Gastos de explotación del Grupo

Passive air samplers were deployed in homes and offices within the Birmingham region, for a period of 30 (±2) days. The samplers were assembled as described in chapter 2 and fitted with a single SIP disk. Field blank sample results were all non-detects for all target PFCs.

Atmospheric concentrations of PFCs in homes and offices are displayed in Table 39 and Table 40 along with descriptive statistics. Results indicate the majority of PFCs are present in indoor air, but with the exception of MeFOSA, which had limited detection. The more volatile compounds: EtFOSA, FOSA, MeFOSE and EtFOSE were the most abundant in air (12 – 3100 pg m^-3). Despite the lower volatility, PFOS, PFOA and PFHxS were detected in > 90 % of the air samples. This may be partly attributable to the passive samples retaining deposited particulates as well as vapour phase PFCs (Hazrati & Harrad, 2007). Similar studies of UK outdoor air (using active HiVol samplers) have indicated the presence of PFOS and PFOA on filters, but none in the gaseous phase. Comparison of the median and mean from both home (1800 pg m^-3, 2000 pg m^-3, respectively) and office (1300 pg m^-3, 1300 pg m^-3, respectively) air samples indicate similar values for ΣPFC concentrations(Table 2). The individual compound descriptive statistics show that office samples have higher mean and median concentrations for the less volatile compounds, PFOS(53pg m^-3,56 pg m^-3), PFHxS (93 pg m^-3, 95 pg m^-3) and EtFOSA (45 pg m^-3, 58 pg m^-3), similar arithmetic means and medians for PFOA (17 pg m^-3, 58 pg m^-3), MeFOSA (<DL, 6 pg m^-3) and FOSA (49 pg m^-3, 74 pg m^-3), but lower office values for MeFOSE (380 pg m^-3, 480 pg m^-3) and EtFOSE (380 pg m^-3, 490 pg m^-3). This relationship could be due to the room usage in offices, with most offices sampled being well ventilated and more sparsely furnished than homes. Better ventilation in offices, facilitates more efficient

removal of more volatile compounds, while the relative lack of furnishing results in reduced PFC emissions.

Table 39 UK home air concentrations (pg m-3)

Table 40 UK office air concentrations (pg m^-3)

The air data was tested for normality of distribution, and differences in the presence of individual PFCs was observed for homes and offices. The Shapiro-Wilk test was used, as opposed to the Kolmogorov - Smirnov test, because each microenvironment category contained less than 50 samples, and the Shapiro-Wilk has proven superior when handling small data sets (Livingston, 2004). While concentrations of most PFCs in both home and office microenvironments were revealed by the Shapiro-Wilk test to be positively skewed (apart from EtFOSE, p < 0.023), there were some exceptions (EtFOSE, W = 0.914, p = 0.076). Analysis of room contents and concentrations was conducted for the air concentrations and no significant relationships (p > 0.05), were identified from the data given on the questionnaire (this included flooring type, textiles, ventilation, electronics and time spent in room).

A t-test was conducted for home and office air samples, to compare the means for statistical differences. All data were either normally distributed or were normalized using a log transformation Significantly different air concentrations were identified (p

< 0 .05) for PFOS, EtFOSA and MeFOSE. Offices may contain fewer volatile compounds (means of PFHxS and PFOA are also raised in office environments) due to differences in room use, and different sources.

PFOS in air concentrations were significantly higher in offices than homes (p = 0.001), where the opposite was true for EtFOSA (p = 0.039) and MeFOSE (p = 0.033). Although not significantly different, concentrations of volatile compounds were higher in homes than offices and the converse is seen in offices for the less volatile compounds.

Figure 10 Comparison of arithmetic mean air concentrations in homes and offices (pg m^-3)

PCA analysis was used to examine patterns among PFC profiles in homes and offices.

Three factors (Table 41) accounted for 61 %) of the variance ((Figure 11). The PCA divides the data into two main groups. The two groups are dependent upon factor 1 (Figure 11). Cluster 1 samples have positive scores on factor 1 due to higher contents of the following compounds in the samples: MeFOSA, PFHxS, PFOS, EtFOSE and FOSA and samples in the cluster were mainly offices. Cluster 2, which contained mostly homes, had negative factor 1 scores with greater contributions of MeFOSE in the samples.

Homes in cluster 2 tended to have carpeting, electronics and general furnishings, with higher concentrations related to newer products within the rooms (as indicated by the questionnaire).Offices with larger contributions of MeFOSE fall within this category could be attributed to newer furnishings within the individual offices, but homes remain greater than offices.

MeFOSE is also present in carpet stain repellent formulae, and may be linked to homes with treated carpets (Beesoon et al. 2010). However, when MeFOSE concentrations in carpeted homes and offices are compared with those in non-carpeted microenvironments, there are no significant differences, probably attributable to many indoor environments containing carpets of different ages and production, and thus differences for microenvironments with and without carpets were indistinguishable.

Comparison with other studies of PFCs in indoor air (Table 42) supports this study‘s finding that concentrations in homes exceed those in offices. De Voogt et al., (2008) collected indoor air samples via HiVol samplers using glass fibre filters (GFF) to collect the particulate phase and a combination of PUF and XAD in columns for the gaseous phase.

Table 41 PCA factor loadings for air samples.

Compound Component

1 2 3

PFOS 0.507 0.488 -0.073

PFOA -0.028 0.782 0.214

PFHxS 0.695 -0.071 -0.086

MeFOSA 0.678 -0.246 0.217

EtFOSA -0.055 0.667 -0.115

FOSA 0.242 -0.061 0.794

MeFOSE -0.777 -0.442 0.077

EtFOSE 0.400 -0.114 -0.751

Figure 11 PCA of home (dark green) and office (light gr een) air samples, with variance represented by factor 1 (27%), factor 2 (18%) and factor 3 (16%).

The results from this study remain within the lower range of concentrations measured in home samples from Canada (290 – 4000 pg m^-3) (Shoeib et al., 2004, 2005) and Europe (100 – 83 000 pg m^-3) (de Voogt et al., 2008). The upper ranges measured by the first Canadian study are within the ranges of values measured from this study for MeFOSE and EtFOSE (77 – 3100 pg m^-3 and 80 – 1900 pg m^-3, respectively) for both homes and offices. Comparisons of this study to a European study (de Voogt et al., 2008) reveals that the concentrations of MeFOSE (max = 83 000 pg m^-3) and EtFOSE (max = 29 000 pg m^-3) are much higher (25 times and 15 times, respectively)than from the UK (MeFOSE = 3100 pg m^-3 and EtFOSE 1900 pg m^-3)compared to Tromso values.

Table 42 Comparison of indoor air samples from other countries (pg m^-3), (^a range of sample concentrations)