RECOMENDACIONES

(mg/L) 0.01 0.01 0.01 0.00 - - 4 2.4 Calcium as Ca (mg/L) 2.76 5.07 2.40 3.02 150 200 200 - Potassium as K (mg/L) 0.29 1.00 0.21 0.19 50 50 - - Magnesium as Mg (mg/L) 0.40 0.82 0.28 0.37 70 30 200 - Sodium as Na (mg/L) 3.09 4.96 1.86 1.91 200 100 180 - Phosphorus as P (mg/L) 0.09 0.11 0.04 0.02 - - - - Silicon as Si (mg/L) 0.20 2.58 0.17 0.41 - - - - Aluminium as Al (µg/L) 5.96 443.29 7.29 15.15 300 150 100 - Chromium as Cr (µg/L) 0.06 0.09 <0.303 <0.303 100 50 50 50 Manganese as Mn (µg/L) 4.76 1.61 0.83 0.68 100 50 500 - Iron as Fe (µg/L) 94.02 30.59 21.23 10.61 200 100 300 - Cobalt as Co (µg/L) 0.06 0.10 0.02 0.09 500 - - - Nickel as Ni (µg/L) 0.15 1.11 0.64 4.76 150 - 20 70 Copper as Cu (µg/L) 1.06 0.49 1.99 1.53 1000 1000 2000 2000 Zinc as Zn (µg/L) 24.94 1.96 22.46 5.73 5000 3000 3000 - Arsenic as As (µg/L) 0.24 7.41 0.23 0.77 10 10 10 10 Selenium as Se (µg/L) 2.36 1.18 1.88 0.68 20 20 10 40 Strontium as Sr (mg/L) 21.76 160.29 18.14 52.56 - - - - Molybdenum as Mo (µg/L) <0.000 <0.00 0.03 0.06 - - 50 - Cadmium as Cd (µg/L) 0.02 0.01 <0.019 <0.019 5 5 2 3 Tin as Sn (µg/L) 0.01 0.02 <0.016 <0.016 - - - - Antimony as Sb µg/L 0.08 10.57 0.08 0.24 - - 3 20 Barium as Ba (µg/L) 21.06 20.37 17.20 31.76 - - 2000 700 Mercury as Hg (µg/L) 0.01 0.03 0.02 <0.021 1 1 1 6 Lead as Pb (µg/L) 0.18 0.17 0.13 0.05 20 10 10 10

149 The average concentration of aluminium observed for the filtered rainwater samples (443.29 µg/L) collected on day one of sampling exceeded the recommended guidelines of 300 µg/L stipulated by SANS 241 (2005), 150 µg/L as stipulated by DWAF (1996) and 100 µg/L as stipulated by the AWDG (NHMRC and NRMMC, 2011). In addition, the average concentration of antimony in the rainwater sample collected on day one after filtration through the activated carbon system, exceeded the AWDG (NHMRC and NRMMC, 2011) guideline of 3 µg/L, with an average concentration of 10.57 µg/L observed. However, the concentrations of aluminium and antimony in the remaining unfiltered and filtered rainwater samples collected on day 5 were within the respective guidelines. Although aluminium has been associated with Alzheimer’s disease, Parkinsonism dementia and amyotrophic lateral sclerosis, it has been concluded that there is insufficient information to link these diseases with the consumption of aluminium through drinking water sources (DWAF, 1996; NHMRC and NRMMC, 2011). In studies performed on rats, the consumption of antimony was also linked to fertility and it was demonstrated that antimony accumulates in the heart, spleen, liver and kidney (NHMRC and NRMMC, 2011).

While it was noted that all other cations detected were within the respective guidelines, the significant variations in the concentrations of the cations as well as aluminium and antimony (that were not within guidelines) will be discussed. The concentration of silicon, aluminium, arsenic, antimony and strontium present in the filtered rainwater samples collected on day one after the formation of the biofilm increased significantly (more than a 10 fold increase) from an average of 0.2 µg/L (unfiltered) to an average of 2.58 µg/L (filtered), from an average of 5.96 µg/L (unfiltered) to an average of 443.29 µg/L (filtered), from an average of 0.24 µg/L (unfiltered) to an average of 7.41 µg/L (filtered), from an average of 0.08 µg/L (unfiltered) to an average of 10.57 µg/L (filtered), and from an average of 21.76 µg/L (unfiltered) to an average of 160.29 µg/L (filtered) respectively. In contrast, in the same rainwater samples the concentrations of manganese, iron, copper and zinc were observed to have decreased on day one after filtration through the activated carbon from an average of 4.76 µg/L (unfiltered) to an average of 1.61 µg/L (filtered), from an average of 94.02 µg/L (unfiltered) to an average of 30.59 µg/L (filtered), from an average of 1.06 µg/L (unfiltered) to an average of 0.49 µg/L (filtered) and from an average of 24.94 µg/L (unfiltered) to an average of 1.96 µg/L (filtered), respectively. On the fifth day after the formation of the biofilm, no significant increases were observed for any of the cations present in the rainwater samples, however potassium, phosphorous, manganese, iron, copper, zinc, selenium and lead exhibited a negligible decrease after filtration through the activated carbon system.

150 The raw materials used in the manufacturing of commercially available activated carbon contains ppm levels of arsenic and antimony and have been shown to leach small fractions of these elements when in direct contact with water (Vaughn and Distefano, 2013). This could account for the increased levels of antimony and arsenic observed in the rainwater samples that have been filtered through the activated carbon. Moreover ash, used in the manufacturing of activated carbon, contains aluminium and silicon (Block and Dams, 2010). Therefore aluminium and silicon could also have leached from the activated carbon particles into the rainwater during the filtration process.

All anions present in the unfiltered and filtered rainwater samples collected on day one and day five after the formation of the biofilm on activated carbon were within drinking water guidelines according to SANS 241 (2005), DWAF (1996), ADWG (NHMRC and NRMMC, 2011) and WHO (2011) and are represented in Table 5.4.

Table 5.4. Average anion concentrations obtained from rainwater samples collected before and after filtration through the activated carbon on the respective days. These values are compared to the recommended concentrations as stipulated by the respective drinking water guidelines (n = 8, average of each duplicate sample).

Anions (mg/L) Before Day 1 After Day 1 Before Day 5 After Day 5 SANS

241 DWAF ADWG WHO

Sulphate as SO4 1.15 1.7 1.4 1.55 200 100 250 -

Chloride as Cl- 6.65 2.9 6.55 4.9 400 200 250 -

Nitrate and Nitrite

as NO3 and NO2 0.2 0.1 0.05 0.05 10 6 50 50

Phosphate as PO4 0.095 0.085 0.025 0.025 - - - - Fluoride as F <0.01 <0.01 0.1 0.1 1 1 1.5 1.5

On the first day of the trial chloride and nitrite and nitrate concentrations decreased from an average of 6.65 mg/L (unfiltered) to an average of 2.9 mg/L (filtered) and from an average of 0.2 mg/L (unfiltered) to an average of 0.1 mg/L (filtered), respectively. On the fifth day of filtration only the concentration of chloride was shown to have decreased from an average of 6.55 mg/L (unfiltered) to an average of 4.9 mg/L (filtered).

The formation of a biofilm layer on granular activated carbon (GAC) is referred to as biological activated carbon (BAC) and has gained interest as a water treatment technology. Biological activated carbon has been shown to remove chemical pollutants, disinfection by-product precursors and organic matter (Simpson, 2008). In this study the biofilm was allowed to establish for a week before the rainwater was filtered through the activated carbon. The efficiency of this system could therefore be increased by lengthening the time of biofilm growth before using the system for the filtering of rainwater.

151 5.3.2.2 Biological Filtration: Slow Sand Filtration

A slow sand filtration system was installed on rainwater tanks B and C. Before the efficiency of the system was monitored a biofilm was allowed to establish for two weeks. The presence of cations was assessed in unfiltered and filtered rainwater samples collected on the first and the twelfth day after the formation of the biofilm within the slow sand filtration system. All cations present in the rainwater samples collected before and after slow sand filtration were within the drinking water guidelines according to SANS 241 (2005), DWAF (1996), ADWG (NHMRC and NRMMC, 2011) and WHO (2011), with the exception of aluminium, manganese and iron as represented in Table 5.5.

The aluminium concentration of the filtered rainwater sample (average 1601.43 µg/L) collected on day one exceeded the recommended guidelines as stipulated by SANS 241 (2005), DWAF (1996), ADWG (NHMRC and NRMMC, 2011) and WHO (2011). However, the concentrations of aluminium in both unfiltered rainwater samples and the filtered rainwater sample collected on day twelve were within the recommended guidelines. The concentration of manganese in the filtered rainwater sample collected on day one had an average concentration of 53.45 µg/L which exceeded the recommended guideline of 50 µg/L stipulated by DWAF (1996). However, the concentrations of manganese in both unfiltered rainwater samples and the filtered rainwater sample collected on day twelve also adhered to recommended guidelines as stipulated by SANS 241 (2005), DWAF (1996), ADWG (NHMRC and NRMMC, 2011) and WHO (2011).

Only the unfiltered rainwater sample, collected on day one, adhered to the recommended guidelines for iron as stipulated by SANS 241 (2005), DWAF (1996), ADWG (NHMRC and NRMMC, 2011) and WHO (2011) with an average concentration of 23.04 µg/L observed. However, the average concentrations of iron in the unfiltered rainwater samples collected on day twelve (143.99 µg/L) and the filtered rainwater sample collected on day twelve (108.31 µg/L) were not within the recommended DWAF (1996) guideline of 100 µg/L. In addition, the concentration of iron in the filtered rainwater sample collected on day one (4083.45 µg/L) was not within the guidelines as stipulated by SANS 241 (2005), DWAF (1996), ADWG (NHMRC and NRMMC, 2011) and WHO (2011).

As mentioned previously (Section 5.3.2.1), aluminium present in drinking water has been linked to certain neurodegenerative diseases. Manganese is however, considered to have very low health risks if present in water, but can have aesthetic and taste effects if present at concentrations exceeding the respective guidelines (DWAF, 1996; NHMRC and NRMMC, 2011).

152 Table 5.5. Average cation concentrations obtained from rainwater samples collected before and after filtration through the slow sand filter on day one and twelve, respectively. These values are compared to the recommended concentrations as stipulated by the respective drinking water guidelines (n = 8, average of each duplicate sample).

Metal Before Day 1 After Day 1 Before Day 12 After Day 12 SANS