British Sugar Plc v Cegelec Ltd

J U D G M E N T

1. LORD JUSTICE HOOPER: on 29th September 1996 the 20MW rated steam turbine driven alternator (or generator as it is also called) at the respondent's Wissington factory failed, with the consequence that the factory had to be closed down temporarily. In 1997 the respondent by letter to the predecessor to the appellant company, GEC Alsthom Electrical Services Ltd, raised a claim for consequential loss. Proceedings, however, were not issued until 2002. On 6th February 2004, by a reserved judgment, HHJ Raynor QC, sitting in the Salford Technology and Construction Court, found in favour of the respondent on the issue of liability. The appellant appeals that decision. One ground of appeal relates to an earlier decision of the trial judge, on Wednesday, 3rd December 2002, not to permit the appellant to rely on much of the evidence of a proposed witness, Mr Regan. The trial had started on Monday, 1st December at which time the appellant unsuccessfully sought an adjournment. The judge however did allow an adjournment until Wednesday and it was on that day that the application to rely on the evidence of Mr Regan was made and rejected. I shall call that ground: the procedural ground. It became clear during the course of argument before us that that was the principal ground upon which the appellant now relies. The other grounds of appeal attack findings of fact in the decision of 6th February, and I shall call those the merits grounds.

2. It was common ground that there was substantial damage to the end windings of the alternator's rotor, which whilst revolving at 1500 rpm inside the stator, creates the electricity. In simple terms, the copper "wire" used in the windings on the rotor passes along a slot on the rotor face and then emerges at the end of the rotor, is bent round and enters another slot to return to the other end. It is then bent around again and this procedure is repeated many times. Those parts of the windings which are at the ends of the rotor are called the end windings. The damage was to part only of the outer end windings, that part forming some 30 per cent of the circumference of the rotor at that point. It was also common ground that five heat sink fins located at the end of the rotor and revolving within the housing had, at some point, broken off. There was no damage to the other 67 fins. There was also some damage to the stator itself.

3. There was no real dispute by the end of the trial that the only two possible causes for the failure of the alternator were: (1) the migration or, to use another word, expansion of part of the end windings (ie 30 per cent) so that it came in contact with, and interfered mechanically with, the stator (the respondent's case); (2) the failure of the five heat sink fins (the appellant's case). Two of the broken fins were associated with one diode (W1) and three with another, W2, which was on the opposite side to W1. All broken fins were at the end of their respective groups of 12.

4. If, and only if, the respondent was able to prove on the balance of probabilities that migration had occurred, then the respondent succeeded, as Mr Sampson conceded at trial and before us today in his attractive submissions. As to the damaged heat sink fins, it was the respondent's case, with which the judge agreed, that the damage had occurred during the dismantling of the exciter following the failure.

5. Although ground 1, the procedural ground, is the principal ground upon which Mr Sampson relies, I shall consider the merits grounds first. I do so only because in my view it is easier to understand the procedural ground once one understands the merits grounds and the factual issues in the case.

6. I turn to the judge's findings and conclusions. In giving judgment the judge said this of the exciter:

"4. The exciter incorporated a rotor, which was attached to the shaft and rotated with it, and a stator which was fixed to the ground. There was a three phase winding on the exciter's rotor armature connected to a series of diodes mounted on a diode hub attached to the rotor armature; each diode was connected to a heat sink -- and there were six diodes and six heat sinks on the hub. The windings comprised a series of electrically insulated turns of copper conductors; at each end of a winding a series of copper conductors leaving one slot were bent around and routed back along another slot, and the areas at the end of a winding are referred to as 'end windings'.

"5. The rotor rotates at a speed of 1500 revolutions per minute and such rotation tends to cause the end windings to splay outward or migrate. Therefore additional bracing or cording was added to the end windings to provide support and reduce potential movement."

7. The generator had been installed in 1991 and had worked perfectly until 1996. In 1993 the appellant had modified the design of this type of rotor. I return to the judgment for a description of what occurred in March 1996:

"On 11th March 1996 the defendant's engineers carried out an electrical and mechanical inspection of the generator. What they discovered regarding the exciter assembly was set out in a report of that date ... Under the heading 'Exciter assembly' it is stated:

'A visual examination of the Exciter and PMG assemblies was carried out ...'

'The PMG was in excellent condition with no signs of overheating or movement. The Exciter Stator was inspected and found to be in good condition showing no signs of overheating or distress. There were blobs of varnish, still in a tacky condition in the bedplate under the Exciter stator. There was no indication where this may have come from. Following the inspection of the Exciter rotor, it was observed that part of the outboard winding adjacent to where the rotor connection leads joined to the winding has begun to migrate in a centrifugal manner causing a part of the outer covering and cording to contact the stator coil washers. This has resulted in two of the cords being completely worn through and two others partly worn through. At this time it was agreed by all parties to call in a CEGA design engineer, Mr S Ross. This resulted in the Exciter rotor being despatched for repair. After the Rotor was rewound the air gaps were checked ...'"

8. The rotor was rewound using Vacuum Pressure Impregnation (VPI) but was not altered to comply with the 1993 modification, which amongst other things allowed a greater space for the end windings.

9. The refurbished rotor was reinstalled in April 1996. Returning to the judgment:

"14. In May 1996 the defendant submitted a report to the claimant (at page 919 of bundle 4); however, no explanation was given as to the cause of the migration and indeed in a draft disclosed for the purpose of these proceedings the author of the report stated that 'there was no indication as to the cause of the winding being larger than the air gap.'"

10. Thereafter the alternator was used at half its maximum rating, all that the factory needed at that time. On 24th September there was a progressive increase in the electrical load to meet the needs at the factory. A fault occurred at 2.15 pm on 29th September, following which the alternator was restarted. At about 5.45 that afternoon there was a further fault and a strong smell of burning was experienced around the exciter. The alternator was stopped. The judge describes what happened next:

"19. On 30th September 1996 the defendant's service engineer, Mr Paul Savory [called by the appellant as a witness], in the presence of various of the claimant's staff, dismantled the exciter so that it could be taken back to the defendant's Bradford works for repair. All appreciated that the work had to be done very urgently, because the claimant's plant was shut down because of the generator failure. The dismantling involved the removal of the stator and its housing -- which required a crane -- and in photograph 494(1) in bundle 2 Mr Savory can be seen manually steadying the stator as it was moved away from the rotor, with another operative operating the crane's control. In evidence Mr Savory described how the crane's motor was (repeatedly) stopped and started and he said (as is no doubt obvious) that the stator movements would have been jerky had he not steadied it. We can see from the photographs (once the stator had been pulled away) that aluminium heat sink fins (which are brittle) had been broken off at some stage (two fins on one side and three on the other: see the photographs in bundle 4 at 867 and 870). At the time, no one would seem to have made any comment about this occurrence, and Mr Savory not only has no recollection of the damage but until he saw photographs disclosing the damage in December 2003, he would 'have sworn that there were no fins broken'. He is adamant firstly that he did not break them in the course of his work, and secondly that had he done so he would (at the time) have noted the event and been surprised. He does not believe that anything that he did could have accounted for the breakage, and cannot say how the damage was caused. Mr Graham Heatrick, the claimant's Operations Manager at Wissington, was intermittently present during the dismantling process; he does not recall there being any aluminium debris in the base of the exciter and there are photographs of the base taken during the dismantling in the defendant's supplemental disclosure bundle at tab 11, photograph 36, and in bundle 2 at tab 4; in none I find is there any aluminium debris visible -- although copper debris is evident.

"20. What was immediately apparent after the stator was removed was:

(i) substantial penetration damage to the end windings in the location shown on the photograph at bundle 3, 527, and

(ii) some damage to the stator including damage to the two stator damper bars shown at 12 noon and 3.00 pm on photograph 495(1) in bundle 2.

"21. The exciter was taken to the defendant's Bradford works for repair and on the night of 30th September and over the one or two days following there were discussions between Mr Heatrick and the claimant's Senior Engineer (Electrical), Mr Richard Morgan, representing the claimant, and Mr T Cooper, the defendant's Bradford Plant Manager, its Bradford foreman, and others; the claimant's case as to these discussions is set out in the Particulars of Claim and in witness statements, and Messrs Heatrick and Morgan gave evidence before me. The defendant has not chosen to call any party to these discussions. Both Mr Heatrick and Mr Morgan were impressive witnesses, whose evidence I accept. I find that in the course of the discussions:

(i) it emerged that the rewind had not been done in accordance with the new design, but in accordance with the old.

(ii) there was an acceptance on the part of the defendant's representatives (and in particular Mr Cooper) that the failure had resulted from a defect in the end windings -- the likelihood being that the use of the original design had caused the problem, Mr Cooper stating that this design involved excessive shaping with a mallet, which process the defendant's foreman said was extremely difficult since there was so much copper to fit into a small space.

(iii) it was agreed that there had been a migration of the windings, and a repeat of the earlier 1996 occurrence -- but this time the problem was not caught at an early stage but resulted in a catastrophic failure.

(iv) there was discussion about the missing heat sink fins (because these had to be repaired or replaced) -- but there was no suggestion that these had contributed to or caused the failure."

11. The judge commented on the failure by the appellant to call those involved in the examination and repair of the exciter following the failure:

"22. The defendant, by its defence of this claim, disputes all of the above matters, and yet has not chosen to call anyone (other than Mr Savory) who was involved in the 1996 works, nor Mr Richard Emery, who was the author of a memorandum dated 4th March 2003 (to which I shall refer presently) and who was heavily involved in the design of exciters for the defendant and was well acquainted with the 1993 design revision (and is employed by the defendant now). It may, of course, be the case that the defendant's representatives were wrong in the views that they expressed, but when weighing all the evidence it seems to me that I am entitled to have regard to what the defendant's employees thought and said at the time."

12. The judgment continues:

"23. At the beginning of October 1996 the defendant carried out the repairs to the exciter (including rewinding) in accordance with the later design, and the generator has worked without problem since then. These works were carried out without charge to the claimant. The defendant suggested (in a letter dated 3rd November 1997) that the work was done in the interests of goodwill and without acceptance of responsibility; however, the fact that no charge was made is consistent with the defendant's stated position at the time of the discussions I have referred to above."

13. The judge then points out that following the respondent's letter of 27th October 1997 setting out the claim, the appellant repudiated liability, stating that the cause of the failure was unknown. In September 2002 the appellant pleaded positively in its defence that the "machine suffered a mechanical failure of the heat sink fins and that debris from that failure damaged the windings".

14. Dr Peter Jowett, an expert witness called by the respondent, concluded that:

"4.4 I consider that, on balance of probabilities, the integrity of the electrical insulation failed because the rotor winding migrated towards the stator, with which it came into contact, damaging insulation on the rotor winding and thereby causing it to fail."

15. Dr Wadsworth, also an expert witness called by the respondent, wrote this as to the failure of the fins:

"9.5 Failures of the diode cooling fins most likely occurred when the exciter was being dismantled rather than when the generator was in operation. Therefore, fin failures were not responsible for the failure of the exciter."

16. Dr Wadsworth continued in his report:

"9.6 On balance of probabilities, failure of the exciter on 29th September 2003 was due to an electrical fault."

17. He attributed the failure to the March/April refurbishment of the rotor.

18. Professor Williamson, an expert witness called by the appellant, looked at various causes of the failure. As to the theory that the cause was the broken fins, in other words the cause advanced by the appellant in its defence, he wrote:

"Against: no record of remains of such an object in the sump."

19. As to the migration theory he noted that the rotor had been rewound in April because of migration but:

"Against: I would have expected the damage to the stator to be more extensive and uniform."

20. His conclusion was that it was not possible to find the cause of the failure with any degree of certainty.

21. The judge roundly rejected the heat sink fins theory.

"34. Having considered all of the evidence:

(a) I reject the suggestion that the damage was inflicted by the heat sink fins or other foreign object -- indeed, I find that it was not.

(b) I find (on the balance of probabilities) that the failure occurred because there was migration of the windings until they interfered mechanically with the stator.

In other words, I accept the expert evidence of Drs Jowett and Wadsworth.

"35. The following are my reasons for finding that the damage was not inflicted by the heat sink fins or other foreign object:

(i) No plausible explanation has been advanced on behalf of the defendant to explain how the heat sink fins could have become detached during the operation of the generator. Dr Wadsworth put the matter clearly in paragraph 39 of the joint statement as follows:

'There is disagreement as to the possibility that the heatsink cooling fins might have fractured during operation. There is no mechanism that IW is aware of that could explain the spontaneous mechanical failure of one, let alone five, diode cooling fins during operation of the generator after such a period of continuous operation. IW and PJ note that no debris from the damaged fins was reported when the exciter cowling was removed, despite five fins being found missing at a later stage. SW is of the opinion that abnormally high speeds will have been reached during the first generator trip, and that this, possibly combined with the violent axial movement of the generator shaft, might have caused the fractures.'

Professor Williamson did at one stage propound a theory that:

'... there will ... have been at least one transient pulse of torque during the "turbine manual trip", and that when transmitted through the gearbox this could have produced an axial movement through the shaft.'

That is paragraph 2.7 of his supplemental report, as amended when he gave evidence. However, as I have said, he made it clear when giving evidence that he did not claim to be an expert on the mechanics of gearboxes and Dr Wadsworth, who did have such expertise, gave convincing evidence, which I accept, demonstrating that the sort of axial movement of the shaft required to bring the end windings into contact with the stator cowling was 'inconceivable'.

(ii) There is no evidence of the consequential damage which I find (accepting the evidence of Dr Wadsworth and applying what (I hope) is common sense), would have been bound to result had the fins become detached whilst the generator was in operation. They would have been travelling at substantial speed (Dr Wadsworth put it at 84 mph) and the opinion expressed by Dr Wadsworth in paragraph 40 of the joint statement 'that the (near) simultaneous fracture of five metal cooling fins rotating at a speed in excess of 1500 rpm should have caused significant consequential damage to nearby items' seems to me to be obviously right. I find that the physical state of the stator cowling in September 1996 was as now (in other words I accept Mr Heatrick's evidence that the cowling was not repainted after the failure -- and I am also satisfied, notwithstanding the suggestion of Mr Regan to the contrary, that it had not been repainted at any time after its initial installation); its state is depicted in photographs and again I accept Dr Wadsworth's evidence that, given my finding as to the physical state of the cowling in September 1996, one can exclude its having been struck by a fin. As previously stated, there is no evidence of any aluminium deposit in the photographs of the base of the exciter -- and there is likely to have been some evidence had the fins been detached when the generator was in operation. Nor is there any evidence of damage to any of the remaining fins -- Dr Wadsworth was of opinion that they would have had to come into contact with the cowling had the same caused the detachment of the fins.

(iii) It plainly never occurred to anyone in September 1996 that the fins were responsible for the failure and that is because (I believe) no-one noted anything to suggest that they were and I find that they were not.

(iv) It is not necessary for me to reach a conclusion as to how the fins became detached (because I am satisfied that they did not play any role in the failure of the exciter). However, on the balance of probabilities I find (notwithstanding Mr Savory's evidence) that they were detached during the dismantling operation. In evidence Dr Wadsworth gave a plausible explanation as to how that could have occurred (involving contact between the fins and stator housing bolts or the rear external component of the stator cores as the stator was being removed with the aid of a crane). Mr Savory is adamant that he would have been aware of such an event, but I find that all were operating under pressure and the detachment of the brittle fins was not seen as in any way significant in the circumstances. It has been said that each of the rival scenarios (viz the detachment of the fins during dismantling and the detachment whilst the exciter was in operation) is unlikely; however, in the event, on all the evidence, I am able to exclude the latter altogether, and find the former proven.

(v) Having excluded the fins as a cause of the failure, there is no evidence that any other foreign body might have inflicted the damage. In his closing submissions, Mr Sampson suggests that there might have been a failure of some other part of the machine 'perhaps the damper plates of one of the stator coils'. This suggestion is simply unsupported by any evidence."

22. Having rejected this theory, he continued:

"36. The following are the reasons for my finding as to the migration of the end windings:

(i) Once a fin or other foreign object is excluded as cause of the damage, then the expert evidence of Professor Williamson (the defendant's witness) is that migration was the most likely scenario.

(ii) There was unexplained migration discovered in March 1996 sufficient to bring the windings in a localised area into contact with the stator. (There is no basis for suggesting that the fault progressively developed over the preceding 5 years -- it was unexplained and Professor Williamson stated in terms when he gave evidence that he had no idea over what period the damage seen in photograph 475(1) in bundle 2 developed.) The rewind was carried out to the original design (as to which more later) and I have no reason to doubt the statement of Mr Emery (the defendant's employee who may be presumed to have knowledge of the matters spoken of by him) that the VPI process 'made little change to the performance ... mechanically': see his e-mail dated 4th March 2003. In the circumstances, it seems to me entirely plausible that what had earlier occurred had simply recurred, but that this time it was not discovered until there had been a catastrophic failure, this being indeed the conclusion reached at the time by the defendant's personnel who were actually involved in the events.

(iii) Dr Jowett furnished an explanation, consistent with the evidence before me, as to how the penetrating damage visible on photograph 527 in bundle 3 could have resulted from migration bringing the end windings in one location into contact with the damper windings which can be seen damaged on photograph 495(1) in bundle 2. That explanation is that at the point at which the damper bar was severed, the bar could have been flipped back and pushed into the very small gap between the end winding and the damper bar. He was of the opinion that the damper windings had caused the damage visible on photograph 527 in bundle 3. I found his evidence not only plausible but convincing, and I accept it.

In his closing submissions, Mr Sampson, in arguing that Dr Jowett's theory is unlikely, ignores part of the evidence of Dr Jowett and indeed of Professor Williamson.

* In paragraph 10, he quotes Professor Williamson as saying that he would have expected to see evidence of bulging around the circumference, but ignores his evidence that the section shown on the photograph has been destroyed (along with any evidence of bulging). Professor Williamson indeed stated that the fact that there was no sign of scuffing or migration did not negative the claimant's case and did not mean that migration was not the effective cause of the failure. Indeed, I note that Professor Williamson in paragraph 4 of his report, stated:

'Photograph 43, which shows a damaged damper winding that has been twisted and bent back on itself. This is consistent with it coming into violent contact with the rotor. The mechanical damage to this damper winding might also have arisen as a result of it bearing the brunt of the reaction force exerted on the stator when some extraneous object became lodged between it and the rotor. The relative locations of the damper winding and the rotor end-windings indicate that the most probable site of such an interaction would be the end-winding.'.

* In paragraph 12(b) Mr Sampson argues that with the mechanism for damage propounded by Dr Jowett 'it is inconceivable that no other damage would have been visible on the circumference of the end windings' but that point was specifically addressed in cross-examination by Dr Jowett, who said that if only one part of the end winding stood proud -- as it might have done because the end windings have elements of irregularity and unevenness -- the damage need not only have been in one location.

In the end, I find that it is more likely than not that the damage occurred in the manner postulated by Dr Jowett."

23. In conclusion the judge ascribed the migration to a failure to use VPI properly and to the use of the 1991 rather than the 1993 design. He said:

"In the event, however, I am satisfied that the process did not achieve its objective, and I agree with the submission of Mr Davies, counsel for the claimant, and find that in the absence of any explanation for the failure, it is right to infer that there was a defect in the process involving a failure to use reasonable skill and care in its application since, as I have said, it is agreed by the experts that the process, properly applied, should have significantly increased the mechanical strength of the end windings."

24. Mr Sampson agreed, during the course of the hearing before us, that the judge was entitled to reach that conclusion. Indeed it was consistent with the concession Mr Sampson made, namely that if the judge found migration then he would also have to find that the appellant was responsible, given that they had restored the rotor just a short time before.

25. I turn to the first of the merits grounds. Before doing so I note that the appellant did not in his skeleton argument challenge the critical finding in paragraph 34 that the damage to the exciter was not caused by the fins or other foreign object. Before us today Mr Sampson argued that there was no reference to debris in any of the notes, and that no-one was looking for it. Given the judge's conclusions I see no merit in that argument.

26. It seems to me that, in the absence of such a challenge (which could not in my view in any event succeed), the appeal on the merits in this case is hopeless. There were only two theories for the failure of the generator. If the appellant's only pleaded theory failed, then, on the facts of this case, the respondent was very likely to succeed, if not, inevitably.

27. I turn to ground 2. Ground 2 reads:

"The learned judge erred in finding that the heat sink fins became detached during dismantling the Exciter when there was no evidence that the same happened and in fact the evidence suggested very strongly that it would have been impossible for such damage to have occurred in that way."

28. It is submitted that the evidence that the breakage of the heat sink fins did not occur during dismantling was compelling. Dr Wadsworth had originally thought that the fins might have been broken off by the use of a spanner or T-bar during the course of dismantling. However he accepted in examination-in-chief, having heard the evidence that had been called, that that could not have happened. Thus it was during examination-in-chief he opined that the damage must have occurred whilst the crane was removing the stator, obviously a very delicate operation.

29. Complaint is made by Mr Sampson of the fact that the judge dismissed the evidence of Mr Savory. It was his evidence that he was sure that no damage had occurred during dismantling. The problem with that argument, as is pointed out by Mr Davies in his skeleton argument, is that Mr Savory could not remember seeing any damage. He had no recollection of any broken fins. It is quite clear that the fins were broken because one can see the broken fins in the photographs taken at the time of dismantling.

30. Mr Davies also submits, rightly in my view, that Mr Sampson overlooks in this argument the critical finding in paragraph 34 that the heat sink fins were not responsible for the failure. In the light of these (unimpeachable) reasons it follows that the heat sink fins must have been damaged during the dismantling phase. I see no merit in this ground.

31. Grounds 3 and 4 can be dealt with together, as indeed Mr Sampson did during the course of argument before us. In ground 3 complaint is made of the judge's finding that the VPI process for securing the end windings had failed when, so it is said, there was no evidence that it had. Likewise, his complaint is made of the finding about the migration of the end windings.

32. Mr Sampson, rightly in my view, accepted during the course of argument that these two grounds, even if made out, could not affect the outcome of the appeal because of what the judge said in the passage which I have already read out from paragraph 40, except to the extent to which it could be said that by making those findings doubt could be cast upon his earlier findings of fact. In my judgment those two findings, which in any event are amply supported by Mr Davies in his skeleton argument, cannot affect the findings reached in paragraph 34. The same applies to ground 5.

33. I turn now to the principal ground argued by Mr Sampson, namely ground 1. As I have said, that relates to a ruling made by the trial judge on the second day of the trial, Wednesday, December 3rd, the trial having started on Monday and having been adjourned until Wednesday:

"The learned judge erred in refusing the Defendant permission to rely upon the expert evidence of Mr Regan of the Defendant company to deal with the Defendant's theory for failure. He had however on the first day of the trial permitted the Claimant to adduce and rely upon expert evidence in the form of supplemental reports from the Claimant's expert witnesses as to that theory. The Defendant was thereby denied a fair trial."

34. Mr Regan started as an apprentice and is now Manager of Central Operations for the appellant. At one time or another, according to his statement, he has been involved in just about every stage of the building of a generator. He was not however involved with either of the repairs to this generator during 1996.

35. The judge did allow, then and later, parts of the evidence to be admitted. Complaint was made then, and is made before us, that the rest of his evidence ought also to have been admitted. In particular, it is submitted that Mr Regan should have been permitted to give evidence in support of the heat sink theory and as to the cause of the exciter failure. Mr Regan should have been permitted to give evidence, so it is argued, as to why he did not think there had been migration, why he did not think that what had been done to the rotor in April 1996 would have caused migration and:

"Since 1989 Alstom have produced 164 of this type and size of generator exciter. There are no known instances of winding migration other than this machine in March 1996. I believe that the only explanation for this is repetitive significant over speed where the speed of the exciter was greater than 1800 rpm. The defendants have no control over the speed of the machine."

36. It was also submitted that the judge should have allowed Mr Regan to explain why, in his opinion, the fins had broken and why that had caused the consequent failure of the exciter. It is important to note, as the statement makes clear, that this opinion had been set out much earlier in a report written by Mr Regan for the appellants' expert, Professor Williamson. I have already referred to the fact that the Professor, when considering the theory of the broken fins, wrote in his report dated 26th September 2003 that the absence of debris tended to contradict the theory. Mr Regan's opinion set out in his statement is:

"The Renk Tacke gearbox controls the shaft axial position from its wheel thrust bearings on the alternator side of the wheel.

"Explaining the loss of 5 heat sink fins from the exciter diode bridge was difficult until it was discovered from British Sugar's logs that there had been a major over speed only 2 3/4 hours earlier. The fins lost are from the leading end of two heat sinks mounted approximately diametrically opposite. (See British Sugar photograph number 1)

"It is presumed that the sequence of events was that the system over speed occurred when the alternator circuit breaker was opened on full load due to a tap changer control fault. The reaction to this would be for the turbine speed to increase very rapidly. With the loss of pinion to wheel tooth contact load, the gear wheel and the alternator shaft solidly coupled to it, would take up new load angles in the bearings and 'pump' the shaft axially due to the helical gear thrust changing rapidly.

"It is presumed that this extreme axial movement caused one or more heat sink fins to just touch some part of the casing or its bolting, damaging these fins or breaking one or more off.

"The machine was then restarted with either cracked and damaged fins or loose fins in the exciter housing. If they were just cracked even moderate vibration on a highly stressed aluminium casting will cause it to fatigue and break off. Again if there were loose fins in the exciter casing they could have lodged anywhere and dropped into the exciter at any time after restart of the system.

"It is unknown if the axial alignment between the alternator base plate and turbine, gearbox base plate has been checked to ensure the axial floats and clearances are correct. It is possible for base plates of this size to move under high vibration conditions. A major over speed may have caused some relative movement.

"The system was again restarted and loaded to approximately 15MW for some 2 3/4 hours before loss of excitation due to damage to the exciter rotor.

"The theory of cracked or loose fins in the exciter casing is evidenced by the 'gouge' in the end winding overhang of the exciter. The conductors on either side of the 'gouge' all trail the rotation demonstrating contact with a stationary body. If the damage had been electrical the conductors might have burnt back a few millimetres but 50% would be pointing with rotation and 50% against rotation roughly in the direction of their original coil form." [Underlining added.]

37. Mr Regan also referred to over 100 separate drawings that he had looked at. Apparently it was necessary to look at 100 separate drawings in order to be able to determine the clearance between the heat sink fins and the housing. Obviously, if one was going to assess the theory being put forward by Mr Regan the clearance was important. Mr Regan went on to comment on Dr Wadsworth's conclusions as follows:

"Dr Wadsworth does not agree with this theory and makes the point that there is no evidence of impact damage from broken heat sink fins. There may well not be. It is possible that the interior of the cowling would not show signs of impact from a fin travelling at such a high speed and it is far more likely that fins would have hit the base plate. The sheet metal casing would absorb a large part of the energy by deflection. I have already stated that there are gaps in the stator into which or through which a damaged fin could have entered. The photographs do not show any loose broken fins but the photographs were not taken by the defendants and they were not present during the initial stages of removing the cover and disconnecting the up shaft leads."

38. In this latter paragraph Mr Regan is not only disagreeing with Dr Wadsworth but also, to a certain extent, with Professor Williamson.

39. Before looking at the trial judge's ruling, it is necessary to look at the chronology. As I have already said, although the failure took place in September 1996, it was not until 2002 that the respondent issued proceedings. I have already also pointed out that the appellant put forward the heat sink theory, for the first time, in its defence.

40. Professor Williamson considered three possible theories for the exciter failure. One was a broken heat sink fin or some other object becoming lodged between the stator poles, impacting the rotor end winding and gouging a portion of the end winding before it was expelled by the mechanical forces present. He gave a reason in support of that theory and then wrote that the absence of a "record of remains of such an object in the sump" was against the theory. The second theory was that a portion of the end winding had bulged outwards and come into contact with the stator. He gave reasons in support of and reasons against the theory, one of which I have already mentioned. The third theory was later rejected by all the experts and can be ignored.

41. There is no reference in that report to Mr Regan's theory as to why the fins had broken or to his theory to explain the absence of any record of remains. The experts met in October and on 28th November, the Friday before the start of the trial on the Monday, produced the memorandum of agreement. There, for the first time in writing, Professor Williamson put forward a theory that the fins might have been broken as a result of what happened earlier that afternoon after the initial fault had been found and the alternator was restarted. As paragraph 20 rightly points out, this was a "hitherto undisclosed theory". The other two experts said they were unable to comment on this hitherto undisclosed theory until further investigations had been undertaken. They did, however, take the view that it was implausible and that there was no evidence to support it.

42. Standing back for a moment, it follows that although the appellant had pleaded heat sink fin failure, it had produced no material in support of the cause of the alleged failure until the Friday before the start of the trial. Then, and for the first time, this theory was disclosed. Where had the theory come from? It had come from Mr Regan and came from that part of his statement to which I have already made reference. Mr Regan has put forward his theory to Professor Williamson. Professor Williamson on the face of it does not accept it -- indeed, he says that it is undermined by the lack of debris -- and yet on the Friday before the trial the theory finds its way to the respondents.

43. There was then a supplementary statement made by Professor Williamson and replies in which he said this, in paragraph 2.3:

"My argument rests on whether or not this event [a reference back to what had happened at 2.15] subjected the generator to a sudden change in acceleration or deceleration, which when transmitted through the gearbox would I believe produce a transient axial movement of the shaft."

44. In other words, the shaft would have moved so much when restarted at about 2.15 that it would have led to the failure at 5.45. I should add that Mr Regan had an explanation in his statement for why, if there was such a "transient axial movement of the shaft", it would not have resulted in failure until 5.45. It was his theory that the fins might have been cracked and then only actually broken off some time later (notwithstanding that they were travelling at a speed of 1500 rpm).

45. That was the situation which the trial judge faced on the Monday. As I have said, first of all Mr Sampson made an application to adjourn the trial. By now Mr Sampson was getting concerned, so he told us, that Professor Williamson did not have enough expertise to bring home the "transient axial movement of the shaft" theory, he not being an expert in gearboxes, whereas one at least of the experts called by the respondent was. The adjournment application not having been successful, he then sought to rely on Mr Regan's evidence.

46. The judge started his ruling with these words:

"For reasons that I shall state, in my judgment, it is not in the interests of fulfilling the overriding objective that permission be granted to adduce the witness statement of Mr Regan."

47. The judge then referred to the claimant's case as pleaded and went on to set out the defendant's pleaded positive case to which I have already made reference. The learned judge then continued:

"3. D irections were given as to the adducing of lay and expert evidence. The lay witness statements of both sides should have been exchanged on 27th June 2003 and a debarring order was made against the defendant on 15th July, giving a final extension to 23rd July, at which time some witness statements were served which, frankly, do not take the matter any further forward at all. There was at all times, I am satisfied, available to the defendant, had it wished to adduce this evidence, the factual evidence of Mr Roger Regan, who has at all material times since the proceedings have commenced, as far as I understand it, been engaged by the defendant in a senior management position, being the manager of central operations of the defendant. The defendant, notwithstanding that it had thought it appropriate to plead a positive case, and had what appears to be relevant evidence that Mr Regan could give, chose not to adduce that evidence.

"4. Moreover, permission was then given later in the year to both sides by order dated 19th September to rely upon the expert evidence of two engineers. The defendant again, notwithstanding that its pleading was averring a positive case, chose not to rely upon Mr Regan (no doubt for its own good reasons) as an expert, even though he was in a position apparently to put forward a theory as to how this failure had occurred.

"5. The expert evidence that was relied on by the defendant was that of Professor Williamson. Professor Williamson in his report did not reach any clear conclusion, or indeed any conclusion on the balance of probabilities. But at some stage between submitting his report and the joint experts' meeting, Professor Williamson felt able to state the conclusion that he has now reached on the balance of probabilities as appears in the joint experts' statement.

"6. Even then, there was no wish and no application by the defendant to adduce late the expert or lay witness evidence of Mr Regan. Only on the first morning of the trial (3rd December) after the defendant had been faced with rebuttal evidence that came inevitably from the theory that had come at a very late stage from Professor Williamson, did the defendant seek an adjournment to obtain further evidence. That evidence has produced a supplemental report from Professor Williamson and there is of course no objection to that going in. But it has also produced a witness statement from Mr Regan that contains a mixture of fact and opinion. There is no good reason that the defendant has advanced before me why it did not produce either factual or expert evidence from Mr Regan at the appropriate time.

"7. Leaving aside paragraph 6 (which does seem to me to flow directly from the evidence with which the defendant has been faced) the expert theory that is now put forward and advanced in Mr Regan's evidence was a theory that, in my judgment, if the defendants had wished to advance through his evidence, they should have done by adducing the evidence at the appropriate time. To my mind, the defendant's action in doing this at this late time inevitably will cause significant prejudice if I allow this evidence to go in.

"8. So far as expert evidence is concerned, the defendant in effect will put the claimant in a position of having to face substantial evidence from an additional witness at this stage (such evidence being produced only for the first time today). So far as factual evidence is concerned, leaving aside paragraph 6, again assertions are made without any documents being produced. For example, in paragraph 16 of the witness statement, oral evidence is given without discovery, and without the claimant having had the opportunity to investigate matters. It seems to me, as I say, the interests of justice are not advanced by allowing such evidence to be admitted now and I refuse permission in the exercise of my discretion to adduce this evidence late."

48. When approving the draft judgment the judge added:

"2. Although I did not say so in my short ex tempore judgment:

a. I felt that if I allowed Mr Regan's evidence to be adduced, it was unlikely that the trial would be concluded in the allotted time, which would have necessitated a significant adjournment, which I felt was undesirable; and

b. I also did not think that the Court would be likely to be assisted or the interests of justice served by allowing to be adduced the non-independent expert opinion of Mr Regan, particularly when the defendant was in a position to adduce independent expert evidence to similar effect from Professor Williamson (in his supplemental report), as was conceded by Mr Sampson. (on 3/12/03 it was assumed then by all three Professor Williamson had the expertise required to give the evidence contained in his report, the true position only emerging when he was called -- see para 33.3 of my Judgment)"

49. It is submitted on behalf of the appellant that what the judge should have done at this stage was to adjourn the trial and set a new date. Mr Sampson accepts that it would not have been possible in the time then available for the experts relied upon by the respondents to examine some hundred drawings to find out what was the gap between the heat sink fins and the housing in which those heat sink fins revolved. He submits that the judge could properly have made an order of costs against the appellant, being the costs thrown away by the very late disclosure of this new theory. He submits that it would have been in the interests of justice to have such an adjournment and allow the matter to be investigated fully. He criticises reliance upon the fact that Mr Regan was not an independent expert.

50. I take the view that if the appellant had sought to adduce evidence from Mr Regan at a much earlier stage of these proceedings it may very well be that Mr Regan would have been allowed to give the evidence, but the passage in the postscript 2b about which complaint is made must be read in the context of the fact that the judge, and indeed everyone, thought that Professor Williamson was going to be able to give much of the evidence which would support the transient axial movement of the shaft theory. It was not until Professor Williamson gave evidence that it became very clear, at least to the respondents and to the judge, that Professor Williamson was quite unable to support the theory in the way that he had advanced it in his supplementary report.

51. It is well established that this court is loth to interfere with rulings by a trial judge on case management issues. Mr Regan had been available to the appellant as a witness for years. Not only was he available to the appellant, he had actually submitted to Professor Williamson the very theory which Professor Williamson was to rely on when the experts met.

52. In my judgment, the learned judge was quite entitled to reach the conclusion that he did and his decision not to adjourn the trial cannot in my view be faulted, and for those reasons, I would dismiss this appeal.

53. LORD JUSTICE PARKER: I agree.

54. LORD JUSTICE PILL: I also agree on the procedural point. I agree with Hooper LJ for the reasons he has given. The judge was entitled to make the decision he did. This is not a case where a new theory had emerged from the claimants to which the defendants were seeking to reply. It emerged and came into the open at the instance of the defendants who then sought to give evidence about it. Nor was it based on facts which emerged at the last moment. The theory had been advocated internally as a possibility by the defendant's operations manager at a considerably earlier time.

55. The evidence in the event before the judge was limited and circumstances did not point overwhelmingly in one direction or the other. This was a case in which there were only two possible causes for the damage which occurred. The judge was entitled to have regard to the expert evidence before him as to the cause of the damage. The judge held that of the two possible causes, that advocated by the claimants was the more likely. If that was a tenable finding the claimants succeed on the balance of probabilities, the civil stand of proof. The judge gave reasons for his finding and they are, in my judgment, tenable reasons. He was entitled to reach the conclusion he did and his conclusion is not rendered unsatisfactory by his findings at a later stage of his judgment when dealing with the allegations of breach of contract. The appeal is accordingly dismissed.

Order: appeal dismissed.