explained in the previous paragraph, any inferred relative heights or depths are independent of a different axial to radial scaling.

3.5.5 The Board thus concludes that features 1g and 1j are also directly and unambiguously disclosed in figure 16a of D17.

3.6 Turning to feature 1e, the inner end of the vane front side is seen to be spaced considerably, at about a third radial length of the impeller from the hub axis. Thus, a significant inner section of the impeller base is at a level higher than the inner vane end and the remaining section beyond is below it. Nonetheless, because that position close to a third is only roughly indicative, it cannot be concluded from the figure that all of the inner third is no more than 1/6 H deep. This feature is thus not directly and unambiguously disclosed.

3.7 Applying the problem solution approach, it is necessary to identify an effect with this sole difference. As indicated above, this feature is said in the patent, paragraph [0008], to cause, surprisingly, a reduction of suction with resulting enlargement of the flow path and reduction of accretion. As explained above, because of the close similarity of the impeller shapes this effect must also be present to some degree already in D17. Indeed, as shown in D17 (see any of its figures), the central hub area of impellers typically projects forward of the surrounding vanes and will be free of vanes, which can but lead to reduced suction in that area. Hence, the specific values, a depth less than 1/6 H in the inner third must be associated not with the reduction per se, but the degree thereof, which according to the patent would be surprising.

However, there is nothing in the patent to substantiate a surprising degree of reduction of suction. It contains neither an explanation why this would be so, any type of research that demonstrates the effect, or any comparative tests that show that the claimed values have particular significance. It is reiterated that 1/6 H is only a lower limit and the claim thus covers any form and shape of the hub body meeting this requirement in the central area. If any effect exists, it must therefore exist across the entire breadth of hub shapes covered. Without further evidence, it is simply not plausible for the Board that this lower limit has technical significance. In this regard, the statement in the various declarations that this is so merely repeats the allegation and adds nothing. Similarly, the comparative tests subsequently submitted fail to convince. These compare performance of a single impeller of similar, but not identical shape to that of figure 4 of the patent, to an impeller that is similar to that of figure 1 of the patent (but no longer claimed). That reference impeller has a very steep drop from the hub body to a mainly flat disk surface and clearly belongs to a different family of impeller shapes than that to which the impeller shapes of figure 4 of the patent and figure 16 of D17 belong. Any difference in flow characteristics may thus be due to the stark differences between the two families of shapes, rather than demonstrate a significant effect associated with feature 1e across its whole breadth within the family of shapes of figure 4 of the patent and figure 16a of D17.

3.8 As feature 1e is without proven technical effect, i.e. is technically arbitrary, it cannot contribute to an inventive step according to established case law. Since there are no further distinguishing features, no technical problem can be identified which is solved by the claimed pump. It thus lacks inventive step, see Case Law of the Boards of Appeal, Ninth Edition, 2019 (CLBA), I.D.9.5, and the decision T 1009/12 cited in there.

3.9 Indeed, similar arguments regarding a lack of proven technical effect can be made against features 1g and 1j. Thus, also there the patent and subsequent argument and evidence have failed to plausibly demonstrate any effect that can be associated with the limit values across the entire breadth of impeller blade shapes covered thereby. The Board adds that for the effects that have been alleged - further reduction of accretion, maintaining pump efficiency - vane height and shape are certainly also critical, features that however do not appear in the claim.

Thus, even if features 1g and 1j were also seen to distinguish the claimed impeller shape from that of figure 16a of D17, then for the same reasons given above they would not contribute to inventive step and the Board would reach the same conclusion.

4. Auxiliary requests - Amendments

4.1 These requests have in common the feature added to claim 1, amongst other features, that the disk surface consists of an inner convexely curved surface portion and an adjoining concavely curved surface portion.

4.2 The Respondent cites original claim 4 and original page 4, line 22 to page 5, line 2, where only an inner convex surface adjoining the hub is described. These passages however do not indicate how the adjoining area of the disk surface is shaped, whether linear, concave or a combination of both, or even with a stepped decline, as described for the previous embodiment of page 4, lines 13-20. The Respondent argues that an adjoining peripheral concave surface is implicit in original claim 4 as dependent on the previous claims, since it is inevitable to fulfil features 1e, 1g and 1j present in those claims and now included in the amended claim 1 of all requests. The same should hold for the cited passage of original page 4, that is the part of the description corresponding to claim 4. The Board is not convinced by this argument. Other combinations of linear, concave and/or stepped declines for the peripheral surface can also provide a surface that meets the limitations imposed by features 1e, 1g, 1j and 1f. Thus, the Board cannot identify an implicit disclosure here of combination of inner convex surface and an adjoining concave surface.

4.3 A disk surface consisting of an inner convexly curved surface portion and an adjoining concavely curved surface portion is only originally disclosed in page 19, lines 30-33, which correspond to the description of the embodiment of figure 4. However, the disk surface for that embodiment is described there in more detail together with other features that are not included and which are structurally and functionally intertwined to define a specific geometrical shape of the disk surface. In the specific embodiment of figure 4, the disk surface extends over the outer two thirds of the impeller radius and the convexly curved surface only corresponds to about a seventh of the radius of disk surface. No convincing argument has been presented why the claimed features might be selected in isolation from that specific shape. The omission of the other features represents therefore an unallowable intermediate generalisation.

4.4 Thus, the subject-matter of claim 1 of all auxiliary requests is deficient under Article 123(2) EPC.

5. Request for referral to the Enlarged Board of Appeal

5.1 According to Article 112(1)(a) EPC, in order to ensure uniform application of the law the Board of Appeal shall either of its own motion or following a party's request refer any question to the Enlarged Board of Appeal if it considers that a decision is required. If the Board of Appeal rejects the request, it shall give the reasons in its final decision.

In order to be admissible, the referred question must concern a non-uniform application of the law by the boards, see CLBA, V.B.2.3.6.

5.2 The Respondent argues that this Board is not applying the principles regarding interpretation of technical and schematic prior art drawings as established in previous case law, for instance in T204/83 and T451/88. A referral is thus necessary to ensure uniform application of the law in this respect.

5.3 The Board here refers to the above discussion of inventive step. As explained there, this Board in its decision does not deviate from the case law developed in T204/83 and T451/88 and later qualified in T748/91. Thus, non-uniformity in the application of the law cannot be established. The application of the same legal principles and criteria may lead to different results in different cases. The Board does not thus consider that a decision on this issue is required and therefore refused the request for referral.

6. Request for remittal

6.1 During the inventive step discussion of the main request, the Respondent suggested (but did not formally request) a remittal for allowing a two instance discussion of the lack of technical effect issue in the framework of inventive step. The Board then expressed its view that remittal did not appear to be justified, since the issue had indeed been discussed in first instance and was also central to the appeal case of the Appellant Opponent. Section 2 of their grounds is headed ".... missing technical effect" ("... fehlende technische Wirkung") and argues the absence of any proven technical effect or advantage, which, first paragraph of section 3, leads it to conclude obviousness. The issue was again discussed in sections 1.3 and 1.4 of their reply.

6.2 Faced with this line of argument, the Respondent did not appear to pursue the matter of remittal any further until the final stages of the oral proceedings before the Board when the parties were asked to confirm their final requests before pronouncing the decision. Only then did the Respondent submit a formal request for remittal on the above grounds. In the light of the above, considering that the Board had expressed a view on this point and that the formal request for remittal was submitted at such a late stage after closure of the debate and the pronouncement of the Board's conclusion of inventive step, the request had to be refused.

7. Oral proceedings held by videoconference

7.1 In the light of the COVID pandemic situation and ongoing lockdown and general travel constraints, the Board decided to hold oral proceedings by videoconference, see Article 15a RPBA, and accordingly informed the parties with communication of 30 November 2021.

In their response dated 23 December 2021, the Respondent wrote that they did not consent to oral proceedings via videoconference, but that they however assumed that a VICO would take place ex-officio under Art. 15a RPBA.

7.2 The Board notes that at the time it decided to hold the oral proceedings by videoconference, a lockdown was in force, travel restrictions applied and the number of infections was on the rise throughout Europe. Thus, at this time the Board's decision was in line with the criteria as set out in decision G1/21 of 16 July 2021, see its headnote: "During a general emergency impairing the parties' possibilities to attend in-person oral proceedings at the EPO premises, the conduct of oral proceedings before the Boards of appeal in the form of a videoconference is compatible with the EPC even if not all of the parties to the proceedings have given their consent to the conduct of oral proceedings in the form of a videoconference."

That situation had not changed when the Respondent filed their letter withholding consent. The Board therefore saw no reason to deviate from its earlier decision and revert to a hearing in person, the more so as the Respondent in their letter also provided contact details for oral proceedings by videoconference.

7.3 The issue of consent was not raised in the Respondent's subsequent letter of 18 February 2022, where they announced the participation of accompanying persons. Nor was the issue raised at the beginning of the oral proceedings, which they attended remotely and wherein they then actively participated. Only after all relevant issues had been discussed and the debate had been closed, and just as the order was about to be pronounced, did the Respondent state that it had never given their consent to oral proceedings by videoconference.

7.4 The Board finds itself in a situation similar to that leading up to the decision in the petition for review case R1/14. There, the petitioner appellant only took issue with an earlier refusal of a request for postponement at the end of the oral proceedings before the Board, after having actively participated in the discussions. As the Enlarged Board held in its decision, reasons 4.3, "the objection to a refusal to postpone the date of oral proceedings only makes sense if it is made before the date of the oral proceedings or at the latest at the beginning of the oral proceedings. How could the Board remedy the alleged procedural defect, when the oral proceedings - with the full participation by the petitioner - have already taken place?".

Also here, raising the issue of consent to a hearing by videoconference would only have made sense shortly before the oral proceedings, for example in its letter of 18 February 2022, or at the latest at the beginning of the oral proceedings. Only then could the Board have dealt with the issue. After having discussed the case and reached a conclusion on all relevant issues for all requests, it was no longer an option for the Board to consider the question of consent to a hearing by videoconference and to possibly decide that the oral proceedings should have been held in person. It is incumbent on the Respondent to raise issues and file requests at the appropriate time so that such requests can be given due consideration. Otherwise a party could wait to raise such an objection until faced with an adverse decision, but refrain from doing so when faced with a favourable decision. This appears inequitable. That certain procedural objections need to be raised at the outset of the hearing is not alien to domestic laws of civil procedure, particularly in regard of jurisdiction (sec. 39 German Code of Civil Procedure and Art. 26 EU Regulation 1215/2012). Under these provisions, once the defendant enters an appearance, the objection can no longer be raised, however well-founded.

7.5 The Board thus takes the view that when it ordered the hearing to be held by videoconference, it was within its discretionary powers under Art. 15a Rules of Procedure of the Boards of Appeal as qualified by decision G 1/21. Should the Respondent have wished the Board to reconsider its decision in the light of changed circumstances, it would have been incumbent on the Respondent to make any such request prior to any substantive discussion of the case. Faced with such a request only at the end of the hearing, the Board was no longer able to take into consideration that the Respondent had not given their consent for oral proceedings by videoconference. The issue was therefore disregarded.

8. The Board concludes that the decision under appeal was wrong in finding the main request allowable. Moreover, the Board finds that the patent as amended according to the auxiliary requests also fails to meet the requirements of the EPC, so that pursuant to Art 101(3)(b) EPC, it must revoke the patent.