Tobacco Documents Online

To See DistTibutJ REF AWC/KPN/40A on: ~jJ FRO~ T A.W. Cronsha~ ~, L ......... I " ~'Z~---ll3th December, 1982. • . . @ VISIT REPORT - B&W I attach a copy of my report on my visit to B&W, which may be of interest to you. I must stress that the report is in the form of a historical record of my visit and some parts ol the report are jottings. No attempt has been made to draw the points together or to form conclusions. Comments would be welcome - particularly on some of the questions tha~ were posed. A.W. CRONSHAW Encl. Distribution: Dr. C.I. Ayres Dr. R. Binns Mr. D.E. Conway/ Dr. R. Oldman cc: Dr. M.J. Hardwick BATCo document for Legal Services • Health Canada 18 May 1999

AUTHOR: A.W. CRONSHAW AWC/KPN/40A 30TH NOVEMBER, 1982 • VISIT TO BROWN & WiLLIAMSONl LOUISVILLE 30TH AND 31ST AUGUST, 1982. My contact was Jim Nall. (My original contact - Dr. Matulis - had left the company). He had produced a programme (copy attached). However, there were several changes (additions). The "orientation", with Jim Nall and John Lauterbach covered the hierarchy (copy attached) (Ed Philpot, who was away, is acting as the Department Head of Research Department) and the identification of the activities of the participants (job title). The opportunity was taken to identify any omissions. The obvious omission was Quality Control in the factories; no plans were made to cover this since the Louisville plant closed in July 1982. A second omission was packaging, which is handled by Manufacturin~ (though R~D may be involved?). Also, I asked about specifica=ions and tolerances since without these, we could not decide om what Test Methods are required; Jim will• enquire. We joined a large meeting for "On-line Instrumentation". Filter Rods The Tennessee Eastman device is used, which is based on a displacement pump and assumes that everything that is pumped is transferred to the tow. The final analysis is by wet-and-dry or by the chemical analysis in the laboratory. Circumference, on-line, appeared to be a problem with B&W because of maintenance and cleanliness of the measuring heads. They have trouble with circumference control on non-porous rods. However, I am not sure if the people around the table were really aware of the factory problems. For example, I asked the question about the control of circumference with porous paper and no-one could answer - principally because they believed that there was a problem with non-porous rods. The filter rod makers are principally PM5 and use threaded rolls. They have some KDFI and KDF2 and it was believed that the ADC worked on these machines. The operator takes sample and tests off-line (for circumference and pressure drop). (A Q.C. operator does the tests?) Though there was said to be a tolerance of ±0.05 mm on circumference, filter rods were not rejected (and, anyhow, may go direct to the cigarette makers). Rods may be rejected at the cigarette maker. B&W said that they had problems with tow that can affect circumference e~ as bale runs low, the tow may be withdrawn differently and this variation may not be taken out in the pre-~ension device. B&W do not do denier /2. ~o BATCo document for Legal Services : Health Canada 18 May 1999

-2- checks but asked about methods for the measurement of crimp. They would like to have a device to measure tension between the threaded rolls. They have problems of differing perfor- mance between different machines and, particularly, speed effects. They commented on the bale monitor device to control weight but said that the action of the control loop had to be slowed. B&W control to a pressure drop and asked about control by pressure drop versus control by weigh~. If blooming and denier are correct, then weight should be the control. B&W would like to have an understanding about the action of the plasticiser on the filter rod: how does it at work at the joints and how does it affect the fibre? I mentioned our study. They would like to understand the curing process and how to shorten the curing time. They asked about the use of PVA - I ducked. B&W do not measure hardness of filter rods and asked if anyone did this measurement. Cigarette Making B&W have a mix of cigarette weight controls. Again, the people round the table did not know what use was being made of the device. They were sure the controls were linked to the computer but not what data was being transferred. B&W asked questions about the CIM. success in Imperial, Canada. I mentioned the imck of The Interiorscope, ex GR&DC, is not used. It was seen in one the laboratories. We had a discussion about the use of the Interiorscope and the results. I mentioned coal-retention as an associated measurement. But, coal retention is not a problem in B&W. B&W use plain ecreteur discs on their cigarette makers. B&W asked about on-line monitoring and control of circumference; however, they did not express a real interes:. B&W asked about the measurement of firmness on the cigarette maker. I suggested that this should be done in the Primary. They mentioned the loss of 20% in the cigarette maker but do not understand where the loss occurs. They claimed that they could predict cigarette firmness from measurements of filling value. I did not question the seeming illogicality of these arguments. Primary We bad a general discussion about measurement and control of the primary, including moisture meters, weighing conveyors and filling value. On the latter, we raised the questions of "Do you use the information?" "Is is necessary?" "Can you use it?" /3. N CD Lm4 "-4 qm4 qm4 BATCo document for Legal Services • Health Canada 18 May 1999

-3- PROCESS DEPARTMENT John Jewell (Department Head) had asked to see me. I spent about thirty minutes with him and Paul Newton (the latter had been an additional member of the morning meeting). This discussion ranged over a variety of items including infra-red measurements for off-line and on-line. TOUR OF LABORATORY I was given a very quick tour of the routine analytical laboratories by Les Shockley and of the research analytical laboratories by John Lauterbach. The former had smoking machines that were linked to computers and a computerised system for handling all the samples and the data from receipt of the samples in the laboratory (including such nice touches as printing the labels for the samples). I saw FPIT Serial No. 47 with the electronic counters and the output to the computer. The Borgwaldt Densimeter is used. Two were being operated side- by-side by one person. Conditioned samples are used. The Borgwaldt ends stability tester is used. (The instrument was said to be liked by the Plant Managers because ends stability was a useful indication of quality). The Filtrona Tape Circum- ference Gauge is available but is not used because the cigar- ette is deformed by the tape. B&W have tested the plastic tape version but got high results (believed to be because the tape did not close onto the cigarette at the overlap point of the tape); Filtrona will supply a heavier load (but will this defeat the object of the plastic tape and cause the paper to be deformed?). The Filtrona Automatic Ventilation Meter is used and B&W had just completed tests on the latest version, which includes pressure drop vents ope, and vents closed. An individual cigarette weigher and classifier (five zones) made by Diversified Engineering was in use. I asked questions about the use of weight-sorted cigarettes but I was obviously talking to the wrong person. The cigarettes are used for tests on cigarettes of different weight for, say, Process Research or Marketing. Yet, in the morning discussion, one person had said that a reason for a requirement for on-line monitorin~ of firmness at the cigarette maker is that a light cigarette may have acceptable firmness - so a weight selection process could produce a biased sample. CD /4 Q_m BATCo document for Legal Services • Health Canada 18 May 1999

-4- REAL TIME/OFF LINE INSTRUMENTATION A Neotech (Model 51) infra-red analyser is available in the laboratory. It is equipped to look at up to 720 lines. The results are sent to the statisticians who run the results on a larEe computer in Head Office to find the important lines. The equipment does not work for menthol. It is used for tobacco (before casing, etc are added). B&W feel that there is a need for an instrument that will indicate nicotine at the GLT plant - at the input or at the output. At the filter rod maker in the fact6ries, the Filtrona Manual " Tape Circumference Gauge and the Celanese pressure drop tester are used. The instruments are used by both the operator and the Q.C. person. In the pilot plant, they have a Filtrona Automatic Circumference Gauge and a Filtrona Automatic Pressure Drop Tester. Both are scaled in millimetres but the Pilot Plant want the pressure drop in inches and are arranging for the conversion to be done. The factory instruments are cross-checked with the instruments in the laboratory in the plant, which are, in turn, cross- checked with the instruments in R&D. Plasticiser is measured by the wet and dry method by Q.C. Some samples are sent to the laboratory for chemical analysis of plasticiser con~ent. The filter rods are not weiEhed by the operator; Quality Control weigh the rods - after the fact. At the cigarette making stage, the operator does the weight checks. Quality Control take samples and test for ventilation, unbound pressure drop and circumference. There was some uncertainty about how the ventilation was measured; the BAT Ventilation Tester was mentioned. There was a long discussion on the measurement of cigarette firmness on the Firmness Profile Integrating Tester. Is the FPIT stable (because of motor speed and wear of the cam). How could the testers in B~W be upgraded (because of wear of the contacts)? I suggested that the test units be sent to the U.K. - but pointed out that we have made improvements to the feed unit. (B&W do not want the feed unit to be changed because their electronic modifications, including the fitment of electronic counters and connections to the external computer, are made in the feed uni~). Firmness is very important to B&W because i% chanEe in firmness and, hence, weight represents many millions of dollars. m._ QJ~ BATCo document for Legal Services • Health Canada 18 May 1999

-5- TOUR OF PILOT PLANT I was taken to the pilot plant by H.G.(Butch) Bryant who accompanied me during ~he tour. My guide was Bruce Bandy. The size of a sample operation is 500 pounds of domestic tobacco with add-ons (WTS, Oriental etc) to 700 - 800 pounds. A Vacudyne is used for conditioning. The Buriey is passed over a weigh conveyor, which is used to control the rate of casing application in the casing cylinder. There are three cylinders one above the o~her for this operation; the tobacco is fed to the top cylinder and drops, by gravity, from cylinder to cylinder. For some cased blends, the cased lamina goes to a Proctor redryer, which dries the lamina down to 6~ and re-orders up to 18~. The Burley then receives the top dressing. The pilot plant has a casin~ preparation room. The strip is screened (screen size not known - possibly 6 mesh) and the small strip is added back after cutting the remainder. Between 2~ and 8% is removed with an average of 4-5~. This strip by-pass process is not used in Macon but will be installed. GLT screen approximately 29 of their throughput, which by-passes the cutters; this process is in use. The tobacco is stored overnight before cutting. About three to four operations are run each day. The tobacco is cut on a modified (two knives) Hauni cutter and dried in a high humidity air dryer. CRS can be added either before or after the dryer. The former arrangement is used in Petersburg and the latter in Macon. The maximum temperature achieved in the electrical beater for the dryer is 700"F. The tobacco temperzture at the inlet to the cooler is 200"F and, a~ the exit, is lO0"F (but is dependent on room conditions). B&W will try a process in which the tobacco is overdried and, ~hen, re-ordered. The tobacco goes to a blending bin to even out the moisture and to blend the tobacco. Samples are taken and the chemical composition is checked. A defined weigh~ of tobacco is taken and put into a cubic box that can rotate across diagonally opposite corners. Here, the flavour, metho!, etc are added. Expanded tobacco is added and, if WTS dried to 14~ is added after the dryer stage, it is added here. A stem process is available in the Pilot Plant, including the process to remove nitrate from the Burley stem; however, this process is not run because it is time- and labour-consuming. Instead WTS is obtained from the plants. /6. rkD *...j BATCo document for Legal Services • Health Canada 18 May 1999

-6- The surplus tobacco was sent to Louisville factory to be blended in. The tobacco is now a problem as it has to be shipped to Petersburg or ~acon. In the filter rod area, there is one KDF2 with AF2 and a bale weigher. The plasticiser monitor was not working. There are three old filter rod makers and one homemade. In the ciEarette making area, there are five Garant 4 makers(one with Max 15 and one with Max S - just received from the factory, well worn). There are five Molins 6-8-6 with Max 3 attachments. All machines have hand catchers. The Max S was the one on which the 50 watt laser had been tested. There had been many troubles at installation; after that, the trial had run satis- factorily. With a mean ventilation.of 20-30%, the standard deviation was 2.5~. The I00 watt laser trial has been completed in Petersburg. A Technical Services laboratory is in the Pilot Plant for various tests. A home-made, paper permeability tester is used; a Phobos instrument is available. A Canadian shaker and a home- made, vibrating compression tester were seen. I asked about instrumentation in the plant. It is, virtually, nil except for the process control mentioned earlier. A moisture meter is not used on lamina because of the wide variatioas between the blends. A moisture meter is used on WTS, which is more uniform, but the Pilo~ Plant do sot process stem. Sample moisture meters, such as Plessey and Kappa, are used for those blends for which the meters have bees calibrated. Instrumentation research is not done in the Pilot Plant but process tests are done. The main function is sample manufacture. PAUL NEWTON Paul Newton had asked for a meeting with me. (He had been added to one session yesterday and there had been the added session, yesterday, with John Jewell. The topic raised by Paul was the GLT plant. B&W have dome a two-year investigation of the plant. (Details were not ~iven and I did not ask). The small strip process has been in use for three years. The problem is to set the specification e~ the particle size, stem in lamina and stem particle size for WTS. Some of these are less important mow with high-speed cigarette makers eg the long-strand tobacco is destroyed in the maker; however, they are important as a measure of the yield of the GLT plant. How do we set up the process? What instrumentation is required: moisture meter and chemical analysis. B&W have a problem with the wide range in moisture /7. 0 BATCo document for Legal Services • Health Canada 18 May 1999

-7- content, which can be 6-8: with Burley in December. What does the manufacturing plant want from the GLT? Very brief meeting. Pmul suddenly terminated the meeting because he bad to go to ano=her meeting. DATA USE Paul Chen and Wallace Crain were the statisticians. Bryant was present briefly. Jim Nall was present. Butch Wallace Crain gave a lecture on the analysis of results with x and R charts, CUSUM charts and moving averages. Slides were used throughout and he said that he gave the lecture to the Tobacco Chemist's Conference two years ago (and, he said that he was surprised that no-one had given a similar talk before then). Routine stuff. B&D are applyiag this to their own work and intend to apply the same principles in the factory; however, they recognise the problems involved in the trainisg of people to understand and use the techniques. The factory has four makers in one group with a filter rod maker. Each group has a circumference gauge, pressure drop tester and a balance. The operators use the instruments but their results are not recorded. Quality Control go from group to group, take samples and record the results. The data is entered into the computer about three hours later so it is historical and is not used by the operator. We had a discussion on the results that should be recorded and put on the computer. Is historical data of any use? However, if we record the operator's results, will there be a reaction from the operator that the computer is a management spy? I described our Quality Control Test Station and its uses. I think the B&W people were surprised that we had advanced so far. Wallace described the use of the smoking results to derive the weekly averages of tar (weighted by market share of the samples that were smoked); the weekly averages were plotted with warning limits and action limits to detec~ trends, etc. From this chart and knowing the age distribution of the cigarettes on the market, it would be possible to predict the results obtained by the Government (FTC). Jim Nall had, obviously, done some homework and had looked up the specification books. Only the ~arget is specified - no tolerances. He said that they had limits in the Branches but these did not appear to be defined. /8. CD ~0 CO BATCo document for Legal Services • Health Canada 18 May 1999

-8- EXPANDED TOBACCO This was a sparcely attended meeting because the engineers had been called to a meeting by John Jewell. Those attending were Spivey, Lauterbach, Nall. B~W use liquid nitrogen in the flotation method to separate expanded tobacco. The method is not sensitive to tobacco moisture content and has the additional advantage that the tobacco is not affected and can be used for subsequent chemical analysis. B&W mentioned the Philip Morris tea-ball test. The tobacco is put in the tea-ball. B&W asked if anyone measured the strength of the ash for ash retention work. I mentioned the filling value/air impedance tester for the measurement of the expansion of DIET - and described the problem with CKS. _WRAP UP / M I SCELLANEOUS Another sparcely attended meeting and was, effectively, a continuation of the previous meeting. Butch Bryant joined the meeting to say that he had received the samples of moulded filters from John Luke and had run some small samples of the plastic paper. I asked one or two questions to fill the gaps; however, these questions were incomplete as I bad not had the time to write-up all my notes. B&W intend to test the Infra-red Engineering SM4, on-line, for the measurement of moisture only; they do not plan to test for chemical constituents. The Firmness Profile Integrating Tester was raised again. This time, it was testing it for long-term stability. I pointed out the difficulties. In any case, the comparison was usually with the competitors' cigarettes. Nevertheless, B&W said that they would like to know how their cigarettes had changed in the year. At this point, I gave up. ~owever, B&W feel that they may have a long-term problem with wear and drift because they use their machines heavily. Nevertheless, I feel that it is a mechanical check tha~ is required on the FPIT. An alternative is to have a "standard" instrument %bat is little used for comparison with the working instr%u~enz. mO /9. BATCo document for Legal Services • Health Canada 18 May 1999

-9- The problem of :he measurement of the firmness of distorted cigarettes removed from packets was raised. I said that we had so experience but I would expect the distortion effects to be a[eraged over t~e sample. A general question was raised on the purpose of the firmness measurements; this was not answered. I was asked if there was an alternative method to the method, used by B&W, in which tobacco is blown from the cigarette when they wish to de~ermine non-tobacco weight (because of the problem of tobacco trapped in the seam). I said "None". We briefly discussed on-line permeability after I had raised the topic. B~W want to know how we will pass on the information ie by blue prin~s or by making instruments? I said that we had to decide but it could be either or both. The question of how to characterise the physical properties of the incoming material by grade was raised and mention was made of the Canadian work. I said that I had little to add and suggested that B~W contact Canada. I added that shattered tobacco could be used in cigarette making and so eliminate the cutting process. B&W asked what moisture measuring devices were there? I mentioned solvent extraction, various oven procedures and the Beaudesson. The latter raised a lot of interest. I asked about coal fall-out measurements. They are done in the Branches for their own production and by E&D for competitor's cigarettes and for samples from the Pilot Plant. I was asked if anyone else did coal fall-out. I said that some did, occasionally. Hot collapse was raised and I mentioned the various instruments. PAUL NEWTON In a subsequent telephone conversation with Paul Newton, I said that, since he was not present at the final meeting, I would like to have his views os the Suture in Quality Control. His comments were as follows. Chemistry of leaf on-line: the infrared method is where om-line moisture was ten years ago. B&W are very interested in our work programme (see my separate note)." Control philosophy in Secondary: what is the driving force? Is is weight? Should it be firmness (but not, necessarily, measured on-line)? I r~o BATCo document for Legal Services • Health Canada 18 May 1999