A comparison of fibre counting stopping rules in the application of the membrane filter method. Final report on CEC Contract 84/396

The method used to test compliance with control limits for airborne asbestos fibre concentrations in the workplace involves sampling dust onto a membrane filter and counting the fibres rendered visible by phase contrast microscopy. The European Reference Method (ERM) specified in the CEC Directive on Asbestos requires fibre counting to cease after 100 fibres have been counted or 100 fields of view have been evaluated, whichever is reached first, with a minimum of 20 fields being examined. With this stopping rule, more of the filter area is evaluated at lower concentrations than is examined when the concentration is around the compliance limit, i.e. more counting is involved at lower concentrations than is necessary. This report describes the first stage of an investigation carried out with financial assistance from the CEC to identify and test alternatives to the ERM stopping rule which would reduce the counting effort at low concentrations whilst retaining a comparable degree of accuracy.Data comprising counts recorded field by field were available from various studies. Of these 425 counts had been made using the ERM counting protocol. This data subset was analysed to identify alternative candidate stopping rules, followed by a short experimental trial to provide some practical indications of how these rules compare with the ERM rule.Simulation studies were first performed on the data, involving random selections of fibres and fields. This, together with theoretical considerations showed that it was possible to derive other rules which in practice classify sampling in terms of compliance at least as efficiently as the ERM rule at fibre concentrations near the compliance limit, but require less counting at lower concentrations. Two candidate rules were defined, applicable to assessing compliance with the European control limit for chrysotile and amosite of 1.0 fibre/ml. The counting rules can be designed differently for different sample volumes. Sample volumes of 240 litres, corresponding to the widely used sampling practice of 1.0 1/min flow rate for 4 hours, were considered in defining and testing the rules in this study. These rules are:(a) ‘135 fibres/30 fields’. Counting continues until at least 135 fibres have been counted or 30 fields evaluated, whichever is reached first, with at least 20 fields being examined.(b) ‘two-stage rule’. The first stage involves counting 20 fields.If fewer than 75 fibres, or more than 105 fibres have been counted, counting ceases. Otherwise, the evaluation continues until at least 180 fibres have been counted or 40 fields examined. Analagous rules appropriate to control limits lower than 1.0 fibre/ml would require more counting than (a) or (b) in order to achieve the performance of the ERM rule.The theoretically expected performance of these rules was derived and compared with that of the ERM rule. Since these performance characteristics apply on the assumption that the data conform to the Poisson distribution, goodness of fit tests to Poisson were conducted on existing data. 72 per cent of the samples did not differ significantly from the Poisson distribution.The two selected candidate rules were compared with the ERM rule in a small experimental trial. Fifteen chrysotile samples were counted by two microscopists who in all other respects followed the ERM counting protocol.Sample densities spanned the range 0-600 fibres/mm2, the upper limit corresponding to 1.0 fibre/ml as sampled at 1.0 litre/min for four hours using the ERM procedure. Evaluated fibre densities obtained by the candidate and ERM rules did not differ systematically. For concentrations below about 0.8 fibres/ml, the time involved in applying the candidate rules progressively decreased as concentration decreased; at low concentrations application of the candidate rules involved only one-fifth of the time of the ERM rule.A more broad-based study, involving government and industrial laboratories throughout Europe, is recommended so that each laboratory can gain first-hand experience on the effects of candidate rules for conditions prevailing in its own country.