Tobacco Documents Online

^ II a PHIiLIP MORRIS U. S. A. INTER-OFFICE C0,°,RESPONDENCE 1 J • RICMMDN0. YIRBINIA OCT?,9'1QRq. Mr. L. F. Meyer ate: October 21, 1982L . B. Goodman subject:.. EFFECT OF REDUCED DILUTION ON TAR,DELIVERY TO A SMOKER In connection with the task force "Pandoras,"'you raised the question of what might happen to deliveries to the smoker when he partially covers the dilution holes. Would the reduced di~lution cuase him to take sufficiently smaliler puffs to result in the same deliveries as if he took larger puffs at the full dilution level? This memo describes the work that was done to answer your question. The first step consisted of modelling the CIi puff para- meters at normal and partially occluded dilution levels for the nine cigarette brands examined by J. Nepomuceno et. al. The existing Simulator APL model gave good agreement with the analyzed data for the five least diluted brands, but started deviating at hiigher dilutions. The Cambrid'ge Box was excluded from all analyses due to its small changes in dilution and delivery. The results were as follows: TABLE I Original FTC1 Dilution Dil!ution (%) Tar m/ci t.) Model TlarZ (mg/cigt.) w/blockage (%) FTC Tarl' m/ci t.) Model TarZ m ci t.) Marlboro KS 10 16.3 15.8 6 16.6 16.4 B&H Multifilter 23 12.7 12.4 15 13.7 13.0 Menthol KS Parliament Lts. 32 10.4 10.6 20 11.5 11.5 100's , Marlboro Lights 25 10.2 10.4 14 10.7 11.1 KS Va. Slims Lights 41 9.8 8.6 28 10.3' 9.6 100"s Merit KS 301 7.9 7.8 18 8.8' 8.5. Merit Ultra 55 4.5 4.5 36 7.0 5.6 Lights Cambridge 80 1.0 0.8 64 2.5 1.2 KS Softpack Cambridge KS 95 0.0 - 92' 0.3 Box 1CI data from J. Nepomuceno. 2Projected from Simulator APL model.

_„ Page 2 Mr. L. F. Meyer October 21, 1982 The next step was to gather information on the correlation between dilution changes and smoker parameters from past Simulator studies. A total of fifty-eight cigarette models that have been smoked on profile recorders were entered in the computer for regression analysis of puff . parameters, cigarette RTD and dilution. The slope for the best fit line was calculated for volume, flow and dilution versus cigarette RTD and for volume and duration versus dilution. An example can be seen in Figure 1. These slopes were themused for calculating changes in puff parameters for the decreases in dil'ution shown in Table I. It was also assumed that the cigarette RTDichanged proportionally with the dilution. The theoretical changes in puff volume, flow and duration were then applied to the puff parameters measured in our General Data collection and/or more recent studies of low delivery cigarettes. The deli'very to the smoker was predicted with the Simulator APL model for the two levels of input para- meters,on each cigarette brand using the July 1982 CI report as calibration: data. The projecte6tar deliveryes are shown in Table IIi along with the amount of increase above CI tar deliveries. Table II Full Dilution HSS Tar (mg) - Increase over CI % Marlboro KS 24.3 54 B&H Multifilter Menthol KS 20.6 66 Parliaments Lights 100's 14.2 31 O Red'uced Dilution ~ C.~ Increase ~ Dilutio (%)' n HSS Tar m ) over CI (q N Cn N 6 24.4 54. 15 20.9 69 Gd 20 .14.7 36 (continued)

Page 3 Mr. L. F. Meyer October 21, 1982 TABLE II (continued) Full Dilution _ Reduced Dilution HSS Tar (mg) Increase over CI W ~ Dilution W HSS Tar (mg) Increase over CI M- _ Marlboro Lts. KS 13.4 26 14 13.5 27 Virginia Slims Lights 100's 12.4 57 28 14.3 81 Merit KS 12.6 62 18 13.0 67 Merit Ultra Lts. 8.6 100 36 9.9 130 Cambridge KS Softpack 1.6 100 64 2.2 175 Conclusions The decrease in dilution from covering a portion of the perforated area can result in an increased delivery to the smoker of highly-diluted cigarettes even though the puff parameters decrease. For cigarettes with lower dilutton levels the increase in tar delivery is minimal. zq_ B. Goodman BG/lad' cc: Mr. W. Claflin Mr. P. Gauvin