Sand as a filter media

When using sand as a filter media two important factors play a role; sand grain size and sand bed depth. Both have important effects on bacteriological and physical water quality.

Sand grain size

Most literature recommends that the effective size of sand used for continually operated slow sand filters (COSSFs) should be in the range of 0.15 – 0.35mm, and that the uniformity coefficient should be in the range of 1.5 – 3, although a coefficient of less than 2 is desirable (Schulz and Okun, 1984 [ref.01]Ref.01: Schulz, C.R.; Okun, D.A. (1984). Surface water Treatment for Communities in Developing Countries. IT, London. p.193. Available from www.developmentbookshop.com).

The sand used for a slow sand filters should preferably be preferably rounded, and free from any clay, soil or organic matter. If necessary, the sand must be washed before being used. If the raw water is expected to have high levels of carbon dioxide, then the sand must contain less than 2% of calcium and magnesium, calculated as carbonates. This is to prevent the formation of voids in the media if the calcium and magnesium are removed by solution (Huisman and Wood, 1974 [ref.02]Ref.02: Huisman, L; Wood, W.E. (1974). Slow Sand Filtration. WHO, Geneva, Switzerland. p.52. Available from WHO).

Effect of sand size on microbiological quality
Results from some studies on continiually-operated slow sand filters have shown that there is scope for the relaxation of typical values that have been used as benchmarks of slow sand filter design. One such study (Muhammad et al, 1996 [ref.03]Ref.03: Muhammad, N.; Ellis, K.; Parr, J.; Smith, M.D (1996). Optimization of slow sand filtration. Reaching the unreached: challenges for the 21st century. 22nd WEDC Conference New Delhi, India, 1996. pp.283-5. Available online here) done on coarser sand with a constant uniformity coefficient of 2, found that the treatment efficiency (for removal of bacteria, turbidity and colour) of slow sand filters was not very sensitive to sand sizes up to 0.45mm, although a slight increase in treatment efficiency was observed with decreasing sand size. They concluded that from the standpoint of removal efficiency the argument for using very fine sand is not strong.

The ideal range for the uniformity coefficient seems to vary – for example, Ellis (1987) [ref.04]Ref.04: Ellis, K.V. (1987). ‘Slow Sand Filtration’, WEDC J. Developing World Water, Vol 2, pp 196-198. recommends a uniformity coefficient in the range of 1.7 – 3, while one of less than 2.7 is preferable. In practice, it seems that sand that is both finer and coarser than the recommended range still provides acceptable results in terms of filtration in continually-operated systems (Barrett, 1989 [ref.05]Ref.05: Barrett, J.M. (1989) Improvement of Slow Sand Filtration of Warm Water by Using Coarse Sand. PhD Thesis, University of Colorado, USA.).

However, most of this research has been carried out only on continually-operated sand filter systems. In contrast, research carried out on intermittently-operated filters does seem to indicate that sand size is important.

Research done by Jenkins et al (2009, [ref.10]Ref.10: Jenkins, M.W.; Tiwari, S.K.; Darby, J.; Nyakash, D.; Saenyi, W.; Langenbach, K. (2009). The BioSand Filter for Improved Drinking Water Quality in High Risk Communities in the Njoro Watershed, Kenya. Research Brief 09-06-SUMAWA, Global Livestock Collaborative Research Support Program. University of California, Davis, USA. Available here.; 2011, [ref.11]Ref.11: Jenkins, M.W.; Tiwari, S.K.; Darby, J. (2011) Bacterial, viral and turbidity removal by intermittent slow sand filtration for household use in developing countries: Experimental investigation and modeling. Final draft of submitted paper. Dept of Civil & Environmental Engineering, University of California, Davis, USA. The draft is available here, and the published article is available at ScienceDirect. A poster submitted at a household water treatment conference in 2008 also illustrates the findings well, and is available here.) found that filters using finer sand (D10 of 0.17 mm) performed significantly better in terms of bacteria and virus removal than filters using coarser sand (D10 of 0.52 mm). More details here about their research of the effect of sand size and hydraulic loading. Research done by Logan et al (2001) [ref.06]Ref.06: Logan, A.J.; Stevik, T.K.; Siegrist, R.L.; Rønn, R.N. (2001). Transport and fate of Cryptosporidium parvum oocysts in intermittent sand filters. Wat. Res. Vol. 35, No. 18, pp.4359-4369. on intermittent sand filter columns of 60cm sand revealed that fine-grained sand columns (D10 0.16mm) effectively removed cryptosporidium oocysts under the variety of conditions examined, with low concentrations of oocysts infrequently detected in the effluent. Coarse-grained media columns (D10 0.90mm) yielded larger numbers of oocysts, which were commonly observed in the effluent regardless of operating conditions. Factorial design analysis indicated that grain size was the variable that most affected the oocyst effluent concentrations in these intermittent filters.

Sand bed depth

In slow sand filtration, the vertical height of the sand bed that the water has to pass through is important in terms of filtration efficiency. The reasons for this are the existence of biological activity in a sand filter, which is known to occur at depths of up to 0.5m within a sand bed, and the available surface area for mechanical filtration and chemical reactions.

The depth of the sand bed designed for a sand filter must at least reflect this zone of biological activity. In coarser sands, an increased sand bed depth is required as the depth of activity will increase.  However an increased bed depth can contribute to better filtration as greater surface area provides a more intimate contact between the constituents of the raw water, thus speeding up chemical reactions (surface catalysis).  Normally this surface area is increased by using smaller-grained sand. However, the surface area could equally well be increased by increasing the depth of the filter bed but, when it is noted that a depth of 0.6 m of sand comprising grains of 0.15 mm effective diameter presents the same surface area as a depth of 1.4 m of sand of 0.35 mm grains, it is obvious that the application of finer sand is economically a better method (Huisman and Wood, 1974 [ref.02]Ref.02: Huisman, L; Wood, W.E. (1974). Slow Sand Filtration. WHO, Geneva, Switzerland. p.52. Available from WHO).

Effect of sand depth on bacteriological quality
Bellamy et al, (1985) [ref.07]Ref.07: Bellamy, W.D.; Hendricks, D.W.; Longsdon, G.S. (1985) Slow Sand Filtration: Influences of Selected Process Variables. American Water Works Association Journal 77 (12), pp 62-66. suggested that sand height could be reduced to 0.48m with no change in bacteriological removal efficiency. However, Muhammad, et al (1996 [ref.03]Ref.03: Muhammad, N.; Ellis, K.; Parr, J.; Smith, M.D (1996). Optimization of slow sand filtration. Reaching the unreached: challenges for the 21st century. 22nd WEDC Conference New Delhi, India, 1996. pp.283-5. Available online here) concluded that most bacteriological purification occurs within the top 400mm of a sand bed.

They found that bacteriological treatment was not highly sensitive to sand bed depth (Table 3), suggesting that a continually operated slow sand filter bed could be reduced even further to 0.40m and still produce a satisfactory bacteriological quality of water. Other research confirms that the majority of biological processes occur in the top 0.4m of the sand bed (ASCE, 1991 [ref.08]Ref.08: ASCE (1991). Slow sand filtration. Logsdon, G.S. (Ed). American Society of Civil Engineers, New York, USA.). However, while this is generally true, bacteriological treatment efficiency does become more sensitive to depth with larger sand sizes because the total surface area within the filter is reduced in a sand bed with larger grains, as well as higher flow rates potentially increasing breakthrough.

Interestingly, Ferdausi and Bolkland (2000) [ref.09]Ref.09: Ferdausi, S.A.; Bolkland, W. (2000). Design improvement for pond sand filter. WEDC Water, Sanitation and Hygiene: Challenges for the Millennium. 26th WEDC Conference, Dhaka, Bangladesh, pp.212-5. Available online at www.wedc.lboro.ac.uk found adequate faecal coliform removals to below 10 per 100ml in pond filters, which only had sand bed depths of around 30cm.

Effect of sand depth on removal of Cryptosporidium oocysts
Research done by Logan et al (2001) [ref.06]Ref.06: Logan, A.J.; Stevik, T.K.; Siegrist, R.L.; Rønn, R.N. (2001). Transport and fate of Cryptosporidium parvum oocysts in intermittent sand filters. Wat. Res. Vol. 35, No. 18, pp.4359-4369. on intermittent sand filter columns of 60cm sand revealed that while grain size was the variable that most affected the oocyst effluent concentrations, the depth of sand was also important in removal, and became more important for coarser sands (D10 0.90mm). Filters with find-grained sand that were run under a variety of hydraulic loadings (4cm to 20cm) still had no oocysts deeper than the top 10-15cm of sand. In comparison, in coarser-grained sand, oocysts were found at depths ranging from 20cm (4cm hydraulic loading) to 60cm (10 and 20cm hydraulic loading). Sand bed depth therefore becomes increasingly important with coarser sands, and becomes critical when coupled with hydraulic loads of 10cm or 20cm.

Effect of sand depth on turbidity and colour removal
Although bacteriological quality of water does not improve drastically after 0.4m of sand bed, turbidity and colour removal efficiencies were found to definitely improve as bed depth increased beyond 0.4m. This shows that adsorption occurs throughout the filter column in purifying water. Consequently, a decrease in sand bed depth causes a reduction in total surface area of the sand grains and ultimately total adsorption capacity is reduced (Muhammad, et al, 1996 [ref.03]Ref.03: Muhammad, N.; Ellis, K.; Parr, J.; Smith, M.D (1996). Optimization of slow sand filtration. Reaching the unreached: challenges for the 21st century. 22nd WEDC Conference New Delhi, India, 1996. pp.283-5. Available online here)

Effect of sand depth on removal of nitrogenous organic compounds
While most bacteriological purification occurs mainly in the top 0.4m of a filter, this does not mean that there is no biological activity below 400mm. While not sensitive to sand sizes and filtration rates, biochemical oxidation of nitrogenous organic compounds was found to be dependent on sufficient sand bed depth. These compounds were not completely oxidized within the top 0.4m (Muhammad et al, 1996 [ref.03]Ref.03: Muhammad, N.; Ellis, K.; Parr, J.; Smith, M.D (1996). Optimization of slow sand filtration. Reaching the unreached: challenges for the 21st century. 22nd WEDC Conference New Delhi, India, 1996. pp.283-5. Available online here).

References:

Ref 01: Schulz, C.R.; Okun, D.A. (1984). Surface water Treatment for Communities in Developing Countries. IT, London. p.193. Available from www.developmentbookshop.com

Ref 02: Huisman, L; Wood, W.E. (1974). Slow Sand Filtration. WHO, Geneva, Switzerland. p.52. Available from WHO

Ref 03: Muhammad, N.; Ellis, K.; Parr, J.; Smith, M.D (1996). Optimization of slow sand filtration. Reaching the unreached: challenges for the 21st century. 22nd WEDC Conference New Delhi, India, 1996. pp.283-5. Available online here

Ref 04: Ellis, K.V. (1987). ‘Slow Sand Filtration’, WEDC J. Developing World Water, Vol 2, pp 196-198.

Ref 05: Barrett, J.M. (1989) Improvement of Slow Sand Filtration of Warm Water by Using Coarse Sand. PhD Thesis, University of Colorado, USA.

Ref 06: Logan, A.J.; Stevik, T.K.; Siegrist, R.L.; Rønn, R.N. (2001). Transport and fate of Cryptosporidium parvum oocysts in intermittent sand filters. Wat. Res. Vol. 35, No. 18, pp.4359-4369.

Ref 07: Bellamy, W.D.; Hendricks, D.W.; Longsdon, G.S. (1985) Slow Sand Filtration: Influences of Selected Process Variables. American Water Works Association Journal 77 (12), pp 62-66.

Ref 08: ASCE (1991). Slow sand filtration. Logsdon, G.S. (Ed). American Society of Civil Engineers, New York, USA.

Ref 09: Ferdausi, S.A.; Bolkland, W. (2000). Design improvement for pond sand filter. WEDC Water, Sanitation and Hygiene: Challenges for the Millennium. 26th WEDC Conference, Dhaka, Bangladesh, pp.212-5. Available online at www.wedc.lboro.ac.uk

Ref 10: Jenkins, M.W.; Tiwari, S.K.; Darby, J.; Nyakash, D.; Saenyi, W.; Langenbach, K. (2009). The BioSand Filter for Improved Drinking Water Quality in High Risk Communities in the Njoro Watershed, Kenya. Research Brief 09-06-SUMAWA, Global Livestock Collaborative Research Support Program. University of California, Davis, USA. Available here.

Ref 11: Jenkins, M.W.; Tiwari, S.K.; Darby, J. (2011) Bacterial, viral and turbidity removal by intermittent slow sand filtration for household use in developing countries: Experimental investigation and modeling. Final draft of submitted paper. Dept of Civil & Environmental Engineering, University of California, Davis, USA. The draft is available here, and the published article is available at ScienceDirect. A poster submitted at a household water treatment conference in 2008 also illustrates the findings well, and is available here.