England and Wales High Court (Patents Court) Decisions

IN THE HIGH COURT OF JUSTICE March 26, 1998

CHANCERY DIVISION

PATENTS COURT CH 1996 H. No. 51

1.   This is the official judgment of the court and I direct that no further note or transcript be made

Before: THE HON MR JUSTICE JACOB

Between

HOECHST CELANESE CORPORATION

(Plaintiff)

-and-

BP CHEMICALS LIMITED

(Defendant)

Mr Anthony Watson QC and Mr Andrew Waugh (instructed by)

Messrs Taylor Joynson Garrett) for the Plaintiff

Mr David Young QC and Mr Guy Burkill (instructed by)

Messrs Bird & Bird) for the Defendant

2. Hearing Dates: 9-20 February 1998

3. Judgment Date: 26 March 1998

2. Hoechst called Dr. Porcelli, Prof. Osborn and Dr. Torrence as viva voce witnesses. Dr Porcelli was an enthusiastic but ultimately fair witness. In addition to his expert evidence he gave some factual evidence which was not challenged. Dr Porcelli did not have experience of the prior art Monsanto process. The company for which he was a senior research chemist, Halcon, were not licensees of Monsanto. Nonetheless he had a good deal of knowledge of the kind of process chemistry concerned, himself having been concerned in devising a process for making acetic acid at around the date of the patent. Prof. Osborn is currently Director of the Chemistry Research Centre of the Université Louis Pasteur, Strasbourg. He gave his evidence fairly, holding strong views as a scientist on a hypothesis advanced by Dr Watson for BP. He had been a consultant to Rhône Poulenc in relation to catalytic carbonylation processes but was not privy to details of the Monsanto process, even though Rhône Poulenc were licensees of it. Dr Torrence was one of the inventors of the patent in suit. He struck me as being very straightforward. Hoechst also relied upon a Civil Evidence Act notice in respect of a sworn statement by M. Perron, a Rhone Poulenc research scientist who was closely concerned with that company’s operation of the Monsanto process. There was also such a notice in respect of a Mr Davis who essentially exhibited Monsanto’s "lesson plans" for their licensees. Finally there was reliance by way of such a notice on some statements in documents emanating from BP.

3. Mr David Young QC, for BP, at an early stage in the proceedings, pointed out that large parts of Prof. Osborn’s expert report overlapped with that of Dr. Porcelli. Both, for instance, gave evidence about the teaching of the patent in suit, the cited prior art and so on. Mr Young asked for a direction that he should not have to put his case on the overlapping matter to both Dr Porcelli and Prof. Osborn. I went further. I directed the plaintiffs to delete the duplicated subject matter from Prof. Osborn’s expert report before he was sworn. Parties in patent (and indeed other) actions are not entitled to prolong proceedings by repetitious expert evidence. Courts do not decide cases by counting how many experts a party can get to say the same thing. There would have been no point in my admitting the duplicated material and acceding to the suggested direction that the case on overlapping matter need be put only once - what would have been the value of the uncrossexamined duplicated material?

4. BP’s viva voce witnesses were Dr. Watson, Mr Smith, Dr Clode and Mr Laughton. Dr Watson was called both as an expert and as witness of fact (on the infringement issue). He was by 1983 the day plant superintendent for BP’s acetic acid plant in Hull (the "A4"). He gave his evidence fairly but was necessarily hampered in relation to the issue of obviousness because he was not concerned with research, and particularly the potential use of stabilisers, until 1989, five years after the priority date of the patent. His evidence in relation to the operation of the plant did not suffer from this defect and in relation to that I found him a reliable and balanced witness. So also were Mr Smith and Dr Clode, both of whom I thought gave careful answers even though their cross-examinations were unnecessarily prolix and must have been tiring. Mr Smith was a graduate chemist who became the A4 plant superintendent in 1986 with promotion to senior plant superintendent in 1989. It was part of his function to supervise operators and to ensure that the plant operation was monitored on a day to day basis. Dr Clode took her doctorate in 1989 and, having joined BP then, moved to join the Close Plant Support team in 1993. The CPS team consisted of about 6 people, including both chemists and chemical engineers whose job was to troubleshoot plant problems. Their function included overseeing operators to see that tests were carried out, and samples taken, as instructed. Mr Laughton was a graduate chemist who formed part of BP’s acetic acid team when it first took a Monsanto licence in 1978. He was Group Superintendent for the A4 and CO plant until 1989 when he took over responsibility for the A5 plant which makes acetic anhydride and also consumes CO. BP also relied on some Civil Evidence Act notices relating to discovered documents. Both in the case of BP and Hoechst the material matter contained in such documents were in effect put in evidence at the same time as they were put to witnesses.

5. The patent is for a process for making acetic acid (CH-3COOH or HAc) by carbonylation of methanol (CH3OH, or MeOH) by carbon monoxide (CO). The reaction can be expressed very simply:

CH3OH + CO CH3COOH

6. But things are not that simple. Methanol does not simply react with carbon monoxide. A catalyst is needed and there are a number of intermediate reactions which go on to achieve the overall result expressed in that equation.

7. By the priority date the manufacture of HAc by carbonylation process was well-known. It was first developed in Germany, particularly by BASF who used a cobalt iodide catalyst. Process conditions were very stringent. In the 1960’s Monsanto made a major improvement by replacing the catalyst with another Group VIII metal, Rh. Process conditions were less stringent so that temperatures below 2000C and pressures below 35 bar could be used. At the same time the yield moved close to 100%. But even so the process could not be run at its theoretical maximum rate.

8. The improved process was patented and widely licensed by Monsanto. There was no better way to make HAc from then on and anyone who built an HAc plant took a licence and used the Monsanto technology. 60% of the world demand for HAc is made by the Monsanto process. The Monsanto technology became part of the common general knowledge of those skilled in the art. Some of this knowledge (detailed know-how rather than the information contained in the Monsanto patents) was designated confidential. This potentially raised the problem of how patent law treats "secret common general knowledge". Fortunately the parties were agreed that I should treat the details of the Monsanto process, so far as relevant to this case, as part of the common general knowledge of the skilled man. I am sure they were right to do so, not only because it would be very artificial to ignore the mental equipment of what BP called "the entire universe of skilled practitioners", but also because it seems that most, if not all, detail as is relevant to obviousness had leaked out to those who were not Monsanto licensees: certainly Dr Porcelli, called by Hoechst, was an "outsider" but knew some of the detail.

9. I therefore turn to summarise the basic knowledge of the Monsanto process which a skilled man at the date of the patent would have had. First there are some basic chemical equations. These describe what is going on in the reactor vessel, both in terms of the equilibria and the cycle which produces HAc. Apart from the ultimate production of HAc by the carbonylation cycle there is a "rival", unwanted, process which also takes place. This is called the water gas shift reaction (WGSR), the net effect of which is to produce hydrogen and carbon dioxide from carbon monoxide and water. The parties agreed the various reactions and equilibria as set out on App. 1 to this judgment.

10. Present in the reaction chamber is a melange. It includes Rh which can be in state I or III and temporarily combined with iodine (I) or other components, water, methyl acetate (CH3COOCH3 or MeAc), water, methanol (MeOH) methyl iodide (MeI), hydrogen iodide (HI) and carbon monoxide (dissolved). In fact the iodine will largely (see later) be present in the ionic form, (I-) The process involves the complex interactions between all these components. You want to so arrange things that you reduce the WGSR in favour of the carbonylation cycle. To keep the process going continuously there is a constant feed to the reaction chamber (which operates at high temperature and pressure) of the primary ingredients, CO and methanol. The Rh and I are recycled within the chemical process. Because the materials are corrosive, you also get some corrosion metals (CMs) from the reaction vessel, particularly iron (Fe), cobalt (Co), molybdenum (Mo) and nickel (Ni). At the date of the patent CMs were seen as unwanted, causing the Rh precipitate to be in an undesirable amorphous form. Steps were taken to remove CMs periodically for that reason.

11. The reaction chamber operates at above 1800C and about 30 bar (I do not give exact figures). Into that chamber is fed MeOH and CO as primary feedstock. Recycled catalyst and some recycled material from the purification stage is also fed to the chamber, together with water. The chamber has two "take-offs". First, there is the primary takeoff of liquid material via a valve to a "flashing chamber". The pipe to the flashing chamber is called the "flashing line". And there is a gas take-off to remove CO2 and H2, the product of the unwanted WGSR. This take-off is also used to control the pressure, the partial pressure of CO of course affecting its rate of dissolution in the liquids. The chamber is agitated in order to make such dissolution take place quickly - I accept Dr Clode’s evidence that without agitation the reaction would stop because it would be starved of CO. The whole reaction is exothermic.

12. The flashing chamber is at a much lower pressure than the reaction chamber. The entering reaction mixture undergoes an adiabatic flash producing liquid and vapour phases. The vapour contains primarily the lighter components, namely MeOAc, MeI, H2O, and the desired HAc. The liquid will also contain some of these lighter components (particularly some HAc which is heavier than the other "lights") and will also contain the involatiles, namely the Rh (the catalyst) and the CMs. The liquid is recycled to the reaction chamber. The vapour phase is then fed to a distillation column. The MeI and MeOAc, along with some water and HAc form the light ends from this column and are recycled. The HAc (which still contains water) is withdrawn from it and goes on to additional purification (including drying).

13. Water is necessary for the desired reaction cycle. Prior to the date of the patent Monsanto and its licensees always used about 14% wt. or above. So much appears from Monsanto’s "lesson plan". It is corroborated by what happened at BP and, independently, at Rhone Poulenc. There was an obvious incentive to use less water if one could for several reasons. First, there is the problem of what happens in the flashing chamber. The more water you have the bigger the flashing operation has to be - you are essentially flashing a component which you do not want. Second there is the problem of the wet HAc: the wetter it is, the more it costs to dry.

14. Why, then was less water not used? It was a problem with the Rh. It had been found that as you decreased the water two things happened to the Rh in the flashing chamber. It actually precipitated out onto the walls and even that which stayed in solution somehow (the chemistry was not precisely understood) lost activity. Rh is a very expensive metal. Some could be reclaimed but this was an expensive operation. And anyway there were Rh losses. Of course there is a trade off between Rh usage and HAc output but the trade-off was widely seen and accepted as happening at 14%.

15. The level of the intermediate, MeOAc, was also seen as important. It was perceived as having an effect on Rh precipitation. Monsanto taught that the level of MeOAc should not be more than 1%wt. BP (licensees at the time) were prepared to go up to 1"% but more was seen as causing too much Rh precipitation.

16. The patent is aimed at a variant of the Monsanto process which uses less water. It is hence called the "low water" patent. It begins by recognising the known Monsanto process, carbonylation of methanol using Rh catalyst with MeI as a catalyst promoter. It goes on to point out that water was seen as necessary to achieve a high reaction rate, a level of 14-15% being typically used. A paper by Hjortkaer and Jensen (not suggested to form part of the common general knowledge of a man skilled in the art) is cited as showing that from 0-14% water increases the reaction rate, but not above that level. The patent then goes on to recite the problem of catalyst precipitation, and says:

4. After citing various other pieces of prior art, the patent then makes its main promises:

5. The patentees then say:

17 I therefore now set forth claim 1, using symbols, adding some numbering for convenience.

(1) a quantity of water which is at least 1% but below 14% by weight of the reaction medium together with

(2) effective amounts of

(a) a catalyst stabiliser selected from iodide salts which are soluble in effective concentration in the reaction medium at the reaction temperature, the concentration of the iodide salt in the case of LiI being 2 to 20% by weight or in the case of other iodide salts being such that the molar equivalent iodide concentration is the same as that obtaining with the said LiI concentration,

(b) MeI in a concentration of 5 to 20% by weight, and

(c) MeOAc in a concentration of 2 to 5% by weight,

18. Claim 5, which BP accept is valid, is dependent on claim 1. It limits its generality so that the upper limit of water is 4%, the catalyst stabiliser is LiI at 10-20% by wt., and the MeI concentration is 14-16%.

19. The patentees then tell the reader about the iodide ion, I- They say, immediately following the statement of invention - which to my mind emphasises to the skilled man the importance of I- :

7. The patentee is here focusing on the significance of a higher level of the ester, in practice of MeOAc. After describing the system in general (reactor, flasher, MeI-HAc splitter column) and a certain amount of repetition he says:

8. He illustrates his point by way of examples at high water (14%) and low water (4%). He concludes from these:

20. That assertion is true. No-where in the cited prior art is any significance attached to the MeOAc level: as I have indicated Monsanto taught it should be at 1%. There is no indication that anyone was prepared to run at 2%, the lower limit of claim 1. Moreover the general view of those who operated the Monsanto process was that a level of 1"%wt. was the upper limit.

21. It is true that the patentee does not investigate to any great extent the extent of synergy with levels of water above about 8%. It is implicit in his teaching that there will be such synergy - but to a lesser extent than at 4% or less. I accept Hoechst’s evidence that this is so and there is no sharp cut-off at about 8% water as suggested by BP.

22. I need say little more about the specification at this stage, save to point out certain further passages which again emphasise the significance of the presence of ionic iodide, as opposed to iodine in some other form. At one point he compares LiI (which he describes as "a representative metal iodide salt") with NMPI, "a representative salt having an organic cation". He concludes that there is no difference for the same molar concentrations and says:

9. And, in his discussion of example 1 he says:

10. Example 2 uses NaI. The patentee says:

12. Table V includes a variety of iodides, including both inorganic organic cations. There are 22 in all. FeI2 (described as "partially soluble"), CoI2 ("soluble") and NiI2 ("insoluble") are all included. The cations of these 3 iodides are corrosion metals. But very large quantities (compared with the amount present in a continuous commercial reactor) are present in the examples.

13. Finally on the significance of the iodide the patentee says this of his figure 9:

23. The patentee’s conclusion from his fig. 9 is of some significance because Mr Watson asked me to draw a different conclusion - that the patentee was unconcerned about the nature of the cation even if it withdrew iodide from the system. I do not think the skilled man would so read fig. 9 in the light of the passage I have just quoted and the remainder of the patent’s teaching about iodide.

24. From fig. 10 of the patent, the patentee emphasises the synergy between LiI and MeOAc. He concludes:

14. This claim is made for the whole range of water up to 14%, as shown in fig. 10. The effects are greater, however, below about 10%.

26. BP say the words mean that the parameters must be kept within the limits of the claim for a significant period of time in the context of the reaction as a whole. A short term fluctuation into the range will not alter the fact that it is being "maintained" outside the range. In relation to MeOAc, for example, "maintained at 1.2 to 1.7" is consistent with short fluctuations ("excursions" as they were sometimes called) above 1.7 or even above 2.0.

27. The patent gives little explicit guidance on the point. In relation to Ex 1 it says:

15. BP suggests the passage helps them. I do not see why. The fact that ranges are given suggests no more than the patentee recognises that reactor conditions do fluctuate, a fact which the skilled man would in any event have known through operation of the Monsanto process.

28. I think there are several difficulties with the BP approach. First, over what period is one to take an "average"? Suppose the reactor is run for a year with one or two days of excursion. On the BP approach there is no infringement. But suppose those two days are at the beginning of the year. If one looks just at the first two days, all the operation is within the specified parameters so there is infringement. How (and when) can it cease it be an infringement because for a later period there is operation outside the parameters? Second, the approach is close to introducing a subjective element: at what level did the alleged infringer intend to operate? That cannot be a relevant matter. Third I cannot think of any reason why the patentee would want to exclude any significant commercial production within his specified parameters. Doubtless he would not be interested in truly transient "spikes" but anything as much as a day’s production he would surely wish to catch.

29. Accordingly I think Hoechst are right on this point. Since the measurements were taken daily, I propose to approach the question of infringement on the basis that, if, on any particular day, BP were within the parameters, they infringed on that day.

30. The problem here arises from the presence of CMs. These are in the solution and build up with time. Steps are taken from time to time to reduce them. When in solution they form iodides. Hoechst say the claim includes all iodine in solution. BP say the claim excludes any corrosion metal iodide. Alternatively they say it only includes corrosion metal iodide to the extent that that iodide is free to take part in the reaction.

32. On the other hand, I think for exactly this reason, that the Hoechst contention is wrong. The patentee is not interested in iodine unavailable for taking part in the reaction. So if you have some of the iodine " stolen" (by being taken up in a complex with rhodium as suggested by Dr Watson or in a hydrate as suggested by Prof Osborn or for some other reason) then it is not useful and does not count for the purposes of the claim. I am reinforced in this view by the actual structure of the claim. This focuses first on "effective amounts". Although the patentee is concerned with amounts in this phrase he is also focusing on effectivity. Unavailable iodine is not effective. Further, the whole thrust of the claim is that in the case of alternatives to LiI one should have a molar equivalent amount of iodide. This was agreed to be 149 millimoles of iodide per kilogram of reactor solution. The claim could have been written more simply with just that limitation instead of the reference first to LiI and then equivalents of other iodide salts. Hoechst’s contention is that it does not matter that provided you have 149 millimoles of iodine it does not matter if some of that has been stolen. This makes no sense. It would mean that the claim covered different amounts of effective iodide, depending on the degree to which a particular cation was responsible for theft. In the case of LiI the problem does not arise, because Li is honest and does not steal. So also for many other cations. But the position of CMs is different, as I shall explore in more detail on the question of infringement.

33. I should mention one further point here. When measuring iodide the total amount of iodine was regularly measured by BP. They were seeking to avoid the claims of the Daicel patent which refers to "iodide". I do not see this as relevant to construction. Moreover, even if it were, it would be sensible to measure total iodine for in that way one could be sure that one was outside the claim. If some (but you don’t know how much) iodide is being stolen you have a margin.

34. My construction may pose a difficulty of fact: if you have a cation which steals iodide how are you to know how much is effective iodide? It seems to me that in the end the question boils to one of fact and onus. It is for the patentee to show, on a balance of probabilities, that there is 149 millimoles or more of iodide acting to stabilise. If he cannot show this then he fails. It is not good enough to show that there are 149 millimoles of iodine. The problem only arises with "stealing cations" - most do not do this.

35. BP say claim 1 is not "new" by reason of UK patent application No.1,146,637 (Daicel). This has an earlier priority date (September 1983) than the patent in suit (May 1994) but was not published until after the latter date. By virtue of the operation of s.2(3) of the Patents Act 1977 it forms part of the "state of the art" but, by virtue of s. 3 may not be relied upon for obviousness. The only question is whether Daicel is novelty-destroying.

"for the purpose of practical utility be equal to that given by the subsequent patent",

38. BP say that the "clear and unmistakable directions" rule must not be applied mechanistically. It is enough if the earlier material must be "as good for practical purposes" as the claim of the patent, but the court is entitled to rely upon what would reasonably be inferred by the skilled man.

42. Diacel’s specific examples are all of batch operation. As such they cannot fall within the Hoechst claims because they are limited to continuous production ("maintained"). Daicel is of course really concerned with a continuous process. So the real question is what does he teach the skilled man about process conditions for a continuous process. The answer is not enough to be a clear teaching to do something within the claim.

43. Attention was focused on examples 29 and 30, particularly the latter. I will assume that BP are entitled to do this, even though Hoechst say "why pick on these?". The example involves putting a charge into an autoclave, heating whilst stirring over a period of 15min to 1800C and, it seems, stopping the reaction when .6 to .8 moles/litre of unreacted starting material remained. The contents of the charge are specified. Information is given as to the final contents of the batch reaction (i.e. when it was stopped). BP calculate from the information given that the MeOAc was 3.6%wt. that there was the equivalent of 13.4%wt. Li and a water content of 5.9%wt. They suggest that Daicel teaches that a continuous reaction should be run under these conditions. I just do not see why. The stopping point appears to be arbitrary and Daicel does not say "the end points of my batch experiments are what you should use for a continuous reaction."

44. What he does say about continuous reactions is to my reading contrary to any suggestion that this is even implied. He says:

45. To my mind this is all saying that one should run a continuous reaction at 1 mol/l, even though the batch examples had a higher "alcohol" (as defined) content. I see no direction to run continuously at the "batch" levels.

46. There are other arguments (set out in Hoechst’s closing submissions), which also have force. I think Daicel is a long way off from providing clear and unambiguous instructions to do something within the claim. It is at best a cryptic "signpost" to the Hoechst invention.

47. Singleton (EPA 055,618A) was published nearly two years before the priority date of the patent. It was a Monsanto patent and it was widely read as soon as it was published. BP scientists had it, though none of those who read it (Drs Ray and Stringer, at least) were called. BP suggest that its teaching made a low water continuous process within the claim obvious. One must approach such a suggestion with care because in fact it did not lead anyone (including those at BP) to do anything within the claim. Further, there is an oddity in the argument: if right it would almost certainly apply to claim 5 also, yet BP accept this is valid.

48. Crudely the case based on Singleton is this: Singleton teaches that you can stabilise the catalyst (particularly in the flash chamber where precipitation occurs) in the Monsanto process by the addition of a stabiliser. Singleton refers to a large range of these. Two examples use LiI and KI.. So you would research with these sooner or later And in any event you would realise that if the catalyst was stabilised you could use lower water because it was only fear of instability which kept you above 14%wt. You would also soon realise (discover) that you could use higher MeOAc levels. And hence you would come within the claim. It is not suggested that you would realise there was a synergistic relationship between iodide and acetate levels. It is suggest you would find this bonus out - stumble upon it - simply as the result of doing what was obvious from Singleton.

49. Being more specific, I can take the summary of the Singleton disclosure from Dr Porcelli’s report:

17. All of Singleton’s examples use water in the 14-17% range and there is no explicit suggestion that you can depart from Monsanto’s standard conditions in this regard. Nor is there any mention of the MeOAc levels, though they were probably of the order of the standard Monsanto level of 1%. There is no mention of the impact of the additives on the rate of reaction or on the decrease in loss of selectivity to by-products.

50. Further I think Dr. Porcelli further comments accurately in saying:

52. That the use of a stabiliser (still less an iodide) enabled you to run at low water, high MeOAc with less WGSR is not pointed up by Singleton. I do not think it was obvious to the skilled man. I think there is ample confirmation of this. So desirable is the use of low water and greater efficiency that it would surely have been the case that Monsanto and its licensees would immediately have reacted to that which is allegedly obvious. Some may have been able to put the idea into practice immediately. Others would have set about altering their plants so as to make use of the major improvement which had been made possible. Yet none of this happened.

53. At BP, for instance, some investigation of a stabiliser was made as a result of the Singleton disclosure. They investigated an imidazole iodide stabiliser. This was also a prospect for BP’s acetic anhydride plant (the A5 reactor). The details of what happened are a little murky. It seems a tarry substance was formed. BP say that at the time their real interest lay in the A5 plant. Moreover since it and A4 shared the CO source which at the time had limited output, the need to improve A4 was not great. The need only arose when there was an adequate CO supply, which was not constructed until 1989. It is not shown that Singleton led anyone at BP to investigate a process within claim 1 of the patent. In particular, even if the iodide level produced by the imidazole additive was high enough, the allegedly obvious consequential step of raising the MeOAc and lowering the water was not taken.

54. Mr Watson fairly comments that those most central to BP’s research in this area were not called. BP relied upon the evidence of Dr Watson, who necessarily was giving hindsight evidence and Mr Laughton whose evidence boiled down to a series of excuses as to why BP did not do (or see) that which they now say is obvious. Cross-examination of Mr Laughton revealed the true situation:

55. Mr Laugton also was not really able to explain BP’s reaction upon learning of the Hoechst invention from the patent in suit. He wrote on 13th March 1986:

24. Mr Laughton wanted the "patent situation" clarified - self-evidently because he now wanted to work within its scope. Mr Laughton suggested that he wrote this memo as a stick to beat management harder to invest in increased CO production. I am sure that this is so. But that does not explain why such a memo was not written when Singleton was read. If BP were right about obviousness then the enhanced "potential for acid plant increases" would have been seen then, and the stick would have been used then.

25. The question has only one answer: that the invention was not obvious. The patent is valid.

57. This involves, on my construction, the question of whether on any given day BP operated within the parameters of the claim. There is no dispute that they were using water at below 14%wt during the relevant period (though they only started doing so after knowledge of the patent). Nor is there any dispute about the MeI concentration. Attention focuses on the iodide catalyst stabiliser and the MeOAc levels. BP clearly thought the patent might be valid and did not want to fall within claim 1. But did they do so, for all or part of the period in question? BP say not. Their basic position is set out on a graph called "A4 Reactor Components" (App.2 to this judgment). Broadly this shows the MeOAc level running at about 1.5% but with fluctuations which include spikes above 2%. The LiI level was always below 2% save for about a month (19th April to about June 1st, 1994). During that period the measured MeOAc level was always well below 2%. There was no time when both parameters satisfied the claim. If right, this graph shows that BP did not infringe.

58. Before dealing with the detailed facts, there are two points I should mention, one of procedure and one of law. Each are predicated on the basis that infringement for at least one period of time is established or admitted.

26. Now it is easy enough to see how the maxim applies in the case of the assessment of damages for an established wrong. The wrongdoer, by hiding the consequences of his wrongful act, cannot be heard to say "you show what those consequences were". The man who steals a jewel can hardly complain if, when he fails to produce or account for it, the court presumes it to have been "of the finest water." But it is another thing to say the maxim applies as some sort of rigid rule of law to establish wrongdoing itself or a greater extent of wrongdoing than has been proved. Take this case. Suppose BP infringed for a period of a day or a week in a year. Why should the principle by which their liability for the remainder of the year be different from the case where there was no infringement proved at all? Hoechst seemed to be submitting that proof of a little bit of infringement cast the onus on BP to establish that there was not a lot more.

63. I have some difficulty with this. It does not depend on any proof of infringement at all. Graham J seems to be creating a duty ("obligation") to keep records of non-infringement. Patent law does not create any such duty. It merely marks out a forbidden territory and says you are liable if you enter it.

27. And, a little later:

68. Up until this last paragraph Judge Newman is saying no more than that inferences arise where a defendant has destroyed documents. Her last passage, which refers to a positive duty to keep and preserve records, relates to the period during which the proceedings were pending. The duty to which she refers is one imposed on a litigant, not one imposed by patent law.

69. Turning to the present case I cannot see that any inference should be drawn against BP. There is no suggestion that any records were destroyed. It is merely suggested that for some of the time their methods of measurement were inept. In respect of this period there is no suggestion that BP were guilty of any dishonesty or sharp practice by way of blind-eyeing or otherwise.

70. BP added LiI to the A4 reactor at various times. They did so with a view to getting some stabilisation. But did they add so much as to bring the level within the claim? If one ignores the CMs the answer is clear. They did for a short period as I have said. If however one were to include all the iodine associated with CMs then essentially BP were above the level for virtually all the period of alleged infringement. For quite a lot of that time the CMs alone would have taken the iodide level into the claim (above 149 mmol/l). So much appears from Dr. Porcelli’s Appendix 11. So infringement turns on construction. Now I have held that the claim is talking about a stabiliser which, if it is not LiI, is an iodide salt which provides an equivalent amount of iodide to that provided by LiI. If I am right then one must ask how much iodide was in fact made available by the CMs?

71. The short answer is "some, but not a lot, and we do not know how much." I elaborate. The witnesses who knew most about the matter were Dr Torrence for Hoechst and Dr Watson for BP. Unlike these two, Drs Porcelli and Osborn did not have the experience of actually operating the Monsanto process. The two witnesses did not differ in their views of what the effect of the CMs was.

31. Q. It steals iodide which would otherwise be available for stability. Is that correct?

A. It steals HI.

32. Q. Yes, which would otherwise be available to stabilise the rhodium.

39. Q. Dr. Smith says destabilised.

74. He was cross-examined on this:

75. Dr Watson attributed the observed facts to the Rh forming a complex with the CMs. Prof Osborn excitedly disputed this. He explained the observations as due to a hydrate being formed. I really have no means of deciding who is right. I do not think it matters. The practical Dr Torrence simply referred to "stolen" iodide or "unionisable iodine" - why does not matter.

76. I should mention that Hoechst complained that BP had not given discovery of documents relating to the stabilising effect of CMs at low levels. However Hoechst did not specifically ask for this before trial and it so happened they adventitiously had some BP documents about this arising from another case. BP did not object to their admission. Since Hoechst place much reliance on them, I quote relevant passages. The first document is a proposal about the A4 plant. It says:

And

50. In fact, of course, the proposal to operate at up to 5% LiI was never implemented, I expect because of the patent.

77. The other document is a fax of advice to BP’s Texas City plant. It says:

51. All of this is consistent with the evidence of Drs Watson and Torrence. It is not consistent with the proposition that the CMs can provide iodide as effectively as Li.

78. In the result though the CMs in BP’s process will provide some stabilisation, I am unable to say how much. Certainly it is unlikely to be anywhere near the amount suggested by Dr Porcelli’s Appendix 11, and, in view of Dr Torrence’s evidence about his inability to provide a comparison with LiI I have to conclude that Hoechst have not proved that BP’s iodide levels fell within the claim except for the admitted period.

79. What BP tried to do was to keep well below the 2.0% level of MeOAc set by claim 1. They set a maximum value of 1.7% - a level at which action was to be taken 1.6 was a "warning" level and the mean value aimed at was 1.4%.

80. Hoechst strained every nerve to show that BP had severely underestimated their MeOAc levels. As a consequence a great deal of time at the trial was taken up with this point. Since this was so central to Hoechst’s position, it is surprising that so much of their case seems to have been developed late in the day. And it is all the more surprising in view of the cryptic answers provided by Hoechst in response to BP’s request made on 20th November 1997. BP asked for admissions to the effect that their measurements were accurate and did not suffer from systematic error. They specifically asked whether Hoechst intended to allege that any such measurements did not provide accurate results or suffered from systematic error and asked for the nature of Hoechst’s case if they intended so to allege. The answers (given over a month later) were cryptic in the extreme:

81. The result of Hoechst’s failure to respond properly to this perfectly reasonable request is that BP (and for that matter Hoechst) have had no opportunity to do even simple experiments. And the experts on either side have been invited to speculate with the result that to a large extent I am faced with speculation and counter-speculation. In all this I remind myself that the onus of proof lies on Hoechst.

82. It is first necessary to describe in general terms how BP measured the levels. App.3 to this judgment is a diagram showing the apparatus used. A T-junction pipe is provided in the flashing line to S.101, the flasher. This passes through some valves (not relevant), through a chiller and goes, via a 3-way valve, into the sample chamber ("bomb"). The other port of that valve is connected to a pipe which can take sample, via the chiller, to the sample cabinet. On the other side of the bomb is another 3-way valve which is linked to the first such valve. There are two ports to that second valve. One is connected to a pipe which goes back to S.101. The other is connected to a source of nitrogen.

83. Normally the 3 way valves are set so that the material simply circulates through the system (chiller - bomb - S101). It flows fast. When one wants to take a sample, the valves are turned simultaneously. The material in the bomb is pushed by the inflowing nitrogen through the chiller again and comes out of a needle in the sample cabinet. That needle is inserted in a sample bottle. It is pushed through a self-sealing membrane (called a "septum"). The septum is fitted as a seal inside a screw-on cap, the cap having a central disk removed. A sample bottle and cap are in evidence. After the sample stops flowing there is a "nitrogen purge", after which the valves are turned again and flow from the flashing line, through the system, back to S.101 resumes. The bottle is taken to a laboratory at some distance away and in due course (on average about 2 hours later) it is analysed for MeOAc and MeI. The analysis is by gas chromatography ("GC"). Dr Clode pointed out that the GC apparatus is not reliable to more than a second significant figure. So a figure of 1.8 should really be regarded as 1.8.1. If you have several readings taken near together at about the same time and they about the same number then that number is more reliable.

84. I begin with systematic errors said to be due to continuing reaction after the liquid leaves the reactor. Three suggestions are made: reaction in the flashing line itself, reaction in the sample line (i.e. the pipework from the flashing line until the liquid is chilled) and, so far as the latter point is concerned, exacerbation because trapped sample may have been allowed to react in the pipework whilst the operator was emptying a bottle. The suggestions have emerged over time. Prof. Osborn raised reaction in the flashing line in his report, Dr. Porcelli reaction in the sample line and, very late in the day (after Dr Watson had given evidence) the reduced flow rate point was raised.

85. Now any sampling system is bound to have pipework for taking the sample, whether the sample is analysed on-line or off line. It is not suggested that the length of BP’s pipework is atypical. According to Dr Watson it was much the same sort of length as recommended by Monsanto (BP started with an on-line analyser as recommended by Monsanto but it clogged up - as similar ones did - and needed cleaning). So the skilled man would expect a certain amount of continuing reaction. The patent says that the contents of the reactor are analysed and the figures given in the patent must be based on analysis of samples taken from the reactor. This must be done by extracting some of the contents, continuously or periodically. One simply does not know whether any continuing reaction is or is not significant. The temperature may fall in the sample line and temperature has a significant effect on reaction rate. You will need dissolved CO to react and it is not clear whether any significant amounts of CO will dissolve in the short time available (particularly because, unlike in the reactor, the liquid is not agitated).

86. As to the suggestion that there would be "stale" liquid in the sample line, it is simply not made out. The operators were instructed to take the third of three samples. They would have had to work very fast between samples (i.e. between successive openings of the 3-way valves) to prevent the natural flushing of the system from clearing it. Dr Clode was clear that the taking of stale samples was most unlikely. Moreover, as she pointed out, if the phenomenon were happening it would do so on a random basis and would show up in the results, which it did not.

87. Hoechst further suggested that there was a particular reason why there might be more continuing reaction in the case of BP than in the case of other Monsanto operators. BP had a particular problem with solids caused by precipitation of lead compounds. Their MeOH supply tank was apparently an old leaded fuel tank and provided some lead contamination. This was significant enough for Mr Smith to try (though he did not succeed) in raising the temperature of the material entering the bomb with a view to reducing the tendency to precipitate. Hoechst say accordingly that BP’s pipes might be more blocked than those of others, with consequential sluggishness which might lead to stale samples. It is no more than a speculation. Dr Clode would have none of it, giving the perfectly good reason that if there were any substance in the point, one would find a difference in measurements before and after the pipework was cleaned out, as it periodically was, yet none was found. Moreover it is not as though other Monsanto operators (using an on-line system) did not experience precipitation problems. They did.

88. The next point relates to loss of volatiles or "lights" as they were called. The reaction mixture has in it water, HAc, and other components, including MeI and MeOAc. The latter two compounds are fairly volatile, with MeI (b.p. 42.40C) being significantly more volatile than the larger moleculed MeOAc (b.p.570C). There is a substantial amount more of MeI than MeOAc - about 12-14%wt as opposed to the 1-2% (the latter figure is the one you want to know precisely). Hoechst say that the BP measurement system was likely to lose, and did lose volatiles in a whole variety of ways: evaporation in the bomb, loss through the nitrogen purge, loss through septa, loss through delay in analysis. Each of these matters were explored in minute detail. And it is clear that there were indeed losses of "lights" for some periods, see below. It is unnecessary to go further into the causes for those periods. It may have been one or more of the factors relied upon. The detail does not matter, notwithstanding the prolonged cross-examination about the possible causes. Since there were, by and large, no losses after Dr Clode did her work, it is probably the tightening up of the actual sample taking procedure which prevented losses.

89. For the beginning of the period in question there were disparities between the measured and calculated MeI levels. BP were not getting consistent results, particularly of measured MeI. This appears most clearly from Mr Smith’s exhibit 8 (App.4 to this judgment). Mr Smith expressed the position graphically in a note of 29th April 1993:

59. A. No. It is all point 1. You are looking at the axis of the ----

92. Accordingly even though BP did lose some volatiles over some of the period in question it is not established that measurable quantities of MeOAc were lost. So on the basis of the lost volatiles point the critical level was not exceeded except where BP’s measurements show it was. I exclude "rogue results" as shown on Appendix 2.