Current status of the rotating belt filtration (RBF) technology for municipal wastewater treatment

In rotating belt filtration (RBF) (Figure 1), suspended solids are removed as wastewater flows through the inclined section of a continuously rotating belt screen filter (Figure 1). Water that is filtered through the belt is conveyed by gravity to the outlet pipe. An overflow weir is located on the upstream side of the belt to prevent flooding of the RBF if the filtration capacity is exceeded. On the effluent side of the belt, water is conveyed by gravity to an effluent pipe or channel. A seal system prevents the carryover of removable solids into the filtered effluent.

Depending on flow rate and the buildup of solids on the belt screen wastewater rises to a certain level (measured by a sensor) on the influent side. Typically, RBFs are equipped with a control mechanism designed to speed up the rotation of the belt – and consequentially cleaning – as the level of water increases.

Belt screens are available with various opening sizes and materials depending on the specifications, and requirements for each application. Typically, polyester belts with mesh sizes between 50 and 500 micron are used for municipal wastewater applications, but finer meshes are also available. Overall RBF filtration performance can be described as a balance of two processes. The rotating belt screen is cleaned at the end of every filtration cycle so that at the start of every new cycle the filtering mesh is clean. Initially, when the belt is clean, solids are removed by sieving. As solids are accumulated on the mesh they create ‘filter mat’ that is thought to enhance filtration performance by removing solids smaller than the belt mesh size.

Sludge that is accumulated on the mesh is conveyed to the upper portion of the belt and then dropped in a collection trough. A dedicated system (options include air knife, water or mechanical) continuously cleans the filter mesh and dislodges any remaining solids as the belt rotates. The mesh can also be backwashed as needed with hot water to remove oil and grease accumulation.

Sludge collected in the trough is conveyed to one end of the collection trough by means of an auger. Depending on application requirements sludge can be further process through a compression and de-watering section to produce a 20–30% dry solid cake (without the need for any additional dewatering equipment); or as a direct feed (2–8% solids) to a sludge stabilization process (e.g. direct digester feed).

The main drivers for considering of RBF include reduced space requirement, ability to support small mesh without clogging, reduced civil engineering site work, and modular construction allowing for reduced design work, faster installation, and ease of plant expansion. Depending on the location of the prospect installation and process configuration additional potential advantages of RBF may include reduced capital and operation costs, limited operation and maintenance requirements, no requirement for further thickening of collected solids, energy savings (e.g. reduced aeration costs following the addition of primary solids removal by RBF), and overall reduction of CO₂ emissions. In addition, when compared with conventional primary clarifiers RBF are not subject to occurrences such as short-circuiting – due to thermal stratification, wind, density currents, and high flow rates – and biological activity within the sludge blanket at the bottom of the clarifier that can negatively affect the performance of conventional clarifiers.

The focus of most existing applications and testing has been the removal of primary solids. However, filtration of other wastewater streams may be possible with the appropriate selection of filtering mesh size, mode of operation and/or (if required) chemical conditioning.

Existing and potential applications for primary solids removal include:

• Primary solids removal in facilities that do not have primary clarifiers (e.g. lagoons, and oxidation ditches);

• Replacement of inefficient primary clarifiers;

• Plant expansions and plant overflow treatment;

• Pretreatment for a range of biological processes including MBR, MBBR, SBR, BAF, CAS;

• Decentralized ‘scalping’ water reuse plants.

Other potential applications under different stages of development may include:

• Filtration of combined sewer overflows;

• Chemical enhanced primary treatment (CEPT) pretreatment for TSS/BOD removal enhancement;

• Upstream protection for particle-sensitive biological processes (ANAMMOX, UASB, etc.);

• Algae and struvite recovery/harvesting;

• VS enrichment for improved Biogas production;

• Waste activated sludge co-thickening;

• Pre-coating for fine solid cake filtration.

The aim of the paper is to review the current status of development of the RBF technology for municipal wastewater treatment and particularly to provide a review of the relevant published literature.

Results from full and pilot-scale testing reported in the literature indicate that RBF can meet and exceed the performance of conventional primary clarifiers in terms of TSS and BOD removal. Rusten & Ødegaard (2006) conducted a study with contributions from the Norwegian University of Science and Technology, national R&D organizations, consulting companies, water and wastewater utilities, and the financial support from the Norwegian State Pollution Control Agency to test the efficiency of sieve filtration technologies in alternative to primary clarifiers. Rusten and Ødegaard reported that four of the five primary clarifiers examined in their study did not meet EU guidelines (Council Directive 1991) for TSS and BOD5 removal unless they were converted to CEPT processes. They reported TSS and BOD5 removal for wastewater plants in using RBF (six plants), rotating disk screen (two plants), and stationary screen plus drum screen (one plant). The mesh sizes used during testing ranged between 80 and 850 microns. Only the plants using RBF (350 micron mesh) where able to meet EU primary treatment requirements – at least 20% organic matter removal (measured as BOD) and 50% total suspended solids (TSS) removal. Where it is was possible to operate the RBF a low filtration rate (25 m³/m²/h) to favor the development of a sufficiently thick filtration mat on the sieve, average TSS removal up to 90% and average BOD5 removal up to 80% were observed. When higher filtration rates were used, lower performance was observed. For example, at Guldholmstranda WWTP (Norway), 78% removal of SS was achieved at 116 m³/m²/h (350 microns mesh). All RBF in this study used the auger for sludge dewatering of primary sludge, with total solids (TS) concentration in the dewatered sludge ranging between 17 and 37% with an average of 27%.

In 2012, Franchi et al. presented results from 11 weeks demonstration-scale operation of an RBF unit at the UC Davis, CA (USA) wastewater treatment plant. During testing, three alternative polyester wire mesh cloths with 180, 250 and 350 micron mesh size were used. These tests were conducted at filtration rates ranging from 39 to 235 (typical 150) m³/m²/h. For the 350 micron belt, TSS test results were comprised between 30 and 65% TSS removal (the average removal over the testing period was 54%). The envelope of TSS removal for the 250 and 180 micron belts was similar to that observed for the 350 microns. However, the average removal was higher, around 60%. This difference was attributed to the finer mesh size providing greater solids retention during the initial stages of the filtration cycle. For the 350 micron belt, the majority of results were within Ten State Standards (‘Primary settling of normal domestic wastewater can be expected to remove approximately one-third of the influent BOD when operating at an overflow rate of 41 m³/(m² d) [1,000 gallons per day/square foot]’) and EU Standards (European Community standard for BOD removal is 20%). For the finer meshes (180 and 250 microns) the average BOD removals was nearly 40% and above.

Franchi et al. (2012) reported that effluent TSS values were strongly correlated with influent TSS. This occurrence was attributed to the formation of a thin layer of particulate material on the surface of the belt during filter operation. It is the presence of this mat which results in a ‘self-filtering sieve’ effect sensitive to the influent TSS concentration but not to flow conditions because of the automatic adjustment of the belt speed. Figure 2 illustrates the removal (indicated by lines of different colors) of TSS for different influent TSS concentration ranges. Clearly, TSS removal increased with increasing influent TSS concentration. Average TSS removal went from 30% at the lowest influent TSS range (<100 mg/L) to 65% at the highest influent TSS range (>300 mg/L). Solids moisture content during testing ranged between 30 and 35% solids content. These results show that effective mechanical dewatering can be achieved by the auger section of the RBF. The dry solids routinely pass the US EPA ‘Paint Test.’

The relationship between TSS removal efficiency and TSS concentration entering the RBF technology was investigated with pilot experiments conducted in London ON, (Canada) by Santoro et al. (2014). The RBF technology was operated with a 350 um mesh and tested under different flowrates. The RBF was controlled automatically by keeping the upstream water level to constant set-point (30 cm) while letting the belt speed vary. It should be noted that, during testing, the applied flowrate was also varied within the 120 and 370 gallons per minute range. Figure 3 illustrates the results obtained during testing. A general increase in removal with increasing TSS concentration can be observed in Figure 3 (also note that Figure 3 includes the data presented in Figure 2 from Franchi et al. (2012)).

Peeters et al. (2014) tested RBF with three different belt opening sizes upstream of an MBR system. The median removal during the course of these experiments was 49% (750 micron), 66% (350 micron) and 72% (154 micron) for TSS and 16% (750 micron), 30% (350 micron) and 39% (micron) for chemical oxygen demand (COD). Based on these results, Peeters et al. (2014) conducted energy balance calculations for RBF + MBR. Their conclusion was that RBF can provide substantial removal of organics from biological treatment resulting in smaller biological reactors (up to −38%) and lower oxygen requirements (up −25%) when compared to a conventional MBR without primary.

RBF performance is site specific and it is thought to be depended on the amount and quality of solids in the raw water. Specifically, higher solids concentration may lead to the formation of a thicker ‘filtering’ mat on the surface of the belt. In terms of the effect of the quality of solids, it is thought instead that the presence of fibrous materials may lead to the formation of a thick and porous filtration mat that, in turn, enhances TSS and organics removal. Moreover, a RBF technology can be operated using different control strategies. For example, hydraulic capacity can be maximized ensuring that the water level upstream the filter is always at the highest set-point. Under those circumstances, the TSS removal performance is minimized as the thin layer of particles accumulated on the belt during the filtration cycle is also minimal.

Similarly to CEPT in primary clarifiers, RBF performance can be enhanced through chemical addition. In this case, we refer to chemically enhanced primary filtration (CEPF). Tests conducted up to date show that polymer addition upstream the RBF is more effective than the addition of metal salts and flocculant aids. This occurrence can be attributed to faster particle growth kinetics and the agglomeration of colloids during the coagulation step. As specified in EPA Method 160.2 (US EPA 1983) and Method 2540 D (Standard Methods 2012), TSS is the retained material on a standard glass fiber filter after filtration of a well-mixed sample. A large fraction of colloids (defined as particles smaller than 1 micron in at least one dimension) are not retained by the filters (0.7–1.5 micron range) approved by Standard Methods. Thus, coagulation upstream or downstream the RBF can ultimately result in increased TSS and giving the false impression of poor filter performance even if solids are effectively removed by the belt. Polymer on the other had tends to facilitate the aggregation of already formed flocs and not so much of colloidal material.

Bench – scale studies (sieve and column tests) conducted by Trojan Technologies in Australia in 2014 (Trojan Technologies 2014) showed that the addition of cationic polymer (Table 1) significantly increased the removal of TSS. Without polymer addition, TSS removal ranged between 38 (350 micron mesh) and 44 (250 and 158 micron mesh). BOD removal with the 350 micron belt (BOD was not measured for the other two mesh sizes) was 20% and COD removal was 23%. Because the TSS and BOD removal did not meet the facility requirements, additional tests were conducted with polymer addition before RBF filtration. Those results are presented in the next section. The polymer dose added in these tests was 2.0 mg/L (when used in conjunction with a coagulant, typical dosage is full-scale ranges 0.5–1 mg/L depending on the type of polymer and wastewater concentration). For the three mesh sizes using in these tests, the increase was: for 350 micron from 38 to 72% (+ 38%) and for both 250 and 158 micron from 44 to 74% (+30%). Polymer addition was also beneficial in increasing organic removal. The use of polymer doubled the removal of COD (from 23 to 46%) and BOD (from 20 to 41%). Following polymer addition, 16% total phosphorus removal was also achieved (no removal of phosphorous without polymer). A summary of these test results is provided (with and without polymer addition) in Table 1 (for TSS filtered through 350, 250 or 158 micron mesh) and Table 2 (for other parameters monitored filtered through 350 micron mesh).

Rusten & Ødegaard (2006) described the results of tests for CEPF for a facility (Bangsund WWTP, Norway) where the characteristics of the wastewater not favorable to RBF (smaller particles) requiring preconditioning of TSS upstream of RBF. The original testing train included a 850 micron RBF followed by a static mixer for coagulant addition and a final 250 micron RBF for the removal of coagulated solids. Best results were obtained when the 850 micron RBF was bypasses and a 1 mg/L dose of cationic polymer (Pemcat 163) was added upstream of a 250 micron RBF. Even in this case, polymer addition was more effective than coagulation by metal salts followed by flocculation. When the 250 micron RBF alone was used, the average TSS removal during six testing events was 66%. In terms of screw press sludge dewatering, Rusten and Ødegaard did not observe differences with or without CEPT.

The RBF process occupies a fraction, approximately 1/10th, of the space requirement of a conventional clarifier. For TSS removals between 40 and 70%, typically achieved by RBF during the filtration of municipal wastewater in North America and Europe, it corresponds to approximately 1.0 and 7.0 hours of detention time in an ideal conventional primary clarifier (M&E 2003). Overflow rates, typically, used in wastewater treatment plants range between 30 and 50 (average 40) m³/m²d which result in nominal detention times between 2.0 and 2.5 (average 2.0) hours (M&E 2003). Current flow rates per square meter of space requirement for RBF range between 300 and 500 m³/m² d (Salsnes 2014) depending on the RBF configuration (larger units have greater filtration capacity on overall space requirement).

Depending on plant size, type of design and characteristics of the wastewater, biological wastewater treatment consumes from 50 to 80% of an entire plant's power. By reducing the TSS and BOD loading in primary treatment, the power use in the biological process can be considerably reduced primarily due to lowered aeration needs.

Table 3 presents power use calculations conducted for a 36,000 m³/d (10 MGD) wastewater treatment plant using activated sludge and aerobic digester (case A) or activated sludge and anaerobic digester (case B). For each case, power uses were calculated for various process units when primary solids were not removed (no primary) versus when RBF (without chemical addition) was used for primary solids removal. Power requirements for the different processes were obtained from the sources listed in Table 3. Particularly, energy use for aeration in activated sludge were estimated using the web-based program Aquifas (Aquifas 2012) and values reported in surveys by EPRI (2002), Energy Center of Wisconsin (2003), NYSERDA (2008), and Pennsylvania Dept. of Environmental Protection (2011). It was assumed that all primary solids would be conveyed to the digesters. No power generation credit was given to the anaerobic digestion system because of the relatively small size of the plant. Calculations were conducted for influent wastewater containing 250 mg/L BOD and 250 mg/L TSS. Removal by the RBF process was assumed to be 30% for BOD and 50% for TSS (these assumptions are based on the performance data reported in this paper).

As one could intuitively expect the addition of RBF to a plant that does not practice primary solids removal resulted in the significant decrease of power use: 22% for the case of aerobic digestion (where a larger fraction of power use is for the digestion process) and 28% for anaerobic digestion. Power savings were calculated for aeration in secondary treatment with an overall plant energy saving of 19% for the case of aerobic digestion and 24% for anaerobic digestion. Smaller savings (4–5%) were also calculated for secondary thickening because of the reduced BOD loading to the activated sludge (creating less biological solids). In summary, the calculations presented in Table 3 suggest that considerable energy savings can be achieved with the addition of RBF to a treatment plant that does not implement primary solids removal.

Regarding the energy potential of the harvested sludge, Paulsrud et al. (2014) reported on one study on the comparison between sludge from fine mesh sieves with sludge from primary clarifiers (primary sludge) with focus on anaerobic digestion and/or incineration. Samples of sludge were analyzed for the content of dry solids, volatile solids, COD, calorific value and methane potential. Results showed that that sieve sludges had significantly higher volatile solids content and higher methane potential than primary sludges. Particularly, methane potential results reported on a dry solids basis showed mean values of 318 NmL CH₄/g DS for the sieve sludges and 229 NmL CH₄/g DS for the primary sludges. According to Paulsrud et al. this was a clear indication of increased sludge energy potential if fine mesh sieves are used for primary treatment instead of primary clarifiers at WWTPs with anaerobic digesters.

Remy et al. (2014) tested the concept of maximum extraction of organic matter into the sludge based on enhance the recovery of energy during anaerobic digestion and decrease aeration demand during the mineralization of carbon. The anaerobic digestion of primary sludge collected through coagulation, flocculation and microsieving (70 to 80% COD removal) was compared to the digestion of mixed sludge from a conventional activated sludge process. The biogas yields per organic dry matter input (oDMin) of the primary sludge was 600 NL/kg oDMin compared to 430 NL/kg oDMin for mixed sludge from a conventional activated sludge process.

It is expected that the effect of RBF on the COD/N ratio will be similar to that of primary clarifiers. Both technologies remove similar amounts of COD and neither RBF nor conventional primary clarifiers are expected to remove a significant amount of readily biodegradable BOD5 or COD (Barnard 1984). Limited testing conducted at the Trojan Technologies’ lab in London have shown nitrogen removal in the range of 20% which is comparable to what is reported in the literature for conventional settling (16%) by Henze & van Loosdrecht (2008).

As long as the readily biodegradable COD/N ratio in the effluent stream is not altered, no impact should be expected for the downstream biological nitrogen removal processes; however, a plant-wide analysis of extreme scenarios when primary processes are designed to remove >80% TSS should be conducted to better understand the impact of nutrient loading coming from return side-streams as well as the carbon balance required for nutrient removal biological processes. It should be also mentioned that, since CEPT or CEPF processes are based on the concept of carbon diversion (rather than carbon degradation), other strategies are also possible to minimize the use of an external carbon source for biological nutrient removal processes; for example, dynamic carbon addition strategies through the use of a nitrate probe in the effluent, or alternatively to filter a portion of the raw wastewater without chemical addition to use the dissolved COD for denitrification (Remy et al. 2014).

Finally, future primary solids removal schemes may also contemplate the use of enhanced primary treatment processes as a pretreatment for biological processes requiring low and well-controlled COD/N ratios such as Anammox, Nereda, algal reactors, or removal of NH₄+ by physical methods such as ion exchange.

RBF, alone or in combination with low-dose polymers, is a technology for the removal of particulate from wastewater that has been recently undergoing intensive testing and technological development, reaching a maturation point and a number of full-scale installations worldwide that it can be confidently recommended as an alternative to primary clarifiers and other solids removal processes. As a matter of fact, RBF performances reported in the literature are comparable with the ones obtained using conventional clarifiers and CEPT. Calculations based on industry design standards show that RBF requires 1/10th of the space requirement of conventional clarifiers. This is of great importance for retrofitting of treatment plants where space limitation or real estate value is an issue.

Full-scale and pilot testing have shown that RBF without polymer addition can remove greater than 30–60% TSS and greater than 15–30% BOD depending on TSS inlet concentration. Additional data from pilot trials reported in this paper suggest that when polymer is added upstream, RBF can reduce TSS by 70% and BOD to up 40%.

As energy savings in plant has become a main focus of treatment plant retrofit and new design, the use of RBF as a primary treatment has gain considerable attention. Wastewater treatment plants account for 3% of the total energy used in America (EPA 2006), and the electricity demand is estimated to increase 20% in the next 10 years with population growth and stricter regulations. The power requirements for the processes in a plant that comprise this energy demand are directly related to the biological and solids loadings delivered to them. When RBF is retrofitted in a plant that does not have a primary solids removal process, considerable energy savings can be achieved mainly because the biological processes are able to operate with lowered aeration requirements. Calculations conducted with a wastewater treatment simulation program (Aquifas) and sources found in the literature can be used to estimate this benefit.