1, F, To ALL MEMBERS OF PACKAGI14G FROM 14. Foster. GROUP. REF NF/UBM/85. DATE 29th December, 1977. B.A.T GROUP SCHME FOR ODOUR AND TAINT DETECTION IN TOBACCO PRODUCTS. As referred to during the last Packaging Group meeting the enclosed document is G.R. & D.C.'s response to an urgent Tobacco Group need for a reliable safeguard against the occurrence of unwanted odour and/or taint on Associate Companies'tobacco products in the market place. The document clearly states the options available to our Operating Companies, one of which is a collaborative system, that we at G.R. & D.C. believe in the longer term would be of most benefit to Group members. Whilst it is likely that the proposed schcme could be operated more easily by,Companies which have their own laboratories, the availability of laboratories is not considered to be an essen- tial pre-requisite to the successful uperation of the scheme. The document is now in final draft form. If you have any constructive comments please send them to me before 16th January 1978, for consideration. Shortly after this date the document will be made available for distribution to Associates. c.c. D.A. Baxter, Esq. ) B A T (U.K. & E) Woking for info: A.R. Cousins, Esq. ) * , (-n (-n (Dl% Cr,% c>~ 13ATCo document for Legal Services : Health Canada 20 October 1999 BAT GROUP SCHEME FOR THE DETERMINATION OF UNWANTED ODOUR ON MATERIALS AND TAINT ON CIGARETTES 1. STATEMENT OF THE PROBLEM There is always a risk that constituents of cigarette packaging may migrate to the contained cigarettes. The constituents could be volatile, or in the case of material in physical contact with the cigarettes they could be non-volatile. The quantity and quality of such contaminants will determine whether or not the taste of the cigarettes concerned becomes perceptibly altered. In the case of low delivery cigarettes, the risk of the cigarette becoming tainted in this way has been shown to be significantly greater than that for medium and high delivery cigarettes. With the trend towards low delivery cigarettes, the problem of tainting is already showing a significant increase. A comprehensive approach for detecting the presence of unwanted odour or/and taint is provided in detail in this document. It is designed for use by Associate Companies, particularly, but not exclusively, for application to low C71% delivery brands under development. BATCo document for Legal Services : Health Canada 20 October 1999 2 - 2. AN OUTLINE OF THE PROPOSED SCHEME 2.1 Introduction In seeking to provide a useful procedure for dealing with unwanted odour and taint* problems it has to be acknowledged that the Group experience of these situations is currently limited and geographically scattered. However, when such situations do arise they 'are potentially damaging and costly. The procedures which are described here incorporate features which will enable the learning value of those experiences to be maximised. Thereby it is intended that a more efficient and precise instrument for controlling odour and taint will evolve. Progress towards a fuller understanding of the problem and better control procedures will be most effec- tively served if some coordination of activities in this area can be established. The proposal embodies a collaborative research and development enterprise in which all Associate Companies are invited to participate. However, the odour and taint panel procedures described may, of course, be adopted for local application without any such commitment. OdOur is defined as a sensation produced by the interaction tp certain volatile materials with the scnsory organs of the notei. ON Taint is any type of off-sensation which can be describe asso, foreign to tobacco. C:) Co BATCo document for Legal Services: Health Canada 20 October 1999 - 3 - It should be noted that, whilst the overt emphasis in these notes is on packaging materials, particularly of the printed sort, the procedures can be readily adapted to allow consideration of other materials. 2.2 The Panel Approach The procedures for the assessment of odour on materials and taint on cigarettes involve two groups of people - the odour panel and the taint panel respectively. in the case of odour assessment the panel's task has been derived from a consideration of a method developed by and extensively used by Cadbury Limited. Because of their considerable experience with certain packaging materials they are able to determine a critical level of odour beyond which a material may be rejected as unacceptable. Over time it has become possible, for the Cadbury panel, to associate a given odour level with a potential product taint. Unfortunately we do not yet have sufficient experience with the odour/taint.interaction to be able to specify an equivalent subjective value to be adopted in a simple transposition of the Cadbury method for use with cigarette packaging materials. Since the taint potential of any odour can only be determined from experience, the procedures described for the development of our awareness of the odour/taint 0`1 CD N10 BATCo document for Legal Services: Health Canada 20 October 1999 - 4 - interaction seek to accelerate our acquisition of such information. It is necessary, therefore, to consider odour and taint separately and to employ separate panels to do so. 2.3 The Local Administrator In the procedures described for the establishment and operation of the odour and taint panels, a key function is served by a person, or persons, nominated as the local administrator(s). The local administrator is charged with the management of the panels and the evaluation of the data. In the case of the odour panel, the administrator gradually acquires an understanding of the subjective evaluation of odours and the description of odours by panel members. He/she is also involved in the basic training of the panel. In the case of the taint panel, the administrator additionally acts as a referee, determining from the panel's responses whether or not a possible taint problem exists and what sort of further action is required. With the benefit of experience, the administrator will seek to integrate his experience of the odour and taint procedures and evolve more precise and efficient assessments. To assist with this process, the administrator may choose to collaborate in the interactive Group elements of the scheme through which it is hoped a more comprehensive (,-n O*N ON BATCo document for Legal Services: Health Canada 20 October 1999 N - 5 - and speedier understanding of the odour/taint problem will be achieved. 2.4 Routine and Research It is, of course, open to Associate Companies to choose not to participate in the exchange of experience suggested here. In this case they may decide to develop their own procedures, whether or not based upon the protocols described below. If, on the other hand, participation is agreed, then the collaboration will allcw both routine and research elements to be incorporated int2 the exercise. The Associate Companies will for themselves decide on the appropriate rate of odour and taint testing, but it is suggested that particular attention be paid to new produc--s, particularly low delivery cigarettes, and to new packagin7 materials. Such testing will be designated as routine. Under research testing it is envisaged that, from time tc time, special panel evaluations are made. These may be suggested to the participating administrators by the Cen--ze, to investigate, for example, the response of different panels to the same process chemical. In this way the Group effort will have access to an international assessment panel through which the universality of odour and taint CD judgements may be determined. 0-% BATCo document for Legal Services -. Health Canada 20 October 1999 6 - 2.5 Research Support Recognising that thelroblems associated with odour and taint are becoming increasingly important, a programme of more fundamental research into the evaluation of odour and taint is being initiated at G.R. & D.C. This will provide an important adjunct to the local and immediate consideration of the subject within the Associate Companies and can be expected to consolidate efforts in this area. 2.6 Collaboration The scheme for odour and taint assessment presented here is distinctive insofar as it proposes a collaborative venture between Associate Companies and the Centre. It is because of the unpredictable rate and type of occurrence of odour and taint problems that it is believed that only through collaboration can the best procedures be eventually determined. The evaluation of both odour and taint require development and, for the latter particularly, a comprehensive feedback and information sharing mechanism has been designed. It is, however, important to stress that the anonymity of information supplied by participating Companies will be preserved by the Group administrators for the project. The apparently elaborate reporting mechanism has been designed to enable the fullest information to be collated. C::) Periodically the local administrator (s) will be requested (-n (-.n to supply copies of the summary data sheets to the Group ON BATCo document for Legal Services : Health Canada 20 October 1999 II - 7 - Project Administrator and to comment generally on experience to date. In addition, particular problems may be referred to the Group Administrator for advice. Feedback from the Centre will include a digest of all participants' experience, further guidance on the operation and development of the methods, and supporting information from the research programme at G.R. & D.C. 3. INTRODUCTION TO PROCEDURES 3.1 The Odour Panel The panellists' assessment of odour is conducted in two stages. Firstly an estimate of the level of odour is made. Secondly the quality of the odour is described. During the early stages of the operation of the odour/taint assessment scheme it will not be possible to correlate a level and type of odour with a subsequent unacceptable taint, therefore the aim wil'l be to familiarise the panellists. with typical odours and the reliable assessment of their level. 3.2 The Taint Panel The taint assessment procedure is more complex and more highly structured. In this case, panellists are required to detect differences, which may or may not suggestc-,,j C71- QIN U-J BATCo document for Legal Services : Health Canada 20 October 1999 - 8 - a tainting, between packaged and non-packaged cigarettes. Even if a sufficient difference is noted this may not be unacceptable. The acceptability criterion is applied by the administrator and/or someone with an intimate knowledge of the product. It is important that the adn.iinistrator is made aware of all the acceptability decisions made on referred products in order that he may gradually refine his judgement of the panel responses. 3.3' Advice to Suppliers Notes for the guidance of suppliers in relation to odour and taint have been prepared and all relevant suppliers should be advised of the procedures being adopted by the Associate Company in this connection. 4. DETERMINATION OF ODOUR AND TAINT LEVELS BY PANEL TECHNIQUES 4.1 Gdour Panel The odour of packaging materials can affect the odour and taste of packed cigarettes. The following procedure is recommended for employment in cigarette factories and to form the basis of similar methods in the suppliers' factories. 4.1.1 Selection_of_22D21 nembers, (a) Members must be sensitive to the cdours being assessed, and should be drawn from 41- BATCo document for Legal Services : Health Canada 20 October 1999 - 9 - a wide range of occupational activities and age. (b) People continually exposed to the types of odour requiring detection should be excluded owing to fatigue and possible loss of sensitivity. (c) People suffering from chronic or recurrer respiratory infection should be excluded. (d) The size of the panel should'comprise a minimum of six members drawn in rotation from a pool of 12 - 15. Note: More specific guide lines to assist in the selection and training of panel members will be provided. 4.1.2 0dour_p2D2j_2pfEajion (a) The odour panel should consist of at least six members. (b) The odour assessment should be conducted in an odour-free room, -ideally with individual members of the panel screened from each other to avoid undue influence from one panel member to another. (c) The following six-point scale should be used to assess the level of odour, with the sample C-M placed close to the nose: C__j UN ON (-n BATCo document for Legal Services: Health Canada 20 October 1999 - 10 - 0 No odour 1 Barely perceptible odour 2 Slight odour 3 Moderate odour 4 Strong odour 5 Very strong odour (d) Each panel member should record his assess- ment of the odour level detected and comment as to its nature (in technical terms or common domestic association) on the written form provided (see Appendix 1). It should be noted that the panel member is not bein2 asked if the sample is acceptable for odour, but merely to report the level and type assessed.. The completed pro formae are returned to the panel administrator (see Section 4.1.5). 4.1.3 SamL3j2_pEfsfntation (a) Samples for assessment should be taken from the pallet, load, reel, case or parcel and wrapped in odour-acceptable (rated 0 on the proposed scale) aluminium foil or waxed glassine (the latter for short periods of time). In the case of hinge lid carton CD blanks or soft cup labels, care should be taken to avoid cutting or tearing, as these C7\ BATCo document for Legal Services : Health Canada 20 October 1999 - il - should be tested in their end-use condition. In the case of samples taken from unheated stores, the wrapped sample should be stored at normal room temperature conditions for a minimum period of one hour prior to assessment. (b) The wrapped samples should be taken to the odour panel room and given to each member for assessment. Sufficient samples from reeled materials should be taken to allow the panel member to place the nose intmediately next to the surface of the packaging material requiring comment. This is best achieved by using at least two layers and placing the nose between them. It is important to have scores and comments about both sides of the material as odour from the printed surface can, through contact or absorption, affect the reverse surface. The odour level of the reverse surface is important since it forms the inside of the pack and is in close proximity to the cigarettes. (c) In order to reduce the possible effects of olfactory fatigue it is recommended where possible to limit the maximum number of samples to ten per session, and to vary the type of packaging materials for consecutive samples. BATCo document for Legal Services : Health Canada 20 October 1999 - 12 - 4.1.4 instructions_to 32Liel_Ei2mbers The following instructions are given to each panellist on each occasion together with a rcsponse sheet: "Here is a sample of packaging material. We would like you to assess it for odour in the following way: open the jar/packet containing the sample. You will be advised which surface to consider. Bring the sample close to the nose, almost touching, and sniff using strictly one inhalation within which you may use as many short sniffs as you wish (though you are, advised not to use one large sniff for the whcle inhalation). In the case of materials with a printed and unprinted side, you will be required to always sample 10 the unprinted side first. This is done by placing the sheets at either side of the nose with their unprinted faces innermost. You are asked to assess each sample in two separate ways: STEP 1 - Take a single inhalation of sniffs and express any od our you detect using the following six-point C__ scale: Q^ ON CC BATCo document for Legal Services : Health Canada 20 October 1999 0 1 2 3 4 5 NO BARELY SLIGHT MODER- STRONG VERY ODOUR PERCEP- ODOUR ATE ODOUR STRONG TIBLE ODOUR ODOUR ODOUR Record your judgement on the response sheet provided. STEP 2 - Take further sniffs and describe the QUALITY of the odour using as many technical terms or common domestic associations as come to mind. Always leave at least two minutes between sniffing of samples, otherwise your acuity will be disturbed." 4.1.5 The role of the administrator (a) The administrator is responsible for the operation of the panel. Thus, he/she will prepare and distribute samples, collate the elicited responses, and evaluate the data obtained. (b) The administrator has to decide, on the basis of the panel evaluation, whether the recorded responses represent an acceptable level and quality of odour. (c) Clearly, and particularly in the early operation of the panel, this will be a difficult decision to make since it is not possible to set a criterion value above which the group mean evaluation unequivocally defines on unacceptable odour level. Equally, the (-n Q^ GN BATCo document for Legal Services : Health Canada 20 October 1999 - 14 - quality of any odour is not readily deter- mined as acceptable or otherwise. When the evidence suggests that the odour rating or the type of odour is unacceptable, the matter should be referred to the suppliers and/or the taint panel (see Section 4.2). (d) In the case of mild or low delivery cigarettes which may be particularly sensitive in absorb- ing odour molecules, it'is recommended that the level of taint should be estimated. (e) Further notes for the guidance of administra- tors in determining the acceptibility of o6our will be Drovided. 4.2 Taint Panel If cigarettes are packed in odorous cartons, the odorous molecules in the vapour state can be absorbed by the cigarette paper or the tobacco. By absorbing the odorous molecules from the air space the vapour present is disturbed and more odorous molecules are released -from the printed materials to be abs orbed by the cigarettes. Eventually, tainting can occur, detectable as an odour from the cigarette or producing a taint when smoked. It is important, however, to note the possibility of taint arising in the absence of odour on packaging materials. To estimate the level of tainting, ,C::) a taint panel is used. 01N r\-) CD BATCo document for Legal Services : Health Canada 20 October 1999 - 15 - 4.2.1 Selection of_pap2l_Ln2mbers (a) Persons selected for panel membership require no special qualifications and should be drawn from as wide a range of occupational activities, age, and usual cigarett e consuji,.ption as possible. No training is required beyond familiarisation with the evaluation proccdure to be used. (b) Persons with an occupational involvement in taste, flavour and other subjective evaluations of cigarettes should be excluded from participa- tion. (c) At least ten, and preferably twenty, panellists are required for each assessment. To avoid fatigue and boredom, panellists should be drawn in rotation from a larger pool. 4.2.2 Taint-panel-operation (a) The taint panel should comprise ten to twenty members. (b) The pair-comparison procedure should be con- ducted in an odour-free environment, but may be undertaken at the individual's point of work where conditions allow. (c) Each panel member is supplied with two cigarettes on any one occasion and required (-r C-F to smoke them according to the instructions QJ_ 0- given (see Section 4.2.4). 0~ rII_ BATCo document for Legal Services : Health Canada 20 October 1999 - 16 - (d) For each pair comparison test, each panel member will complete the pro forma supplied (see Appendix 2). Completed pro formae are returned to the panel administrator (see Section 4.2.5). 4.2.3 Sampl2_L)EEE2ntation (a) Samples for panel assessment are obtained from three sources: U) from the production line immediately before packing, having been stored at an appropriate temperature* for an appropriate time (either x or y weeks) in air-tight odour-free con- tainers; (ii) from packages after x weeks' storage*; (iii) from packages after y weeks' storage*. (b) The pair comparison procedure on any one occasion is conducted with source W material and either source (ii) or source (iii) material. That is, a cigarette which has never been packaged is paired with one packaged for either a short (ii) or a long (iii) period. Storage temperature should be relevant to marl;et concliticns, where x represents a realistic short-term peri.od and y represents a long-term period equivalent to the anticipated ON maximum market life. ON rII-) N-) BATCo document for Legal Services : Health Canada 20 October 1999 - 17 - (c) Because the panel is requir6d only to detect differences, it is not necessary for the individual cigarettes to be coded. Each cigarette is placed in its own screw-top jar, and it may be a wise precaution to label the jars according to the number of the test, panellist code number, date of test, etc., but not to identify the source of the cigarett contained within. (d) Before being placed in odouy-free air-tight containers, each tobacco rod is marked off in thirds using a soft (5B) pencil. 4.2.4 Instructions to_p2nf.I_Lr,2mbers The following written instructions are given to each panellist on each occasion, together with a response sheet: "Here is a pair of cigarettes which we would like you to smoke. "You will observe that the tobacco rod of each cigarette has been marked off into three equal lengths. This is because we want you to record on three occasions during your smoking any differences which there may be between these cigarettes. "We want you to report any differences which you may notice: BATCo document for Legal Services : Health Canada 20 October 1999 here ----------- I ----------------------- here ------------------------------ ------ and here --- ----------- ------ "In making your assessment, you are asked to indicate on the response sheet the magnitude of any Is difference which you observe. "You are not required to make judgements on the quality of any differences. "To express any differences, use the following six-point scale: 0 1 2 NO BARELY SLIGHT DIFFER- PERCEP- DIFFER- ENCE TIBLE ENCE DIFFER- ENCE Procedure "Smoke both the cigarettes at the same time, alternating puff-taking from each and trying to keep the unburnt rod lengths as equal as possible. Try to make your assessment of difference for each one-third of the cigarettes, at about the midpoint of your smoking through those thirds. The assessment, in ON BATCo document for Legal Services : Health Canada 20 October 1999 - 19 - the form of a number from 0 to 5 for each third, should be entered on the response sheet. Please return your response sheet and empty jars to (administrator)". 4.2.5 The role of the administrator (a) The administrator is responsible for the operation of the panel. Thus he/she will prepare and distribute the samples, collate the elicited responses, and evaluate the data obtained. (b) The panel is, in effect, simply asked: "Is there a difference. between these two cigarettes?"., The administrator has to decide, on the basis of the panel evaluation, whether any recorded differences are sufficient to warrant further evaluation of the product/packaging interaction. (c) Clearly, and particularly in the early operation of the panel, this will be a difficult decision to make since it is not possible to set a criterion value above which the group mean difference assessment unequivocally defines a sufficient difference. W If a sufficient difference is determined, such a difference may yet be acceptable. It is not the role of the panel to assess the CD C7\ C7\ r1l) U7 BATCo document for Legal Services : Health Canada 20 October 1999 I - 20 - acceptability of the difference. Accepta- bility is assessed by persons with an inti- mate knowledge of the product, probably the product developer(s), and perhaps the administrator. (e) Further notes for the guidance of administra- tors in determining the sufficiency of difference will be provided. 4.3 Further Notes for the Guidance of Administrators Frcm time to time administrators will receive further notes tc assist them in developing improved administrative procedures, more reliable data analysis methods, etc. Through the network of collaborating companies, experience will be shared. Clearly the 'person designated as a local administrator is confronted with a number of potentially difficult decisions to make. How best is the panel trained to respond to odour? What is a sufficient difference from the taint panel to warrant further action? Assistance with these questions will 01 be provided but it is to be hoped that local administrators will be motivated to attempt their own solutions within the overall framework of the scheme. To assist with the collating of data from the panels, data surmary sheets have been designed (see Appendices 3, 4). These will form the basis of the administrator's own data - storage system and copies will be requested, periodically, by the Group project administrators in order that information Q^ C7\ on varicus local experiences may be compared. ON N.) ON BATCo document for Legal Services: Health Canada 20 October 1999 5. ADVICE TO SUPPLIERS 5.1 I-troduction The following should form the basis of information on odour and taint supplied to materials suppliers. I 5.2 BIT Procedures for the Control of Odour of Printed Packaging by Suppliers 5.2.1 Intention Residual odours from printed packaging matcrials can produce taint problems in packed cigarettes. We are, therefore, anxious to ensurc that an accepLable, o4-our level is achieved in all printed wrapping materials supplied to us. However, at present we are unable to specify what level or qualities of odour are considered acceptable. We therefore wish suppliers to always ensure that printed mazerial of a minimal odour level is delivered to us. In order to ensure a satisfactory level of odour in our pr--nted wrap-ping materials we propose to monitor samples as prescribed in the attached procedure. Suppliers may wish to adopt similar techniques to ensure that our standards are met. 5.2.2 Quality_Ln~intenance Please note that, although in this document we are principally concerned with odour, we request ON ON BATCo docurnent for Legal Services : Health Canada 20 October 1999