AUTHOR: A. W. CRONSUAW 5TH OCTOBER, 1982
AWC/KPN/40A
VISIT TO CANADA
16TH AUGUST TO 27TI] AUGUST, 1982
MONTREAL - INTRODUCTION
On my first day in Montreal, Jay Perry (my contact) had arranged
& meeting. Those attending are given in the attached list
(loth August, 1982). Jay Perry gave an introduction on the
purpose of my visit; this summarised the information that had
been circulated before the meeting. A copy of the proposed
programme for my visit aud this supporting information is
attached.
I was asked to give a review of "GR&DI-Is recent test instrument
development activities". This was based on the review that I
gave at the Technical Exchange Meeting in June 1982. There were
only a few questions because later discussion periods were
scheduled.
Also, I amplified the purpose of my visit - an examination of
methods and of the need for now methods. I emphasised the need
to know the characteristics of the cigarette that were important
to the customer.
HEAD OFFICE - CENTRAL MANUFACTUILING
The meeting was with Mrs. Denise Henry aud Don Schroeder (late
morning and all afternoon).
Imperial are moving to the ventilation of existing brands with
the aim of reduced delivery without a change of taste. Ventilated
brands have about 25% volume-----ne big sellers are not ventilated
and there is a reluctance to ventilate the brands whilst they are
selling well. No new brands have been launched for about two
years. -There is an agreement with the Government to have a
maximum delivery of 12 mg.
There is a move from Canadian flavour to U.S. flLvour.
There is a move to cheaper prices. There is little that can
be done with the cigarettes. Ideas are:- cheaper pack-aging,
give 25 cigarettes for the price of 20 cigarettes (tax is by
the cigarette so the extra 5 cigarettes would be given LS &
premium) or even put 9 cigarettes in a pack. All these raise
problems with manufacturing I@ecause of the machinery changes
that would be required.
12
CD
co
BATCo document for Province of British Columbia 16 April 1999

-2-
There is a move to make Production responsible for Quality
Control and to get rid of inspectors. The new system has been
used in filter rod manufacture in Quebec factory for two years
and is just starting at Guelph. It is not in use in Montreal,
which is equipped with Mark 6 makers and hand-catching is used
- so the operators do not have free time. (KDF2 machines will
be installed). The argument is that the operator has about
40 seconds to spare between tray-operations. The operators
have responded well: they took it as a vote of confidence.
Similarly, in cigarette manufacture, the operators take cigar-
ettes for group weight tests.
I expressed my doubts on this - usually, Production and Quality
Control had to be separate to &void conflict. What about
cheating by the operators?
Because of this philosophy and the move to ventilated cigarettes,
Imperial would like to have ventilation-only testers for use by
the operator. Imperial have one of our testers and have tested
it in Quebec for the past four months. There is a concern about
the frequency of adjustment of the instrument but it is possible
that the Production people are making unnecessary adjustments.
See below. (When we started this discussion, Imperial had not
seen the Filtrona literature on the two BAT Ventilation Testers;
however, the literature dropped into Don's in-tray during the
discussion and we examined It.)
Imperial want a Ventilation-only meter with print-out. I
explained why we had not fitted it - because the instrument was
to act as a guide to the operator and was backed up by Q. C.
with the BAT Ventilation Tester.
Imperial want a print-out because the operator samples would be
the only samples and there would not be a back-up from Q. C.
Imperial raised the question of whether instruments designed
for the laboratory are suitable for factory use. Well maintained
(absence of maintenance?). No need to adjust.
The Imperial Ventilation Tester was not suitable for use in the
factory because of its sensitivity to vibration. Imperial have
done a four-month trial an the BAT Ventilation Tester in the
plant. They had found that the factory people did far too many
adjustments and, as a result, they had asked for records of the
drift if calibration adjustments were not made; we examined
these records and confirmed that the drift Was within acceptable
limits. Imperial will continue their examination.
/3.
Q4
BATCo document for Province of British Columbia 16 April 1999

Imperial commented on the price from Filtrcna of the BAT
Ventilation Tester (approx-05VO + pump) wherc&s the Ventila-
tion only version is a reasonable price (approx E1550). It
looks as though Filtrona are trying to price the former out-
of-the-market so that customers will buy the Filtrona AVM -
I found that Denise Henry and Don Schroeder had not seen the
circular letter on the Ventilation Tester. So, I arranged for
a TELEX to be sent to D. H. Mummery.
Problems - such as the effect of hot melt on filter rods and,
thence, bleeding, which affects tip ventilation - require
instruments.
A monitor of filling value in the process would be useful
because WTS is suspect as a process and has not been fully
accepted; hence, full advantage is not being taken of the gain
in filling value.
A monitor of ends stability could be useful. Imperial suffered
from bad ends stability (about two years ago?). As a result,
all the factories were equipped with Turbula Schatz testers.
However, the problem was traced to the leaf and the problem
disappeared with the change in crop. The ends stability test
has been discontinued in the factory.
A rapid test for moisture is wanted. The Bemudesson oven has
been tested and found to be very good for cut tobacco only.
The oven is in use in Montreal. Further ovens will be bought
for the other plants. The oven will be used as a back-up for
the Quadrabeam. If the Quadrabeam fails. the philosophy is to
plug in & spare, calibrated unit (only possible in the larger
plant) or to use the Beaudesson.
There was a discussion on other methods for the measurement of
moisture content.
OPERATION TESTING
In discussion, Mrs. Henry mentioned a special, intensive,
monitoring test once per month in the primary and secondary to
test the final product in each factory and hence to compare
factories, which use different equipment. I expressed interest,
so, Pierre Leblond was brought in.
In each test, an operation of CRS and an operation of lamina
will be examined. The following measurements will be made:
Abb
NJ
C)
C=)
BATCo document for Province of British Columbia 16 April 1999

-4-
Bale temper&ture(because of the cold weather)
Continuous conditioner
Conditioning cylinder water addition
temperature
shaker test
cuts per inch
Cigarettes filling value
ends
nicotinelsugars
I noticed that t@e drying process was not included. So Pierre
added tobacco temperature ex-dryer - but not process air flow
(because temperature and air flow are linked: see page 17).
The work is for long term interest - not for control. The
standard conditions will be used: there will be no changes.
Two main recipes will be chosen in each plant: e.g., a standard
CRS and a low-delivery CRS.
The reason for the proposal is that existing results are mon-
thly averages; in grder lo gxgming the process, we want to
look at particular recipes/operations. However, the work has
not sf-X-r1-e_d_ViF`caus5-no one is available.
MISCELLANEOUS
Imperial have one Hauni tunnel - which they do not like because
the holes become blocked. They will fit cylinders.
Ends stabil
_Lly was & problem in Guelph plant. The problem was
traced to the continuous conditioners. A lot of work was done
on the units but the solution was found in external changes:
improve the uniformity nf dengity of tobacco in the case and
warm the tot;'a-cco 6'efore it enters the primary process. One
improvement in packing density was obtained by turning the
case upside down immediately after the case was filled and
storing the case upside down; this facility was seen at Aylmer.
The problem is much reduced by the use of the Mark 9 cigarette
J;
maker. 4.A* r e
However, Pierre Leblond said that, in his opinion, the
Dickenson conditioner could never be as good as the CTCM.
Because the probe went between the layers, the distribution of
moisture across the face could not be uniform since air has to
flow through the layers; in the CTCM, the air flows between
the layers. Further, the doff ers operate across the laminj_.6,,6;@,
and must break the longer lamina particles.
Imperial have looked at the Rothmans TOP (Tobacco Opening
Process) machine (with Eric Dennis). He will let me have
details.
/5.
C)
Xb.
__4
CD
BATCo document for Province of British Columbia 16 April 1999

I was asked about the use of the Borgwaldt for filling value.
I said that I could recommend.it. However, Pierre said he bLdJJ
read a report, by Roger Hedge, in which the Borgwaldt was not
recommended. #6
Imperial face the problem that they will purchase DIET tobacco
from their competitor (Philip Morris). How can they monitor
the quality? I recommended the Filling Value/Air Impedance
tester.
QUALITY CONTROL IN I.T.L. PLANTS
.PHILOSOPHY OF IMPERIAL TOBACCO
About three years ago, there were major changes in the Senior
Management. Now, Imperial have a working President and exe-
cutive Vice Presidents. The Organisation chart for the
company is attached.
About two years ago, a Business Plan identified the Company
Objectives in each division. For example, productivity, im-
provement in the utilisation of material, quality. The Company
started to use the term "Product Integrity" - and will not
due-t--thuT-Tv-o-u-t-s-i-Ee--Ege--s-p-e-ci-T3-cation. Because of
sell a pro
this, more errors are reported. The errors can apply to any
item e.g. incorrect blend, incorrect packing material. Thus,
if the cigarette has been made with the wrong paper, and the
fault is found by the factory, a check is made to find it the
error is gross-, if so, destroy th @@ i-s
igarettes. If the error
small,--tszrc@-wbu,.I(J-D-e-eoiisultitions with the relevant depart-
ment and the ELppropriate decision taken. If the error is
found by the central laboratory, it may be too late to take
action, since the goods would be in the shops, and no action
would be taken (because of the resulting bad publicity) unless
the error was a dangerous one.
Many specifications are "Not worse than the competitor's
product".
The other constraint is an industry agreement that the five
month running average for tar should be within :to. 5 mg and
for nicotine within :tO. 05 mg of declared value. The samples
&re generated in the plant (see Inspection Standards and Class-
ifications) and are used for several tests. There is a requir-
ement, made within Imperial, to bring the CO down until it is
equal to the tar.
The work in the Head X!atory will be discussed, later,
with Bruce Stirl-in-g-L y; the first is responsible
/6.
C)
`%4
BATCo document for Province of British Columbia 16 April 1999

-6-
for the measurements; the latter is responsible for reporting
the results.
Ends stability is reported weekly to the factory. The other
results are reported monthly and are shown on graphs that cover
a five-year period. A separate graph shows the five-month
moving average. The results also showed the results obtained
on competitor's brands - and showed deviations from the de-
clared values. I. asked what happened if Imperial disagreed
with a competitor. On one occasion, this happened: the com-
petitor found a disagreement on a new Imperial brand so the
competitor wrote to Imperial. An inter-company standard cig-
arette "MONITOR" is used. The cigarette is provided in rota-
tion by the various companies.
The Cigarette Specification is issued by Product Development
and Specification (Audrey Bishop). It includes every possible
detail of every component. Then, Central Manufacturing (Don
Schroeder) prepares the Material Specification from the cig-
arette specification; the Material Specification contains all
the information required by the factory to produce the cig-
arette. All materials have a commodity code (which is marked
on the material) to avoid mix-ups because of the similarities
of some materials.
There is a detailed specification for all components. Some is
based on information received from the supplier. It Imperial
find deviations from the specification, in practice, they
discuss them with their supplier.
When a change is made in the specification, a note is sent out
to all holders of the specifications. The first samples are
monitored.. Then. the change is entered into the book.
For a new brand in a particular factory, say, & brand that is
already made in another factory, the first five days manufacture
are checked in H.O., the laboratory and marketing.
It is recogaised that the Specification is based on information
provided by the supplier. If deviations are found, and a.re
accepted, the specification is changed. Similarly, for a brand,
the specification may be changed to meet what is achieved in
practice.
All the specifications are held on Word Processors so that the
implementation of changes is simplified.
('Note. Only one highly-porous plugwrap (26000 CORESTA) is used
for all tip-ventilated brands and only one plugwrap for all non-
ventilated brands. The filter rod makers run at 400 m min-'.
If there is trouble, the speed may be dropped to 350 m mia-1 -
/7.
CD
ZIM
__j
(04
r@j
CD
BATCo document for Province of British Columbia 16 April 1999

-7-
but not lower. The tow-oppnere are being changed from E10 to
AF2. There are some threaded rolls on the old machines.)
The Quality Control in the plant is described in the attached
note prepared by Imperial Tobacco.
The question of Q.C. of mean weight was raised. (This procedure
is not in the note.). In Guelph and Quebec, where all the
machines (Mark 9 and 6-8-6) are equipped with either tray
fillers or OSCARS, and the mean weight is controlled by AccuRay
C1700 or 700, the operators take three or four samples per
shift, weigh the samples on a Shadowgraph located near the ma-
chine and complete a control chart lbcated on the machine.
-The samples are one hundred cigarettes taken as a grab sample.
Standard deviation is 0.35; action limits are +1.1g. The
adjustment of the mean weight control is by the AcQuRay eng-
ineer (who checks that the adjustment is necessary). In the
Montreal plant, equipped with Accufty C4, the standard devia-
tion is larger.
With the AccuRay 7000, there is very little need for check-
weighing, so what checks should be done? I said that it de-
pends on the effect of errors.
The group weight charts are used only in making - and are not
used elsewhere, for example, Accounting. Is this satisfactory?
Should there be an ultimate check on the weight of tobacco? I
pointed out that Analytical Services derive density values -
but this was not used, either.
Plain cigarettes are, still, made but the proportion is declin-
ing.
Mrs. Henry described the Production Organisation in Head Office
and in the factories (attached - page 33). The Quality Ser-
vices in the fLtories was given (attached - page 34).
For moisture tests, four samples are taken in each operation
(ie each hour) from some parts of the process (the lamina
stream) and two samples from the CRS line. The mean and range
are plotted and the results are circulated.
For shaker tests, samples are taken once per day at certain
points - smalls, CRS, cooler, cigarette making; for ecretage
level, one sample each week; for strand width, one sample each
week. No ends stability tests are done in the factory. The
ecretage level is measured with a vacuum cleaner. The test is
done mainly in Quebec. This testing will_b duced with Ike I
_Ma4_ 9. Also, the Micro 7o0b-VTT1-7rvet0e on
of ividual weight.
C)
%D
LP4
r@j
CD
BATCo document for Province of British Columbia 16 April 1999

The inspection system looks at 30,000 to 50,000 cigarettes each
year (depends on the factory)-- In Montreal, the inspection is
done at the cigarette maker and there is direct feedback; in
the other factories, the samples are taken at the packer (on
both shifts) and inspected remotely (on one shift only - so
second shift samples are examined on the following day). The
inspection is, therefore, an audit at a central location.
I
There was concern about the efficiency of inspection - especially
with OSCARS and tray filling; also, the operator does act as a
material feeder rather than as & machine minder. Tn Montreal, it
is said that the operator tends to rely on the Quality Services
to provide the inspection that he should do.
Approximately 70% of customer complaints were covered by three
faults: missing (or detached) filter, detective tipping and
damaged.
A large improvement has been achieved by improved control of
the circumference of the filter rod.
Though factory audit rovement in product quality,
t6 hich Is said
to be because:
- & carton is given to each complainant
- the price of cigarettes has increased
- reaction against smoking
Only the important fault categories from the factory audit are
reported to Read Office (Vice President); these are: open
laps, detached filters, defective tipping. leakers, packages
with damaged cigarettes (torn, badly buckled - i.e. non smokeable
cigarettes). The less important are plotted only. The results
are shown as a single value for each month.
The maintenance of common visual standards between plants is
done by bringing people together twice each year.
The Primary Process Engineer does not control the primary but
helps the primary in the solution of problems, which he does by
exmialug data, contacts with Head office, and installation of
Dew equipment. - ere is a real specification with control
limits for the p: s is a joo Tor Day Fe
COMPUTER
There is a five-year programme to introduce computers Into
imperial Tobacco. Two years ago, consultants were called in.
Imperial will probably buy a bigger computer - perhaps with
peripheral computers. All three divisions had to form a three-
C)
BATCo document for Province of British Columbia 16 April 1999

year plan. The sub-committee for the Production Division
decided to do production measurements first: production quantity
(cigarettes produced), production quality, production costs.
Production planning will come later.
Production quality is the first priority. The audit information
will be entered manually. It will give the information that
will identify the problem area - by machine, by operator - more
quickly with a report to the Plant Manager and to Read Office.
Also first priority is the filter rod Suality information on
pressure dro-p-,--cyr--cu-m--f-er--en-c-e--a-n-Cwe-icht.
The work will start in Quebec this week and will take about
eight months.
The work on Production Quantity has already started. The data
will be entered manually (not from the 7000 MICRO - using the
philosophy "walk before run").
The second priority items are:
Making and packing cigarette weight
circumference
pressure drop
ventilation
(eirnrette ends stability)
Primary Process mois re 97
partiL size Jones 4n6s
Consumer Complaints requir -n interface
with U, -ting fo@ lalysis of
the s-- ,- of com; nt eg
i categ of person area, etc.
The third priority is:
Material quality.
There will also be a project to integrate specification, material
change, advice notice --A pachin, 12 7.teet . This project
links the three divisi ,oncei 1.
I asked what is the VLlUt the ograzffa- to install the
computer? It has not been anti -@-d - moi-@ qualified. Imperial
hope to be able to identi par ular maci, nes, particular
Materials that are & sour. 04 iroblem an:,,, therefore, identify
nd production
a fault - with a conseque iinp iement in @,uality a
efficiency.
Mrs. Henry asked if we h examij d the AccuRay CIM. Canada
have examined it; they I. d that "-hey-could not set the rej6etion
definition
limits nor could Accult help. This related to a
L4
BATCo document for Province of British Columbia 16 April 1999

_10-
of a fault). So, now, AccuRay are producing a Mark 2 version.
Mr. Yves Primeau came in and asked some question's about details
in the ANTIC specification. (1 was intrigued to find that this
document is-completely anonymous). Yves went through all the
-items to be measured or monitored. I was able to answer some
questions but could not answer some others (because I had not
seen the document before). I was also asked & question by
Mrs. Henry that I could not answer: "'Who will evaluate ANTIC'?
DR. P. DUNN
On Wednesday, 18th August, 1982, I had a brief meeting with
Pat Dunn - mainly to meet the new V. P. He expressed the concern
ielt about the lack of cooperation from the other two partic-
ipants (Germany and B&W) in the Q. C. Project.
Dr. Dunn said that a Product Integrity Group (PIG) was initiated
about one to two years ago. However, it "never took off". Now,
we have Quality Integrity. it involves Production, Marketing
and R&D. The question is 'Who should police this?"
The new Organisation chart for R4D is attached.
RhD PROCESS DEVELOPMENT
I joined a meeting with the people given in the attached list
(18th August, 1982). Note. Ga. eta Levert is now in Process
Development. fie is involved in such activities as installation
and testing of the new weigh conveyors (K-tron and Weycon) in
Guelph plant. His role in Instruments has been taken by Gilles
Prefontaine.
Wilf Gray asked about the effect of moisture and temperature
on the cigarette making process. I talked for some time on
this. A question was asked: "is Germany working on this?"
Bob Rice asked if ends stability was affected by strand width
I mentioned the work we did some years ago on the effect of cu t
width but there was no work on ends stability.
Bob Rice described the work that they had done on the differ-
ence in ends stability between leading and trailing ends. The
also existed when the cigarette rods were reversed
difference
and reassembled. Also, when rods were cut and reassembled.
Peter Benn talked about the work where the imperial Tobacco
Impedance instrument Was used. The test parameters, e.g.,
weight of sample, bad been examined (use constant weight).
Can the instrument be used for process control? Samples had
been taken at various points in the process and tested both
direct and after conditioning. The results are attached.
/11.
Ah
-4
BATCo document for Province of British Columbia 16 April 1999

The problems of sampling were discussed - particularly the
avoidance of segregation; this.applies to both on-line and
off-line instrumentation.
The use of the flotation method for the measurement of the
inclusion of expanded tobacco was discussed at length.
Imperial bad obtained low results for tobacco from the hopper
but correct values for tobacco from Cigarettes. Yet, tobacco
mixed by hand gave the correct values. I suggested that
separ&tion/segragation had occurred.
The filling value/air impedance was described and I gave some
results - including the strange results on CRS. I recommended
the instrument to meet Imperial's requirement for a monitor for
DIET tobacco.
During my visit, I looked at the Imperial Tobacco Impedance
Gauge. It works from compressed air. Imperial have found that
the readings drift and they wait 15 - 20 seconds before the
reading is noted. They are puzzled by this. I commented that
the tobacco might compact under the action of the air (but
this is unlikely to be as obvious ms in our work) but the more
likely reason is a loss in moisture content because of the use
of factory air. Imperial Will Check if this is occurring.
AJAX
I had a very brief visit to the AJAX plant - made brief by a
long lunch with C. Brown (the Plant manager) Peter Bean (recently
transferred - 3 weeks ago - to AJAX) and Bob Rice.
I had a quick tour of the plant and a tour of the laboratory
facilities with James Sinclair.
The tests &re:-
Moisture tests by Freas oven
Sand tests
solids on the binder and the slurry by drying at
140'C for 45 minutes
Flour particle size by screening - usually 270 US mesh
Net thickness gauge
Energy reduction by computer simulation. (On the dryer,
this shoved that eounter-curreat was the most efficient).
Fibre length
Moisture after dryer is controlled through a TI 550 by varying
the thickness (within certain limits).
Viscosity - on line
Thickness - off line
Breakload - Instron
Flex test - M. I T. folding endurance tester (relates to
what is wanted in the cigarette).
112.
C)
r%j
CD
00
BATCo document for Province of British Columbia 16 April 1999

-12-
In the above tests, it is important to cut a clean edge.
Apha-cellulc@se is now added io' customer's stem.
There is pressure from customers (incuding competitors) to use
100% of the material. This is difficult because of the high sand
cc;ntent of the-fine fractions. In Imperial, AJAX is about 6%
of the blend.
i The Quadrabeam is used to measure the output moisture content
of the sheet, wh'ich is overdried and, then, re-ordered.
A Honeywell computer is used for the inventory for the several
customers. The computer also has an interface with the T. 1.
Controller and records several items.
R&D ANALYTICAL SERVICES AND ELECTRONICS/INSTRUMENTS
A small meeting (see list 19th August, 1982).
The requirement is to link instruments to the computer. Thus,
the Filtrona automatic circumference gauge and Filtrona pressure
drop testers will be linked. Imperial have two Firmness Profile
Integrating Testers (one 1980-type and one much older) and would
like to link those. (Later, I was shown the modification to the
cigarette pusher to overcome the friction problem with the
existing slider system).
Other tests are:
Centrifuge for filling value of single grades
Ends stability by Turbula Schatz
Porosity by Wiggins Teape tester for cigarette
paper using the 2.5 cm2 test head. The high
flow section is not used so the tester is
used unmodified. A Filtrota tester is used
for higher permeability. The Phobos is used
for the highest permeability but using the
Filtrona test head for perforated tipping(?)
where the Phobos head cannot be used. The
Phobos head Is used for some plug wraps
(small test area to obtain highest permeability).
The Phobos tester is linked to an UP97S.
Ventilation Tester - the BAT Ventilation Tester
is more precise and repeatable than the imperial
version - particularly below 15% ventilation.
Use only one head to measure tip ventilation
only (overlap leaks not included).
I was asked about my work on puff profile and puff volume.
Imperial would like to have the equipment.
CD
BATCo document for Province of British Columbia 16 April 1999

-13-
Imperial want a device to measure paper permeability on the
cigarette. I recommended the BAT Ventilation Tester.
In'the tour of the laboratory, I was shown the Phipps and Baird
smoking machine modified by the fitment of the Filtrona cigarette
lighter - more reliable and cheaper. The machines are fitted
with the Filtrona CO meter. The smoking machine is connected
to a computer for puff number (which counts each cycle, 6OHz,
and, therefore, almost fully occupies the computer). The
balance is also connected for weighing the smoke condensate.
The Mettler balance can have two-way communication and messages
can be put on the display. Nicotine and water determinations
are also linked to the computer.
The conditioning cabinets-for cigarettes and tobacco use air
sucked from the conditioned laboratory. The air is measured
and controlled at the entry to the cabinets. The cigarettes
are placed in holders (about six cigarettes deep) except for
a few that are in single layers.
In the earlier meeting, I have a long talk about the conditioning
of cigarettes because Imperial bad spoken about drift in the
firMusee-rt-j"44.s._and wanted to blame the PPIT. From what I
have seen, the Imperial conditioning practice is poor in sever
respects.
R&D TECHNICAL SERVICES
I had a discussion with Louis Bouchard. Technical Services
are not responsible for Quality Control of materials except for
cigarette paper, highly-porous plugwrap and perforated tipping.
The suppliers send samples to Technical Services. The pallets of
materials are sent to the factories but are not accepted for
use until Technical Services issue a clearance note. Three per
cent of the material is checked for sixteen characteristics.
The work of Technical Services is:-
Trouble shooting
Packaging (P.D. materials testing)
Homogeneous leaf for cigars
Filters
Routine - eg film analysis
Health and safety - effect of materials in general
The trouble-shooting covers complaints-and work in the plant on
problems. Many tests may be used.
/14.
C)
I.-D
Jh.
-4
BATCo document for Province of British Columbia 16 April 1999

-14-
The P.D. Material Testing is done:
Phase 1 Preliminary tests on
(coordinated by R&D) samples made in Develop-
men Centre (Laboratory
only)
Remainder of batch for
runnkbility trials
Phase 2 Runnability trials on
(coordinated by large batches
Manufacturing)
At times, Distribution
Centre. (depends on
material).
Then. regular orders.
Specifications with tolerances are prepared and are sent to the
suppliers. There &re technical meetings with the supplier when
there is a problem or a potential problem because drift has
been observed in the results.
t
Some characteristics are difficult to measure and specify
because the process can be adapted to accept some variation
e.g., rigidity is affected by scoring.
Sometimes, the question is: "Do suppliers have test facilities?"
Dave English said that there is no filter hardness measurement
in the plant. Instead, samples are tested once a year in the
laboratory on an Eastman tester, which looks like a mixture of
the old and new Filtrona instruments. The weight of filter
rods is not controlled but it is monitored; on one occasion,
heavy weight was found and a claim was made on the supplier for
a refund. I was shown the Laser Mike. It is not used for
filter rods but is used for the calibration of the metal standard
rods.
Tom Rehes talked on safety. There are three concerns:
Government - control within industry so that the
Government will not produce legislation.
Own employees
Customer
Odour and taste transfer are watched carefully. There is a
rigid inspection procedure for new materials and new machinery.
For the same reason, care is taken when machines are overhauled
that there are no tobacco materials or products nearby.
/15.
C:)
rQ
BATCo document for Province of British Columbia 16 April 1999

-15-
New materials are put into jars with-cigarettes; the Cigarettes
are smoked by a panel. New materials on machines and belts are
examined. The AGP is used but it is too slow. With existing
materials, which are --- is made. Imperial
have a newly-acquired travel kit to identify plastics, etc.
Some polymers are banned (PTFE and neoprene). Only a few have
complete clearance. Imperial have a list of acceptability for
materials.
Health and Safety in the Workplace is achieved by education of
the supervisors, who are made aware Of all hazards and are
made responsible for the safety of employees.
Material Safety Data Sheets are used to give employees the
essential information. Care is taken to avoid extraneous
information to avoid an upsurge of mysterious illnesses.
Eddy Sondermeyer talked on paper tests and specifications
cigarette paper, plugwrap and tipping paper.
Imperial use only laser-perforated tipping paper but are looking
at other types of perforated paper. They know of the Wattens
paper. Wattens have made the claims about their monitoring
device (I was offered a copy of the patent) which Is used to
control electrostatic perforation with a claimed coefficient of
variation of 3%.
The paper suppliers sample 10% of the bobbins sent to Imperial.
Thirty strips are said to be taken at random across the width
of the set and represent forty to fifty bobbins. The paper
samples are stapled to a sample card in four bundles. Imperial
test one of the bundles. A test report, with the accept/reject
verdict. is sent to the factory. The report goes to all
factories because bobbins might be transferred between factories.
The specification includes the target and limits, between
batches and within batches, including limits on individual
results, for permeability.
The specification is supplied to and agreed by the supplier; it
is prepared by Soadermeyer and is put into a purchasing
specification. Audrey Bishop prepares the specification for use
in the plants. The packing material specification is on a
separate sheet. Examples are attached.
On the general problem of specificatioa of materials, it is
difficult to know what tests are important and should be done.
Imperial do cross-checks with suppliers to check on instruments.
/16.
C)
NJ
r*J
BATCo document for Province of British Columbia 16 April 1999

-16-
Imperial recognise the problem with microlaser perforation
higher permeability at the beginning of a run (at the centre of
the bobbin) and lower permeability at the end of a run (at the
outer layers of the bobbin) associated with the accumulation of
dirt on the lens. They showed me the results of work they had
done on this effect.
Imperial attempt to apply limits that reflect their importance
on the "dilution" in the cigarette. The tipping paper represents
about half the final Variation.
I had a quick tour of the laboratory and examined the various
instruments. These have been mentioned earlier. A micro-fiche
device is used to examine the perforation in the paper - very
cheap device.
HEAD OFFICE - 1. T. M. DRYER CONTROL
This meeting with Robert Robitaille of Engineering was requested
by Bob Gibb. Jay Perry pointed out the essential links between
his group in R&D and the work in Engineering. Normally, this
activity would have been done in R&D but R. R. had done this in
his spare time.
In Imperial, the responsibility for Primary also includes the
V/
Green Leaf Threshing (G -L. T.) plant.
The I -T. M. dryer has been in use for many years and the mechanical
problems have been solved. However, the control system in use
was simple. The problem was that the dryer was being used to
correct for faults that originated in the earlier part of the
process. Imperial have, now, installed automatic control for
the dryer; the next step will be to go to the earlier parts of
the process and to fit instrumentation and control. The first
candidate is the conditioning cylinder on the CHS line.
The problem in this work has been the reliability of sensors.
The two measurements are temperature and moisture content of
the tobacco. Moisture meters have been available for many
years but the aims (expectations) were too high. imperial
re he need in
an been made in the
area o we must educate people to maintain
and calibrate the Instruments and then people will rely on the
instruments. Ease of maintenance of controls is important.
Imperial chose to use a programmable controller for the work
and have used the Texas Instrument family (PM550). One of
117.
C)
LA4
U4
BATCo document for Province of British Columbia 16 April 1999

- 17 -
these was put in the worst environment and tested as a
monitoring device; then, the control loops were closed one by
one. This procedure ensured that the first installation was
reliable and was accepted.
The I T. 'A. dryer and its controls were described. The control
loop for the water pump had been & "wait and see" control loop;
this has been changed to a continuous control. The second
control - more of an integrating control - adjusts the shell
temperature by adjustment of the burner so that the water pump
is maintained in the middle of its range. The flow of air
through the dryer is adjusted to control the tobacco temperature.
The standard deviation of-moisture content was said to be 0.2%
and that for temperature was 0.1'C. I did not ask how these
results were obtained.
The only control for the operator is & thumb-wheel switch for
moisture content. (In my tour through the Montreal plant, I
found that there was a rotary -switch on the moisture meter f or
the various blends because the moisture meter was sensitive to
the blend.)
But, Imperial pointed out that the system depends on the moisture
meter. The Quadrabeam meter is used. Imperial have had visits
from Inf ra Red Engineering and recognise that the I Gauge
may be better. However, servicing might be a problem from the
U. K; Quadrabeam is in Boston so Imperial have stayed with the
Quadra - -
An M.S.C: Quadrabeam is installed at the outlet of the dryer
for control purposes; a second Quadrabeam is installed at the
outlet of the cooler for monitoring purposes.
In the final version installed in the plant, the interaction
between system and operator is through a C.R.T. In the Montreal
installation (remember, Montreal is atypical of the Imperial
plants), the set points and actual readings of temperature,
etc., are through the programmable controller. ?
e effect of tobacco emperature on fillinz value wa diidussed.
h to my surprise, Canada have no evidence on, this but they
tend to do tests. -BlIr-Wi@Lvepf'oposed. that tobacco should be
erdried and then re-ordered. Imperial are interested in the
W high humidity drying system.
Imperial pointed out that evaluation takes a long time because
of the availability of people. The development and testing of
C)
S.0
4:3.
LF4
BATCo document for Province of British Columbia 16 April 1999

the control for the ITM dryer had taken two years.
Robert Robitaille then talked in more general terms (as & young,
enthusiastic newcomer to the business - my words). We should
aim to bring the industry from the 1900, to 1980 1 S. But we
need to obtain acceptance in the plant for new equipment and
computers; this is the problem in the primary where the factory
has a choice in 'the details of the equipment. The same problem
does not &rise in secondary where the equipment is bought in as
a package. (I think that this demonstrates the weakness of
it. R. , ideas because packages come in primary as well; there
is a choice in secondary). There must be a move to se I
9 o8is
in the primary control system e.g.. is a belt moving, is air
moving, so that the operator can be directed in fault f indin
FIL ,go.
A simulator has been built to tr&in'the operator in the use of Cal
the dryer control system. The results of the simulation are
plotted on a digital chart recorder.
The future work is high humidity drying, small strips by-pass,
installation of moisture meters and weigh-conveyors. An
examination will be made of the GLT plant to determine what
variables in the process affect the uniformity of the product;
then, control the variables.
Over the years, the strip dryer in the GLT has been modified.
All the exhausts except one have been removed. Counter-current
air flow and high-humidity drying are used. The air flow is
controlled by & wet-bulb thermometer in the exhaust. The
changes were made to reduce energy consumption but taste is MT-U-
said to be affected.
Work has to be done on the stem dryer to reduce the dust from
the dryer and to improve the quality of stem fed to the silo.
The problem of mould on the bottom of cardboard cases in the
warehouse was mentioned. The problem had been found when
Imperial went from hogsheads to cardboard cases for their
domestic tobacco. The mould occurred on the bottom case, which
was placed on the pallet. After much investigation and work
Polythene cover, 12 inch high, had been put on the outside oi?
the bottom of the case to prevent condens at the
pallets ha-d-E-e-enscr-apped: the cases are, now, put down directIA It
on the concrete or wooden floors. A Polythene liner, such as
that used on overseas tobacco had been proposed and tested but,(
though the tobacco was protected, the cardboard box had gone
mouldly -
Imperial have received a request from BAT to reduce the
of stem at packing because high moisture content
temperature
been found at the outside of cases of stem.
4CA@L
CD
-4
L14
r1_j
_2
Ln
BATCo document for Province of British Columbia 16 April 1999

GUELPH PLANT
Mrs. Denise Henry travelled with me. We joined up with Don
Schroeder (Read Office) and Jacques Roy (Technical Services,
Quebec plant) at Toronto. Our contact in Guelph plant was Tom
Vhitehouse (Technical Services).
Tom Whitehouse had gained the Impression that I was interested
only in instruments so he had listed the I that they
use; this added-little to what I have recorded elsewhere mud
will be discussed later, where appropriate. However, one new
feature was that they have \an a-pREral,s determine menthol,
which is added to the inner foil.
Guelph are slowly changing over (the programme goes into 19U5,
possibly, 1986) to a module consisting of a Mark 9 and a Focke
pp.cker (linked by Oscar). All grou s are treated as individual
machines because of problems with t9e unions on the group of
four machi nes concept that was being tried on my previous visit
(and, still, remains with the Mark 9 machines feeding tray
fillers - and Guelph hope to go entirely to that arrangement
eventually.)
In the old arrangement, with Mark 6-8-6 machines, one set of
the quality control instruments (weigher and circumference
eftuge ) served large troups of machines; however, with the new
layout, there will be difficulty of access to instruments. The
main limitation will be that the machine operators will have
limited time to walk to instruments; hence the proposal to put
instruments on each machine (or pairs of machines) - and, also,
the need for automatically-fed instruments with automatic re-
cording of the results rather than m&nually-fad instruments.
During my visit, I found that the AccuRay Micro 7000 weight
controls are being used as weight controls only. The other
data is not used. For example, production data is taken from
the cards. (See page 9). Even the Accuftay cigarette detectors
on the Molins packers, which, of course, give 100% inspection,
are not linked to the Micro 7000. (So. what is Imperial's
real interest in ANTIC?). It required, a printer is
taken from machine to machine to record the data from the
MICRO 7000.
The stoppages, also, are recorded manually - with reason for
stoppage; this information will eventually go on the computer,
again, manually.
Imperial said that there is a need to measure; control will be
in the future. Do we need control? This was a negative
reference to features of ANTIC but was also a positive question
to examine features that would provide preventative quality -
ie would keep the machine running with good quality rather than
shut the machine down because of poor quality and, then, put
the fault right. ORA
DA ;S
/20.
re
BATCo document for Province of British Columbia 16 April 1999

-20-
In Imperial, they control the final weight of the cigarette;
so, filter rod weight variations, which they do not control,
are important. Hence, they have two warning limits: 1.2g/100
filter rods for control within Manufacturing (Don Schroeder)
and 1.8g/100 filter rods, which they discuss with Ruth Smith
because of the product design implications.
The tour of the plant (with & group of people including Denise
Henry, Tom Whitehouse and Jacques Roy) looked at the new facility
for DIET addition: a weigh conveyor in the main cut tobacco
line after the coolers, a DIET feed line with a weigh conveyor
slaved to the first weigh conveyor, and a cylinder to mix the
two lines. The DIET line was located in the old tobacco store;
a new, adjacent, building had been built for tobacco storage.
The Dickinson continuous conditioners are, now, working s&tis-
factorily. (The main improvement was to strengthen the doffers
and to pack the tobacco more uniformly in the case, top to bottom,
by inverting immediately after packing at the GLT plant.) As
a result, the CTClAs had been removed during the summer shut-
down.
I noticed that one snag of the pre-blending bin for the lamin
before the normal blending bin is that there has to be gap
between operations to allow the bin to empty. A speed-up
arrangement can be used, of course. However, this is not a
problem because Guelph primary is not fully used.
The CRS line is a conventional line. WTS is still imported
from Montreal or Quebec. A WTS line will be installed in the
near future.
in the filter rod department, there are seven. KDF2; three with
AF2 and four with E10. The latter will be changed to AF2, event-
ually. I found that the Filtrona ATCG and APD, though linked
to printers at the machines, were used as stand-alone: the
machine number could not be entered on some (the program had
to be changed) and in the others was not used. The reason was
the unre-liability of the printers (the old type) and the prob-
lems with the unions in the servicing of the instruments. (See
later). Imperial will try hot-melt for the lap seal adhesive
to improve uniformity. Only normal porosity paper was being
run whilst I was there; they Said that they run at 400m min-I
with high porosity paper but, if an operator has trouble, the
first move is to reduce machine speed.
On the problem of the weight of filter.rods, (see above),
Imperial have found that different machines run with different
filter rod weight. The nominal target used in the Product
Design was derived empirically (having defined pressure drop
&ad circumference).
/21.
BATCo document for Province of British Columbia 16 April 1999

-21-
I raised the question of circumference control of filter rod
makers with highly porous wrap. Jacques Roy said we need & new
control; Fred Thomson (Guelph) said "No".
The Q.C. equipment for filter' rod manufacture is in two stations.
The sampling and testing is done by a Manufacturing person -
no,tQuality Services. The girl uses the results to plot the
Q. C. charts. The paper tapes from the instruments are discarded
(no machine identification). So, subsequent analysis of results
is based on the Q.C. charts, which may be illegible or incorrect.
Remote printers are located on each filter rod maker but are not
used. This is because of the unreli&bility of the Filtrona
instruments in the Guelph Plant. (The problem is the printer,
which is the early model; probably mishandling the paper when
tearing the paper; this is complicated by the way that the
instrument repair is handled by any one of a group of technicians,
who may, as & result, see the instrument at long intervals -
a union problem and the hierarchy system for skilled jobs, -
which demands expensive, time-consuming training for all
technicians rather than one specialist). r" 41- 1.414
I @E&St`mLa m'onitor for plasticiser addi7ion__@
L
The Quality Control limits for circumference are that the average.
for five rods should be within tO.05mm of target; however,
this is an action limit to the operator to adjust his machine:
the rods are not rejected. For pressure drop, the limits
are 10mm 1i2O for all rods except the high pressure drop
rods where the limits are 25MMH20- Again, the rods are not
rejected because they would have been rejected for other
defects eg recessed ends.
In cigarette making, the Shadowgraph weighers and Solex gauges
are located on each line of cigarette making machines. The
machine operators do the Quality Control. The reason for using
the operators to plot the results is that it ensures that the
operator is aware of the results. (See p 19 for proposed changes).
Three MSE moisture meters (just brought-up-to-date) are installed
to monitor the moisture at the output of the three cut-tobacco
dryers.
A weigh conveyor has been fitted recently to control CHS addition
(but did not seem to be working?)
QUALITY AUDIT
This area is run by Lee Redsell. She had prepared the sampling
scheme for the entire plant.
Faults (defects) are those categories that produce customer
Complaints. The standards are set by machine Capability Lad
may not be those wanted by Marketing.
/22.
CD
_J
Lri
co
BATCo document for Province of British Columbia 16 April 1999

- 22 -
One person on each shift goes to each packer, four times on
each shift, takes two cartons to m test station, located at the
edge of the floor, and inspects the cartons for carton faults;
& "Carton Inspection" form '- attached - is completed. (The form
shows 'line"; is there only one carton packer for each line?).
Lee Redsell asked the inspector it this information was fed
back to the operator of the packer. The answer was that it all
depends on the operator: some will accept the information;
other will "not want to know". From one carton, one pack is taken;
from the second carton, two packs are taken. Random numbers
are used for this selection. There are 25 Cigarettes in each
pack. The packs are taken to the Quality Audit room and are
placed randomly in the ten boxes for each shift; however, at
the end of the shift, all the boxes have the same number of
packets of each brand to balance the work load and to ensure
that each inspector sees each brand. (Remember, Guelph makes
only a few brands. What happens in, say, Montreal factory
with its many brands?). The ten inspectors, who work on the
day shift only, are allocated randomly to each column of three
boxes, (one box for each of the three shifts). The Cigarette
and Package Inspection form is completed with maker number mad
packer number. The faults, together with machine number and
shift, are identified on the cigarette with a pen and collected,
placed in another series of boxes by making machine number and
examined by two more inspectors for Intensive Inspection as
Part of the Frequent Major Cigarette and Package Defect Prog-
ramme. A Frequent Major Defect Notice form (a
completed ice Number
is completed so that, if the fault persits, this is identified
by the number of the notice (1,2,3,4 .... ) for the particular
fault on the particular machine. L-ba-
taken when a notice is received. The number of the notice
3. WMatifies the level 4 t at which action is taken.
Two clerks analyse all the results and make reports and complete
graphs, weekly, monthly and quarterly. which are sent out within
the Plant and to Montreal (Head Office).
The results also go to Fred Turner (Manager - Quality Statistics)
who plots the data in various ways to identify trends, fault-
producing machines etc. Thus', the plant defects (percentage
weekly totals) are plotted for each week, by week over a period;
the defects that exceed the Plant Average are picked out.
Then, the week's results are examined by machine and those that
exceed the Upper Confidence Limit (UCL) are picked out. Then,
analyse each machine over several months. The work ignores
brand.
One person in the primary takes the moisture samples and the
shaker samples (see sheet). Approximwtely 120 moisture samples
each day. (No W.T.S. process).
The moistures are sent to the Primary Process Engineer and to
supervison. Some results go to Montreal. (M in the chart
indicates that these results go to Montreal).
/23.
_@J
LIN
BATCo document for Province of British Columbia 16 April 1999

-23-
Note. Each operation is approximately 10,000 kgs. It consists
Z_f54 cases of lamina, plus PCL plus CRS. Approximately ten
operations a- day - less than two shifts are required. Friday is
kept free for maintenance.
R.C. LABORATORY
The laboratory is a tiny place with few instruments but many,
under-occupied, girls - but, who do have other responsibilities
such as calibration of instruments. The BAT Ventilation Tester
is located here - and is used in preference to the I.T.L. ver -sion
because of ease of use (insertion of the cigarette into the
test head is difficult on the I.T.L. Tester, which uses the
Filtrona labyrinth se&ls). The dental dam perishes quickly
and is replaced almost daily. (Later, I recommended that
Imperial use the sleeve instead of the dental dam; however
Imperial have always used diaphragm se&ls and are reluctant
to change. See Final Meeting.) An I.T.L. version is located
on a trolley so that it can be taken to the machines (which
surprised other people In the party because the I.T.L. tester
-has proved to be unsatisfactory on the factory floor because
of vibration). An EEPD Filtrona tester was used. A Turbula
.Schatz tester was available for Ends Stability but was not
used. A Borgwaldt Densimeter had been delivered but not put
into use.
A Mettler Top-pan electronic, digital balance was used for
moisture tests. I't was said to have interface capability but
no money had been allocated for the computer, etc. I expressed
surprise at this. (It was on Denise Henry's list of items to be
discussed).
The visit finished at lunch time on Tuesday 24th August, 1982.
Denise Henry was not present on Tuesday and I was accompanied
by Don Schroeder and Jacques Roy. NOTE. During my visit, I
had a brief talk with the Plant Manager (Mike Courtney).
Jay Perry (who should have escorted me during this visit but
had been requested to see Pat Dunn Instead) arrived before
lunch.
.AYLMER
I travelled to Aylmer with Jay Perry after lunch on Tuesday
24th August, 1982; however, he did not accompany me during
my discussions and tour. The afternoon was spent in a discussion
with Alec Nyalis (Plant Manager) and Vic West (Quality Control).
The grade characteristics change from day-to-day and through
the season; it depends on the buyer, the particular auction
room, buying policy and price. _lo within n@,,@de
uniformml. & r ru'@
@ I is
an attempt is made to produce a e ine e
124.
.0
BATCo document for Province of BritiSh Columbia 16 April 1999

-24-
from twelve racks simultaneously (used to be six). At 20,000
pounds in hour, a grade run will last 1i hours for a short run. YjA-
Ideas. Can we get longer process runs? A larger inventory
would be needed. Blend whole leaf. Use a controlled condi-
tioning unit. Set C. and C. Weigh the output from each dis.
charger, which monitors the efficiency of each section. Air
flow measurements in the classifier. Better analysis of data.
The standard deviation of particle size has improved over tile
past four years. Why?- The market is more stable; the process
line has been made more efficient less blockages (and
stoppages).
On Wednesday, I met Cy Bossy - Manager Leaf Buying - and Howard
Good - Manager, Overseas. I did not meet George I
Manager, Domestic. (The above was changed during my visit: Cy
Bossy was made V.p. Leaf to replace Peter Gage who was retiring).
There is a need for an instrument to help the buyer - particularl?
for chemical analysis but, also, for moisture. I @
to1dTSDt-__t ere might 5-e-lo-n-e--that could-go in the boot of a car. bg
The start of the season liad been brought forward another week;
now, late September for first sales.
I had a discussion with Vic Test and a tour of the plant.
For moisture testing of over: as tobacco, a grab sample of stem
is taken from every second r: qe (or the jar is held under the
delivery chute) and a core -_;--ole is taken every fourth case
of lamina.
For domestic tobacco, & W", -)le is taken sixth case.
(ti- tiermostat is
The older Freas oven is tt
erratic) and replaced by a The er oven is, now,
eight years old. The old 1 alanc stem is, still,
used. I had a long discus! Vic he automatic weighing
system (at his request) ant that 1 lild send further
information.
I had & discussion with Harry mell. Tht; -untral control
room is "under a cloud". The proposed remote control of the
driers and the C. and C. was over-s-ibitious and did not give
the information rcquiree, Moisture lieters were put everywhere
but have been removed need 1, calibration was not
recognised. A; C-eters f. -e th- rs did not give the
required information on og er; of the threshers.
The modifications to the -Yer - d ussed earlier (with single
exhaust and contra-flow to giv iergy savings but, which
had also given taste imp) meitts) aLd been finished and
/25.
BATCo document for Province of British Columbia 16 April 1999

-25-
tested in the previous season; 1982 would be the first season with
the rebuilt equipment. The steam pressure is n nch-2 _wu.,,e
instead of 100 lbs inch-2, -11-1 @the sections are balanced A.:.f
(gauges are provided) to prevent air flow between sections.
There is no hold-down section so air velocities are limited. A
thLanIr b uniform
drying. meter is used in the cooler section for
feed forward control of the hot-water sprays in the re-ordering
section with a second Posk.et meter after the redryer for feedback
control. A BAT meter, also, is located after the redryer as a
monitor. Finally, a Kappa meter is used (every tea minutes) on
grab samples. (This room, with the Kappa, has been moved to
the end of the redryer). The variation in moisture content at
the output of the dryer (S.d. - 0.85%) is twice that at the
input to the dryer (s. d. = 0.46%).
Harry said that %hey have _deve1apaA_11LejX own wet and dry bulb
thermometer and claimed that it worked Y. The we
bulb ia the dryer exhaust is used to control the air flow ti
through the dryer. A heat exchanger in the dryer exhaust is
used to recover some of the heat, which is used to heat the
bale storage.
The temperature of the lamina at the exit of the redryer is
controlled to ensure that Lhe temperature of the lamina in the
case is correct. A sensor located in the case filler monitors
the temperature of the air forced out of the case filler and
the peak temperature - obtained when the ram moves down - is
used, with & correction, for case temperature.
On the stem line, a stem cooler/heater will be installed. It
has the appearance of a SML11 redryer and can be used either to
cool the stem that goes into cases or to heat the stem that
goes to the silos. BAT(UK&E) have asked for the stem to be
packed cooler than before because of high moisture content
found in the cases when opened. On the other hand, the stem
must be hot, to reduce breakage, when it goes to the silos.
Harry Cartmell has developed his own weigh conveyor; it is &
simple counterbalanced roller that presses on the underside of
the belt conveyor and operates a small load-cell. It is crude
but is expected to be better than 10%, which is all that is
required for the particular job - to operate the flLp-type
divider that splits the feed of leaf tobacco for the two C.
and C. lines. One unit would be in each of the two lines.
Remember that Aylmer runs two C. and C. lines in parallel
on one feed line of tobacco.
A control has been installed for the ventilation system in th
packed case warehouses. This ensures that ventilation occurs
only when the dew point of the outside air is less than the
floor temperature and the absolute temperature is above a
pre-set limit.
/26.
%-0
--4
LrJ
I"..)
BATCo document for Province of British Columbia 16 April 1999

I was shown the elaborate recording system in a test bay
of the new w-&rehouse where measurements were being made for
a normal section and for a floor-heated section.
MONTREAL PLANT
I-had & short visit to the Montreal plant on the morning of
Thursday 26th August, 1982. 1 was accompanied by Don Schreoder.
I was met by Jacques Gaboury who has been promoted recently to
Assistant Plant IdLnlger Lad by Arthur Tasan who has been recently
transferred from Quality Services to Manager, Quality statistics.
I did not meet Jacques Thibault who has been recently promoted '
to be Manager, Quality Services.
Jacques Gaboury gave me a talk on the Montreal factory, its
problems and its philosophy. He gave me a folder (attached) on
the important items. Arthur Tasan talked on the Quality Services
and acted as the guide around the plant.
Montreal factory is equipped with the oldest machines (often,
Passed from Guelph or Quebec). It makes only 15% of Imperial I 8
cigarettes and specialises in the small brands and the odd
brands eg. 100 mm length cigarettes, special packs, American
packs. Some brands are known as frozen brands; they are made
in batches, pre-dated at intervals of, say, three months, frozen
and released on the appropriate date.
Montreal make many other tobacco products (see list). The
Manufacture of snuff has been discontinued; it will be made In
the U.S. from Canadian tobacco. Over the next two years, the
bland will move to American tobacco so that & common blend will
be used for the Canadian and American market. The U.S. will
continue to import the Canadian tobacco but it will be for
other uses.
A major problem in the Montreal factory is that the average
age of the employee is high. Of the 1022 people, 270 have
over 35 years of service. Imperial are encouraging these
people to leave with the offer of good pension. However, as
they leave, they take their expertise with them; so, Imperial
try to retain this expertise by the use of video, which is
used in the training programme. On the other hand, because of
modern toohnololgy and automation, the intake of people has
been reduced so there are very few, young qualified people In
Imperial. Further, these new people have to start at the
bottom - usually, on manual labour - so they go to other jobs.
Imperial cannot recruit qualified people direct to, say, machine
operator. (This restriction does 00t'Lpply for certain, limited,
specialised trades, eg electrician who come from the Montreal
/27.
CD
BATCo document for Province of British Columbia 16 April 1999

-27-
equivalent of a Technical Training College at the age of 20
years). There are pressures to send people out to be trained.
The nou-specialised people-that are recruited are low grade
(year) - lower that at Guelph. Of the 1022 people, 65% are
non-skilled. There are no restrictions on job transfers
(approximately 400 job changes each year) which makes training
important. There are 125 absentees each day; this is the
problem with the large number of older employees: many are on
long-term sickness. Also, because the older employees have
long holidays, their absences must be covered. This requires
surplus of suit.ably trained people who can be transferred
and is particularly a problem in Quality Services.
There are twenty-five foremen in Production to control the
many sections. There is a problem of unqualified persons in
supervision who cannot, therefore, control their mechanics and
operators; this problem is-said to include higher management eg
Assistant Plant Manager.
People are being moved to staff. It is hoped that this will
improve the quality of the product. The number of professionals
in Quality Services and computer analysis will be increased and
there will be a training programme.
At present, to Join Quality Services, though the job is allocated
on a seniority basis, the applicant must, in addition, pass a
small test (the form of the test is dependent on the job) and
must pass an eye test.
All sampling and testing is done by Quality Services. The
attached diagram indicates the tests and people.
Moisture sampling is done separately from moisture testing,
which is done on the day shift, only. Approximately 350 tests
each day, range from 200 up to 400 samples each day, cover all
tobacco products.
Materials Inspection Program works on the samples provided by
the supplier so it depends on the honesty of the supplier;
however, Imperial do occasional checks on the batch.
The Quality Control inspection is done at the machine by Quality
Services; also, they do occasional weight checks on the cigarette
makert the results are recorded on the same chart as that used
by the machine operator - but, using a different symbol. In
the Montreal system, the quality inspection at the machine is
quality control as well as quality audit; this is necessary
because the machines have hand catchers so the operators do not
have time to do inspection. This meth6d of inspection is not
/28.
(_Aj
r@j
DJ
BATCo document for Province of British Columbia 16 April 1999

-28-
liked by management because the operator relies on the Quality
Services inspection Instead of making his own inspection.
There will be a move to the audit system when mechanical catching
is installed. However, the problem will be to make the operator
responsible for quality.
For the inspection of cigarettes, 50 cigarettes are taken from
the tray to &a inspection station, where there are examples of the
faults, for inspection; the information is recorded and is fed
back to the operator. A notice (see attached) is given to the
foremen with Cigarettes (slipped through two slits in the form).
to show the fault. (The form says "25 cigarettes" but the
sample is 50 cigarettes).
For the inspection of packets, separate samples are taken.
One person (and test) that is additional to the attached chart
is that one person in each shift does ventilation tests in the
laboratory.
In the filter rod room, old machines with hand catchers are
used; all are equipped with the Filtrona continuous circumference-
monitor. The sampling and testing is done by Quality Services
on manually-fed instruments for pressure drop, circumference,
weight and pl&sticiser. The KDr2 is used for the production of
tubes because of the high demand.
highly-porous filter rods are made on the old filter rod makers
but some work is being transferred to a KDF2. There are some
problems with the KDF2 on both porous and non-porous rods. The
main problem is circumference control but weight is also a
problem; the weights are too high in comparison with the older
machines. There are, also, problems with tow denier (as mentioned
earlier). Now, an old bale has been found that has been tested
and found to have the correct denier. This bale will be run in
all three plants to measure the capability.
As part of the education programme, Montreal will have meetings
of Engineering, Production, Quality Services and Training. The
meetings will have foremen, mechanics, operators etc. They
will be taught the limits on quality and how to trace the source
of the defect - rather than just run the machine until someone
else finds the problem. The intention is to go further and to
train the people in fault prevention. A further aim is to
bring together the mechanic Lad the operator and to make them
aware of each others' role so that they will work together to
the common goal Of the production Of a quality product - and
avoid the problems of isolated individuals who will attempt to
blame the other person, the machine or the material. Them,
Quality Control can become Quality Services.
/29.
LOW
Ln
BATCo document for Province of British Columbia 16 April 1999

-29-
MONTREAL HEAD OFFICE
An afternoon (Thursday 26th August, l9a2) was spent with Don
Schroeder and Luke Hireault on materials.
A jumbo roll of paper gives three sets (front, middle and back)
of 40 - 45 bobbins each. A jumbo roll has a length of about
20,000 metres and will provide about three bobbins in each
set. There are approximately four sets on a pallet. An order
might be for twenty pallets.
The supplier will take about I metre from four or five bobbins
out of each set ie 10% of bobbins.
The specification was originally supplied by the supplier. It
gives average values : sometimes with range, sometimes with
minimum. However, the values are based on averages of tests;
there are no requirements for variations of the individual
values.
If there is a problem in the running of a material that can be
allocated to the supplier, the matter is dealt with by Manu-
facturing.
The visual examination of printed materials is done on samples
provided by the manufacturer. Because of the improvement in
printing techniques, rejection is rare. Manufacturing would
like to reduce this inspection but Marketing will not agree.
The residue of the samples is sent for the runnability tests.
Runnability tests are not done on printed stocks.
Some years ago, Imperial gave examples of faults, with acceptable
and unacceptable examples, to the printers. This guide estab-
lished standards and was said to be the first in the industry.
Imperial wish to have less than 1.5% total defects. The insi-
stence on the quality product has shown to the suppliers that it
is important that they maintain their machines; otherwise,
because of the cost of rejects - either in sorting the product
or scrapping the product -.it is expensive to produce a poor.
quality product. Now, the printers have well maintained
machines and produce a better product.
On cellophane, etc., no samples are tested on a routine basis.
However, occasionally, a pallet is selected and run.
There are no tests, by Technical Services, on adhesives.
There may be occasional viscosity tests-on hot-melt adhesive.
For flavours and casings, Imperial rely on the integrity of
the supplier. For menthol, samples are sent to the laboratory.
For triacetiu, a hand saccharimeter is used to give a crude
measure of refractive index. There are subjective assessments
for colour and smell. Samples are sent to Technical Services
but the results are too late. The supplier provides a certificate
of analysis with each tank load.
/30.
L'4
NJ
BATCo document for Province of British Columbia 16 April 1999

MONTREAL FRIDAY 27TH AUGUST, .1982
INSTRUMENTATION
I had a discussion with Gaeten Levert on his work. Gaeten -
who has been with Imperial for ten years - has been transferred
and now works for Jay Perry. However, he appears to be used in
many of his earlier projects.
mined the OM Lad aj he correct
of 1.5% that I reported - except with DIET where the correction
is 1.3%. (We do not test DIET in the Beaudesson).
He has checked the calibration of he minar flow element in
the BAT Ventilation Tester against a Miriam element and has
found an excellent 1:1 relationship. The Imperial tester shows
marked deviation at low flows. Hence, Gaeten is recommending
the BAT Ventilation Tester.
In his new job, he is involved in the use and calibration of
weigh-conveyors and moisture meters. Imperial have about
seventeen of the Quadrabeam meters - a large investment and,
hence, they are reluctant to change to another instrument.
Gaeten seemed to be unaware of the Infr&gage SM4. All the
Quadrabeam instruments have been returned one-by-one to the
manufacturer. in Boston, for modernisation.
At my request, Gaeten checked the results on the Imperial
Impedance Gauge. The drift is to lower values - which
supports my suggestion that the tobacco dries during the
measurement because of the use of the factory air supply.
FINAL MEETING
Jay Perry had organised a final meeting, which was attended by
Denise Henry, Don Schroeder, Louis Bouchard, Ga.eteu Levert and
Jay Perry.
I had no questions to ask since these had been dealt with during
my visit.
The use of the dental dam test head in the Ventilation Tester
was discussed. Gaeten described the various test-head arrange-
meats for the benefit of the audience and Said that the use of
the dental dam had &risen because of Imperial's original ideas
on measurement of tip ventilation only with minimum enclosure
of the envelope. The problems of changeover to the sleeve-
type head were discussed and it Was agreed that there should
be no problem except for a change in the numbers in the speci-
fication.
Denise Henry raised the point that 1 had made on the large
/31.
C)
-4
BATCo document for Province of British Columbia 16 April 1999

-31-
number of people in Quality Audit. I said that quality Audit
could fulfill two roles --& Police activity and an Information
activity. Denise stressed that it Was not a Police activity but
for Management Information. She said that far more people
would be required it this Inspection activity was transferred
to the machines - because the operators would cease to inspect.
She asked about the practices in other Companies. I said that
in Germany there is little inspection they rely on the Machine
operator.
132.
C)
U4
N)
N)
co
BATCo document for Province of British Columbia 16 April 1999

-32-
PEOPLE AT MEETINGS
16TH AUGUST, 1982
Wilf Gray Process Development, Primary
Jay Perry Process Development, Primary
Audrey Bishop Prod. Design & Spec.
Don Schroeder Manufacturing
Louis Bouchard Technical Service Manager
Denise Henry Manufacturing
Dave English Technical Service
Gilles Prefontaine Instrument
Ed Sondermeyer Technical.Service
Stan Candlish
Bruce Stirling Analytical Service
Bob Rice Research
Gaeten Levert Instruments Proe. Dev.
18TH AUGUST, 1982
J. Perry
W. Gray
T. Smith
A. Porter
G. Prefontaine
B. Rice
P. Bean
G. Levert
19TH AUGUST, 1982
Bruce Stirling
Carol Roy (Mrs)
Robert Robitaille
27TH AUGUST. 1982.
J. Perry
D. Henry
D. Schroeder
L. Bouchard
G. Levert
/33.
CD
N)
N)
BATCo document for Province of British Columbia 16 April 1999

- 3 3 -
ORGANISATION IN MANUFACTURING
HEAD OFFICE
V. P. OPERATIONS - C. WARREN
MATERIALS CENTRAL LEAP BLENDING CENTRAL ENGINEERING
(PURCHASING MANUFACTURING Peter Gage Bill Ross
CENTRAL
SUPPLY
TRAFFIC AND
DISTRIBUTION)
V. P. - ANDRE LAPORTE
PRODUCTION ENGINEERING TECHNICAL SERVICES
(long term Y. Primeau D. Henry
planning) (optimisation D. Schroeder
of machinery) (cigarette components
and quality control)
A similar arrangement in the factories:
Plant Manager
Production Manager
Engineer
Office Manager
(NOTE. Other Head Office functions are: Personnel, Finance,
Marketing. Information Services, industrial Relations,
Research and Development).
C=)
/34.
CD
BATCo document for Province of British Columbia 16 April 1999

-34-
QUALITY SERVICES IN THE PLANT
Guelph Montreal Quebec
Staff 5 3 3
Union 26 46 14
Total in 882 1114 432
Factory
Q. C. (S) 3.5 4.4 3.7
(includes cigar
factory)
Breakdown of Quality Services
Primary Filters Audit Others
Guelph 4 3 10 9
Montreal 4 2 10 8
Quebec 2 Operator 7 3
Quality Services Manager
Superviser Superviser Process Engineer
Incoming Materials Secondary Process
(New Post)
(Discussed elsewhere)
M. I P. First P.D.
(Material Shipment (Purchasing
Inspection Check Department)
Programme)
Visual Inspection Runnability tests New materials.
of incoming on those items Make packs (small
material (said that are a problem number) in Develop-
to be less eg corrugated cases, ment Centre* for
important because paper, tipping, laboratory to
of improved print- adhesives. evaluate properties
ing technique). (including odour) .
Then, remainder
for runna.-
bility tests.
Phase 2.
New full title is:- Product Development and Optimisation
Centre.
Head:- Roger Lafleur (Manufacturing)
BATCo document for Province of British Columbia 16 April 1999

C@
-@O
p1b.
-@j
U-j
ri
L.@j
BATCo document for Province of British Columbia 16 April 1999