Scaling filtration time initial dependencies of wastewater sludges

Abstract

The filtration time, t_f, during constant pressure dead-end filtration testing of wastewater sludge is dependant on the initial height, h₀, and the initial solids concentration, φ₀. The theoretical dependencies of these initial conditions are explored: t_f varies with $h_0^2$ and $c{\phi }_0^2$, where c is a material dependant parameter that is also dependant on φ₀ and the applied pressure. Empirical values for c relative to a given φ₀ are determined from phenomenological filtration theory to give a qualitative scaling method to compare the filtration behaviour of highly compressible materials under differing initial conditions. The method is validated using filtration testing of municipal wastewater sludge. This new scaling method is applied to the filtration results of a range of different wastewater sludges, additives and treatments to illustrate its application for plant comparisons, polyelectrolyte comparisons, dose optimisation of polyelectrolyte and ferric chloride and combinations thereof, and the effects of two physicochemical treatments.

Introduction

Wastewater sludges are often filtered prior to disposal to increase the solids content and therefore reduce transport and disposal costs. However, they are highly impermeable materials requiring long process times to reach desired solids concentrations. Maximising the throughput and final cake solids of unit operations by choosing the right treatments and additives to give easily dewaterable sludge is becoming increasingly important due to escalating disposal costs. Constant pressure dead-end filtration testing is often performed in order to study sludge dewatering behaviour. At a given applied pressure, ΔP, the time to filtration, t_f, to reach an average solids volume fraction, φ_f (or specific volume of filtrate, V_f), is dependant on the sludge material properties, the initial height, h₀, and the initial solids concentration, φ₀ (Landman and White, 1997). The material properties of wastewater sludge are dependant on the method of biological treatment, physicochemical treatment prior to filtration, and the type and dose of dewatering aids such as polyelectrolytes.

The prediction of the filtration performance of compressible suspensions requires material characterisation using constant pressure tests over a range of pressures (Green et al., 1998). Permeability information is extracted from the slope of t vs. V² and compressibility information from the equilibrium height. The slope is constant during cake formation and increasing to infinity (signifying equilibrium) during cake compression. Potable water treatment sludges, which do follow traditional sludge behaviour (an example is shown in Fig. 1(a)), have been modelled successfully with the Landman and White fundamental theory (Harbour et al., 2004; Stickland et al., 2006).

Conversely, constant pressure dead-end filtration of wastewater treatment sludge shows short linear cake formation compared with extremely long cake compression (Scales et al., 2004). An example of such ‘non-traditional’ behaviour where non-linear cake compression dominates the process is shown in Fig. 1(b). The extraction of meaningful material characteristics is difficult since the short cake formation competes with initial pressure fluctuations and the long cake compression competes with biological activity. Experimental difficulties aside, the characteristics of wastewater sludges can be elucidated from the cake compression regime of single-pressure tests (Stickland, 2005), but the multiple single-pressure tests required for full characterisation are time consuming due to the long filtration times required.

A method is required that allows unambiguous comparisons of wastewater sludge without relying on time consuming and tedious parameter extraction procedures. Capillary suction time (CST) tests (Baskerville and Gale, 1968) are often used to compare wastewater sludges, but such tests represent the behaviour at very low pressures and are often affected by sedimentation. Wastewater sludges are very compressible materials, such that high solids concentrations are possible at medium to high pressures, and the permeability may change with solids concentration by orders of magnitude. Therefore, extrapolation of CST results to filtration behaviour is inaccurate. If the high-pressure filtration behaviour is to be studied, high-pressure tests (filtration or centrifugation) must be performed. Also, CST tests do not provide fundamental information since they are dependant on a range of empirical factors (Smiles, 1998). As with filtration time, CST is also dependant on the initial height and solids concentration.

Normally, if sludges from different treatment plants or the same sludge with different treatment or conditioning are to be compared, they must be tested at the same initial conditions or else one sludge may appear to be more dewaterable than another. For example, when polyelectrolyte is used to condition sludge, the solids consolidate and produced large quantities of free water, such that transfer of the sludge to filtration equipment may result in changes to φ₀. Addition of solids (such as ash or lime as conditioning agents) or coagulants (such as ferric salts) that form solids in situ, and physicochemical processes such as thermal hydrolysis may also change φ₀. Thus, any differences observed in filtration experiments that would ordinarily be attributed to treatment or conditioning may be primarily due to changes in φ₀ rather than changes to material properties.

The purpose of this work is to detail a method for scaling filtration results that allows for changes in initial conditions. In the theory section, the h₀ and φ₀ dependencies of the constant pressure filtration theory of Landman and White (1997) are investigated and used to provide an empirical method of scaling these dependencies. This theory has been shown (Landman and White, 1994; Stickland, et al. 2005) to be related to conventional filtration theories such as specific resistance to filtration (SRF, Ruth, 1946) and cake consolidation (Shirato et al., 1986). Therefore, the theoretical relationships developed here are applicable conventional theories.

In the first part of the results and discussion section, the scaling method is validated using filtration tests of the same activated wastewater sludge at different initial conditions. The second part presents results for using the scaling method to qualitatively compare municipal sludge from different plants with a range of treatment conditions, additives and doses in terms of their dewaterability. Finding the best treatments and additives to produce a dewaterable wastewater sludge is the subject of much research (for example, Koppers et al., 1990; Novak et al., 1999;Chen et al., 2001; Neyens and Baeyens, 2003; Buyukkamaci, 2004), although the aim here is limited to illustrating the use of the scaling method as a useful comparative tool under a range of variables rather than exhaustively covering all treatments and additives.