Suspended Substance Removal in Wastewater via Sand and Roughing Filtration

Introduction

Water supplied for public use must be potable — that is, satisfactory for drinking purposes from the standpoints of chemical, physical, and biological characteristics. Drinking water should preferably be obtained from a source free from pollution. The raw water normally available from surface water sources is, however, not directly suitable for drinking purposes. The objective of water treatment is to produce safe and potable drinking water. Common treatment processes used for water purification include plain sedimentation, slow sand filtration, and rapid sand filtration, with coagulation-flocculation units serving as essential pre-treatment steps. Pressure filters and diatomaceous filters have been used, though very rarely. Roughing filters are used, under certain circumstances, as pre-treatment units for conventional filters. This paper specifically addresses the removal of suspended substances in domestic wastewater by coagulation using slow sand filtration and roughing filtration.

The application of microfiltration (MF) and ultrafiltration (UF) via slow sand and roughing filtration for drinking water purification became a standard practice during the past two decades. This microfiltration process, which uses slow sand and roughing filtration, will henceforth be referred to as MFP in this paper. The first full-scale applications in this field were reported in 1988 (Amy, 2006). Since then, MF and UF applications have developed into widely established methods of primary treatment, with a steady increase in installed production capacity in recent decades (Furukawa, 2002), both in the European Union and the United States.

By replacing conventional treatment steps — coagulation, sedimentation, and rapid filtration — with microfiltration, a more reliable, robust, effective, and cost-efficient treatment method is introduced (Mallevialle et al., 1996). Compared to traditional treatment methods, further advantages include: (a) stable process performance under varying feed water quality, (b) a smaller physical footprint, and (c) highly automated operation. Most full-scale treatment plants are designed with polymeric MF/UF membranes. On a technical lifetime basis, the purchase of ceramic membranes for an entire drinking water treatment plant was not a competitive alternative due to higher cost. However, recent pilot-scale studies (Loi-Brugger et al., 2006; 2006a; Lerch et al., 2005; 2006) suggested that purifying coagulated surface water with monolithic ceramic microfiltration membranes in constant flux and dead-end mode can be optimised to such an extent that their use becomes competitive with polymeric hollow-fibre membranes. Higher flux, less frequent cleaning, and longer membrane lifetime form the basis for this recent leap in productivity.

Slow Sand Filtration

Due to their size, slow sand filtration systems are typically designed specifically for each site and application. Package slow sand treatment units are available but are not commonly used.

The primary design parameter for slow sand filtration is the filtration rate. Design filtration rates typically range from 0.05 gpm/ft² to 0.1 gpm/ft², although rates as high as 0.15 gpm/ft² may be tolerated for short periods during filter scraping or ripening. Filtration rates can have a significant impact on filter run lengths; lower filtration rates may provide longer filter runs. The appropriate filtration rate should be determined by a pilot study conducted on the raw water to be treated. Using the design filtration rate, the required filter area can be determined for the design flow rate.

Since slow sand filtration requires that a filter be taken off-line for up to two weeks for scraping and filter ripening, more than one filter basin is typically necessary. State regulations require multiple filter units that provide redundant capacity when filters are out of service for backwash or maintenance. This requirement may be waived for non-community water systems that provide engineering justification acceptable to the relevant authority. Each filter basin that can be operated independently is considered an individual filter unit.

The number of filter basins provided will depend on the difference between average and peak flows, the anticipated filter run time, and available storage within the water system. The most conservative system design criterion is to use the maximum day demand as the design filtration rate with one filter basin out of service.

Filter basins can be constructed using concrete or earthen berm construction. For very small systems (under 25 gpm), basins can be constructed from alternative materials such as polyethylene or fibreglass tanks. Except for these very small systems, surface area requirements for slow sand filters make such tank types impractical.

Regardless of the construction material, the tank should be made as watertight as practical because filtered water is collected in the bottom of the tank. For concrete tanks, water-stop material should be used at all construction joints, and hydrostatic relief valves should not be used. For earthen berm construction, continuous geomembrane liners should be used. Integrity testing should be performed on all geomembrane liner seams to verify that no leak paths are present. Care should be taken when installing underdrains and gravel materials on geomembrane liners so as not to damage the liner material. Common wall construction should not be used between basins containing filtered water and unfiltered water due to the potential for contamination.

Geomembrane-lined earthen berms are typically less expensive to construct than concrete basins, but they have a shorter design life. The design life of a geomembrane is typically not greater than 20 years, while the design life for a concrete basin is typically 40–50 years. Geomembrane liners are also less durable than concrete basins, as they can be damaged by activities such as sand scraping and resanding.

In order to fully understand the operating principles of a slow sand filter, it is necessary to have a basic knowledge of how the filter acts both biologically and physically. The treatment process is entirely natural and is simply dependent on the maintenance of the correct environment for the growth of certain beneficial micro-organisms on or near the surface of the sand filter. Soon after the start of the treatment process, a film of these biologically active micro-organisms develops in the filter fabric and at the top of the sand. This film breaks down incoming disease-carrying organisms, converting them into water, carbon dioxide, and other harmless chemicals. At the same time, a large amount of suspended matter — which causes the cloudy or turbid appearance of raw water — is retained in the fabric and sand by simple straining (Amy, 2006).

The continuous straining process will gradually block the pores in both the fabric and sand, which allow water to pass through. This is indicated by a lowering of the water level in the head-loss indicator tube on the outside of the filter tank, while the water level above the sand remains the same. In order to maintain the same flow through the filter, it is necessary to open the outlet valve further. After a certain period — generally 3 to 12 weeks, depending on raw-water quality — the valve will be fully open and the filter so blocked that it is no longer possible to obtain sufficient clean water. At that point it becomes necessary to clean the filter (Amy, 2006).