Tobacco Documents Online

MEMORANDUM May 13, 1977 REF:RLT-25-77 1 ( TO: Mr. M. S. Block FROM: R. L. Thomas SUBJECT: Engineering,Progress Report, CONAP Project Preliminary data is now available to permit design feasibility of extraction methods for flavorant B-100. Waste tobacco usable as a source of B-100 is 5 x 106 lbs. per year at a weighted average of 0.9% B-100 for a total`of 45,000 lbs./year. Note,that this quantity will vary depending on tobacco quality, purchases, etc. B-100 analysis in.the waste is probably accurate to +.05 or so, but the uniformity of the stems is uncertain. Current waste tobacco is 2/3 stems (Burley, Penn., and Wisc.) and 1/3 dust and fines (primarily BP-7 from Greensboro.) - %,_ Two approaches to producing a flavor enhanced RL appear reasonable: direct application to RL slurry of B-100 either commercially obtained or extracted internally and production of a reasonably concentrated "blackwater" solution at 1% or less made from the waste tobacco and used directly in the RL process to replace the makeup city water currently used. Direct use of concentrated blackwater solution certainly appears to be the most attractive choice since it eliminates or minimizes the difficulties involved in handling,B-100. An experiment is scheduled May 17 at Danville to make 2% RL from concentrated blackwater obtained from the waste mix and existing RL blackwater. It is f er v e ntly hoped that this 2% RL will be suitable for the flavor enhanced product. In any event, extraction of tobacco waste either to produce B-100 or to pro-. duce concentrated blackwater is certain. Note that waste extraction will produce around 45,000 lbs./year of B-100 at current waste levels. _Other sources such as the Burley dryers are uncertain. Amounts over 45,000 lbs./year must come from outside sources, extraction and reformulation of existing brands, or ~ extraction of,tobaccos bought or grown strictly for B-100. ~ O Two promising routes have been investigated for extraction: (1) direct ~ extraction with FC-11 and tobacco to which ammonia has been added; (2) water ~ extraction from untreated tobacco. The water extraction is preferred, ~ particularly if concentrated blackwater can be used, and tobacco so washed ~ potentially can be reduced to a lower level of B-100 and still be usable. The FC-11 extraction requires 10-20 lbs. of solvent per lb. of tobacco, extensive solvent recovery equipment, and apparently produces an unusable tobacco after extracting.

.6 . Mr. M. S. Block Extraction Discussion Page 2 Design parameters have been identified as follows: May 13, 1977 REF:RLT-25-77 Target production quantities = 5 to 10 x 109 cigarettes per year Cigarettes per pound of tobacco = 700 Enriched RL requirement, 1010 cigarettes @ 25% = 3.5 x 106 lbs./year Production time (one end of building) = 50 to 100 days Target RL concentration = 2% to 4% B-100 Target slurry concentration = 0.5 to 1% B-100 B-100 requirements for 1010 cigarettes = 49,000 lbs./year @ 2% RL B-100 requirements range = 24,500 lbs./year to 100,000 lbs./year Blackwater evaporation rate = 3500 lbs./batch Conc. blackwater generation rate = 3500 lbs./batch :,. C C The production of blackwater is a leaching or washing operation with the washing solution recovered for reuse. Since the strongest blackwater solutions are obtained in countercurrent extraction techniques and since high efficiency of B-100 recovery is necessary, the following schematic is proposed: The extractor can be any of several multi-stage type extractors commonly used for leaching or extraction of oils from plant material. . Each stage requires a means of contacting the tobacco with liquids from the adjacent stage and a means of separating the liquids after contact. The separation can be draining, suction, centrifuging, pressing, etc. The degree of_separation directly affects the number of stages required but not necessarily the overall effi- ciency of B-100 recovery. The following calculations are based on 1 lb. of waste and assume that the entire year's accumulation of waste will be processed within 50 days and the concentration of blackwater adjusted to whatever level is required so that the final cigarette production will be between 5 and 10 x-109 cigarettes. Extraction data we currently have'developed indicates that repeated washings will reduce the B-100 content of at least some of the waste to 0 and that in the simplest case the B-100 in the tobacco dissolves in the water independ- ently of any B-100 concentration already in the water, at least up to the concentrations projected for direct use of the concentrated blackwater. Q C Reasonable extraction efficiencies have been obtained using 5:1 and higher ~ weight ratios of water to tobacco and the water retained on the tobacco ~ waste (in this case RL dry mix was used to simulate the waste mix) after ~ draining was 2.5:1 weight ratio. ~j

C Mr. M. S. Block Page 3 May 13, 1977 REF:RLT-25-77 Countercurrent Washing 5 lbs. water ditto ditto ditto water 1 lb. tobacco "dry" 4 1 lb. tobacco 2.5 lbs. water ~ ater, / 1 5 lbsI. / / ( 1 lb. tobacco .2.5 lbs. iI f 14 I i I , ditt•4, ) 6ttd`~~ ~ , 1 ! t I i i 2.5 lbs. c blackwater > f. C The above schematic traces 1 lb. of tobacco•and 5 lbs. of water through a typical 5 stage cascade of a countercurrent wash. The only B-100 loss from this system is residual in the tobacco and a very low concentration in the adhering water. Laboratory extraction based on this technique resulted in essentially 0 residue in the stems used. Losses in the water are dire(~4ly controlled by the number of washing stages. This system seems capable of a 90% recovery of B-100. The concentrated blackwater in that case would be about 0.32%. If the water feed rate is reduced to 4 lbs., the blackwater concentration will be about 0.54%. Higher concentrations in blackwater can be produced by recycling a portion of the concentrated blackwater back a few stages, adding additional wash stages, and reducing the amount of feed water by the amount of recycle. Using the most favorable equilibrium assumptions, the number of stages required for 5:1 water to tobacco washing is 4 and for 4:1 water to tobacco is 5. See attached stage calculations. The extraction conditions, based on data obtained on RL dry mix, are 185°F with a contact time of 5 to 30 minutes, indicate 75% extraction in one pass. The key equilibrium assumption is that subsequent wash passes will remove substantially all of the remaining undissolved B-100 from the tobacco. Waste Disposal Quantities Using 5:1 ratio or 4:1 ratio: - Tobacco processed = 5 x 106 lbs. 4 50 days = 100,000 lbs./day Waste tobacco = .9 x 100,000 = 90,000 lbs./day Waste water on tobacco = 90,000 x 2-1/2 = 225,000 lbs./day Waste water to sewer from process = 0 Treatment of waste tobacco to remove water prior to land filling may be required. RLT:bs Attachments cc: Mr. J. S. Bullock Dr. H. J. Minnemeyer R. L. T.