B.n.f. Metals Technology Centre v. the Broken Hill Associated Smelters Proprietary Limited, (Bhas)

[1988] APO 33

In the Matter of the Patents Act 1952 - and - In the Matter of Patent Application No. 540860 in the name of B.N.F. METALS TECHNOLOGY CENTRE - and - In the Matter of Opposition thereto under Section 59 by the BROKEN HILL ASSOCIATED SMELTERS PROPRIETARY LIMITED.

 

DECISION OF A SUPERVISING EXAMINER OF PATENTS:

Background

Application No. 71913/81 was lodged on 17 June 1981 by B.N.F. Metals Technology Centre for a patent for an invention entitled "Continuous Method for Removing Copper from Lead". The application was advertised as accepted on 6 December 1984 and given the serial No. 540860. A notice of opposition was lodged on 4 March 1985 by the Broken Hill Associated Smelters Pty. Ltd. (BHAS); the grounds of opposition were those specified in paragraphs (c) to (i) of sub-section 59(1), although only matters relating to section 40 were pursued by the opponent. The matter came to hearing in Melbourne on 9 August 1988; Mr B. Caine of Counsel, instructed by Mr P. Pearson, patent attorney, of Edwd Waters, appeared for B.N.F. and Mr R. Macaw of Counsel, instructed by Ms V. Santer, patent attorney, of Griffith Hack & Co., and Professor F. Lawson from Monash University, appeared for BHAS.

The Specification

The specification relates to the use of sulphur to remove dissolved copper from molten lead by formation of a copper sulphide dross which floats to the surface of lead. Conventionally this "sulphur drossing" process has been practised as a batch process with inherent disadvantages such as arduous labour, hygiene hazards, and difficulty in achieving consistent output. The specification refers to, as prior art, a continuous process described in British Patent No. 1524474 (AU 13564/76) which discloses a series of agitated reaction vessels used to avoid back-mixing, the charge being continuously transferred from one stage to the next. The specification suggests that a disadvantage of this known process is that each agitated reaction stage is homogeneous. This is explained as follows:

"Now the rate of reaction of copper with sulphur in molten lead is initially rapid but slows down greatly as the concentrations of free sulphur and free copper are reduced. A homogeneous mixture therefore reacts more slowly than one whose composition is continuously changing as reaction takes place. Moreover, the selectivity of the reaction, as well as removal rate, is better when the copper concentration is high. If the output is to be of a low copper content and the reactor is homogeneous, the reaction occurs in low copper content lead; this produces a high lead content dross and is thus less efficient than reacting a high copper lead. In order to avoid these problems, the Patentees use a series of reaction stages. But this is not very efficient, since the major part of the reaction probably takes place in the first stage, and requires relatively expensive equipment."

The invention, which allows the process to be carried out on a small scale, involves the use of a single stirred vertical reaction vessel in which a stream of lead is caused to follow a spiral path from top to bottom. Claim I reads as follows:

"A continuous method of removing copper from lead, by reaction of copper with sulphur in a single stirred reaction vessel maintained under non-homogeneous conditions, which method comprises introducing a stream of molten lead containing copper as an impurity to the upper end of a vertical stirred reaction vessel, feeding sulphur into the stream of lead at the upper end of the vessel, maintaining a dispersion of sulphur in the stream without substantial back-mixing for a time sufficient to effect reaction between the sulphur and the copper, recovering the stream of lead from the lower end of the vessel, and allowing the formed copper sulphide to float to the surface of the recovered molten lead."

Claim 8 is directed to the features of the reactor and reads as follows:

"A method as claimed in any one of claims 1 to 7 carried out in a generally U-shaped reactor having an upstream arm joined to a downstream arm at their lower ends, the said upstream arm comprising an elongated vertical stirred reaction vessel of circular cross-section there being provided an axial impeller to cause the stream of molten lead to follow a generally spiral path down the vessel without substantial back-mixing, and the said downstream arm comprising a vessel extending to approximately the same height as the upstream arm and having an outlet at the upper end thereof."

The stirring impeller is an important feature of the invention. The specification indicates that a minimum rotation speed is necessary to keep the sulphur in suspension. A range of 100 r.p.m. to 3000 r.p.m. is specified for "steady state operation", i.e. "the body of molten metal in the vessel circulates at a rate approaching that of the impeller". The specification goes on, at the bottom of page 4:

"However, friction at the walls leads to continuous shearing of the streams of metal and continuously introduces the dispersed sulphur to new regions of molten metal."

In the embodiment shown in the drawings the preferred speed is 700 r.p.m.

The shape of the impeller is also significant: paragraph 2 on page 5 reads as follows:

"It is preferred to use an impeller which imparts horizontal rotational impetus to the molten lead, but little or no vertical impetus. Under these circumstances, the vertical movement of the lead in the vessel is controlled mainly by the rate at which it is introduced at the top and removed from the bottom. The stream of lead follows a generally spiral downward path with no tendency for back-mixing. If an impeller is used which imparts a degree of vertical impetus to the molten metal, then other parameters may need to be adjusted to avoid back-mixing."

On page 7, in referring to the embodiment of the invention, the description states:

"the impeller is inclined at only 100 to the vertical so that there is little downward thrust."

The Evidence

Evidence-in-support consists of a declaration by Professor Lawson in which he deposes that the "without substantial back-mixing" feature of the B.N.F. claim "was recognised to be desirable although no practical manner of achieving it in a single vessel had been devised". Professor Lawson presents calculations in his declaration to prove that there would be turbulent flow and thus back-mixing in the B.N.F. reaction vessel. The calculations, firstly relating to the Reynolds Number, take into account the impeller diameter and speed and the density and viscosity of the molten lead. Several approximations are made in the calculations, but even in the extreme case of minimum impeller diameter and speed the Reynolds Number calculated is in the range of turbulent flow.

The calculations secondly relate to the pumping rate of the impeller, and based on the examples given in the specification, show that the pumping rate is greater than the rate of feeding charge into the top of the vessel. Professor Lawson's declaration also suggests that the deep vortex at the surface of the charge in the vessel is a further indication of turbulent flow.

The more relevant exhibits accompanying Professor Lawson's declaration may be summarised as follows:

FL 3 Copy of an article by J.F. Castle and J.H. Richards from "Advances in Extractive Metallurgy 1977" pp. 217-234. This article discloses (in relation to "possible developments" in continuous lead refining processes) that a single stage reactor is a feasible concept, (p. 227) viz:

"A once-through reactor was designed and sulphur addition through a tube or down the stirrer shaft was considered. Even on a laboratory scale, the rate of reaction was very high, which indicated that a single-stage continuous reaction should decopperize lead, provided that the dross could be removed sufficiently quickly."

However the experimental example shown and discussed in some detail is not a top to bottom flow through reactor of the type claimed by B.N.F., even though it does have an impeller near the bottom of a mixing kettle.

FL 4 Copy of Australian Patent No. 482917 (AU13564/76, GB 1524474, previously referred to as prior art) to BHAS and Monash University, inventors: Professor Lawson and Messrs D.H. Ward and R.G. Kelly. This patent discloses a series of stirred reaction stages whereby lead bullion, dross and unreacted sulphur is transferred from one stage to the next, consequently without back-mixing. This system minimises the reversion of copper sulphide to copper which occurs if the reaction is allowed to continue for longer than an optimum period of about 25 minutes. Stirrers produce a vortex at the top of each pot stage, outflow from one pot is via a channel at the top of the pot which runs into the next pot at a lower level. There is a dross separation chamber at the end of the series.

FL 5 Copy of an article by Professor Lawson and Messrs Ward and Kelly from "Advances in Extractive Metallurgy 1977" pp. 185-189. This article indicates that a continuous process is desirable and explains some of the difficulties in achieving such. Graphs are shown of the variation of copper concentration against time during sulphur drossing of molten lead; these show that within a period of 3-10 minutes the concentration level is at a minimum, after which it again rises and approaches an equilibrium value. The article goes on:

"In the batch process, reaction can be suspended at the desired copper content merely by the termination of stirring at the appropriate time. The low viscosity of the lead and the large difference in specific gravity between lead and dross ensure good separation of the reaction products. For a continuous process, however, the shape of the curve indicates that at the desired end concentration of copper in lead, the rate of removal of copper is effectively zero. This means that a simple single-stage reactor in which reaction takes place at the discharge concentration cannot be used."

Professor Lawson and his associates suggest that there are two observations that lead to the setting of conditions for a continuous process; firstly, that the "sulphur should be added when the copper concentration is high to prevent the formation of excessive amounts of PbS", and secondly, that "each element of liquid lead and dross should remain in the reaction system for the same length of time corresponding to the minimum of [the above-said graphs]". The article then discusses development of the series of stirred reactor stages:

"The reactor type in which these conditions can be achieved is the ideal "plug flow" reactor. A difficulty arises in the system under consideration, however, as the great density difference between sulphur and lead, and between dross and lead, makes intense agitation essential for adequate contact of reactants to be achieved. This means that a simple plug flow reactor cannot be used, but a train of stirred reactors through which reactants and products flow concurrently with no back-mixing of reaction products of later stages, has similar material retention characteristics. Such a stirred reactor train will approximate more closely to the plug flow reactor if a large number of stirred reactors are used in the train."

A "plug-flow" reactor is essentially a flow-through reactor where the charge is acted upon or reacts in plugs or sections; the term implies that there is no back-mixing.

FL6 Videotape recording of a perspex model of the B.N.F. apparatus, constructed by Professor Lawson. The vessel contains water, which according to Professor Lawson has a similar Reynolds Number to molten lead, and a dye is introduced near the impeller to show flow patterns in the vessel. Professor Lawson deposes that "the dye ... is rapidly mixed and carried up into the vortex, demonstrating that back-mixing does in fact occur".

The evidence-in-answer consists of a declaration by J.E. Bowers, the inventor. He criticises Professor Lawson's "theoretical approach" and comments, on page 7, firstly, that the reaction vessel of the invention is not a regular shape and secondly that

"even though the flow at and around an impeller itself may be turbulent, lamellar [sic] flow can still occur in parts of a vessel which are remote from the impeller."

Mr Bowers emphasises the importance of the pitch of the impeller. On page 6 he refers to the examples in the B.N.F. specification, in which good results were achieved using a pitch of 100 to the vertical. On page 7 he indicates that an impeller angled at 100 or less

"will produce very little downward movement of the lead and its vertical flow rate will depend almost entirely on the rate of addition to and removal from the reaction vessel."

On page 11, referring to the process being most "efficient" using a 100 pitch, he states

"I believe that this clearly demonstrates that there is an absence of substantial back-mixing in the reaction vessel."

Mr Bowers also refers to back-mixing on page 5; he states:

"I do not claim a complete absence of back-mixing in the reaction vessel ... there is almost certainly some, but it is not enough to make the contents of the vessel homogeneous. It is submitted that the amount of back-mixing cannot be "substantial" because the process does in fact work."

Mr Bowers comments on some of the Lawson exhibits as follows. The vessels shown in FL3 and FL4, because of the stirring and removal of the charge at the top, are such that the charge therein would be homogeneous. With regard to FL5, he points out that even though an ideal plug-flow reactor is mentioned "the use of a single simple reactor of that type was dismissed as unworkable" and "in particular there is no teaching of the way in which such a vessel might be stirred in order to avoid a substantial amount of back-mixing of the contents" (page 5). With regard to the model shown in FL6, Mr Bowers suggested that there were too many differences between his invention and the model for the behaviour of the latter to be significant. He goes on to discuss his own model where cork particles were used to represent the sulphur.

The evidence-in-reply comprises a second declaration from Professor Lawson. He reiterates his arguments from the evidence-in-support, finds discrepancies between the examples in the B.N.F. specification and experiments described in Mr Bowers' declaration, and emphasises that back mixing must occur if there is turbulent flow; he also suggests that "laminar flow would occur only in the region directly adjacent the walls of the vessel". Some further exhibits FL8 and FL9 give details of flow patterns in tanks with low viscosity liquids therein mixed by impellers; substantial mixing at the tips of the impeller blades is evident; I note however that all the examples in these exhibits relate to tanks closed at the bottom. FL7 is a copy of an article by T.R.A. Davey in a monograph book entitled "Research in Chemical and Extraction Metallurgy". This article was used as a basis for the Castle & Richards article FL3, and discloses the use of a single stirred pot reactor, typical of the type which could be used for a batch mode.

The article also discloses that

"the stirrer should produce a pronounced vortex to ensure that the sulphur is rapidly swept into the lead"

and

"tin acts by forming a double sulphide with copper, but kinetic factors also operate, since prolonged treatment, even while adding further sulphur, leads to re-solution of copper."

Section 40

Mr Macaw submitted that the opponent's objections to the B.N.F. specification were as follows:

"There are fundamental section forty problems in the specification. If you can believe the directions as providing the method by which the process is to be carried out then the claims simply can't define the invention, because the directions don't produce something which works as claimed. If you on the other hand believe the claim, that is that you've got all that's happening without substantial back-mixing (assume that has a definite meaning for the moment) then the directions are insufficient, because by following them you don't get it. You can't have it both ways."

Putting this another way, Mr Macaw said that Claim 1 (in particular) merely defines what was known in the prior art to be the perceived ideal; on the other hand, if what the claim defines is truly the invention, there is insufficiency of description through failure to reduce the invention to practice.

He also stated the problem in more specific terms thus:

"You simply cannot have the degree of agitation required to maintain the sulphur in dispersion without substantial back-mixing, or putting it another way, if you had in fact achieved a plug or spiral flow down the path of the vessel, there is simply no way in which you could, in the meantime, have achieved the degree of agitation necessary to maintain the sulphur in dispersion."

Mr Macaw further submitted that as well as Professor Lawson's evidence showing that there must be substantial back-mixing in the B.N.F. invention, the specification offered no proof that there is no substantial back-mixing; e.g. there was no discussion of a tracer test or tests on steady-state sampling at different points of the vessel. Also he pointed to a supposed "admission" by the inventor: the last paragraph on page 10 of Mr Bowers' declaration, referring to Professor Lawson's model, reads as follows:

"It is also noted that Professor Lawson adds the dye at the base of the impeller rather than at the top of (sic) vortex at the upper end of the reaction vessel. Addition at this point of turbulence might be expected to result in some of the dye being carried back towards the top of the reaction vessel".

Mr Macaw suggested that this is an admission by the inventor that substantial back-mixing does occur in his apparatus: but in any case, all the features of claim 1 are well-known except the "without substantial back-mixing" feature, and this feature is the perceived ideal.

On the matter of common general knowledge in the art, Mr Macaw submitted that the documents accompanying the evidence-in-reply i.e. exhibits FL7, FL8 and FL9 were extracted from well-known publications, but that the documents accompanying the evidence-in-support, i.e. exhibits FL3, FL4 and FL5 were not matters of common general knowledge. Thus the ideal solution to the problems of the prior art being a plug-flow reactor was known to Professor Lawson and his associates but was not a matter of general knowledge. Mr Macaw suggested that "what was well-known was that you needed vigorous agitation and that was inconsistent with the ability in a single reaction vessel to achieve a continuous process."

Mr Caine submitted that the use of a single stage reactor under non-homogeneous conditions was not taught by the opponent's evidence, and that the ideal plug-flow reactor was not a matter of common general knowledge. He suggested that Professor Lawson's calculations show only that there is turbulent flow near the impeller, that laminar flow occurs in other parts of the vessel, and that the model shown in FL6 merely shows that there is some back-mixing. In reply to Professor Lawson's comment that the results of the invention are no better than those achieved with Professor Lawson's multi-stage apparatus, Mr Caine pointed out that the reactor of the invention is simpler and less expensive.

In this case I am initially led to the conclusion that the invention is involved with a matter of degree of avoidance of back-mixing. The evidence suggests certainly that there is some back-mixing in the B.N.F. method; also I think it is unlikely that the methods of the prior art would have been completely homogeneous, since the sulphur was introduced into the charge generally at one point; however, since BHAS is opposing grant of the patent, the onus is on BHAS to prove that the invention does involve substantial back-mixing. In this regard I will briefly comment on some of the evidence and submissions.

From an analysis of Professor Lawson's calculations, firstly it appears certain that there is turbulence near the impeller, but I note that these calculations did not take into account any factors involving other features of the vessel, so the flow pattern over the whole vessel cannot be determined. Similarly the calculations of pumping rate do not appear to take all relevant factors into account, for example, the angle of the impeller blades.

The prior art, I conclude, does not teach the practical use of a single vessel flow-through reactor. Professor Lawson's article FL5 does refer to "the ideal plug-flow reactor", but this is truly an ideal only, a theoretical model. The prior art dismissed the single stage reactor and approached the theoretical model with a multi-stage solution, FL4, whereas the applicant's invention is another way of approaching the ideal plug flow reactor, which appears to have some merit in the light of the prior art. The evidence-in-reply relating to flow patterns in impeller stirred vessels does not seem particularly significant to me, since the vessels discussed were open only at the top; thus one would expect a considerable amount of back-mixing in such vessels. The B.N.F. method on the other hand involves continuously "recovering the stream of lead from the lower end of the vessel", and it appears to me that the flow-through action of the inventive process would considerably reduce the amount of back-mixing produced by the impeller.

Despite Professor Lawson's assertions that the Reynolds Numbers of the molten lead in the B.N.F. apparatus and the water in his model shown in FL6 are similar, I do not consider the evidence of his model to be significant. I agree with Mr Bowers that there are too many differences between the water model and the molten lead process of the invention. I have no way of confirming for example whether the physical features of the model, such as the impeller and flow-through hole, conform to those of the B.N.F. vessel. (I note for example that in the model the impeller shaft has some radial wobble, which could contribute to turbulence).

Thus I conclude that although there appears to be some back-mixing in the applicant's method, the opponent has not offered sufficient proof that there is substantial back-mixing. Thus despite the doubts expressed by the opponent, I am of the opinion that the process of the invention is effective, in that the percentage of copper in the molten lead is reduced; this is at least indicated by the examples in the specification. Such effectiveness may be no greater than that of the prior art processes, but it does not have to be, the public is entitled to a useful choice. See e.g. Renlita Products Pty. Ltd. v Lees (1967) AOJP at 3270. However, in this case I conclude that there is some doubt as to how, the effectiveness of the invention is achieved. Mr Bowers' comment, previously referred to, "that the amount of back-mixing cannot be "substantial" because the process does in fact work" is a non-sequitur; there may be some chemical or physical process at work which leads to reduction in the copper concentration.

Notwithstanding the chemical or physical processes which may be occurring within the contents of the reactor it is evident from the B.N.F. specification and Mr Bowers' evidence that several features of the invention contribute to the reduction of copper content in the molten lead. Two significant features are: firstly, the near vertical orientation of the impeller blades, and secondly the continuous top-to-bottom flow through of the contents of the vessel. These two features ostensibly contribute to minimal back-mixing and are the crucial features of the inventor's consideration in his approach to solving the problems of the prior art.

Turning now to the applicant's, claim 1, I note with regard to the first feature referred to above that the orientation of the impeller blades is not a characteristic of the claim; and with regard to the second feature: in the claim there are references to "introducing a stream of molten lead ... to the upper end of a vessel", "feeding sulphur into the stream" and "recovering the stream of lead from the lower end of the vessel", but it is not clear how these features relate to the "continuous method of removing copper from lead" referred to in line 1; there is some doubt as to whether the "introducing", "feeding" and "recovering" steps are continuous features of the supposedly continuous method.

In fact, claim 1 is broad enough to include any means of creating non-homogeneous conditions without substantial back-mixing within the vertical stirred reaction vessel, whereas the invention described achieves this condition by the use of apparatus specifically designed for the purpose. Thus I consider that firstly, claim 1 is not fairly based as it fails to define as a limitation the first feature referred to above, (see e.g. The Mullard Radio Valve Co. Ltd. v. Philco Radio and Television Corp. of Great Britain Ltd. (1936) 53 RPC 323 at 347); and secondly, claim 1 does not clearly define the second feature referred to above.

Conclusion

I find that the opposition succeeds on the grounds of non-compliance with Section 40. There is, however, patentable subject matter disclosed in the specification and consequently I allow B.N.F. 60 days from the date of this decision to propose amendments with a view to causing the specifications to claim that matter.

I award costs against the applicant.

(J L Roveta)Supervising Examiner of Patents

10 OCT 1988

Patent Attorneys for the Applicant: Edwd. Waters & Sons Melbourne

Patent Attorneys for the Opponent: Griffith Hack & Co., Melbourne