technical report 1st collaborative study on air

Physical Test Methods Sub-Group

Technical Report

1st Collaborative Study on

Air Permeability in Accordance with ISO 2965:2019

November 2019

Study Coordination Group:

Bernhard Eitzinger, delfortgroup AG, Austria

Philippe Le Men, SWM Intl., France

Patricia Müller, delfortgroup AG, Austria

James Vincent, Cerulean, U.K.

Table of Contents

1. INTRODUCTION ............................................................................................................ 3

1.1 Purpose and Scope ................................................................................................... 3

1.2 Study Protocol ......................................................................................................... 3

1.3 Products and Measurements .................................................................................... 4

1.4 Study Participants .................................................................................................... 4

2. STATISTICAL EVALUATION ...................................................................................... 5

2.1 Raw Data Treatment ................................................................................................ 5

2.2 Outlier Analysis and Removal ................................................................................ 5

2.3 Robust Mean Values and Standard Deviations ....................................................... 6

2.4 Evaluation of Repeatability and Reproducibility .................................................... 7

2.5 Evaluation of Laboratory Performance (z-Scores) .................................................. 7

3. DATA INTERPRETATION ............................................................................................ 8

3.1 Repeatability and Reproducibility ........................................................................... 8

3.2 Laboratory Performance .......................................................................................... 9

4. REFERENCES ................................................................................................................. 9

5. APPENDICES ................................................................................................................ 10

APPENDIX A – Protocol ............................................................................................... 10

APPENDIX B – Data Summary ..................................................................................... 12

APPENDIX C – Diagrams ............................................................................................. 16

PTM-217-CTR 1st Collaborative Study on Air Permeability – November 2019 3/25

1. INTRODUCTION

1.1 Purpose and Scope

It is an objective of the Physical Test Methods (PTM) Sub-Group of CORESTA to carry out

inter-laboratory studies on physical parameters of tobacco products and their components. As

air permeability is well-known as an important parameter of wrapping papers for tobacco

products, the PTM Sub-Group decided to carry out a first collaborative study related to this

parameter. The method for measuring air permeability is specified in ISO 2965. In section 9 of

this standard two collaborative studies are mentioned to determine repeatability and

reproducibility of this method. A first collaborative study was carried out in 1994 and a second

collaborative study was carried out in 2005 including banded cigarette paper for lower ignition

propensity cigarettes. ISO 2965 allows two measuring heads to be used for the measurement of

air permeability on cigarette paper, the 10×20 mm² measuring head, which is the more common

head, and the 2×15 mm² measuring head, which is mainly used for measurements of air

permeability on lower ignition propensity bands. The revised version of ISO 2965, published

in 2019, now allows the use of the 2×15 mm² measuring head for air permeability measurement

on cigarette paper, whenever the available area on the paper sample is too small for the

10×20 mm² measuring head. This may be the case for measurements on paper samples taken

from cigarettes.

At this time no repeatability and reproducibility data for the 2×15 mm² measuring head used on

conventional cigarette paper are available and thus this third collaborative study serves to

provide that data. It further serves to compare permeability values obtained with the two

different measuring heads.

Furthermore this study allows the participating laboratories to monitor their performance in

comparison to other laboratories, to derive actions for improvement of their internal processes

and to fulfil accreditation requirements.

All results will be presented in anonymized form.

1.2 Study Protocol

The test protocol used for this collaborative study is given in Appendix A and will be briefly

summarized below.

The protocol contained information on the products to be tested, the preparation of samples,

set-up and calibration of instruments and the procedure for carrying out the measurements. The

study participants were required to provide data to identify their laboratory, data on the

instruments and calibration methods used and all measurement results obtained. The results

were collected in an Excel-sheet and sent to the Study Coordination Group for further

processing and evaluation.

Once received, the data set was inspected for obvious inconsistencies and for deviations from

the protocol. While no substantial deviations from the protocol were found, not all laboratories

were able to measure all samples due to the wide range of air permeability values and the

requirement for using different measuring heads.

The distribution of sample materials, the protocol and the data collection sheet started in June

2019 and laboratories were asked to report the results until end of August 2019.


1.3 Products and Measurements

For this collaborative study a conventional (non-LIP) cigarette paper, a plug wrap paper and

two perforated tipping papers were used as sample materials. Their details are given in Table 1.

The permeability data in Table 1 do not necessarily represent actual measured values but just

serve as an indication of the permeability values to be expected. The paper samples were

provided by Schweitzer-Mauduit Intl., Tannpapier GmbH and Papierfabrik Wattens GmbH &

Co KG.

Table 1 – Characteristics of the paper samples

ID Product description Permeability Measuring head

Type Typical Value CU

A Cigarette Paper Natural 60 2 mm × 15 mm

B Cigarette Paper Natural 60 10 mm × 20 mm

C Plug Wrap Paper Natural 10000 10 mm × 20 mm

D Tipping Paper Electro-

perforated 340 10 mm × 20 mm

E Tipping Paper Laser-perforated 1400 10 mm × 20 mm

The cigarette papers A and B are identical, they only differ with respect to the measuring head

that shall be used for the measurement.

For each paper sample A to E three replicate measurements had to be made, with one replicate

consisting of 20 individual measurements. Thus a laboratory completing the full set of samples

had to make 3×5×20 = 300 individual measurements of air permeability.

For the measurements the laboratories had to follow ISO 2965.

1.4 Study Participants

In total 9 laboratories participated in the study by submitting data, with the list of participants

in alphabetical order given in Table 2. A code was assigned to each laboratory, thus the order

of laboratories in Table 2 does not agree with the order of the laboratories in other tables. Not

all laboratories were able to measure all samples. Some laboratories did not have a 2×15 mm²

measuring head available and some were not able to measure perforated paper or the high

permeability of the plug wrap paper.

Table 2 – List of participants

Participant name Number of sample sets Country

British American Tobacco 2 Germany

Cerulean 1 United Kingdom

Godfrey Phillips India Ltd. 1 India

Japan Tobacco Inc. 1 Japan

Papierfabrik Wattens GmbH & Co KG 7 Austria

PT. HM Sampoerna Tbk. 1 Indonesia

R.J. Reynolds Tobacco Company 1 USA

SWM Intl. 1 France

Zhengzhou Tobacco Research Institute 1 China


One laboratory did not have the 2×15 mm² measuring head available and measured paper A

with the 10×20 mm² measuring head. This data set had to be excluded. Another laboratory used

the 2×15 mm² measuring head but the head was not automatically detected by the measurement

instrument. Thus the permeability values were calculated manually from the values reported by

the instrument. This was found to be acceptable and the data set was included. A third laboratory

wanted to provide new results for sample A after the first draft of this report had been circulated.

This request was granted and the results provided herein now reflect this new data set.

2. STATISTICAL EVALUATION

2.1 Raw Data Treatment

In total 16 data sets were received and after an initial screening for inconsistencies the data were

prepared for statistical analysis. Mean values (MV) over all laboratories, the average within-

laboratory standard deviation (SDw), that is, the standard deviation of a single replicate

averaged over all laboratories and the between-laboratory standard deviation (SDb), that is the

standard deviation of the single replicate values obtained by each laboratory on the respective

day of measurement, are provided in Table 3. The number (N) of data sets is also given.

Table 3 – Summary data for air permeability per day over all labs, outliers included

Day 1 Day 2 Day 3

ID MV SDb SDw N MV SDb SDw N MV SDb SDw N

CU CU Cu CU CU CU CU CU CU

A 69,99 1,90 3,35 11 70,73 2,24 3,26 11 70,21 2,11 3,34 11

B 60,64 3,00 2,31 14 61,03 2,43 2,33 14 61,16 2,83 2,15 14

C 9510,51 353,32 615,27 10 9489,95 253,29 584,36 10 9461,56 303,49 676,41 10

D 342,22 14,97 8,55 9 340,40 12,85 8,68 9 340,77 12,87 9,20 9

E 1401,14 46,16 28,20 8 1396,91 50,03 27,49 8 1399,05 48,67 27,75 8

2.2 Outlier Analysis and Removal

Repeatability and reproducibility data were determined following ISO 5725-2, whereby outlier

testing according to Cochran’s test and Grubbs’ test was used. In a first step each set of 20

individual measurements was tested by Grubbs’ test with respect to the maximum, minimum

and the two highest and lowest values. Any outliers that were found were excluded, but the

reduced data set was kept in the analysis.

Second, the three replicates of each laboratory were tested for their internal consistency using

Grubbs’ test and then the mean values and standard deviations of each laboratory were checked

in comparison to the mean values and standard deviations of other laboratories using Cochran’s

and Grubbs’ test.

After elimination of outliers, global statistics, in particular mean values and standard deviations,

were calculated and the repeatability and reproducibility statistics were determined.

In order to evaluate laboratory proficiency in the form of z-scores, as described in

ISO 13528:2015, a ‘true’ value and standard deviation need to be assigned to each paper sample,

which form the basis for the calculation of z-scores. In contrast to other studies, where the ‘true’

value is known or can be easily assigned, such values are not available in this study.

Consequently, the ‘true’ mean value and standard deviation were determined as the global


average and standard deviation obtained by the above outlier elimination procedure used for

the determination of repeatability and reproducibility. The z-scores were then calculated for all

laboratories, which reported data, irrespective of whether their results were excluded in the

calculation of the global mean value and standard deviation. The z-scores are based on the

originally reported data set but after elimination of individual values by Grubbs’ test.

The laboratories or data points which were excluded are listed for each sample in Table 4. The

outliers are coded in the following manner. LxDy means that results from laboratory x on day

y were detected as outlier. The addition of “Max1”, “Min1” and “Max2” for Grubbs’ test

indicates that the maximum, minimum or the two highest individual values were excluded from

further analysis.

Outliers were found for laboratories 1, 2, 9, 13, 15 and 16.

Table 4 – Laboratories that were excluded as outliers by Cochran’s test or Grubbs’ test

based on individual values, within laboratory replicates and between laboratory mean

values.

Test A B C D E

Grubbs’ Test on Individual Values L9D2Max1

L1D3Max2

L2D1Min1 L13D3Max1 L1D2Min1

Grubbs’ Test on Replicates None

Grubbs’ Test on Mean Values None

Cochran’s Test on Standard Deviation L15 L16

The remaining data sets were then used to calculate a global mean and standard deviation.

2.3 Robust Mean Values and Standard Deviations

After the removal of outliers robust mean values and between-laboratory standard deviations

were calculated using ISO 5725-2. The results are given in Table 5. The number of laboratories

is denoted by N.

Table 5 – Robust mean values (MV), between-laboratory standard deviations (SDb) and

within-laboratory standard deviations (SDw)

Air Permeability

ID MV SDb SDw N

CU CU CU

A 70,25 1,94 1,02 11

B 61,16 2,66 0,67 13

C 9462,92 242,19 141,61 10

D 341,00 13,94 3,80 8

E 1399,03 47,73 8,83 8


2.4 Evaluation of Repeatability and Reproducibility

Based on the robust mean value and the between-laboratory and within-laboratory standard

deviations, repeatability and reproducibility statistics were calculated according to ISO 5725-2.

The results are given in Table 6. The table shows the standard deviation (SD), the limit and the

coefficient of variation (CoV) relative to the global mean value for repeatability and

reproducibility.

Table 6 – Repeatability and reproducibility statistics for air permeability

Repeatability Reproducibility

ID SD Limit CoV SD Limit CoV

CU CU % CU CU %

A 1,02 2,88 1,45 2,11 5,97 3,01

B 0,67 1,90 1,10 2,71 7,68 4,44

C 141,61 400,52 1,50 268,37 759,08 2,84

D 3,80 10,75 1,11 14,28 40,38 4,19

E 8,83 24,98 0,63 48,27 136,53 3,45

2.5 Evaluation of Laboratory Performance (z-Scores)

Based on the robust mean value and the between-laboratory standard deviation the z-scores

were calculated as described in ISO 13528:2015 and are provided in Table 7. In the table fields

marked in orange are z-scores with 2<|z|<3. There were no z-scores with |z|≥3.

Table 7 – Z-Scores for all laboratories and samples

Paper Sample

Lab A B C D E

1 -0,38 0,28 0,17 0,92 1,16

2 0,25 0,69

3 -0,41 1,46

4 -0,79 0,92

5 0,78

6 0,64 1,07

7 1,55 0,91

8 0,96

9 -0,15 -0,32 0,02 -0,12 -0,44

10 0,79 0,19 0,01

11 -1,37 -1,03 -0,27 0,85

12 -0,85 -1,18 1,18 0,35 0,31

13 1,18 -0,37 -0,07 -0,35 -0,16

14 -1,56 -1,05 -1,96 -2,10

15 -1,06 1,68 1,41 0,45

16 -1,85 -0,71 -1,79 0,12 -0,08


3. DATA INTERPRETATION

3.1 Repeatability and Reproducibility

For comparison the repeatability and reproducibility limits as a percentage relative to the mean

values as given in ISO 2965:2019 and as obtained in this study are provided in Table 8.

Table 8 – Comparison of repeatability (r) and reproducibility (R) limits as percentage of

the mean air permeability (AP) reported in ISO 2965, Table 3, with results from this study

ISO 2965 This Study

Paper AP r R ID AP r R

CU % % % %

A 70,25 4,11 8,50

Cigarette Paper 31,75 10,45 11,72 B 61,16 3,11 12,55

Plug Wrap Paper 11171,14 12,74 15,95 C 9462,92 4,23 8,02

Tipping Paper 1006,50 4,38 7,20 D 341,00 3,15 11,84

Tipping Paper 1006,50 4,38 7,20 E 1399,03 1,79 9,76

It has to be taken into account that the testing of air permeability as carried out in this study is

considered as “destructive” so that each individual measurement has to be performed on a

different area of the paper sample. Consequently, repeatability and reproducibility data also

include the product variability, which is likely to differ between the samples used in ISO 2965,

Table 3, and in this study.

The repeatability and reproducibility statistics reported in ISO 2965 agree in general with the

results found in this study, considering that different paper samples with different nominal air

permeabilities have been used.

As expected the repeatability statistics of measurements with the 2×15 mm² measuring head

(sample A) are slightly worse than those for the 10×20 mm² measuring head (sample B), which

may be due to the smaller measured area. Surprisingly the reproducibility statistics are better

for the 2×15 mm² measuring head than for the 10×20 mm² measuring head. But more

importantly the values obtained with the 2×15 mm² measuring head are on average 9,09 CU

higher than those obtained with the 10×20 mm² measuring head for the same paper. This is a

statistically significant difference, which means that laboratories should exercise great caution

when comparing air permeability values obtained with different measuring heads. Further

measurements of this kind may be needed to confirm the result.

Furthermore it has to be noted that seven data sets were coming from one participating

laboratory using seven different instruments but the same operator and the same set of

calibration standards. Thus these sets were not obtained under full reproducibility conditions.

However, eliminating these seven data sets had no substantial effect on the reproducibility

results and also the conclusion regarding the influence of the 2×15 mm² measuring head vs. the

10×20 mm² measuring head remains unchanged. Indeed after elimination of these data sets the

difference between the global mean values for samples A and B becomes even larger

(10,36 CU).

The PTM Sub-Group proposes to add repeatability and reproducibility data for the

measurements with the 2×15 mm² measuring head on conventional cigarette paper and, if the

influence of the measuring head can be confirmed, to include an appropriate note in this respect

in the next revision of ISO 2965.


3.2 Laboratory Performance

As described in ISO 13528:2015, in normal circumstances about 95 % of all z-scores will be in

the range between -2 and 2. Occasionally, absolute z-scores equal to or greater than 2 may be

expected at a rate of about 5 %, while absolute z-scores equal to or greater than 3 will occur

only at a rate of about 0,3 %.

Thus for absolute z-scores between 2 and 3 it is up to the laboratory to decide if these

exceptional values are of importance and require any corrective action or review of the

laboratory procedures. For absolute z-scores of 3 or higher it is strongly recommended that the

laboratory investigates the reasons for the deviation and derives appropriate actions from these

investigations.

In the present study, only laboratory 14, had z-scores below -2, which is not unusual in a data

set of 52 values. However, all z-scores of laboratory 14 are between -1,05 and -2,10 which

means that laboratory 14 generally measures lower permeability values than the other

laboratories. There were no absolute z-scores above 3, which indicates that the participating

laboratories have established reliable procedures for measuring air permeability.

4. REFERENCES

• ISO 2965:2019, Materials used as cigarette papers, filter plug wrap and filter joining

paper, including materials having a discrete or oriented permeable zone and materials

with bands of differing permeability. Determination of air permeability.

• ISO 3402:1999, Tobacco and tobacco products – Atmosphere for conditioning and

testing.

• ISO 187:1990, Paper, board and pulps; standard atmosphere for conditioning and testing

and procedure for monitoring the atmosphere and conditioning of samples.

• ISO 5725-2:1994, Accuracy (trueness and precision) of measurement methods and results

- Part 2: Basic method for the determination of repeatability and reproducibility of a

standard measurement method.

• ISO/IEC 17043:2010, Conformity assessment – General requirements for proficiency

testing.

• ISO 13528:2015, Statistical methods for use in proficiency testing by interlaboratory

comparison.

• CORESTA Technical Report PTM-065: Recommendation of Measurement Area for Air

Permeability Determination of Super-Slim Cigarette Papers, CORESTA PTM Sub-Group,

September 2016.


5. APPENDICES

APPENDIX A – Protocol

The protocol is reproduced in its original form. Minor typographical errors were corrected and

e-mail addresses were removed.

Protocol for the 1st Collaborative Study on Air Permeability

Note: The purpose of a Collaborative Study is to assess r&R data of a method and it can also

be used by laboratories to evaluate their performance with respect to this method in comparison

to other laboratories. Therefore it is important that you strictly follow ISO 2965.

1. Preparation of Samples

1.1 Where possible, all measurements should take place in an environment according to ISO

187: testing atmosphere: (23 ± 1) °C and (50 ± 2) %rH. If this is impossible, make a note in the

comments column.

1.2 You will find 5 separate samples called from "Sample A" to "Sample E". For each sample,

1 sheet with 3 series of strips of paper for day 1, day 2 and day 3. Inspect prepared samples

visually to avoid measurements of damaged samples.

1.3 Samples should be conditioned for 48 hours regarding to ISO 187 prior to the measurement.

1.4 To avoid changes in sample moisture, samples should be transferred from the conditioning

environment to the instrument in tightly closed plastic boxes.

2. Instrument Setup

2.1 Prior to a set of measurements, all instruments have to be checked and calibrated.

2.2 According to ISO 2965, calibration and instrument settings as used in your laboratory.

2.3 All specific instrument information should be recorded in the instrument configuration sheet.

3. Measurement Plan

3.1 Measurements on each sample (and each instrument, if applicable) should take place on 3

consecutive days (e.g. Monday to Wednesday).

3.2 On each day determine the air permeability of each of the samples from A to E by

performing 20 individual measurements in your laboratory.

3.3 On each day a new series of sample strips shall be used. You have 3 series.

3.4 Do not touch areas to be measured with fingers.

3.5 Never measure on the same area twice.

3.6 Visually check the samples.


Example of sample A with 3 series of paper strips, 1 per day of measurement.

3.7 Samples should be as flat as possible.

3.8 Please use a new sample strip for each measurement series and realize 20 measurements per

day (you have 20 strips of paper)

4. Recording of Results

Please record the results using the yellow cells as defined in data collection sheets Day 1, Day

2 & Day 3.

It is not permitted to make any changes to the data recording sheets. Results, which are not

reported correctly, may be disregarded in the collaborative study.

E-Mail completed spreadsheets (as Excel file) to:

Bernhard Eitzinger

Philippe Le Men

Results must be submitted until August 31st, 2019.


APPENDIX B – Data Summary

Appendix B.1: Mean values of air permeability over all laboratories and days, outliers included. The mean value is given in CU.

Day 1 Day 2 Day 3

Lab A B C D E A B C D E A B C D E

1 69,7 61,2 9419,6 354,0 1448,7 70,3 62,9 9498,1 353,2 1461,3 69,6 61,6 9596,2 352,8 1453,3

2 70,4 62,5 70,0 62,6 71,7 63,4

3 70,0 65,0 69,0 64,6 69,4 65,5

4 69,2 64,0 69,2 63,3 67,8 63,6

5 9779,6 9379,6 9795,5

6 71,0 64,2 72,1 63,8 71,3 64,0

7 71,5 64,0 73,8 63,4 74,4 63,3

8 9559,1 9748,1 9777,5

9 70,4 60,6 9364,5 344,5 1389,7 71,2 60,0 9535,9 340,9 1378,9 69,1 60,2 9506,4 332,6 1365,1

10 72,0 60,4 343,0 71,4 61,8 338,3 71,9 62,7 342,2

11 57,3 9241,6 340,7 1452,2 58,5 9219,6 337,7 1428,4 56,7 9180,8 333,4 1438,6

12 68,7 57,3 9761,2 345,5 1413,7 68,5 58,1 9743,3 345,1 1416,3 68,6 58,6 9738,9 347,1 1412,1

13 71,6 60,0 9419,2 336,1 1390,5 74,8 59,8 9737,6 335,3 1394,4 71,2 60,7 9294,1 337,1 1389,7

14 58,1 9116,1 313,8 1305,6 57,1 9257,5 312,6 1291,9 55,9 9254,6 314,8 1299,2

15 55,5 10283,6 368,9 1420,2 58,9 9724,5 358,3 1419,3 60,6 9602,6 354,8 1421,8

16 65,2 58,8 9160,7 333,6 1388,5 67,5 59,7 9055,5 342,3 1384,8 67,3 59,4 8869,1 352,2 1412,7


Appendix B.2: Standard deviations of air permeability over all laboratories and days, outliers included. The standard deviation is given in CU.

Day 1 Day 2 Day 3


1 4,46 2,16 792,59 8,48 27,98 4,02 2,79 675,09 9,06 25,98 4,59 1,98 714,10 8,86 28,62

2 3,80 3,02 1,19 1,46 2,38 2,21

3 2,16 1,46 2,14 2,46 1,76 3,10

4 1,25 1,88 1,72 1,74 2,52 2,21

5 419,70 605,09 728,63

6 3,88 3,92 3,32 1,46 4,05 2,60

7 2,71 1,83 3,05 2,47 3,74 3,08

8 594,20 638,49 715,23

9 3,85 2,25 706,61 9,94 21,88 4,92 1,18 573,49 10,46 32,51 4,05 2,12 775,12 9,29 20,99

10 3,78 1,99 9,36 4,00 2,64 8,91 4,78 2,14 10,02 0,00

11 1,65 803,39 8,44 37,21 2,53 588,86 9,22 30,44 1,05 859,38 9,41 27,25

12 1,52 1,93 503,07 8,37 23,92 1,43 2,76 541,90 6,51 16,08 2,01 1,50 394,59 7,30 19,46

13 2,90 2,82 478,66 8,93 32,51 3,43 1,96 526,44 8,83 19,22 1,91 1,77 709,66 10,51 19,30

14 2,14 652,46 6,73 27,06 2,12 639,22 7,31 33,25 1,73 605,31 10,45 26,55

15 1,89 246,73 6,93 24,96 3,63 533,68 9,00 30,86 1,98 479,44 7,43 40,63

16 4,57 2,27 717,40 9,23 27,05 4,17 2,21 495,38 8,17 26,49 3,05 1,56 654,23 8,99 32,43


Appendix B.3: Mean values of air permeability over all laboratories and days, outliers according to Grubbs’ test on individual values

excluded. The mean value is given in CU. Highlighted values have changed due to outlier elimination.

Day 1 Day 2 Day 3


1 69,7 61,2 9419,6 354,0 1448,7 70,3 62,9 9498,1 354,6 1461,3 68,5 61,6 9596,2 352,8 1453,3

2 70,4 63,0 70,0 62,6 71,7 63,4

3 70,0 65,0 69,0 64,6 69,4 65,5

4 69,2 64,0 69,2 63,3 67,8 63,6

5 9779,6 9379,6 9795,5

6 71,0 64,2 72,1 63,8 71,3 64,0

7 71,5 64,0 73,8 63,4 74,4 63,3

8 9559,1 9748,1 9777,5

9 70,4 60,6 9364,5 344,5 1389,7 70,4 60,0 9535,9 340,9 1378,9 69,1 60,2 9506,4 332,6 1365,1

10 72,0 60,4 343,0 71,4 61,8 338,3 71,9 62,7 342,2

11 57,3 9241,6 340,7 1452,2 58,5 9219,6 337,7 1428,4 56,7 9180,8 333,4 1438,6

12 68,7 57,3 9761,2 345,5 1413,7 68,5 58,1 9743,3 345,1 1416,3 68,6 58,6 9738,9 347,1 1412,1

13 71,6 60,0 9419,2 336,1 1390,5 74,8 59,8 9737,6 335,3 1394,4 71,2 60,7 9181,4 337,1 1389,7

14 58,1 9116,1 313,8 1305,6 57,1 9257,5 312,6 1291,9 55,9 9254,6 314,8 1299,2

15 55,5 10283,6 368,9 1420,2 58,9 9724,5 358,3 1419,3 60,6 9602,6 354,8 1421,8

16 65,2 58,8 9160,7 333,6 1388,5 67,5 59,7 9055,5 342,3 1384,8 67,3 59,4 8869,1 352,2 1412,7


Appendix B.4: Standard deviation of air permeability over all laboratories and days, outliers according to Grubbs’ test on individual

values excluded. The standard deviation is given in CU. Highlighted values have changed due to outlier elimination.

Day 1 Day 2 Day 3


1 4,46 2,16 792,59 8,48 27,98 4,02 2,79 675,09 6,52 25,98 2,51 1,98 714,10 8,86 28,62

2 3,52 1,00 1,68 0,64 2,71 2,39

3 2,50 1,58 2,26 1,17 1,51 2,04

4 1,69 2,18 0,59 1,31 3,10 2,73

5 477,59 594,86 753,19

6 3,60 1,05 4,53 0,71 3,83 1,96

7 3,89 1,62 2,42 2,09 3,63 3,03

8 380,41 803,14 342,14

9 5,25 1,79 759,32 8,77 22,11 5,38 1,67 253,27 11,27 29,30 2,64 3,03 1104,40 12,02 6,91

10 3,85 2,59 8,54 3,56 3,76 9,94 5,83 3,09 7,00

11 1,08 820,03 4,99 26,66 3,03 612,38 13,35 23,92 1,69 361,98 8,64 17,54

12 1,52 1,93 503,07 8,37 23,92 0,53 2,76 541,90 6,51 16,08 2,01 1,50 394,59 7,30 19,46

13 4,50 2,58 580,20 11,61 29,62 3,67 0,93 59,38 7,09 12,24 1,31 1,87 600,95 10,64 25,80

14 2,24 202,67 6,87 32,24 2,83 675,40 4,80 45,85 1,88 763,89 6,97 37,50

15 1,08 155,93 9,06 29,62 1,34 156,01 9,95 31,00 2,27 401,26 8,67 40,03

16 5,72 1,79 858,74 14,96 26,11 6,38 3,24 582,86 8,72 29,11 2,56 1,53 790,02 7,46 15,78


APPENDIX C – Diagrams

Appendix C.1: Mean value and confidence interval for air permeability over all

laboratories for Sample A (cigarette paper).


laboratories for Sample B (cigarette paper).

50,0

55,0

60,0

65,0

70,0

75,0

80,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID

Sample A (Mean95% CI)

50,0

55,0

60,0

65,0

70,0

75,0

80,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID

Sample B (Mean95% CI)



laboratories for Sample C (plug wrap paper).


laboratories for Sample D (tipping paper).

7000,0

7500,0

8000,0

8500,0

9000,0

9500,0

10000,0

10500,0

11000,0

11500,0

12000,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID

Sample C (Mean95% CI)

300,0

310,0

320,0

330,0

340,0

350,0

360,0

370,0

380,0

390,0

400,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID

Sample D (Mean95% CI)



laboratories for Sample E (tipping paper).


laboratories for Sample A (cigarette paper) for Day 1.

1200,0

1250,0

1300,0

1350,0

1400,0

1450,0

1500,0

1550,0

1600,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID

Sample E (Mean95% CI)

50,0

55,0

60,0

65,0

70,0

75,0

80,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID

Day 1 / Sample A (Mean95% CI)



laboratories for Sample B (cigarette paper) for Day 1.


laboratories for Sample C (plug wrap paper) for Day 1.

50,0

55,0

60,0

65,0

70,0

75,0

80,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID

Day 1 / Sample B (Mean95% CI)

7000,0

7500,0

8000,0

8500,0

9000,0

9500,0

10000,0

10500,0

11000,0

11500,0

12000,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID

Day 1 / Sample C (Mean95% CI)



laboratories for Sample D (tipping paper) for Day 1.


laboratories for Sample E (tipping paper) for Day 1.

300,0

310,0

320,0

330,0

340,0

350,0

360,0

370,0

380,0

390,0

400,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID

Day 1 / Sample D (Mean95% CI)

1200,0

1250,0

1300,0

1350,0

1400,0

1450,0

1500,0

1550,0

1600,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID

Day 1 / Sample E (Mean95% CI)






50,0

55,0

60,0

65,0

70,0

75,0

80,0

85,0

90,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID


50,0

55,0

60,0

65,0

70,0

75,0

80,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID







7000,0

7500,0

8000,0

8500,0

9000,0

9500,0

10000,0

10500,0

11000,0

11500,0

12000,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID


300,0

310,0

320,0

330,0

340,0

350,0

360,0

370,0

380,0

390,0

400,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID







1200,0

1250,0

1300,0

1350,0

1400,0

1450,0

1500,0

1550,0

1600,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID


50,0

55,0

60,0

65,0

70,0

75,0

80,0

85,0

90,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID







50,0

55,0

60,0

65,0

70,0

75,0

80,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID


7000,0

7500,0

8000,0

8500,0

9000,0

9500,0

10000,0

10500,0

11000,0

11500,0

12000,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID







300,0

310,0

320,0

330,0

340,0

350,0

360,0

370,0

380,0

390,0

400,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID


1200,0

1250,0

1300,0

1350,0

1400,0

1450,0

1500,0

1550,0

1600,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Air

Pe

rme

ab

ilit

y [C

U]

Laboratory ID


technical report 1st collaborative study on air

Documents