of GNPs that can be detachedby one batchof milkis roughly similarto the
number of bacteriathat can be detectedby a swabsample: 50 000 cfu. The
poorly cleanable annulargap around a temperaturesensor placedin a T-piece
was foundto contain 10 000 cfu of GNPs,whichwould result in a contamination
of 0.05 cfu/Lif these10 000 GNPswereto be releasedoverone batch.
An empty buffertankcontains air withaerosols contaminatedwith about
100 cfu/m^3 GNPs(Table 20.1).Thesebacteria movetowardsthe walls and
bottom of the tank by sedimentationand diffusionand then stickto the wet walls.
Uponfilling the buffertankwithmilk,thesebacteriawill be released into the
milk,thuscontaminating the pasteurisedmilkwith100 cfu/m^3 , i.e. 0.1 cfu/L.
During filling,the milkcan be recontaminated by bacteriaoriginatingfromthe
air, fromthe rinsing water or from the packaging material. The level of
recontamination is moredifficult to estimate directly fromthe data in Table20.1.
Based on literaturedata(e.g.Eneroth, 1999 ), roughestimations can be made:
eachof thesethree sourcesaround the fillerwill contribute about0.2 cfu/LGNPs.
In therisk characterisationstepthe results of exposure assessment and
hazard characterisation are integrated and compared. As explained above,
Table20.1 Microbiologicaltest resultsfor a risk assessmentof a pasteurisedmilk
productionline (eachnumberis a roundedaverageof a largenumberof samplestaken,or
givesan approximaterange)
Samplelocation Samplelocation B. cereus Gram-negative
of liquidmilk swab psychrotrophs
cfu/L cfu/swab cfu/L cfu/swabSilo tank 100 5 million
Directlydown- 200 < 1
streamof
pasteuriser
Diaphragmof < 10 50,000
constantpressure
valve
T-pieceof < 10 10,000
temperaturesensor
Buffertank 200 0... 3
Packagematerial < 10 < 10
beforeusage
Filledand sealed 200 0... 4
consumerpackage
Estimatednumberof GNPsin air n.d. 100 cfu/m^3
nearbuffertank
Estimatednumberof GNPsin air n.d. 1000 cfu/m^3
nearfillingmachine
Watersamplesat fillingmachine n.d. 100 cfu/m^3
320 Handbookof hygiene controlin the foodindustry