Effect of Powdered Activated Carbon to Reduce Fouling in Membrane Bioreactors: A Sustainable Solution. Case Study

Abstract

Membrane Bio Reactors (MBRs) are mainly used for industrial wastewaters applications where their costs can be more easily afforded. High costs are basically due to energy consumption and membrane cleaning or replacement. Membrane fouling is responsible for reducing treated water production and increasing maintenance as well as operation costs. According to previous researches, the addition of Powdered Activated Carbon (PAC) in high dosages could reduce membrane fouling; but such concentrations are economically unsustainable for operative conditions. A MBR pilot plant, fed by mixed liquor of a full-scale activated sludge process from a municipal wastewater treatment plant, was operated dosing low PAC concentrations (0, 2, 5, 10 and 20 mg·L⁻¹, respectively). Experiments were also carried out at two different temperatures corresponding to summer and winter conditions. Results indicated that PAC addition was effective at the low dosages (2 and 5 mg·L⁻¹) by reducing the permeate flux loss (from 16 up to 27%, respectively) while higher PAC concentrations turns out in a useless cost increase.

1. Introduction

Although membrane bioreactor (MBR) applications allow several advantages (e.g., higher performances, lower space requirements, lower sludge production) with respect to conventional activated sludge [1,2]. Their extensive application to urban wastewater treatment is still restrained due to the capital as well as the operation and maintenance (O&M) costs (e.g., energy and membrane replacement). Energy consumptions span from 0.50–0.80 kWh·m⁻³ for flat sheet membrane to around 0.15 kWh·m⁻³ for tubular membranes. Membranes and system configurations costs (approximately 43–47€ Equivalent inhabitant⁻¹ for flat sheet and 42–43€ Equivalent inhabitant⁻¹ for tubular) decreased over the past 10 years as a result of their increased diffusion, improvements in process design, more sophisticated control of the operating parameters and backwashing operation strategy. Also their life time has been expanded [3]. One important issue, still affecting the O&M cost, is the membrane replacement due to excessive fouling. Fouling causes significant increase in hydraulic head loss, manifested as permeate flux decline or transmembrane pressure (TMP) increase, depending on whether the treatment is operated under constant-TMP or constant-flux conditions. Therefore, fouling brings to an increase of MBR systems energy demand. Frequent membrane cleaning is therefore required, increasing significantly the operating costs as a result of cleaning agents and production downtime. More frequent membrane replacement is also expected. Many authors [4,5,6,7] have shown that membrane fouling (and deriving energy costs increase) remains the most adverse barrier to the MBR implementation both in urban and industrial wastewater treatment sectors, remaining one of the most challenging issues to face further MBR development.

Many factors can influence membrane fouling [4]. The main factors are: biomass characteristics; extracellular polymers; inorganic precipitates or scalants; colloids; operative conditions [8,9].

Several researches [5,10,11,12] have shown that the addition of Powdered Activated Carbon (PAC) to sludge contributes in reducing membrane fouling. However, most of these contributions regard relatively narrow PAC concentration ranges and mainly industrial wastewater applications, as shown in Table 1. Moreover, economic issue is not fully investigated even if some authors indicate it as the most important criteria to assess MBR applicability.

Remy et al. [5] proposed a very comprehensive resume of the main research contributions on this issue and reported that the addition of low PAC concentrations can increase the permeate flux of about 10% by improving the membrane filtration performance.

Considering other experiences, Pirbazari et al. [10] observed that a PAC concentration of 10 g·L⁻¹, in a cross-flow ultrafiltration-MBR treating high strength landfill leachate, resulted in less fouling. This effect was explained by the deposition of a dynamic and permeable PAC layer on the membrane surface, protecting it from the deposition of foulants.

Table 1. Example of some results from previous researches.

Author	Wastewater	PAC type	Dosage (g·L−1)	Flux reduction or other benefit
[13]	High strength wastewater from an alcohol distillery	Commercial (steam activated wood charcoal)	2.0	PAC addition allowed continuous operation at a constant flux for 20 d without filter change or cleaning. This duration was shorter (8 d) without PAC addition.
[13]	Sugarcane molasses based distillery wastewater (spentwash)		2.0
[14]	Municipal secondary effluent from a traditional active sludge process	Generic	0.75	Sustainable operating time was extended by up to 2 times through PAC addition, reducing membrane fouling.
[15]	Synthetic wastewater	Generic	1.20	Effective flux reduction control was accomplished by adding PAC. The near-critical flux for the PAC system could be raised by about 32%. Operating intervals could be extended about 1.8 times.
[16]	Various	SA Super Picahydro LP27 (Norit)	5.0	Different and interesting results

Table 1. Example of some results from previous researches.

Author	Wastewater	PAC type	Dosage (g·L−1)	Flux reduction or other benefit
[13]	High strength wastewater from an alcohol distillery	Commercial (steam activated wood charcoal)	2.0	PAC addition allowed continuous operation at a constant flux for 20 d without filter change or cleaning. This duration was shorter (8 d) without PAC addition.
[13]	Sugarcane molasses based distillery wastewater (spentwash)		2.0
[14]	Municipal secondary effluent from a traditional active sludge process	Generic	0.75	Sustainable operating time was extended by up to 2 times through PAC addition, reducing membrane fouling.
[15]	Synthetic wastewater	Generic	1.20	Effective flux reduction control was accomplished by adding PAC. The near-critical flux for the PAC system could be raised by about 32%. Operating intervals could be extended about 1.8 times.
[16]	Various	SA Super Picahydro LP27 (Norit)	5.0	Different and interesting results

Ying and Ping [11] reported a similar effect when dosing 0.75 and 1.5 g·PAC·L⁻¹. The scouring effect that permits the removal of deposited foulants from the membrane surface was also reported by Park et al. [17] with a PAC concentration of 5 g·L⁻¹ applied in anaerobic MBRs.

Fang et al. [18] and [19] indicated the adsorption of foulants to the PAC particles as the responsible mechanism (2–5 g·PAC·L⁻¹ activated sludge) of fouling reduction, but also observed as frequent refreshing of the PACs was necessary because foulants saturate them, while operation at an infinite solid retention time (SRT) did not exhibit a positive effect on filterability. Fouling reduction was explained by a stronger sludge floc structure [20,21].

Remy et al. [5,6,7] analyzed the course of the TMP during the critical permeate flux determination. No PAC sludge exhibited a higher TMP than sludge with PAC. It was also shown as PAC-added sludge had a 19% higher critical flux. PAC addition shows an increase in the biggest particle size but a reduction in the mean particle dimension (−30%). In the sludge without PAC the extra shear also caused an increase of supernatant composed of Chemical Oxygen Demand (COD), polysaccharides and multivalent cations (Ca²⁺ and Mg²⁺). The release of the polysaccharides could explain the higher fouling [5].

Fan et al. [22] considered the effects of sludge characteristics on critical flux using a submerged MBR pilot plant applied to urban wastewater working at different operative conditions. Similar results were obtained by Wang et al. [23] in a submerged membrane bioreactor under sub-critical flux operation.

The need to reduce membrane fouling appears to be a critical technical challenge affecting MBR process performance and economics.

The main goal of the present research was to analyze effects of PAC addition in a pilot scale MBR plant, in order to evaluate the most suitable concentrations able to guarantee an improvement of the treatment yield in terms of permeate flux loss. In particular, the results are discussed in order to determine the minimum concentrations reducing the negative effects of fouling, also considering temperature effect due to seasonal change. The approach aims to find sustainable solutions to remove micropollutants from wastewater, reducing the charge on environment and the risks for human health [24].

2. Materials and Methods

All membrane bioreactor experiments were carried out in a pilot plant fed with mixed liquor coming from the activated sludge tank of a full scale wastewater treatment plant (WWTP).

The WWTP attends to a very large basin in an area with very high density of population and industrial activities. Specifically, the municipal to industrial wastewater ratio is about the 70/30. As a consequence, a significant presence of micropollutants and PAHs characterizes the influent [25,26]. The averaged inflow/outflow data in the months of January and July are reported in Table 2.

Table 2. Wastewater treatment plant (WWTP) data during experimental activities (inflow, inlet and outlet quality parameters).

Parameter	January		July
	inlet	outlet	inlet	outlet
Inflow (m3·d−1)	30,000	-	28,000	-
COD (mg·L−1)	105	21	153	17
BOD5 (mg·L−1)	49	6.4	48	4.6
N-NO3 (mg·L−1)	1.9	6.7	4.9	4.9
N-NH4 (mg·L−1)	14.9	0.50	14.6	0.25
TKN (mg·L−1)	17.50	1.59	19.45	1.77
Total phosphorus (mg·L−1)	2.90	0.48	3.20	0.29

Table 2. Wastewater treatment plant (WWTP) data during experimental activities (inflow, inlet and outlet quality parameters).

Parameter	January		July
	inlet	outlet	inlet	outlet
Inflow (m3·d−1)	30,000	-	28,000	-
COD (mg·L−1)	105	21	153	17
BOD5 (mg·L−1)	49	6.4	48	4.6
N-NO3 (mg·L−1)	1.9	6.7	4.9	4.9
N-NH4 (mg·L−1)	14.9	0.50	14.6	0.25
TKN (mg·L−1)	17.50	1.59	19.45	1.77
Total phosphorus (mg·L−1)	2.90	0.48	3.20	0.29

The pilot plant layout is shown in Figure 1. It was located near the full scale activated sludge reactor for minimizing the head loss due to mixed liquor pumping.

Figure 1. Process layout.

Figure 1. Process layout.

The maximum tank volume was 0.050 m³, with an air diffuser system applied to the bottom of the tank. The MBR was a TMP system characterized by a tubular inorganic membrane. MBR system operating conditions and membrane main characteristics are shown, respectively, in Table 3 and Table 4.

Table 3. Membrane bioreactor (MBR) main characteristics and operating conditions.

Parameter	Unit	Value
Reactor volume	L	50
Hydraulic retention time, HRT	h	10
Solid retention time, SRT	d	50
Mixed liquor suspended solids, MLSS	mg·L−1	4
Average temperature	°C	12 (January) and 22 (July)

Table 3. Membrane bioreactor (MBR) main characteristics and operating conditions.

Parameter	Unit	Value
Reactor volume	L	50
Hydraulic retention time, HRT	h	10
Solid retention time, SRT	d	50
Mixed liquor suspended solids, MLSS	mg·L−1	4
Average temperature	°C	12 (January) and 22 (July)

Table 4. Membrane main characteristics.

Parameter	Unit	Value
Membranes type and module model	-	Tubular inorganic membrane porous carbon support (Dow FILMTECTM)
Frame support material	-	AISI 304
Internal diameter	mm	6
External diameter	mm	10
Membrane pores size	µm	0.05
Trans-membrane pressure, TMP	bar	0.8
Range of working temperature	°C	10–40
Max backwashing TMP	bar	1.1
Backwashing period	min	30 (duration: 30 s)

Table 4. Membrane main characteristics.

Parameter	Unit	Value
Membranes type and module model	-	Tubular inorganic membrane porous carbon support (Dow FILMTECTM)
Frame support material	-	AISI 304
Internal diameter	mm	6
External diameter	mm	10
Membrane pores size	µm	0.05
Trans-membrane pressure, TMP	bar	0.8
Range of working temperature	°C	10–40
Max backwashing TMP	bar	1.1
Backwashing period	min	30 (duration: 30 s)

Two sampling campaigns were carried out at different periods (January and July) with the aim of evaluating temperature influence on results. Experiments had the duration of 168 h, and plant operative conditions were kept constant. Tests were carried out both with and without the PAC addition at different concentrations (2, 5, 10 and 20 mg·L⁻¹ respectively). PAC was added to water flux through a preparer-batcher. PAC main characteristics and cost are reported in Table 5.

Table 5. Powdered activated carbon (PAC) main characteristics and cost.

Parameter	Unit	Value
Brunauer-Emmett-Teller (BET) surface area	m2·g−1	600–800
Iodine number	mg·g−1	760
Humidity	%	15.6
Density	kg·m−3	400
Granulometry (refusal on a sieve with a 20 μm diameter)	%	85
Current cost	€·t−1	1,230–1,550

Table 5. Powdered activated carbon (PAC) main characteristics and cost.

Parameter	Unit	Value
Brunauer-Emmett-Teller (BET) surface area	m2·g−1	600–800
Iodine number	mg·g−1	760
Humidity	%	15.6
Density	kg·m−3	400
Granulometry (refusal on a sieve with a 20 μm diameter)	%	85
Current cost	€·t−1	1,230–1,550

Permeate flux J was monitored during the test and used as an indicator of the filtration process performance.

3. Results and Discussion

The results, obtained with a high SRT (50 d), are expressed as percentage of the permeate flux with clean membrane, J_clean (see Figure 2 and Figure 3) .

Figure 2. Influence of PAC concentration on permeate flux—winter conditions.

Figure 2. Influence of PAC concentration on permeate flux—winter conditions.

Figure 3. Influence of PAC concentration on permeate flux—summer conditions.

Figure 3. Influence of PAC concentration on permeate flux—summer conditions.

It can be noticed that MBR filtration performance inevitably decreases with filtration time and PAC addition brings to positive immediate effects (within about 6 h) in permeate flux loss. Moreover, PAC maintains permeate flux more stable over time.

PAC significantly improves the permeate flux loss both in summer and winter conditions in a similar manner, but the performances related to adding 5 mg·PAC·L⁻¹ are higher than dosing 2 mg·PAC·L⁻¹ (about 26–27% with respect to 16–17%).

Starting from the obtained results, a second set of experiments using higher concentrations of PAC (10 and 20 mg·L⁻¹) was carried out for a period of 96 h. Results are shown in Figure 4.

Obtained results do not encourage the application of PAC concentration higher than 5 mg·L⁻¹ because no further improvements in terms of permeate flux are achieved. In fact the permeate flux loss is almost the same (after 4 days the measured value is around 62–65%).

Figure 4. Influence of high PAC concentration on permeate flux—summer conditions.

Figure 4. Influence of high PAC concentration on permeate flux—summer conditions.

4. Conclusions

Results from this research enforce previous experiences reported in technical literature, confirming that PAC addition, in low concentrations, can contribute to reduce the membrane fouling in MBR systems. The enhanced performances have been evaluated through a decrease in the permeate flux loss over time. PAC addition in low dosage (5 mg·L⁻¹) makes possible to halve the permeate flux loss while other tests carried out with higher concentrations did not reveal significant efficiency improvements. Moreover, the temperature influence (considering two series of tests carried out at 12 and 22 °C, respectively) is negligible.

The benefit of PAC addition improves the MBR filtration performances, such as the energy consumption reduction due to mitigation of TMP increase (or flow rate decrease), elongation of cleaning in place as well as physical cleaning intervals.

In conclusion, the solution, considering both material cost and benefits regarding the increased yields, presents a significant level of economic and environmental sustainability for mixed civil and industrial wastewater treatment.