7578409707

BIOLOGICAL AEROSOL MEASUREMF.NT IN DAIRY PLANTS

of nitrogen and ozonc are produced which may be loxic lo microorganisms. Thcse samplers havc a high sampling rale, high collection efficiency, and Iow rcsisiance lo air flow. They are mcchanically complcx and rrnisl bc handled carcfully. Allhough scveral cicclrosiatic prccipitators are manufaciurcd spccifically for sampling microbial acrosols. ihcy arc not widcly uscd for (his purposc (74).

Thermo! prccipiioiion mcihods

Thermal prccipit.nion mcihods rccover pariiclcs based on ihcrmophoresis principlcs in which particie* move away from a hol surface ioward a coldcr surfacc by a force proportional lo Ihc nfmpcrature gradient. Thcsc samplers can bc uscd for ihc dcicrminntion of paniclc size distribu-lion. lhough Ihcy are morc cffeciive when collecting smali panicles (less Ihan 1 prn). Howcvcr. ihey are not in com-mon use sińce prccise adjustmenis arc rcquircd and air sampling raies are quiic Iow (300*400 ml/min). The aerosol particles are usually collectcd on a glass covcrslip or elec-iron microscopc grid, and arc subscquent!y sized and counicd microscopically (74).

Addilional Information on aerosol samplers can be found in ihe following references (J J 20,23 Jl J5J0.69,74).

Comparison studies on aerosol samplers

Millipore and absorbent cocton samplers morc efficiently rccovered mold spores ihan the AGI-30 sampler (65). Fields et al. (27) recommcnded ihe use of ihc membranę filtcr Field (MF) monitor for estimating aiibome microorganisms aficr comparing it with ihe Reynicrs slit air sampler for microbiological laboratory and clean room environment tesiing. These resulls showed ihai a significamly higher number of microorganisms were recovcrcd by the Reynicrs slit sampler with a high degree of consistcncy. The MF sampler detccied 79% of ihe concentration measured by ihe Reyniers slit sampler and the typcs of microorganisms identified from boch sampling mcihods were similar. Chaligny (17) concluded ihal .4 pm Millipore filier as an air sampler of smali panicles may be somewhat morę effi-cieni ihan the AGI-30 sampler, bul viable rccovery will usually be lowcr exccpt in ihe case of bacterial spores or fungi. Similarly, ihe Litton (LVS) large volumc electro-static sampler is from 40 to 70% as efficient as the AGł-30 sampler, but ii has a sampling ratę approximatcły 100 limes greater than ihc AGI-30 sampler.

The Andersen 2-stagc impactor was morc effcctive ihan ihe May 3-stage glass impinger for recovering Escherichia coli from aerosol in a wasie water plam environment (75). Curtis ei al. (21) compared the Andersen 8-siage and 2-siagc air samplers for recovery of viablc organisms. They found ihat the 2-siage disposable air sampler gavc lower values for airbome bacierial colony-forming panicles ihan did ihc 8-siage viablc air sampler in eilher a swinc bam or a classroom. When Lembke ei al. (46) tried lo dcvise a method lo determine ihe precision of the AGf-30 sampler and ihc Andersen 6-stage air sampler over a wide rangę of aerosol concenirations insidc a municipal solid-waste re-covery system, ihcy found a high degree of variability associaied with both typcs of air sampling deviccs._ They indicated ihai slippage of panicles from one stage to an-Olher siage, paniclc fragmentation or agglomcraiion, and wali losses in Andersen 6-stagc air sampler, and Processing techniquc in AGI-30 sampler may accouni for sorne of Ihc vnrinncc. An Andersen 6-stagc sampler. a Casclla slit sampler, an AGI-30 sampler. and a filtcr sampler with gelalin membranę filters or ordinary membranę filters were lestcd for collcction cfficicncy with a bacterial aerosol in laboratory cxpcrimenl. in field expcrimenls, and in experi-ments wiih skin fragment sampling (48). The Andersen sampler gavc the highest bacterial counts in all environ-ments leslcd. The slit sampler gavc significantły lowcr counts only in the aerosol cxpcrimcnts and in one of Ihc field expcrimcnts. 'fhc filters performed efficient sampling in skin fragment expcrimcnts only.

Radmore and Liick (6!) compared exposure plale, liquid impinger and gelatin membranę filtration (GMF) mcihods. Air counts determined by exposure of agar plates were not at all, or only weakly corrclatcd with łhc counts determined by iwo other methods. The rclationship betwcen the liquid impinger and GMF mcihods were also not very consistent (r=0.75). Al levels above 1000 microorganisms/m³, ihc impinger method yiejdcd counts up lo 6 limes higher Ihan the GMF method. They explaincd ihis was probably a resuli of the dispersion of dusi particles and the breaking up of chains and clumps of bacteria during the bubbling of air* ihrough the impingement liquid.

In comparativc studies of airbome microbial recovcry raie (22J5859), ihe RCS sampler was found lo bc signifi-canily morę efficient ihan a slil sampler or a liquid impinger. The RCS sampler samples air the shape of a sphere with a diameter of 1.3 ft-representing an air volume of about 1.2 cu fi versus only 0.5 cu ft dimension of air sampled by ihe slil sampler (22).

Comparison studies of air sampling deviccs mdicaie ihat there is often no obvious choicc of ihe conreci sampler to usc. A multislage sicve sampler such as the Andersen may be most cfficienl al viable particie rccovery bul it is not suitabłe for laking repeated sampling on a routine basis and rcquires a vacuum source. Filier samplers work well for quality control monitoring of molds and bacterial spores. but bacierial rccovcry is qucstionablc. depending on ihc cxlent of dehydraiion ihat occurs during sampling. In addi-tion, a vacuum source is requircd. The RCS sampler is con-vcnieni to usc, creaics iis own air flow and recovers bacteria as well as molds. Even lhough the RCS sampler does not rccover the smallcst viablc particles, it is still uscful for dcicrmining rclativc air quality on a routine basis. Slil samplers may not bc as convcnient lo usc as the RCS sampler, cspccially if a vacuum source is rcquircd. How-cver. slil samplers are morę cfficienl at recovering smali particles.

SAMPLING AND MEASUREMENT STANDARDS

Standard Methods for the Exominati<m of Duiry Products

The I5th cdition of the Standard Mcihods for ihc

JOURNAL Oh hOOD PROIUCIION. VOL. 52. 3UI.Y IVXV