Enumeration by Viable Count

The experimental procedure described in 3.5.1.1 was followed. However in these experiments cell numbers were estimated by plating the samples on growth media. Colonies were counted after the plates were incubated.

Table 3.19: Recovery of S. cerevisiae Aerosolised in Ringers' Solution

Aerosol 1.09 X 10* 1.09 X 10* 1.09 x io l®

Mean recovery in cyclone (%)

1.3 ± 1.9 0 2.3 ±3.7

Mean Recovery from the floor (%)

Chapter Three: Mass Balancing in Contained Environments

Table 3.20: Recovery of S. cerevisiae Aerosolised in Broth

Aerosol 1.23 X 108 1.23 X 10* 1.23 X 10*0

Mean recovery in cyclone (%)

0 0.7 ±0.8 5.2 ± 0.7

Mean Recovery from the floor (%)

1 5 + 4 1 8 ± 6 25 ± 2 0

Table 3.21: Recovery of E. coli Aerosolised in Ringers' Solution

Aerosol 9.36 X 10* 9.36 X 10* 9.36 X 109

Mean recovery in cyclone (%)

34 ± 2 8 1 ± 1 4 ± 1

Mean Recovery from the floor (%)

42 ± 4 0 52 ±63 1 3 ± 3

Table 3.22: Recovery of E. coli Aerosolised in Broth

Aerosol 2.6 X 10* 2.6 X 109 2.6x lOlO

Mean recovery in cyclone (%)

1 0 ± 4 1 ± 1 4 ± 1

Mean Recovery from the floor (%)

4 ± 3 3 ± 3 5 ± 2

When cell recoveries are calculated based on viable counts rather than total cell counts, the effect of the cyclone on cell viability can be seen. In tables 3.19-3.22 the recoveries from the floor are greater than those from the cyclone. This is because cells recovered from the floor are more likely to retain their viability than those recovered in the cyclone. It was seen in table 3.9 that the viability of E. coli

cells is reduced far more when the cells are collected in the cyclone compared to floor plates. The particles falling to the floor are probably the large ones and

Chapter Three: Mass Balancing in Contained Environments

The recovery o f 1.1 x 10^ 5'. cerevisiae cells in Ringers' solution from the floor of the cabinet is greater than 100%. This may be due to a particularly large particle containing many cells not having been broken up before it was spread on the agar. This effect could also explain the overall variability in these viable count data. Although viability is used frequently to count cells, plate counts are difficult to repeat accurately and viability itself is a very difficult state to define.

The standard deviations shown in the tables are very large, as large as the means in some cases. This implies that these data are not very consistent. This lack of consistency probably reflects on the cell enumeration method used rather than the efficiency of the cyclone in sampling the cabinet air. The cyclone samples the total cabinet volume in each minute that it is operated.

3.5.2 Soft Film Cabinet

Table 3.23: Recovery of S. cerevisiae and E. coli Aerosolised in Ringers' solution

The experimental procedure described in section 3.5.1.1 was followed for these experiments, except that the aerosols were released into the soft film cabinet. The numbers in the table are the mean recoveries, based on total counts, for 5 repeats.

Aerosol S. cerevisiae E. coli

Total cells released 1.4 X 10^ 1.4 X 109 1.4 X lOlO 8.9x lOlO

Mean recovery in cyclone (%)

3 1 + 6 19 + 5 15 + 3 17 + 2

Mean recovery from floor (%) tftc (<20% of the release) tftc (<2% of the release) - 3 0 14 ± 5

A Student' t-test was carried out to test if the difference between the mean recoveries was significant. The recovery for 10^ total cells released into the cabinet was found to be significantly different from that for 1 0^ cells at the 1.0%

confidence level and significantly different from that for IQlO cells at the 0.1%

confidence level. Again, this indicates that there is a decrease in recovery in the cyclone as the concentration of the released cells increases. The mean cyclone recovery for E. coli of 17.1% compares well with that of 15.4% for a release of

Chapter Three: Mass Balancing in Contained Environments

IQlO yeast cells. Although the mean recovery from the floor is much lower for E. coli compared with yeast cells at the same concentration.

When the cell recoveries for aerosols released into the soft film cabinet (table 3.23 above) are compared with those in table 3.15 for aerosols released into the Bassaire cabinet, it can be seen that the recoveries are approximately halved. This is interesting because the soft film cabinet has a volume 23 x that of the Bassaire cabinet. When the cyclone samples from the Bassaire cabinet, the whole volume of air is sampled each minute. This is not the case when sampling from the soft film cabinet and for this reason collection data from the soft film cabinet are expected to be less consistent. It is likely that more cells will be recovered from the floor of the soft film cabinet because less than 2 air changes occur inside the

cabinet, during the whole sampling period.

The experiment described above was repeated with the APV 30CD bomogeniser inside the soft film cabinet. The results are shown in table 3.24.

Table 3.24: Recovery of S. cerevisiae from Soft Film Cabinet Containing the APV 30CD Homogeniser

Cells Released {S. cerevisiae

in R ingers' solution)

1.3 X 10*0 total 1.1 X 10*0 viable

Mean recovery in cyclone (%) 21 ± 2 6±2

Mean recovery from floor (%) 21 ± 5 16 + 3

These results show that a significant proportion of a microbial aerosol can be recovered from the soft film cabinet and that the results are consistent when the mass balance is repeated. This then suggests that if an aerosol of significant proportions was released from a piece of bioprocessing equipment, it would be possible to measure that release using the mass balancing technique developed in this chapter. This is investigated in section 4.3, using the APV 30CD high pressure homogeniser.

Chapter Three: Mass Balancing in Contained Environments 3.6 Particle Size Analysis o f M icrobial Aerosols