**1. Introduction**

Belt filter presses (BFPs) are mechanical dewatering equipment commonly used in municipal wastewater treatment plants (WWTP)<sup>1</sup> . The main reasons for sludge dewatering are to reduce the volume as it relates to the cost of sludge handling, especially transportation cost, increase in sludge heating capacity<sup>2</sup> , and reduction of leachate [1].

<sup>1</sup> Plate Filter Press installations present a greater capital cost, but have comparatively low running costs. The high running costs of belt presses are mainly attributable to the cost of polymer dosing, and it is the optimization of this cost that is the predominant motive for this chapter.

<sup>2</sup> Dewatered sludge is sometimes incinerated to provide heat energy.

#### *Advances in Slurry Technology*

The mechanical components of a BFP include dewatering belts, rollers and bearings, belt tracking and tensioning system, controls and drives, and a belt washing system. The operation of the BFP includes a polymer conditioning zone, gravity drainage zone, low-pressure zone, and high-pressure zone [2]. A schematic diagram of a BFP is provided in **Figure 1** [1].

The water is released in three stages: (a) free water is allowed to drain through belt pores in the gravity zone and then it is passed through (b) the low-pressure zone where it is gently compressed, and (c) the interstitial water is released in the highpressure zone where it is further compressed through a series of rollers of different sizes [3]. The dewatered sludge also known as the cake is removed by scrapers. The belts are washed by high-pressure jets to remove sludge and polymer deposits from the belt.

The process parameters that affect the BFP performance, that is, the final cake solids concentration and filtrate suspended solids, are the influent sludge characteristics and flow rate, solid feed rate, polymer concentration, dosage and mixing energy, belt speed, and belt tension [4]. There are no fundamental models relating the process conditions to the final cake solid concentration, filtrate suspended solids, or cake solids.

Johnson and coworkers conducted tests to evaluate the effect of process conditions on BFP performance [4]. They evaluated the effect of sludge flow, solid feed, polymer dosage, belt speed, belt tension, and polymer mixing energy. Their tests were conducted over a 6-month period.

The solid feed rate was found to be one of the most critical variables for the optimization of BFP performance. They also emphasized the importance of using polymer prepared at plant scale as mixing energy influences the performance of the polymer. They used both cationic and nonionic polymers. Under-dosing resulted in poor flocculation causing a thin, dry cake. Johnson and coworkers determined the optimum value for each parameter individually and then ran the BFP at the optimized value for each condition. They found that there was no correlation between the

**Figure 1.** *Schematic diagram of the operation of a BFP [1].*

process parameters and the sludge production or cake solids and that it was not predictable and that similar optimization studies should be done when parameters need to change.

Kholisa and coworkers used a factorial trial design and response surface methodology to evaluate the interaction between process parameters for BFP optimization [5]. The work was also conducted at the WWTP at full scale. The plant co-thickens secondary sludge using a cationic poly-electrolyte Flopam 4800, holds it in an aerated holding tank, and dewaters using belt filter press. A four-factor three-level Box– Behnken response surface experimental design (BBD) was used to determine the effect of process conditions on BFP performance. The operating parameters were polymer concentration, polymer dosing, sludge flow rate, and belt speed. The performance measures were sludge cake solid percentage, filtrate suspended solids, and solid capture. The sludge cake solid concentration was found to be constant over the wide range of experimental conditions tested. They found that the sludge flow rate was the most important parameter affecting the filtrate suspended solids and solid capture while the belt speed was insignificant. The results showed clearly that the interaction between sludge flow rate and either polymer dosing rate or polymer concentration was consistently more significant than polymer concentration and polymer dosing rate individually. This was important information that cannot be established by varying single parameters at a time, as in the work of Johnson et al. [4].

Kholisa and coworkers [5] proposed quadratic models to describe the relationship between operating parameters and the filtrate suspended solids (FSSs) and solid capture (SC) based on response surface model as shown in Eqs. (1) and (2)

$$\begin{aligned} \sqrt{\text{FSS}} &= -5.805 - 29.906 \ast \text{A} + 17.013 \ast \text{B} + 0.5495 \ast \text{C} - 1.286 \ast \text{AC} \\ &- 1.325 \ast \text{BC} + 87.184 \ast \text{A}^2 + 0.034 \ast \text{C}^2 \end{aligned} \tag{1}$$

$$\begin{aligned} \text{SC} &= 81.03 + 855.42^\ast \text{A} - 85.85^\ast \text{B} - 8.35^\ast \text{C} - 592.68^\ast \text{AB} + 21.96^\ast \text{AC} \\ &+ 18.14^\ast \text{BC} - 1274.35^\ast \text{A}^2 - 0.47^\ast \text{C}^2 \end{aligned} \tag{2}$$

Where

A = polymer concentration g/L; B = polymer dosing m<sup>3</sup> /hr; C = sludge feed m<sup>3</sup> /hr; D = belt speed Hz; and E = linear screen speed Hz

In addition, Kholisa et al. [5] found a relationship between the yield stress of the sludge and the filtrate suspended solids. An optimum yield stress of 90Pa of sludge was found that was suitable to produce filtrate suspended solid concentrations that were within acceptable limits.

It is important to verify the trend with other filter belt presses and another WWTP especially a plant where there is sludge thickening on linear screen before filter press, which was the objective of this work.
