FLAMMABILITY OF PROPELLANTS 491 FLAMMABILITY OF PROPELLANTS K. DIXON, B.Sc*. Presented at the Symposium on "Aerosols", organised by the Society, at Southport, Lancs., on 25th April 1963. Possible tyl•es of accidents with aerosols in the user's hanfls are consiflerefl. A metho½l for assessing the fire risk associateel with a bursting disl•enser is •lescribe•l. Results obtaineel •rom this metho½l and •rom flame extension tests are comlmre•l with some l•ublishe•l figures. It is shown that any fire hazard is associateel with the comlflete •lisl•enser rather than with any l•ar•icular ingre½lient. THE AEROSOL industry is continually searching for a method of reducing the cost of its products, and it is well known that the use of flammable propel- lants in place of the conventional chlorofluorohydrocarbor•s would help in this direction--if it could be done safely. Practices differ from country to country. For example butane propelled aerosols, giving an extensive flame when ignited, are common in France, and in Germany we believe opinions to be divided, with some interest in the use of vinyl chloride. Since the end of 1961, a leading marketer in the U.S.A. has apparently made a success of products depending on a butane/water emulsion as propellant and in doing so has precipitated a maj or controversy with regard to their safety. In this country we are trying to decide, on the evidence available, what are the risks in usir•g a flarnmable propellant per se. In this paper the evidence is reviewed and the possible hazards in the user's hands have been considered. Possible hazards existing in the manufacture and bulk storage and transport also require careful cor•sidera- tion, but are beyond the scope of this paper. As little information on accidents appeared to be available, a review has been made of relevant incidents covering 12 years' experience, and the sale of over 70,000,000 aerosols. As a result of this review, a method of testing the fire hazard associated with a bursting dispenser was developed and this is described below. The results obtained are compared with those obtained by some of the standard C.S.M.A. methods, and with some published results from the literature. It is concluded that the most likely hazard in which fire may be involved is explosion of a dispenser due to mis-use, and that the intrinsic fire risk associated with a dispenser is rnore properly related to the *The Cooper Technical Bureau, Berkhamsted, Herts.

492 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS total amount of flammable material contained in it than to the specific nature of the flammable material concerned. REVIEW OF fi•CCIDENTS During the period under review (1950--1962) our Company has sold over 70,000,000 aerosols, the majority direct to the consumer market, with others on a contract filling basis. This has given us some sight into consumer reaction, not only towards our own products but also towards a wide range of other types. The bulk of them have been distributed within the United Kingdom, but a proportion have been exported to many parts of the world, including most climatic and cultural conditions. In response to a question- naire sent out, covering accidents involving fire or explosion, we obtained reports of four relevant incidents, which were as follows: (a) A 20 oz insecticide pressure pack was placed in the steam from a boiling kettle, and exploded. There were, fortunately, no injuries nor fire. (b) A 6 oz insecticide pressure pack was placed on a domestic boiler. The ends inverted. There was neither fire nor explosion. (c) A 6 oz insecticide pack was exposed to strong sunlight in a shop window, and exploded. There was no fire. (d) A 12 oz insecticide left in strong sunlight in the back of a car failed at the sideseam. There was no fire. In addition to these specific reports, we have information from a few tropical or semi-tropical countries, of explosions occurring due to exposure either to sunlight or to other forms of heat. We have not been able to obtain any reports of accidents involving fire where the contents of an aerosol were the first material ignited, nor of any where, in the hands of a consumer, an aerosol has contributed effectively to a fire. During most of the period under review we have been selling, as a sig- nificant proportion of the total, a product containing over 50% w/w of flammable material, and giving a flame extension by the C.S.M.A. method of about 20". Products with similar fire characteristics have been widely marketed in this country for the last eight or ten years, and to the best of our knowledge no accidents involving fire have occurred. We can only conclude that these products, which bear legends such as "Do not spray near a naked flame or while smoking" are safe in a consumer's hands. The fact that few accidents have occurred in the hands of users would suggest that aerosol products are reasonably safe, and we are left with the two questions "Are those propelled by a butane-water emulsion any less safe?", and "In what way, if any, are they less safe?".