CHAPTER SIX - CAUSATION 2
6.1. Flotta recordings
Various statistics reflecting oil production from time to time on the Piper Alpha platform were relayed to the Flotta terminal and recorded there as well as being noted by the operatives working at the terminal. One of these records was produced and witnesses spoke to what they had observed. The implications of this material was commented on by the parties at some length. It was principally the defenders who sought to place importance on this material. Thus the defenders averred that "there was a sudden and material drop in the flow of oil from Piper Alpha at least 7 minutes before the initial explosion". The defenders sought to get support for this aspect of their case from what had been noted at Flotta at the time of the accident. However in argument they were content to modify the time span so as to argue that production at the platform had diminished substantially or ceased about 2 minutes before the explosion rather than seven. Their contentions in this respect are important since if there was a major process upset at the platform some minutes before the accident (particularly affecting the production of oil) this may afford an alternative explanation of the accident and indeed eliminate a leak of a relatively limited amount of hydrocarbon as being the cause of the explosion. In fact Senior Counsel for the pursuers accepted that if it were proved that there had been a sudden and material drop in oil production minutes before the explosion this could cause him difficulty. The pursuers accept that there was some loss of oil production minutes before the explosion but not the major upset suggested by the defenders and they say this is explicable as being caused by loss of production following upon the tripping of the condensate injection pump B. This brought about the need for the operators to unload and recycle the reciprocating compressors. This of course would cause a loss of gas lift which in turn would bring about a loss of production. Indeed in my view there is no doubt that a loss of gas lift would bring about a certain loss of oil production. Moreover if the reciprocating compressors were recycled gas which would otherwise have been available for gas lift would escape to flare. The operators had installed a system of alarms and process trips designed to alert the Control Room to any serious change in process conditions but in respect of most of these we were not told in evidence just at what levels of disturbance they had been set to respond. I think that the amount of hydrocarbon which could have possibly escaped from PSV 504 was restricted in relation to the very considerable quantities of condensate produced by the platform and therefore may not have had much effect on the systems for identifying process upsets (other than gas alarms). It is therefore unlikely that any phenomena at Flotta which indicated an upset equivalent to a shutdown could have been in consequence of the mere escape of condensate from a jagged condensate injection pump. On the other hand the process disturbance recorded at Flotta, had this been due to an escape of crude oil, must have been reflecting the loss of a substantial amount of oil. Certainly enough crude oil to have created a large explosion. The scale of the operations is illustrated by the fact that loss of gas lift alone could produce a production loss of about 25,000 barrels of oil per day and yet this loss would reflect nothing like a shut down. On the other hand one would have expected any escape of a large amount of oil sufficient to register a significant drop at Flotta (as distinct from a mere drop in production) would have triggered some of the various pressure and level drop alarms on the platform and the evidence was to the effect that this did not happen. The starting point for the pursuers’ submissions on this matter is that the reciprocating compressors were unloaded at least 10 minutes before the explosion and I agree that in fact that happened. Mr Vernon reported as much to Bollands the Control Room Operator. I further think that it has been established that the full effect of the loss of gas lift would have been noticed in about 10 minutes although some effect would be noticed after about 3 minutes.
6.1.2 The Spectra-Tek System
Details of the operators’ metering and telemetry system was given by the witness Bruce Lawson. Mr Lawson seemed to me to be a reasonably reliable witness. He was a senior metering engineer and had been employed by OPCAL in 1988 as a metering engineer. He had detailed knowledge of OPCAL’s metering and telemetry systems in their offshore installations. OPCAL has a telemetry system. Insofar as this was based on the Claymore platform it was the system known as the Spectra-Tek system and Mr Lawson had been the project engineer responsible for the installation of this. This system involved the automatic transmission of information between the pursuers’ platforms and also from the platforms to the Flotta Terminal. The communication network used microwaves and tropospheric systems. The microwave system operated on line of sight and the other did not. In relation to Piper Alpha the transmission dishes required by these systems were situated at the north face of the platform. As was explained to me a communications system is one which can ferry information from point A to point B. Telemetry is the information that is carried along the communication system. There were two functions of the operators’ telemetry system. The first of these is the offshore function. The Operators offshore are concerned primarily with the operations which are current and the information which at any point of time may affect these. This information can affect their productive capacity. Thus for example the pressure at Claymore can affect the other platforms’ ability to export products. The onshore interest on the other hand is not so focused on immediate data as on overall trends. The system on Piper Alpha was not as modern as that on Claymore and was known as the Solartron system. The advantage of the Spectra-Tek system on Claymore was that it had a trending facility. The system on Piper was only a real-time system. It is perhaps important to note that the Solartron system on Piper was situated in the Control Room there. It had an alarm panel which would annunciate among other matters serious interruptions of the process flow. It could also annunciate what were regarded as minor alarms and this would include indications of low or high pressure. This alarm had not annunciated prior to the accident. The telemetry systems would also relay from one platform to another or to Flotta alarms relating to a communications failure or a shutdown. However Piper Alpha was at the hub of the system so that any communication from another platform to Flotta had to be relayed through Piper. This means that if Piper were blown up the alarm system at Flotta would indicate not only a loss of communication at Piper but a loss of communications with the other platforms and alarms at Flotta would show this.
The pursuers argued strongly that if prior to the explosion a serious loss of oil pressure had occurred this would have registered on the alarms in the Piper Control Room, including the Solartron alarms. The defenders contended on the other hand that in the absence of evidence about the levels at which alarms were set there could be no conclusion arrived at as to what was required to set-off an alarm. In my view this can only be so within relatively fine limits. For example there was a substantial volume of hydrocarbon circulating through the process system at any point of time. It certainly cannot be assumed that the operators would want to be alerted to every variation in the process flow. An oil-well might be withdrawn from production causing some diminution in production but it may be that this would not give rise to a flurry of alarms. However one must pay some regard to the purpose in having an alarm system. It would be most surprising if the alarm system was not set to register any change in the process gross enough to require urgent action and in particular one which might signify the development of a serious danger. I think that it is clear that any change in the process approximating to an involuntary shutdown would be in that category and I find it difficult to believe that there could be a loss of flow equivalent to such a shutdown without alarms annunciating in the Control Room. If this could happen the elaborate alarm and control system would be relatively pointless.
The computer graphic number 56/2A of process illustrates the lay out of the Control Room on the Piper Alpha platform. A person sitting at the Control Room operator’s desk could see the alarm annunciator and in any event would hear the audible alarm. Number 42/1 of process illustrates the Control Room at Flotta. If there were to be a communication fault at Piper Alpha there would be flashing lights and an audible alarm at the control panel at Flotta and the alarm could be converted into a solid state light by operating the acceptance button.
The Spectra-Tek system on Claymore in the first instance recorded local data from the metering skid on Claymore itself and this data would include the flow from the Scapa Field which converged on Claymore. The said data is collected by what is known as a multi-drop loop. This means that to collect data from each of the nine microflow computers it takes 3 seconds. Thus to go round the whole loop takes 27 seconds. Therefore if a particular piece of data just misses its place on this chain it may take approximately 30 seconds before that data can be displayed. The computer is able to set out trends in the data it collects and indeed the data for the nine streams is amalgamated. The data collected for each stream would include readings for a number of matters such as pressure differentials between say Claymore and Flotta. The data used for this purpose is collected on a 1 minute basis. This means that if a particular value is shown on the output documentation at a specified time it is possible that such refers to data actually recorded 59 seconds earlier. The gathering of data by the Spectra-Tek system from sources external to Claymore, such as Piper, does not require to go through the multi-drop loop and thus the input is much faster, perhaps by a second or two. Only the trending data is recorded. In making comparisons between the different kinds of data allowance has to be made for the time skew caused by the multi-drop system. Eighteen different trends were configured. The trending data was also collected on a hard disk at Flotta. This meant that after the accident information was available of the trending records relating to Piper and in this respect two trends 1 and 4 were particularly relevant. The trending data is timed according to the system’s own clock and this is not synchronised with any other clock. After the accident the hard disk containing the Spectra-Tek data was brought to Mr Lewis, a software engineer, (and a witness in the case) and he produced the document setting out the data which is number 41/15 of process.
Trend 1 relates to the differential pressure between Piper and Flotta. Trend 4 relates to the flow rate from Piper to the Tee junction on the MOL where the Piper flow meets the Claymore flow. Thus the oil flow leaving Piper was relayed to Spectra-Tek and recorded there at minute intervals. Trend 5 was the flow rate from Claymore to the said junction. Piper had an indirect input into this trend because the data from the Tartan platform went through the telemetry system on Piper. Trend 18 relates to the main oil line pressure as measured at Claymore.
If there was a failure of the telemetry system on Piper certain trends would be lost and the data for any such trend would become invalid. For example because of the inter-relationship of the data affecting the various trends, trends 1, 2, 3, and 4 would all become invalid if telemetry at Piper were to fail. If the equipment in the Control Room at Piper were damaged by the effects of an explosion then failure of the telemetry system could result either by disconnection of cables or disconnection of energy supplies.
If there was a problem with the telemetry system on Piper then there would be a short delay before this registered on the Spectra-Tek system because of what the witness Lawson referred to as the "handshake". This is in effect attempts made by the stations under the system to establish contact with one another before an alarm is annunciated. This process could take as little as 20 seconds but could also take longer.
On the day of the accident in relation to the recording of trend 1 at 2200 hours there is a pressure value of 43.4 bar brought out but at 2202 there is a value of 42.30 and that value thereafter continues to be repeated. The freezing of values indicates that telemetry is no longer working. Although it is not impossible that the system would record the same value on successive readings this is unlikely. Thus subject to adjustments on the time scale the time when the telemetry on Piper failed can be worked out. In relation to trend 1 the telemetry at Piper failed no earlier than 2202 hours as printed although failure a minute or so later is just possible. However the trend information also shows that there was a gradual reduction in oil flow after about 2156 hours and this could be attributable to the loss of gas lift. Certainly in my view the trend information would not justify a finding of a sudden and material drop in oil production 7 minutes before the explosion. Trend 4 also shows a gradual loss of production from about 2158 hours followed by a constant value after 2202 hours. A particular value may in normal circumstances repeat 2 or 3 times but given that the values being discussed continued to repeat in my view it is likely that the beginning of this repetition marked the loss of telemetry. Indeed the values at 2201 hours could also have been a sign of the loss of telemetry since they show a reduction in pressure.
At the time of the accident there were about 15 wells in production and the effects of a loss of gas lift would have begun to be felt after about 3 to 5 minutes with the full effect happening after about 10 minutes. The effect of the loss of gas lift would vary from one well to another. The wells which required gas lift would have suffered a loss of production of about 50%. Thus given that the recycling and unloading of the reciprocating compressors occurred between say 2145 and 2150 hours there should have been a notable loss of pressure about 2200 hours or slightly earlier. It takes about 40 seconds for the pressure pulse to travel from Piper to Claymore so that the latter platform would see the change of pressure after that interval. A further important consideration is that it takes about 3 minutes for the pressure pulse to travel from Piper to Flotta so that a pressure drop at Piper would not be noted in the pressure measuring equipment at Flotta until about 3 minutes after it had occurred. The loss of pressure at Piper would result in a pressure drop in the MOL and this would affect the production at Claymore since the oil from that platform would meet a different pressure as it traversed the line. The production at Claymore was in fact recorded in trend 18. The pressure at Flotta would not drop below about 230 psi because there were pressure control valves (PSV 90 A and B) at Flotta designed to maintain the MOL pressure at that value. The graphs in number 41/16A and B of process show the recordings of trend 18 and a loss of Claymore pressure some minutes before 2200 hours to be followed by a sudden drop of pressure at 2202 hours. The advantage that trend 18 has over trend 2 is that the material in the latter contains a combination of locally and remotely derived data which causes a certain degree of time skewing. The graphs relating to trend 1 (the pressure differential between Piper and Flotta) also show the commencement of a gradual drop in pressure about 2156 hours to be followed by a drastic drop between about 2201 hours and 2202 hours. The loss of pressure due to the cessation of gas lift would be compounded slightly by the loss of condensate production when the condensate injection pump system failed.
6.1.3 Process at Flotta
The flow into Flotta first went through the feed preheater the function of which was to heat up the fluid. The flow of hot oil was controlled by temperature control valves (TCVs). If the flow coming into Flotta were to diminish the temperature of the oil flow would increase because the oil was itself heated by hot oil and the amount of this heating oil would not diminish. Thus a reduction of oil coming into Flotta would require process adjustment by an operator because otherwise the TCVs would automatically close if the pre-set pressures were exceeded. Such closure would cause alarms to annunciate in the Control Room. Alarms of this sort were not in themselves an unusual occurrence. After it was heated the oil flow went through a separator and then a desalter. The flow emerging from the desalter was split into 2 streams. One representing 40% of the fluid (called the cold stream) went directly to the stabiliser while the remaining 60% (called the hot stream) went through a vessel called the feed bottom exchanger before then going into the stabiliser. These 2 streams had first to pass flow elements which measured the flow going into the stabilisers and these measurements were recorded in a series of pen charts in the Control Room.
6.1.4 Evidence of Flotta Operators
Mr Kelly was the Process Shift Controller at Flotta on the evening of the accident and as such was responsible for the process side of the Flotta operation during his shift. The schematic number 42/1 of process shows his office within the Control Room where he was when the accident occurred. The Lead Operator that evening (working under Mr Kelly ) was Mr Stockan, and a Mr Slater was the Process Operator which is the equivalent of the Control Room Operator on Piper. In the Control Room (as shown in the said schematic) there were four stabiliser trains (A, B, C, and D) and these represented each of the trains in the process flow. The alarms for the TCVs were on the stabiliser train to which they related. There was a pen chart for each train and this also would be situated on the train to which it related. The pen charts were number of process 56/1. On these charts the red pen represented the vapour draw, the blue pen the hot feed and the green pen the cold feed. The rectangular boxes were intended each to represent the passage of an hour and the square subdivisions within the rectangles a period of 15 minutes. However I think it was agreed that no accurate timings could be taken from the charts. For some reason the instrumentation of train C was rather more sensitive than that on the other trains.
At a point in the evening approximate to 2200 hours Mr Stockan came to the window of Mr Kelly’s office and asked him to come out and have a look. He told Mr Kelly that telemetry alarms had come up offshore and that the feed had started to drop off. When Mr Kelly came out of his office and walked past the telemetry display units he could see that communication alarms had come up for every platform. He claimed that as he walked through the Control Room he had looked at the clock there and noted that the time was 2202 hours. It is not absolutely clear that such a precise recollection was not perhaps prompted by a degree of hindsight because it was at 2200 hours that the telemetry system signalled a clear loss of telemetry. Nevertheless he did testify that the Control Room clock did keep accurate time. Because all the telemetry signals were relayed through Piper a telemetry failure there would of course affect the telemetry relating to every platform. Mr Kelly then went to look at the pen charts. He reckoned that it took about 10 seconds from the time Mr Stockan approached him until he was consulting the charts and there is no reason to doubt the approximate accuracy of this recollection. When Mr Kelly looked at the chart relating to train A he saw at once that the blue and green lines on the chart had begun to drop markedly which indicates that there was a serious diminution of the flow. This drop continued as he watched. Before the point where the traces had begun to drop markedly they had been within normal limits. The movement in the traces from relative normality to clear abnormality as observed by Mr Kelly according to his evidence took from about thirty seconds to about a minute. The extent of the drop in terms of Mr Kelly’s experience was sufficiently marked to signal a total shutdown at Piper and I accept his judgment on this. When the traces on the various trains had dropped to their base levels he instructed Mr Stockan to contact Piper but the telephones were out of order.
The witnesses who gave evidence on the matter of what was observed at Flotta did not always agree on matters of timings. I did not form the impression that any of these witnesses were doing other that their best but given the compressed time scale of the relevant events it is only to be expected that there could be a degree of inaccuracy. Thus there may be some uncertainty as to just how long it took before Mr Kelly could be certain that he was observing a drop in flows equivalent to a shutdown at Piper but it certainly was relatively short.
It has to be observed that the pen charts are not related to the telemetric system but rather record the actual flow into Flotta and this would explain why the pen charts continued to show some flow because flow would be continuing from Claymore. It should also be noted that it was accepted by parties that the pressure pulse would take about 3 minutes to travel from Piper to Flotta and it would therefore take that time before the flow elements at Flotta would record pressure changes at Piper.
Some evidence about timings can be derived from the entries in the Control Room log book. This was in general kept by Mr Slater but on the relevant occasion because the operators were under pressure as the critical events occurred much of it was not written up until the morning following the accident. However the entries were made by reference to a scratch pad kept at the time of events and Mr Kelly, Mr Stockan, and Mr Slater consulted before the log was written up. The scratch pad was not produced and indeed no one was asked if it was still available. The log had an entry that at 2200 hours there was a communication fault for all rigs. On the other hand Mr Stockan gave evidence that there was a clock forming part of the telemetry system and that this clock froze at 2202 hours. He may well be right about this because there are other indications in the telemetry records that this was the time at which telemetry failed. If so the log would be inaccurate as to the time of the communication failure and this would not be surprising. On the other hand on one view Mr Slater would have had the best opportunity to note the time when the communication alarms came up. Mr Kelly said that he personally did not make any contribution to the entry relating to these alarms.
The loss of telemetry would not in itself have prompted Mr Stockan to report to Mr Kelly since this happened from time to time for a variety of relatively innocuous causes. What on the other hand did motivate Mr Stockan to report events was signs of an obvious drop off in production and inability to contact Piper.
Mr Slater asserted that the first thing out of the ordinary that he noticed was that the pen charts were showing a serious drop in production. This was a larger drop than might have been brought about by loss of gas lift. He had been in the vicinity of the stabiliser trains in the Control Room. Shortly afterwards alarms indicated that the TCVs had closed which also indicated a serious drop in the flow. Mr Slater’s evidence that he had not noticed the communication alarms is not too convincing since it was apparently he who made the log entry that such alarms had gone off at 2200 hours. However he claimed that his recollection was that he had noticed a drop in the flow before he heard any alarms - such alarms relating to the TCVs. It should be noted in relation to the log that in the morning after the accident there was a debate between Mr Kelly, Mr Slater, and Mr Stockan as to what time ought to be entered in the log in respect of the time when the feed rate began to drop. At the time Mr Stockan was taking the view that the drop in feed rate did not occur until after 2200 hours but when he gave evidence he claimed that the drop had occurred about 2155 hours - that is before the accident. He claimed that the low flow alarms had gone off about that time. Such alarms are likely to have indicated at least a 25% drop in feed rate although the precise figure would depend on the settings. However if this were so it is curious that he did not refer to Mr Kelly until as he said the communication alarms had gone off. Moreover when Mr Kelly came out of his office he noted that the pen charts were still dropping which might be odd if a significant drop had been noticed at least 5 minutes earlier. Moreover Mr Kelly noted that the pen charts were dropping rapidly. Nevertheless Mr Stockan explained that during the periods in question it was not at all unusual for low flow alarms to go off because the production processes on the platforms were somewhat unreliable. When pressed by Counsel Mr Stockan was prepared to admit that his timings may not have been accurate and given the whole circumstances I should have thought that this concession was not surprising. Mr Slater on the other hand claims that he did not notice the TCV alarms until after such time as he had observed a serious loss in gas flow. For some reason the sounding of the TCV alarms does not appear to have been logged. Number 94 of Process is a rather curious document. It is dated 7 July 1988 and Mr Stockan accepts that it is in his handwriting although he cannot remember how it originated. It would appear to state that all the communication alarms went off at 2201 hours. The next entry is that the feed to the plant started to drop off immediately and this is certainly not consistent with the evidence Mr Stockan gave about the sequence of events. Although Mr Stockan cannot remember writing this document it must have been written shortly after the accident and I think it places a question mark over his subsequent memory of detail. It is to be noted that there was a shutdown alarm on the telemetry panel on Flotta. This initiates an alarm if the combined flow rate from Piper falls below a certain level and the purpose of the alarm is to give a signal if Piper shuts down. It is important to note that none of the Flotta witnesses speak to this alarm annunciating before the break-down in telemetry at the time of the accident. The absence of such an alarm is difficult to reconcile with any suggestion that there was a major upset on Piper before the explosion.
6.1.5 Defenders’ submissions on Flotta
The defenders sought to argue that on a proper analysis of the Flotta evidence it was clear that there had been a major drop in production at Piper Alpha - indeed one equivalent to a shutdown - some minutes before the telemetry broke down. Since it must be assumed that the failure of the telemetry system came at the time of the explosion then there must have been a major upset of the production at Piper before the explosion. A leak from PSV 504 could not have caused such a response and therefore it must be inferred that some cause other that a leak from the said blind flange was responsible for the shutdown conditions and if this happened then such a breakdown would have been a pre-eminent source of the explosion. I would certainly agree that if conditions on Piper were such that a major general shutdown of production was signalled before the explosion then it would be difficult to reject the possibility that the shutdown was occasioned by a large escape of gas and this event could of course have caused the accident. The defenders contended that the conclusion they relied upon was supported not only by Mr Stockan but by the pursuers’ own witnesses. The defenders accepted that the computer clock at Claymore from which the Spectra-Tek records were taken was 38 seconds fast and since that data was recorded only at minute intervals a recording shown could fall anywhere within that period. Another important consideration is that the Spectra-Tek data might not freeze until about 20 seconds after the loss of telemetry. This is the result of what is described as the "handshake" between the computers as they tried to adjust to the signals indicating a loss of telemetry. The defenders also agreed with the pursuers that a fall in pressure at Piper would not be detected by the metering system for flow at Flotta until about 3 minutes after the event because of the time the pressure pulse would take to travel from Piper to Flotta. Subject to a possible time lag of about one and a half minutes the Spectra-Tek system is recording real-time data whereas the pen chart system at Flotta indicates a time lag of about 3 minutes because of the pressure pulse. Senior Counsel for the defenders argued that on the most favourable view for the pursuers the loss of Piper production pre-dated the explosion by 2 minutes. Mr Lawson gave evidence that a failure of the Spectra-Tek system could occur if the communication disks on the north face of the platform were damaged or if the communications between these disks and the communication room were damaged. Mr Lawson also explained that the system could also fail if the equipment in the platform Control Room was damaged or there was a disconnection of cables or supplies. There certainly was evidence that the explosion caused damage in the Control Room and this included damage to the computer VDUs. The defenders argued that the explosion was certainly not earlier than 2200 and on that point I agree with them since the preponderance of the evidence places it close to 2200 hours but not before that time. Number 41/17 of process was said by the pursuers’ witnesses to be a print-out from Claymore showing various readings from the Spectra-Tek system at about the time of the accident. The defenders founded on this document. This duplicates what would have been recorded at Piper. The document shows that there was a loss of telemetry from Piper between about 22O2:08 hours and 2202:50 hours. Moreover the trends shown in number 41/15 of process indicate that telemetry was lost just after 2200 hours. When the values on the trends freeze this is an indication that the telemetry has failed. Two or three constant values may occur in normal course but if these are followed by a string of the same numbers this in my view suggests strongly that the constant numbers have from the beginning indicated loss of telemetry. Trends 1 and 4 may be particularly significant because they do not include any locally derived data. The pressure control valve at Flotta seeks to maintain pressure in the flow from the MOLs at the set level which means that the pressure differential between Piper and Flotta should not vary much (although it may be influenced by flow rates from the other platforms). Looking to the whole evidence I have concluded that the loss of telemetry at Piper occurred at 2201 hours or, allowing for the intrinsic delays in the system, seconds before. However even in attempting to interpret the data care must be taken not to rely too much on the apparent precision since the time scales are very narrow. The defenders’ Senior Counsel was at pains to point out that the experts had conceded that a telemetry failure as late as say 2203 was within the bounds of possibility. However possible a slightly later time may be as I have indicated I think that it is unlikely that normally recurring constants would immediately precede the failure. This would be quite a coincidence. Mr Stockan testified that he had noticed that the computer clock in his control room had failed at 2202 hours but I doubt if his memory is sufficiently reliable to cope with fairly fine detail. The defenders argued that it can therefore be concluded that the explosion occurred at the time of the telemetry failure. I think this is stating matters rather too baldly. I prefer the way the pursuers put it which is to contend that the telemetry failed at the time of the explosion or shortly after. It is not obvious that the telemetry would fail instantaneously with the explosion. For example not only the damage to the basic equipment could have caused the telemetry to totally fail but also damage to the transmission disks or interruption of the energy supply. On the other hand it would appear unlikely on other grounds that the explosion occurred materially before 2200 hours. For example the witness Captain Morton remembered that just before the explosion the 10-o’clock news had started.
After the loss of telemetry and he had noticed some drop in pressure Mr Stockan brought Mr Kelly out of his office. Mr Kelly claimed that it was 2202 hours when he looked at the clock on his way to look at the pen charts. This was not a digital clock. The clock, he said, was always kept accurate. Mr Currie argued that the pursuers had not challenged Mr Kelly in respect of this time. This is true and it may be that the pursuers accepted that Mr Kelly was giving times to the best of his recollection. If so I think that they were right. Nevertheless I am faced with the task of deciding whether Mr Kelly’s precise timings can be reconciled with other acceptable evidence. Mr Kelly certainly accepted that there was a drop in the feed rates shown on the pen charts and expressed the view that this drop must have started about 30 seconds before he looked at the charts. I think there is little doubt that the pen charts are perhaps the most reliable record we are left with of the flow rate from Piper about the time of the explosion. Moreover I accept that they show that at or about the time of the explosion there was a rapid drop of pressure at Piper equivalent to a total loss of production. Unfortunately however the charts do not show timings that could be regarded as reliable. Relying on Mr Kelly’s timings the defenders argued strongly that if the major drop in production at Piper was noted to have occurred by about 2202 hours then allowing for the delay of 3 minutes required for the Piper pressure pulse to reach Flotta then the drop must have occurred 2 minutes or so before the explosion.
It should be noted that as Mr Kelly walked through to look at the pen charts he saw that the communication alarms were up in the control room. The Log number 42/3 of process is interesting. It shows that at 2200 hours communication faults were noted on all the rigs. Then there is an entry that at 2202 hours the feed rate was dropping off fast. The Log was partly made up the next morning from discussion between the operators and perhaps from entries on the scratch log. The significant thing is that at that time close to the incident the impression of those concerned was that the communication fault occurred 2 minutes before a pressure drop was noted. If the entry about the time of the pressure drop should have read 3 minutes rather than 2 then that would be reasonably consistent with the explosion having caused the pressure drop.
According to the evidence given by Mr Stockan the first thing he noticed were the TCV alarms which he thought had annunciated about 2155 hours and it was not until 2202 that the telemetry alarm went. It was only after that point that he communicated with Mr Kelly. Mr Stockan had noted a drop in pressure shortly after the TCV alarms. This programme would mean that the oil flow at Piper had reached a critical level of reduction about 10 minutes before the explosion. This evidence certainly does not square with what is written in the Log nor indeed is it possible to reconcile the timings with those of Mr Kelly. Mr Stockan accepts that his timings may be wrong and Mr Currie did not seem to be too keen to rely on them preferring those of Mr Kelly. Mr Stockan also accepted that it is possible that he is mistaken in attributing the drop in oil flow to a time preceding the alarms. On the other hand apart from the niceties of the times Mr Stockan has the pressure showing a pronounced drop before he summoned Mr Kelly.
Mr Wottge took data of oil production of the platforms from the Spectra-Tek records at Claymore and he averaged this out for the period just before the accident. Looking at the average position his analysis shows that there was a downward trend in production starting at about 2148 hours and becoming more significant about 2158. The drop is in percentage terms from about 102% to about 89%. Some of this drop was attributable to Piper and some to Claymore. However this loss of production was before the system shows frozen values and is likely to have been attributable to loss of gas lift on Piper and of course there is also a loss of condensate production from the time the pump failed. Defenders’ Senior Counsel argued that the drop noted by Mr Wottge at 2158 hours is consistent with his thesis that the pressure was dropping drastically before the explosion. However what the defenders seek to take from the pen charts is not a drop of a gradual character but a drop signifying something similar to a shutdown. Indeed Mr Wottge’s analysis of the Spectra-Tek data discloses what he calls a "drastic reduction" in the flow from Piper about 2202 hours. The pressures coming from all the platforms interact with one another so that fluctuations within normal ranges cannot necessarily be attributed to a particular platform.
6.1.6 Conclusion on Flotta recordings
The defenders urged me to conclude that several minutes before the explosion the production processes at Piper had effectively shutdown. It is certainly true that it is impossible to reconcile all the evidence spoken to by Flotta witnesses. Indeed it was in recognition of that fact that the defenders seemed ready to abandon the evidence of Mr Stockan. On the other hand the defenders’ case on this Chapter depends very much on the evidence of Mr Kelly. Given that the telemetry system clearly failed about 2201 or even 2202 then if this represents the time of the explosion and the production breakdown can be attributed to some minutes earlier then this would be difficult to reconcile with the pursuers’ hypothesis about the cause of the accident. A point made by the defenders, and it has a measure of force, was that the pursuers did not appear to challenge any of Mr Kelly’s evidence particularly as it related to timings. If there was evidence to support the view that there was a breakdown in production at Piper some time before the explosion then in that situation the evidence from Flotta may have given support to the defenders’ contentions. However there is a strong body of evidence that suggests that the inferences the defenders seek from the position at Flotta would be ill-founded. In respect of the position at the Piper control board at the time of the accident Mr Bollands, the Control Room Operator, gave his evidence about the state of the alarms at the time of the accident in a clear and convincing manner. His observation about the incidence of alarms was materially supported by Mr Clark, the Maintenance Lead Hand. Prior to the accident Mr Bollands did not observe any process alarms other than those relating to the tripping of the Compressors and the Condensate Injection Pump. In relation to oil flow the platform operators had an elaborate system of safety features and alarms. Thus anything untoward in the oil flow should have produced trips and alarms. It is true as the defenders argued that we were not given details of the settings for the various safety flow and level meters and alarms. However these features were highly important. Indeed several witnesses spoke to the fact that at about the period of the accident process failures were a frequent occurrence. The pen charts were indicating a sharp and rapid drop in the flow rate equivalent to a shutdown. In my view it is inconceivable that the safety features would be set in a way other than that they would respond in the event of conditions equal to a total failure in production. If the alarm and safety system did not respond to an accidental production shutdown it would have scarcely been worth having. Thus if the situation suggested by the defenders had arisen appropriate alarms would have been annunciating in the Piper Control Room and this did not happen. Moreover it is difficult to believe that the operators working in the Modules such as for example the Oil and Water Operators who had duties in Module B would not have been alerted to a process breakdown. It is of course possible that the module B operators were not in the module at the time but given that the reciprocating compressors were being re-cycled and unloading at the time it may have been surprising if they had not duties to perform there. Moreover I should have expected that any mass leakage of gas in Module B would have triggered at least some of the gas alarms in that module. There is of course evidence that might signify that some of the alarms in that module were not operative at the time but it would certainly be another coincidence if the absence of any alarms at all in the module could be explained in that way.
In my opinion the evidence supporting the view that there was no shutdown in process preceding the explosion is far stronger than the rather tendentious inference which the defenders sought to extract from the evidence concerning Flotta. In the circumstance it is not necessary for me to make findings as to the source of the weakness in the Flotta evidence. However there may be a variety of explanations for the apparent anomalies in the Flotta evidence. For example the defenders’ theory depends to a degree on Mr Kelly’s timings being very accurate. It is certainly the case that the pursuers did not challenge these but as I have said this may have been due to the realisation that Mr Kelly was giving his timings to the best of his recollection. The defenders for their part led Mr Stockan who in some respects differed from other witnesses in regard to timings. As the evidence in this part of the case indicated not surprisingly at least some witnesses had difficulty in being accurate about the very precise timings on which the defenders’ case depends. Some of the clocks either on the Spectra-Tek system or in the control room may unknown to witnesses have been inaccurate. The loss of telemetry may not have coincided exactly with the explosion but occurred shortly after it. Finally I think it could be very significant that in the Log compiled in the hours immediately following on the explosion the loss in production is thought to have occurred some minutes after the loss of telemetry.
6.2. Gas Dispersion
The question of gas dispersion is central to the pursuers’ case as to the cause of the explosion. The question which they must answer is whether an escape through PSV 504 of such quantity of gas as the condensate pump system could have accumulated could have caused the explosion which took place. Moreover could such escaping gas have dispersed in such manner as to account for the gas alarms which were noted by Mr Bollands? Could the gas which may have thus accumulated have exploded so as to create the forces necessary to destroy the B/C and C/D firewalls and to have hurtled the projectiles into Module B which are the pursuers’ explanation for the damage to equipment in that module which damage they say released hydrocarbon?
6.2.2 Dr Davies
The pursuers’ main witness on this branch of their case was Dr Davies who gave evidence over a period of 4 weeks. The defenders led a witness, Dr Bruun, whose evidence was restricted to a discussion of certain aspects of Dr Davies’ modelling methodology and only lasted 2 days.
At the time of giving his evidence Dr Davies was 47 years old. He was the managing director of BMT Fluid Mechanics. Before it was privatised this organisation had been the National Maritime Institute and then it merged with the British Research Association. Dr Davies was also the Chairman of an associated company BMT Offshore Ltd based in Aberdeen. He held a BSc Honours degree in applied mathematics and had a Masters Degree in Aeronautics from Imperial College. That work was concerned with airflows as they relate to structures. He had worked for the British Aircraft Corporation as an Aerodynamicist. In 1971 he returned to Imperial College where he carried out research on the behaviour of flows round structures and thus earned a PhD. In 1974 he went to work at the National Physical Laboratory where he did work on the low of bodies around circular objects particularly in relation to turbulent flows. This work arose from the requirements of the offshore oil industry. Thereafter he had a continuing involvement with that industry. From about 1976 he was working on topside aerodynamics including fluid dynamics. This was concerned very much with the nature of airflow around platforms. Thereafter he had been involved for 18 years in consultancy work for the offshore industry. His experience included wind tunnel modelling. He had carried out advisory work in relation to the Piper and also the Claymore platforms. Some of his work was the subject of publications. In the 1980s he was considerably involved in heavy gas dispersion and this included the modelling associated with the work. He did work in connection with the British Gas Terminal at Canvey Island and in that work he was looking at the behaviour of heavy flammable clouds of gas. Thereafter he was engaged in a number of further major studies into gas behaviour. He was given a contract by the US Gas Research Institute to carry out a wide range of wind tunnel exercises. Among his other work was a study into high pressure releases for a consortium of European companies. He has been involved in consultancy work in relation to about 30 or 40 different oil platforms and this covered work into the dispersion of gases within Modules. Since the Piper Alpha disaster he had carried out ten studies into the kind of problems that arose there. Since 1972 he has been continuously involved in the setting up measurement systems. Since 1982 he has been a Chartered Engineer. He has been the author of a number of publications all within the field of fluid mechanics. One of these was in connection with Froude number scaling a problem that arose in his evidence. In my view there was no question about Dr Davies’ qualification to give the evidence he gave and indeed the defenders never doubted those.
Dr Bruun was also well qualified. He was 55 years old when he gave his evidence and held the post of lecturer in the Department of Mechanical and Manufacturing Engineering at the University of Bradford. He was awarded a first class degree in Mechanical Engineering in 1964 and a PhD in 1967. This latter degree was in the general area of fluid mechanics. After qualifying he had periods of research work at first Southhampton and then Cambridge Universities. He was a guest professor at Karlsruhe University before taking up his present post. He has a number of publications to his name. Some of these concern the behaviour of flows but he has done a considerable amount of specialist work into hot wire anemometry. Hot wire anemometry involves the use of a wire heated up by means of passing an electric current through it. The device can be used to measure quantities such as velocity and also concentration measurements. There is no doubt that Dr Bruun has a lot of experience of the use of these. On the other hand most of his experience in fluid mechanics has been in the academic field. Dr Bruun sought to challenge a certain amount of Dr Davies’ methodology particularly in regard to the use of hot wire anemometry in his modelling runs. However a lot of the detail in his evidence was not pursued by the defenders in their submissions no doubt because at the end of the day they were happy enough to adopt some of Dr Davies’ evidence.
In cross-examining Dr Davies, the defenders sought to attack his use of Froude numbers and Reynold numbers in his modelling. The problem was of considerable technical difficulty. At the end of the day the defenders did not support their criticisms by evidence - even from Dr Bruun who had relevant experience about such matters.
Dr Davies concluded that a release of flammable material within Module C which affected only the gas detectors within Module C at gas zones C2, C3, C4 and C5 would have had to be released at the eastern half of Module C. I did not understand the defenders, at the end of the day, to challenge that particular assertion. Moreover he opined that any release within Module B that could have passed to the east end of B would have had to initiate gas alarms within B to get there. This of course assumes that the alarms within B were working correctly. In 1989 Dr Davies had carried out certain wind tunnel work to investigate a variety of gas release scenarios in Module C. This work was done on the basis of a number of assumptions which had been put to him and in particular assumptions relating to the pattern of gas alarms which had annunciated at the accident (these latter taken from the evidence of Mr Bollands). He concluded from his 1989 work that releases from the vicinity of PSV 504 frequently produced the first alarm at zone C3. However the gap between the first alarm and the second alarm which had annunciated was about 2 minutes and this was never sufficiently long to explain the alarm pattern that had been given to him. Because of this he carried out further work in 1993 in relation to small releases. He also carried out in 1993 work on gas escapes within Module B. The pursuers contended that the evidence of Dr Davies confirmed that a pattern of alarms such as was experienced was consistent with an escape of gas in the general vicinity of PSV 504. Mr MacAulay submitted that the evidence of Dr Davies showed that the initial alarm at C3 could have been caused by a leak from the area of PSV 504 if the release had been a small continuous release or a curtailed release such as a puff type release. A second release from the same site, if more substantial could explain the remaining alarms. The pursuers argued that it would in normal circumstances be highly coincidental to have two separate gas leaks separated by about 2 minutes. However this is just what one would expect if the escape was due to a pump jagging operation. Certainly I think that it would be reasonable to conclude that there was some connection between the gas escape which caused the first alarm and that which caused the second. It would be rather a coincidence if there were two totally unrelated escapes within minutes of each other.
Dr Davies’s work showed that any gas released in Module B would dilute with air as the wind moved it through B to the east face. This process of dilution would continue at the east face of the platform even before gas could be ingested into C. Thus to transmit gas generating from B which could be absorbed into C and form a flammable mass there the gas released in B would have to constitute a relatively large mass. Such a mass would necessarily initiate any gas alarms in B which were working. Mr MacAulay emphasised that no alarms had gone off to indicate any process disturbance in B and this might have been expected if there had been a large release of gas there. Moreover Dr Davies could envisage no circumstances where it would be possible to ingest gas into Module C so as to trigger an alarm at zone C2 and the defenders did not lead any expert evidence adverse to this view.
Dr Davies did not consider that the precise sequence of the final flurry of low level alarms made any difference. In any event the alarms were so close together that it is difficult to suppose that their sequence accurately displayed the sequence of the arrival of critical quantities of gas at the detectors.
The natural wind over the sea is a turbulent wind with a speed that varies with height above the platform. When the wind encounters the platform it will be disturbed by the presence of the platform and this will cause it to flow around, over, under, and to a degree through the platform. The wind external to the platform is the primary flow and any wind internal to the platform is the secondary flow. It is called secondary because it is generally much smaller than the primary flow. The secondary flow is principally created by the pressure difference across the Modules and these differences are generated by the primary wind flow. The amount of secondary flow through the Module will depend on congestion within it and the resistance to flow caused by the congestion is measured as a pressure drop which can be expressed as a pressure drop coefficient. Principally the windward faces of the platform will have positive pressures and the leeward will have negative pressures. Thus on the occasion of the accident there would be pressure driven ventilation through Modules C and B from the positive pressure side (the west) to the negative pressure side (the east). However the equipment causing obstruction to the wind in each Module will restrict this flow. The summation of all the drag the wind flow experiences as it goes through the Module can be measured and expressed as the pressure drop coefficient. Of course large objects such as compressors and tanks (known as "bluff objects") disturb the wind significantly as compared with more streamlined objects. Since such objects will cause variations in the air flow through the Module they are in effect creating turbulence. Thus the flow through the Module becomes dominated by local effects. The ventilation flow through the Module can be expressed as a ventilation rate in cubic metres per second. It will not necessarily be constant throughout the module and indeed can vary by a factor of 2 or even 4. The secondary flow can also be expressed as the number of air changes per hour. The foregoing matters were spoken to by Dr Davies and I did not understand that the defenders disputed any of them. Of course in determining how an escape of hydrocarbon within a Module would behave account has to be taken of the secondary air flow which would have the effect both of moving the gas and providing a source of air for the formation of a gas/air mixture. Moreover assuming an escape of 100% gas this would dilute as it moved through the module and mixed with air. If the escape is slightly dense and into a strong air flow it will act much as a neutrally buoyant gas. The defenders argued that this would be the case if the gas that had escaped was predominantly composed of the lighter ends and this they say must be so. Air of course represents neutrality. If you have a very dense release into still air the gas will bear all the characteristics of its heaviness and slump. As such gas dilutes its heaviness is progressively reduced. Thus the characteristics of the cloud may change during the course of its travel. Moreover if a heavy gas was released in the west part of either Module B or C as it travelled through the module some of the gas would become lighter and rise and thus could trigger an alarm in a high location at the east end of the module. As the released gas impinges on the equipment within the module that would also affect its dispersal. If a cloud or jet of gas is released it will take a finite time to arrive at a particular point in the module and to build up the concentration of gas there. However if we are dealing with a continuous release then once the maximum concentration at the point in question is attained that concentration will remain steady for so long as the release continues. That plateau of concentration is known as the steady state value. If that value is not sufficient to trigger a gas alarm then such alarm will never be triggered at that point. Moreover the arrival at the steady state condition does not necessarily reflect the maximum volume of gas within the module since some gas could escape out the end of the module before the steady state condition is reached. Thus the steady state is not necessarily the maximum concentration. With a transient type of release if this is terminated abruptly the gas may never reach the steady state level. Moreover because the gas detectors have specific response time a situation could arise with a transient release that the steady state was reached for such a short period that the detector did not have the time to register what otherwise would have been a concentration sufficient to trigger the alarm. If a liquid jet is released and then flashes the flashing process itself will accentuate the dispersion.
Dr Davies looked at a number of hypothetical release sources in Module C in order to ascertain how a gas cloud might behave in that module and in that respect he had regard to the location of the gas detectors. In so doing he used a model that replicated in basic form the obstructions in the module as these were in fact portrayed in the model of the platform which was in Court. Because the obstructions in the module offer considerable resistance to the air flow particularly towards the east end the initial wind speed as the air first enters the module will be considerably less than the wind speed attained once turbulence develops. Thus the wind speed out of the east end of the module would be greater than the initial speed of the air flow within the module particularly because of the substantial obstruction represented by the centrifugal compressors.
The location of the gas detectors in Module C are shown in number 12/112 of process. The first alarm to annunciate at the time of the accident was a C3 alarm which puts the first concentration of gas observed in the south east quadrant of the module. Dr Davies pointed out that generally in relation to the problem facing him the difference between a release of a neutrally buoyant hydrocarbon and one that is heavier primarily relates to the behaviour of the gas at the early stage after the release. A neutrally buoyant release would tend to spread very rapidly in all directions. A heavy cloud would tend to slump at the early stage of the release dependent on how large it is in relation to the ventilation. A heavier cloud will tend to stratify. Thus the density and concentration will tend to be greater towards the bottom of the cloud. Not only will the cloud be lower but it will be more packed and not spread out so much sideways to the airflow. Essentially it will be more contained. This is because nearer the top of the cloud it will be better mixed with air. The cloud will eventually disperse widely but not so fast as a lighter cloud. If a cloud of gas be it neutrally buoyant or heavier had escaped from the west end of Module C then Dr Davies would have expected it to trigger other alarms before arriving at the east end and the defenders make much of this opinion. Therefore he discounts an escape of gas from the west end of Module C. The gas would have become dispersed as it moved the length of Module C and thus since it would have become lighter a response from the higher placed detectors in the Module would also have been expected.
Dr Davies considered a release from points just to the east of reciprocating compressor A. He assumes a release somewhere along the line just to the east of that compressor and also at mid-height. In relation to a neutrally buoyant gas if it was released at the point being considered then his view is that the detector G101/2 would be unlikely to avoid being triggered. The same applies to G101/3 another detector in the C2 area. Indeed most of the detectors at the east end of the module would see the gas cloud. The point is that a neutrally buoyant gas might be expected to spread out quickly in all directions. He considered that a similarly sized release of a heavier gas from the same point would spread at a lower level and because of the configuration of the module would be more likely to move towards the south east corner. He would expect G101/2 to alarm under these circumstance but G101/1 may not. On the other hand all the gas detectors located at low level would certainly see a low level gas alarm. His assumptions about the spread of different weights of gas is illustrated by his modelling
A release to the north and to the east of the reciprocating compressors of a neutrally buoyant gas was also analysed by Dr Davies. He assumed that such a release would result in a relatively strong flow because of the space contours. It would encounter the process skid quickly and thereafter disperse widely as it mixes with air. G101/2 would alarm but G101/1 would be unlikely to detect. The earliest alarms would probably be in the C5 area. Alarms in the south sector may also trigger depending on the size and spread of the cloud. The release of a heavier gas at the same location, if it was released in such a way that it was not spread at source, would tend to move in a more southerly direction round the process skid. However it would also go through the process skid and would mix to a degree that it would rise and trigger G101/2. It would also trigger low placed detectors in the C5 and C4 areas and possibly also in the C3 area.
Regarding a release of gas just to the east and south of the reciprocating compressors, then if the gas was neutrally buoyant Dr Davies would expect the flow to be highly turbulent. He would have expected the flow to move from west to east not encountering many obstacles. Spreading would be encouraged by the fact that the flow would be faster on the north side. G101/1 would be a strong candidate to respond. Because there is some air trapped in the detector itself even very strong concentration of gas would lead to an alarm response and the stronger the concentration of gas the faster the response. Moreover with the spreading with height G101/3 would also detect. All the detectors in the C3 zone would detect irrespective of height and G 101/2 would also almost certainly detect. Because of the distances involved the timings would be short. If the release was of a heavier gas the gas would have a relatively easy path to exit out of the east of the module. He would expect low placed detectors in C3 to respond and then progressively detectors would respond across C4 and C5 depending on the size of the cloud. G101/2 would be a likely candidate to respond but G101/1 may not. Senior Counsel for the pursuers accepted that in the light of the foregoing if a lighter gas had escaped from PSV 504 then one would have expected an alarm at G101/1. However he contended that the escape from PSV 504 had been of a heavier type gas in which case G101/1 may not have detected it. He maintained that on the evidence that detector may have been located at a height of about 20 feet in which case there was more possibility that the gas could have escaped detection by it. These considerations represent a vital difference between the pursuers and the defenders since the defenders claim that on their analysis of the evidence of Dr Richardson the first stage release upon jagging would in fact have been effectively a neutrally buoyant gas cloud. This is in contra-distinction to Dr Davies who modelled the alleged gas escape as a propane equivalent. It is clear that as Dr Davies stated the dispersion of a gas release in general terms will depend on its initial location in terms of how far it will travel in Module C and also the precise dispersive nature of the environment in which it starts from. It will also depend on the composition and temperature of the gas.
The defenders contended that for Dr Davies’ views to have validity then at best to explain the fact that the C2 detectors did not annunciate in relation to the first stage release one would require conditions where there was a release of a heavy gas at PSV 504 with the release pointed downwards. This they say is not what the evidence points to. Dr Davies said that if you can get about 20 to 23 kilograms of gas out either in one puff or a low continuous release then if it is a heavy gas that comes down and under the C2 gas detector that could cause the C3 alarm. This particular approach seems to be postulated on the basis of escape of a heavy gas. The evidence of Dr Davies was that even a much lower level release such as 2-4 kilograms per minute could trigger a C3 alarm if continuous or even terminated after a short time provided that a steady state is attained. It is a puff burst of gas at a low level rate of release that could not trigger a C3 alarm. For such a short release a much higher rate of release would be required.
In relation to the natural gas as the gas that he uses in some of his tests Dr Davies defines this as being in relation to his work a mixture of methane, ethane, and propane which was so mixed as to be neutrally buoyant. He seems in his work to use ‘neutrally buoyant gas’ and ‘natural gas’ interchangeably. It is perhaps a pity that we do not know that the natural gas he used contained components in the same proportions as condensate.
If gas is introduced into a still atmosphere the gas can spread by the process of diffusion. If on the other hand it is introduced to moving air there is only diffusion at the interface between hydrocarbon and air but the turbulence or ventilation can also accelerate the diffusion process. In any calculation of flammable mass of gas at a particular location ventilation and turbulence have to be taken into account. The continuity of the source of gas is also an important factor. As his approximate starting point Dr Davies considers that you need a continuous release at the leak source of about 100 kilograms per minute to develop a significant flammable mass at the east end of the module. Moreover if there was a release as low as 4 kilograms a minute there would be no alarms at all. The concentration of gas to air will vary as the gas gets more distant from the leak source and becomes more diluted. If the release is continuous then the gas cloud at any point will increase in concentration to reach a steady state and this should not change so long as the release continues at the same rate. On the other hand if the release is short lived a particular location may never see the steady state value of the gas because such gas as there is will move on too quickly. Of course at different points the steady state concentration will not be the same since the further away from the source the concentration is going to decay. The point that the defenders seek to make out of this is that if any detector is going to see a particular level of gas it is likely to be the one nearest to the leak source.
It should be noted that the above views are those which Dr Davies also expressed on the basis of his general experience so that they rest independent of his methodology in modelling. Moreover although he was asked to assume that the gas which escaped was a heavy gas he was highly experienced in the behaviour of gas and he never doubted that a leak of condensate could conform to the assumption given to him.
Dr Davies also expressed general views as to how he would expect gas to behave if it escaped in Module B. Number of process 41/1 shows the location of gas detectors in Module B. He is of the view that as Module B is a more open module it would be likely to have a higher ventilation rate than C. The flow would be largely a turbulent mixing flow. There are open corridors in B which would speed up the flow. He opined that if a gas cloud was released towards the west end of B then he would expect many of the detectors in the module to see gas concentrations above the alarm levels. If material proceed to the east end of B so as to permit ingestion into C at alarm levels then the original escape would have required to be well above these levels because the gas would have been substantially diluted as it moved along Module B. Thus a release in Module B that could result in serious ingestion into C would have required to be a major release that would have triggered gas alarms in B (assuming that these were operational). I have no difficulty in accepting that view. The gas would stratify as it progressed so that both low level and high level alarms would be triggered and the alarms triggered would include G99/1, G99/3 and G99/4. G19, G20, and G21 would also see alarm concentrations. He specifically considered what the gas behaviour might have been if a natural gas (that is a neutrally buoyant gas) had been released about the middle of Module B but towards the north. He would have expected such a cloud to have been seen by a number of detectors including G99/3 among others. The lower detectors in the module had been placed in a low position particularly to enable detection of the heavier hydrocarbons.
If a dense cloud of gas had come out of the east end of Module B it would tend to drop and Dr Davies would have expected detector G127 at the 68-foot level to have seen gas. However that detector is external. In the present case since there certainly were men working at the 68-foot level when the accident occurred one might have expected that they would have noticed any material intrusion of gas into the platform at that level particularly if it had exploded.
The quantitative work that Dr Davies did in 1989 is summarised in his report number 12/362A of process. He did not at that time perform his quantitative exercise in relation to Module B because as he stated he considered that the possibility of a release there not triggering alarms was extremely remote. Moreover because the wind pattern would have spread the gas along the east face quite widely he thought that the creation of a significant inventory of flammable gas within Module C by such a process most unlikely. However in the light of certain averments of the defenders in the present actions he did in 1993 carry out some quantitative tests relating to Module B. Dr Davies also approached his tests on the basis that between the first alarm and the second flurry of alarms noted by Mr Bollands a period of 2 or 3 minutes had lapsed. Mr Bollands’ evidence in this respect was indeed not challenged. In the view of Dr Davies the fact that there was a gap between alarms of longer than about a minute would be sufficient to require two separate events. Dr Davies also said that if there was a flurry of alarms that event would signify a largish release of about 100 to 200 kilograms per minute. Thus in the first instance the flurry of alarms noted by Mr Bollands might point to a gas release of that size.
Dr Davies assumed in his 1989 tests that he could regard a concentration of 0.75% of the neutrally buoyant streams as being sufficient to trigger a low-level alarm and a concentration of 3.75% to trigger a high-level alarm. These figures are based on the lower explosive limit for methane. Indeed the detectors were calibrated for methane. When he came to do further work in 1993 he used the figure of 3.3% as the lower explosive limit of the neutrally buoyant stream. This was taken from Dr Balfour and the change should not affect the implication of his results. In relation to heavier gases, such as one would find in the condensate streams Dr Davies, used a lower explosive limit of 2.15% again taken from Dr Balfour. He used 1:100 and 1:33 scale models for his work and these were said to be accurate representations of the salient aerodynamic features of the platform and Module C. Because the detectors were calibrated for methane calculations were required to adjust the explosive limits for the heavier gases. These calculations were done by Dr Balfour and appear in his report. Thus for a calibration set for the low level alarm setting, namely 15% of the LEL of methane, the equivalent value required to generate a low level alarm for condensate would be 23%. The high level value for the heavier gases was taken by Dr Balfour as 115% of the LEL of 2.15% to give him 2.5% as being the level at which a high level alarm would be triggered.
Regarding the response times of the gas alarms Dr Davies again used the figures in Dr Balfour’s report. Thus for example for the Sieger 910 detector the time to alarm for the first alarm set at 15% LEL would be 1.5 seconds and the time to the high level alarm would be about 14 seconds. The applicable response times of the detectors are set out at page 5 of Dr Davies’ report. It will be seen that the response times for a low level alarm are in practice very quick.
PSV 504 was located about 15 feet above the deck and to the south and east of the reciprocating compressors and Dr Davies worked on that assumption. To adjust for possible imprecision in the exact location of the valve, in his 1993 experiments Dr Davies changed the assumed position of the valve about a metre northwards from the position that he had used in 1989. Dr Davies also worked on the basis that the flammable mass at the initiation of the explosion was about 40 to 60 kilograms. By flammable mass I mean the amount of hydrocarbon needed to be within flammable limits for the purpose of the initial explosion. These values for the amount of gas in the explosion were taken from the evidence of Dr Mitcheson and Dr Cubbage. Of course the flammable mass may not represent all the gas that has escaped from a leak source since some gas might for example escape or not be absorbed in the gas and air mixture. In any event a proportion of the leaked gas will decay. Indeed Dr Davies said that if a flammable mass in the region of 40 to 60 kilograms is required you would need a hydrocarbon release of about 140 to 160 kilograms per minute over 30 seconds. However the defenders contend that although a release rate of 180 kilograms per minute is quite possible according to Dr Richardson this rate is only going to last for about 6 seconds because at that point the flow will revert from liquid to gas. Dr Richardson’s calculations are of course all related to total mass and it is Dr Davies who translates this into flammable mass. There is of course a practicable limit in the release rate that would be consistent with the alarm patterns observed. Thus if the release rate was much higher than say 200 kilograms more alarms would have annunciated. However Dr Davies considered that the alarms could be accommodated with a release rate of between 150 to 200 kilograms per minute over about 30 seconds. Of course these views of Dr Davies were based on tests with propane. It should also be noted that the flammable mass is considerably less in a natural cloud than in a propane cloud because of differing dispersion and LEL levels. These broad estimates were generally based on the size of a flammable cloud that at the time of the explosion would not have generated a flame or hot combustion products in the Control Room. The witnesses who had been in that location when the explosion occurred did not experience either of these phenomena and this gives some support that the gas cloud was restricted in mass. Dr Mitcheson had worked on the assumption that the flammable cloud which exploded after expansion would have occupied about three-quarters of the volume of the module (taking the equipment at the east end into account) and that the explosion would generate an expansion of 7.5 to 1. He assumes that the module had a capacity of about 4,500 cubic metres. The total volume of the module was about 5,175 cubic metres and Mr Wottge had estimated that about 15% of the module was occupied by piping and equipment. That would leave about 85% of the volume of the module available for the gas. Dr Mitcheson indicated that in his calculations he did not take into account any material that might have escaped but been outside the explosive limits. Essentially he was looking for the mass of the flammable cloud. He expressed the view that in order to generate the sort of over pressures which occurred he would have expected the minimum mass of the flammable cloud to be about 15 kilograms. He is of course talking about condensate gas. He thought the upper bound would be about 50 to 60 kilograms. Thus he arrives at what can be no more than a broad yardstick. Mr Cubbage agrees that Dr Mitcheson’s figures would give a good view of the mass of flammable gas likely to have caused the explosion. On this part of his evidence Dr Mitcheson was not challenged and the defenders led no evidence to contradict it.
Dr Davies had to take account of ingestion of gas into the centrifugal compressors. With the tripping of these pumps combustion gas would no longer be ingested but that may not have stopped the ingestion of air for ventilation purposes. Combustion air when taken into the compressors was taken in via intake hoods positioned on either side of each of the compressors. The situation is shown in the schematic which is number 12/126 of process and the plan number 12/106. The air for combustion goes into the turbine section where combustion occurs. Most of the air for the combustion process is drawn in outside the module from beneath the external grating but about 1o to 25% is drawn in from Module C itself. When the compressors were working there was a total requirement of combustion air of about 27,000 standard cubic feet per minute. If gas were to emerge from Module B whether or not it might be ingested into the combustion system would depend on the ambient wind conditions and its weight. But in any event such gas would have to drop down and then be drawn up. There was a gas detector, G102/2, sited within the air intake hood of compressor A. There would be another detector, G 102/1, at the south intake in a equivalent position. The other two compressors had similar arrangements. In addition to air for combustion purposes these compressors ingested air for ventilation purposes. The gas turbine ran at relatively high temperature producing radiant heat so that coolant air was required for both the turbine and compressor compartments of the compressor enclosures. The air intakes for ventilation purposes is shown in number of process 12/126 where this aspect of the system is illustrated in red. The air is coming in from the outside of Module C and at the same level. Gas emerging from Module C would flow directly into the paths of these intakes. The air is drawn into the intakes by fans. After being drawn in the air goes to the compressor and turbine compartments. In relation to Compressor A there is a gas detector, G26, positioned almost at the mouth of the intake. That would probably be the first detector to register any gas drawn into the compressor through the ventilation system and it is in the C3 zone. However five detectors were available in C3 which might have triggered the first alarm noted by Mr Bollands. The ventilation air drawn into the compressors totalled 8,000 standard cubic feet per minute and of this 75% went into the turbine compartments and 25% to the compressor compartments. The compartments were pressurised though the pressure difference between the turbine compartment and the module was greater than in the case of the compressor compartment. The ventilation fans would run for about 2 hours after a compressor trip in order to ventilate and cool the compartments. The detector G27 is situated on the outside wall of the turbine compartment and G28 within the compressor compartment. Since the alarm panel in the Control Room can only accept one C3 low level alarm at a time it is possible that what Mr Bollands experienced as the first low level alarm in fact indicated responses from more than one of the C3 detectors. If high gas alarms were to occur in G27 or G28 the compressors would trip but not the ventilation. On the other hand if there was a high concentration alarm perceived at G26 then not only would the compressor immediately shut down but the ventilation would shut down as well. Thus it would appear that the centrifugal compressors were not caused to trip by a high concentration of gas because they seem in the case of the accident to have tripped before a high level alarm annunciated. Ventilation from the turbine compartment in each compressor vented through ducting to the south of the compressor as shown in number 12/126 of process. Ventilation air for the compressor compartment was taken into Module C through louvres at each side of the west end of the compressor compartments. The fact that each compartment had 2 louvres can be deduced from the said schematic and by the video film number 41/ 21 of process. Dr Davies proceeded on the basis that the exhaust amounted to 2000 standard cubic feet per minute and that this would exhaust equally between the 2 louvres. Thus 6,000 scf would be exhausted by the turbine compartment exhaust system.
With regard to the combustion air intakes Dr Davies was of the view that they had little influence on the air flows within Module C but he did not consider what effect if any these may have had on gas detectors. In his 1993 tests Dr Davies could not find a mechanism, even assuming that all the centrifugal compressors were running and the compressor compartment louvres were emitting 100% gas, which would account for the annunciation of a C2 detector. Before anything approaching a high level of gas could be expelled from the louvres the G26 alarm at the ingestion point would have to register. There was of course a C2 detector above the central centrifugal compressor. The closest C2 detector to the east of Module C was G101/3 which was situated above and outside the C compressor and was about 15 to 20 feet above the deck level. There are three alarms altogether in the C2 zone. The other two are G101/2 and G101/1 which is more inboard and towards the south. Their positions are shown in number 12/112 of process. Another alarm that may be critical to the pursuers’ hypothesis and which Dr Davies required to consider was G103/1 which the pursuers suggested was the C3 detector most likely to have triggered the first gas alarm noted by Mr Bollands. That detector was located within the fuel gas valve compartment of the centrifugal compressor C and was located beneath the walkway grating. Dr Davies considered that if a quantity of gas did get into the Centrifugal Compressor this would probably dilute with time because of the air being ingested.
In modelling one has to adjust the results of the modelling to take account of the effect produced by the scale of the modelling. Certain pragmatically derived compensatory numbers can be applied to the result and these are known as Froude numbers and Reynolds numbers respectively. Dr Davies considered that in the case he was dealing with it was only necessary to take account of the Froude number. This was as he explained because Froude and Reynolds numbers are incompatible. He was tested at length as to why he had omitted the Reynolds number but although his reasoning was very complex in the event the defenders led no evidence about this matter and I have no difficulty in accepting the reliability of Dr Davies on the subject. In testing he used two kinds of wind tunnel. One was a blower tunnel which was used essentially to blow air into the model. The second tunnel was much larger and was know as the number 7 environmental tunnel. Dr Davies was able to work out a pressure drop coefficient for Module C by attaching a 1:33 model of the module to the blower tunnel. Having got that figure he measured the coefficient of Module C for the 1:100 scale model and then adjusted the model until he could get a similar pressure drop coefficient. Thus he had secured a coefficient value for the larger more reliable model. Having got a ventilation rate for the 1:100 model he replicated that in the 1:33 model which was then attached to the blower model for the gas dispersion tests. His value for ventilation rate was 46 cubic metres per second which is equivalent to a wind speed of about half a metre per second. This can be compared with the external wind speed of about 8 metres per second. Dr Davies considered that on a fair estimate there would be about 50 air changes an hour in Module C. The significance of such speed is that a particular cloud of gas will have a finite time to reach the end of the Module and escape into the atmosphere. The average rate of air movement can be calculated form the air changes. Thus Dr Davies considered that this would give a ventilation rate of about 54 metres per second which compares well with his measured value of 46 metres per second. As I shall consider later as hydrocarbon moves from the source of a leak there will be concentration decay. As the defenders argued if you have a release source 5 metres from a downwind detector and there is another detector 10 metres downwind then not only should the gas cloud hit the nearer detector sooner but it should also have a higher concentration of gas when this happens. The defenders contend that on this supposition there is no way that there could be an alarm at C3 without an alarm earlier having been triggered at C2. The defenders point out that the pursuers’ hypothesis would involve the gas cloud triggering a low level detector, somehow developing so that it later triggered an alarm at C2, then somehow increasing to trigger a high level alarm at the compressors. However it was accepted that this analysis may be more suspect if in fact the gas cloud is composed of heavier ends.
The defenders accepted that a puff type release would not necessarily attain the steady state concentration that would result from a continuous release.
For Module B his values would be higher because that module is not so congested. For Module B he would expect the air speed to be about one-tenth to one-twentieth of the external air speed. He was certainly cross-examined about the accuracy of his estimates in relation to the modules for the air flow was introduced straight-on to the model whereas in reality the wind was blowing at a slight angle to the platform. Dr Davies explained that unlike the situation at the very mouth of the modules within them the air flows would be largely dominated by the turbulence and pressure differentials rather than the external wind. Moreover it was suggested that the model used in the 1993 experiments may not have represented the modules with sufficient accuracy. However no evidence was led to support the cross-examination. For his 1993 experiments Dr Davies decided to adjust the 1:33 model to improve its salient dynamic features and it was this model that he used for his work on the interaction between Module B and Module C. It was this model that he placed in his wind tunnel. His approach was to apply to Module C the air flow that would reproduce the ventilation flow he had found was applicable to that module. With that assuming the model gives an accurate representation of the aerodynamic features of the modules then the experiments should give valid values for both modules. It was suggested to him that heated equipment within the modules could influence air flow patterns. Dr Davies was also cross-examined on the question of how convective heat may affect air flows. His response was by reference to scientific literature to the effect that convective currents would have to be exceptionally strong to overcome the dominance of pressure differential and turbulence. Dr Davies was very experienced in the use of models for air flow experiments and he also made reference to the literature on the questions now being considered to back his views. In the absence of contradictory evidence I see no reason not to accept his judgment on these matters.
In his experiments Dr Davies sought to simulate both natural gas and condensate. For natural gas he used a mixture of carbon dioxide and helium to give him a neutrally buoyant gas. This had a molecular weight of about 18 to 20 kgs. For condensate vapour he used a mixture of argon and freon to give him a heavier than air gas approximate to propane. This had a molecular weight of about 42 to 44 kgs. In relation to the probes so far as affects the heavier gases the calibration of these probes was challenged by the defenders both in cross-examination and through their witness, Dr Bruun. To carry out his tests Dr Davies had manufactured devices which would produce gas releases both of the circumferential type and partial circumferential type. The amount of gas released could be controlled. These release mechanisms were improved for the 1993 experiments. He had various release points in his representation of Module C. Position 1 was in the general area of PSV 504, position 2 being the equivalent position opposite near the north wall, position 3 which is generally within the centrifugal skid area, and position 4 which is in the west of Module C. His thermal conductivity aspirating probes were placed in positions representing the positions of gas detectors in the module (G probes). He also had a number of additional probes (designated as B probes) and they were intended to test additional points in the module. The probes operated on the hot wire system. That is as air or gas passed over them the loss of heat would produce an output by way of an electrical signal. These probes would respond to gas quickly and to that extent they differ from detectors on the field which have response factors. The electrical signals transmitted by the probes are converted into a number which reflects it size. Those numbers are the essential output and the number relates to voltages. As far as Dr Davies is concerned his methodology was established and well tried. His 1989 results are set out in reports 12/362 A and B. He divided his tests into different series. A series would set out the applicable scenario as for example the fact that there was a release of a particular gas from a certain location and simulating certain conditions of release. Thus for example he carried out tests for a variety of release conditions such as jet leaks or circumferential leaks. For each series there were a number of runs. Thus in effect a series was a group of runs at the same conditions. He also sought to model the ingestion and exhaust of gas from the centrifugal compressors. He tested the various possibilities for neutrally buoyant gas leaks and for heavy gas leaks. From his data Dr Davies worked out the consequential traces manually.
In respect of the runs at position 2 Dr Davies released simulations of propane (series 43) and natural gas (series 46) respectively. In the first case the supposed rate of release was 37 kilograms per minute and the sequence of low level alarms he discovered to be C2; C4 ; C3 and C5; whereas the only high level alarms which would have responded was in C2. The release was designed to simulate a partially circumferential leak. He sought to measure the steady state concentrations as a basis for his results. Thus Dr Davies was able to measure the time to particular detectors in particular conditions and from that to know the sequence of alarms. Dr Davies was able to conclude that neither of the position 2 scenarios resembled the alarm patterns reported by Mr Bollands. In fact he becomes rather more positive than that for he asserts that in the light of his general experience he would not expect any scenario constructed round the position 2 release point to provide the necessary alarm sequence.
Position 3 was intended to represent a point on the centrifugal compressor B skid - that is a point just to the west of the compressor itself. Again he concludes that a leak from that point would not fit Mr Bollands’ gas alarm pattern. He likewise found that any leak from position 4, at the west of Module C would produce a well dispersed cloud of gas which would affect a number of alarms within the module and not in the pattern observed at the accident. On the other hand in relation to position 1 which seeks to replicates the site of PSV 504 he is able to conclude that with this site generally C3 was the first zone to detect gas when compared with C2; C4; and C5. However it is necessary for his results that the leak is generally downward-pointing and involves a heavier than air material like condensate. This is the circumstance that the defenders say was not proved. The need for a downward-pointing leak is said to be the fact that only such a leak of a heavier material could avoid setting off the G 101/1 detector which would be in the path of a eastward moving cloud from position 1. Thus subject to his qualification Dr Davies tends to favour the site at position 1 as being a likelier source of the leak than the other positions he explored. He found from his tests that looking to a release rate of about 90 kilograms per minute he could conclude that longer release times like 17 seconds were too short and that a value such as 13 seconds was appropriate to set off an alarm at C3. He accepted that if the release rate were 180 kilograms per minute then a puff release of about 6 seconds could trigger the C3 alarm. However one could have a higher rate for a shorter period of time or a lower rate for a longer period of time. He makes the point that there is a definite correlation between rate and time. To get to his result he of course requires a shrunken cloud of heavier gas. Indeed the defenders did not challenge that the foregoing results were appropriate if dealing with a propane cloud and they accepted that much of the force of their challenge would depend on their hypothesis that the first stage release would not in fact behave like a propane cloud. Moreover Dr Davies does accept that a neutral buoyancy release such as he had been considering would not miss a C2 alarm. For his opinion Dr Davies relies not only on his quantitative material but on his general experience. As he declares "moving the release position to anywhere else never produced C3 as the first alarm". However he acknowledges that although his results fit in well with a first alarm generally there should be a shorter time than 2 minutes before the second series of alarms. Senior Counsel for the pursuers sought to urge me to accept the view that the problem highlighted by Dr Davies would fit in well with a scenario where there had been a two-stage release. Indeed his view was that the pattern of alarms which occurred could only be explained if the first release was a relatively small release. By that he means a release of 2 to 4 kilograms per minute were the escape to be continuous. As we shall see there are certain difficulties with that view. On the other hand a puff type release of a restricted quantity of gas could cause the same effect. The gas cloud would require to be just enough to set-off the C3 alarm. However Dr Davies could not envisage circumstances where a continuous release of gas could account for the presumed alarm pattern. In his 1993 report Dr Davies gives further consideration to the possibility that there was a puff type release. The essence of a puff is that there is a sudden release and then it is curtailed. He explained that the key features of a puff are the release rate when it rises to its maximum and the duration of the release. Thus 75 kilograms of gas a minute at 17 seconds may produce the same sort of effect as 90 kilograms a minute at 13 seconds. If the release rate is put up then to produce the sort of cloud that would explain the alarm pattern one has to shorten the duration. In fact it was possible for Dr Davies to simulate a single C3 alarm with a puff type release from the locality of PSV 504. He concludes that for a release time of 15 to 20 seconds no low alarm would be triggered if the release rate was less than about 70 kilograms per minute but that one alarm would generally register for releases greater than 100 kilograms per minute. Whatever the correlation between release rate and time it would be necessary to generate a mass of propane type gas of about 20 to 25 kilograms to set off the C3 alarm. With some rates of release and times the critical mass is never reached because the ventilation has removed some of the gas from the module before that mass is reached. To consist with the alarm pattern the release has to be relatively quick. A first release of the kind necessary to create one low alarm at C3 would not generate a sufficient gas cloud to have caused the explosion. In Dr Davies’ view the quantity of gas necessary to generate an explosion would result from the second stage release of a two stage release. He also accepts that the dynamics of the situations he has been modelling are complex so that some care is needed.
Because of the phenomenon of pressure drop any escape of gas would not be at a steady rate. Thus Dr Davies was asked to calculate what would be the escape if the initial release rate was 30 kilograms a minute and after the first 9 seconds the release rate halved and continued to do so thereafter at the same rate. This is to take account of the fact that escape rates drop with time. The calculation brought out a total discharge of hydrocarbon of 6.33 kilograms which the defenders said would not produce a single C3 alarm. Even if the leak is continued for 36 seconds the release rate drops to 1.8 kilograms per minute. To arrive at a discharge of 20 kilograms the escape would have to last 10 minutes and no gas leave the module. The calculation of Dr Davies is set out in number 68/1 of process. The calculation above represents an eventual revisal by Dr Davies of his original opinion.
It has to be noted that in terms of the ventilation rate it would take about 12 to 13 seconds from any gas release for that gas to reach the compressor detectors and then of course perhaps 15 seconds for a high alarm to respond. Perhaps another 5 seconds or so would pass before the explosion. This means that if there is about 2 minutes between the first C3 alarm and the explosion only about 30 seconds was occupied in the time from release to explosion. Thus there must have been at least about one and a half minutes between the final jagging and the one that preceded it. It should perhaps also be noted that Dr Davies used his general engineering judgment to determine the proportion of total gas mass that would be flammable. In Dr Davies’ tests it was generally 103/1 that was the forerunner of the C3 alarms.
In relation to the first stage release the defenders spent some time developing their thesis that Dr Davies was wrong to suppose that the gas escaping at this stage would behave as a propane gas. The defenders suggested that Mr Clark’s flow chart shows that the stream coming off the JT Flash drum had a molecular weight of about 22 and they asked why the stream leaking from the pump would be any different. The condensate that goes into the injection pump comes from the JT Flash Drum. However a larger proportion of the lighter gas and in particular the methane will have flashed off in the Drum so that the material proceeding to the injection pump should have a higher proportion of the heavier ends. The defenders maintained, and in this they were correct, that with heavy hydrocarbon such as the crude oil (with a molecular weight of 38.7) there is still a capacity to flash off lighter ends such as methane. However the gas that flashes off in such circumstances would be slightly heavier than air. The defenders’ point is that one would expect the gas flashing off the condensate entering the pump during the first stage of jagging to have a molecular weight falling between that of the flash drum and the oil flow. Dr Drysdale and Dr Balfour indicated that the proportions of components representing 80% of the condensate flow in stream 360 and in relation to the stream as a whole he finds that it contains 19.6 % methane, 18.3% ethane and 30.7 % propane. Dr Balfour further indicates that if the condensate sustains a substantial release of pressure about 50% of the propane will flash off giving a percentage of 15.3% of the total. Of course flashing into a restricted space such as a pump chest is different to a situation where there can be unrestricted flashing but the defenders point out that Dr Richardson said that the lighter ends will flash off first as the condensate flows into the largely empty pump. The reduction in temperature will inhibit the heavier ends from flashing. The defenders claimed that if there is more methane than propane the mixture will be marginally lighter than air. It seems clear that the flash fraction of the condensate is roughly 50%. Dr Drysdale calculated that if the pressure in the pump was 20 bar the density of the gas that would flash off would be 20 kilograms per cubic metre. The evidence submitted the defenders is that the gas that flashed off in the pump would be approximately neutrally buoyant. Defenders’ Counsel maintained that the average molecular weight of the flashing mixture can be looked at but I was not sure that the gases would not stratify and although there was some evidence that the mixture will have certain characteristics the question of stratification was not specifically explored. Attractive as the defenders’ argument may appear at first sight the experts seemed quite content to work on the supposition that the leaked gas would be similar as a propane gas and they were not really asked if this could not be the case. The defenders construct their hypothesis from fairly general evidence.
It should be noted that Dr Davies seemed to concede that his "natural gas" represented a mixture of propane, methane and ethane. Dr Richardson it was contended gave evidence that it was only after a pressure in excess of 43 bar that liquid would escape because before that point there would be a gas plug which would permit further flashing. I certainly consider that at 43 bar it is eminently possible that liquid would emerge from any leak since one is at least about the area of sufficient pressure to form condensate. On the other hand it is doubtful if you would be beyond the gas bubble threshold at 40 bar. The argument was that one needs liquid to escape before escapes at a rate of 100 kilograms per minute or more are possible. It has to be noted that the flashing as the condensate enters the pump takes place in a fraction of a millisecond. Certainly Dr Davies thinks that if the first stage leak is neutrally buoyant G101/1 would be "a very strong candidate" but that view is expressed "unless the source is not diffuse in the way I am imagining it". That last remark may indicate a somewhat tentative quality in Dr Davies’ opinion. He also expected G101/3 to detect and indeed all the C2 detectors. However it is fair to say that Dr Davies says that if the gas escape is neutrally buoyant his expectation would be that the cloud would rise and hit G101/1 first. I think the defenders can derive some comfort from the fact that the pursuers seemed to accept that it was central to their assertion that the first alarm to go off was caused by a leak during the jagging process resulting in the first gas which triggered the alarm being heavier than air.
Dr Davies deals with the suggested second stage release and he extended his work on this aspect of the case in 1993. He deals with the additional work in his report 14/53. He indicated that to trigger the multiple alarms said to have occurred at this stage a increased leakage rate was necessary. Moreover such an increased rate would have been needed to generate the mass of gas required to cause the type of explosion which occurred. Indeed he considers that leakage rates of 100 kilograms per minute may have been necessary to develop a flammable mass in the range from 40 to 60 kilograms. If the time scale of the second release was about 30 seconds then a release of perhaps 140 to 200 kilograms per minute would have been necessary to cause a flammable cloud with a mass of 40 to 60 kilograms. He thought that to avoid a second high alarm requires a release duration of less than 55 seconds but to ensure a first high level alarm requires a duration of approximately 25 seconds. He concludes that these are the time bounds of the second stage and nobody challenged him on this.
Generally Dr Davies was of the view that there was no mechanism that would have resulted in a high level alarm from gas ingested into Module C from the east end. The detectors G27, G30, and G33 were located beside the turbine compartments of the Centrifugal compressors and these were devised so as to detect gas within the turbine compartments by way of a tube which went from the detector into the compartment. A question was raised by the defenders as to whether the said detectors would have been able to detect gas within Module C because of differential pressure. However Dr Davies did not think that these alarms were a significant factor in affecting his conclusions. Senior Counsel contended that alarms such as 103/1, 102/1 and 102/2 could have responded to gas from Module C.
Dr Davies considered that on the basis of his modelling work the probability of a significant inventory of gas accumulating in Module B without triggering a gas alarm was very low. That of course assumes that the alarms were operational. Thus I consider that if the alarms were working properly then apart from other considerations it is unlikely that Module B was the source of the gas that caused the explosion. That means that the evidence about the effectiveness of the detectors in Module B at the time of the accident is central to any decision about the possible role of Module B in the accident. Dr Davies did some work in 1993 on the topic of Module B this being contained in his Report number 65/1 of process (and the appendices number 65/3). In his experiments he set-up a plane of probes at the east face of Module B to replicate gas detectors. To explore the interaction between the modules he set-up a similar plane at the east face of Module C. These planes each consisted of 25 probes to give extensive coverage of the gas concentrations at different points. He set-up probes to replicate detectors G21 and G99/4 on the basis that any gas leaving Module B from the east was likely to be felt by these detectors because they were the two eastmost. In his experiments he simulated gas releases near the MOLs at the west end of B and also at two points towards the east of the module. He tested for both a natural gas and a heavy gas. His view is that if the release had been towards the west end dispersion would have been fairly rapid so that most of the detectors would have seen gas. If the release was towards the east of the module from Position 5 then G99/3 and G20 would receive the gas cloud, whereas the position was less sure in relation to G19. The release rate of gas from Position 5 was based on a supposed rate of 100 kilograms per minute. The pursuers argued that if a cloud of H2S sufficient to poison the detectors in Module B had been at the east face of B then it is likely that some of this cloud would have been ingested into the centrifugal compressors and also poisoned the detectors there. The principal consideration in Dr Davies’ analysis is that any gas originating in B will be substantially diluted by the time it can enter C. Thus where a gas leaving the east end of Module B has an average concentration of 1.63 % then the east face of C would show an average of 0.3 % for the same run. Another run, 107, involved a release from Position 5 at a rate of 100 kilograms per minute and a sucking mechanism was operated to simulate ingestion into Module C by the Compressors’ intakes. The average concentration at the east face of C was only slightly higher at 0.52%. Thus Dr Davies did not consider that the effect of operating the compressor exhausts was material. Dr Davies maintained that as gas moves through air the concentration of gas can only decay. In fact he found that gas decays at approximately the square of the distance from its source. However within a module the amount of decay may be limited by the restricted availability of air. He asserts that the concentration limit of gas equals the release rate of gas over ventilation rate. He reckoned that if a neutrally buoyant gas was released at the west end of Module B at the rate of 100 kilograms of gas per minute then the dilution rate by the time it reached the east end would be about 2.6 %. If the release rate was much smaller then correspondingly the dilution would be greater and the gas would not trigger an alarm at the east end of the module. He also says that a heavier gas would also mix and separate rapidly to become dilute. The gas cannot be taken to a higher concentration of gas by concentrating the mixture at one point. The difference in concentration from a gas leaving B and what would enter C is about one-fourth or even higher. Thus if gas coming from B were to set-off a high level alarm in C the original cloud in B would be well above the explosive limit. Dr Davies found that even a release rate of 150 kilograms per minute in Module B would not create a flammable cloud in C.
He did further work to look specifically into ingestion into the centrifugal compressors. For this purpose he set up his equipment so as to have probes replicating the air ventilation detectors (G26, G29, and G32). He also had probes representing the detectors at the combustion air intakes and positioned a probe to represent detector G101/3 in the C2 area directly above the B compressor. He placed a plane of twelve detectors just to the west of the centrifugal compressors. He found that the average ingested material at the plane of detectors was about 10% to 20% of what had emerged from Module B. Releases of heavier gas and of natural gas at the rate of 200 kilograms per minute and from a point to the north east of Module B showed no concentration at all at the plane of probes in Module C (series 109 and 110). This conclusion was related to the assumed flow patterns at the eastern exterior of the modules and was what Dr Davies considered to be what he would expect. In other experiments he simulates the ventilation that would come into Module C by the louvres at the end of the compressor compartments even assuming 100% gas was being exhausted. He tests on the basis of a flow of gas at the rate of 1000 kilograms per minute and measures the effect of final dilution as it is expelled through the louvres into Module C. Moreover he tests in relation to a neutrally buoyant gas since he considers that this is what would be encountered in the situation he is envisaging. Even in this worst case scenario no alarms trigger at C2. The explanation is that the gas emerging from the louvres is rapidly swept downstream to the east by the strong airflows between the compressors. Thus Dr Davies finally arrives at a view that he could not envisage a scenario that would result in gas generated in Module B setting off an alarm in C2 zone in C. The defenders led no expert evidence to a different effect.
Dr Davies also did work to test the hypothesis that gas originating in Module B might have been ingested in the ventilation or compressor intakes of the centrifugal compressor and thus set off alarms. Series 115 and 119 relate to releases from Position 6 in Module B at the rate of 1000 kilograms per minute. The detectors at the combustion air intakes of the compressors would see gas which would trigger both high and low level alarms. On the other hand the detectors at the ventilation air intakes would not see flammable levels of gas. Of course a heavier gas would sink and therefore be more affected by the combustion air intakes which are lower than the ventilation air intakes. The combustion air intake route does not provide a mechanism for getting gas into Module C. Moreover to have gas in Module C at the C2 detectors such high rates of release in B would be required that high level alarms in the combustion air intake would inevitably be triggered before the gas encountered C2.
It has to be noted that Mr Cubbage and Dr Mitcheson thought that the characteristics of the explosion pointed to it having been caused by a heavier than air cloud.
6.2.3 Dr Bruun’s Criticism
Of course the accuracy of the work of Dr Davies was attacked by Dr Bruun who challenged his calibrations of the probes used. It has to be appreciated that the fundamental methodology used by Dr Davies in his general experimental work is what is being attacked.
The first issue raised was whether the probes which were being employed by Dr Davies would respond to the gases used in his modelling in a linear manner or not. Dr Davies for his part described the relationship between the concentration fed into the probe and the voltage output by the probe as a straight-line relationship particularly for the heavy gas releases. Dr Davies when asked why he had been assuming linearity for all his calibrations stated that this was an empirical decision based on the work his group had been performing for years. He claimed that at the very outset of his work he investigated the suitability of a linear approach with the authors of the British Gas Team paper (Birch, Brown, Dodson, and Swaffield). That assumption had been used in their joint work during the 1980s. Other studies supported the same techniques. The proposition put to Dr Davies in cross-examination was that for heavy gases linearity of concentration cannot be assumed between concentrations of 0 and 100%. Dr Davies disagrees with that on the basis of the empirical experience of himself and other scientists over an extensive period. Moreover his decision to use linearity was practical for it simplifies his experiments. Dr Bruun on the other hand thought that Dr Davies in his calibrations for heavy gas had used a most unusual procedure. He said that in doing experiments one should never just assume a priori a linear relationship. Moreover instead of considering concentrations between 1 and 100% Dr Davies should have investigated the range he was actually using which was 0 to 5%. The first of these points is scarcely fair because Dr Davies’ approach is not that of a person with no experience of the problem arriving at an a priori solution but rather a reliance on personal and general experience on the matter. 0n the other hand Dr Bruun asserts specifically that there are people who have carried out the calibration of what he describes as "certain gases" and that the relationship is not linear. However Dr Bruun does not himself appear to have carried out work to demonstrate that the relationship between concentration of an argon/freon mixture is non linear nor does he refer to any research with that combination of gases. Indeed he declares that he himself is not an expert in doing measurements in gas concentrations. He refers to research by others into freon showing that a curved relationship develops rather than a straight one but the mixture used by Dr Davies was 82% argon which seems to develop in a straight line. Dr Bruun’s evidence on this matter appears to be speculative and academic. Dr Davies and others have used the linear approach in their work for many years apparently without noticing any distortion in their results. Even if Dr Bruun were correct that the introduction of freon may cause a degree of curve there was no evidence to show that this would materially affect results. I do not consider that the issue of linearity affects the utility of Dr Davies’ results.
The second criticism made by Dr Bruun concerns what is generally referred to as drift. This arises from the tendency of the voltage from a probe to vary with time. The system of calibration employed was that readings of voltages were taken of a mixture with zero gas and then a mixture of 100% gas and from such readings the calibration co-efficients were derived by computer process. The matter is complicated by the fact that the signal from the probes goes to a digital converter which converts the electrical signal into numbers. The risk is that after the calibration is completed voltage drift will occur and that should invalidate the results. To allow for drift Dr Davies would do a number of zero runs before the actual test runs. It seemed to be agreed by the experts that at 100% concentration the significance of drift is negligible. Originally Dr Davies had one calculation in cross-examination that appeared to show that even at 100% concentration drift would be a significant factor. However he repeated the calculations and found that they had been incorrectly done in cross-examination. This was because he had done his calculation by reference only to the lower co-efficient whereas the proper approach is to regard the co-efficients as being interdependent. Doing the calculation correctly confirmed that the 100% concentration was not materially affected by drift. Dr Davies considered that the question of drift at zero concentration had been properly dealt with by the zero runs. The degree of drift established by these runs can be off-set. Dr Bruun gave evidence that there was a better equation for calculating the effect of drift than that used by Dr Davies but I am handicapped in accepting that evidence because the matter of the alternative equation was not put to Dr Davies. In any event Dr Bruun eventually accepted that his equation and that of Dr Davies were mathematically essentially the same.
Dr Davies’ calibration tests are recorded in number 65/5 of process. It can be seen that six readings are taken for each probe and averaged. The two co-efficients are then calculated on the basis of the calibration exercise and fed into the computer. In the said calculation it is necessary to look at the interrelationship of the readings at zero concentration and 100% concentration rather than seek to calculate each co-efficient individually. When the computer is fed a voltage for that probe it uses the co-efficients to determine what the voltage represents by way of a gas concentration. Dr Davies gave evidence that in relation to his experience with probes he would expect his methods to produce accurate results to within a tolerance of about 0.1 %. The suggestion derived from the calculations that were put to Dr Davies in cross-examination that the probes were susceptible to considerable error because of voltage drift can be explained by the fact (later corrected) that in his first attempt at the exercise there was an inadvertent failure to take account of the 100% co-efficient. This is perhaps explained by the fact that the calculations he was asked to do in Court are quite complex. The result of the recalculated figures was that the observed voltage variation produced a variation in concentration between .49% and .47%. Such a variation would be within the tolerances that Dr Davies would expect and be prepared to accept. At the end of the day when he gave his evidence Dr Bruun did not challenge Dr Davies’ revised way of doing his calculations. In any event it would only be the variation of voltage in relation to the lower co-efficient that would have to be watched since mathematically the upper coefficient is relatively insensitive to voltage changes. This is because the range of concentrations being tested were low values of concentration (about 0.5 %) so that changes at the upper end of the linear scale do no affect the results so drastically as those at the lower end of the scale.
Dr Bruun for his part claimed originally that a calibration that only took account of variations of the zero concentrations was deficient. He claimed that the equation set out in 96/101 of process was the proper and usual one to apply to hot wire calibrations. Dr Bruun’s equation begins by placing the emphasis on deriving a value for voltages whereas Dr Davies concentrates on deriving the gas concentration. However, as I have said, in cross-examination Dr Bruun seems to agree that the equations in mathematical terms are identical and thus should produce the same results albeit through different routes. A comparison can be made through numbers 44/145 and 96/101 of process. The equations both express the same relationship between concentration and voltage but one is the inverse of the other. But even Dr Bruun’s calculations show that the effect of drift towards the 100% values is negligible when the experimental situation is dealing with a low concentration of gas. Dr Bruun was probably more of a theoretician than Dr Davies but although his equation may have been theoretically more attractive than that of Dr Davies it emerged that after he had used it on the actual data of the test probes the results obtained were not materially different from those of Dr Davies. Dr Bruun’s final position was that although Dr Davies did not use the best method his methods were sufficient to produce results that can be taken as approximately correct. One difference between the two experts was that Dr Davies calculates his offset by reference to his original co-efficients whereas Dr Bruun declares that the preferable method is to calculate new co-efficients before introducing the off-set. I would have difficulty in deciding which of the two possible methods is theoretically preferable but ultimately Senior Counsel for the pursuers conceded that Dr Bruun’s method may in fact be more correct. On the other hand given that it seems to be conceded that any deficiency in Dr Davies’ method would produce a very small margin of error and that his method has been used without difficulty by a number of experimenters over a period of years then I consider that the difference in methodology on this matter is of little importance. This particular question of methodology was not put to Dr Davies in cross-examination and had that been done he may have been able to produce practical reasons for the adoption of his own rather less precise methods.
Dr Bruun also questioned the way that Dr Davies had used the analogue to digital converter. The system which Dr Davies used involved the electrical signal from the probes going to a converter where it was translated into digital numbers. The converter used had a total range of 10 volts that is from plus 5 volts to minus 5 volts. This allowed for a range of 4,096 voltage steps. Each of these steps was equivalent to an input voltage of 0.0024 volts. The output generated by the probes varied generally between minus 1 volt at 0% and minus 0.5 at 100%. Thus in terms of the potential range of the A to D converter Dr Davies was using output from the probes amounting to some 205 voltage steps. The concentration of 0.5 which Dr Davies was using was equivalent to 0.0025 volts which in turn is almost equivalent to one voltage step. It follows therefore that one voltage step almost represents the difference between no gas and a critical concentration. Thus it was suggested to Dr Davies that it was not possible to get proper measurement of low concentrations of gas. Dr Bruun suggested that the proper measurement approach was to apply gain - that is to magnify the signal being sent to the A to D converter - so that more refined measurements could thus be taken. Under Dr Davies’ system it was claimed there was a complete step from zero to 0.5 % concentration and nothing in between could be measured. However Dr Davies maintained that this problem can be overcome by what he referred to as the smoothing process. This was in fact a measurement averaging process. In the course of one second of model use 21 voltage readings are received by the A to D converter. Thus every twentieth of a second or so there is a digit reading. What Dr Davies did was have the computer look at the twenty or so digital readings per second and average them together. This process gave what he called a smoothed concentration value. Averages are obtainable because there are inevitably peaks and troughs in the signals. Thus the difference between the experts is that Dr Davies was content to take his readings on the basis of averages whereas Dr Bruun would seek to take actual readings. One justification of Dr Davies’ method is that looking to the detectors they would never see more than what in effect is the average because their response time would not permit measurement of the very short measurement times issued by the converter. The averaging that is obtained in the smoothing process is essentially a sliding average obtained on a continuous basis. It should be noted that with regard to the detectors one requires to have the critical concentration of gas present over the response period to get the alarm triggered. Dr Bruun says at one point that "when you do scientific experiments the whole objective is to have a true signal". However it is clear in my view that Dr Davies was not concerned with measurements of absolute scientific accuracy. Indeed that kind of accuracy was probably not available over the whole scheme of his tests which inevitably had to depend on a degree of judgment and simplification. The question really is did Dr Davies apply methods that would give reasonable results looking to the practical questions the experiments were designed to answer. Indeed Dr Bruun acknowledged that he was not an expert in the actual taking of measurements. Dr Davies was a practical modeller with years of experience in modelling and I have no reason to conclude that his methods were faulty in any practical sense.
As has been seen Dr Davies is basically concerned with the likely behaviour of gas clouds within Module C and to a lesser extent Module B. As he put it he starts after the gas is out. The nature and characteristics of any prospective leak from PSV 504 was dealt with in the evidence that pivoted rounds the evidence of Dr Richardson. Since what interests us is how any gas that escaped from PSV 504 might have behaved the evidence of Dr Davies and Dr Richardson requires to be viewed in tandem. This is a matter that particularly concerned the defenders since they argued that the assumptions that had been given to Dr Davies and which formed the foundation of his opinions did not really match the facts spoken to by Dr Richardson. Of course a link between the two witnesses I have mentioned is the evidence of Mr Standen as to the what kind of leak orifice there would be with a finger tight blind flange. There are of course witnesses such as Dr Mitcheson, Mr Cubbage, Dr Bakke and others who express views on the nature of gas cloud needed to explain the accident.
The defenders also submitted that the two stage incidence of gas alarms could be explained if the tripping of the first compressor had occasioned a leak at PCV 51/1 and 2. Of course these valves are in Module B. When there was second trip of a compressor this could have caused a second leak about one and a half minutes after the first leak. This may seem logically possible but it was not explored with the expert witnesses and therefore rests on speculation by Counsel. For example PSV 51/1 is situated at the north-east corner of Module B and whether or not a leak there could have possibly accounted for the alarm pattern in C was not really explored. Nor was the possible extent of such a leak investigated. The PCV in question could not leak until the valve opened to relieve pressure and whether a leak would be likely in such circumstances would also have to be considered by experts.
The evidence of Dr Davies depends on balancing complex factors. He has to deliver the correct mass at the detectors at the right time and this is affected by many variables such as the inventory of the pumps at various stages, the process of jagging, the leak rate, the rate of release, the nature of the material released, and ventilation patterns.
The question of the presence of scaffolding affects the defenders’ contentions about Dr Davies’ evidence. The exact position and height of the scaffolding in relation to PSV 504 is not known but is clear that at the time of the valve removal there was a scaffolding erected to give the fitters access to the PSV. It was said that this scaffolding was about 1 metre below the PSV and this is certainly the height that one would expect. Whether it was directly below the blind flange or to one side or another is not known but definitely it could accommodate a number of men and the platform must have been close enough to the PSV to allow access to the valve. Moreover I think it is plain that the scaffolding remained in place at the time of the accident. One of the tests carried out by Dr Davies (series 27) included a scaffolding but the results of this test did not seem to me to have been conclusive. The jet release introduced to the test was directly downwards at 180 degrees. The release rate used was 37 kilograms per minute. The test seemed to suggest that the C2 alarms would detect first. Dr Davies is reluctant to be too dogmatic but accepts that he can only get the required alarm pattern if the gas is not too dispersed. The defenders argued that Dr Davies was ignoring facts to fit his hypothesis. He definitely accepts that scaffolding directly below the leak could create a gas dispersion that would not fit his hypothesis but he does not specifically suggest that such dispersion would occur with every type of leak even with the scaffolding in place. When Dr Davies was asked if a downward stream of condensate leaking from the PSV would strike the platform of the scaffolding the only concession he would make was that it was possible. He accepts that if the leak struck the scaffolding the effect would probably be to disperse the gas. It also has to be noted that all Dr Davies’ evidence in relation to the effect of scaffolding is based on the assumption that the scaffolding was directly below the blind flange. Moreover he accepted that if there had been a partly circumferential release rather than a jet release but in a generally downward direction then different considerations could apply. As he puts it the "interconnections would be different". The defenders argue that the evidence of Dr Davies about alarm patterns with a dispersed gas would eliminate not only the alleged first stage leak but also the second stage.
6.2.4 Dr Richardson and Dr Saville
For the pursuers a key question was if the condensate injection pump had been isolated and depressurised whether the process of re-pressuring by jagging (which involved a staggered introduction of condensate to the pump to protect it against a too drastic change of pressure) could have resulted in a leak at the PSV 504 which accounted for the gas alarms and subsequently the explosion. In this regard the pursuers sought to found on the evidence of their expert witnesses Dr Richardson and Dr Saville.
The witness Dr Richardson was aged 42 at the time of his evidence and was a Reader in Chemical Engineering at Imperial College, London (in fact since giving his evidence he has become a professor). He graduated with a Bachelor of Science first class honours degree in Engineering in 1972 and gained a Doctorate of Philosophy in 1976. Both these degrees were from Imperial College. Most of his research was into various aspects of Fluid Mechanics and he obtained a Fellowship at Cambridge University in that subject. In 1978 he took up a position in the Chemical Engineering Department of Imperial College and has been connected with that institution since. Until the mid-1980s his main activity was in the area of injection moulding of plastic. At about that time he did certain work along with John Saville in connection with blowdown problems in the design of a platform in the Norwegian sector of the North Sea. He uses the word "Blowdown" as being synonymous with "Rapid Depressurisation". Thereafter he developed a close relationship with the oil industry. Along with Dr Saville he developed a computer programme known as "BLOWDOWN" which has been used in connection with 40 to 50 platforms since it became available. The development of this programme was financed by associated companies of Shell and Esso. After about 1985 the depressurisation of vessels became a main interest. This interest involved material flow, heat transfer, mass transfer, and thermo-dynamics. His colleague Dr Saville tended to concentrate on thermo-dynamics whereas Dr Richardson dealt with the other three matters which are essentially what is known as fluid mechanics. However he is aware of the problems in the whole field of fluid engineering since he teaches the subject. Dr Richardson has been involved in extensive research into his area of interest. He is the author of a number of publications including a text book on Fluid Dynamics. He carries out a considerable volume of consultancy work for the oil industry. He does not look at the structural or mechanical engineering aspects of platforms but rather the process or chemical engineering aspects. His experience extends to the processing implications of pumps, valves, and compressors. His experience has been practical and theoretical. As well as appearing four times at the Cullen Inquiry he gave evidence at the Ocean Odyssey Inquiry. In general there could be no doubting of Dr Richardson’s capacity to deal with the evidence he gave and in fact he was an impressive witness.
The pursuers also led as a witness Dr Saville the colleague of Dr Richardson. When he gave his evidence he was aged 58. His evidence was largely technical and in much shorter compass. He was a Senior Lecturer in the Chemical Engineering Department of Imperial College. In 1988 he attained the degree of Master of Arts at Oxford University in Chemistry and in 1961 he was awarded the degree of Doctor of Philosophy (his research being in the field of thermo-dynamics). He worked for a period as a research assistant at the University of Minnesota and in 1963 obtained a research fellowship at Imperial College. He developed his interest in thermo-dynamics under the auspices of Professor Rowlinson acknowledged as a world expert in the subject. He has lectured extensively in thermo-dynamics and mechanical engineering. Throughout his working life he has also been interested in Statistical Mechanics. He has also been extensively engaged in experimental work covering in particular high pressure engineering. He was involved in the development of the PREPROP computer programme which is a thermo-dynamics computer programme. Thereafter he was involved along with Dr Richardson in the development of the BLOWDOWN programme. He is the author of a book on thermo-dynamics and a code called The High Pressure Code" as well as being the author of about 50 to 60 other publications. Like Dr Richardson he carries out consultancy work. Like him also he gave evidence at the Cullen Inquiry and at the Ocean Odyssey Inquiry. He is a member of the High Pressure Technology Association and of the Royal Society of Chemistry.
The defenders for their part led in connection with this matter the evidence of Dr Crofton. When he gave his evidence he was 44 years of age. He held the post of Senior Lecturer in Strength of Materials in the Mechanical Engineering Department at Imperial College so that in fact he was a colleague of Dr Richardson and Dr Saville. In 1971 he obtained a Bachelor of Science Degree in engineering from the Royal School of Mines at Imperial College. In 1973 he obtained an MSc in Materials Technology again from Imperial College and that degree was awarded with distinction. In 1978 he took up the post of lecturer in the Mechanical Engineering Department of Imperial College. His MSc involved the study if the mechanical properties and physical characteristics of materials. In 1981 he was awarded a PhD which included research into High Pressure Engineering. His expertise at that time was more in Mechanical Engineering although with more knowledge of the processing aspect of that discipline than would normally be the case. He acted as designer to a German company manufacturing high pressure equipment. Although he was in the design branch of the company he also was involved in the commissioning of certain new equipment. He considered that since most high pressure equipment is inherently dangerous a sound understanding of stress analysis and material behaviour is required from those who work in the field. He has been involved in the High Pressure Fatigue Laboratory at Imperial College. Thus his experience involves an understanding of how the components in pressure systems work. His publication include about 20 papers on the fatigue strength of components. In his consultancy work he has provided about 200 reports concerning failure analysis of components. He has given evidence in a number of litigations. He is a member of the Institute of Metals and a Chartered Engineer. He was the Vice-Chairman of the High Pressure Technology Group. Among his numerous consultancy commissions he had worked for a German company called Polyflex Schwarz and the interest of that is it involved positive displacement pumps like the condensate injection pumps in the present cases. He also had worked for manufacturers of pneumatic actuators which of course are relevant to the operations of the condensate injection pumps and particularly the operation of jagging. There was little doubt about this witness’ general experience of pumps, pump chests, actuators, and the process of pressure drop. He has given expert evidence on the failure of oilfield equipment and has numerous publications to his name. His experience in relation to Dr Richardson and Dr Saville is rather more concerned with what can be called the hardware side of the matters being considered although all are concerned with problems associated with the containment of high pressure fluids.
As foundation material all three of the witnesses I have here detailed relied on the flow chart prepared by the witness Mr Martin Clark. This is set out at number 13/40 of process. At the time he gave his evidence Mr Clark was the Chief Process Engineer with the pursuers in Aberdeen. 13/40 of process was prepared by Mr Clark as a result of a computer simulation of the Piper Alpha phase 1 operations as at 6 July 1988. Mr Clark was a Chemical Engineer with a BSc and also a MSc in Chemical Engineering. Since 1984 he had worked for oil companies as a process engineer both at home and abroad. He knew the process conditions on the Piper Alpha platform since he had worked on it for some months in the early 1980s had visited it regularly thereafter when problems arose. I had no doubt about his capacity to carry out the sort of exercise his simulation involved. Such simulations are well recognised in the oil industry (as well as being regularly used) and, as Mr Clark explained, are based on mathematical calculation which use correlations developed as a result of experimental work on hydrocarbon mixtures. For a number of years this has been done largely by the use of certain commercially available software packages. He personally in preparing 13/40 has relied on such a package called CHEMSHARE DESIGN II. This programme had been on the market for about 10 years and over a period of about 5 years Mr Clark had himself used it regularly (at times every day). From time to time samples were taken from the actual process and tested and these confirmed the reliability of the model. He tells us that the process flows on Piper were regularly re-assessed. The document 13/40 had been prepared by himself in this way for the Cullen Inquiry.
The chart under the title "stream name" identifies parts of the flow by virtue of a number and this is then contained within a diamond in the flow diagram. The reference to phase is identification of the particular phase that the flow will be namely liquid, two phase, or vapour (it should perhaps be mentioned at this stage that there can also be a substance described as aerosol which is a collection of tiny little liquid droplets forming a mist). The chart has a section for the temperature of the flow at particular points and likewise to the pressure. A further section gives the average molecular weight of the relevant flow process. The reference to "Comp flows" indicates the composition of the various flows. We see that air has a molecular weight of 29 compared with the molecular weight of methane which is 16 and accordingly lighter than air. This would mean that it would be expected to rise in the atmosphere. Molecular weight in this context is a constant. On the other hand if a flow is shown as having an average molecular weight of 42 that would show it was heavier than air and would be expected to fall in the atmosphere. Gases can also be compared by reference to density. Against the foregoing background the defenders contended that if gas were admitted into an empty pump so as to say half fill it the lighter gases would predominate in any flashing off process and if this gas escaped through a blind flange it would have an average molecular weight lighter than air and would thus tend to rise. It will therefore be seen that the foregoing analysis assumes some importance for the defenders. Indeed defenders’ Counsel argued that the document 13/40 of process was perhaps the most important single document in the case. However although he spent considerable time expressing his view of what the document meant he accepted that he was not challenging the input data nor the information extrapolated from it.
The experts seemed broadly to agree that if condensate liquid was released into an empty pump chamber about 50% would flash off as gas so that the subsequent behaviour of this gas could be important.
The flows shown in the chart may be measured in relation to time as being pounds per hour which is a reference to mass or pounds moles per hour which itself is a reference to the volume of molecules passing per hour.
It should be noted that all molecules occupy the same volume so that a mole of methane gas will occupy the same volume as a mole of pentane gas. On the other hand the molecular weight of a mole of methane is about half of the molecular weight of a mole of ethane. The difference in molecular weight represents a difference in density not a difference in volume assuming the substances are at the same pressure and temperature. Density of course is mass per unit volume. If the temperature of mass of gas is increased or the temperature reduced the density of the gas is increased and the volume reduced. Thus with such changes molecules will not change in volume but may come closer together or further apart as their density changes and this of course will vary the volume of the gas if it is gas. A mole of methane and a mole of ethane will occupy the same volume if they are at the same temperature and pressure. but not otherwise. If components are mixed and at the same state (that is pressure and temperature) then it is possible to calculate an average molecular weight for the mixture albeit there is a two-phase flow.
In the foregoing circumstances Mr Clark is able to demonstrate the proportions of the different components that will be in a particular flow and he does this. From his information it is possible to work out the volume that each component will occupy by applying the molecular weight of the component. The average molecular weight can be worked out by dividing the mass flow per hour by the volume flow per hour. The flow chart also provides information on the density of a particular stream under certain conditions of temperature and pressure.
In the situation where there is a two-phase stream in the pump being pressurised some of the liquid will emerge from any leak as aerosol. The lighter ends of these droplets will then evaporate and become gas but not the heavier ends. However the leak itself would be expected to appear as a visible mist.
The chart itself demonstrates that the composition of condensate will vary as the process develops. Thus the proportion of the constituent parts of the stream are very different in stream 350 compared with stream 300. Because of this the propensity of stream 350 to flash is greater than in the case of stream 300 because the former has a higher proportion of lighter ends. Thus the defenders contended that the detector system at the east end of Module C had been designed to cope with escapes of the lighter ends the bulk of the heavier ends having by that time been removed. It was suggested that the only reason there are low lying gas detectors at that part of Module C is because they are related to the various parts of the centrifugal compressor zones. Certainly it can be said that assuming there was an escape from the blind flange at the missing PSV 504 this may not have been a contingency that OPCAL considered that they particularly had to guard against.
The defenders argued that if the effect of Dr Davies’ evidence is to establish that the projected escape at the blind flange was in fact neutrally buoyant then not only have the pursuers failed to prove a critical assumption but by reference to Dr Davies’ evidence as to how a neutrally buoyant gas would behave they have proved that their hypothesis is not possible. It was said that the assumptions given to the pursuers’ experts proceeded on the molecular weight of the allegedly escaped gas as derived from Mr Clark’s chart. However it was said that this does not take into account the effect of flashing.
The starting point for the process flow is the reservoir fluid and with regard to the Piper Field there was a homogenous reservoir which meant that the composition of fluid would not change much over time. In fact the defenders did not appear to challenge the composition of the reservoir fluid. The amount of fluid coming from the reservoir can be worked out from the fluid known to be exported along the MOL. This amount can give an indication of the amount of water removed from the oil because the water content of the oil is also known. Once the separator conditions are taken into account the gas formation in the system can be calculated. Allowance has to be made for the circulating gas used in gas lift. What concerns the amount of condensate being produced are the pressures and temperatures in the system. Mr Clark knew the amount of condensate being exported in the MOL. Each day a report, known as the 4 o’clock report was sent from the platform to the beach and this included particulars of that day’s production. This includes information such as the water cut, the MOL pressure values, and the condensate going into the MOL. Thus on the day of the accident the condensate being put into the MOL was 7548 barrels per day. Mr Clark made use of the reports in preparing his flow chart. However initially he made a mistake in that he assumed that the total production from the platform was 138275 barrels per day including the condensate volume of 7548 barrels. However the condensate figure had been separated so that the total production was in excess of 145,000 about 5% more than he actually assumed. However it is doubtful if this discrepancy makes any material difference to his results. Gas production was dealt with by a different Report showing results at midnight each day and usually communicated to the beach in the morning. 13/42 of process is the Report sent on 6 July 1988 which of course relates to the preceding midnight. Although there is a time discrepancy between the oil and gas Reports they are sufficiently accurate to give a fair account of the overall picture. On the basis of the foregoing data Mr Clark was able to derive the input for his computer exercise. To complete that information he required information about pressures and temperatures at various points in the system. With regard to the centrifugal compressors he had regard to certain documents and as far as the reciprocating compressors were concerned he was able to get information from operators after the accident and the same applied to other parameters. In other words by researching the best material available to him he was able to derive the requisite parameters for his exercises. What he produced was a process flow diagram intended to give an overview as to the way the various process vessels on the platform were configured on the day of the accident. The production he prepared was also intended to show the main process conditions (and in particular the pressure and temperature compositions of the major streams within the process and the relevant flow rates). This was geared to matching a total production rate to the MOL of 130,000 or so barrels a day. Although the defenders took exception to the evidence relating to certain computer simulations this was not so in relation to Mr Clark’s simulation. Nor did they attempt to produce any competing simulation. It was accepted that the calculations of Dr Richardson depended on the accuracy of the flow chart and particularly on the assumption that condensate would flow into the injection pump if it was being re-pressurised. However Dr Richardson said that this is in fact the situation one would expect in practice. In general Dr Richardson considered that the accuracy of the pressure shown by the flow chart was more critical than the accuracy of the temperatures. In particular he thought that a 5 degree variation in temperature would make no difference to his conclusions. Mr Clark of necessity had to work with figures involving a degree of approximation but so far as possible he cross-checked his figures. Thus he was able to check the discharge pressures from the centrifugal compressors by reference to compressor performance curves which were available. Accordingly he was able to conclude that a pressure of 675 psia was not an unreasonable pressure to attribute to the discharge pressure of the centrifugal compressors. The pressure attributed to the reciprocating compressor, namely 1465 psia, was the mid-point of the information he got from various operators. For the pressure in the JT flash drum Mr Clark took a pressure of 635 psia. This was based on the differential pressure of 30 to 40 psia kept between the condensate suction vessel and the flash drum so that condensate could flow into the JT drum. The line which leads from the JT flash drum to the condensate booster pumps is shown on Mr Clark’s Chart as line 330 and it is this which has a pressure of 635 psia. It has a temperature of 55 degrees Fahrenheit. The purpose of the condensate booster pumps was to raise this pressure to 670 psia and this is shown in line 340. This pressure rise is significant because at 635 psia the pressure would not be sufficient to recondense any gas which has flashed off. The stream would then enter condensate injection pump A and B in Phase 2 operation but only one pump in Phase 1 and the purpose of this part of the operation was to increase the pressure to 1100 psia. The effect on density of this was marginal. The stream then is sent by way of a pressure control valve 511 as stream 350. This drops the pressure slightly when the stream becomes 360. The purpose of this is to maintain the pressure for the measurement of the flow. It will be noted that during the part of the process being discussed the contents of the streams do not change. If the injection pump is not functioning the pressure upstream of the suction GOV which then closes will be 670 psia. However since the pump was pressuring the stream up to 1100 psia when the pump stops the pressure on the discharge valve will be that. Even if the booster pump were not functioning the hydrocarbon trapped in the material downstream of it would maintain its pressure until the GOV of the condensate injection pump opened and released it into the pump. During the running of the pump pressure peaks can arise which exceed 1700 psia and bring the relief line and PSV into action. The pump itself trips at the same pressure but it takes a finite time for this to happen. That is why it was the practice to open the manual valve before starting the pump.
If during the initial stage of jagging a proportion of stream 340 enters the pump some of this will flash off and this will draw some of the energy from the surrounding atmosphere in the pump chest which will be drawn off as heat. Such heavier ends as then may be left face reduced temperature and this would inhibit them further from flashing off. If the pressure is 20 bar then the temperature will be about minus 20 degrees centigrade. Such gas as has flashed off will go to the top of the pump and one is left with gas at the top of the pump and liquid at the bottom. The defenders maintained that the gas at this stage would have a lower molecular weight than the molecular weight of stream 340. It is this which is likely to determine how the gas at the top of the pump will behave if it escapes. The pressure hitting the blind flange at this stage will be materially less than 670 psia since it will to a degree dissipate as it enters and partly fills the pump chamber. The defenders argue that the pursuers case is that an orifice exists at the finger tight blind flange before any pressure is placed on it. I think this is rather a narrow view since if a flange is not sufficiently tightened it is obvious to me that the application of pressure is likely to cause strain which will expose any insufficiency in the sealing effect of the plate as fitted. This could perhaps make a difference because Dr Richardson explained the two-phased leak in terms that supposed an enlargement of the orifice during the second stage. Of course at the second stage the pressure on the flange would be increased further.
A point raised in cross-examination was that Mr Clark produced for Lord Cullen’s inquiry a drawing containing information additional to that on the drawing produced in this proof. In particular certain process notes before Lord Cullen had been omitted from the present document. Indeed in the Cullen Inquiry document process note 5 stated "all data should be considered approximate. Data was produced from a computer simulation of the operation based on limited data and the best estimates available". Mr Clark’s explanation was that the discrepancy had not been deliberate. After the original drawing had been produced extra notes were added to it for the Cullen Inquiry. These were not added to the original copy so that when that was copied for the present cases the Notes added for the earlier Inquiry did not appear. I saw no reason not to accept the truth of what Mr Clark told me about this matter.
It was suggested to Mr Clark in cross-examination that it would have been more accurate if he had used a figure for the temperature of the JT flash drum that was about 5 degrees lower than the figure used. Mr Clark accepted that the lower figure would have been better but as Dr Richardson explained that particular inaccuracy does not affect the position.
It was pointed out to Mr Clark that the figure that he had used for flaring was about 20% lower than the figure on the daily report for 5 July. However Mr Clark pointed out that the figure in the report was itself a calculated figure and therefore could not be taken as being too precise. The difference between one set of figures and the other is the difference between 29 million scf a day and 27.5 million scf per day so that for such levels the difference is unlikely to be significant. Mr Clark all along accepted that the data he worked from was in some respects incomplete but was the best that was available. I am able to conclude that 13/40 of process is a drawing and flow chart which is accurate within the broad limits required by the evidence and as I have indicated it appeared that at the end of the day the defenders accepted this for practical purposes.
The pursuers argued that the evidence of Dr Richardson established that the explosion could have been generated by an escape of gas introduced in the condensate injection pump A during re-pressurisation by jagging. It was contended that Dr Richardson as a result of his BLOWDOWN exercise had established that the pump could be re-pressurised within seconds by the opening of the suction GOV. Since both the suction (GOV 5005) and discharge GOV (GOV 5006) are closed after depressurisation if the pump is repressurised by opening the suction valve at that point the fluid cannot escape through the discharge GOV but it would flow into the relief line connected to PSV 504. It would therefore flow up to the blind flange assembly. Because of the action of the pump when it is operating properly normally the fluid would enter the pump at a pressure of about 670 psi and when it left through the discharge line it would have a pressure of about 1,100 psi. If the pressure reaching PSV 504 is above such pressure value as is set (in practice about 1,700 psi)the PSV would open and permit the condensate to the condensate suction drum where it is collected. If the pump is depressurised then the pressure which can be introduced to the pump during re-pressurisation cannot exceed the suction pressure of the condensate because the pump itself is not working to increase the pressure.
The practice of jagging involved opening the suction GOV partly and briefly so that condensate and pressure was introduced into the pump in stages. If the pressure is introduced suddenly and in one stage the internal valves of the pump may be damaged. If the pump not being re-pressurised is actually working then the equalisation line that connects the pumps could be used for re-pressurisation but of course that possibility was not available at the time of the accident. Although jagging could re-pressurise the pump quite quickly the actual time it would take would vary between different operators. The pursuers’ main case of course is that the first jagging of the pump could have caused a leak which initiated the first gas alarm and that the flurry of alarms about 2 minutes later would have been caused by the second or subsequent stage of jagging. We know from Mr Clark’s flow chart that the pressure of the condensate upstream of the pump is about 46 bar. Thus if the orifice of the GOV is partly opened the difference between the fluid pressure and the ambient pressure of the depressurised pump will cause the condensate to rush through the orifice to the pump chest. Since the pump will not initially be fully occupied by the condensate some of this will flash which means that the pump will contain a quantity of gas and also some condensate. There will of course also be some air. If additional condensate was jagged into the pump this would quickly compress the condensate and the pump would eventually be largely full of liquid condensate. The number of times an operator would jag to complete repressurisation could vary from 2 to 5 times depending on the operator. There does not appear to have been a set practice in that regard. One difference that arose between the evidence of Dr Richardson is in relation to pressure drop across the internals of the pump. Pressure drop can occur if the configurations of the pump slow down the build up of pressure. Dr Richardson did not take this factor into account because he thought it could be discounted. It was accepted that there would always be a degree of pressure drop but the question is whether this would be material in the present case. Dr Crofton on the other hand thought that pressure drop was a critical factor and required to be taken into consideration. The pursuers seek to explain this divergence of view by claiming that Dr Crofton was making false assumptions in relation to the internal structure of the pump. One is also of course able to test different theories about the time it would take to re-pressurise the pump by reference to the operators who did the job. Dr Crofton who was very experienced in the workings of actuators also thought that the actuator would take longer to open the GOV than Dr Richardson had assumed. However against this is the fact that Dr Richardson while perhaps not as involved with valves as Dr Crofton nevertheless had considerable practical experience of such valves. It is difficult to be too precise as to how long it would have taken to re-pressurise the pump particularly having reference to differences in the practice of operators. Moreover the witnesses were asked to estimate times based on a very narrow timescale. Thus Dr Richardson assumed that the valve would open in about 8 to 10 seconds. This was based on a view that as a rough guide you could take the opening speed of an 8-inch valve as being about one second an inch. The operator, Mr Henderson, thought that the valve could open in about 5 to 10 seconds which must be taken to support Dr Richardson. The operator Mr Murray talked about 20 to 30 seconds but he was considering the whole process of re-pressurisation rather than the mere opening of the GOVs. Mr Lloyd thought that it would take about 30 seconds for the valve to open but he was not an operator. Dr Richardson assumed for his practical purposes that the valve would open in linear fashion but Dr Crofton took a different view. The defenders were interested in the question of the opening time of the valve because they contended that this had a bearing on the likely time of a single jagging operation. One thing is perfectly clear and that is that there would be differences in the time taken to jag by individual operators because the techniques used were in a broad sense particular to each operator. In order to introduce gas slowly the GOV was not opened by a steady pressure of the button but by pulling and pushing the button so that the valve opened slowly. Moreover the operators varied in the number of jagging operations they employed to repressurise the pump.
Mass is the product of the density of the volume and the density of condensate is half the density of water. Thus if there is 400 litres volume one would have 200 kilograms of condensate. It follows that it is possible to relate the volume of the whole system to an identifiable mass of condensate. Allowing for the occupation of the whole system 200 kilograms of condensate is the maximum amount that there could be. However this assumes that the pulsation dampeners are not pre-charged so that the whole of their volume is available to condensate. With the dampeners pre-charged to about 50% the total volume of condensate would be about 150 kilograms.
Dr Richardson explains that with a leak from an orifice with a diameter equivalent to 10mm there would be a leak of 200 kilograms per minute (according to Dr Davies there could not be more). Dr Richardson explains his exercise in terms of this situation. Nevertheless he is assuming that the pump has been repressurised and the jagging stopped when full pressure is obtained. He says that about 20 kilograms would escape in about 6 seconds. After about 650 seconds the total efflux is about 60 kilograms. The explanation for the difference between the first 6 seconds and the period following is that for the former it would be liquid that is released. After that gas is released and the flow rate comes down considerably. This is also because the pressure in the system is decreasing. However before the flow rate changes dramatically at 100 seconds about 38 kilograms of hydrocarbon are released, say 20 from liquid and about 18 from gas. The defenders of course claimed that these facts which come form Dr Richardson demonstrate that the necessary flammable mass is not attainable. However Dr Richardson was more indulgent than either Dr Mitcheson or Mr Cubbage and thought that 15 kilograms of flammable material could have produced the explosion. Of course with a 15 kilogram mass there may not be enough gas to account for the alarm pattern. The size of the orifice must be influenced by the factors spoken to by Mr Standen in relation to water leaks from the finger tight flanges he tested. He was only able to achieve escapes of about 70 kilograms per minute. The defenders submitted that the effect of Mr Richardson’s evidence must be that the leak source was somewhere other than PSV 504.
Another question for consideration was the point raised by Dr Crofton that if through process upset there was depressurisation at the JT flash drum then a two-phase flow (gas and liquid) would develop upstream of the pump. If the material going into the pump at jagging were two flow and not merely condensate this would affect the pressure calculations. Dr Richardson based his calculations on what was described as the Bernoulli Equation and the pursuers accepted that this equation would not be appropriate to a two-phase flow. However the pursuers argued that there was no actual evidence that the flow into the pump had or could become two-phase and certainly this was not developed with the pursuers’ witnesses.
Another criticism made by the defenders of Dr Richardson’s approach is based on the view that when the valve was opening there would in fact be two orifices - a suction orifice and a discharge orifice - so that as condensate passed through the space between these orifices it would flash and become a two-phase flow. Dr Richardson accepted that there would be flashing of gas in the bore of the valve but thought that because of the minimal timescale involved this would have no effect on his calculations.
The internal structure of the condensate injection pump is detailed in the Schematic 12/136 of process. Very small pressures (about 0.15 bar) would cause the internal valves to shift and admit condensate to the valve chest. The pressure values needed to open the discharge valves is about the same as that required to shift the suction valves. Dr Crofton when he began his evidence was under the impression that there was only one cylinder suction valve, but as the note on the schematic shows there were three and this fact affects the question of pressure drop through the system.
There were pulsation dampeners on the pumps, one for the suction side and one for the discharge side. These were designed to even out pressure surges that might arise in the operation of the pump. These dampeners were pre-charged to operate in response to a particular pressure and a diaphragm within each dampener would move to allow fluid into the chamber up to the precharged pressure. The diaphragms were regulated in their response by nitrogen introduced behind them. The total capacity of each dampener was 75.7 litres. They were located between the relative GOV and the pump. In any calculation of the holding volume of the pump system allowance has to be made for the pulsation dampeners and their effect would depend on the pressure they are exposed to relative to their pre-set value. It is only when the pressure reaches the pre-set value of the dampener that the diaphragm expands and the flow enters it. The pre-charged pressures for these dampeners varies between about 350 and 700 psi. With the former value being typical. This would mean that if the pre-set value was 350 psi when full pressurisation had been achieved the dampener would be about half full. The holding volume of the pipework has also to be calculated. The full capacity of the each pump system was about 400 litres but if the dampeners were not filled that could reduce the volume to about 250 litres. Since condensate has about half the density of water the mass of condensate contained by 400 litres would be 200 kilograms. These figures of volume and mass were calculated both by Dr Richardson and Mr Wottge and I did not understand them to have been seriously challenged. Dr Richardson deponed that without the pulsation dampeners only a small amount of liquid would be expelled through a leak (as compared to hydrocarbon gas). However if the pulsation dampeners came in to operation they would create extra pressure on the liquid and a higher proportion would be discharged. In other words the material in the pump is squeezed. The effect of the dampeners of course must depend on their pre-charge pressure. In the situation being postulated all the liquid will escape first.
With regard to the operation of the suction GOV the pilot latch which operated the system is shown in number 12/186 of process and this is also detailed in Mr Wottge’s Report (12/2). The valve is opened by the effect of an valve actuator and this in turn is brought into operation by the effect of compressed air on a spring. The actuator works on Scotch Yoke principle so that the movement of a lever turns the ball of the GOV (the GOVs being ball valves). The ball can be turned 90 degrees and when this happens the valve is fully open since the hole in the ball meets with the flow. When the push/pull button which controls the valve is pulled compressed air is admitted to the system and in turn operates a piston which when moved admits the compressed air to the actuator. This push/pull button is shown in number 12/186 of process. It follows that there has to be a sufficient pressure of air admitted to the valve opening system to overcome the spring that when depressed admits the air to the actuator and consequently causes the GOV to open. Mr Wottge reckoned that with the operator pulling the push/pull button it would take about 8 seconds for the GOV to open fully. Of course during that time as the valve opens it is allowing an increasing amount of fluid into the pump. Mr Wottge’s estimate of the valve opening time is supported by Dr Richardson who stated that as a general rule of thumb such a valve will open at the rate of about one inch per second and the GOV had an 8-inch ball valve. He claimed that he had spoken to the very manufacturers of the actual actuators used on the GOV and that they had informed him that the rule of thumb figure was applicable to these GOVs. However as the valve begins to open there will a period before the valve presents any degree of orifice (known as dead time). Moreover because of inertial effect the speed of opening may not be linear. However Mr Wottge was not challenged in respect of his estimate of 8 seconds as the time to open the valve. The push/pull button was situated on a stanchion some 2 to 3 feet to the south of the GOV valves themselves. Each of these pumps had its own button and this would have to be operated in any jagging operation. The location of these buttons is shown in number 13/49 of process. If the electrical supply was connected when the button was pulled it would remain open (which meant of course that the GOV would remain open and a supply of condensate to the pump could be maintained) but if the pump was electrically racked out then if the button was not held pulled-out by the operator it would spring shut and the GOV would close. There was some question as to whether the particular pump on the night of the accident had been locked out or racked out but Mr Lloyd, a senior electrical engineer on the platform, indicated that if the pump had been withdrawn for planned maintenance the normal procedure would be to rack it out. In any event the operator would know from the response of the button whether the equipment was racked out or locked out and could if necessary have pushed the button closed.
Mr Grieve described a jagging operation. He indicated that once the button had been pulled noise would indicate when air was going into the valve and when pressure was entering the pump. He explained that his practice would be to let the button go as soon as pressure was heard to enter the pump. He thought it would take one to 3 minutes to repressurise a pump by jagging. The witness, Mr Murray, was when he gave his evidence the equivalent of a Lead Operator but at the time of the accident he was an experienced production operator who had worked on Piper Alpha for 10 years. There were gauges available to the operator to indicate the extent to which a GOV was open and also the pressure in the pump. It was certainly the practice of Mr Grieve to consult these gauges as he jagged a pump. Mr Murray explained that in operating the latch button one would hold it pulled out until the gauge indicated that the valve was just open. He stated that it would perhaps take three of four jags to start the pump. He thought it would take 20 to 30 seconds to complete the jagging. This is of course a different timescale to that indicated by Mr Grieve. Since I have no reason to suppose that these witnesses were doing other that trying their best to give reliable evidence, it may be that the discrepancies between them is due to different jagging techniques or differences in their ability to judge such short time scales. One fact of possible significance is that Mr Murray’s jagging experience was confined to situations where it was possible to use the equalisation line during jagging. This possibility existed when the alternative pump was operating. The equalisation line can then be used to pump some condensate from the duty pump to the off-line pump. This is done into the discharge side of the pump to be re-pressurised and this to an extent moderates the effect of introducing the suction pressure. It follows that Mr Murray’ observation about jagging times may possibly not be based on experience relevant to the circumstances at the time of the accident. On the other hand the pursuers argued that the jagging Mr Murray carried out may be equivalent to a second stage jagging operation when the pump had been partly pressurised already and there is some attraction in that argument because with the equalisation line working Mr Murray would be trying to fill just part of the pump. Undoubtedly then it was within the experience of some operators that a jagging sequence should take about 20 to 30 seconds. Mr Henderson was another operator who gave evidence about jagging. He unlike Mr Murray never used the equalisation line during jagging. He indicated that in jagging one could at the time look at the gauges to see how pressure was building up. He claimed that you could actually see the valve moving from an indicator. However it is clear from his evidence that he was referring to the situation after a planned maintenance when the pump had actually been isolated by spading so that he was dealing with a situation where some attention had to be paid to the joints which had been spaded. Mr Henderson thought that it would take the GOV about 5 to 10 seconds to open and broadly this fits in with the evidence I referred to earlier. He claims that in a jagging operation one never allows the GOV to fully open. Mr Henderson thought that jagging if done carefully could take about 15 minutes but this is not very useful to me because he is talking about a pump that had been spaded so that it is necessary to check joints by radio messages as the jagging proceeds. On the other hand he thought that if flanges did not have to be checked the jagging would take about 4 or 5 minutes. He also thought that jagging should be done slowly and in controlled stages. Mr Henderson confirms the evidence of Mr Grieve that during jagging there would be a noise when the condensate was entering the pump. The witness Mr Lloyd thought that the opening time of the GOV was probably about 30 seconds but he was an electrical engineer and his evidence on this point was in any event somewhat tentative. On the other hand Dr Crofton also claimed to have spoken to the manufacturers and he opined that unless special measures were taken it would take about 30 seconds for the valve to open. This certainly seems to be consistent with Mr Lloyd. However Dr Crofton appeared to suggest that if a good air supply is available shorter opening times that 30 seconds could be achieved. It certainly seems that it would take a fairly short time for the GOV to open and this would range somewhere between 30 seconds and 8 seconds with the exact time probably short of the higher figure. The valve does not necessarily open in a linear progression because at the initial stages of opening there is inertia to be overcome. Of course when the valve begins to open condensate is going to pass the ball valve as soon as some aperture appears in it. If Mr Henderson is right in saying that the valve would not be allowed to open fully then each jagging operation would last for less than 30 seconds and maybe much less.
The defenders submitted that it was clear that a normal jagging operation tended to take minutes rather than seconds. That would certainly be true of the practice followed by many of the operators. This of course may be relevant to the suggestion that Mr Vernon may have jagged the pump without Mr Grieve having noticed. Also many of the operators watched the gauge as they jagged and this would make sense. I think the problem with Mr Richardson is that he assumed that the jagging operators opened the valve whereas they manipulated the button slowly so as not to open the valve completely. If the pump was filled in a small number of seconds it would be difficult to fill the pump slowly as was required by the principle of jagging.
The conclusion which Dr Richardson drew from his BLOWDOWN exercise was that re-pressurisation could be achieved very quickly once the valve orifice was available perhaps within a few seconds. He repeated his calculations by hand in Court. Indeed if Dr Richardson’s calculations are accurate the pump would be fully re-pressurised before the valve arrived at the fully open position. What the pursuers were anxious to establish from this material is that during jagging it would be possible to get a restricted leak which could set off one alarm and in another introduction of condensate get a much larger leak which would release enough material at the leak to set of a number of alarms. In deriving data about the characteristics of the flow which would have arrived at the pump Dr Richardson relied on Mr Clark’s Flow Chart. This in his view showed that the material that would arrive at the pump would be condensate. This would arrive at the GOV at a pressure of 670 psi and at a temperature of 59.9 degrees Fahrenheit. When PSV 504 was removed then inevitably some air would enter the pump. This would mix with any vapour which remained in the line and just how much air would enter would depend on how long it would take to remove the PSV and fit the blind flange. Moreover as the condensate first enters the pump it will experience a sudden pressure drop and flash. He thought that the flow reaching the GOV would unquestionably be a liquid. However when the valve is first jagged quite a lot of the liquid will flash to form gas-perhaps 50 to 60%. Not all the liquid will flash. This means that there would not be liquid all the way to the PSV. There would be an air plug and the gas that has flashed off the condensate. The second stage of the jagging would have more condensate in the pump because by this stage there would be less flashing. However it is only when the pump is close to being fully repressurised that the flashing will stop. As the gas will rise initially it will occupy the relief line at the point nearest to the blind flange. However as the pressure continues to rise and more liquid is entering the pump system the pressure on the gas in the relief line will increase and this will convert gas to condensate. Thus at the end of the pressurisation process the pump system will largely be full of condensate. With partial pressurisation there will be gas at the top of the relief line. Gas will be converted to condensate at about 43 bar. However since the air which has entered the top of the relief line will mix with some gas that element will not convert to condensate but will on the other hand compress. The result is that there will always be a small air plug at the top of the relief line. To calculate how long it would take to pressurise the system if the valve is assumed to open in an uninterrupted way involves a complicated calculation but with a simple calculation the time can be derived within say an approximation of 10 or 20%. In doing his complete calculation Dr Richardson applied the Bernoulli equation which seeks to relate pressure drop to the velocity of fluid through an orifice. His mode of calculation is set out in number 44/88 of process. Of course in this respect there is a divergence between Dr Richardson and Dr Crofton who unlike Dr Richardson considers that in the application of Bernoulli, pressure drop is an important factor. Of course there has to be some pressure drop because before pressurisation takes place the pressure in the pump is considerably less than that upstream. Dr Crofton’s reference to the complexity of pressure drop arises from his view that the internal obstacles and restrictions within the pump would affect pressure drop. In his equation Dr Richardson arrives at an average velocity for the flow of fluid through the GOV of 67 metres per second. In production number 44/89 of process Dr Richardson in a further calculation works out that the pump could be pressurised in 1.5 seconds. I find I have to regard this result as only giving the broadest of indications as to the time required to pressurise the pump since such a short time span would make jagging impracticable. However some account can be taken of the dead zone when no fluid at all flows so that on any view the time taken for the valve to overcome inertia and move past the dead zone would have to be added to Dr Richardson’ estimate of the time that pressurisation would take. Of course this factor would occur at each jagging stage. I think the most that can be taken from Dr Richardson’s calculation is that once an orifice appears in the GOV the pump will fill up very quickly. Dr Richardson assumes for the purposes of his calculation that the valve opens immediately and then continues to open in a linear manner. Moreover I cannot ignore the difference between the experts on the question of pressure drop. This is what may introduce a measure of distortion to his calculations but it is difficult to believe that this would invalidate the general picture they present. Dr Crofton’s view of pressure drop is not backed up by detailed calculations of his own but rather on calculations which Dr Richardson is applying to a different situation. Dr Crofton claims that through not taking sufficient account of pressure drop Dr Richardson arrives at a flow rate which is too high. Furthermore he claims that Dr Richardson’s calculations if proper account is taken of pressure drop would bring out a figure of 75 bar which with an upstream pressure flow of only 46 bar would be impossible. However this suggestion is based on assuming a mass flow rate of 130 kilograms per second rather than the mass flow rate of 3 kilograms per second which Dr Richardson’s use of the equation brings out for the rate of leakage at the assumed leak site. The 130 kilogram flow rate per second is based rather on the assumed flow rate at the GOV into the pump. What Dr Crofton has done is to assume that if Dr Richardson with a flow rate of 3 kilogram per second calculates a pressure difference of 0.04 bar then with a flow rate of 130 kilograms one would come up with a flow rate of 75.11 bar. In arithmetical terms this would be so. However Dr Richardson’s calculations are based on an assumption that the flow is through one suction valve and one discharge valve in one cylinder. This assumption may be adequate as a means of simplifying the reduced leak calculation which Dr Richardson was doing but may not be at all appropriate for the different exercise which Dr Crofton did. Dr Richardson is considering a depressurisation process whereas Dr Crofton is considering a re-pressurisation process. Dr Crofton is postulating an empty pump because it is only on that basis that there could be a flow rate of 130 kilograms. Dr Richardson’s calculations are based on a pump that has hydrocarbon within it. When Dr Crofton seeks to adjust Dr Richardson’s calculations by assuming that the flow is going through three cylinders but still through one suction valve and one discharge valve he brings out the much reduced figure for pressure drop of 8.33 bar. This is still a distortion of the actual situation where the flow is going through three suction valves in each cylinder. Indeed it is not at all clear that Dr Crofton was aware that one is not dealing with sets of three valves but in fact of nine. Allowing for that fact the pressure drop diminishes to less than one bar. This is because the differential pressure is proportional to the velocity squared. Dr Crofton expresses the view that the discharge valve would become the critical constriction point for the flow and therefore asserted that some allowance had to be made for the fact that the valve is larger than the suction valve in the ratio of 1.81 to 1. He recalculated on that basis that the pressure drop would be 22.8 bar. However that was still on the basis of Dr Richardson’s calculation based on just one of each valve, in a depressurisation situation, and for a flow rate much less than 130 kilograms per second. Even accepting his figures if adjustment is made for three cylinders the pressure drop would be reduced to something like 2.5 bar. Even assuming that 2.5 bar was a significant figure Dr Richardson was not asked if that figure would have made any radical difference to his calculations. It cannot be assumed that a pressure drop of 2.5 bar would be significant in relation to a pressure of 46 bar. Moreover BLOWDOWN proceeded on an assumed flow rate of 100 kilograms and not 130 kilograms so that the 2.5 pressure drop would have to be reduced further to take account of that. There is a degree of artificiality in Dr Crofton’s use of Dr Richardson’s calculations. He does express the generalised view that there would be pressure drop because of friction between the moving fluid and the walls of the valves. Of course the fact that there would be such friction to a degree can hardly be disputed. The question is how significant is this factor. Dr Richardson is also an experienced Fluid Engineer and it is his experience that in terms of achieving a practical result a detailed investigation of pressure drop would not in this case materially affect his results. If regard be had to the Report which is number 12/390 of process by Dr Richardson and Dr Saville it will be seen that at least in respect of depressurisation pressure drop is shown on calculation not to be an important matter and that is on the worst case scenario that the fluid is only passing through one valve rather than the three there actually were. Given that Dr Richardson’s calculations are being taken out of context by Dr Crofton and that otherwise there is really nothing conclusive with which to challenge Dr Richardson’s considered view on this question I am content to accept that once the jagging operation begins the pump would fill up in a matter of seconds after the orifice appears rather than some substantial period later. The result of the BLOWDOWN exercise bears out this position. It also has to be noted that Dr Crofton was not instructed until some considerable time after Dr Richardson had given his evidence so that the latter was not given an opportunity to comment on the approach which Dr Crofton ultimately adopted.
For clarification the velocity of a fluid is the speed at which it moves through the system whereas the mass flow rate is the amount of matter that can be transported over a defined time.
Dr Crofton also advanced a criticism of Dr Richardson’s calculation in relation to the latter’s assumption that the ball valve on the GOVs would open at a linear rate. Dr Richardson’s opinion on this was at least to a degree influenced by work he had done with 2-inch ball valve at his laboratory. He claimed that one material dimension is the diameter of the ball and the other the bore of the hole through he centre of the ball. Assuming the ratio of those two numbers is the same for a 2-inch and an 8-inch valve the results for one can be adapted to the other. Measurement of valves of different sizes have confirmed to Dr Richardson that the different sizes of valve operate on more or less the same ratios. In such valves the dead area (the portion before any aperture appears) is about 13 degrees of the turn. Dr Crofton thought that an opening time of about 8 to 10 seconds could only be achieved by what he termed "special measures" and this he based on information given to him by the manufacturers Messrs Bettas. However by "special measures" he meant a good instrument air supply. It should be noted that the matter of the times of 8 to 10 seconds was not taken up with Dr Richardson by the defenders although he was challenged on the question of linearity. Mr Wottge described the instrument air supply system on the platform and it was not suggested to him that this was other than adequate. However whatever the precise position on the platform the times Dr Richardson used seemed to fit in with his general understanding of an opening rate of about one inch per second. The question of the platform air supply first became an issue with the evidence of Dr Crofton and he had no direct experience or detailed knowledge of the supply arrangements on Piper Alpha. However knowing the outline of the air supply system he thought that there would be a degree of pressure drop in the air supply system that might haven fallen short of the arrangements needed to operate the ball valve to best effect. His point is that it is difficult to achieve the optimum pressure when the pneumatic equipment is situated some distance from the air supply. However he accepted that other factors such as the configuration of the pipework have to be taken into account. Indeed he thought that it may have taken as long as 30 seconds to open the GOV. Dr Crofton is an experienced mechanical engineer and in term of generalities I have no doubt that he is well familiar with the problems of achieving swift opening times for ball valves. The problem is that in the case of the GOVs we are concerned with no attempt was made to get the views of Mr Wottge. He may have been able to tell how, if at all, the efficiency of the air supply system was tested or monitored. Moreover the 100 psi of instrument air supply which was required for the actuator appeared to be the pressure which Mr Wottge took to have actually arrived at the site. There is a degree of support for the view that the valve took more than 10 seconds to open from the witness Mr Lloyd but he was an electrical engineer whereas Mr Henderson who was an experienced operator thought that the valve opened in 5 to 10 seconds. I think in the light of the total situation it is difficult to ignore the recollection of Mr Henderson who actually worked with these valves. Dr Crofton is speaking of generalities and had never himself actually been offshore on an oil platform. Indeed if the issue had been raised at an earlier stage it may well have been possible to settle the matter by measuring how long equivalent valves on the Claymore platform took to open. In many respects (though not all) the system on Claymore mirrored that on Piper. The precise time it would have taken the GOVs involved in this case to fully open will never be known but I am prepared to hold that they would not have taken significantly more than 10 seconds and certainly under 30 seconds. This overall view may be more important than a strict analysis of the linearity of the process. According to Dr Crofton’s calculation if the valve opening time is extended to 30 seconds then on the basis of Dr Richardson’s results the pressurisation of the machine can be done in about 3 seconds. In fact the pursuers accepted that it may well have taken 3 seconds of flow to repressurise the pump. The total capacity of the pump was about 200 kilograms of condensate. I think it was clear from the evidence that the pump could be repressurised in a very short time when dead time is excluded. Even Dr Crofton agrees that it would probably have taken about 3 seconds to repressurise the pump.
Dr Crofton’s view of the non-linearity of the opening time of the valve is dependent on the fact that initially before the valve begins to open the friction and inertial forces have to be overcome. Moreover in the case of a Scotch Yoke type of actuator there would be asymmetric torque characteristic. There is greater torque at the beginning and this levels out to be again followed by greater torque at the end of the process. Thus the valve will open more slowly at the beginning of the opening process. Dr Richardson did not claim that the opening times would be linear in any absolute sense but he stated that he had had conversations with the manufacturers and they had asserted that the opening could be regarded as linear for practical purposes. At the end of the day he has taken what is essentially a broad axe view of the matter. However even if the opening of the valve is less linear than Dr Richardson supposed we have total opening times from the operators and the effect on his results in relation to pressurisation times is only going to be a matter of seconds.
Dr Richardson accepts that as the condensate arrived at the suction orifice of the suction GOV it would have the diameter of the ball to travel before arriving at the discharge orifice and that this factor would cause some flashing of the condensate. Perhaps as much as 50% of the liquid would flash as gas. Dr Richardson accepted that during the flashing process the Bernoulli equation would not strictly be applicable but he considered that the time scale is so microscopic, a few milliseconds, that this would make no practical difference. The flashing liquid is then overcome by the passage of liquid through the valve as the bore fills. At the end of the day Dr Crofton did not give contrary evidence on this point.
Dr Richardson accepted that if there were pressure or temperature fluctuations upstream of the GOVs which reduced pressure below about 43 bar then this could affect his BLOWDOWN results. If the stream went below bubble point pressure then there would be a change not only in the pressure of the condensate but also a change in the composition of the flow in that gas would be produced. He had already accepted that the Bernoulli equation would not be appropriate to a two-phase flow. I may say that the experts never suggested that if the reciprocal compressors tripped this would diminish the pressure of the flow to the GOV. The flow of condensate into the JT flash drum came not only from the reciprocating compressors but from the condensate suction vessel and it was not suggested that the pressure from the drum would alter if compressors were unloading or recycling. Indeed because the flow from the suction vessel to the JT flash drum was regulated by differential pressure the system would seek to maintain this differential pressure. Moreover although there was no evidence of the specifics of low level pressure alarms affecting the JT flash drum because most other critical elements in the process system appears to have been protected by such alarms it may not be difficult to suppose that they would be set to respond to any serious change in pressure levels. This was not really explored. The final position therefore is that there was no evidence to show that such process changes as may have occurred upstream of the condensate injection pumps would have caused the flow to the GOVs to change significantly in pressure. Indeed the evidence of Mr Grieve was that liquid level in the JT flash drum was rising in the period prior to the explosion.
As Dr Richardson said, and I think it is obvious, if the first jagging step were for example to fill the pump with half its capacity, that is about 20 kilograms of condensate, then if it encounters a leak at the blind flange with the passage of time both the mass and pressure of the condensate will decrease. With this situation we do not have the pump filled with liquid condensate since gas will have flashed off within the pump. Dr Richardson was able to do calculations as to the escape of material through a loose blind flange given different dimensions for the assumed aperture caused by the leak. I did not understand these calculations to have been challenged (see 44/92 of process). If there was a mixture of gas and condensate in the system the gas would escape first. He calculates a leakage rate of 15 kilograms per minute given a orifice equivalent to 10 millimetres. Since there will be some air in the empty pump before repressurisation this will form a small plug at the end of the relief line adjacent to the blind flange. This would have to be pushed out of the leak orifice by the liquid before any liquid could escape. It can be noted that any hydrocarbon gas could diffuse through the air. Dr Richardson describes this air as a "small plug" because it will be compressed. However it would take about 17 seconds to discharge from the relief line through a 10 mm leak orifice. Moreover the original pressure on the leak would not be maintained as gas escapes. The result is that the escape will be almost independent of orifice size and will be more related to the length of the jagging step. He makes what he describes as a crude analysis of how the pressure would decay in 44/93 of process. Effectively what he calculates is that every 8 seconds the pressure in the system would be half. When about 2 bar is reached a condition equal to stability would be reached. This would happen after about 24 seconds. Dr Richardson explains that this type of calculation is simply the practical calculation an engineer will use to get a feel for a situation so that Dr Richardson’s calculation is essentially an illustrative exercise. The point is that the pressure and therefore the release drops all the time so that quite quickly one gets to a rate of release which would be in the region of 1.8 kilograms per minute. According to Dr Davies that class of release could generate a single gas alarm in Zone C3. Under the circumstances which the pursuers sought to urge upon me there would be a partial pressurisation about 3 minutes prior to the explosion and after about 24 seconds any surge of gas which escaped would be effectively exhausted. Under the calculations of Dr Davies in 68/1 starting with a 40 bar figure the total release after 36 seconds would be 6.3 kilograms but that of course is with the higher initial pressure. With Dr Richardson’s figures it would take at least 3 minutes to release 100 kilograms of material even not allowing for decay. What the pursuers claim to be clear is that with the range of calculations provided by Dr Davies the low rate of release that he postulates might go and only go to trigger a C3 alarm. Although with a release rate of 30 (or half rate of 15 kilograms) there is initially a sizeable leak this quickly degenerates to a leak of 1 to 4 kilograms per minute. Thus in total for an initial release rate of 30 kilograms one would expect a total release of about 6.33 grams for the first jag and for 15 kilograms 2 to 4 kilograms would be released. Of course the whole exercise is dependent on the leak orifice being 10 millimetres. The calculations are of necessity imprecise but given a hole size of 10 millimetres (itself at best an assumption) then all other factors being equal a small amount of gas could have been released by the first jagging process which would have been capable of setting off the C3 alarm. Whether the flow patterns of the gas could have produced such a result without setting off other alarms is of course a different question. There was evidence that if the bolts on the blind flange had not been properly tightened a leak could have occurred but the size of such a leak could not be quantified.
The pursuers maintained that after the first jagging operation subsequent jagging took place which generated sufficient leak material to account for the second flurry of alarms. This hypothesis supposes that after the first jagging episode the second or at least a subsequent jagging took place which lasted for longer and could have released a larger quantity of condensate. Of course in carrying out what was anticipated as the final jagging operation the operator may have expected to complete the repressurisation of the pump and therefore may have held the latch open for the time needed to achieve this. If material was escaping this period would have been rather longer than would otherwise have been the case. Dr Davies postulates that if there was a second stage release of about 140 to 160 kilograms per minute and this lasted for about 30 seconds then such a release could bring about the flurry of alarms which was noted. He takes the 30 second period as being a compromise between 10 seconds and a minute. If the last stage release did in fact last about 30 seconds then something like 70 kilograms of gas or condensate would have entered the pump and if a substantial proportion of this escaped there could be the gas cloud of about 40 to 50 kilograms which both Dr Mitchison and Mr Cubbage were agreed could account for the explosion which eventuated. I did not understand their evidence on the size of the requisite cloud to be challenged. Indeed Dr Davies also agreed that with the sort of release being considered a flammable cloud of about 40 to 60 kilograms would be generated. In 44/94 of process Dr Richardson carries out an exercise to demonstrate the escape that might be expected from the second stage release if there was a 10 millimetre aperture. He used a velocity of 134 metres per second rather than a mean velocity because at this stage he is not pressurising a closed-in system. Thus as the condensate enters the pump there is not the same degree of back pressure as there would be with a closed pump. That particular approach was not challenged. He takes the pressure entering the pump as 46 bar and allows for a discharge co-efficient of 0.58 to take account of the fact that there is a flashing flow immediately downstream of the valve orifice. This co-efficient was derived from his experimental work and was a refinement of his earlier BLOWDOWN work where he had used a co-efficient of 0.65. Dr Richardson accepted that none of this experimental was related to an irregular orifice such as would occur with an ill-fitting blind flange but he had done considerable work on a number of similar non-circular orifices. His results were invariably within the range of 0.56 to 0.6. He produces a mass flow rate of about 3 kilograms per second. Although his use of the particular discharge co-efficient was challenged in cross-examination there was no expert evidence to support the challenge. I can accept from Dr Richardson’s evidence that even independently of BLOWDOWN it would be possible to have leaks of varying size from different jagging operations just depending how they are carried out.
Dr Richardson did an exercise to demonstrate what would have happened if the electricity had been restored to the pump and thus the GOV had been held open. Since this did not happen any application of the exercise to the situation of repressurisation by manual means would require care. On the other hand Dr Richardson declares that using BLOWDOWN he found that the rate of flow of condensate through an orifice of 10 millimetres with an upstream pressure of 46.2 psia was 3 kilograms per second which of course is 180 kilograms per minute. With regard to that particular calculation it is not altogether made clear why the GOV being latched open should be different to a very full jagging operation when the valve was held open for a period of seconds. The defenders suggest that the jagging operation would mean that the pump was brought to full pressure and then after 6 seconds the flow rate would dramatically decline. This may happen if the GOV closes immediately upon re-pressurisation but the situation may be different if the GOV is held open by hand even after the pump is full.
The matter of the precise range of release that would have triggered the C3 alarm and that alone was considered by Dr Davies. The figures he considers are essentially imprecise because the results depend not only on the flow rate but on the length of the release (which of course is dependent on the length of each jagging operation) thus if there was a release rate of something like 180 kilograms per minute but for only a few seconds this might have generated a puff sufficient to set off a C3 alarm. The scenario which might have arisen is variable since the total picture will depend on factors such as the size of the leak aperture, the precise pressure in the pump and the length of the jagging operation. The defenders submitted that a further factor which has to be taken into account is whether the escaped gas is neutrally buoyant or a heavy gas. The pursuers were content to rely on their submission that the scientific evidence showed that it was possible that there had been a smallish leak bringing out only single alarm and then a larger leak which could have brought about a larger inflammable cloud and a series of alarms.
The defenders place considerable store on what they consider to be unresolvable inconsistencies between the evidence of Dr Davies and Dr Richardson. The former made certain assumptions about flow rates and on the basis of these he postulated what alarm patterns could be achieved. However it was said that on a careful scrutiny of Dr Richardson’s evidence the flow rates needed to sustain Dr Davies’ views could not be attained. To have an inventory satisfactory to cause the necessary flammable mass the problem is not so much getting hydrocarbon into the pump but rather in getting it out in the right quantity and timescale.
In relation to the initial stage of jagging I think that Dr Richardson’s work shows that there could not have been a sufficient size of puff release to trigger the C3 alarm unless the release had been of liquid condensate. On the other hand a low release sustained over a period could have triggered this alarm as Dr Davies accepts. It is easy to get confused about rates of release since the flammable mass is of course dependent on the relationship between rate of release and time. Moreover Dr Davies mentions that there could be a combination of puff release and low continuous release and the defenders do not really address this. The prospect of confusion is of course increased because in his 1989 study Dr Davies is concentrating on continuous release whereas in his more recent study he is exploring puff types of releases. It is clear from the conclusion to his 1993 Report that in his later study he is considering an alternative not a substitute approach.
At the end of the day the pursuers seemed to accept that if pressure was only raised to about 40 bar there would only be a gas release. This means that in relation to the first stage release a puff release would only fit the bill if re-pressurisation had been taken to the point where the pump was fully repressurised or nearly so.
The pursuers contended that if there was a liquid release then the defenders’ molecular weight argument would have no application. I think this must be right. It would follow that in that situation the material that escaped would not be neutrally buoyant initially but at least some would have the weight of propane. In any event the pursuers made the point that the unchallenged evidence in the case was that the gas associated with condensate was heavier than air. Indeed it is true that although Counsel for the defenders sought to analyse the evidence to suggest that all the gas that would flash from a condensate mixture as it entered the empty pump would be neutrally buoyant, in relation to the flashing process in the pump, the matter of what gases would emerge from the pump was not put specifically to the experts. By this I mean the pursuers’ experts, for the defenders did not attempt to lead their own expert evidence on the matter.
The pursuers in their reply rejected the defenders’ contention that Dr Davies had wholly rejected the possibility that a neutrally buoyant gas could not set off a C3 alarm without triggering a C2 detector. They referred me to Test Series No 49 in the Report 12/362 B in Table A1.5. In this test with a neutrally buoyant gas the only detectors which saw gas were those in C3. It must be mentioned of course that the concentrations were low because the test was concerned with a small release. The pursuers also submitted that Dr Davies had not concerned himself with the question of a neutrally buoyant gas in any detail because he was primarily concerned with exploring whether it was possible to generate a sufficient flammable mass and for that a heavy type gas would be required. The pursuers however do accept that to miss G101/1 the gas would have to be generally down pointing.
In contra-distinction to the argument of defenders’ Counsel Mr Wottge considered that the gas that flashed of condensate was heavier than air. The matter was of some importance for him because the detectors designed to detect a leak of condensate were those located at a low level. If Mr Wottge is in error then an important part of the detection system may have been misconceived. Mr Wottge’s evidence was of course relating to the general situation and not to gas entering an empty pump. Mr Wottge was not asked for his view on that nor cross-examined on his general opinion. Dr Mitcheson also declared that the gases associated with condensate are denser than air and would tend to fall to the ground. Dr Richardson has no difficulty with the view that the components of condensate together are heavier than air.
The pursuers also argue that the defenders’ analysis of the reaction of the condensate when it enters the pump is erroneous because it makes no allowance for the effect of temperature. There was no in depth exploration with the experts of the effect of temperature although Dr Richardson thought that the kind of variations in temperature that might have been expected would not have a particularly material effect on his results. The defenders based their submissions on extrapolations from a number of sources. They sought to set boundaries for the molecular weight of gas flashing of condensate. They began by taking the molecular weight from the gas in the JT Flashdrum which was taken to be 22.6. This is used as the lower boundary. However it has to be observed that no flashing takes place at the JT flashdrum. The formation of condensate takes place at the JT Valve and the flow into the Flashdrum is two phase. Vapour enters the JT Valve and what results in the formation of condensate is the reduction both of pressure and of temperature. Thus the stream being routed out of the Flashdrum, stream 211 is a vapour stream composed of the lighter gases. There is no doubt that the drop of temperature at the Valve contributes to the formation of condensate. The upper bound for the molecular weight of the gas flashing off condensate was taken from the flashing of gas from crude oil. However Dr Richardson makes it perfectly clear that oil is not a condensate at all. Crude oil contains liquid that has never been gas and thus differs from condensate. It may well be difficult to get a heavier gas like propane to flash from crude oil but the critical thing is that these hypotheses were not put to the experts. Thus in my view I am unable to conclude that the molecular weight from a condensate derived gas would fall somewhere between the boundaries which the defenders attempted to set. The defenders sought to extrapolate from Dr Richardson’s exercise (numbers 44/90-92 of process) that because he assumes that condensate has a density of 20 kilograms per cubic metre when exposed to pressure of 20 bar the substance will have a density of 1 kilogram per cubic metre when exposed to atmospheric pressure of 1 bar. However this hypothesis involves an assumption the correlation is linear and there was no evidence to that effect. The extrapolation that the defenders sought to draw is that since air has a density of 1.2 kilograms per cubic metre then the gas flashing from condensate must be lighter. I do not think this was proved.
The pursuers also contended in reply that in relation to the exercise carried out by the defenders applying flash fractions one cannot assume that although only 50% of the propane will flash all of the methane and ethane will flash. Certainly no one supported the view that all the lighter gases would flash. As I understood the evidence the process is differential. Some of the propane will flash, some of the ethane, and some of the methane although the lighter gases have a more pronounced tendency to flash. However before the gas was converted into condensate in the first place it must have contained all the components later found in the condensate.
The pursuers contended that the density of a substance is affected by temperature. They cited the case of air. Although this maintains a molecular weight of 29 warm air will rise whereas cold air will drop. If condensate enters the injection pump and then flashes both gas and liquid will get cooler. It should be noted that the evidence of Dr Richardson was that as soon as the liquid enters the pump perhaps as much as 60% of the liquid will flash to form gas. But not all the liquid will gasify because there will be a serious drop in temperature. He speaks of a balance between the liquid that flashes and the liquid that does not flash because of the cooling effect. What matters is that any gas that escapes will be cold gas perhaps having a temperature of minus 20 degrees or less and not "hot gas" as the defenders suggested. It would appear that in those circumstances that any heating up of the gas is only relative. Of course if there were no leak by the time the pump is up to full pressure all the gas has recondensed to liquid. A further point is that as the cold gas and liquid form in the pump they will extract warmth from the walls of the pumps and piping. This will cause some evaporation and thus the quantity of gas produced may be more that the flashing fraction. Flashing is largely dominated by pressure fluctuations whereas evaporation depends on temperature. Thus what actually happens within the pump depends not only on flashing fractions but on the delicate interaction between pressures and temperatures. The point the pursuers made however is that if the gas that flashes off is cold gas it will be heavier than air at ambient temperature irrespective of molecular weight.
It has to be observed that even when a pump is de-pressurised a small quantity of residual heavy gas remains in the pump and would be there when re-pressurisation begins. The pursuers also asked me to notice that when gas flashes of the condensate and later leaks from the pump if will not do so as separately stratified gases. It will be a mixture with the various components mixed together. Air itself is a mixture of nitrogen, oxygen and other gases. The pursuers also wanted me to conclude that as the gas escapes through the leak orifice there would be a Joule Thomson effect. Dr Richardson spoke to this and I think it would be a factor cooling the gas although Dr Richardson could not quantify the amount of cooling since this would require complex thermo-dynamic sums. The pursuers took issue with the defenders’ submission that there would be a significant air plug in the pump piping adjacent to the blind flange. It was however accepted that there would be some air. However the defenders had not taken account of the fact that when the flange came off during the removal of the PCV the contents of the pump would have been exposed to air for a short time and air could have mixed with residual hydrocarbon. The actual effect of the line being exposed to air is a difficult matter and would depend on the length of time the pipework was open and the relative densities of the air to the residual gases. As Dr Richardson said one difficulty is to know the temperature of the air. However he accepts that some air will have got into the pump. He also accepts that there will be a certain amount of mixing with other gases. Dr Richardson had said the removal of air in a pump full of air would take about 17 seconds in a leak but this was an upper bound and he said it was inconceivable that the whole pump would have been full of air. His calculation (done on BLOWDOWN) was essentially a worst case scenario and did not even take account of the actual geometry of the line. Instead it was assumed that the relief line was a straight line which in fact was far from the case. However the pursuers’ main point is that by the time of the stage of the leaking which caused the second flurry of alarms the whole air plug would have leaked out. I think on the balance of the evidence at least some of the air will have leaked out. The defenders argued that the air will not have been compressed until the liquid condensate fills the whole pump. I find this difficult to follow because if say half the pump is filled with liquid, and a substantial amount is occupied by gas that has flashed off, where is the air to go if it not to a degree compressed.
I think it can be concluded that after depressurisation had been completed the vent lines would not have been left open. Indeed we heard evidence that one of the purposes of attaching a blind flange was to prevent residual gases from escaping.
In relation to the second stage release I do not think it was established that it would have taken a material time to eject the airplug. If any time was required for this purpose it would have been substantially less that the 17 seconds referred to by the defenders.
The pursuers pointed out, correctly I think, that on the evidence even supposing that the booster pump had ceased to function there would be enough pressure to maintain a liquid flow to the pump provided that there was sufficient pressure in the JT Drum.
An important matter that the pursuers rely upon is the suggestion that with a jagging process of 10 to 15 seconds there could be a liquid release lasting about 25 to 30 seconds. This is because the pulsation dampeners would continue to push material forward even after the GOV had closed again.
Dr Richardson carried out two exercises in relation to leakage rates which he designated in his Report Case 1 and Case 2. Case 2 requires a constant pressure to be maintained for a finite time and the defenders sought to construe this as meaning that it could only apply if the pump was electrically de-isolated. However Dr Richardson is only interested in having the requisite relationship between time and pressure and it does not matter to him whether the GOV is kept open by being latched under power or simply held open for the necessary period. One must therefore examine whether the necessary flow can be allowed to enter the pump irrespective of the state of the electrical connection. The significance of having the valve open in Dr Richardson’s Case 2 calculation is that it enables him to ensure the necessary quantity of material entering the pump. What is important is to have more pressure coming in after you have reached what normally would be full pressure. Moreover as I have said the effect of the pulsation dampeners would have to be taken into account. These would maintain pressure after the operator has stopped manipulating the GOV button. I think the defenders are mistaken in thinking that the Case 1 hypothesis is the only one available to the pursuers.
The pursuers addressed the defenders’ submission that the presence of scaffolding rendered the tests of Dr Davies valueless. The defenders had argued, the pursuers claimed, that with a flashing release at the leak source there would not be a jet effect but a wide dispersion of the gas which flashed off. Thus it was argued that G101/1 would be an obvious candidate for the first alarm. Dr Richardson illustrates the character of a jet release of gas in the drawing number 44/103 of process. He is dealing with what he calls turbulent flashing releases but then he adds that most releases will be of this description. I note that one reservation he has is that there is no metalwork to block the path of the release. However what he shows is a cone shape with a semi-angle of about 7 degrees if a line is drawn down the middle of the one. If liquid leaks so that there is a non-flashing leak the jet is expanded significantly so that the semi-angle can be 180 degrees. However a little way downstream the jet resumes shape as a normal turbulent jet with a 7 degrees semi-angle. Dr Richardson does not explain why this happens but he relates it to experiments. The pursuers contended that contrary to the defenders’ submissions Dr Davies did consider the possibility of there being scaffolding present although the only scenario he specifically modelled was a block type structure directly below a vertical downwards jet. However he could not arrive at specific conclusions from this because he did not have the precise geometry of the scaffolding nor the piping and other equipment that surrounded the flange. His conclusion was that he could provide no precise answers for the geometry of the equipment surrounding the PSV because it was impossible to have precise details of this. Thus he concentrated on the general position. The general position was that a downwards pointing leak if conditions were otherwise favourable could have provided the alarm patterns preceding the explosion. The pursuers accepted that if the release had been directly downwards, if the plate had been shaped and placed equivalent to a scaffold, and if the surrounding equipment and pipes had not affected the position, then in accordance with series 27 the C3 would not be the first alarm to annunciate but this would depend on a number of dubious assumptions. Dr Davies himself said in evidence that he had not attempted to replicate the scaffolding in detail since there would have been no point in trying to do so. He lists the significant details that were unknown to him. In fact it is clear that he has kept an open mind in relation to the effect of scaffolding on his tests. As the pursuers pointed out it is perfectly possible that jets were coming out not in one but in various directions. The pursuers also pointed out that the flange was set in an east/west plane. We cannot know whether the leak jet was directed eastwards, westwards, downwards, or even upwards. In the foregoing circumstances the pursuers argued that it was unlikely that all the segments of any release would impinge upon the scaffolding the geometry of which is in any event unknown. As an illustration mention was made of a hosepipe playing on a motor car. The water would splash-off according to precisely what was happening.
The pursuers’ Senior Counsel accepted that with the evidence now being considered alone he had not positively proved that a leak from PSV 504 was the cause of the accident but what he claimed in relation to the relevant evidence, particularly that of Dr Richardson and Dr Davies, was that he had shown that it was quite feasible that a leak from the blind flange had caused the alarm pattern observed by Mr Bollands and the flammable mass need to account for the explosion. The defenders had suggested that the evidence of the pursuers’ experts showed that it was not possible for a leak from the blind flange to have caused the alarm pattern and explosion and I do not accept this contention. It would be a different matter if the contention was that the pursuers have not definitely proved from this expert evidence alone that the accident was caused by a leak from the blind flange. Of course the pursuers’ case does not depend on that expert evidence viewed as in a vacuum.
6.2.5 Dr Bakke
The witness Dr Bakke was the Norwegian explosives expert who gave evidence on the sort of explosion overpressure which a cloud of gas in the centrifugal compressor skid in Module C might have been expected to generate. He gave evidence based on general experience that given that the firewalls were not constructed as blast walls, the kind of overpressures which would have been generated by the explosion might have been expected to bring about the failure of these walls.
When he gave his evidence Dr Bakke was 39 years of age and he was a Staff Engineer with Statoil the Norwegian state oil company. He had joined them about 1993 and this was to strengthen their expertise within the field of gas dispersion, fire and explosion analysis. He was involved in development projects for new fields where he would for example look after safety. He did work on safety technology and investigated accidents. His work involved visits to oil platforms and indeed he was familiar with about 20 platforms. This involved looking at the explosion factor in about 50 modules. He was concerned to reduce the risk of serious explosion in these modules. He had graduated from the University of Bergen in 1979 with a B.Sc degree in physics and mathematics. In 1981 he was awarded a MSc in applied mathematics and in 1986 he obtained a PhD in a subject connected with gas explosions. He had worked as a Research assistant with the Christian Michelsen Institute in Bergen doing work on gas explosions. He had worked for some years doing experiments into the propagation of gas explosions. He was promoted to the post of Scientist and then to Senior Scientist. His Institute was very geared towards the industrial applications of his work. He carried out various gas explosion research programmes. He also had experience of consultancy work. In relation to his PhD he had been the author of a computer programme called the FLACS 2D. This was essentially a two dimensional approach so that further work was done to develop a three-dimensional Code and in this work Dr Bakke was a co-author. Much of his work was concerned with the prediction of explosion propagation and the resulting over-pressures. He had given evidence at Lord Cullen’s inquiry on two occasions. During his evidence I was referred to certain video material showing explosions in a number of simulated modules. The FLACs programme is what is called a validated programme in that its performance has been checked empirically. The defenders sought to persuade me that of all the pursuers’ experts Dr Bakke was the one who had the most practical experience of gas dynamics and there is little doubt that he was well qualified to express views on the matters he spoke to. The defenders sought to draw a distinction between the evidence of Dr Bakke and Dr Palmer who was an expert on structural mechanics rather than gas dynamics. This difference in specialist experience is certainly a factor to be taken into account although both men are experts of wide practical and theoretical engineering experience relevant to the problems at hand. Thus for example Dr Bakke has general experience of the effect of explosions on containing walls but he may have to defer to a structural engineer on a detailed analysis as to how a particular wall would be likely to break up. Thus to illustrate this although Dr Bakke was the pursuers’ expert they led Dr Palmer to speak in detail to the dynamic response of the firewall to the assumed explosion.
Dr Bakke was prepared to accept that understanding the overpressures that would be generated by an explosion is a complicated matter and quantitative views require careful calculation. He was further of the view that the volumetric size of the gas cloud could influence an explosion but that the relationship is not simply size because the geometry of the vessel will also determine what kind of explosion occurs. A smaller cloud in a congested area could generate the same overpressures as a larger cloud in a less congested area. The area of the centrifugal compressor skid in Module C was certainly to be regarded as a relatively congested area having as it did large machines and a considerable amount of pipework. The point of ignition also has an effect on the explosion. An ignition source at the edge of the gas cloud could create a different effect to an ignition source in the centre of the cloud. The strength of the ignition is also an important factor. He has seen situations where a change in ignition from a spark to a flame jet has multiplied pressures by 5. A stronger ignition source will tend to create a greater over-pressure. The shape of the cloud and its confinement are also important factors. A spark could be extended as an ignition source if for example it ignited gas within a motor cover and the ignited gas created a flame which was able to come out of the motor casing at high velocity. Venting through the walls of the enclosure is an additional factor that has to be considered. The effect of venting is not merely a matter of the size of the vent but is also influenced by the location of the vent. Lighter walls that fail more readily will of course produce better venting in the event of an explosion. Even such a wall is not likely to fail until there is an overpressure of about 0.2 bar. On the other hand a firewall is likelier to be of a heavier structure than a release wall because of its primary function and of course this raises the question of added inertia. Nor can it even be assumed that a lighter wall will provide relief from overpressure since the time factor enters into it. The wall will only provide venting if it opens up during the rise of the pressure pulse. Even a lightly bolted and less heavy structure may on a general overview have limited movement with the sort of pressure pulses being considered. Moreover the composition of a hydrocarbon cloud will affect the explosion which could be generated. Thus he considered that relatively heavy walls like the firewalls would not, if the explosive cloud was of the size being postulated, move with sufficient rapidity to create much of a venting effect. The defenders in this connection relied on the equation that force equals mass times acceleration. The point they made is that if force is constant the greater the mass the lower the acceleration. The logical consequence of this is that if you have a heavier panel and apply the same force the lighter panel will accelerate faster. Thus the defenders say that the pursuers cannot have it both ways. If the wall fragments so as to vent early, explosive force is lost. On the other hand if the wall does not vent then it is difficult to contend that a fragment would develop the necessary velocity to affect the condensate pipe. The earlier in the progress of the pressure pulse that the walls fragment the more residual force is available to impart velocity to them. The firewalls cannot move a material distance without venting. Even after the pressure pulse has peaked a fragment that has acquired freedom will not stop moving after that because before the peak it has acquired momentum. In order to get propulsion of a fragment one takes account of the movement of the wall as it distorts prior to break up. Then there is the propulsive effect of the pulse itself. The final factor is the propulsive effect of the gas stream as it passes through the vent. In my view it was clear from the evidence that the interaction between venting and propulsion of a fragment represents a delicate balance so that the defenders’ attempt to dichotomise the matter is perhaps rather too simplistic.
At various points in their submission the defenders claimed that, relative to Dr Palmer, Dr Bakke was to be given greater pre-eminence as an expert on gas dynamics because of his considerable specialist experience in that field. However it is important to note that Dr Palmer gave his evidence after Dr Bakke and it was not suggested to him that his evidence was contradicted by the latter.
It should also be noted that the heavier hydrocarbons require less volume. From his general experience Dr Bakke focuses on the area around the centrifugal compressor skid as being an area within Module C that would be likely to provide high order pressure. On the other hand if the cloud had been situated at the west end of Module C because this is much less congested he would have expected explosion over-pressures to be relatively much less. If a high velocity gas flow is not set up ahead of the flame there is not much turbulence and the combustion is not very fast. Moreover if the explosion was at the west end of C it could vent relatively freely out of the module. The worst case scenario for Module C would be to have the module full of a stoichiometric mixture of heavy gas. Then several bar over-pressures would be created even with ignition at the west end. Dr Bakke expresses the opinion that if the whole module has been full of gas the blue flame that Captain Clegg claimed to see would not have occurred but he would have seen substantial flame coming out the end of the module. The same would apply if Module B had been full of gas and the explosion had taken place there. Dr Mitcheson makes the same point. This evidence is in my view acceptable and is important for it may well run counter to any suggestion that a leak at the west end of Module B proceeded to the east end of the module and into Module C with sufficient concentration to set off alarms there. If there is one matter about which Captain Clegg is particularly likely to be correct it is that he saw a blue flame at the west side of the platform. The blue flame which emerged from either Module B or C is consistent with it being the tail end of an explosion at the east end of the module. Dr Bakke is asked about the effect of an explosion in Module C if the east end of the module in the general area of the compressors there had a gas cloud occupying about 10% of the module. His reply is very tentative since obviously the result will depend considerably on the geometric configuration of the congestion. On the basis of his experimental work he thinks that such a cloud could generate an overpressure of about 0. 2 to 0.7 bar although this is he stressed a rough estimate. The cloud he was considering would have a volume of about 500 cubic metres. Of course an actual cloud would be likely to be rather more or less than that volume so that this emphasises that only approximate guidance can be obtained from such estimates. I think in such matters Dr Bakke must be regarded as an important witness since more than any other witness he was involved in the precise quantification of explosive effects. His above views are not based entirely on his FLACS model but also on his experiments. Indeed he claims that such is his experience over many years that he can get a rough intuitive feel for what would be a correct result in the situations he is considering. He also told us that in Module C it would in certain circumstances be possible to have an explosion which would generate much larger over pressures than those which the pursuers seek to rely on but of course that would require a leak source quite different to that which is postulated for the PSV 504 site. The kind of explosion which a particular escape of hydrocarbon could generate would in his view be sensitive to a number of factors particularly turbulence and congestion. A stoichiometric mixture of hydrocarbon occupying about 10% of the Module C would contain about 40 kilograms of the substance. A stoichiometric mixture would contain about 4% hydrocarbon in air. To fill the whole module with a stoichiometric mixture would require about 400 kilograms of hydrocarbon. Dr Bakke hesitated to give a quantitative opinion based on experience alone but he did express the view declared to be approximate that a cloud at the east end of the module occupying rather more than 10% might be expected to generate an overpressure of about 0.6 to 0.7 millibars. At one point he suggests that 0.7 millibars might be an extreme of his range but he all along emphasises that these figures are a rough estimate. It seems to have been conceded by the defenders that if an overpressure of 0.2 bar was generated then it is possible that wall panels would be projected away from the firewall.
It should be noted that more is required to create a given stoichiometric mixture from an escape of oil than an escape of condensate since the latter has a higher flash factor.
Dr Bakke did express one opinion which I had little difficulty in finding highly persuasive. He declared that it is no simple task to estimate the consequences of a gas explosion.
Dr Mitcheson thought that as a simple rule of thumb cloud which occupies about one-tenth of the module would cause an overpressure of about one-tenth of the maximum exploding pressure. He did no precise calculation but on the basis of his experience he considered that such volume could generate an explosion force in excess of half a bar. He thought that overpressures of 0.7 of a bar were possible but perhaps unlikely because of venting. He knew from his experience what sort of effect such an over-pressure might be expected to produce and he thought that an over-pressure of 0.5 could have caused the damage assumed to have occurred on the platform. He resisted the suggestion made in cross-examination that his experience did not justify the expression of such a view. Not only had he kept abreast of the literature on experiments since the 1980s but he had been involved in giving advice to colleagues in his Loss Prevention Department about the overpressure which may be expected in offshore modules. He has also discussed the problems which arise with other experts. There was no expert evidence that the overpressures which might have been expected would have been substantially lower than those discussed by Dr Bakke and Dr Mitcheson.
In relation to ordinary firewalls (as distinct from lightly built blast relief walls) Dr Bakke was of the opinion that they would not affect the rise of the pressure pulse generally. However venting may occur and affect the situation after the pressure peak has passed. He thought that the duration of the pressure pulse is also dependent on the energy contained in the gas which burns. He considered that given a cloud of about 500 cubic metres the pressure pulse would last for about say 20 to 40 milliseconds, although in test site experiments he had seen pulses of 50 to 100 milliseconds. Dr Bakke said that with the small cloud being considered if the cloud was ignited at the east end of Module C because of the congestion at that end he would expect the flame to burn westwards but assuming that the cloud was stoichiometric he would not expect much flame at the west end. This I accept.
The results of Dr Bakke’s use of FLACS are to be found in his Report, number 14/46 of process. Dr Bakke explained how he had fed into that programme what he considered to be details of Module C and its equipment. To do this he had to use two other specialist programmes HIDDEN and CASD. However the particulars of the configurations within Module C which these programmes threw up were checked directly by Dr Bakke by measurement. Dr Bakke’s conclusion was that he would have expected FLACS to produce similar over-pressures to the Court Model of the Modules. Dr Bakke proceeds on the basis of certain assumptions in regard to the porosity (ability to let gas pass by venting) of the firewall but these levels were found not to arise at the phase of the explosion when pressure was being developed. He made the assumption that the maximum porosity of the B/C firewall would be 20% and of the C/D wall 40%. However his simulation suggested that these levels were not reached anyway and higher levels could have been used without affecting the results. Pages 33 and 34 of his report show the porosity situation thrown up by his simulation. It also has to be borne in mind that because different sections of the wall had different structures the porosity of the wall would have varied at different points. Dr Bakke assumed that the wall failed at 0.1 bar and this may be an important factor for the venting potential of the wall would depend on its break up point. He did not quarrel with that assumption. Dr Bakke claimed that it required a structural engineer to determine the break-up. He seemed to say that the peak of the pressure pulse would have passed before there had been enough movement of the wall to cause significant venting. On the other hand the arrival of venting might affect the point at which the pressure pulse peaked. However the overpressure developed would not be significantly affected by venting. The porosity time curves in Dr Bakke’s Report (number 14/46 of process) confirm his conclusions from his simulation that the overpressures generated would not be affected by the failure of the firewall and also that the wall would begin to fail during the rise of the pressure pulse. He also stated that he had considerable faith in all his results. With regard to the weight of the firewalls Dr Palmer calculated the B/C wall to weigh 32 kilograms per square metre and that of the C/D firewall to be 106 kilograms per square metre. However Dr Bakke assumed that both firewalls had a weight of 63 kilograms since any changes in weight in his view would tend to balance out. In his computer simulation he assumed that the fuel in the gas cloud was condensate and that it had a stoichiometric composition. He put a cloud representing a flammable mass of 46.1 kilograms into the east end of Module C and this would have occupied about one-twelfth of the volume of the module. The cloud was assumed to be stagnant. In fact if he had allowed for the actual air flow conditions he thought that higher over-pressures may have been brought out. He assumed a weak ignition source at the worst location (that is the one which would cause the highest over-pressure) so that some sort of balance was struck. He monitored the results at a number of pressure points along the firewalls as illustrated in his Report. He explains that because the simulation of the Module was probably showing less congestion than existed in reality the computer results would show pressures rather on the low side. Although the graphic results of the exercise were produced by the pursuers the actual binary output of the computer was not. One fact demonstrated by the exercise was the isobaric nature of the over-pressures generated on either side as the explosion proceeded along the firewalls. Dr Mitcheson had said that he would have expected this to be the case with a gradation of pressure from the source of the explosion. The over-pressures developed by the exercise rose to just 0.302 of a bar. Dr Palmer was eventually asked to do his firewall failure calculations in relation to pressure pulse results at Point 1 shown in the Report because this corresponded to where there was a condensate line at the south side of the B/C firewall. The peak pressure shown at Point 1 is just under .2 of a bar. However it must be remembered that Dr Bakke was carrying out a relatively conservative exercise. His exercise certainly shows that with the type of explosion he postulates one could find the kind of over-pressures spoken to by Dr Mitcheson and Mr Cubbage. Moreover he has assumed a cloud of the size and location consistent broadly with the evidence of Dr Davies and Dr Richardson.
In terms of Dr Bakke’s simulation parts of the B/C and C/D firewalls would fail upon the sorts of overpressures he has calculated. This is of course assuming the strength of the wall to be as calculated by Dr Palmer. The B/C firewall would fail at points close to the west end of Module C and the C/D firewall at points adjacent to the Control Room. However care had to be taken not to take Dr Bakke in isolation on this matter because his results depend on input derived from Dr Palmer’ calculations of the failure points of the firewalls which is approximately 0.1 bar. Equally Dr Palmer’s figures and results are predicated on Dr Bakke’s simulation. One point that Dr Bakke confirms is that with a stratified cloud and a leak source about the centre of Module C an explosion in such circumstances could account for a phenomenon such as the low blue flash at the west end of the module and in a general sense this would be consistent with what Captain Clegg claimed to see. The computer simulation did not work on the basis of a stratified cloud such as would have in reality existed if there had been a leak. Indeed this leads Dr Bakke to the conclusion that the gas cloud did not fill the module but rather was local to the east end of Module C.
This in turn fits in well with Dr Mitcheson’s analysis of the accident. He like Dr Bakke is assuming that the explosion over-pressures caused part of the B/C firewall to collapse. Although he had not quantified the matter he thought that the fragments of the fractured B/C firewall would have possessed a significant amount of energy and would thus have been capable of severing pipework in Module B. Thereafter a consequent release of flammable material would have caused the smoke and yellow flames which had been observed in B. The propagation of flame through the unburned gases will eventually cause a fuel lean blue flame at the west end of the module.
Dr Bakke affirms that the results of his computer simulation seem to be quite probable values for the scenario he is dealing with.
6.3. The Status of the Condensate Injection Pumps
6.3.1 Procedures Affecting Condensate Injection Pumps
The pursuers of course attribute the accident to a chain of events arising from procedures taken by their employees to restore the operation of a Condensate Injection Pump which had been withdrawn from service for Maintenance and to the effect this had on the relief piping with the PSV 504 removed. In particular they aver that at the time of the accident Condensate Injection Pump A had been taken out of service for maintenance and repair. At the beginning of the shift pump B had been in operation. Pump A had been electrically isolated and depressurised. The GOVs required for the operation of the pump had had their air supply disconnected. The defenders do not really dispute the averments regarding the status of the pumps.
The maintenance to be carried out on pump A was intended to be in the first instance a routine maintenance programme carried out at 24 months intervals. Evidence about the work the pursuers were intending to do on the pump was given by Mr Lynch the Lead Production Operator. I have no reason to doubt that much of Mr Lynch’s evidence was accurate on this aspect of the case although on other matters I was not convinced that he was wholly reliable on all questions of detail. On the day of the accident he had just completed an 8 day shift and in fact he had left the platform in the late morning of that day. Mr Lynch had come on duty early, about 4 am, on the said day because he wanted to allow his back-to back Lockwood to get some sleep. The day Lead Production Operator, Mr Flook, had come on duty about 6pm and for a period seems to have shared the responsibility with him. In any event Mr Lynch got a telephone message in the morning from Mr Curtis the Production Superintendent informing him that the Maintenance Lead Technician, Smith, wanted to do a 24 months planned maintenance on Condensate Injection Pump A. With production being in Phase 1 rather than Phase 2, condensate production was reduced and this was thought to be an opportunity to take one of the two Injection Pumps out of line. Mr Lynch instructed Mr Grant who was the Phase 1 Operator on duty to switch from Pump A to Pump B and give the latter an extended run to test that it was working properly and that it was safe to take A out of production for an extended period. Mr Lynch explains that once he and his colleagues were satisfied with the standby pump B his responsibility was to start the isolations which were necessary for a pump going for maintenance. The pump would require to be isolated electrically and depressurised to zero by removing the condensate filling it. Since the GOVs governing the pump were air operated the air supply had to be disconnected by unscrewing it. Thus they would be locked in the closed position. Mr Lynch said that the depressurisation process took a long time and in general this was not disputed. From the evidence it was clear that it could take up to 2 hours. The relief line was protected by a manual isolation valve (which was just upstream of the PSV 504 in C module and between it and the condensate collection vessel) which also had to be isolated by closing the manual valve. Mr Lynch then says that after the isolations had been completed and checked he would issue the permit as the Designated Authority. Indeed he explained that the spading of the pump (putting metal plugs in the line), which had to be carried out by the maintenance technicians, could not be begun until the permit had been issued. Moreover he explains that the issue of a permit would also permit instrument work to proceed. On the other hand he explains that if after isolations there was no manpower available to carry out the spading he would probably leave the permit on the Lead Operator’s desk until the performing authority was ready to start. Mr Lynch was asked what time would be required from the instruction of the isolation work until it was completed and he thought about 6 hours.
Mr Henderson also gave evidence about the process of de-isolation and he was generally a reliable witness. He confirmed that a hot-work permit would be needed for the planned maintenance. The performing authority would be the Lead Maintenance Technician because the job would be multi-disciplinary. Mr Henderson makes the interesting observation that if it was proposed to work on the maintenance work during the night the Lead Maintenance Technician would come into the Control Room at the shift change to have a live permit re-validated or in different circumstances he would have it suspended. Indeed he thought that the Lead Maintenance Technician would discuss the status of the job with the Lead Production Operator. In detailing the isolation procedures Mr Henderson does not differ materially from Mr Lynch. His view was that it would be appropriate to run Pump B for perhaps 3 to 4 hours before decommissioning Pump A. He indicated that one could tell from the suction and discharge pressure gauges if a pump was pressurised. He confirms that to disconnect the air supply to the GOVs or to reconnect it again would only take a couple of minutes. Indeed he thought that it would take about 15 minutes to complete the manual isolations. He considered that de-pressurisation could be completed in about 2 hours unless there were problems.
In the present case after Mr Lynch’s intervention the depressurisation of Pump A was carried out and completed. Mr Rankine confirms that Mr Grant did this work and tested it before handing over to Mr Rankine for the valve maintenance.
6.4. The Firewalls
A substantial part of this proof was occupied in considering the structure of the firewalls dividing the modules on the platform and the effect the explosion would have had on them. The pursuers’ case is of course that the disaster was as result of an escape of condensate in Module C which subsequently exploded and caused the ensuing tragedy. The escape of condensate was attributed to an ill-fitted blind flange on the piping leading to PSV 504 and the accumulation of gas was said accordingly to have been principally in the south east quadrant of Module C. There is no doubt that shortly after the first explosion a major fire developed at the west end of Module B. This is clearly shown in the photographs taken almost immediately after the explosion. Accordingly if the explosion occurred initially in Module C the pursuers have to explain how the worst effect in the early stages seems to have been in Module B. The pursuers seek to explain this by suggesting that the initial explosion destroyed the B/C firewall which divides Module C from Module B. Thus they seek to show that the limited inventory of condensate, which on the basis of their own hypothesis could possibly have escaped through the blind flange, could have accumulated to cause an explosion which could have resulted in the requisite damage to the B/C firewall. Thus they have to show that about 40 to 60 kilograms of hydrocarbon could have caused an explosion sufficient to explain the catastrophe which ensued. Moreover the question of fuel for the fire in Module B also has to be addressed. Module B had of course a considerable inventory of crude oil and indeed the MOLs were situated at the west end of the module. The pursuers’ main hypothesis however is that the main fuel for the development of the initial fire was an escape of condensate under pressure after the explosion. A 4-inch diameter pipeline carrying condensate was situated close to the north wall of Module B at the west end of the module. The pursuers’ case is that this pipeline was damaged so as to allow an escape of condensate after being struck by a fragment of the firewall when the initial explosion took place. It follows from the foregoing that the pursuers were concerned to prove that the condensate which they say escaped through the blind flange could have generated sufficient overpressure to shatter the B/C firewall and furthermore to result in a projectile with sufficient energy to break the condensate pipe. It was therefore central to the pursuers’ approach to the case to show the strength of the firewall structure and the explosive overpressure which would have caused it to break up. If to the layman this might appear to be a relatively modest scientific problem it proved to be anything but that and indeed a period of many months was spent exploring this question. Moreover much of the evidence was extremely technical - or at least so it seemed to me.
6.4.2 Dr Cox
Much of the protraction of this branch of the case may well have sprung from the difficulty the pursuers experienced with their own witness Dr Cox. The evidence of Dr Cox got off to a somewhat ominous start when through no apparent blame on his part he was unable to begin his evidence at the proper time and I required to adjourn the proof from Thursday 9 December 1993 until Tuesday 14 December 1993. Dr Cox accordingly began his evidence on day 88 of the proof on 14 December. However even on that day the pursuers had moved me to adjourn because one of their advisory experts was missing. Nevertheless I refused this motion because the Christmas vacation was imminent but in the event it made no difference. When Dr Cox began his evidence he seemed to be very well qualified as a engineer holding as he did a BSc in Mechanical Engineering, a first class MA in Engineering Science from Cambridge and a PhD in Aeronautical Engineering. After leaving his studies he worked for an aeronautical firm and specialised in the dynamic problems arising from the break-up of aircraft. He had considerable experience of computer programmes and it was clear that he was principally called by the pursuers to speak to his computer exercises into the effect of explosion on the structure of the platform firewalls. He had carried out research at Imperial College into the dynamic response of tall buildings to extreme atmospheric winds. In his work he acquired experience of mathematical and physical modelling associated with his interests. He had worked for 5 years in the Mathematical Department of Imperial College. He then worked with Consulting Engineering companies specialising in the assessment of industrial hazards and gas escapes from pipelines in particular. He had been involved extensively with the oil and gas industry. He was when he gave his evidence the Chief Executive of a consultancy company called Four Elements Limited. I was quite happy to conclude that he had a very wide experience of practical and theoretical engineering with particular reference to computer modelling. Indeed he had been President of the European section of the Society for Risk Analysis. He had given evidence at the Cullen Enquiry and must presumably have made a sufficient impact there to persuade the pursuers that he could be useful to them as a witness in the present cases. This is perhaps not surprising because he had been approached by OPCAL shortly after the disaster and had led a team of four experts investigating the accident. Reports he had prepared (numbers 13/76, 13/76 and 14/55 of process) were produced. In the preparation for these reports he had collaborated with a Dr Trbojevic. The Report 13/76 purported to be concerned with the investigation of blast resistance of firewalls and the Control Room Wall. The Report 14/55 is headed "Projectile Effects of Firewall Disintegration".
Dr Cox gave general evidence for 3 days and much of his evidence was extremely technical. There were a number of objections to aspects of this evidence but these were disposed of by me and as will be seen are no longer significant.
Dr Cox described the difference between a static and a dynamic analysis. A static analysis is concerned with a static load which is one which is constant and invariable with time. A steady pressure may be one way of describing it. A dynamic load on the other hand is one which varies with time. A transient explosive pressure would be an example. In respect of his dynamic analysis of the firewall structures he intended to use a finite element analysis and this would have involved taking account of the individual components of the structure. In fact he had divided for the purposes of such an exercise the B/C firewall into 1,000 different finite elements. The computer programme he claimed to have used is know as DYNA 3D (which is a finite element code) and this he claimed was well tried and recognised. Dr Cox thought that in theory a dynamic analysis could be done by hand but this would have involved such repetitious calculations as to be impracticable and in any event to render such an exercise tractable it would have been necessary to employ gross simplifications which has the effect of reducing accuracy. He was an experienced analyst and whatever the quality of his evidence of the contents of his particular analyses I have no reason not to accept his evidence on the foregoing general matters.
At the end of day 89 of the proof (Wednesday 15 December 1993) the defenders’ Counsel objected to the fact that the witness had not produced the details of the input to his computer exercise. He claimed that this problem would normally be dealt with by the lodging of the input files so that the defenders could examine and check what had been put into the programme. I reserved this objection. Later towards the end of day 90 of the proof (Tuesday 18 January 1994) it emerged from Dr Cox’s evidence that his computing exercise in relation to the production 61/6A involved input material which was not that which he had initially used but the same or similar material stored in an archival tape. When Dr Cox was asked what difference there was between the original material and what had been put onto the archival tape Counsel for the defenders objected on the basis that any evidence that Dr Cox could give on that matter was hearsay. He argued that the material from the archival tape should at least be lodged along with the original data. I continued the matter overnight. At the beginning of day 91 of the proof Counsel for the pursuers announced that he had agreed to make the archival material sought by the defenders available to them. He sought an adjournment until Friday 21 January and this was granted although it meant not only the waste of the Wednesday but that of the Thursday as well. In fact it was not possible to arrange the transfer of the material in the time I had allowed and the proof had to be further adjourned until Tuesday 25 January. At the commencement of the resumed proof Counsel for the defenders complained that the material which the pursuers had made available to them was not in fact the original material which had been promised. This resulted in lengthy argument which occupied the morning of day 92. Counsel for the defenders pointed out certain deficiencies in the material that had been presented to the defenders and sought a further adjournment to allow his experts an opportunity to examine this material. The pursuers intimated that they would not be pursuing the evidence in relation to the DYNA 3D runs which appear in the production number 13/76 of process and they did not oppose the suggested further adjournment. I adjourned the proof further until 27 January and awarded the defenders expenses against the pursuers on an Agent and Client basis from the 15 December 1993 until the date of the motion. I took the view that if the pursuers had prepared their documentation properly in the first place then the expense generated between the dates I have mentioned would not have been necessary. When the proof resumed on Thursday 27 January 1994 the defenders immediately addressed me on the documents which had been produced to them and made various complaints about them. After further lengthy debate the pursuers asked for a further adjournment until Tuesday 1 February so that they could consider their position further. The adjournment was allowed and expenses reserved. On the last-mentioned date the pursuers sought yet a further adjournment until Wednesday 2 February (day 95). This was to enable them to consider further doubts that the defenders had been expressing about the further material relating to Dr Cox’s evidence which the pursuers had produced. The proof resumed on the Wednesday and the pursuers immediately intimated that they did not want to proceed further with the evidence of Dr Cox and indeed would not be founding on this evidence. This put both parties in some difficulty in regard to the availability of further witnesses because it had been anticipated that Dr Cox would occupy some substantial further time. The proof accordingly required to be adjourned until Tuesday 7 February. In the circumstances I awarded expenses to the defenders from 13 December to and including 1 February (so far as not already dealt with) and that on an Agent to Client basis. I also allowed expenses on the same basis in respect of the various experts’ reports which the defenders had had prepared to deal with the evidence of Dr Cox. Such evidence as Dr Cox had given was merely laying the background for the substantial evidence which never materialised and other than the vacation the proof from 13 December 1993 until 1 February 1994 had been a complete and very expensive waste of time. The pursuers never gave an explanation as to why they were abandoning the evidence of Dr Cox.
The defenders made some reference to the evidence of Dr Cox as they were entitled to do. They referred to his view that if there is in fact a dynamic situation one cannot say that the wall would fail under a specific static pressure you can only say that the wall will fail under a specific pressure time history. In the present case the firewalls were not being exposed to a single pressure but different pressures over a period of time. The Code that Dr Cox had proposed to use for his computer exercise was based on the assumption that the effects on the firewall would be dynamic. It may be significant that Dr Cox does not refer to the firewall as being substantial but as being relatively speaking rather flimsy. Of course the wall was only there for fire protection and not for blast protection. Dr Cox said that he expected the firewall to behave in a non-linear manner. To make the calculations he was required to do by manual calculation it would be necessary if the problem was to be tractable to make what he describes as gross simplifications.
The pursuers sought leave to lodge further productions which related to a computer modelling exercise and the matter requires to be viewed against the history relating to Dr Cox which I have set out above. It has to be noted that the motion to which I am about to refer was made on 28 April 1994 (day 123) and that between Dr Cox’s appearance in court and this motion the pursuers had led Professor Fenner as a witness. Professor Fenner was an expert on structural analysis and moreover was expert on computational analysis.
On day 123 the pursuers sought leave to lodge a file relating to a Report by Ove Arup dated March 1984. This Report purported to be a finite element analysis of the B/C firewall under blast loading conditions. The pursuers made no secret of the fact that this report was intended to form the basis of evidence about a modelling study which had been instructed by the pursuers in order to provide an alternative to the evidence of Dr Cox which they had abandoned. This is borne out by the fact that the Report had only been prepared after the attempt to lead Dr Cox had failed. A copy of the Report had been sent to the defenders on 30 March. The pursuers proposed that a Dr Myles would be called as a witness to speak to this Report. The proposed production also contained substantial input material relating to the exercise said to be detailed in the Report. There was a considerable amount of further related material.
On the same day, that is day 123, the pursuers moved me to allow them to lodge material relating to a Report by Dr Bakke. In this case the Report itself was already in process and although the defenders opposed the motion I had no difficulty in deciding that the material should be lodged. In fact it was not suggested at a later stage that these productions had caused the defenders any difficulty.
However the defenders vigorously opposed the lodging of the Ove Arup documents. It was said that the Report related not only to the strength of the firewall but to the velocities of the fragments that would be generated when the firewall breaks up. The defenders maintained that they would be seriously prejudiced if the pursuers were allowed at the stage in question to set up the documents for a further computer exercise. The pursuers had already had some years to prepare their evidence for the case. They had wasted at least a month with the evidence of Dr Cox and had not given any satisfactory explanation as to why they had not proceeded with this evidence. The assumptions put to those who compiled the Ove Arup Report were not identical with those which had been put to Dr Cox for his report. Since the evidence of Dr Cox the pursuers had led a number of witnesses to fact who spoke to matters having a bearing on the factual assumptions now emerging in the Ove Arup Report. If the defenders had known of the proposed new Report they would have required to consider what further cross-examination of these witnesses was required in the light of that. For example witnesses had been called to give evidence about the construction and structure of the firewall. Professor Fenner had been called as a witness on 1 March 1994 and seeing that he was presented as a computational analyst the defenders might well have wanted to test the validity of the Ove Arup study with him. Professor Fenner had finished his evidence and it was now too late to seek his views. Moreover the defenders would have had the opportunity to exploit differences between the Ove Arup Report and Professor Fenner’s hand calculations which were directed at the capacity of the B/C firewall to bear a static load. It was the pursuers’ own decision to lead Professor Fenner after they had instructed the Ove Arup Report but before that Report was available. In fact it is clear that Professor Fenner was instructed to prepare the ground for a computational analysis and that may be the reason that there is a slightly uneasy relationship between his evidence and that of Dr Palmer. The defenders also intimated that if the new Report was allowed they would require an adjournment of some weeks so that their experts could consider it. I decided in the whole circumstances not to allow the Ove Arup Report to be lodged since in my view this could have been prejudicial to the defenders. I considered that I had already shown considerable latitude to the pursuers with regard to the evidence of Dr Cox. About a month of time had apparently been wasted over that matter and even with the protracted timescale of this case as it appeared at that stage this seemed a material loss of time. No satisfactory explanation had been offered for the difficulties experienced with the evidence of Dr Cox. If I allowed the Report to be received a further substantial adjournment would be required. In addition to this factor I could not be sure that the defenders had not suffered prejudice because of the course that the evidence had taken. Certain evidence about the firewall might well have been handled differently by the defenders and of course they could have explored the new material with Professor Fenner. I could not be sure that the witnesses could conveniently be recalled and of course this procedure would in itself have generated more delay. I also awarded to the defenders expenses against the pursuers resulting from the expense rendered futile by the pursuers’ conduct.
As it happens my decision on the Ove Arup report may not have been as fair to the defenders as I intended. At no stage during the debate on the pursuers’ motion was it indicated to me that in the absence of the Report it would be necessary for the pursuers to present a structural analysis based on hand calculations. This is in fact what happened and the matter gave rise to many months of evidence, a large amount of somewhat arid mathematical calculation, and much technical debate in the submissions. However this could not be predicted by me when I made my decision on the Ove Arup Report. I have set out the history of Dr Cox’s evidence at length because it clearly had an influence on the profile of the proof.
It should be noted that although the pursuers discounted the value of the evidence of Dr Cox he was an experienced engineer and indeed as they were entitled to do the defenders sought to rely on him. The defenders had a copy of Dr Cox’ report available to them. Despite the pursuers’ approach they would have been entitled to cross-examine him. We do not know the results of Dr Cox’ investigations. However the defenders did not suggest to him that at any stage he had found anything antagonistic to the pursuers’ hypothesis.
6.4.5 Break up of the B/C Firewall
The pursuers contended that they had established that the force that could have been generated from a leak of gas from PSV 504 would have been sufficient to cause a breach of the B/C firewall (and indeed the C/D firewall too). They also maintained that they had proved that such an explosion would have created sufficient energy to produce projectiles which would have ruptured the 4-inch condensate pipe in Module B. The main witnesses on the structural mechanics side of the problem were Professor Fenner and Dr Palmer for the pursuers and Professor Reid and Professor Stolllery for the defenders. The pursuers’ argument was that Professor Reid was seeking a perfect research type accuracy that was not practicable even with a computer exercise. On the other hand it was contended that the evidence of Professor Fenner and Dr Palmer even if based on practical approximations was sufficient to show that the pursuers’ analysis of the accident was the most probable explanation. It has to be pointed out that the defenders’ approach was to attempt to destroy the validity of the pursuers’ exercise but at no time did they offer, either on the basis of a computer exercise or otherwise, to disprove the pursuers’ thesis. The pursuers for their part accept that in order to make hand calculations manageable it is necessary to make certain simplifications and idealisations which make a degree of approximation inevitable. The main difference between the parties is that whereas the pursuers claimed that even allowing for approximation and simplification the manually calculated exercises their experts carried out can provide valuable guidance, the defenders for their part submitted that the simplifications and compromises inherent in hand calculations are so radical as to render any conclusions derived therefrom worthless.
Professor Fenner began the analysis by doing a static analysis of the firewall B/C. This he did by idealising part of the firewall as a beam. This of course has to be recognised as a limited approach because from it one could derive nothing other than a static failure pressure. He also did work on the comparative strengths of various components of the firewall. Professor Fenner and Dr Palmer were at one in believing that the firewall would break up by a separation of the panel units one from the other. The panels consisted of Durasteel units within frames composed of angle frames. The panels would separate by the failure of the frame bolts holding the panel frames together. Professor Reid on the other hand declared that he could not say if the frame bolts would be the first component to go without carrying out a finite element analysis. He was not even sure that such an exercise would give a satisfactory result.
With regard to the pursuers’ case on the matter of the firewalls there is little doubt that they proffered Dr Palmer as their main witness (given the situation that had arisen in relation to Dr Cox and Mr Myles). He was 56 years old when he gave his evidence. He was also at that time Technical Director of SAIC Science and Engineering Ltd who were engaged in the offshore engineering business in connection with pipelines. He was Visiting Professor in the Principles of Engineering Design at Cambridge University. He had graduated MA at Cambridge in 1961 and later obtained a PhD at Brown University in the USA. He was a Fellow of the Royal Academy of Engineering and a Fellow of the Institute of Civil Engineers and Chartered Engineers. He was a Fellow of the Royal Society. After he graduated he had for two years been a lecturer at Liverpool University in the Department of Mechanical Engineering and then a lecturer and Senior Research assistant at Cambridge. He was thereafter Head of Pipeline Design Group with R J Brown Associates an engineering group in the Netherlands. In 1979 he was appointed Professor of Civil Engineering at UMIST University at Manchester. In 1982 he returned to RJ Brown as Vice President of Engineering. It can fairly be said that Dr Palmer has a good mix of academic and practical experience acquired while working at a high level. He has published a great amount of material including a book "Structural Mechanics". He has given evidence at various important Public Inquiries including the Cullen Inquiry.
Since 1991 Professor Fenner has been the Professor of Computational Engineering at Imperial College, London University. He has worked extensively in boundary integral equations as distinct from finite element equations. He has a BSc in Mechanical Engineering. Before obtaining his chair he was a Research Assistant and then a Reader at Imperial College. He is a Fellow of the Institute of Mechanical Engineers and a Fellow of the Institute of Materials. He has published 7 books and over a 100 papers some dealing with stress analysis. He appeared to be familiar with the development of computer analysis procedures to solve structural problems. He was instructed in early February 1994 just before Dr Cox disappeared from these cases. His remit was to assess the strengths of the various firewalls involved. He was also asked to estimate the absolute strengths of the firewalls. However the figures he produced as he says pre-suppose that the walls were only subject to a static overpressure.
Professor Reid, was 49 years of age when he gave his evidence and at the time held the Conoco Chair of Mechanical Engineering at the University of Manchester Institute of Science and Technology, a chair which he has held since 1985. He holds the degrees of BSc and PhD both in Mathematics. He had been the author of many papers in relation to engineering research. He is a Fellow of the Institute of Mathematics and a Fellow of the American Society of Mechanical Engineers. He is a Chartered Engineer and a Member of the Institute of Structural Engineers. He is a Fellow of the Royal Society of Arts and also a Fellow of the Institute of Mechanical Engineers. He is a Fellow of the Royal Society of Engineers and held the editorship of a number of engineering journals. The pursuers claimed that Dr Palmer had an advantage over Professor Reid in having had more direct practical experience.
There seemed to me to be little doubt that all the engineers I have mentioned in connection with the problems now being discussed were very able men highly respected in their respective spheres. This does not make my task any the easier since they did not always agree. If there was any discernible difference between Dr Palmer and Professor Reid it was that the former tended to take a more pragmatic approach whereas the latter had adopted an iconoclast role. He criticised the technical accuracy of a number of approaches adopted by Dr Palmer without carrying out alternative exercises of his own. This was no doubt in accordance with his instructions. One further problem was that Professor Reid was instructed relatively late on in the case so that some of his criticisms of Dr Palmer’s methodology were not put to Dr Palmer when he gave his evidence.
Like Professor Fenner, Dr Palmer carried out a static analysis of the firewall structure. This he did by hand calculation. He used a different approach to Professor Fenner and got a somewhat different answer. He extended his static results to arrive at a dynamic analysis. Both Professor Fenner and Dr Palmer worked on the assumption that the firewall would fail in bending. Unfortunately Professor Reid’s view that membrane forces would be relevant was not put to Professor Fenner.
6.4.6 Dr Palmer’s Static Analysis
In carrying out his various analyses Dr Palmer conceded that he was restricted by not being able to utilise finite element analysis. Because of this he had to use certain broad idealisations and approximations to render his hand calculations reasonably manageable. He accepted that in certain areas he may have been able to obtain more precise results by a finite elements analysis. However it should be noted that even a finite element analysis would have posed difficult questions and that its use would have required fine judgments. It certainly cannot be assumed that a finite element analysis of the firewall problem would have removed all points of contention between the parties. The defenders’ witness Professor Reid doubted if any useful results could be obtained from a hand analysis and he even doubted if a finite element analysis would be reliable. He thought that computer exercises were largely devised to predict collapse in the sense of yield rather than break-up.
A static analysis assumes no impact on the firewall and no inertia. In carrying out this analysis Dr Palmer analyses the wall as a series of triangular plates. He reckons that he could break the wall down into triangular plates because of the way the wall is restrained and its symmetry. The particular triangle that he analysed is shown in number 101 of process. This he regards as representative. He also (in triangle 2) looked at some of the larger triangles to the right of the firewall. His purpose is to regard a triangular segment as a plate and to apply plate theory. He applied what is known as the yield line method. This involves calculating the work done by a load on a mechanism as it deforms. The ultimate question is how much energy is needed in the structure to cause its collapse although Dr Palmer then seeks to extend collapse to failure. To effect his calculation he examined the local strength of the plate segment that he took from the firewall. In this regard he took as the governing structural parameter what he called the moment stress resultant and this he showed as little m in all his calculations. The moment stress resultant is the bending moment per unit length across any line on the plate. This value was calculated for five separate modes of failure. Modes 1 and 2 were for the failure of the frame bolts depending on whether the angle irons were prised apart at their toes or heels. The failure of the composite sheet regarded in isolation was another mode. The failure of the angle irons was yet another mode. The failure of the composite sheet of Durasteel material in combination with the angle irons was another. These sheets consisted of two skins with a composite material in the middle. Dr Palmer concluded that the two modes involving the frame bolts were the governing modes because they were involving the weakest components. In fact of these two modes he concluded that Mode 1 was the critical mode. This Mode involved the failure of the bolts as a result of the prying action around the toes of the angle irons. This question caused a degree of difference between Dr Palmer and Professor Reid since initially at least the latter claimed that he could not see a mechanism which involved Mode 1 or Mode 3 alone. Dr Palmer thought that the structure of the wall would provide mechanisms for the relevant modes. Next Dr Palmer attempted to relate the values he had for the relevant moment stress resultant to the load on his triangular segment in order to determine if the strength of the plate would be exceeded locally. Thus you arrive at the force that the frame bolts would take before they began to fail. In order to arrive at his bending moments Dr Palmer relied upon methodology set out in a paper by a Dr Mansfield and also he used a dimensional analysis equation. This approach was disputed by the defenders. The third step taken by Dr Palmer was to bring his first and second steps together to get a static failure pressure. This helps him to estimate the maximum deflection in the plate at the point when it breaks up. This is in order to relate this factor to the dynamic situation. His vision of how the plate would break up is that it would break up along the vertical and horizontal lines of the frame bolts.
Dr Palmer thereafter moved on to his dynamic analysis to consider the response of the firewall when the overpressure caused by the explosion was caused by fast loading. These dynamic considerations are important unless the loading time of the overpressure was long compared to the natural frequency of the system. He therefore began his analysis by determining the natural frequency of the triangular segment of the firewall that he was using. In order to do this he used a method known as Rayleigh’s method. This depended on some rather complicated mathematical calculation but I did not understand it to be seriously challenged at the end of the day. Having calculated the natural frequency Dr Palmer took the deflection at the centre of his idealised segment and applied a calculated multiplier to it in order to get a time element. In this regard he required to use the pressure pulse observed in the tests of the pursuers’ witness Dr Bakke. Once Dr Palmer had a time for the maximum deflection of the wall before break up he could from Dr Bakke’s pressure pulse work out the pressure at which the wall would break up. Looking to the time of about 42 milliseconds he could calculate by applying this time to the idealised graph representing Dr Bakke’s pressure pulse that the wall would fail at an overpressure of about .1 of a bar. Once this figure is known Dr Palmer reckoned that by having regard to how much of the pressure pulse is left after the wall fails he can calculate how much of the rise and fall of the pressure pulse is left to provide kinetic energy for projectiles. Dr Palmer then attempts to relate the velocities and kinetic energy available after the break up of the wall to the energy which would be required to rupture the 4-inch condensate line.
One point of coincidence between the evidence of Professor Fenner and Dr Palmer is that they both considered that the firewall would break up by fracture of the frame bolts and also that the frame bolts along the horizontal lines would go first. This differs from Professor Reid who stated that in a dynamic situation he could not tell if the frame bolts would go first or not.
There were criticisms made of some of the idealisations made by Dr Palmer. No one suggested that even in a finite element analysis some degree of idealisation would not be necessary but the exigencies of hand calculation make the use of such all the more essential if a result is to be obtained. Certainly it is clear that engineers will in practice use certain idealisations in their calculations so that at best their results will be approximate. The defenders argued that it is not good enough to accept that idealisations have been used to achieve an approximate result if the tolerances within which the result may be regarded are not established. At one point in his evidence Professor Reid seemed to disclaim having had any experience of analysing engineering structures such as firewalls. He also claimed that he would not have the competence to analyse a complex structure such as the Forth Road Bridge although it is obvious that the bridge was designed and built without recourse to computers. On the other hand practical engineers are often calculating safety tolerances rather than precise failure points. Professor Reid accepts that as a practical engineer one would often begin to analyse a situation by a simple and approximate method and then work up to a more sophisticated methodology. I think to be fair to Professor Reid that although he accepts he could have tried to do an approximate dynamic analysis using methods similar to those used by Dr Palmer he may well have discovered that the only satisfactory method required the use of a computer. Indeed the point must be stressed that Professor Reid thought that to do a satisfactory analysis of the firewall was an extremely difficult, if not impossible, matter even using a finite element code. I think what he meant was that there would be serious problems if a precise result was required. He accepts that in any analysis of any utility it would be necessary to employ idealisations and simplifications because not every feature of the real firewall can be accurately represented. He accepts that the matter of which idealisations are tolerable would be a question of engineering judgment.
In carrying out his analysis Dr Palmer took the view that it was necessary to consider both the static and dynamic response of the firewall but that the proper approach would be to consider the static response first. He began to elaborate on his approach by pointing out that the firewall is supported at the top and the bottom and also has lateral support. Since it transmits loads in two directions the firewall can be treated as a plate and in this respect Dr Palmer diverged from Professor Fenner who had all along confined himself to a one dimensional beam analysis. One reason why Dr Palmer carried out his whole exercise as a plate analysis was to have consistency when he conducted his dynamic analysis with the plate method. Pursuers’ Counsel only relied on Professor Fenner in relation to the transmission of loads and the comparative strengths of components but in terms of the analysis of the wall he founded his case on Dr Palmer. However this was not to the liking of the defenders who pointed out that there were inconsistencies between these two experts and claimed that the pursuers were not entitled simply to ignore one of the experts. However it is not clear that the experts are in fact inconsistent. Professor Fenner accepted that by using the beam analysis method he could not take account of the lateral support from the vertical edges of adjoining panels. In re-examination he accepted that if he had taken account of such lateral support this would have increased the failure pressure that he calculated. It needs to be remembered that Professor Fenner was preparing his analysis for use, as was then anticipated, with a further computational analysis. Whether or not such a computational analysis could have made adjustment for any deficiencies in a static beam analysis was not explored. Dr Palmer’s method of calculation avoided the particular difficulty that Professor Fenner referred to. The pursuers submitted that the main importance of Professor Fenner’s evidence was not to arrive at absolute results but rather to draw a comparison between the strengths of the various components of the firewall. Indeed the defenders did not challenge Dr Palmer’s conclusions on the basis that Professor Fenner had produced different results. Dr Palmer said that the existence of intermediate supports justified him in regarding the firewall as a series of segments. Because of the clamps and trusses there are supports along triangular sections of the wall and these sections appear between diagonal and vertical trusses. Each of his triangular sections he found to be approximately of the same height, breadth, and shape. There is continuity between the triangular segments because the Durasteel panels and angle frames cross the triangular sections. Each section he considered to be representative of the wall as a whole. The first thing to be considered is the local strength of the plate.
6.4.7 Assumptions on the Modes of Failure
The pursuers’ experts worked on the assumption spoken to by the experts on explosions that the resultant overpressure on the wall would be isobaric. The wall it was said reacts to this pressure by bending. The question of the transmission of the pressure received by the firewall to the different components of the wall has to be considered. Professor Fenner thought that the Durasteel panels would see the pressure first and then transmit it to the angle frames. It must be noted that shear stress causes about half the value of tensile strength so with such stress it is necessary to double the value to get a comparison of direct effective stress. It was accepted by all the experts that the welds at the top and bottom of the firewall before they would fail would take greater pressures than other components. Professor Fenner prepared a comparative table (44/39 of process ) in which he sets out the comparative strengths of the different components of the firewall. The calculations which went into that table led him to the view that the frame bolts were the weakest points of the structure. He had applied in his work a pressure of 0.1 of a bar to the various components to discover the comparative stresses that would be seen by the individual components of the firewall and in particular the horizontal frame bolts were the weakest connection. The weight and distribution of the force on components in the wall will of course depend on which side of the wall receives the pressure after the explosion. According to Professor Fenner’s calculations the pressures on the frame bolts will be weaker under a north to south loading than would be the case with a south to north loading. He gave the general opinion that if the wall saw north to south loading then the mode of failure would in the first place be the failure of the frame bolts in the horizontal connection. He was assuming that all the components would be made of a uniform grade of mild steel. This opinion was not directly challenged in cross-examination. Professor Fenner considered that the load transmitted to the bolts on the clamps holding the firewall to the trusses would be taken in tension. He assumed that these bolts under a north to south pressure would carry half the load on that piece of wall and Professor Reid accepted that assumption. He calculated the stress that would be placed on the clamp bolts as being 550 megapascals and if it happened that one of the bolts was less tightened than the other then it could be that one of the two clamp bolts would carry all the load in which case the load would be about 1100 megapascals. The point he makes is that the clamp bolts would not go before the frame bolts. It should be noted that Professor Reid would not accept that the frame bolts would necessarily go first because of the question of the redistribution of the load through the wall. Although Professor Fenner arrived at comparative figures for the strengths of the components of the wall by a one-dimensional beam static analysis he declared that he would not have expected a dynamic analysis to give a different result. His approach was essentially a simplification of the problems he was considering in order to get a manageable method and what was hoped would be an approximately correct result. It should also be noted that he declares that he has had no experience in the analysis of explosive overpressures. He also said that he personally was not equipped to make a dynamic analysis of a structure by hand calculation. It is clear I think that because Professor Fenner held the chair of Computational Engineering at Imperial College the emphasis in his work and experience was computational calculation.
Dr Palmer took mode 1 of failure (the prising apart of the iron angles at the toes) as being the critical failure mode in relation to the particular structure of the firewall. The mode would produce a prising action and generate rotation about the angle toes. The prying apart of the angle irons would lead to tension on the bolts. This would enable him to work out when the local component would fail. By referring to British Standards in relation to the strength of the steel bolts he calculates that if the bolts could be stretched more than 14% of their length they would break (this conclusion is based on the ultimate tensile strength of the bolts). The effect of the mode 1 mechanism is not a matter that Dr Palmer regards as an idealisation but rather as something that would happen in relation to an explosion. Indeed he concludes that it is the critical mechanism. In his second step in order to arrive at the bending moment stress resultant for the mode he uses a formulation from a scientific Paper by a Dr Mansfield. This was in fact a dimensional analysis equation. The equation involves the introduction of a constant value k and this value is taken from Dr Mansfield’s Paper. The third step is to bring his two figures together and see if the stress which the plate can bear will locally be exceeded. Thus he will seek to ascertain how much pressure can be applied to the firewall segment before it breaks up or deforms severely. The pursuers submitted that at the end of the day Professor Reid accepted that Dr Palmer’s use of the Mansfield formula was a reasonable thing to do. Dr Palmer indicates that in the case of a plate unlike a beam, it cannot be idealised in only one direction so that to understand how force is being transmitted it is necessary to look for ‘moment stress resultant’, which is the measurement of how much moment is transmitted along a line within the plate. The bending moment is not necessarily uniformly distributed across the line. In his calculations Dr Palmer used the symbols m times ds to denote the moment stress resultant per unit length (m being the notation for the moment stress resultant and ds being the length of the line being considered) and he uses this approach to work out how much force can be transmitted across a local part of the plate to determine the local strength of the plate. For example one would want to be able to calculate how much moment stress resultant will be applied to a frame bolt. It will be seen that Dr Palmer’s approach differed from that of Professor Fenner since the latter simply worked out the pressure along his beam at which the frame bolt would reach its ultimate tensile strength and fail. On the other hand Dr Palmer had to seek to counter Professor Reid’s criticism that in a dynamic approach it would not necessarily be the weakest component which would fail first. This would depend on how the load was transmitted across the plate. A component of the wall will fail when the force along a line directed towards that component reaches the failure point of that component. Moreover Dr Palmer ultimately reaches the conclusion that the frame bolts would reach their ultimate tensile strength before other components reach their yield point. Dr Palmer took his material for ultimate tensile strengths from Professor Fenner’s calculations indeed the various modes he had regard to are set out and described in number 44/156 of process. As I have indicated mode 1 is the rotation of the angle irons about the toes putting the frame bolts into tension. He calculated the failure value of the frame bolts with mode 1 operating as being 829 Newtons. He used Professor Fenner’s figures for the ultimate tensile strength of mild steel (which he only claimed to be approximately accurate) and the figures given by the witness Mr Cole for the spacing between bolts. Thus he used the distance of 20 millimetres as the distance from the toe of the angle iron to the centre of the bolt. He would expect the mode 1 action to operate first at the point where angle frames are joined about 8 feet above the floor and then the next response about 5 feet above that. Eventually he would expect to see mode 1 response in all the triangular sections of the firewall.
The second mode that Dr Palmer considered was bending within a composite panel (mode 2). He calculated the strength of a Durasteel panel (allowing for the structure of such a panel and the holes that were designed into it). The value he then calculated for the force required to make the panel fail was 945 Newtons. However he considered that these panels would not act on their own because they were bounded by angle irons.
He regarded mode 3 as being similar to mode 1 but differing in that the angle irons would rotate about the heel rather than the toe. He thought that the assumption of an overpressure from Module C there would be bending corresponding to mode 3 at the top and bottom of the firewall.
Mode 4 is not in Dr Palmer’s view critical because it involves the angles comprising the frames themselves to bend about lines parallel to their axis. This is described in number 44/156 0f process and Dr Palmer calculates it to be the second strongest of the five modes looked at. In fact the other modes would cause failure before this mode could operate. The yield strength he calculates for the mode is 255 megapascals. It should be noted that for modes 1 and 3 Dr Palmer uses the ultimate tensile strength whereas for the other three modes he calculates yield strength.
In mode 5 Dr Palmer extends mode 2 to incorporate the frames enclosing the composite panels and he accepts that he was looking at that mode only to complete the exercise. He calculates the yield strength of the mode would be 3344 Newtons.
In considering the implications of the values he calculated for his five modes Dr Palmer thought that modes 4 and 5 could be left out of account as strength governing modes because of the high values ascribed to them. Nor does he favour mode 2 as an important mode because his results would only be accurate in the absence of enclosing frames. This leaves him with modes 1 and 3 which he considered to be the governing modes. On the other hand Professor Reid considered that one could not tell from a plate analysis if these in fact were the dominant modes but Professor Fenner affords Dr Palmer some support in respect of his general conclusions. Dr Palmer also declared that applying his general experience as an engineer he did not find the conclusions of his calculations surprising. Dr Palmer accepts that at best his total approach is approximate and that he would have got a more precise result using finite element analysis. On the other hand he was not using his method as a tool for construction work (where accuracy could be critical) but merely to get an approximate practical result. Professor Reid accepted that even with finite element analysis the problem of analysing a structure like the firewall is so complex that at best any results will be approximate. What the pursuers were attempting to do was merely to be in a position to say that taking a rough check there was nothing apparently wrong with their thesis.
In completing his analysis Dr Palmer indicated that he considered certain other modes of deformation and did calculations to check his view that the five that he concentrated on were the only ones likely to be applicable to the circumstances of the case. He used the results of the tests spoken to by the witness Cole to exclude on an empirical basis certain possibilities. He excluded the possibility of shear failure in that such was an unusual cause of failure in bolted structures. However the important factor is that the defenders themselves did not specify what alternative modes of failure could be looked at.
Professor Reid in his evidence challenged, at least to a degree, the likelihood of certain modes of failure having been applicable but it has to be noticed that his criticisms in this respect were not put to Dr Palmer. Professor Reid did not challenge Dr Palmer’s values assuming that the modes chosen were in fact the causes of the failure of the firewall. Professor Reid took the view that within the plate there will come a point that the moment stress resultants generated in the plate will be limited by factors determined by the construction of the plate. Once a particular limiting factor is reached then the load will be redistributed within the plate. Thus when the limiting value is reached, he said that you cannot use the ultimate tensile strength of a particular component to determine the failure point of the structure. It must be noted however that it was not disputed specifically that load distribution through the structure must essentially occur through the frame bolts. Professor Reid accepted that the bolts at the outer end of the panel unit frames would see more load than those in the middle and that once they went a greater load would result in those in the middle. One of Professor Reid’s most important points was that before the Mansfield Paper can be applied there needs to be a mechanism to get the load to the bolt and he considered that Dr Palmer had not adequately delineated such a mechanism. Professor Reid makes use of the concept of perfectly plastic response after the yield point where the steel loses its elasticity. That is to say it is assumed that after the yield point the plastic response of the component to added stress does not vary and this is in effect an idealisation to simplify calculation. He produced a graph (number 96/9 of process) to illustrate the concept of the rigid perfectly plastic idealisation. That is an idealisation which for convenience is often used in the collapse analysis of structures. The pursuers laid emphasis on the fact that the bolts being smaller components than say the angle irons would irrespective of comparative strengths be likely to fail before the larger component. The point however that Professor Reid was making was that for collapse analysis one cannot use the perfectly plastic approximation to take the analysis beyond the point of yield to break-up. He accepted that in the case of mild steel if one was only dealing with collapse the linear relationship between stress and strain can then be exploited. On the other hand once yield is reached a strain hardening effect develops and there is a large increase in strain for a small increase in stress so that the relationship between the two ceases to be linear. Because, according to Professor Reid, a redistribution of the load would occur once the bolts get into the plastic range it is not at all obvious that this redistribution would cause failure at a point close to where the plastic regime originated. He concluded that to get to the effect of a movement into the plastic range it would be necessary to follow the process of redistribution through the relevant components. On the other hand at other parts of his evidence Professor Reid seems to accept that if one frame bolt failed adjacent bolts would thereafter fail. Moreover Dr Palmer thought it was perfectly reasonable to take an ultimate tensile strength value to determine when bolts would in fact break. The difference between Professor Reid and Dr Palmer was not as to the appropriate upper tensile strengths but as to whether it had been shown that particular bolts ever reach those strengths. The pursuers argued that it was the view of an experienced practical engineer that once a bolt reaches yield point any redistribution of load will be to the other bolts. Professor Reid’ s view was an academic one and did not, it was said, relate to the actual situation of the firewall structure. No witness claimed that the transmission of the load could bypass the bolts. Thus once the bolts arrive at their yield points they will continue to receive the load until they reach their failure points. The bolts are the only connection. Professor Reid did eventually accept in cross-examination that the load had to go through the bolts. Dr Palmer as a practical engineer supposed on the basis of his experience that the bolts were the weakest part of the structure and he did his calculations to discover if his assumption was correct. The opposite view stated by Professor Reid was that if one considers the chain of transmission a stronger component may fail before the load reaches the weaker. This would depend on the structure as a whole, the particular loading upon it and the interconnections. The loading which a particular part bears is brought about by the way in which other parts deform. Professor Reid is disinclined to assume that the parts of the firewall will each receive an equal distribution of load. One problem in Professor Reid’s evidence in this matter is that although he attacks Dr Palmer’s assumptions about the transmission of the load he does not himself suggest a mechanism whereby the load might be transmitted in a manner which would relieve the bolts. He may be correct when he says that one cannot always assume that the weakest point in a structure will go first but the important issue is does that generality have any implication for the actual structure of the firewall. However Professor Reid did launch two specific attacks on Dr Palmer’s employment of mode 1 and they were not really concerned particularly with the transmission of load. One was that the mode is kinematically inadmissible in the firewall and the other concerns the effect of compressive forces. In relation to the first of these he expressed no quantitative opinion as to the actual effect whereas with regard to the second he accepted that his view was based on assumptions which he could not test. Indeed at one point in his evidence Professor Reid declares that modes 1 and 3 may or may not be capable of mobilisation during the deformation of the firewall. Moreover he accepts that Dr Palmer’s modes are a reasonable method of calculating bending stress resultant if the modes existed in the firewall. Professor Reid suggested that an alternative way of proceeding was to take a value between modes 1 and 5 and once a middle value was thus obtained to apply a limit theory that is using an upper-bound analysis to arrive at a failure point. However this implies the possibility of the load being diverted from modes 1 and 3 and Professor Reid does not actually explain how this would be achieved in relation to this specific firewall structure. There is however this definite difference of opinion between Dr Palmer and Professor Reid in that whereas Dr Palmer looks at the structure and from his experience works out the modes of transmitting load Professor Reid for his part would prefer an upper bound analysis. It is certainly true that the validity of Dr Palmer’s whole analysis depends on his judgment concerning the dominant effect of mode 1 being correct.
On the other hand the criticism by Professor Reid to the effect that Dr Palmer had misapplied an upper-bound analysis has to be viewed against the fact that Dr Palmer was not actually using an upper-bound approach since he had considered that from his experience he could proceed directly to the assumption that the load would be transmitted to the frame bolts and that they were the governing mechanism in the structure. Of course Professor Reid’s approach is based on an idealised plate and is not so dependent on views about the actual construction of the firewall. He considers that the advantage of the upper-bound approach is that one does not have to make broad assumptions about load transmission but rather one postulates a global mechanism and calculates from that. Dr Palmer does make use of plate theory but only as an aid to his analysis based on the actual construction. It seems to me that the upper-bound approach would have an advantage if the task at hand was not to arrive at approximate values but to find relatively precise upper limits of endurance in say a bridge building design.
In productions numbers 101 and 102 the pursuers sketched out what they claimed was a series of hinge lines which would provide a reasonable mechanism for the collapse of the idealised plate and these were put to Professor Reid. Hinge lines are in terms of collapse theory lines of weakness in the structure and lines that would have to deform before the structure collapsed or broke up. Professor Reid accepted that 101 would justify the view that modes 1, 2 and 3 contributed to the situation but he also thought that mode 5 would be implicated. However the point is that Professor Reid seemed to accept that the collapse mechanisms shown in these two productions were physically possible. In these mechanisms one gets deformation at the mode 1 prising points. Moreover Professor Reid accepted that what was shown in 101 and 102 was kinematically admissible and involved having mode1 and mode 3 along the bolted frame connections. Professor Reid thought that the only way to determine the admissibility of the mechanisms he was considering was to link the angular velocities and to link the velocities across the hinges. After doing calculations he accepted that what was shown on the productions was possible. However this did not imply acceptance that the mechanisms delineated were the actual mechanisms which had developed at the time of the explosion. Thus the question remains could the mechanisms shown in 101 and 103 come into play in the actual firewall structure when it receives overpressure.
I agree with the point made by the pursuers that this does not appear to be a case where the frame bolts are acting in parallel with the panel units but rather they are acting in series so that the frame bolts could carry all the load in bending mode 1 or 3. Both Dr Palmer and Professor Fenner appear to agree that once one frame bolt goes the distribution of load that takes place is that the extra load will be transmitted to adjacent bolts so that they will go in turn as the force builds up. However Professor Reid was not willing to accept that anything is certain without a finite element analysis. When Professor Reid was asked if he could visualise a mechanism leading to the break-up of the firewall that did not involve the frame bolts he cited the possibility of the panels being blown out in a way that did not involve the frame bolts but he accepts that that would be "an extreme view" which was not even a likelihood. Indeed he does not offer any realistic case for the deformation of the wall that does not bring the frame bolts into play and yet he hesitates to agree that the frame bolts are the dominant factor in the break-up of the wall. However his hesitancy has a certain tentative quality. He was asked in cross-examination that if the only means of transmitting tension from one frame unit to another was through the frame bolts was it not the case that the mode of failure of the wall will involve the frame bolts. His answer was "That could be the case. As an engineer I would not venture to specify the mode of failure or detailed deformation without doing calculations." This perhaps emphasises the very high standard of validation he expected before he would express any opinion. Indeed he all along seemed reluctant to express any general view even if founded on experience as an engineer unless he had backed up his opinion with detailed calculations. It has to be noted that at present he spends from one-half to one-third of his time doing academic research. At a later stage in his cross-examination when again asked if as an engineer he would have to say that the bolts which connect the frames together would be the only means of transmitting tension between one frame and another his answer was "I expect that they might be". He eventually comes to the point where he unequivocally agrees that the frame bolts would be involved in the deformation process but again cautiously insists that the extent of their involvement would be a matter for calculation. When pressed he specifies his doubts as largely resulting from the matter of kinematic admissibility and also the means by which any of the modes being discussed is actually mobilised. He accepts however that if modes 1 and 3 in fact exist in the firewall then Dr Palmer’s approach would be a reasonable way of isolating the strength of local components. He also accepts that for the break up of the wall the bolts would have to fail. He seemed eventually to come to the view that at least in bending if the outer horizontal frame bolts failed the tension would be transmitted to the inner frame bolts and thereafter to the vertical bolts. Professor Reid accepts that if the angle irons had been welded together rather than bolted this would have provided a stronger connection perhaps stronger even than the angle irons themselves. In fact after much hesitancy he eventually agrees that if the lines of bolts were involved in the deformation process then fewer bolts would be weaker than a larger number. I think Professor Reid’s general approach to the case is again brought out when he is asked if a line is left with fewer bolts is it not likely to be weaker and he answers "If you wish to use the word ‘likely’ you can use that but it does not seem to me irrefutable that this is likely to be the case". The pursuers urged me to conclude that looked at globally the whole of Dr Reid’s evidence amounted to an admission that the lines of frame bolts are the lines of weakness and that modes 1 and 3 would always come into play.
Professor Fenner effectively reaches the same position as Dr Palmer in respect of the components most likely to go in the event of an explosion and indeed his opinions were not challenged when he expressed them. Moreover he reached his opinion as a result of a beam analysis and not by employing an upper-bound plate analysis as favoured by Professor Reid. He began his analysis by illustrating the position if his beam had been unsupported and he then translated his results to take account of the supports actually experienced by the notional beam. Indeed he calculated that in the supported section of the firewall the stresses seen under a given pressure will be 44% of what would be the case if the wall had been unsupported from top to bottom. Thus he is able to calculate figures for the wall with its actual supports by applying the appropriate multiplier to his results for a simply supported wall. Essentially the fully supported wall shows values which are about a quarter less. He calculated comparative stresses on the different components and incorporated his results in the chart number 44/39 of process. In this chart he has two values for the frame bolts depending on whether the force is coming from the north or south. The lower value is when the force is from the south. His results were also worked out with the beam method and he originally based them on the idealised concept of the unsupported wall. He idealised the height of the wall as a beam and balanced the forces working on that height of wall with the forces that would resist the bending moment to keep the wall in equilibrium. Then through bending equations he was able to work out the stresses on the components of the firewall. To arrive at comparative strength for the components he takes their actual physical properties into account. As I have said in his calculations he looked at the position of the supported wall to get a correction factor. This approach of course must essentially be a simplification of the real situation. For example in a beam analysis the lateral support for the wall is not taken into account. His comparisons were not challenged nor was his conclusion that the frame bolts would be the mechanism by which the firewall failed. He expressed the same opinion as Dr Palmer that the Durasteel panels would make little contribution to resisting bending. He accepted that the outer horizontal frame bolts would go first unless perhaps they happened to be supported by a clamp. He did however agree that when he first looked at the wall he considered that he had to do calculations to arrive at reliable comparative strengths. He did not think that in regard to relative component strengths a dynamic analysis would give materially different results to a static one. He accepted that just because a component reached its yield point this would not signify inevitably that the ultimate tensile strength would be reached and that it would fail. It should be noted that he concluded that not all the horizontal frame bolts would take the same stress. In fact he considered that the corner bolts would experience something like 50% more loading than the average.
I do not think it was disputed by any of the experts that if all the bolts failed then the panels would be propelled away from the firewall.
It may require to be noted that Dr Palmer’s view was that if one frame bolt failed then the whole wall is likely to fail since the moments on the remaining bolts would increase rather rapidly so that the bolts would fail with what could be described as an unzipping effect.
6.4.8 Kinematic Admissibility and Compressive forces
Professor Reid said that if you have panel units and angle frames bolted together then a question arises whether it would be possible to get the type of rotation between angle frames which would give you mode 1 or mode 3 behaviour in the firewall. According to Dr Palmer modes 1, 2 and 3 require substantially less loads to collapse them than modes 4 and 5. The defenders argued that it is not enough to define how the weakest components of the structure will respond to a load because the various components are interrelated and therefore the question is not only how will, for example, bolts react to a particular load but rather how will the reaction of the rest of the structure affect the bolts. Since the tension on the frame bolts must be transmitted by the angle irons it follows that these must at least receive some of the load and indeed it is the resultant leverage which causes the prising action which puts tension on the bolts. Professor Reid indicated that because frames were bolted to adjacent frames it might not be possible to get the kind of rotation assumed in mode 1 without deformation of the angle irons. Such deformation would mean the introduction of mode 5 as a factor. The experts seem to agree that mode 1 would be an operative factor but the question is the degree of force needed to make the mode critical. Professor Reid’s opinion arises because the outside edges of the two adjacent panels are restrained. It was suggested that in these circumstances it would be necessary before there was rotation at the joint between two panel frames that other parts of the plate would deform across panels and frames. Professor Reid illustrated his argument by reference to number 96/14 of process. However this submission proceeds on the assumption that there is full fixity at the edges of the panels and this does not seem to have been the case. Professor Reid’s evidence would appear to add up to no more than that kinematic admissibility could be a point although it was obviously not a point that troubled Dr Palmer or Professor Fenner who ignored it and they were not asked why. When Professor Reid analyses a plate then he claims that if the bottom and top horizontal irons were expected to prise apart this could not happen without the vertical angles deforming out of their plane. He did some overnight calculations in an attempt to demonstrate that the pursuers’ approach was kinematically difficult and these are incorporated in Productions 103 and 104. The defenders complain that the pursuers did not thereafter tackle these calculations in detail with Professor Reid. The calculations seem to demonstrate that mode 1 is kinematically admissible but that it would require an element of mode 5 to be involved. Indeed the defenders claimed that the pursuers had accepted that the effective operation of mode 1 would involve the introduction of mode 5 to the point when the frame angles would yield. I find it difficult to construe any apparent concession of the pursuers in this way. For such a concession the pursuers would effectively be abandoning their whole position in regard to mode 1 and I am sure they did not mean to do this. If what they said was intended as a concession it could only have meant that a degree of bending of the frame angles must have entered into the situation. This is very different to a position where the frame irons are required to distort up to yield. I think the position of the pursuers must embrace their whole submissions on this area and as I have said their own experts got no opportunity to comment on what Professor Reid said.
The question of compressive forces was also not raised in the proof until Professor Reid mentioned it and again it was not put to Dr Palmer nor Professor Fenner.
The pursuers made the point that even if the angle irons are deforming this would not mean that they would bring about the failure of the wall because they are much stronger than the bolts. The firewall panels, unlike Professor Reid’s idealised wall, are not fully fixed at the edges and in any event the very deformation of the frame bolts introduces a degree of free play.
Professor Reid refers to the schematic number 73/1 of process and from that seeks to conclude that if the angle irons were fully clamped and fully fixed then a cross-over of angle irons could occur. Dr Palmer and Professor Fenner were not asked about this point either. He also maintained that if the end conditions were fully fixed as he had been assuming then because modes 1 and 3 then become kinematically impossible, compressive forces would arise between the angle frames in a bending action before the bending tension would proceed through the frame bolts. This view was based on an experiment in which the ends were fully clamped irons. These experiments were different from the condition of the firewall. The dimensions of the plate used were materially different to those of the firewall. Nor was the degree of compressive force that might be seen ever quantified. The fixation in the experiment he refers to has the ends of the plate fixed against inward and outward movement. It was said that the degree of deformation in the angle irons which would remove this effect was 0.045 %. Moreover if the compressive forces were to develop this would remove any question of membrane action. But in fact the very small degree of displacement that would be sufficient to remove membrane effects would also be enough to eliminate the compressive effect and this against a background where even Professor Reid accepts that the connections at the ends of the plate are flexible. Essentially any compressive forces to be considered would be second order effects.
In number 96/17 of process Professor Reid further elaborates upon the question of kinematic inadmissibility. He says that when the movement from the horizontal positions of the angle irons take place they cannot move outwards or inwards. This he says will lead to a change in the length of the angle iron of .82 millimetres. However this calculation is on the assumption that the ends do not move out or in. If however if there is a change in length (assuming that the assumption about fixed ends is correct) then this is kinematically impossible without compressive forces at the toes. If these compressive forces arise then instead of the force being transmitted to the bolt it would be absorbed by the toes pushing one another. Moreover if the toes of the angle irons cannot cross each other and their ends are fixed and therefore have nowhere to go then there would have to be some deformation of the angle iron itself. Professor Reid thought that the deformation encountered would be elastic-plastic. He said that if the situation he is postulating were to happen then, unless there is enough load to deform the beam, rotation involving the frame bolts simply could not arise.
The Ove Arup Report which Professor Reid refers to in this part of his evidence is number 110/4 of process. This report makes clear that the experiments were carried out with an idealised plate supposed to be fully fixed at the edges. Indeed the beams used in the experiment were fixed by clamping which was not the case in the firewall. Professor Reid was unwilling to express a concluded view about the fixity of the actual firewall plates without doing calculations. Indeed at one part of his evidence he admits that the firewall top connection of the frames could be a flexible connection which would allow movement inwards when under tension.
The beam used in the Ove Arup experiment was 65 millimetres long as opposed to 5 feet long in the actual firewall. There was evidence in the case that it is not always possible to transpose experimental results to a different scale because of the effects of scaling and indeed this problem in itself opens up difficult questions. The purpose of the Ove Arup experiments was to validate a finite element programme. Although in these experiments there was a clear indication of compression this does not appear to have been quantified. Professor Reid claims that the Ove Arup experiments are interesting because they raise the question of the extent to which one needs to refer to the actual structural detail in determining what happens to the frames over relatively small deflections. This was hardly disputed by the pursuers when viewed as a generality. In the experiments the difference in the length of the iron beam was 4.5% compared with the 0.045% which as I indicated above Professor Reid had previously calculated in respect of the 5 foot beam. He was reluctant to say that such difference would be reflected in the compressive forces without doing calculations. But at the end of the day his evidence on the matters I have been discussing was that the possibility of kinematic inadmissibility and compressive secondary order effects are elements in the analysis which have to be thought about. The problem is that we do not know what consideration Dr Palmer and Professor Palmer gave to these matters because they were not asked.
6.4.9 Bending and Membrane Stress
Dr Palmer’s conclusion was that the firewall had failed in bending. This assumption was to a degree challenged by the defenders who contended that the pursuers had failed to take account of membrane stress. Dr Palmer did not agree with this criticism. The question of membrane stress was put by the defenders to Dr Palmer but not to Professor Fenner. This may have to do with the timing of Professor Reid’s introduction to the case or to the fact that Professor Fenner approached his analysis as a beam analysis whereas Dr Palmer used plate theory where membrane stress is a possible factor. When Professor Fenner said that according to his analysis the wall components would fail in bending he was not challenged. Moreover this information was not merely given as an illustration of the kind of answer a static analysis throws up but as an approximation of what according to his analysis he thought would happen. Certainly if he had been the only expert then membrane stress as a factor would not have arisen but by employing plate analysis Dr Palmer opened up the subject. Professor Reid expressed the view that when one regards the problem on a two dimensional basis as a plate then membrane stresses will arise but the pursuers sought to counter this by arguing that what applies to an idealised plate may not apply to a specific structure. It was claimed that when one looks at the actual plate being considered by Dr Palmer then the fact that the bolted connections could allow the triangular section to move at the edges would remove the effect of membrane stresses. As the pursuers maintain Professor Reid did not quantify the effect of allowing for membrane stresses. Dr Palmer opined that if contrary to his view membrane stress is relevant to the firewall then this factor would not significantly affect his conclusions.
Professor Fenner’s analysis of how individual components of the firewall would respond to an overpressure on the wall emanating from Module C was to a degree dependent on his reading of the schematic 12/93 of process and the drawing 12/83 which illustrate the construction of the B/C firewall. He concluded from this that an overpressure in Module C would be transmitted to the angle frames of the firewall panels by the impact of the force on the Durasteel panels. The impact on the frames would thereafter be transmitted to the frame bolts. Some of the force would thus be transmitted to the cleat attached to the frame and through the clamp bolts to the cleat at the back of the first cleat. The stud bolts attaching this second cleat to the truss would be faced with a compressive direct stress and the load would be carried to the truss itself. The cleats would carry the load partly in bending and partly in tension whereas the welds would carry it in shear. In relation to the angle frame members they will experience bending because they are forced away from the trusses and since they are attached to the deck plate at the bottom and to the chord at the top they will bend in the vertical plane. In other words under this approach the wall would bow out into Module B. There will be bending moments throughout the vertical frame members and since these will be transmitted across the bolted connections the frame bolts will experience the force. The bolts will be put in tension both in the vertical and the horizontal plane. The mechanism described by Professor Fenner was not challenged as being inapplicable to the kind of analysis he was doing. The same comment applies to his conclusion that the top and bottom connections of the wall would hold. He considered that the wall would respond in the way he described along its whole length assuming that the pressure is uniform. To some extent of course the wall will be restrained by the clamps securing it to the trusses although the effect of this will vary with the direction of the overpressure applied to the firewall.. It was never suggested to him that the wall would behave as a membrane.
On calculation Professor Fenner worked out that the panels themselves would not fail before the bolts failed. The frame bolts bear pressure because they are being pushed away. If the loading on the B/C firewall was coming from Module B rather than C then the panels would be pushed into the angle irons and no tensile forces would be experienced by the panel bolts. This would transmit the load to the frame members. In this situation the angle frames would be pushed onto the trusses. The frame bolts would see tensile forces. Again the wall would hold at the top and the bottom.
In relation to an overpressure from Module C to the C/D firewall this of course, unlike the B/C firewall, is a triple skinned sandwich and the load would first be taken on the immediately adjacent layer and then transmitted to the inner material and afterwards to the frames to which the Durasteel panels are attached. The load would be resisted in bending by the frame members and by the cleats that in this case provide the attachment to the truss members. The precise effect will not be the same as with an overpressure from Module B on the B/C firewall since in the case of the C/D wall the panels are smaller being 3 foot square rather than 8 foot by 5 foot.
Membrane stress where applicable has the effect of adding strength to a wall. As I have indicated this matter was first raised by the defenders in their cross-examination of Dr Palmer and in relation to a plate analysis. Dr Palmer agreed that in general when deflections of a plate get relatively large then behaviour is dominated by membrane action rather than by bending. However his view is that the critical criterion is span of the plate whereas Professor Reid’s approach is that it is the thickness of the plate that counts. The essential point is that the more the plate bends then the more the mid-surface of the plate has to stretch and this creates a resistance to further bending. Thus the plate has not only to resist bending but also stretching. The defenders argued that the thicker the plate the more resistance will be offered relative to a thinner plate. Certainly from the publications before me some authorities consider that the relationship between the thickness of a plate and deflection is a relevant factor. Professor Reid suggested that, in relation to a plate, span is an imprecise measurement. It is perhaps a pity that Professor Reid’s view on these points were not put to Dr Palmer. Moreover this view is predicated on the plate being fully fixed at its edges and the deflection equalling or exceeding the thickness of the plate. Dr Palmer also concludes that if membrane action were to occur then the tension on the frame bolts would increase and they would in fact by more liable to fail. The witness Mr Cole was Technical Director of Durasteel and he carried out in connection with his responsibilities for quality testing a test of the strength of a Durasteel panel (the results of which are shown in 13/26 of process). The test was carried out on 9 August 1987 and was designed to determine the maximum pressure that could be applied to a single sheet of Durasteel 3DF2. The plate tested was 1 metre square. The bolting applied to the plate was similar to that employed in the B/C firewall structure. The witness found that when the pressure on the plate was increased to between 4 and 5 psi there was slight doming of the plate. As the pressure increased so did the doming. Just above 10 psi the plate failed when the sheet tore off the bolts - that is the sheet went. The bolts in question were panel bolts and not frame bolts. Before failure the doming had reached about 67 millimetres which can be regarded as severe doming for the size of the plate. This test (the results of which were not challenged) was regarded by Dr Palmer as the type of situation where membrane stresses would dominate. In the test the plate deflection observed far exceeded the thickness of the sheet tested. The membrane action is really a stretching of the material rather than a bending of it. Dr Palmer pointed out that membrane stresses are not linear. The deflections for the wall that his calculations indicated were in his view relatively small and unlikely to introduce membrane stresses. He had calculated deflection of about 30 millimetres at the centre of the triangular plate and about another 60 millimetres associated with the stretching of the bolts giving a total deflection of about 90 millimetres. He thought that the thickness of the firewall was about 50 millimetres. Dr Reid on the other hand calculated that the thickness of the firewall was about 8 millimetres although he accepted that the thickness one arrives at depends on the calculation method used. His calculation is based on the assumption that the idealised triangular plate is of uniform thickness (although in other respects he is not prepared to accept that the idealisation of the plate is appropriate). In fact it was generally accepted that the real firewall differs from the idealisation. The panels in the wall are not isotropic, they are not homogeneous and they are not fully fixed at its edges. Thus for the purposes of membrane effect Dr Palmer argued that the real attributes of the plate should be used rather than the idealised qualities. It must also be noted that although Professor Reid thought that membrane stress was an important factor he at no stage attempts any quantification of the effect. However his view was that if one is calculating the strength of a plate where membrane stress applies and in the calculation account is taken of bending only then the strength of the plate will be under-estimated. Moreover he produced published material that suggested that if membrane stress becomes a serious factor the effect of the under-estimate of strength will be important. In all Dr Palmer for his part thought that in his calculations membrane stress could safely be ignored. Although Dr Palmer accepted that the relationship between plate thickness and deflection affects the significance of the membrane stress factor he emphasised that this is not the only important factor and he drew attention to the fact that the span of the plate and the rigidity of edge attachment could be others. However he concluded that in relation to the B/C firewall up to the moment at which the wall came apart one was not in the regime dominated by membrane action. Engineers in practice use plate theory widely in such situations.
Professor Reid accepted that the ductility of a plate could be a factor in relation to the effect of membrane stress.
Professor Reid opined that if the level of deflection is beyond half to one thickness of the plate then membrane effects begin to appear and as the deflection increases then these stresses become very significant. However it has to be noticed that he is referring to a uniform isotropic plate. With such a plate however he thought that if the plate was fully clamped membrane stresses would be a factor once the deflection surpassed the thickness of the plate even if the triangular plate was clamped on all three sides. However Professor Reid did not go so far as to confirm that the necessary conditions for bending stresses being important attached to the B/C firewall. I think he accepted that the actual plates in the wall would not be uniform, isotropic, nor fully clamped. However I accept, as did the pursuers, that with the proper conditions membrane stress would appear if you had deflection beyond the thickness of the plate. In such a situation a bending analysis would become unreliable since with membrane stress you increase the resistance of the plate to bending. As already noted Professor Reid does not attempt to quantify the difference membrane stresses could make to the calculations in the present cases except to suggest that it could be significant. This is because he all along takes a very strict scientific approach and refuses to venture approximate figures unless he has done the calculations. Nevertheless it must be noted that the Professor thought that the necessary calculations would have required a computer exercise. However he later referred to empirical data which was based on a rectangular plate which he claimed bore similar features to Dr Palmer’s idealised triangular plate. The membrane stress effect which this data demonstrated was illustrated by Professor Reid in the graph number 96/5 of process. The data comes from a book by a Professor Norman Jones (number 99/1 of process) an acknowledged expert in this field of structural engineering and in this book Professor Jones treats the problems presented by the effects on plastic collapse on membrane action. Insofar as the exercise presented by Professor Reid may be relevant it shows that when the deflection increases to about one and a half times the thickness the collapse pressure approximately doubles. Moreover he sought to extract from his reference that as the deflection approaches the thickness of the plate membrane stress may strengthen the resistance of the plate to pressure by a factor of about 30%. Effectively this is because the plate becomes stiffer. Professor Reid accepts that the results he had been delineating would be affected by the geometry of the plate and the actual clamping conditions. He also accepts that the results would not necessarily be applicable to a plate composed partly of Durasteel rather than an idealised steel plate. Moreover he was not asked specifically about the implications of membrane effects for Durasteel panels. Thus Professor Jones’ work confirms that with appropriate conditions membrane effect can add significantly to the strength of a plate.
The pursuers argued that the information given by Professor Jones has little significance when the actual firewall is considered. Professor Reid accepts that the maximum thickness of the actual firewall plate would be the thickness of the frame itself which would be about 50 millimetres. However he idealised the plate as having an equivalent thickness of only 8 millimetres and he accepts that this value may not correspond to the thickness of the actual firewall since his calculation is based on a uniform, isotropic plate fully clamped. He took the flexural rigidity (that is the ability of the wall to resist the loads imposed on it) of the plate as 10,000 Newton Metres. This value was based on a calculation of Dr Palmer and was accepted by the expert witnesses as being the appropriate value in various calculations that were done. It was ascribed the notation D. Counsel for the pursuers argued that to apply deflection of 90 millimetres to the idealised plate (the total deflection at break-up calculated by Dr Palmer) as was done by Professor Reid is a pointless exercise since it does not reflect on the membrane stresses in the actual firewall. All that Professor Reid has done (so it was argued) is to have worked out membrane stresses in an idealised plate that bears no relationship to the actual firewall. On the other hand if regard is paid to the elastic deflection alone (30 millimetres) and that is applied to a plate with an actual maximum thickness of 50 millimetres then the deflection is comfortably within the limit of the thickness of the plate. On the other hand if Professor Reid had taken account of Dr Jones’ formula for calculating equivalent thickness (and this takes account of the plastic element as well as the plastic) he would come up with an equivalent thickness of 80 to 120 millimetres (number 96/42 of process) and that would have to be compared with Dr Palmer’s figure for total deflection of about 90 millimetres. Dr Jones’ work on this matter is supplemented by a paper which is number 92/2 of process. However Professor Reid disagrees that modes of collapse used in this exercise (modes 1 and 3) actually would have occurred and also that a bolted system can be the equivalent of a plate. Nevertheless although hesitating to apply Professor Jones’ formula in calculating equivalent thickness he relies on him in respect of illustrating the effect of membrane strengthening. Perhaps unfortunately the defenders did not ask Professor Fenner about this matter at all. Professor Reid seems to accept that if the equivalent thickness of the idealised plate could properly be calculated as in excess of 80 millimetres this would place a question mark over the significance in these cases of membrane strengthening. Even Professor Reid agrees that the move from the real firewall to an idealised representation of it on a plate is not an easy one.
Of course it has to be considered if when Dr Palmer did his calculations with reference to an idealised steel plate (which is the only approach that would have been practicable) he must also take into account the membrane stresses that would be applicable to such idealisation. Counsel for the pursuers countered such view by arguing that Dr Palmer’s approach had been first to look at the actual structure and at that point he concluded that membrane effects would not seriously affect the collapse of that structure. Thereafter he could carry out his exercise with regard to bending alone albeit that at that point he required to idealise the plate. His exercise only begins after he has concluded that membrane effects do not enter into the actual situation. There seems to be some logic in that approach. Indeed the pursuers urged me to conclude that Professor Reid’s attack on Dr Palmer’s conclusions in respect of membrane effects was academic and theoretical rather than realistic. Indeed at one point Professor Reid states that he has raised the matter of membrane effects because he thought that it ought to be discussed. The real effect of his evidence may be found in his answer when he was asked if he knew if the effect of membrane stress on the actual firewall might be significant and he answered "No I don’t. It would require a complex calculation to determine that". Moreover Professor Fenner also based his calculations on a bending analysis and it was not suggested to him that he ought to have taken account of membrane analysis. His views on that matter would clearly have been of great interest. Indeed it can also be said that Dr Palmer was entitled to a degree to make the assumption that the firewall would fail in bending because Professor Fenner had arrived at that conclusion without challenge. Of course it is contended by the defenders that Professor Fenner was only expressing a view as to the result of his static analysis and that in the real situation it would be the response of the wall to a dynamic situation which would count. In any event applying his own engineering experience Professor Fenner appeared to conclude that membrane effects did not apply but in relation to the general position he conceded that his results might not have applied to a dynamic load. It should perhaps be noted that Dr Palmer differed from Professor Reid in that he considered that if membrane action had in fact been relevant the firewall would if anything have collapsed sooner rather than later because the bolts could not have taken the strain. He thought that the effect of membrane stress resulting would be to increase the pressure on the bolts. Professor Reid accepted that he had not calculated this factor and that there was doubt as to whether the bolts could withstand membrane effects.
Dr Palmer accepted that if he had considered that in the present situation membrane stresses had been an important factor he would not have sought to apply Rayleigh’s method.
It perhaps should be noted that when one is considering the general theory of membrane stresses there is not much difference between Professor Reid and Dr Palmer. It is only in relation to the application of the theory to a particular situation that they may differ.
In his report number 13/76 of process Dr Cox made certain observations about membrane stress but he was not examined about the matter either by the pursuers or by the defenders. The defenders used the comment in the report to suggest that when he did his calculation Dr Palmer must have been aware of the possible significance of membrane stress as a factor. I do not think that Dr Palmer would dispute that. He had simply concluded that the matter could be discounted.
It must also be noted that Professor Reid eventually agreed with the conclusion of Professor Jones that if there is the in-plane displacement of the plate by as little as the width of a playing card then the membrane effect is removed altogether (the relevant publication is "Influence of In-plane Displacements at the Boundaries of Rigid-Plastic Beams and Plates" which is number 99/2 of process). Professor Reid seemed in cross-examination to accept that if there were a displacement of one sixty-fourth of an inch then it would remove the membrane effect. Professor Reid was asked to do an exercise to calculate the in-plane displacement at the edges of the panels of the actual firewall. This is shown in the flip-chart which is 96/43 of process. He comes up with a figure for displacement of one sixty-fourth off an inch for a deflection of about 50 millimetres. Professor Reid did not at the end of the day claim that the triangular sections of the firewall could be regarded as perfectly clamped although he had a degree of reservation as to whether the results of Professor Jones which emerged from a study of a rectangular plate would necessarily apply to a triangular plate. This seemed simply to be because he had not performed the calculations for a triangular plate. However he had earlier appeared to be perfectly happy to apply the results Professor Jones arrived at for the effects of membrane stress to a triangular plate. He accepted that he had not considered the work of Professor Jones on the effect of displacement on membrane stress. Moreover he had based his opinion about the possibility of membrane stress being important on a consideration of the theoretical model used by Dr Palmer rather than on consideration of the potential of the real firewall to be affected by such stress. Dr Palmer had taken an idealised plate fully clamped and in advancing his criticism Professor Reid had proceeded on the basis of the same assumption. He accepted a Paper showing experimental work by a Mr Clarkson (number 99/3 of process and entitled "Uniform Pressure Tests on Plates with Edges Free to Slide Inwards") which demonstrated that when a plate is to a degree free to move inwards at its edges then even if the deflection exceeds the thickness of the plate membrane stress effects will be removed. In these experiments the level of deflection has been ten to fifteen times the thickness of the plate. In relation to Mr Clarkson’s paper the symbol alpha is used to denote the level of fixture of the edges of the plate and alpha equals zero represents full clamping. In submissions the defenders argued that in the present case alpha in respect of the idealised triangle must have a value lying somewhere between 0 and minus 0.2. This is because 0.2 is said by them to represent the value for absolutely no restraint at all at the edges. They argued that if the proper value were minus 2 then this would result in membrane stresses contributing about 685 of the strength of the plate.
Professor Reid accepts that to ascertain the level of membrane force that could in an appropriate case by experienced by the triangular plate would require complicated calculations in excess of what could be done by hand and he had not performed these. Nevertheless he was asked to carry out calculations sufficient to give a broad feel for the position and he carried out the calculations shown in 96/44 of process and related sheets. These were on the assumption that the whole load on the firewall would be carried by membrane action and that the assumptions used by Dr Palmer were valid. The result he obtained from this exercise was 354, 600 Newtons per metre. He made the same calculation for the membrane stresses which a frame bolt would receive (96/47) and came up with the figure of about 41,000 Newtons per metre. He considered the membrane force which would be on the plate and opined that this was a very large force (35 tonnes) and he did not think it could be withstood by a frame bolt. Thus the bolts could not take the force that would be imparted to them by membrane action. Thus if Professor Reid is right to suggest that membrane action might be the force which affects the wall rather than bending then the firewall, as the pursuers point out, would fail anyway. This is of course on the assumption that the pressure pulse spoken to by Dr Bakke can be relied upon because this was assumed to be so in Professor Reid’s calculations.
The matter of membrane stress raises questions but I certainly do not feel able to conclude that the pursuers’ calculations of the strength of the firewall would be destroyed by this factor. Even Professor Reid only advances the matter as an issue for consideration. I am seriously handicapped by the fact that much of the material put to me in submission was not presented to Professor Fenner nor even to Dr Palmer. Dr Palmer for his part with considerable practical experience behind him was satisfied that membrane effects were not significant in the kind of exercise he was conducting. Professor Reid has made no quantified assessment of the precise effect of any membrane stress which may apply to the firewall. The actual relationship of the thickness of the triangular segment to the estimated deflection is a matter of dispute and it is not at all clear that in reality the thickness of the plate would be materially less than the deflection. The implication of the paper by Mr Clarkson is not free from doubt particularly in the light of Professor Reid’s original response to this paper in cross-examination. Certainly the pursuers introduced the relevant paper to the case but that seemed to be in a late attempt to meet certain arguments and material that had not been fully canvassed with their experts. I am not satisfied that the implications of the Clarkson paper can be regarded as having been effectively explored by the experts. There is also the point that the pursuers have made and which does not seem to have been satisfactorily answered in relation to the fact that if membrane stresses had been operational as was suggested by the defenders then the frame bolts could not have met the additional stress.
6.4.10 Dimensional Analysis Equation in the Static Analysis
After defining his modes of deformation Dr Palmer attempts to consider the values of the moment stress resultants that static pressure on the plate would produce. He uses a dimensional analysis and this means in effect that the physical relationship between different kinds of quantities are expressible in forms which do not bring into play their units so that they are dimensionally consistent. The equation which Dr Palmer later applied, and which was extensively discussed in the evidence, was accepted as being dimensionally correct. In order to operate the equation it is necessary to ascribe a value to the constant (k) in the equation and Dr Palmer derives this value from the Paper by Dr Mansfield. This value in Mansfield’s Paper relates to a uniform, isotropic, fully clamped triangular plate. He also used a particular mechanism with particular hinge lines. However in his Paper he describes how his method has been to use arbitrary forms for the distribution of plastic hinges in the collapse mechanism and to minimise the resulting collapse load with respect to these distributions. It is for this reason that he feels justified in calling his method an upper-bound one. He of course works on the basis of an idealisation and Professor Reid made the criticism that the requisite attributes do not apply to the actual firewall plates. Mansfield’s Paper gives a variety of values for the plastic collapse of different shapes of plate with different edge conditions. Dr Palmer in his exercise is seeking to discover an expression that would enable him to look at any particular point on the plate and discover the effect of a pressure on that point. The parameter k depends on the shape of the area of the plate and on there being no internal supports. It also depends on the fixing at the edges of the plate and the condition of its collapse. The calculation of a constant is to find one that would enable the comparison between one triangular shape of a particular area and another triangular shape of a different area. Dr Palmer opines that the value does not depend very strongly on shape which increases its usefulness. He indicates that his calculations have been essentially within the framework of plasticity theory which is what tells us what happens when structures collapse. He treats the triangular section he was considering as if it were a uniform plate clamped at all the edges and made of a material which would fail with a bending stress resultant in mode 1. He applies from Dr Mansfield’s Paper an upper-bound value for k. Thus the value is actually an exact value or above that. Once he has extracted a value for k from that Paper he can apply his expression because he knows the other values. He is in effect relating the strength in the critical mode he has identified to the overall area of the plate. Thus he is seeking to discover what pressure the triangular section of the wall would have to receive to transmit the critical pressure to the bolts. Dr Mansfield’s Paper is in fact number 72 of process. The Paper is dated 1957 and the value for k which Dr Palmer deduces from it for his idealised plate is 50. The value was taken from Table 2 on page 330 of the Paper. The real question which arises in respect of Dr Palmer’s use of the Mansfield Paper is whether it is appropriate to idealise a triangular section of the firewall to correspond to the idealised triangular section which Dr Mansfield uses in his Paper. Dr Mansfield was postulating a fully clamped, uniform, isotropic, homogeneous plate and nobody disputed that his calculations would be valid for such a plate. Professor Reid also argued that Mansfield only gave a true upper-bound value if the collapse mechanism was the same as used in the Paper. However if the value is truly an upper-bound one this should not matter. An upper-bound approach should introduce every possible mechanism and a question arises if it is admissible to apply the calculations where mode 1 is taken as the only relevant mode and the bending moment stress resultant applied is that for that mode. Thus Dr Palmer had chosen to apply not a bending moment stress resultant for failure that could occur at any point of the idealised plate but one that could occur only at the frame bolts in a mode 1 situation. In Mansfield’s Paper the collapse pressure could arise at any point on the plate because he assumes the plate to have the same strength all over. Dr Palmer seemed to be in no doubt that his approach was a valid one and this particularly because the calculations are designed to bring out an upper-bound result. Thus any difference in the boundary conditions between Mansfield’s model and the firewall would not matter because we are dealing with an upper-bound result. Nor would homogeneity or isotropy affect the result. The critical question is whether Dr Palmer’s idealisation is correct in applying the mode 1 value for strength as representing the critical strength of the plate. I perhaps ought to mention that isotropic was defined by Dr Palmer as meaning that if hinge lines are drawn on a plate in different directions the value of m (bending moment stress resultant) along one line would be the same as the value of m along any other line. Dr Palmer accepts that the triangular plate he used in his idealisation was not in practice isotropic because m varies along different lines. The basis of his use of Mansfield was to represent the plate he was using as having values which average out. He considers that he can use the values applicable to mode 1 as his limiting values because the plate will always collapse at those values. Thus he considers that he can idealise the triangular plate as a uniform plate with a material of m equal to mode1. Dr Palmer therefore is not applying Mansfield to the firewall but is only using the Paper to derive a value for k as an upper-bound value in an idealised plate of the attributes he is considering. In regarding his plate as uniform for an application of the Mansfield formula for k he is relying on a judgment based on his general engineering experience.
Dr Palmer does attempt to describe the essence of Mansfield’s method in some detail and since the matter is not without difficulty it could be helpful to repeat what he says:
"The upper-bound theorem says this. We have a structure, we postulate a mode of deformation within the structure. As the structure moves in that mode the various points to which the loads are applied move themselves and therefore the loads do the work on the movements. So we can estimate the work done by the loads on the deformation by adding up the contributions from the modes. Then also when the structure deforms plastically work is dissipated within it because different bits are stretched by different amounts and bent and so on and always in the plastic range. Knowing the mode of deformation we can work out how much work is dissipated internally; and then we can arrive at an estimate of the modes required to produce collapse by equating the work done by the loads in the postulated mode to work dissipated internally. Now that is always a way in which we can make an estimate. What the upper-bound theorem tells us is that under quite general conditions the estimate we arrive at will never be on the low side, it will always be an over estimate of the collapse load or, if we have been lucky to have chosen the exact mode of collapse then our estimate will be exact, but it will never be low".
The lower bound method on the other hand requires that one works with stresses that never break the yield condition. Professor Reid did eventually accept that one does not need to have Mansfield’s mechanism to use his value for k because the result has to be regarded as an upper bound. Indeed I did not understand him to disagree with Dr Palmer’s explanation of upper-bound analysis. Dr Palmer further pronounces that "Collapse to use Mansfield’s value the requirements for the collapse mechanism that you are using, m, is the same all along the hinge lines and that there will yield along all the hinges depicted by the hinge lines". In fact Dr Palmer declares that one can use the figure of Mansfield without defining a deformation mechanism for the plate which corresponds to the deformation mechanisms illustrated by Mansfield.
One of the prerequisites for the use of the Mansfield theorem is that the plate should not be supported by internal supports. Dr Palmer accepted this. Professor Reid however made the point that the grating which supported the deck at a height of about 2 feet above the deck level was equivalent to an internal support. However there was no evidence in the case that the gratings were contiguous with the firewall or that they would have been able to resist a lateral force. Moreover the point was not raised with Dr Palmer. Professor Reid for his part accepted that he had not done any calculations to show the effect of the grating. Professor Reid from his evidence seems to assume that the grating was "in contact" with the firewall and as I have suggested that may not be a valid assumption. Indeed the evidence showed that the grating had to be lifted from time to time which may indicate some space between it and the firewall. After some calculation Professor Reid worked out that if the grating had been rigidly connected to the firewall at a height of about 2 feet then that would increase the failure pressure at the foot of the wall by 20%. On the other hand if the grating was only about 18 inches above the wall (which is consistent with some of the evidence) the effect on the failure pressure would be 16%. If the gap between the wall and the grating had been one inch then there would have been no effect at all.
In my view then there is no factual basis for the introduction of the grating into the matter and even if it had some effect this would probably be slight.
In relation to the fixing of the triangular plates to the structure Professor Reid agrees that where there were frame bolts the edges of the wall could not be regarded as clamped but should be regarded as discretely supported. This may be in contrast with the view he took in respect of kinematic inadmissibility that the edges of the plate should be regarded as firmly fixed. In any event in relation to the application of the Mansfield theorem he claimed that it was not appropriate to regard the plate as fully clamped for the purposes of that theorem. By fully clamped he means that no room is allowed at the edge of the plate for displacement or rotation. However he accepts that if the plate was not fully clamped the k value would in fact be lower. He also makes the point that Mansfield was working on the basis of a rigid perfectly plastic moment/curvature relationship and points out that a square yield criterion is used extensively within plastic plate theory because of its simplicity. The triangle used in the Mansfield Paper is an equilateral triangle and that idealised by Dr Palmer is a right angled triangle but although Professor Reid notes this difference he seems to accept that the Mansfield theorem is not very sensitive to shape. Professor Reid appears to regard the lack of homogeneity as significant although he is less concerned about the lack of isotropy. Professor Reid accepts that when one is regarding a continuous structure such as sections contained in a plate then the symmetry can operate to create the effect of clamping. In this part of his analysis Dr Palmer made use of this factor. Professor Reid does not dispute that the use that can be made of the symmetry point is a matter of judgment. However he differs from Dr Palmer as to whether or not it is suitable to treat the triangular segment as fully clamped. Dr Palmer accepts that his approach arises to a degree from the need to compromise to make the hand calculations he was carrying out tractable. He thought any errors involved would be rather small and in relation to natural frequency which arises in his dynamic analysis the effect of less than perfect clamping would lead to an over-estimate of the frequency. In fact his conclusion was that it is a small detail that makes little practical difference.
Professor Reid does accept that if the edge conditions for the plate are weaker this would make the collapse load for the plate smaller.
The third step in Dr Palmer’s static analysis is to seek to relate the size of the plate and the loading to see if its strength will locally be exceeded. The chart which is number 44/156 of process shows that the collapse values for modes 1, 2, and 3 are relatively close together although the calculations for modes 1 and 3 are for the ultimate tensile strengths of the components affected whereas the other values refer to the state of yield. However the values for modes 4 and 5 are substantially higher so that unless the other modes can be discounted the higher modes can be ignored. Thus in completing his equation Dr Palmer takes the value of 829 Newtons for m because he considered mode 1 to be the critical mode. Then calculating the areas of the triangular sections he gets a static failure pressure for them on 0.02566 of a bar for the larger of his two sizes of triangle and 0.02857 for the smaller. It has to be noted that Dr Palmer does not use limit theory throughout his analysis although he does use it to derive a value for k. His actual equation is not dependent on Mansfield’s plate theory. Professor Reid take a different approach and proceeds as if all Dr Palmer’s calculation is derived from Mansfield’s approach. Thus he says that if you are using limit analysis (Mansfield’s method) then you cannot use only one bending moment stress resultant. You have to use all five separately and do an upper-bound and a lower bound or alternatively combine all five. Now he might be right about that if Dr Palmer was in fact dependent wholly on limit analysis but this of course is the question. Dr Palmer explains that he took the ultimate tensile strength as his limit value for mode 1 since he was interested in break-up and if he had used yield he may not have got to break-up since after yield some of the load would be redistributed to other bolts. He accepts that collapse theory as used by Mansfield is not concerned with break-up but with yield. However he points out that this does not mean that certain applications of Mansfield cannot be applied to break-up. As he says practical engineers have all the time to apply a piece of theoretical work developed for an idealised situation to a much more complicated actuality.
The defenders attack the fact that Dr Palmer extends results obtained from collapse theory to break-up. They say this cannot be done because beyond yield point and failure point response is non-linear. When stress is applied to an object then beyond the yield point the strain which results in extension of the object will manifest itself in a non-linear fashion. This is illustrated by numbers 44/32 and 44/33 of process. Steel has a limited potential for elastic extension if it is subjected to stress. Thus additional force is needed to carry steel from yield to failure. Thus the yield point for typical mild steel is 300 megapascals whereas the ultimate tensile strength would be 450 pascals. Steel components are capable of carrying quite a substantial load after the have reached yield. In order to elevate his conclusion to a failure value Dr Palmer does not introduce yield values to his equation but rather uses the Ultimate Tensile Strength values. It was contended that if this is done all that has been achieved is to predict the yield point of a stronger material. This was because beyond yield the linear relationship is lost. It was also contended that collapse theory is related only to yield. Moreover collapse of a material and collapse of a structure have to be distinguished. It was said by the defenders that generally, particularly in relation to a structure, engineers are interested in collapse rather than failure. This is why the methodology is focused on predicting collapse. Thus collapse methodology assumed that once yield is reached plastic extension will continue infinitely. This idealisation assumes that once there is full yield there can be an infinite stretching or bending of the steel without further stress being applied. In relation to steel this idealisation is in fact an approximation of what actually happens. Thus the defenders argue that collapse theory proceeds upon a calculation related to plastic moment. This is illustrated in numbers 96/8, 96/9 and 96/10 of process. The theory also presupposes no elastic bending of the steel until yield is attained. This assumption is based on the fact that there is limited elastic bending in steel before yield.
Dr Palmer objects to the defenders’ approach and claims that what he has done is to make the kind of practical compromise that is regularly used by experienced engineers. Not surprisingly Professor Reid takes a different view. On the other hand it may be observed that to use collapse theory to get to the point where a bolt would fail probably simply meant that if anything the value obtained for necessary stress is too high, since once the bolt extends to the critical length it of course breaks whether you have calculated the force needed for the extension by collapse theory or failure theory. The mathematics of the matter is complicated but I have difficulty in seeing why if the force on the plate needed to take a bolt up to an assumed yield point of 829 Newtons can be calculated by the application of collapse theory then if the fact is that the bolt would not simply yield but would break at that value the result of the exercise should not be significant. This depends on your knowing that the force on the plate required to take you up to a particular yield point is in fact a force that would also cause your component to break. Of course the assumed yield point fed into the calculation may in fact be well beyond the actual yield point of the component. The question is of course was Dr Palmer’s judgment sound when he concluded that an exercise based on a method appropriate to collapse theory as he used it will give a useful result. The defenders for their part advanced a lengthy technical submission designed to show that Dr Palmer was wrong in thinking that to aim the exercise at ultimate tensile strength rather than a yield point was an appropriate way to proceed. They contended correctly I think that the behaviour of steel after the yield point will be different as it proceeds to its failure point. However as I have already indicated I do not understand why a bolt should not break if a force on a plate which if transmitted through the structure to the critical component (that is the bolt) is known to be such as would cause it to break (even assuming that such force is calculated on a linear basis) provided of course that it is clear that such critical force will arrive at the bolt. The tension needed to take the bolt to its Ultimate Tensile Strength is known before the exercise begins and will take into account the intrinsic quality of steel. What I have expressed seems to be consistent with Dr Palmer’s view. Failure to understand the defenders’ contentions antagonistic to this view may be my fault but the situation was not helped by the fact that the defenders’ detailed arguments (based on the evidence of Professor Reid) were never put to Dr Palmer or Professor Fenner. Dr Palmer can of course work out how much force requires to be applied to the bolt by the mode 1 mechanism before it reaches its ultimate tensile value and this he can do without reference to collapse theory. He arrives at the value 829 Newtons. However he requires calculations based on collapse theory methods to work out how the force he calculates as being necessary to break the bolts, that is 829 Newtons, can be achieved by the application of a load to the wall.
It has to be noticed that Professor Fenner who used a one dimensional analysis to get his static failure pressure gets a materially lower result than Dr Palmer. Thus Professor Fenner brings out a pressure of 0.0072 bar whereas Dr Palmer arrives at the pressure of 0.028 bar but by way of a two dimensional analysis. The implication of this is that Professor Fenner, using a beam method which concentrates on a highly stressed bolt, gets a certain result but Dr Palmer’s result is based on a wider look at the situation since he pays regard to the whole plate. Moreover the results are not quite commensurate in that Professor Fenner was looking at the outer bolts which would see in his view about double the strain of the other bolts averaged out. Dr Palmer on the other hand looked at the average for all the frame bolts. Nevertheless Dr Palmer’s result was used for his dynamic analysis and the defenders can argue correctly that no attempt was made to do a dynamic analysis using Professor Fenner’s static pressure. In a plate analysis one can of course take account of the whole plate which includes the lateral support so that it can be argued that this is a more sophisticated analytical technique than a beam analysis. It should also be noted that Dr Palmer’s method of calculating a static value was an upper bound method which should mean that if the value calculated is not correct then the values are in fact higher than would be needed to collapse a plate.
Professor Reid said that to talk about an upper-bound for part of an engineering structure, as distinct from the whole structure, is not an approach recognised in engineering. However this may come from Professor Reid’s general approach which is that Dr Palmer is seeking to perform a complete exercise based on collapse theory and derived from Mansfield. Dr Palmer maintained that he simply used Mansfield for one purpose - that is to derive an appropriate value for the constant k.
Professor Reid emphasised that the plastic aspect of collapse theory was important. As parts of the structure enter the plastic zone the load is distributed and without taking account of this one never gets to break-up. This would perhaps be true in an idealised exercise. As I have noted what is perhaps the key question is whether as Professor Reid suggests one could not use a uniform bending stress moment unless the collapse mechanism of the structure is known, or as Dr Palmer and Professor Fenner say an experienced engineer can tell that the strength of the whole structure must be governed by the weakest mechanism namely mode 1. The pursuers have perhaps been put at some disadvantage here because Professor Fenner’s evidence that the break-up of the structures could be said to be governed by mode 1 was not challenged. Indeed the same could be said about many of the criticisms later raised by Professor Reid. Moreover one problem with Professor Reid’s evidence is that if his criticisms are valid he was unable to indicate to what extent calculations employing his methods would be substantially different to what Dr Palmer derived. This is particularly pertinent when he says that if the only bending moment which is relevant is that at the weakest point then Dr Palmer’s equation would be a reasonable procedure but that steps would have to be taken before it could be said that it is the correct view. Presumably the same steps could be taken to work out that Dr Palmer’s view of the collapse mechanism is the incorrect view and if that is so Professor Reid does not claim to have established this and clearly not with any degree of certainty.
Professor Reid argued that in a plate plastic moments will develop over the whole plate and then the plate will reach collapse point. However if his idealisation is reduced to a practical situation why should a constant plastic moment be maintained until the angle irons for example reach the plastic yield point. Therefore the pursuers contended that in the actual firewall the plastic load would not develop over the whole of the firewall because the bolts would fail before one arrived at this point. There in fact is a vast difference between the yield strength of the bolts and that of the angle irons. The pursuers refer to a Paper by Professor Kudo as an illustration of a situation where the author uses the perfectly plastic idealisation to deal with the situation of strain-hardening. This illustration is used to suggest that engineers will be flexible and will use idealisation to cope with practical problems where the underlying situation may not be too close to the idealisation. The key to Dr Palmer’s view is that the bolts will reach their failure point before there is any plastic hinge line anywhere else in the structure.
Perhaps the final point I should make on this matter is that Professor Reid agrees that irrespective of calculation the application of the principle we have been discussing cannot avoid sound engineering judgment.
6.4.11 Dr Palmer’s Dynamic Analysis
Having carried out his static analysis Dr Palmer then proceeded to consider the dynamic situation. He accepted that the m value he had produced for his static analysis would have to be adjusted for a dynamic analysis. This analysis is needed because the full overpressure on the firewall exists for a relatively short time (in fact milliseconds). The important interest is to follow the response of the wall to the dynamic situation - that is the development and decline of the overpressure. Indeed if it is desired to proceed beyond the pressure required to break-up the wall to consider what happens to the wall after it breaks up then a dynamic analysis becomes essential. In static loading because the loading develops relatively slowly the deformation of the firewall keeps in step with the load. With the fast loading generated for example by an explosion the wall cannot keep up with the passage of the overpressure. This raises a complication since the overpressure may have passed before the wall has been displaced. Then overshoot may occur after the load has been removed because the movement of the wall acquires momentum. Because the response of the wall in a dynamic analysis is essentially related to time, factors such as inertia can become important.
The defenders argued that the dynamic analysis carried out by Dr Palmer would depend on his views as to Mode 1 being generally correct and this is so. If there is not a mechanism of failure of the firewall dependent on the frame bolts alone then it would be difficult to extract much help from Dr Palmer’s analysis. If the framebolts could not be stretched to the failure point without the need to deform other components then of course this would seriously affect Dr Palmer’s conclusions.
None of the experts disagreed that to understand properly the behaviour of the firewall if there had been an explosion a dynamic analysis would be required.
The first step Dr Palmer did in his analysis was to work out the natural period of the B/C firewall so that he could determine the natural frequency of the wall. To determine this he used a method called Rayleigh’s method. This method is categorised as a single-degree-of-freedom system and an example of this was said to be a mass on a spring. Dr Palmer attempted to apply Rayleigh’s method to his idealised triangular plate on this occasion assuming that it had been fully clamped. He equated the kinetic energy at the centre of the plate as it moves through its central position when deflected with the potential energy at the extreme positions of the plate. Rayleigh’s method was said to be an energy method which seeks to compare the strain energy at the extremes of the plate with the kinetic energy at its centre. The plate of course oscillates under pressure from one extreme to its resting position and then back to another extreme and so on just like the mass on the spring which I have referred to. The mathematical formula applicable to the situation is to be found at 73/1 of process. In order to apply the formula set out there Dr Palmer had to calculate the area of his idealised plate, its mass, and its flexural rigidity. The necessary pressure pulse he extracts from the experiments of the witness Dr Bakke. Thus the rise time of Dr Bakke’s pressure pulse is compared with the natural period of the firewall.
With regard to the C/D firewall Dr Palmer considered that a dynamic analysis was not necessary because it was a much heavier wall which resulted in the inertia of the wall being greater. He also checks this assumption by using the application of formulae derived from Blevin to infer the natural frequency of the C/D firewall on the basis of Dr Bakke’s pressure pulse. He concludes that the response of the firewall will be equivalent to a static loading because the period of its oscillation is much shorter than the period of the pressure pulse. This latter factor is of course derived from Dr Bakke. The point that the defenders make in relation to this matter is that even in relation to the C/D firewall Dr Palmer’s results cannot be entirely divorced from the work of Dr Bakke.
In relation to the B/C wall the static failure pressure of the firewall as calculated by Dr Palmer is fed into the analysis to determine the point to which the wall will have deflected at the time when it breaks up and this degree of deflection is related to the pressure pulse in terms of time. The dynamic pressure at which the firewall breaks up should be different to the static pressure because the response of the wall lags behind the application of the pressure pulse load. Dr Palmer calculates how far the centre of his plate would have moved if it had remained intact up to the peak of the pressure pulse and then he calculates how far the wall would have deflected before break-up occurs due to mode 1 behaviour. He combines the elastic deflection and the plastic component in order to arrive at a total deflection for the firewall. The elastic deflection is derived from the single-degree-of-freedom system and the plastic deflection is calculated from the stretching of the frame bolt. He calculates the elastic component to be about 25 millimetres and the plastic component at about 60 millimetres. This demonstrates that the ultimate tensile strength value is a much larger value than the yield value albeit that the stress required to stretch it the required distance is not so great. Applying the time element as shown in the pressure pulse he calculates a multiplier which he can apply to his static deflection values in order to obtain the deflection at various times. To use the pressure pulse in this way he requires to idealise it. He can thereafter relate the pressure at the time the wall reaches the deflection at which it breaks up to discover the dynamic pressure at the point of break-up. His conclusion was that the break-up pressure was about 0.1 of a bar and his can be compared with his calculated static break-up pressure which was 0.028 of a bar. The time to break-up he calculates at 42 milliseconds. To arrive at his conclusion he treats his triangular sections as clamped, homogeneous, and isotropic because these attributes are claimed to make no difference to the natural frequency of the plate. To justify treating the plates as fully clamped Dr Palmer relies on the symmetry point which I have already discussed.
Dr Palmer makes the point that a system may have more than one natural frequency and that in a complicated structure like the firewall the determination of the natural frequency may be complicated. Insofar as the structure may have several natural frequencies its response to loading is generally determined by the lowest frequency and that would be the case in respect of the firewall. The factors that determine natural frequency include the stiffness and mass of the system as well as area. The natural period is the period the mass takes to go through a complete cycle of oscillation. There are various ways to determine the natural frequency of a structure but the most commonly used is to use approximate methods based on energy principles and in particular the Rayleigh method. Under this method it is assumed that during the oscillation of the structure energy will be exchanged backwards and forwards between potential energy at the extreme positions and the kinetic energy in the middle. At the extremes the structure notionally stops as the oscillation changes direction so at that moment the energy is stored and potential rather than kinetic. However as the structure passes through the middle point of the oscillation there is movement and accordingly kinetic energy. Rayleigh’s method equates the potential energy to the kinetic energy. If Rayleigh is used for one degree of freedom systems it will produce an exact result but otherwise what is produced is an upper bound. Once the natural frequency is derived it can be compared with the pressure pulse. The deformations to be expected in the notional triangular section have to be derived and they have to be chosen in a way that will satisfy the edge conditions. The shape function selected by Dr Palmer is set out in number 73/1 of process and there was no quarrel with his mathematics nor with the applicability of his shape to a fully clamped section. He seeks to take advantage of the fact that the different triangular sections he is idealising are approximately the same and each will restrain neighbouring segments. Because each of the adjacent segments should respond to the load in the same way there should be no rotation at the edges even if the clamping is not complete. The triangular sections used by Dr Palmer are delineated by reference to the trusses of the firewall which provide clamping along two sides of each triangle. The third side of the triangle gains support from symmetry.
Dr Palmer’s calculations for deriving the natural frequency of the firewall are set out comprehensively in number 73/1 of process and fortunately I am not asked to decide any point relating to the considerable and complex mathematical detail.
In general however the situation is that in order to apply Rayleigh’s method the velocity across the plate has to be related to the deflection of the plate at its extreme position. This is calculated in equation 1 which is shown in number 44/173 of process and the equations which result. Equation 2 is used to determine the velocity of the plate as a function of time. The velocity of the plate is the velocity at which any point of the plate is moving as it deforms. It is not a fixed factor since it will vary from point to point on the plate. The velocity will be highest at the centroid of the plate and it is this point which Dr Palmer takes to measure velocity. It should be noted that in this exercise by Dr Palmer the defenders did not attack the mathematics as such. Equation 3 completes the calculation of the kinetic energy at the centroid. In these equations Dr Palmer takes the mass of the plate to be averaged out which he refers to as "smearing" it over the plate. This approach enables him to keep mass outside the integral of his differential equation. There was some considerable discussion at the proof as to whether or not Dr Palmer was justified in adopting this procedure with mass but from the ultimate position of parties it appears that it did not make much difference. In equation 5 we get a formula for natural frequency by putting the equations together. In this equation the mass stiffness and area of the plate are all involved and the equation is refined to a constant. Equation 6 specifies shape. From working through from equation 5 Dr Palmer eventually arrives at equation 35 and I do not consider it necessary to detail the working sheets that bridge this gap since the calculations themselves are not disputed. However questions that do arise are the method of calculating mass, the procedure for determining the flexural stiffness, and whether or not it is appropriate to regard the plate as isotropic, homogeneous and fully clamped. From equation 35 to find the lowest natural frequency one jumps to equation 41 which is the formula for ascertainment of that value. The formula involves values for the mass per unit area, values for the area of the triangular segment, a value for flexural stiffness and the application of the constant which he has calculated at page 8 of 73/1 of process. This constant comes out of the process as 3674 and was not disputed. Dr Palmer calculated the area for his triangle of 14.55 square metres and the value of mass as being 21 kilograms per square metre. The larger panels measuring 8 by 5 foot had a mass of 32.33 kilograms per metre squared.
Dr Palmer concedes that Rayleigh’s method is limited to linear systems such as elastic deflections. He also accepts that to use the method the boundary conditions as relating to fixity have to be determined. Dr Palmer accepts that the plate is only intermittently clamped so that it is a practical compromise to regard it as fully clamped. He is in fact making another idealisation in the interests of simplicity. He also accepts that there will not be exact symmetry. Dr Palmer makes the distinction between arriving at a wrong figure and an imperfect figure. His idealisations seek to achieve the latter rather than the former. His overall view is that the errors arising from his methodology will be small. He supposes that this would be verified if a numerical analysis were done. The natural frequency of a simply supported plate would be about half of a clamped one. If one assumes a higher degree of clamping than in fact exists then there will be an over-estimation of the natural frequency. In general however he considers that the effect of variations on the clamping conditions are not large. A textbook on mathematics by Blevin (which was produced and which is entitled "Formulas for Natural Frequency and Mode Shape") gives results for various shapes of segment and with various support conditions. Indeed some of his conclusions were incorporated into a graph number 95/2 of process and on the basis of this it seemed that the relationship between a partially clamped triangle and a fully clamped triangle was somewhere of the order of 87 to 140. However in terms of Blevin’s book the effect of clamping on results is small and Professor Reid himself did not seek to extract too much from the illustration I have just referred to. Moreover Dr Palmer says that the effect of skew symmetry (a further matter raised by the defenders) would be small. At one point Professor Reid suggested that the actual discrete clamping on the firewall might be weaker than simply supported conditions. He claimed that he would have regarded the plate not as representing a clamped triangle but as one that was clamped at one boundary and discretely fixed at the other two boundaries. He could not quantify the difference this would make to the results and claimed that to do so would require very complicated calculations. However he ultimately accepted that the triangular section would not rotate around its vertical edges. The rotation along the diagonals he maintains is not zero but this may not be significant because he accepts that positive and negative rotation may cancel itself out. However in terms of the natural frequency the results would have been lower. Professor Reid in fact ultimately accepts in accordance with Blevin that the effect of having an intermediate condition between full clamping and no clamping would be relatively small. All his evidence on the matter of symmetry has to be read against the fact that he intimated that he had not really considered the implication of that particular matter. In order to arrive at the above conclusion he has regard to what was described as St Venant’s principle which relates to symmetry in what may be regarded as an infinite structure. This principle was fully explained in the evidence and was not disputed. It should also be noted that Professor Reid did not see that it was his role to do calculations but rather to assess the assumptions taken into account by Dr Palmer. Thus his criticisms of Dr Palmer’s engineering judgments have not been tested by actual calculation. He agrees that in terms of Blevin the difference in changing the clamping conditions at the edge of a rectangular plate is about 12%. This seems to coincide with his general view that the effect of changes in clamping would be small.
If the natural frequency is slower then the wall takes longer to break up. In the various calculations I have been referring to natural frequency is denoted by the symbol omega and the upshot of Professor Reid’s evidence is that if omega has a smaller value as a result of the plate not being fully clamped it is going to take longer for the wall to break up. Dr Palmer’s calculated value for omega was 41 milliseconds and if the value of omega is halved it appears that this change alters the time to break-up to 61 seconds. Thus the effect of a 12% reduction in omega would be much less. The pursuers calculate the effect on time to be to raise the break-up time from 41 milliseconds to about 48 milliseconds. In Dr Bakke’s simulation the pressure cycle rose over 81 milliseconds to a pressure of 0.195 of a bar. Thus even allowing for reduced natural frequency of the order that the actual clamping variations might produce there should be ample time in the pressure rise to accommodate this difference and still leave further energy at break-up to project missiles.
In relation to mass Dr Palmer accepts that if he were not justified in regarding the mass of the plate as smeared in his kinetic energy equation then his equation is defective to the extent that in that event mass could not properly be taken outside the integral sign. If the mass is taken inside the integral sign then separate calculations are being done for the sections of the plate which are variable in terms of mass. Initially Professor Reid indicated that in his view to smear mass was a crude way to proceed. He refers to grillages where the density of the elements is such that you could imagine them to be smeared but he opines that this would not be the case with the firewall. Professor Reid did not think that the relevant calculations could be done by hand. However after being taken through an extensive hand calculation Professor Reid agreed that it does not much matter if the mass is taken as smeared (as Dr Palmer considered it to be) or taken as a weighted average. Indeed he concludes that far from being a crude approximation the smearing of mass gives a good approximation. Thus although the smearing issue was raised as a serious issue it seems at the end of the day to have unravelled itself in relation to mass.
Questions were also raised as to whether Dr Palmer was correct in idealising his triangular sections as being uniform, isotropic and homogeneous. There seems to be no doubt that the firewall did not perfectly reflect these qualities. However Professor Reid did agree that in the firewall the angle irons and the Durasteel panels were, as individual components, each isotropic for practical purposes. These of course represent a considerable proportion of the frewall. Moreover he had done no calculation to discover the effect, if any, of the lack of general isotropy on matters such as the flexural stiffness of the firewall.
Although Professor Reid accepts that the angle irons are homogeneous he does not accept that the same applies to the Durasteel panels or indeed to the combination of components. He accepts that in calculating homogeneity an engineer would seek to take the flexural stiffness of the different components into account. Dr Palmer attempted to do this in calculating the value of D. This involved a degree of smearing the individual values over the triangular segments.
A further issue was raised by the defenders with regard to the correctness of smearing in relation to flexural rigidity (denoted by the notation D). Professor Palmer was strongly of the view that rigidity did not have a great effect on the calculation of natural frequency. He claimed to have done a sensitivity exercise and to have found that the value ascribed to D did not affect pressure much. He maintained that the smearing idealisation is one that is employed by practical engineers all the time. In equations 4 and 5 we again find that in accordance with Dr Palmer’s opinion the value D has been left outside the integral. He does agree that in fact flexural stiffness would not be even over the whole of each of the triangular segments. The smearing of this value he describes as a standard procedure. It has to be observed that there was no challenge to Dr Palmer’s calculation of D when he gave evidence but Professor Reid made a particular issue of this calculation. Indeed Dr Palmer calculated the value of D as being 10,000 Newton Metres whereas Professor Reid proposed a value of 39,000 Newton Metres. In order to arrive at his figure Dr Palmer took two 5 foot lengths of angle irons with a bolted joint between them and then he calculated flexural rigidity by applying moments at either end. This procedure was not challenged with him at the time. Of course flexural rigidity is a factor which enters into the calculation of the appropriate dynamic multiplier. Dr Palmer specifically refutes the suggestion that he arrived at his multiplier by dividing the static deflection at failure by the static deflection at peak overpressure. He maintained that he had calculated deflection as a function of time.
After solving equations 35 and 41 Dr Palmer arrives at an upper-bound for the natural period of 73.7 radians per second. If the radians per second are known the cycles per second can be worked out. This results in a cycle of 2 pi radians. Thus the structure goes through 11.73 cycles per second which is the natural frequency.
Dr Palmer’s general method in respect of his dynamic analysis was first of all to calculate natural frequency to discover whether he required a dynamic analysis. Then he worked out how far the firewall would deflect under static pressure and at what displacement the wall would break up with the involvement of mode 1. This gave him a displacement value for static pressure. The static part of his analysis can then be converted to a dynamic result by applying the dynamic multiplier he has calculated. Of course the argument that was employed against him was that if his calculations were wrong he could get the wrong multiplier. Dr Bakke’s pressure pulse appears in his Report, number 14/46 of process. The relevant pulse is that at point P1 the furthermost west point on the B/C firewall. To achieve his simulation Dr Bakke has filled the east end of C module with gas and ignited it at point IGM as shown in number 15/55 of process. The ignition point is at the south east of the module. The resultant overpressure he concludes to be 0.196 of a bar. This can be contrasted with the overpressure at point P2 which is 0.251 of a bar. The overpressure increases to 0.295 as we proceed east to point P3. At 25 of 55 of the Report the pressure pulse for point P1 is shown. As Dr Palmer states, the magnitude of any dynamic response depends on the pressure applied and on the time over which it is applied. To assist in his application of the pressure pulse Dr Palmer idealised it (number 44/178 of process). Thus there is a simple dynamic form reproducing the essential features of the pressure pulse. It can be analysed as a triangle. By measurement he found an overall duration time of 127 milliseconds, a time of rise to the peak of 81 milliseconds, and a fall time of 46 milliseconds. If the natural period is relatively similar to the time over which the load is applied then a dynamic analysis of the situation becomes important. The natural period calculated and applied by Dr Palmer was 85.2 milliseconds. The natural period is significant for working out if dynamic effects are likely to arise.
The pursuers argued that whatever view was taken of the flexural stiffness question the firewall would always fail before the peak of the pressure pulse thus liberating energy for missiles.
It should be noted in connection with this area of the case that a velocity can continue even after there is no force behind the structure and this is because of the effect of inertia.
In dynamic analysis the lowest mode (mode 1) dominates so that a single-degree-of-freedom analysis becomes feasible. Having calculated the static failure pressure Dr Palmer has to calculate how the wall will deflect statically to that pressure. The static failure pressure is then applied to get the deflection value at the point where the wall would break-up elastically. The static deflection is worked out mathematically from a comparison between the applied load which causes the wall to deflect and the work stored in the wall as strain energy. The calculation applies to the elastic bending and for it to work requires correct idealisation of the same factors as apply to Rayleigh’s method. The equation used is number 43 and the value calculated is W subscript zero which represents the deflection at the centroid of the triangular plate. Flexural rigidity (D) also enters into this calculation. The static deflection at 0.195 bar is calculated thus at 192.8 millimetres. It is of course not expected that the wall would reach this value before break-up since the deflection referred to is at the centroid. Dr Palmer’s next step is to add together the elastic component which is the deflection contributed by the wall and the plastic component which is the part contributed by the part of the wall that is going to fail (the stretching of the frame bolts). To effect this calculation he discovers from tables the minimum allowable elongation for the bolt (which is 14%) and applies this to the length of the bolt shank (12.7 millimetres). Dr Palmer arrives at a total length of the elongation of the bolt of 67.7 millimetres. He then works out what the deflection of the wall will be when the static failure pressure of 0.028 of a bar is applied to it. For that displacement he comes out with a deflection of 28.25 millimetres. Adding this to the figure for the deflection of the bolts he arrives at a total figure of 95.99 millimetres. This will be the deflection of the centroid of the wall at the instant when it breaks-up. However it then becomes necessary to apply the dynamic multiplier. The object is to find the time relative to the pressure pulse at which break-up occurs and from that to work out the relevant overpressure. The dynamic multiplier itself changes with time and Dr Palmer calculates it for cases 1, 2, and 3. The figures were arrived at by way of calculation on spreadsheets. These were not lodged although his relevant calculations for the cases he works on were (number 73/4 of process). He is looking for a time on the pressure pulse when the wall will have deflected to 95.99. When seeking to work out a deflection for 41.5 milliseconds he found a deflection which corresponded to the break-up deflection. He works out his appropriate multiplier at 0.498, that is to say his multiplier for his dynamic analysis. The multiplier is then related to the total deflection if the wall were to deflect intact up to the peak of the pressure pulse. That had been calculated at 192.8 millimetres. He comes up with an answer of 96 millimetres which equates well with his calculated deflection at break-up of 95.90 millimetres. The wall would therefore break-up after about one-half the time of the pressure pulse time rise. This would leave a failure pressure of about 0.1 bar. Calculations showing the effect of applying the other two cases from his spreadsheet show that the case which he did select is the only one appropriate to the situation. The important result of his exercise is that the pressure at break-up was about 0.1 of a bar.
In cross-examination Dr Palmer agrees that in arriving at the static break-up deflection he has added an elastic deflection attributable to the centroid of the idealised plate and the plastic deformation which is based on the deformation mechanism of a non-isotropic firewall section. Dr Palmer also agrees that this is not an ideal procedure because the bolts and the centroid are at different places but considers that it is a reasonable approach to take in the sense of it being practical. He considers that the effect of any misplacement of the centroid in his idealised segment would be very small. He accepts that if his calculations for the computation of the multiplier do not state the true relationship between time and deflection then they fail. The criticism was made that the deflections applied were not truly elastic but incorporated a plastic component. Dr Palmer replies that as a result of mathematics the introduction of a small plastic component does not matter much. Indeed he has only used an elastic model as a practical device to get a reasonable result. This is partly because at the beginning of the pressure pulse it is only inertia that counts and not natural frequency nor for that matter stiffness. Although Dr Palmer accepts that the total static deflection of 192.8 millimetres is based on elastic deformation that is not the same thing as saying that the firewall would behave totally elastically up to 192.8 millimetres because the time span of interest is up to the first 40 milliseconds. The preponderance of deflection would be linear up to the relevant time. There seems to be no doubt that Dr Palmer’s requires his multiplier to be largely based on a linear elastic single-degree-of-freedom system and the question would be is his result materially invalidated if he has added a plastic component to it. However when Professor Reid does the calculation for an elastic/plastic single-degree-of-freedom model he gets results which are similar to those obtained by Dr Palmer’s method. Dr Palmer claims that by the application of power series mathematics it is clear that the first term of his equation only depends on the rate of rise of pressure and the mass per unit area. The value D has dropped out altogether. Moreover it would make no difference if the plate was deforming in an elastic manner or a plastic manner or a visco-plastic manner. The first part of the equation is essentially dependent on mass. The second part of the equation is affected by flexural rigidity but this becomes increasingly important as time lengthens. Thus because Dr Palmer is looking at a time span of 40 milliseconds the effect of flexural rigidity is diluted. Professor Reid for his part agrees that for times up to about 20 milliseconds the first term in the dynamic equation is not significant.
Dr Palmer of course claims to have done a sensitivity study to demonstrate that his calculations are not sensitive to flexural rigidity and the defenders attack this.
There was some attack eventually developed by Professor Reid as to Dr Palmer’s use of a single-degree-of-freedom approach (although as I have indicated this was not put to Dr Palmer directly). To clarify the matter a single--degree-of-freedom system (a concept which I have already referred to at various parts of my opinion) is a system whose response can be described in terms of a single parameter. He raises the question as to whether you can take the lowest or dominant mode as governing the response of the firewall. This I have already discussed in relation to the operation of various modes of deformation. Professor Reid, however does accept that provided the shape of the structure can be maintained one can look at the problem on the basis of a single-degree-of-freedom. He uses as a query the issue of what he calls transient phases before the plate reaches a modal shape (steady state shape). However I did not understand this aspect of Dr Palmer’s work to be fully developed and certainly it was in no way quantified.
Professor Reid contended that a sharp load could cause a deformation which can be different from the kind of deformation that can otherwise be anticipated. It’s a matter of judgment as to how severe a transient phase (that is before the steady state shape) would be. Of course once one arrives at the modal phase and the magnitude of deformation is increasing the shape will remain constant. He accepts that at that point one can describe the behaviour of the plate in terms of a single parameter. An attempt is made to illustrate the shapes assumed in the different phases in 96/18 of process. It is suggested that in transient deformation there is not the same deflection in the centre of the beam as would arise in modal deformation because there would be the same level of deformation over a wider area. The implication which Professor Reid seeks to extract from the situation is that Dr Palmer’s use of the shape function is not accurate. Professor Reid comes to the conclusion that Dr Palmer assumed that the majority of the deformation would be in the modal phase and then he adds "that is often done, it depends on the severity of the loading function." Dr Palmer was not himself asked about this matter. The issue is therefore raised but left somewhat hanging in the air. In fact in cross-examination Professor Reid indicates that he is not in a position to categorise the load on the firewall in this case as being the sort of sharp, severe load that would give rise to transient deformation. As he said "I am simply raising it as an issue, that’s all". He then states that the fact that there is a dominant lower mode does not mean that one can proceed on a single-degree-of-freedom basis but equally it does not mean that you cannot. He agrees that it would be relevant to consider if the calculation required precision or if it would suffice if it were approximate. He also agrees that in such an exercise many people would work on the basis of the lowest mode.
Professor Reid all along maintains that the equation used by Dr Palmer to calculate when the plate would break-up is in fact based on linearity and is therefore only apposite to elastic collapse. Once we get outside the elastic range it is maintained that there is a departure from the straight line relationship. If the plate is going to break-up it will be deflecting through both the elastic and plastic ranges. Thus the relationship between pressure load and deflection is not as simple as the elastic linear relationship. However Dr Palmer only uses equation 43 for the elastic part of his calculation and calculates the plastic element separately. When Dr Palmer uses the elastic equation to calculate total deflection at 192.8 millimetres he makes it clear that such value is not critical because he would expect the plate to break up long before that degree of deflection at the centroid. Nevertheless Professor Reid claims that to get a reliable calculation the equation would have to be changed. The characteristics of the idealised spring would have to be defined. This equation would be much more complicated than that used by Dr Palmer. A Report is put to Professor Reid which is number 109 of process. This is a Report prepared for the Health & Safety Executive by the Steel Construction Institute and it is entitled "Explicit Analytical Methods for Determining Structural Response". The Report seems to suggest that one can do a dynamic analysis of a structure with a single-degree-of-freedom system and that it can be used to deal not only with elasticity but with plasticity provided that the linear quality is maintained by making an idealisation of perfect plasticity. However Dr Palmer does not assume perfect plasticity but works out a separate calculation for the plastic response. Dr Palmer accepts that his method in this respect only leads to an approximation. The Report gives an alternative method in that it suggests it is possible to account for plasticity by the use of a bi-linear single-degree-of-freedom system which binds you to the perfectly plastic idealisation. Professor Reid’s view is that a specifically elastic/plastic model is the way to proceed. He does however accept that with an elastic/plastic model the influence of stiffness is not material in the first part of the elastic motion and moreover is not significant once the structure has reached yield providing the unloading of pressure is not reached. It would of course not be reached if the structure failed. Professor Reid indicates that with the perfectly plastic model as exemplified in number 108/2 of process there should be a correction factor to allow for the fact that in a structure different parts of the wall will accelerate at different rates. That production contains calculations that were put to Professor Reid to test his views. The suggestion was that the model shows that one can use the single-degree-of-freedom model despite the plastic element in the situation assumed. There was in fact a correction factor in these calculations for equivalent mass but Professor Reid argued that there should be a correction factor for equivalent resistance and the rate of increase of force. However without calculation (which he did not do despite having had the production for several weeks) he could not say what effect these correction factors would have on the resolution of the equation. Nor indeed could he exclude the possibility that they would in effect cancel each other out because of the mathematics. For example that would happen if the ratios of the equivalent forces to the total force is the same as the ratio of the equivalent mass to the total mass. He was unable to say if a correction factor of 0.5 would make any difference.
As I have indicated the manner in which Dr Palmer allows for flexural rigidity is attacked and the defenders claim that this factor can distort his conclusions. If the plate had perfect attributes then the calculation of flexural rigidity (D) would be straightforward. However the composite nature of the triangular segment as relating to the actual firewall gives rise to complication. In doing his calculation Dr Palmer took account of the flexural stiffness of the Durasteel sheets, the torsional stiffness of the angle irons, and the axial stiffness of the frame bolts. Number 73/2 of process are the relevant worksheets of Dr Palmer. He calculates the flexural stiffness of the individual components of the firewall. The torsional stiffness of the angle irons is their response to being twisted. In his calculation Dr Palmer calls the rotation in the panels and angles phi 1, the torsional contribution of the angles phi 2 and the contribution of the bolts phi 3. From combining these elements he calculates the value D at 10,000 Newton Metres. The combination process takes account of the fact that the addition of the rotation of the plate components gives a result inversely proportional to the value. He claims as a result of his sensitivity studies to have confirmed that the value of D makes little difference to the pressure and the instant at which the wall breaks-up. Thus there would be failure of the wall even if D was introduced at double the value. If the value was 40,000 the wall would not break-up at all but would simply bounce back. Professor Reid indicates that he would expect the correct value of D to rest somewhere between 10,000 and 39,000 but he could not say where. However he is saying plainly that Dr Palmer arrived at a value which was too low. The pursuers further argue that the value of D would have no bearing on the pressure at which the firewall would break-up. They made the point that it is unfortunate that the value he calculated for D was not challenged with Dr Palmer given that the defenders are claiming that his calculation vitiates his work. The defenders (as I have stated) in their submissions challenged the calculation of D in that because of the supposed smearing it was taken outwith the integration sign but the pursuers answer that Dr Palmer has in fact taken account of the inhomogenities in the Durasteel, angle irons and bolts. He claims that the plate is less stiff because you do not have the welded connection nor the continuity from one part of the structure to the other.
Professor Reid produced spreadsheets number 95/1 of process which set out the results of various exercises he performed. This was produced by the defenders after the pursuers had closed their case and of course neither Dr Palmer nor Professor Fenner had any opportunity to review it. Moreover Professor Reid mounted an attack on Dr Palmer’s computation of D on the basis that no account had been taken of the phi 2 element but this again was not put to Dr Palmer.
Professor Reid in developing his criticisms of Dr Palmer’s methods indicated that D should have been inside the integration sign in the equation rather than outside it since the latter approach was only correct if the triangular segment of wall was uniform. As a mathematical fact this was not disputed nor was it disputed that the triangular segment would not have been strictly uniform in that it was composed of components whose stiffness varied. However as I have indicated Dr Palmer considered that his device of regarding the segment as uniform for practical purposes was justified. In his view the counterbalancing factor produced by the bolted joints entitle him in taking an average view of the stiffness of the segment. According to Dr Palmer’s model there are effectively two half panels bending relative to each other along a bolted joint. Thus his opinion pays regard to the structure of the firewall. Professor Reid indicates that the proper approach is to assume that the edges of the panel are free and this involves concluding that the actual constraints have no effect. He seems to accept that if he was given a uniform rectangular plate made of a homogeneous, isotropic material like steel then experimentally he could calculate flexural stiffness by applying bending moments at the opposite edges of the plate and then by measuring the rotation produced between them although he qualifies this by saying that it is not the way he would do it because it is difficult to measure at the edges. He agrees that the value of flexural stiffness is the ratio between the applied moments and the angle of rotation produced. Professor Reid does not explain why assuming the foregoing method is adopted it is necessary that the plate should be uniform and homogeneous since it is the actual bending across the structure that is being measured. Number 73/2 of process illustrates Dr Palmer’s approach. What is being measured is not an idealised plate but the rigidity of the plate as an actual plate. What Dr Palmer did was after calculating a value for the flexural stiffness of the panel independently of the bolted idealisation, then a value for the flexural rigidity of the iron angles, (along with a separate value for their torsional rigidity), and then a value for the axial stiffness of the bolts he then completed his calculation by measuring the relative rotation in the panel and angles irons as result of the bending moments being applied to the ends of the plates. Then he measured the relative rotation due to the torsion in the angle sections when they twisted and the relative rotation due to the elastic extension of the frame bolts as a result of the bending moments being applied. Then by considering the effect of his exercise on the three phi values he was able to arrive at his result of 10,000 Newton Metres. He did not take his structure as being solid between the frames (which would have given him a higher value for D) but rather took account of the bolted connections to reduce the D value. Under Dr Palmer’s method of calculation the support in the direction opposite to that on which the bending motions are applied would not enter into his calculation. Professor Reid makes a criticism of a calculation made by Professor Fenner on the basis that it was a purported calculation of stiffness but the pursuers argue that it is plain that he was not calculating stiffness at all but the stresses which the individual components would see. Certainly there does not appear to be any evidence which suggests that Professor Fenner set out to calculate the flexural stiffness of the plate (or beam as it was in his case) as a distinct exercise rather than as a matter incidental to a different calculation. He was in fact calculating the second moment of area for the angle irons. The pursuers contend that Professor Reid has taken the result of this limited exercise and sought to convert Professor Fenner’s result not into a value for the stiffness of the angle irons (which possibly may have been a valid exercise) but into a value for D for the whole plate. In any event unlike Dr Palmer, Professor Fenner was not considering the effect of a bolted join. The existence of a bolted join would not have much effect on the calculation of bending moments or stresses but the situation could be very different in relation to a calculation of bending stiffness. Nevertheless if Professor Fenner’s calculation was relevant it would bring out a value for D (as Professor Reid tells us and is not disputed) of 39,000 Newton Metres which is close to the value of phi 1 that was calculated by Dr Palmer. Of course phi 2 requires to be introduced as Dr Palmer recognises and this is the torsional effect which in effect is a limitation on rotation and therefore would increase the stiffness of the notional firewall section. Thus the greatest impact on the accuracy of the final calculation is phi 2. Dr Palmer accepts that 40,000 Newton Metres is a result that can only be produced by ignoring phi 2. It should be noted that Professor Reid accepts that he does not know the detail of Professor Fenner’s evidence. In doing a calculation to test Dr Palmer’s evidence Professor Reid incorporates Professor Fenner’s evidence supposedly on stiffness but I think that the pursuers are right to question the correctness of that procedure particularly as the matter was not tested with Professor Fenner. Indeed Professor Reid accepts himself that Professor Fenner’s method make no allowance for other than phi 1. He accepts that there should be a phi 2 element taken into account but is unable to quantify the effect of this. Dr Palmer of course considers that the effect of phi 2 is to require a reduction of the value of phi 1 by a factor of 4. Thus Professor Fenner and Dr Palmer are in broad agreement as to what would be the proper value for phi 1 (although they arrive at their respective values as a result of different exercises) and this figure is in any event not disputed. Any serious query relates to the value Dr Palmer ascribes to phi 2. However Professor Reid disputes that the problem of phi 2 can be resolved by the simple application of torsion theory as Dr Palmer has attempted. For example in the actual firewall there are not just two half panels bolted along the horizontal (as Dr Palmer posits) but these panels are also bolted along the vertical. That it was said by the defenders would go to limit twisting. Professor Reid attempts to illustrate what he regards as the real life situation by reference to number 96/16 of process. If the extent of the bolting actually limits torsional effect then this would increase stiffness. Professor Reid claims that Dr Palmer has fallen into a trap because he requires to over-simplify to keep his hand calculations practicable. Professor Reid’s view was that the value of phi 2 must lie somewhere between 40,000 and 10, 000. Thus Dr Palmer has understated the stiffness value because he has over-estimated the amount of torsion that would be possible on the actual plate. Unfortunately these points were not put to Dr Palmer nor was it suggested to him that the problem could be solved in any other particular way. If the actual firewall is made up of components some of which are more flexible than angle irons then the correct value for D should be materially less than 39,000 Newton Metres. The pursuers resisted any suggestion that on the matter being considered Dr Palmer and Professor Fenner contradict one another because they were doing different things. Of course in any event Dr Palmer’s fall back position was that with regard to the value of the multiplier X at least in the first part of his calculations (which is the early stage of the pressure pulse) the value of D is not significant. This viewpoint arises from his mathematics so that a question must arise as to the accuracy of their application. In fact the defenders accept that D is less important during the early part of the pressure pulse (which they take to be the first 20 milliseconds or so) but contend that it becomes more important as the time increases. During the initial period the wall does not have the capacity to move much but after that there is a question as to whether it has the capacity to respond by further movement and it was said that at such point the question of stiffness becomes important. Dr Palmer on the other hand says that up to 41 milliseconds when he concludes that the wall breaks up his view on the negligible value of D prevails.
95/1 are spreadsheets prepared by Professor Reid to test out his evidence about the sensitivity of Dr Palmer’s equation to the value of D. In this document it would appear that the result of Dr Palmer’s calculation is only sensitive to the value of D after about 20,000 Newton Metres. The point is that if the stiffness of the wall is low then this will not seriously affect the degree of deflection of the wall but if the stiffness rises beyond a certain value the wall will not deflect so readily. With a stiff wall this would mean, assuming the pressure pulse remains constant that the firewall would fail at a later stage. On the other hand the position in relation to a elastic analysis is not the same as in regard to an elastic perfectly plastic idealisation. This is because the value of D does not matter so much during the first part of Dr Palmer’s equation whereas once the perfectly plastic phase of the deformation is begun then flexural rigidity does not come so effectively into the position. On the basis of Dr Palmer’s calculations 10,000 Newton Metres gives rise to a deflection value of 28.85 millimetres but if the value of D introduced into the spreadsheet is raised to 20,000 Newton Metres then the elastic deflection at break-up is reduced by about one-half. If the value of D is quadrupled from Dr Palmer’s value then this will materially affect the deflection value. Thus if the proper value of D lies between about 20,000 and 40,000 Newton Metres the value of D is going to affect the results. No party quarrels with that generality so that the main issue seems to be whether the value of D is about the figure Dr Palmer calculates or rather falls within the higher area Professor Reid suggests as appropriate. If the values of D between about 20,000 and 40.000 are in fact the proper values then the wall would not fail but would bounce back after the passage of the pressure pulse. However the pursuers say that this cannot be so because it is clear on other evidence that the wall did fail. This would mean that the attribution by Professor Reid of higher values for D (without the benefit of his own calculations) must be mistaken because it does not fit in with the known facts (assuming of course that Dr Bakke’s pressure pulse derived from empirical data is correctly determined). Thus it can be inferred that whatever the correct value of D it must be less than about 25,000 Newton Metres. Of course the defenders say that the answer to the puzzle may be that the explosion was considerably larger than could have been caused by 45 kilograms or so of hydrocarbon a much larger explosion could have resulted in a different pressure pulse that could have accommodated a greater degree of flexural rigidity. Professor Reid agrees that for the range of values for D 6,000 to 14,000 Dr Palmer’s model would show that the variation in D makes little difference to the result although Professor Reid has not included these boundary values in his spreadsheet.
In relation to strain hardening Professor Reid agrees that engineers use plasticity theory for situations where strain hardening exists. Such hardening is the effect which changes the yield point if the structure is loaded rapidly. He also agrees that no model exactly fits the idealised position. Indeed he was taken to Professor Kudo’s Paper and accepts that Professor Kudo was taking a practical approach to strain hardening and proceeding as if the model was perfectly plastic. Professor Reid produced the paper which is number 110/1 of process. It is called "The Effects of High Strain Rates on Material Properties". On the basis of that Paper Professor Reid accepts that if a load is applied very quickly you would have to add about 10% to allow for strain rate. He further ultimately accepts if the assumption is accepted that the frame bolts are deformed and stretched by 14% from the beginning of the pressure pulse then the largest difference of the ultimate tensile strength would be about 10%
6.4.12 The C/D Firewall
Dr Palmer spoke to the attributes of this without cross-examination or contrary evidence being led by the defenders. Dr Palmer’s approach to the C/D firewall was essentially the same as his approach to the B/C wall. Of course Professor Reid made comments on the validity of that approach to the B/C wall but he never said that the same points would apply to the C/D wall. The essence of Dr Palmer’s approach was that he took the local strength of the firewall, then considered how the local actions related to the overall loading whereas the third step was to bring the first two steps together. Dr Bakke’s assumed failure pressure for the C/D firewall was 0.12. Dr Palmer by his calculation arrives at the figure of 0.16 for the failure pressure. Thus in either event it required a lower pressure to cause the C/D firewall to fail than would be the case with the B/C wall. Since at any point along the firewall the overpressure would be isobaric if the B/C wall failed this may reflect on the actual overpressure that must have developed at the C/D wall. Dr Palmer looked at the modes that might operate to cause the break-up of the C/D wall and some of those were rather different to those which emerged in the analysis of the B/C wall. The results of his analysis for the C/D wall are set out in the production number 44/181 of process. The mode 1 he calculates has strength of 1243 Newtons and is similar to the mode 1 for the other wall. Mode 2 corresponds to bending within one of the angles and the value attributed to it is 2571 Newtons. This corresponds to Mode 4 in the B/C firewall. Mode 3 is the prying of the angle irons but about the heels rather than the toes and the value there is 9562. Mode 4 corresponds to Mode 2 in the B/C firewall and has a high value. This is because of the structure of the firewall with panels which consisted of three sheets of Durasteel with composite sandwiched between them. Mode 5 is where the wall bends at the junction between two frames but on the assumption that the bolts do not break and with bending in the legs of the irons to permit rotation. Again the mode has a high value. Dr Palmer indicates that with the blast on this occasion coming from the south he would expect failure in mode 3 around the centres of the triangular sections both in the horizontal joins and the vertical joins. Moreover he would expect failure in mode 1. He would expect break-up to occur by a combination of mode 1 and mode 3 but not mode 2 because it corresponds to the same direction of bending as mode 1 with a considerably higher moment stress resultant. Having reached a view of the critical failure modes in the structure he then applies a dimensional analysis taking Mansfield’s value of 50 for the constant k. This was not challenged. However the validity of this must to a degree be dependent on my views of the same matter in relation to the B/C firewall and the pursuers accepted this. The value of the relevant moment stress resultant was here taken as a mean between modes 1 and 3 and this approach was not cross-examined. Dr Palmer as a result of his static analysis calculated a failure pressure of 0.186 for a triangular area of 14.55 square metres and of 0.167 of a bar for the larger triangular area of 16.2 square metres. He used Blevin’s book to get a natural frequency of the triangular segment which he related to the firewall and treated the edges as simply supported because the wall has a different construction. He calculated the equivalent stiffness of the plate and he concluded that the firewall had a natural frequency of 29.5 milliseconds and that if it were clamped it would have a natural frequency of 15 milliseconds. Because the natural frequency of the C/D firewall is three times higher than is the case with the B/C wall the natural period is about one-third lower. Thus the time of 81 milliseconds is about three times the period of a complete oscillation. This means that the loading time is greater than the natural period and this means that the deflection of the wall is more nearly linear with time than is the case with the B/C wall. Thus in this case the dynamic effects are not significant. He then calculated the multiplier for a time of 40 milliseconds. The net result of his calculations is that the firewall would fail at the lower pressure of 0.16 of a bar. One matter however that the defenders sought to rely on is that Professor Fenner also did a static analysis for the C/D firewall and his result differs somewhat from that of Dr Palmer in that he finds the failure pressure of the wall upon static analysis to be 16 lbws per square inch. If the C/D firewall had a lower break-up point than that postulated by Dr Palmer then (so the defenders argue) this would have had an effect on Dr Bakke’s calculation of the pressure pulse since the C/D wall would have created venting at an early stage.
6.4.13 Comparison between the Strength of Firewalls
A question that arose was the relative strengths of the various firewalls. Professor Fenner looked at the relative strengths of the various firewalls with specific reference to the A/B and B/C walls. The A/B firewall was similar in construction to the B/C wall except insofar as the arrangement of panels was different. Professor Fenner did calculations aimed at comparing the bolt strengths in the two firewalls. He got different values for the two walls in that in the B/C firewall the frame bolts would pry open about their toes whereas in the A/B wall the prying action would be about their heels. He concluded that if the overpressure on the firewall came from the south it would create significantly lower stresses in the frame bolts than would be the case if the pressure came from the north on the basis that the walls were unsupported. He concludes that the firewalls should be weaker under a north to south loading than under a south to north loading. Professor Fenner accepted in his analysis that if the explosion was coming from the south one would not get the mode 3 prying action because of the clamping (although the mode 1 action would still apply). After being referred to specific calculations by defenders’ Counsel Professor Fenner agrees that the values he had for the upper horizontal bolts for the A/B and B/C walls respectively should be reversed so that the relevant frame bolts should see less stress when the firewall is loaded from north to south and the bolts should see more when the wall is loaded from south to north. The strengths are 2,600 and 4,000 respectively and the comments relate only to the bolts where there is a clamp in the height of the wall. However Professor Fenner had indicated that in terms of failure one is looking for an unsupported part of the wall. Professor Fenner also accepted that since he was doing a beam analysis the matter of lateral support was not taken into account in his calculations. He agrees that if there were upper and lower supports at the upper and lower frame bolts the bending moment at the top and lower ends of the region of analysis would be reduced which would also reduce the stresses in the frame bolts along the horizontal connections. He accepts that the vertical bolts would have the effect of providing lateral support which would reduce the stress on the frame bolts. He concluded that clamping would increase the failure pressures he had calculated. He also accepts that long beams would make the firewall less stiff. Because of questions raised about clamps and beam lengths the pursuers accept that Professor Fenner’s evidence is relatively inconclusive. However his conclusion was that with an explosion in Module C the B/C firewall would on a static analysis fail at a pressure of 0.011 bar. If the explosion had been in the opposite direction (that is in Module B) the static failure pressure of the B/C firewall would be 0.0072 bar. The defenders made the point that no one challenged Professor Fenner’s quantitative analysis with him directly. Dr Palmer did not analyse the A/B firewall. Dr Mitcheson took the preliminary view that the B/C firewall would be more vulnerable to an explosion in Module B because of its relationship to the trusses. However detailed analysis suggested that bolt strength could be more significant than the trusses and certainly it was not suggested to Professor Fenner that the firewalls were weaker in an explosion from south to north. On the other hand his conclusions are based on static analysis alone. There is no analytical evidence of the relative strengths of the walls in relation to the direction of the explosion which is based on dynamic analysis. On the other hand the important factor may be the strength of the C/D firewall for Professor Reid accepted that if this failed one would expect the B/C wall to fail as well. On the other hand in the light of the evidence I have been discussing it would be difficult to make a conclusive finding that given an explosion in Module B the A/B firewall would be weaker than the B/C.
Professor Reid accepts that he had given no detailed consideration to the A/B firewall. He also accepts that one might expect the firewall to be stronger when it is being pushed against the module trusses but cautions against relying too much on that because one is dealing with a dynamic situation. Thus his general conclusion is that he could not tell as between A/B and B/ C which firewall was the stronger
Dr Palmer had not really analysed the relative strengths of the different firewalls and any relevant comments he made were rather indirect. However he did opine that in relation to his triangular segments there would be mode1 and mode 3 operating in the firewall whether the pressure was coming from the north or the south.
Dr Mitcheson was asked if there was an explosion in Module B which damaged the B/C firewall what would he expect to happen to the A/B wall. In his reply he observed that in the event of such an explosion the firewall would be forced north against the truss members which would give a degree of support. On the other hand the A/B wall is attached to the truss members on the north of it and would rely on the strength of the attachments if the wall were blown away from the trusses. He would expect the trusses to offer stronger support than the clamps. He concludes that if the B/C wall were pierced he would expect severe destruction of the A/B firewall. He had made calculations as to how the firewall might be expected to behave in the event of an explosion but these were not produced and we were not told his conclusion after calculation. However his general conclusion, at least before calculations, was that he would expect the A/B firewall to be weaker.
Counsel for the pursuers accepted that there was no satisfactorily calculated material that would justify the conclusive finding that the B/C wall would have been stronger than the A/B wall in the event of an explosion in Module B. However experienced experts had at least an impression that the B/C wall might be expected to be stronger. At first sight to the lay person it might seem a sensible suggestion that the wall that would be blown into the heavy trusses would be stronger than a wall blown away from the trusses and onto its clamps. However if the evidence relating to the firewall establishes one certainty it is that the response of the wall to an explosive overpressure is a matter of exceptional difficulty and complexity. The behaviour of the bolted joints for example enter into the position as well as what may be expected from viewing the wall as a whole. On the other hand it seems to be beyond dispute that the C/D firewall was stronger than the B/C wall because it was a bigger, heavier and stiffer structure. At least at worst for the pursuers there is no evidence which would justify a conclusion that the A/B wall was in any material way stronger than the B/C wall so that if the latter wall was destroyed by an explosion in Module B there is no convincing evidence to explain a situation where the A/B wall remained intact.
The defenders of course seek to rely on the fact that Professor Fenner has arrived at results for the strength of the firewalls based on static analysis different from the results achieved by Dr Palmer when he carried out a static analysis. They point out that these experts did not explain discrepancies relating to their respective calculations. Moreover in carrying out his own analysis Dr Palmer did rely on certain on the work done by Professor Fenner. The defenders ask why should Dr Palmer be prepared to accept certain of the assumptions applied by Professor Fenner and yet not recognise the calculations used by Professor Fenner to arrive at results. Of course the method of Professor Fenner was to analyse the firewall as a beam. He then assumes a force is applied to it and failure occurs in bending. Because of the difference between applying pressure to a beam which can only distort in one direction as compared with a plate which distorts in a number of directions and stretches rather than merely bends the question of membrane stress has to be considered in relation to plate analysis. In beam analysis stretching can be ignored because there is an assumed point midway through a beam which does not receive stress because it is the interface between stretching and bending with the result that stretching and compression are equal to each other. Elsewhere I have dealt with the general considerations arising from the phenomenon of membrane stress. Nevertheless the fact remains that Dr Palmer arrives at his results by regarding an idealised triangular plate as being representative of the firewall whereas Professor Fenner takes a slice of the wall and analyses as a beam. Irrespective of the validity of each method then given the approximations and idealisations involved it is not surprising that the results differ somewhat.
6.5.1 Dr Palmer’s Kinetic Energy Method
A matter that the pursuers considered important was to demonstrate that the force that would be generated and break-up the B/C firewall would have sufficient residual energy to cause a projectile upon the break-up of the wall with force enough to damage the 4-inch condensate line in Module B. According to Dr Palmer the kinetic energy of a projectile is calculated by calculating the pressure acting on possible projectiles and then calculate the movement of the projectiles in response to that pressure. Thus the causation requires knowledge of the mass of the fragment, the frontal area, and the pressure acting on it. I did not understand these particular points to be challenged. He is given the assumption that the projectiles would be panel units of 8 feet by 5 feet or 5 feet by 5 feet. He worked out the mass of these units as being 115 kilograms for the larger units and 77 kilograms for the smaller. He was also able to calculate the areas of these units. In order to calculate velocity and kinetic energy of the projectile fragments Dr Palmer set up an equation describing the pressure history and he idealised it as a triangular pulse as he had done before with the firewall dynamic. He then applied Newton’s Law of Motion. He dropped the air resistance term after calculating that it was small in comparison with other terms. He by process of mathematical calculation arrives at a general expression for velocity and displacement as functions of time. These processes are set out in the productions 75/1 to 75/4 of process. In the foregoing context "displacement" is the distance which a projectile will have moved at a point of time. He works out that at the end of the pressure pulse at a time of 127 milliseconds the large panel unit will be moving at 40 metres a second. It will have a kinetic energy of 92 kilojoules and will be 2.3 metres from its original position. The defenders attack this result on the basis of Dr Bakke’s general evidence to the effect that at the end of the pressure pulse even a light panel will only have moved a short distance such as 10 centimetres. However it has to be noted that Dr Bakke is considering the light frangible walls that he used in his experiments and he specifically makes the point that the behaviour of the wall will depend on its particular construction subject to the general comment that if a wall is "very heavy" it will move more slowly when subjected to pressure. Of course light relief walls are designed to relieve overpressure whereas firewalls have a different function, that is to contain fire. Moreover at one point in his evidence he stated that he had done experiments with firewalls and that these had moved as much as 25 millimetres when pressure approached 150 millibars. It was submitted that if this displacement is taken as the basis of a calculation such as that employed by Dr Palmer then the velocity attained will not be 40 metres a second but about 4 metres per second. However it should be noted that Dr Bakke found that the deflection which Dr Palmer calculates at the centre of his triangular plate is 97 millimetre (number 73/1 of process). It was said that if that velocity is substituted then the resultant kinetic energy would be 0.9 kilojoules instead of 92. Dr Bakke is of course only considering generalities. Nor was he asked how he had calculated his 10 centimetres. Dr Palmer is bringing the matter from generalities about light unattached walls to a specific situation founded on Dr Bakke’s conclusion after a computer exercise. Moreover this computer exercise sought to model the particular physical features and configurations of Module C. That would involve as we heard a marked propensity for considerable turbulence. Dr Bakke did not claim that he had experience of a model with the same features.
There is an interdependence between Dr Palmer’ calculations and the validity of Dr Bakke’s pressure pulse. He used the whole of Dr Bakke’s pressure pulse to calculate the kinetic energies and in respect of this approach Professor Reid contends that he should only have taken the proportion of the pulse after 41 seconds which was the point when the wall is said to have broken-up. Dr Palmer’s justification for his approach is that the wall will have been moving before it actually breaks-up and that therefore it will have had a velocity before the failure point. The difference between the two approaches is about 10%. Dr Palmer’s calculation for the smaller panel unit brings out that after 127 milliseconds the panel will be moving at 37 metres per second, that the kinetic energy is 53 kilojoules and that the panel will have moved 2.14 metres from its initial position. As an illustration of the scale of the velocities involved Dr Palmer comments that St Mary’s Church at the end of Melville Street, in Edinburgh has a spire at height of 90 metres and that an object dropped form there would have a velocity of just over 40 metres per second when it hits the ground. The explanation for the larger panels moving faster is that they have more mass in relation to their area. In his calculations Dr Palmer assumes that the fragments do not rotate but move perpendicular to their original plane. It is accepted that tumbling movements could in reality occur (especially if the fragment travels any material distance) but no account is taken of this. Some rotation in the panels is also likely to occur. A finite element analysis could have taken better account of these factors. Dr Palmer did not think rotation would make much difference to the pressure acting on the panel. The pursuers’ objective is not to show precisely how the projectile would have damaged the condensate pipe but to show that there would have been enough energy to have caused damage to the pipe. Questions were raised about the effect of the pressure on the panel if it rotated to become parallel to the floor but the greatest increase in kinetic energy occurs when the panel is less than one metre from the wall. This is because the kinetic energy increases slowly at first then accelerates rapidly before losing energy. In any event rotation before the panel moves more than a metre from the wall could involve the need for some of the panel to turn back into the wall to accommodate it. Thus Dr Palmer considers that it is realistic to use the frontal area of the panel to calculate kinetic energy. When an exercise is done with Professor Reid to arrive at the effect of rotation of the panel it emerged that this factor would only affect the position by reducing the kinetic energy by about 3%. Dr Palmer thought it unlikely that the panel would have rotated so as to present its edge to the pressure at the actual point of breaking free of the firewall. In any event if a number of panels broke away only one of them needed to have the kinetic energy to damage the condensate line. As Dr Palmer graphically stated one does not need to predict the trajectory of the bull through the china shop to arrive at a responsible estimate that damage will be done. The defenders on the other hand suggested that since most of the panels in the wall were affected by welds, clamps, or pipe penetrations the actual number of panels which would leave the wall would be small. However the unchallenged evidence of Professor Fenner was to the effect that clamp connections would fail after the frame bolts. Professor Reid makes the point that in relation to the top panels since the bolts would go before the top connection the panel could be free to rotate before complete severance from the firewall. In fact however the question of the effect of rotation is not central since on Professor Reid’s own calculations the effect would only be a low percentage of the supposed energy. According to Dr Palmer at break-up once one bolt goes the others will go so fast that total break-up will be almost simultaneous. Even if pressure declines before all bolts have gone less pressure will be required to free the balance. It should be noted that to get a time for the break-up of the wall through the bolts an average failure value for the bolts is used.
Since the effect of such rotation as should take place is likely to be small I do not find Dr Palmer’s assumption that the pressure will apply over the whole frontal area of the panel to be unreasonable.
The defenders of course also argue that Dr Palmer should not have taken a point before break-up time to calculate kinetic energy, and that if he had done this and used a higher value of D then the energy values would have been much lower, and that in any event the kinetic energy approach was not apposite and a gas dynamics approach should have been adopted.
Dr Palmer observes that to take the pressure pulse at the time of failure means supposing that at break-up the motion of the panel stops and that it then starts to move again. This he thinks is unrealistic. This matter is important because when considering shock tube theory the defenders’ expert takes the overpressure calculated as at break-up. The pursuers contend that this is not a measure of the actual overpressure which hit the projectile initially because the pressure at failure has been reduced by venting. The defenders for their part submit that it is only the reduced pressure when the fragment is free from the firewall which ought to be regarded as the propellant force. Dr Palmer considers that the firewall is not in fact going to neglect the initial kinetic energy. However he carried out an exercise discounting the time to break-up and this reduced velocities by about 10 metres per second. This evidence does not seem to have been challenged in cross-examination.
The defenders maintain that Dr Bakke’s exercise is founded on a pressure history of the whole firewall. Thus venting begins to have an effect even before point P1 is disturbed. There will be some venting available at the east end of the Module even as the explosion begins. Such factors will go towards limiting the overpressure. They also observe that in his own exercise Dr Palmer applies Dr Bakke’s pressure pulse not some notional higher value. Indeed he assumes that the wall is free to travel at the very beginning of the break-up process. On the other hand if Dr Palmer is right and the initial pressure before venting ought to be taken into account then the shock tube calculation would produce a higher velocity than Professor Stollery acknowledges. Moreover even on the basis of Dr Bakke’s exercise there are higher overpressures than 0.2 bar at other points on the firewall. Whether fragments projected at such points could affect the condensate line is another matter. It would have been interesting to hear Dr Bakke’s views on these questions.
6.5.2 The Gas Dynamic Approach
The defenders also claimed that the kinetic energy method was not appropriate. They also criticised Dr Palmer’s evidence on the basis of Professor Stollery’s view that he had not taken into account the compressibility that would arise when the projectile begins its travel. It was suggested to Dr Palmer that in fact the wall is not moved along by the stored kinetic energy but by the moving gas which surrounds it that is something equivalent to a drag force. Dr Palmer does not think that to be the case. Even the defenders’ witness on gas mechanics, Professor Stollery, agreed that there are three phases governing the gas reaction. In the first the wall is intact but is being accelerated by the rise in pressure. In the second phase after break-up has occurred and venting has begun there is still pressure on the upstream side as calculated by Dr Bakke but there is also some venting of gas through gaps in the firewall. In the third stage although the pressure pulse has exhausted itself the gas is still blowing into Module B away from the explosion and the fragments are flying in that gas. Dr Palmer accepted that the third stage may cause some additional acceleration but did not consider that it was necessary to calculate it. Dr Palmer in fact does not favour the gas dynamics approach since he thinks that we are not considering the case of an explosion acting on isolated fragments and in any event the gas effect would only develop late in the process. However he accepts that one would arrive at lower figures using the gas dynamic method. What Dr Palmer is in fact saying is that since it is not only isolated fragments that are affected by the gas but the whole wall the gas cannot cause any effect until the wall is removed. At the point where the wall first breaks-up there will be gas venting through the openings but still pressure acting directly on the panels. Even at the third stage there would be gas pushing on the panels as well as to a degree acting as a drag force. However Dr Palmer is in effect ignoring stage 3. Professor Stollery indicated that the drag force of an object will be proportional to the square of the relative velocity.
When Professor Stollery was cross-examined he agreed that before the gas breaks beyond the firewall a degree of movement will have been experienced by it and it will gain some momentum. His evidence was circumscribed by certain assumptions. He assumed that the overpressure was 0.2 of a bar. He assumed that the panels were free to fly in a planar position at time zero. He then used shock tube theory to get an upper-bound velocity and then makes another calculation to work out what he called a more realistic velocity and kinetic energy. Shock tube theory is based on equations of motion and Newton’s laws and it has been validated by shock-tube work. However it arises from this approach that there is some divergence from the real position in that there would be overpressure on the firewall before the fragments were free to fly. He had not read Dr Bakke’s Report. Moreover the result of an overpressure which Dr Bakke gets takes into account venting and the movement of the wall so that the use of an assumed pressure of 0.2 does not in fact correspond with what Dr Bakke concludes. It also has to be noted that Dr Bakke’s overpressure at point 1 is not the highest pressure on the wall but only the pressure at that point.
The witness Professor Stollery was well qualified in aerodynamics. He was 64 years old and professor of aerodynamics and head of the College of Aeronautics at Cranfield. He has held this since 1973. Although aerodynamics is concerned with the science of flight it includes the study of motion of air and gases and gas dynamics. He holds BSc, M Sc, and DSc all in Engineering. He is a Chartered Engineer and a Fellow of the American Institute of Aeronautics and Astronautics. He is also a Fellow of the Royal Academy of Engineering. He is an expert on shock tube research and has written a book on it. Because his evidence had not been put to the pursuers’ witnesses it was objected to and allowed under reservation.
He explains that with a shock tube exercise one begins by supposing a long pipe divided into two sections with a piston in the middle which is held. There is a very high pressure behind the piston and a much lower one in front of it. Thus when the piston is released the pressure will drive the piston into the low pressure region and ahead of the piston will be waves. If the mass of the piston is reduced then the piston will accelerate more quickly and the waves ahead will quickly catch up with each other and compression waves will be formed. The waves behind will never catch up and a rarefaction fan is formed. The mass of the piston can be removed altogether and the gases divided by a membrane. If this membrane is released the waves ahead would catch each other up almost immediately to form a shock travelling down the tube. Behind are the rarefaction waves telling the gas ahead to accelerate. The greater the mass of the piston the slower the rate of acceleration will be but not the ultimate velocity. If one goes to the idealised situation for zero mass then you can have optimum velocity. That is the optimum velocity that can be achieved so that whatever velocity the missile may have it will never be higher than that. As well as zero mass, zero inertia and zero friction is assumed. Thus whatever the situation is in regard to mass there is an upper bound in respect of velocity. The shock waves can be measured. What we have is compression of the air ahead of the piston and an attempt to equalise the pressure ahead with the pressure behind. The defenders proceeded on the basis that the relevant pressure pulse is that point where Dr Bakke derived his pulse from since there was no evidence that the condensate line could have been damaged by a projectile from anywhere else.
He was asked to apply this technique to the firewall regarding this as the separating membrane. Indeed it was contended that this is the most effective way of measuring velocity. There is a one-dimensional flow of gas behind the assumed piston and it should be possible to derive an upper-bound for the velocity of gases behind the piston given an assumed overpressure of 0.2 bar. However the velocity of the gas behind the assumed piston and the mass of the piston will affect the velocity of the flow. Professor Stollery indicates that the stage is reached when the piston has a constant velocity. That happens when the pressure behind the piston comes to equal the pressure in front of it. The way in which conditions along the tube vary with time can be measured. To do this Newton’s Law of Motion is applied. Thus if the actual piston is a firewall with material mass rather than one with a zero mass the velocities are bound to be less. However when the gas is travelling along Module C points along that wall may break-up before more western parts and in his calculations Professor Bakke allows for the effect of venting and porosity. The venting factors are stated as a function of time that is to say the time at which each point on the wall will reach the required break-up pressure reaches 0.1 of a bar. The assumption that Professor Stollery was asked to make that would render shock tube theory admissible is that the panels can be assumed to hang in air with a pressure acting on them from Module C. The shock wave pulls surrounding air into motion but at a velocity considerably less than its own velocity. He said that if there was atmospheric pressure of one bar (close to Dr Bakke’s pressure at point P1 of 0.192 bar) and a zero mass piston the velocity would come out at 22 metres per second. This comes from standard tables. The suggestion is that using shock tube theory that is the fastest velocity that is possible because it assumes a massless piston. In fact it was further suggested that less will be attained in practice because as the propellant gas enters Module B it will spread out. The maximum velocity can only be achieved with a one-dimensional flow and not by a three dimensional flow such as would have happened in Module B. If this reasoning is applicable then Dr Palmer’s results must be wrong. The pursuers contend that the flaw in the defenders’ use of shock tube theory is that the assumed pressure of Dr Bakke is not the exact original pressure because an allowance is built in for venting. Professor Stollery’s application of shock tube theory to the firewall situation is set out in number 114/2 of process. There is no dispute that if one uses those equations one gets the results shown. The velocity of the piston cannot exceed the velocity of the flow of gas because the gas cannot get past the piston from behind. Once again this important evidence from Professor Stollery was not tested with any of the pursuers’ witnesses. It was not put to Dr Palmer, nor Dr Bakke nor Dr Mitcheson. Dr Bakke and Dr Mitcheson referred incidentally to shock wave theory in their evidence so that they are obviously qualified to say something about it but they were not asked about its specific application to the firewall and the projectile question. When Professor Stollery does an additional calculation of the kinetic energy in the firewall panel he actually gets a velocity which is higher than his upper-bound. He was not able to explain this very well. In relation to the venting factor Dr Bakke allowed for in his calculations he reckoned that the wall would vent by about 20% at the monitoring points.
By the time the overpressure reaches point 1 it will have lost pressure. If shock wave theory does apply to the firewall this would reduce the kinetic energy to about 20 kilojoules rather than 90. One feature of the firewall theory that may make a difference is that the wall does not represent a stationary piston but before it is released it is effectively moving. Professor Stollery accepts that if one is visualising free standing panels then in terms of basic theory they would if planar be propelled by the difference in force between the pressure behind them and that in front. Complicated questions such as drag would also come into account. He agrees that the velocity of the panels could never exceed the velocity of the flow. Moreover the pursuers contended that at stage 2 of the three stage process the wall is only partly vented and therefore the compressed pressure at the vents could create added velocity. This is equivalent to a shock wave piston with a hole in it and Professor Stollery accepts that such a situation could create higher velocities. Indeed he accepted that shock tube theory makes no allowance for venting but as is plain from the above Dr Bakke’s results take venting into account although he accepts that the porosity will not have reached 20% by the peak of the pressure pulse. The wall is allowed to move away from its original position to the extent of about 1 metre. Dr Bakke in effect says that if there was no venting there would be greater overpressures in Module C at point P1. This emerges particularly from the graphs in his Report number 14/46 of process. Of course since Professor Stollery had not considered Dr Bakke’s Report he would not have known how the pressure pulse was derived. He accepts that with no venting the overpressure in the module would be approximately doubled. Indeed Dr Mitcheson expressed the opinion that with no venting at all in the module the overpressure might have reached 8 bar. An overpressure of 0.4 of a bar such as might well have attacked the wall had there been no venting could have doubled the upper-bound velocity to over 40 metres per second.
Professor Stollery’s alternative calculation for velocities was based on a Workbook from NASA which was 114/1 of process. He was again seeking for a precise calculation of velocities rather than a mere upper bound. The said Workbook is entitled "Workbook for Predicting Pressure Wave and Fragment Effects of Exploding Propellant Tanks and Gas Storage Vessels". It was compiled by Mr Wilfred Baker an expert in explosives. In Appendix IV .D of this document there is a section entitled "Estimate of Velocities Attained by Appurtenances Subjected to Blast Loading". According to Professor Stollery an appurtenance is merely a thing which gets in the way and would include vehicles, cars, and people. He accepted that he could use the method in the Workbook to calculate the velocities attained by a panel of the firewall. In considering this he says that immediately the panel leaves the firewall and is hanging in the air as it were the high pressure of air will rush through the hole around the plate and move it downstream. However since the creation of projectiles as we have seen involves three stages it is not the case that the panel initiates its journey, so to speak, as if it were hanging stationary in mid-air. His calculation reaches a velocity of 11 to 12 metres per second with maximum energy of 8.43 kilojoules but the problem from my point of view is that he is treating the firewall (a fixed structure) as if it were a free standing appurtenance. When asked to do a different calculation for a lighter projectile with a third of the mass he arrives at a much higher velocity (namely 34.5 metres per second) which he declares gives him some concern for the use of this particular equation. In terms of shock tube methodology no velocity should exceed the upper-bound. Moreover the NASA method is designed to deal with secondary missiles whereas the firewall panels it was said are in fact primary missiles. As Mr Baker observes the term "primary fragment" denotes a fragment from the casing or container for an explosive source. The pursuers argued that if Professor Stollery had treated the projected panel as a primary fragment and applied an appropriate level of mass he would substantially have exceeded the velocities he calculated. There is no direct evidence that the panel was in fact a primary missile other than what can be extracted from Mr Baker but then this question was not raised with the pursuers’ witnesses. There is also a question as to what is the appropriate mass to apply for the projectile. The defenders submit that the problem with velocity for a lighter fragment which the pursuers raise arises because the NASA paper has been taken beyond its proper parameters. Nevertheless Professor Stollery accepted that the matter of the lighter fragment caused him some concern. The defenders argue that Professor Stollery was able to get to his 22 metres per second velocity before he attempted to test it with the NASA material. The NASA exercise was an attempt to get beyond an upper bound to an exact figure. They contend that even if the NASA formula is not considered to be reliable it does not affect the separate question of an upper bound.
It should perhaps be noted that even Professor Stollery’s calculation of low velocity could leave sufficient energy to damage the condensate pipe in terms of certain of the modes of deformation spoken to by Dr Palmer assuming of course that he is right about these.
Professor Stollery seemed to accept that he was not an expert in explosions. Dr Palmer claimed to have expertise in fluid mechanics and this was not challenged. Dr Bakke is an acknowledged expert in gas dynamics. He was also familiar with shock tube methodology but the defenders did not see fit to seek his views on Professor Stollery’s approach.
Professor Reid did not profess to be an expert on the calculation of velocities and kinetic energies but nevertheless was asked to express certain views on the matter. He certainly had the basic engineering experience to express a significant view. He claimed that he could judge the method of calculation used by Dr Palmer in working out his equation for velocity on the basis that the underlying assumptions were true. He maintained that in completing the equation only the part of the pressure loading applied after the panel is completely separated from the firewall should be taken into account. Doing the calculation in what he considers to be the correct manner he works out a velocity of the panel at break-up of 35.65 metres per second. If the break-up had occurred not at 42 milliseconds but at the peak of the pressure pulse then a velocity of 15.19 metres per second would be achieved according to his calculations. That calculation might be apt if the value of D was changed from 10,000 to 25,000 as Professor Reid says ought to have been done. This is the best case the pursuers can advance because beyond 25,000 as the value for D the wall would not break-up at all (assuming that a leak from PSV504 caused the explosion). In another calculation Professor Reid is asked to calculate velocity using the perfectly plastic model that had been used in the firewall analysis and there he came out in respect of the larger firewall panel with a result of a velocity of 37.6 metres per second. In relation to this calculation he applies a value for D of 10,000 but accepts that to take D at 39,000 would not have made much difference to the result. Thus he accepts that calculating velocity according to that model is relatively independent of the value attributed to D. The defenders claimed that in respect of projectiles velocity was a critical parameter. It was argued (correctly) that energy is the product of half the mass multiplied by the velocity squared. Thus if you reduce energy by a factor of 10 you are reducing velocity by a factor of 100.
6.5.3 Missile Impact and Pipe Penetration
Perhaps the pursuers’ main objective in seeking to establish the forces likely to have destroyed the B/C firewall was to reinforce the possibility that the wall would have broken-up with sufficient force to propel fragments into Module B that would have been able to rupture the 4-inch condensate line there. Of course in deciding their strategy the pursuers required also to forestall any attempt by the defenders in their evidence to prove that the pursuers’ explanation of the accident was not possible. There is little doubt that the initial explosion was shortly followed by a massive fire in Module B and if the pursuers could show that the condensate pipe was likely to suffer damage as a result of an explosion in Module C then this could provide an explanation for the fuel necessary to feed this fire.
The pursuers’ principal case is that the 4-inch condensate pipe was ruptured by a fragment projected from the B/C firewall and in this connection a rupture is a leak path from the inside to the outside of the condensate line. Elastic deformation will not cause a pipe to rupture as it requires a large plastic deformation to cause this result. The relevant factors are the wall thickness of the pipe and its diameter, the yield stress of the pipe, applicable strain rates, and velocity impacts. The support of the pipe is also relevant. With regard to the impacter the material it is made of, its shape, its kinetic energy, and its velocity are all important. The energy transmitted to the pipe upon impact is absorbed by the plastic deformation of the pipe, any fracture of the pipe, the energy radiated along the pipe to its supporting structure, and sometimes the energy transmitted along the fluid contained within the pipe. Thus if a pipe is filled with a liquid such as condensate and struck a blow some of the energy of the impact is transmitted to the fluid as compression waves and thereby transmitted away from the point of impact. Dr Palmer is asked to calculate various modes of deformation of the pipe on the assumption that the schedule of pipe thickness had been as shown in Schedule 80 or Schedule 120 which were apparently those applicable to the pipe in question. Dr Palmer describes possible modes of rupture. The pipe could suffer a highly localised puncture or it could be severely bent upon impact with extensive deformation at the bend. The pipe could be severely dented or it could be cut. Thus the possible modes of damage could be indentation, bending, denting, or cutting. Dr Palmer considered each mode of deformation separately and worked out the energies required for each of these using calculation or experiment.
Dr Palmer found that the energy need to bend a Schedule 120 pipe through a right angle was 48.3 kilojoules. With regard to rupture by indentation, allowing for strain rates, he found that 7.7 kilojoules were necessary to rupture the Schedule 120 pipe and 3.7 to rupture the Schedule 80 pipe. To cut a pipe over one-half of its circumference the energy required is 29 kilojoules for the 120 pipe and to cut such pipe by one-eighth of its circumference the figure would be 4.8 kilojoules. In relation to denting the energy required to rupture the Schedule 120 pipe he worked out at 23 kilojoules with the figure of 17 kilojoules for the Schedule 80 pipe. Thus he concludes that kinetic energy of less than 10 kilojoules could rupture the pipe by one mode and that rupture by other modes could occur if the kinetic energy was between 10 and 50 kilojoules. He adhered to his calculation that an 8 foot by 5 foot frame could generate kinetic energy of 92 kilojoules in consequence of the assumed explosion even allowing for the effect of air resistance. The smaller panels of course would have rather less energy. Thus the energy available to damage the pipe would not merely be marginal but more than ample. The corner of an angle frame would be just the sort of missile that could cause a damaging indentation. Dr Palmer cannot say exactly how the pipe was damaged (and I am satisfied that nobody can) but his point is that with the energy available it is possible that the pipe was ruptured in a number of ways.
The witness Mr Neilson was responsible for a number of tests on projectiles called the Winfrith tests and the results of these is recorded in the Report which is number 63 of process. The Report was published in March 1988. Dr Palmer finds that these tests cast light on the problem he has been considering and indeed claims that they support his conclusions. The outline of the tests was that rod-like missiles (referred to as billets) were fired at a pipe. The pipes used in these tests were 6 inches in diameter rather than 4 and the Schedule was 80. The projectiles were lighter than those that may have come from the firewall but Dr Palmer sees sufficient correspondence for the tests to be useful. The Winfrith pipes were filled with water rather than condensate but Dr Palmer claims that the tests show that less energy is required to perforate a liquid filled pipe than an empty pipe. This is because the liquid offers a resistance to plasticity and indentation so that the pipe breaks more readily. Professor Reid agreed with this. The damage also tends to be more localised.
Mr Neilson’s experimental work was in general not challenged. He was aged 44. He graduated BSc in 1971 from Exeter University and between then and 1993 he had worked with the atomic Energy Authority at Winfrith. He had held the post of Principal Scientific Officer of Impact Analysis from 1987 and then was promoted to the Head of Experimental Safety. In 1992 he became Product Manager and he left the post in 1993 taking early retirement. Since leaving that body he had worked as an independent consultant.
Dr Palmer was particularly interested in Tests 477 and 479 recorded in the said Report. The pipes in these tests had a wall thickness of about 10.5 millimetres compared with a wall thickness of 11.1 for a Schedule 120 pipe. The target pipes were held in clamps. The Tests confirm what would be expected that the kinetic energy required to penetrate a Schedule 80 pipe is less than is required for a Schedule 120 pipe. Dr Palmer concludes from the tests done that with a missile of 25.4 millimetres diameter the pipe can be penetrated with impact kinetic energy of less than 15 kilojoules. This would correspond to 18 kilojoules for a Schedule 120 4-inch pipe. Dr Palmer indicates that the perforation energy is in proportion to the inverse square root of the outside diameter. He is comparing pipes of the same wall thickness but of different diameters and in this situation the pipe of narrower diameter will require more energy to penetrate it. Dr Palmer refers to tests 271, 272, and 276 and he deduces that the effect of a liquid fill in the pipes is to reduce the energy required for penetration by about 30% so that filled pipes would bring the energy required to pierce them down to about 13 kilojoules.
In his Report Mr Neilson sets out a correlation which he arrived at by dimensional analysis and by means of which he sought to correlate his experimental results and the results of other published tests to pipes of different types and sizes. Dr Palmer also claimed to find support for his results in those correlations. Mr Neilson found that the larger the diameter of the impacter the more energy is required. Calculating on the basis of these correlation Dr Palmer confirmed his view that the explosion in Module C would have caused projectiles of sufficient kinetic energy to have had the capacity to penetrate the condensate pipe. It is I think important that he states that in accordance with his experience as an engineer the results he arrived at are just what he would have expected. Although Professor Reid attacks aspects of Dr Palmer’s approach he accept that as a general engineer if fragments were flying about Module B he would expect damage. This seems an obvious conclusion and it certainly is in any event an important concession.
It was put to Dr Palmer that not too much can be taken from the Winfrith correlations because the velocities were high and the target pipes were short and rigidly fixed. He accepted that the results of the correlation were sensitive to the diameter and shape of the projectile. However he did not think the velocities and the fixing were important parameters. He concedes that the parameters of the Winfrith tests are further from the accident situation than ideally he would have preferred but his assessment of Winfrith he claims is partly dependent on the work of British Gas in the Paper I have referred to already. Thus in the British Gas experiments there are much lower velocities. It was found in the British Gas Paper that their conclusions although based on much lower velocities matched the Winfrith correlations quite well. The formula for translating velocity into kinetic energy is one-half times mass times velocity squared. Thus velocity is always squared in relation to mass. This illustrates the significance of velocity as a parameter. Mr Neilson himself recognised that his tests were devised for particular parameters and that the further one gets away from these parameters the more scope for error in his correlations. He accepts that his results can be applied to other geometry’s but the degree of error likely is a matter of judgment in particular cases. Within a reasonable range of divergence he estimates that the deviation from his own results might be expected to be no more than about 30%. However caution is clearly required when using Mr Neilson’s results outwith the range of his own experiments.
In terms of general engineering judgment both Dr Mitcheson and Mr Cubbage indicate that they would expect the assumed explosion to damage the pipes.
Mr Wottge gives some detail of the specification of the support arrangement for the 4-inch condensate pipe. He organised the replacement of about 5 feet of the line and the part replaced was Schedule 80 piping the rest being Schedule 120. The Schedule 80 section is the part that runs along the B/C firewall. That is the weakest part of the condensate line. Dr Drysdale for his part thought that the obvious place for damage would be downstream of the non-return valve and he puts it between the non-return valve and the tie-in with the MOL. Of course Dr Drysdale is not a general engineer but it can be noticed that the stretch of line in question is of a lower Schedule and has a relatively short span.
From Dr Palmer’s analysis of the bending mode it is probably the case that there would not be enough energy in a fragment of the wall to rupture the pipe in that manner. Professor Reid suggested that Dr Palmer’s analysis in relation to bending is erroneous in that it does not take account of the actual pipe support conditions but then as I have said the implication of Dr Palmer’s evidence is that the mode in question is not likely to have been significant in the circumstances of the cases.
Dr Palmer deals with the mode of indentation and illustrates this at page 3 of number 75/2 of process. He expresses the opinion that this type of indentation is highly localised and he would expect it to be produced by a sharp ended corner of say the angle frames. There are large strains involved in producing this kind of indentation so that allowance has to be made for strain hardening. A finely tuned estimate of this effect could only be achieved by a finite element analysis but an approximate estimate of the effect can be achieved manually. Dr Palmer does a complicated calculation and concluded that the energy needed to produce perforation by indentation is 7.7 kilojoules. For a Schedule 80 pipe the requisite energy would be lower at 3.7 kilojoules. No specific attack on Dr Palmer’s calculation of indentation is developed by the defenders. However there is some challenge of the possibility that in the case of the condensate pipe perforation by indentation is likely although Professor Reid cannot completely rule it out. There is a suggestion that the projectiles being postulated as being potential causes of the indentation might themselves deform on impact thus absorbing some of the energy. Dr Palmer thinks that the frames are probably rather stronger than the walls of the pipe. However he is not claiming that the energy absorbed by the deformation of the angle irons could be left out of account. Thus a question is raised as to the amount of kinetic energy that might be absorbed by the angle irons themselves but there is no quantification of this and Dr Palmer thinks there would be enough residual energy to penetrate the pipe
With regard to cutting Dr Palmer makes use of a Paper "On the Cutting of a Plate by a Wedge" by a Professor Calladine and a Mr Lu. On the basis of this Paper using a yield stress value of 260 (which is about half the yield stress value for a plate of steel) he calculates the value of 29 kilojoules as being the kinetic energy required to cut the pipe. If the one assumes the cutting of one-eighth of the circumference cut instead of one-half the energy needed is according to his calculation reduced to 4.8 kilojoules. These values are for Schedule 120 pipes. For Schedule 80 pipes the values would of course be significantly lower being calculated at 19 kilojoules for a cut of half the circumference and 3.2 kilojoules for the one-eighth cut. He allowed for strain rate effects by multiplying by a factor of two. He categorises the cutting as what one would expect if the pipe was struck by the right-angled edge of the frame. Dr Palmer was criticised by Professor Reid for using the Calladine and Lu Paper. Dr Palmer accepts that he has only used that Paper on one occasion before but asserts that the relevant equation is based on empirical data. He accepted that in using the equation for the one-eighth cut he was using it rather outside its intended range of validity but claimed that the consequential error would be small. He agrees that the idealised situation upon which the Paper was based involved the impact of wedge shapes on the pipe but contended that in practical engineering the practice is to relate idealised solutions to real live situations. Engineers do this all the time and it requires some judgment. Professor Reid on the other hand took the view that the Paper should only be used within the boundaries of its empirical parameters.
Another mode of deformation referred to by Dr Palmer is denting. Denting is much more spread out than indentation. In respect of this matter Dr Palmer makes use of a Paper by Wierzbicki and Suh and he used the relationship between the mechanical properties of the type utilised in their experiments with the dent depth and force required to produce the dent. He got a value for the rupture of a dent of 1 radius of 11.9 kilojoules for the Schedule 120 pipe and 8.4 kilojoules for the Schedule 80 pipe. Applying the multiplier of 2 for strain rates he got values of 16 to 23 kilojoules.
The application of the Wierzbicki and Suh Paper was criticised by Professor Reid but Dr Palmer claimed that in the case of that particular Paper he had compared it with experimental work he himself had done and found that it compares rather well. Professor Reid has claimed that he was unaware of any instance when the formula in the Paper had been applied to a 4-inch pipe. He thought that the formula would have a tendency to produce an under-estimate of the energy under scrutiny but Dr Palmer took the opposite view and thought that in practice the formula was likely to produce an over-estimate. In any event the values produced by applying the formula are about four times less than the energy which might have been available according to Dr Palmer’s calculations.
Professor Reid effected an elaborate series of calculations to test the results obtained when applying the modes outlined by Dr Palmer and the results appear in 96/28 of process. The results he obtains are in relationship to the length of pipe between supports. He does not take issue with Dr Palmer on the amount of force needed for failure of the pipe but disputes if this can ever be reached. Professor Reid claims that the calculations he did are based on Dr Palmer’s models and that according to his results the energy needed to rupture the pipe by the modes Dr Palmer describes would never be reached. However applying Professor Reid’s calculations to the Winfrith tests penetration could never occur but penetration did occur and that anomaly requires to be addressed. Professor Reid explains this by saying that he has used a quasi-static model as had Dr Palmer. However Dr Palmer did not accept that his model was quasi-static and indeed he had made use of dynamic factors such as strain hardening. It is intrinsic to strain rate effect that it occurs under rapid loading. Professor Reid at one point says that he believed that a plugging mechanism was responsible for the failure of the condensate pipe. This is an uncharacteristic remark for him to make because in general he hesitated to come to specific conclusions unless he had backed them up with calculations although at one point in his evidence he does a calculation suggesting that an energy force of 11.7 kilojoules may be enough to have caused failure by plugging in the Winfrith tests. However the point was not developed with him.
The defenders placed emphasis on the submission that in a real engineering situation where the velocities are relatively low one would not be dealing with a single response of the target upon impact but a multiplicity of responses in the form of denting, tearing, and bending so that all these factors have to be taken into account. Not all the energy striking the pipe would be available for bending and some would be absorbed through indentation. It was argued that there could of course also be a multiplicity of missiles striking the target even from different angles. However this could create its own problems because on the analysis of Dr Palmer himself few of the panels would be free to act as projectiles at the correct place and height. The defenders also sought comfort from the fact that after the pressure pulse has passed the missile still retains velocity but cannot accelerate. Thus once a fragment has left the firewall it cannot increase its velocity. It was said that the velocity and distance travelled that Dr Palmer attributes to fragments of the firewall are totally inconsistent with the general evidence given by Dr Bakke about the restricted movement of light panels under the kind of explosive forces being considered. It is also said that the views of Dr Palmer contradict the views of Dr Cox as contained in the latter’s report. However Dr Cox was not really examined about his opinions on the relevant matter.
The defenders also attacked the competence of Dr Palmer to express a view on the effect of projectiles which they claimed involved serious issues of gas dynamics. However this ignores the fact that Dr Palmer had been involved for 25 years with pipelines and the sea. His expertise could be classified as being that of a fluid engineer but the distinction between a fluid engineer and a gas engineer is not always critical. Gas dynamics is concerned with compressible fluids namely gases. Fluid dynamics would encompass liquids but also gases. The main area where the two disciplines vary is in regard to the specific questions associated with compressibility. It seemed obvious to me that as a highly experienced engineer he had practical familiarity with the matters he was dealing with in relation to projectiles. Many of the experts in this case had very wide experience and to an extent it was not always useful to seek to limit them to very narrow areas based on technical definitions of their work in respect of matters that it was perfectly obvious they were well qualified to talk about. Thus for example Professor Stollery’s main expertise was aeronautical engineering but this had given him expertise in gas dynamics. On the other hand there were areas of gas dynamics that Dr Palmer has conceded he was not qualified to opine on but the projectile forces generated on the disintegration of a firewall (if this indeed is solely a question of gas dynamics) did not seem to be among them. Special questions such as the compressibility of gases may be in a different position and Dr Palmer accepted that he would be less well qualified to deal with situations where there was high velocity and compressibility was likely to play a major role. He did not think the matter he was dealing with was such a case. Each expert could bring his own special experience to a problem and there was often overlap. Professor Reid had a lot of experience of impact mechanics but this seems to have developed from his experience as a structural engineer and no one doubted his general competence to deal with the structural behaviour of a firewall. The defenders also argued that Dr Palmer was not in a position to express views on the effect of a missile on the condensate pipe without considering the support conditions affecting the pipe. Professor Reid clearly supported such a view. Dr Palmer did not accept that this was an important criticism. Professor Reid also said that the way in which a pipe is joined can affects its vulnerability to damage. This seems to make sense but certainly the joins in the pipe were not discussed in the evidence. Moreover in general if a precise result was important then I should have thought that the support arrangements for a pipe would be a factor in its vulnerability to damage. Indeed there seems to be no difficulty in Professor Reid’s view that the better a pipe is supported the less it will bend given a particular force and the more prone it will be to indentation. The support arrangements for the actual pipe were in fact available from piping diagrams (see for example numbers 13/56 and 12/96 of process). The distance between supports was 2 to 4 metres. However a lot must depend on the supposed force of the impact and although Professor Reid refers in some detail to the supports for the pipes the matter was not explored with Dr Palmer. Just as Professor Stollery claimed that Dr Palmer was not an expert in gas mechanics Professor Reid claimed that Dr Palmer was not an expert in impact mechanics. As I have already observed (and theses cases have borne it out on numerous occasions) it may not be possible for experts to form any view about certain technical matters well within their general experience without trespassing to a degree on each other’s favoured specialist approach. For example it was never said that Professor Reid was an expert on firewalls but it cannot be said that his views on the structure of these are not relevant. Professor Reid also criticises Dr Palmer’s analysis for being a static analysis. Moreover because a missile may well cause various modes of damage to the target one cannot assume that in the case of a relatively low velocity missile all the energy of the missile will go into one particular mode of damage. Another point of possible importance that was not explored with any of the structural engineers was the possibility of damage to the non-return valve. Dr Drysdale tentatively suggests that this might have explained any delay in the escape of condensate but he was not a structural engineer although he has experience of the kind of circumstance that can fuel a fire.
In the British Gas Paper tests the velocities generated were in the order of 4.4 to 14 metres per second. Professor Reid refuses to relate this fact to the likely velocity of any fragments of the firewall on the ground that he has no reliable information about the latter velocity. He had also shown a reluctance to accept the kinetic energy approach to assess the question of pipe rupture and was again referred to the Paper 110/8 of process where a number of the formulae set out use kinetic energy as a measure for penetration. Nevertheless he insists that kinetic energy formulae are not suited to solving pipe penetration problems in the case of an "arbitrary impact". By that he explains that he means a projectile of a general shape moving in an arbitrary way and striking the pipe at some general orientation.
In relation to the question of liquid fill, Professor Reid is taken to his own Paper "Response of Fluid-backed Metal Beams to Central Impact". He acknowledges that the conclusion of his own Paper is that filling a pipe with fluid has the effect of producing more localised deformation and lowering penetration velocities.
In confronting the whole question of projectiles the defenders were eager to emphasise that the issue they are addressing is not whether a projectile from the firewall could have caused any damage to the condensate pipe but rather whether the sort of explosion that 45 kilograms of hydrocarbon could have generated could have created the projectile energy needed to pierce the condensate pipe.
It should be noted that Professor Reid does not advance the opinion that fracture of the pipe by the postulated explosion is impossible. However he seeks to cast doubt on the reliability of Dr Palmer’s calculations. In any event he would expect bending to predominate and in that situation according to Dr Palmer himself about 48 kilojoules of energy is required. Professor Reid expresses no opinion about the velocity of the projectile but relies on Professor Stollery for evidence about this.
The defenders claim that the analysis employed by Dr Palmer is not fully adequate because it is a static or quasi static analysis rather than a dynamic analysis. The pursuers dispute this and indicated that Dr Palmer in his analysis had taken account of strain load effects. Indeed he doubled certain of his values to take account of this phenomenon. The defenders retort that this in itself shows confusion because it is not clear if the pursuers are relying on strain rate effects or strain hardening. Strain rate effects are the product of a material whereas strain hardening effect is what occurs in a material which has gone beyond yield into the plastic range. However it was said that Dr Palmer had taken no account of inertia which is perhaps the main dynamic effect. Whatever the value of Dr Palmer’s analysis of the projectile potential of the postulated explosion I am left is some doubt as to whether his analysis can be regarded as fully dynamic and indeed I am not certain that he (as distinct from the Counsel employing him as a witness) actually made that claim for it. The question of course is whether the velocities he was suggesting are such as would require a full dynamic analysis and I think that is a complex matter that could trouble engineers. Professor Reid thought that at velocities beyond about 20 metres per second a dynamic analysis is required whereas Dr Palmer considered that you can rise to velocities of 40 metres per second or over without encountering the need for a dynamic analysis. It was generally agreed that at velocities of 100 metres per second a dynamic analysis would be called for.