A comparative examination of MBR and SBR performance for the treatment of high-strength landfill leachate

Abstract

The management of landfill leachate is challenging, with relatively limited work targeting high-strength leachate. In this study, the performance of the membrane bioreactor (MBR) and sequencing batch reactor (SBR) technologies are compared in treating high-strength landfill leachate. The MBR exhibited a superior performance with removal efficiencies exceeding 95% for BOD5, TN, and NH₃ and an improvement on SBR efficiencies ranging between 21 and 34%. The coupled experimental results contribute in filling a gap toward improving the management of high-strength landfill leachate and providing comparative guidelines or selection criteria and limitations for MBR and SBR applications.

Implications

While the sequencing batch reactor (SBR) technology offers some flexibility in terms of cycle time and sequence, its performance is constrained when considering landfill leachate associated with significant variations in quality and quantity. Combining membrane separation and biodegradation processes or the membrane bioreactor (MBR) technology improved removal efficiencies significantly. In the context of leachate management using the MBR technology, more efforts have targeted low-strength leachate with limited attempts at moderate to high strength leachate. In this study, the SBR and MBR technologies were tested under different operating conditions to compare and evaluate their feasibility for the management of high-strength leachate from a full-scale operating landfill. Such a comparison has not been reported for high-strength leachate.

Introduction

Controlled landfilling remains the most economic and commonly practiced method for municipal solid waste (MSW) disposal, with leachate generation being an inevitable consequence. Leachate is formed when the refuse moisture content exceeds its field capacity, which is defined as the maximum moisture that is retained in a porous medium without producing downward percolation. The quantity of leachate formed at a landfill is influenced by factors that contribute to landfill moisture (such as rainfall, snowmelt, groundwater intrusion, initial moisture content, irrigation, recirculation, liquid waste co-disposal, and refuse decomposition) and those that affect moisture distribution within the landfill (such as refuse age and pretreatment, compaction, permeability, particle size, density, settlement, vegetation, cover, sidewall and liner material, and gas and heat generation and transport) (El-Fadel et al., 1997, 1999, 2000, 2002, 2003). Similarly, the leachate quality is affected by several factors, such as the refuse soluble content, as well as by-products of chemical and biological processes within the landfill. Leachate quality can exhibit considerable spatial and temporal variations depending on site operations, refuse characteristics (composition and age), and internal landfill processes requiring varied management options. For this purpose, several options can be adopted, including discharge into sewer systems for subsequent treatment with municipal wastewater (Çeçen and Çakiroğlu, 2001; Trabelsi et al., 2011; Puszczało et al., 2010), recirculation (Rodríguez et al., 2004; Sethi et al., 2013), evaporation followed by sludge disposal (Afsharnia et al., 2012), and on-site treatment and disposal (Bodzek et al., 2006; Nurisepehr et al., 2012).

Historically, various physical, biological, and chemical processes have been applied for leachate management, depending on its quality, discharge requirements, and financial constraints (Renou et al., 2008). In this context, the sequencing batch reactor (SBR) technology is a development of the conventional activated-sludge process whereby the treatment steps are combined into a single basin or tank, often referred to as an activated sludge process occurring in time rather than in space (Singh and Srivastava, 2011). The technology offers a distinctive flexibility in terms of cycle time and sequence (Monclús et al., 2009). Thus the SBR technology can be desirable when considering landfill leachate associated with significant variations in quality and quantity.

More recently, combining membrane separation and biodegradation processes led to the development of the membrane bioreactor (MBR) technology, which is increasingly recognized as the process of choice for the treatment of wastewater streams containing complex and recalcitrant compounds (Bilad et al., 2011; Ahmed and Lan, 2012). An MBR can be considered as a conventional activated sludge system with efficient membrane filtration that holds small pore sizes of less than 0.1 μm (Santos and Judd, 2010). In the context of leachate management using the MBR technology, more efforts have targeted low-strength leachate (chemical oxygen demand [COD]: 1970–2430, 1391–3977, and 1200–1600 mg L^–1; Laitinen et al., 2006; Tsilogeorgis et al., 2008; Zhang et al., 2013) with fewer attempts at moderate to high strength leachate (COD: 4113–9257, 10,435–31,318, and 16,360 mg L^–1; Xiu-Fen et al., 2011; Bai et al., 2011; Campagna et al., 2013). In general, considerably high biochemical oxygen demand (BOD) removals (90–99%) were attained with the MBR technology irrespective of experimental conditions and leachate maturity. In contrast, the efficiency of the MBR in removing COD varied significantly from as low as 25% to as high as >98% (Puszczało et al., 2010, Aloui et al., 2009, Jia et al., 2009; Jakopović, et al., 2008; Chen and Liu, 2006). Further work has been reported regarding the suitability and viability of the MBR technology for the treatment of leachate mixed with municipal wastewater (Ahmed and Lan, 2012), although potential biological inhibition and plant overloading are of concern in this context (Çeçen and Aktaş, 2004). In this study, the SBR and MBR technologies were tested under different operating conditions to evaluate their feasibility for the management of high-strength leachate from a full-scale operating landfill.

Material and Methods

Leachate characterization and infeed preparation

Leachate was collected from a landfill located in the Bekaa valley (33.83° N, 35.91° E) at 70 km east of Beirut, Lebanon. The landfill site occupies ˜180,000 m² at an altitude of ˜1000 m above mean sea level. The facility receives ˜200 tons per day of MSW with ˜70% as organic food waste. Samples of leachate were first analyzed for a wide range of indicators using standard methods for the examination of water and wastewater (Table 1) (APHA 2005). The results defined the main indicators to be used in the evaluation of the performance of the experimental systems, whereby those present at significantly high levels (BOD₅, COD, total nitrogen [TN], ammonia [NH₃-N]) were selected as performance indicators. Following the initial screening, a larger volume of leachate (from the landfill and the evaporation pond) was collected and transported to the laboratory using 20-L containers, which were stored at 4ºC to feed the system at a later stage. Selected indicators (volatile suspended solids [VSS], BOD₅, COD, TN, NH₃-N) were monitored periodically to evaluate the systems’ performance.

Table 1. Screening of leachate quality

Parameter	Unit	L01^a	L02^a	L03^b	Analysis method	MoE, 2001	US EPA, 1992^e	FAO, 1976^f
pH	pH units at 25°C	7.26	7.91	7.56	Electrometry	6–9
EC	μSiemens/cm at 25°C	38,200	50,400	45,500	Electrometry	—
Turbidity	NTU	175	175	250	Nephelometry	—
Apparent color	PtCo units	13,550	14,300	9250	Colorimetry	—
Total Alkalinity	mg L^–1 as CaCO3	15,460	16,380	16,220	Volumetric titration	—
TDS	mg L^–1	19,100	25,200	22,800	Electrometry	—
TSS	mg L^–1	635	928	798	Gravimetriy	60
Total phosphorus	mg L^–1 P	39.5	25	27	Persulfate digestion/colorimetry	10
COD	mg L^–1	17,760	16,520	9520	Closed reflux, colorimetry	125
BOD5	mg L^–1	10,935	9720	4890	Membrane electrometry	25
Ammonia	mg L^–1 NH3	3600	1800	4700	Salicylate method/colorimetry	10
Total nitrogen	mg L^–1 N	4200	2200	5250	Persulfate digestion/colorimetry	30
Chlorides	mg L^–1 Cl^–	7800	14,700	9300	Volumetric titration	—
Ortho-phosphates	mg L^–1 PO4^3–	25.5	11.0	28.0	Ascorbic acid/colorimetry	—
Total hardness	mg L^–1 as CaCO3	2450	2,150	810	Volumetric titration	—
Fecal coliforms	CFU/50 ml	ND^c	ND	ND	Membrane filtration	—
Total coliforms	CFU/50 ml	ND	ND	ND	Membrane filtration	2,000
Li	ppm	0.012	0.233	0.215	ICP-MS^d	—	2.5	2.5
Be	ppm	<0.002	<0.002	<0.002	ICP-MS	—	0.5	0.5
Ti	ppm	0.707	0.526	0.934	ICP-MS	—
V	ppm	0.258	0.323	0.361	ICP-MS	—	1	1
Cr	ppm	0.461	0.685	0.481	ICP-MS	2	1	1
Mn	ppm	0.193	0.329	0.234	ICP-MS	1	10	10
Co	ppm	0.112	0.119	0.116	ICP-MS	0.5	5	5
Ni	ppm	0.481	0.75	0.541	ICP-MS	0.5	2	2
Cu	ppm	0.007	0.039	0.031	ICP-MS	0.5	5	5
As	ppm	0.05	0.176	0.104	ICP-MS	0.1	2	2
Se	ppm	0.042	0.086	0.029	ICP-MS	—	0.02	0.02
Sr	ppm	0.516	0.675	0.438	ICP-MS	—
Mo	ppm	0.027	0.116	0.035	ICP-MS	—	0.05	0.05
Ag	ppm	<0.002	<0.002	<0.002	ICP-MS	0.1
Cd	ppm	0.008	0.003	0.009	ICP-MS	0.2	0.05	0.05
Sb	ppm	0.004	0.029	0.013	ICP-MS	0.3
Ba	ppm	0.779	0.135	0.235	ICP-MS	—
W	ppm	0.015	0.021	0.024	ICP-MS	—
Tl	ppm	<0.002	<0.002	<0.002	ICP-MS	—
Pb	ppm	0.007	0.023	0.018	ICP-MS	0.5	10	10
U	ppm	<0.002	<0.002	<0.002	ICP-MS	—
Zn	ppm	0.05	0.07	0.053	ICP-MS	5	10	10

Notes: ^aL01 and L02: fresh leachate samples.

^bL03: sample from evaporation pond.

^cND: not detected.

^dInductively coupled plasma–mass spectrophotometry.

^eUS EPA recommended limits for reclaimed water for irrigation (short term use).

^fRecommended maximum concentrations of trace elements in irrigation waters, for use up to 20 years on fine textured soils of pH 6.0 to 8.5.

MBR setup

The experimental setup of the MBR system consisted of an anoxic tank followed by an aeration tank integrated with an ultrafiltration (UF) unit to clarify the effluent. The UF membrane ZW-10 was provided by ZENON and has a contact area of ˜10 ft² (0.929 m²) divided in several hollow-fiber membranes with nominal pore size of 0.03 μm, a dry weight of ˜4.2 lb (1.9 kg), and a wet weight of ˜4.6 lb (2.1 kg). The membrane was kept in its original package soaked with glycerine until the experimental program was initiated.

The system components included a Plexiglas tank for housing the membrane with a capacity of ˜20 L preceded by two larger plastic process tanks with the first one utilized as an anoxic tank and the second as an aeration tank (Figure 1). A pressure sensor was connected to a Sensotec digital display with an accuracy of ± 0.05% to record pressure variations in the membrane and to define the backwash frequency for periodic cleaning. The system was equipped with several pumps with variable speed and reverse operation modes to ensure continuous flow: a permeate and backpulse pump (Masterflex series 7554-20; Cole-Parmer, US) with a flow range of 0–60 L hr^–1; a recirculation pump (Masterflex series 7524; Cole-Parmer, US) with a flow range of 0 to 140 L hr^–1; and a feed pump (Masterflex series 7550; Cole-Parmer, US) with a flow range of 0-20 L hr^–1. An aeration blower that can provide an airflow of up to 3000 L hr^–1 was used to help in scrubbing the membrane, preventing potential fouling, and increasing the backpulsing rate. The aeration was supplied through five air diffusers distributed equidistantly in the aeration tank to provide uniform diffusion in the tank while the temperature was maintained at about 22 ± 1ºC in a temperature-controlled room.

Figure 1. Experimental setup of the MBR system. (a) Infeed/raw leachate; (b) Mixer; (c) Anoxic tank; (d) Aeration tank; (e) Air diffusers; (f) Air blower; (g) Pressure sensor; (h) Permeate/Backpulse pump; (i) Hollow fiber MBR; (j) Recirculation, 5-10Q; (k) Return sludge; (l) Drain; (m) Permeate tank, (s) Sampling location.

The tanks were first filled with the influent leachate while opening the aeration valves in the aerobic tank, turning on the mixer in the anoxic tank at a low speed (˜100 rpm), and recirculating the flow between all tanks without feed or permeation. This was recorded as phase 1 of the experimental program and was intended for the acclimation of bacteria. The bacterial activity was observed through the temporal increase of mixed liquor volatile suspended solids (MLVSS) in the aerobic tank. The second phase required increasing the influent flow rate. During each phase, the system was operated without sludge wastage while changing several parameters to maintain a reasonable variation in the transmembrane pressure and a constant flux. Table 2 summarizes the experimental program with the duration and range of parameters applied. Parameter variation was based primarily on the observed flow rate. For instance, when the flow increased in the permeation, the backpulse frequency and scour airflow were increased to maintain a constant flow and clean the membrane. Membrane cleaning was conducted periodically to help reduce fouling and maintain a constant membrane flux. This was achieved by reversing the flow through the membrane using the temporarily stored permeate. Based on the recommendation of the manufacturer (ZENON), periodic chemical cleaning was performed on a need basis using sodium hypochlorite diluted with water at a concentration of 250 mg L^–1 for a minimum of 3 backpulses.

Table 2. MBR experimental program

Indicator	Phase 1, acclimation	Phase 2, optimization
Duration, week	2	14
Membrane scour airflow, m^3/hr	1–2	1–3
Hydraulic retention time, hr	0–20	0–20
Backpulse frequency, minutes	5–120	5–30
Backpulse duration, seconds	30	30
Backpulse flow, L/hr	5–20	5–20

SBR setup

The bench-scale SBR experimental setup (Figure 2) comprised two cylindrical tanks of 17.4 cm diameter, 40 cm depth, and a total volume of 9.5 L each. The first tank served as the leachate feeding container, while the second tank served as the SBR. An air diffuser was installed at the base of the cylinder to maintain the dissolved oxygen (DO) level between 6 and 8 mg L^–1 during the aerated phases.

Figure 2. Experimental setup of the SBR system. (a) Fill volume; (b) Settled sludge; (c) Magnetic mixers; (d) Peristaltic pump; (e) Air inflow; (f) Effluent tank; (g) Feeding tank; (h) SBR.

Bacterial inoculum was obtained from the aeration tank of the MBR setup for biomass acclimation, which was initiated by allowing 5.2 L of the seed to settle in the reactor for 1 hr, and taking 2.6 L of the supernatant and replacing it with an equal amount of raw leachate. The reactor was then operated under continuous mixing and aeration for 21 days, during which settling, decantation, and replacement with leachate were performed every 2 or 3 days. Several scenarios were tested (Table 3), in a temperature-controlled room at 22 ± 1ºC, with an effective volume of 5.2 L, and an influent/effluent volume of 2.6 L. The first three experiments served as starting trials to evaluate the performance of the SBR at low HRT under variable aeration patterns of the reacting phase, whereas the last two experiments targeted the assessment of the effect of longer HRTs with different aeration patterns as well as the effect of pH adjustment.

Table 3. SBR experimental program

			Experiment I	Experiment II			Experiment III			Experiment IV	Experiment V
Duration of SBR phases (hr)	tC		8	8			8			72	72
	tF		0.75	0.75			0.75			0.75–1.5	0.75–1.5
	tR		6	6			6			68	68
	tS		1	1			1			2–3	2–3
	tD		0.25	0.25			0.25			0.25–0.5	0.25–0.5
	HRT		16	16			16			144	144
	SRT (days)		1.2	1			1			12.9	14.1
Operating conditions	Fill	M	√	√			√			√	√
		A	√	√			√			√	×
	React	M	√	√			√			√	√
		A	√	√	×	√	×	√	×	√	√	×	√
	Other										pH Adjustment to 7.08 and 7.25 at t = 0 and t = 26.5 min using H2SO4

Notes: t_C, Cycle time_; t_F, fill time; t_R, react time; t_S, settle time; t_D, decant time; HRT, hydraulic retention time; M, mixing; A, aeration.

Results and Discussion

Leachate characterization

Table 1 compares the leachate characteristics with national standards for discharge in water bodies (Ministry of Environment [MoE], 2001) as well as with the EPA recommended limits for reclaimed water for irrigation (short term use), and the FAO Recommended maximum concentrations of trace elements in irrigation waters. The limits were exceeded for most parameters except metals, which exhibited low levels, indicating that metal treatment was deemed unnecessary prior to the bioreactor. Low metal concentrations can be attributed to the type of waste received at the landfill, that is, municipal waste with a high fraction of food waste (up to 70%) (Tatsi and Zouboulis, 2002).

Ammonia levels in the landfill leachate were high, ranging from 1800 to 4700 mg L^–1, which could be attributed to the high percentage of food waste deposited in the landfill, yet these levels were comparable to and at times lower than reported concentrations in nearby Mediterranean regions (Clement, 1995; Loizidou et al., 1992; Tsonis, 1998; Tatsi and Zouboulis, 2002). Similarly, high levels of COD (9520–17,760 mg L^–1), BOD₅ (4890–10,935 mg L^–1), and TN (2200–5250 mg L^–1) were recorded, and hence these parameters were selected as the main indicators in the performance assessment of the experimental systems.

SBR experiments

The SBR removal efficiency reached 84.6% for BOD, 46.7% for COD, 71.4% for NH₃, and 72.5% for TN, resulting in considerably low levels of contaminants in the final effluent (Figure 3). Nevertheless, the discharge standards of the World Health Organization (WHO) were not attained, necessitating additional treatment for the residual COD and nitrogen concentrations in particular. The effectiveness of the SBR results becomes more pronounced when put in the context of the literature (Table 4). For instance, the solid retention time (SRT) in this study varied between 12.9 and 14.1 days, which is on the lower range of literature-reported values of 8 to 80 days (Klimiuc and Kulikowska, 2006; Kulikowska et al., 2007), although the quality of the leachate used is at times stronger by severalfold than the leachate used in the literature. Similarly, the corresponding hydraulic retention time (HRT) ranged between 0.66 and 6 days, which is also on the lower range of literature-reported values of 0.19 to 20 days (Neczaj et al., 2005; Klimiuc and Kulikowska, 2006; Kulikowska et al., 2007). While the leachate strength (COD: 9095 mg L^–1; BOD: 3114 mg L^–1; NH₃: 1400 mg L^–1) was on the higher range of those reported in the literature, similar, and at times better, attenuation was achieved, albeit with the relatively lower HRT and SRT. Zouboulis et al. (2001) improved the SBR performance by adding enzymes, whereby removal rates of COD, BOD, and NH₃ increased to 40, 70, and 60%, which were still lower than the efficiencies obtained in this study (47, 82, and 71.4%, respectively). Although they used a notably lower strength leachate (COD: 1396 vs. 9095 mg L^–1, and NH₃: 579 vs. 1400 mg L^–1), Aziz et al. (2011) managed to raise the COD removal from 27 to 50% with the application of the double react–settle scenario. Finally, slightly higher COD removal (≪56.9 vs. 47%) and lower BOD removal (<79.6 vs. 82%) were reported by Guo et al. (2010) while treating young landfill leachate mixed with wastewater corresponding to a significantly lower strength leachate in terms of COD (1625 vs. 9095 mg L^–1 in this study) and BOD (619 vs. 3114 mg L^–1). When the total equivalent substrate removed is used as an indicator, the efficiency of the treatment adopted in this study becomes more pronounced whereby the load of COD, BOD₅, and NH₃ removed are higher than those reported in the literature. For instance, the total equivalent COD, BOD₅, and ammonia removed were 4365.6, 2553.5, and 999.6 mg/L, respectively, while as low as 211.7, 442, and 5 mg/L were reported in previous studies for those parameters (Table 4).

Figure 3. Variation of parameters tested throughout the SBR experimental program.

Table 4. SBR removal efficiencies in comparison with literature reported values

	Operating conditions				Leachate characteristics, mg L^– ^1				Removal, %			Total equiv. removed, mg L^– ^1
Reference	HRT, days	SRT, days	Scale	T, °C	COD	BOD/COD; age	BOD	NH3-N	COD	BOD	NH3-N	COD	BOD	NH3-N
Lin and Chang (2000) ^1			L		<200–350 ^14	0.1 to >5 years			77	—	—	211.75
Zouboulis et al. (2001) ^2	20		L		15,000	0.373	5,600	1,800 ^16	40 75 ^15	70 95 ^15	60 70 ^15	6000 11,250	3920 5320	1080 1260
Uygur and Kargi (2004) ^3	9.33	10	L	25	5750		—	185 ^16	62	—	31	3565	—	57.35
Neczaj et al. (2005) ^4	0.19	10	L		3500	old	—	800 ^16	85	—	70	2975	—	560
Laitinen et al. (2006) ^5	4–8	10–40	L		2200	0.59	1300	210	77	93 (BOD7)	99	1694	1209	207.9
Klimiuc and Kulikowska (2006) ^6	2–12	8–80	L	25	1348	0.383	517	—	≤ 83	98	—	≤1,118	506	—
Kulikowska et al. (2007) ^7	2–12	11–33	B	20–22	1380	0.39	539		76.7–83.1	>98		1,102	>528
Kulikowska et al. (2007) ^8	2–12	30–80	B	20–22	1188	0.379	451		75.7–79.6	>98		922	442
Guo et al. (2010) ^9	5	15	L	25 ± 2	1625 ^17	0.38	619.3 ^17	30 ^17	≪56.9 ^18	≪79.6 ^18	≪16.6 ^18	≪924	≪493	≪5
Aziz et al. (2011) ^10	3.34		L	25	1396			579	–42.89 ^19		89.4			517.6
Aziz et al. (2011) ^11	3.34		L	25	1396			579	27.25 ^20		89.9	380.4		520
Aziz et al. (2011) ^12	3.34		L	25	1396			579	50.34 ^21		85.5	702.7		495
This study^13,^22	6	12.9	B	20-25	9095	0.32–0.75	3114	1400	46.7	81.8	71.4	4365.6	2553.5	999.6

Notes: L, lab; B, bench.

¹12 hr, cycle time. Leachate is then mixed with wastewater at a ratio of 1:3.

²2 hr, mixed fill; 6 hr, aerobic; 1 hr, anoxic; 6 hr, aerobic; 1 hr, anoxic; 6 hr, aerobic; 1 hr, anoxic; 1 hr, settle and draw.

³1 hr, anaerobic; 1 hr, anoxic; 2 hr, oxic; 1 hr, anoxic; 2 hr, oxic.

⁴0.15 hr, aerobic fill; 2 hr, aerobic react; 1 hr anoxic react; 0.30 hr, settle; 0.08 hr, draw.

⁵0.25 hr, fill; 21.5 hr, aerobic react; 2 hr, settle; 0.25 hr, draw.

⁶0.08 hr, fill; 3 hr, mixing; 18 hr, aerobic react; 2.83 hr, settle; 0.08 hr, draw.

⁷3 hr, anoxic; 18 hr, aeration; 2.75 hr, settling; 0.75 hr, discharge.

⁸2 hr, aeration; 2.75 hr, settling; 0.75 hr, discharge.

⁹24 hr, cycle time; 20 hr, aeration time.

¹⁰0.5 hr, filling and mixing; 5.56 hr, react; 1.5 hr, settle; 0.43 hr, draw and idle.

¹¹0.5 hr, filling and mixing; 5.5 hr, react; 1.5 hr, settle; 0.5 hr, draw and idle.

¹²0.5 hr, filling and mixing; 2 hr, react I; 1.5 hr, settle I; 2 hr, react II; 1.5 hr, settle II; 0.5 hr, draw and idle.

¹³0.75–1.5 hr, filling and mixing; 68 hr, aerobic mixed react; 2.25-3 hr, settle; 0.25 hr, decant.

¹⁴Leachate pretreated by coagulation–flocculation and electro-Fenton oxidation.

¹⁵After addition of enzymes.

¹⁶NH₄⁺ –N.

¹⁷Leachate previously subjected to air stripping and Fenton’s oxidation.

¹⁸Results are reported after mixing the leachate with wastewater at a ratio of 1:3.

¹⁹Nonpowdered activated carbon addition 10 mg L^–1.

²⁰Powdered activated carbon addition 10 mg L^–1.

²¹Double-react-settle process.

²²Data are taken from a representative experiment trial showing optimum performance.

MBR experiments

In the MBR experiments, greater than 95% removal efficiencies were attained for all indicators (VSS, BOD₅, TN, NH₃-N), while for COD the removal efficiency reached 70% (Figure 4). The relatively high effluent concentrations of COD can be attributed to the presence of non-biodegradable compounds in the leachate. Operation with no sludge wasting is another factor that can contribute to low COD removal, whereby only 50% COD removal was achieved when treating old landfill leachate using an MBR at infinite SRT (Tsilogeorgis et al., 2008). Overall average removal efficiencies for various indicators (VSS: 95%; BOD₅: 99.5%; COD: 70%; TN: 95%; NH3-N: 96%) were higher or within the range reported in the literature (Table 5) (BOD₅: 85–100%; COD: 38–84%; TN: 43–88%; NH₃-N: 31–75%), although the initial concentrations of leachate used in this study (BOD₅: 4000–6000 mg L^–1; COD: 9000–11,000 mg L^–1; TN: 2000–6000 mg L^–1; NH₃-N: 1800–4000 mg L^–1) were generally higher than literature-reported values (BOD₅: 100–1300 mg L^–1; COD: 235–2456 mg L^–1; TN: 240–1141 mg L^–1; NH₃-N: 780–2180 mg L^–1) with the exception of Visvanathan et al. (2007), who used relatively young leachate (BOD/COD = 0.39–0.65) at similar or stronger characteristics for BOD₅ (4714–9498 mg L^–1) and COD (12,000 m L^–1) but lower for NH₃-N (1000–1700 mg L^–1). Visvanathan et al. (2007) operated at a temperature of 45ºC, which is twofold the temperature used in this study (22ºC), with removal efficiencies similar for BOD₅ (99 vs. 99.5%), lower for NH₃-N (75 vs. 96%) but higher for COD (79 vs. 70%).

Figure 4. Variation of parameters tested throughout the MBR experimental program. S1: Raw leachate, S2: Effluent from anoxic tank, S3: Effluent from aeration tank, and S4: Effluent from MBR.

Table 5. MBR removal efficiencies in comparison with literature reported values

	Operating conditions					Leachate characteristics, mg L^– ^1				Removal, %			Total equiv. removed, mg L^– ^1
Reference	HRT, days	SRT, days	Scale	T, °C	Treatment type	COD	COD/BOD; age	BOD	NH3-N	COD	BOD	NH3-N	COD	BOD	NH3-N
Trebouet et al. (2001)	10	Inf.	P	25	NF	500				74–80			385
Ahn et al. (2002) ^1			F		MBR	1017	0.294	100–500	200–1400	38	97	68.25	386.46	291	577.2
Bodzek et al. (2006)			L		MBR	442	0.656	290		82.4	98.3		364.2	285.1
Laitainen et al. (2006)	3	35–60	L		MBR	2200	0.59	1300		84.5	99 (BOD7)		1859	1287
Visvanathan et al. (2007)	1		L	45	MBR	12,000	0.39–0.65	4714–9498	1000–1700	Up to 79	Up to 99	Up to 75	Up to 9480	Up to 7034.9	Up to 1012.5
Tsilogeorgis et al. (2008)	10	Inf.	L		M-SBR	2456				40–60			1228
Feki et al. (2009) ^2	2–3		L	30	MBR	700	0.92	650		61	100		427	650
Ince et al. (2010)			L	25 ± 2	NF	2070			2180	42–51		37–53	962.5		981
Ince et al. (2010)			L	25 ± 2	MF+PAC	2070			2180	66–70		31
This study ^3	0.83	Inf.	L	22 ± 1	MBR	10,400	0.506	5,270	3000	70	99.5	96	7,280	5243.6	2880

Notes: L, lab; P, pilot; F, full; Inf., infinite; NF, nanofiltration; MF-PAC, microfiltration + powdered activated carbon.

¹Also reported MBR+RO (reverse osmosis) to improve efficiency.

²Also reported MBR + electrochemical oxidation to improve efficiency.

³Data are taken from a representative experiment trial.

MBR vs. SBR

While the SBR technology offers some flexibility in terms of cycle time and sequence, its performance is constrained when considering landfill leachate associated with significant variations in quality and quantity. Combining membrane separation and biodegradation processes or the MBR technology improved removal efficiencies significantly. When comparing the SBR and MBR results in this study, the MBR exhibited significantly better performance in removing BOD, COD, NH₃, and TN, with corresponding improvement in removal efficiencies of 21, 33, 34, and 30%, respectively (Table 6). This could be attributed to the ability of the MBR to operate at higher SRT and MLVSS levels than those of the SBR. Such high levels of biomass in the MBR allow slowly biodegradable substrates to be consumed in the reactor, which could be explained by eq 1 (Bérubé, 2010):

Table 6. Summary of SBR and MBR overall comparative performance

	Influent, mg.L^–1				Effluent, mg.L^–1				Removal, %
Description	COD	BOD	NH3	TN	COD	BOD	NH3	TN	COD	BOD	NH3	TN
SBR (this study)	8665–11,770	2766–3569	900–1100	980–1600	4740–7770	427–1299	400	440–750	19–46.7	63.6–84.6	55.6–71.4	23.5–72.5
MBR (this study)	9000–11,000	4000–6000	1800–4000	2000–6000	3000– 4000	15–30	50–150	70–200	70	99.5	96	95
WHO discharge standards, mg L^–1 ^a					100–150	20	5–15	15–45

Note: ^aStandards set by the World Health Organization (WHO) in a compendium of standards for wastewater reuse in the Eastern Mediterranean Region. The range differs between countries for wastewater reuse or discharge into water bodies.

(1)

where r_S is the rate at which the substrate is consumed (M/L³-T), k is the maximum rate of substrate consumption per unit mass of biomass (1/T), X the concentration of biomass in the bioreactor (M/L³), and k_S the half saturation constant for the substrate being consumed by the biomass (M/L³).

Easily biodegradable compounds have a high maximum rate of substrate consumption (k) and hence can be rapidly consumed. By maintaining a high MLVSS level in the reactor, slowly biodegradable compounds can be equally consumed at a fast rate ( eq 1). In MBR systems, the membrane module preserves much of the biomass within the bioreactor, thus allowing for relatively high biomass concentrations. In contrast, high SRTs, accompanied by high MLVSS levels, cannot be maintained in SBR systems because the latter rely on gravity settling for the separation of biomass, which is less effective than the membrane especially at high MLVSS since the sludge loses its ability to settle into a separate layer. The improved performance of the MBR can be attributed to the membrane separation of the mixed liquor from the treated water whereby the fine pores increase the removal of organics relative to gravity settling in the SBR system.

The presence of the membrane for biomass separation is the main difference between SBR and MBR systems. It provides several benefits, including reduced footprint, higher effluent quality, less sludge production, and ease of retrofitting. When operating at full scale, the elimination of phase separation using gravity settling reduces the burden of continuous monitoring of the sludge quality, which otherwise may vary daily, necessitating more process oversight and on-site adjustment in the SBR operation to maintain good settling characteristics for sustaining high effluent quality. Furthermore, for the same quality effluent, the SBR system will inevitably require a tertiary process when treating high-strength wastewater such as leachate, thus increasing the overall capital investment as well as operation and maintenance costs. When constraints associated with footprint and high-quality effluent are relaxed, the SBR may become economically more effective, which is often the case for low-strength wastewater in rural communities.

Summary

The performance of the SBR and MBR technologies for the management of high-strength landfill leachate was demonstrated at laboratory scale using leachate with COD (9000–11,000 mg L^–1), BOD₅ (4000–6000 mg L^–1), TN (2000–6000 mg L^–1), and NH₃-N (1800–4000 mg L^–1) as indicators. While the MBR exhibited higher average removal rates than the SBR (BOD₅: 99.5 vs. 82%; COD: 70 vs. 47%; TN: 95 vs. 73%; and NH₃-N: 96 vs. 71.4%), post-reatment is invariably required for high-strength leachate, depending on influent levels, discharge standards, or effluent end use. For COD in particular, discharge standards are seldom attained, indicating that depending on the intended use of the leachate, the design of a full-scale treatment facility must account for the non-biodegradable fraction by introducing an additional treatment process to lower the effluent residual COD and attain discharge standards. The comparative assessment assists in technology selection depending on the desired effluent quality in terms of allowable pollution load. The footprint, technical expertise, and membrane lifetime constitute important decision factors to consider when assessing the economic burden of adopting the MBR technology for high strength leachate.

Acknowledgment

Special thanks are extended to the U.S. Agency for International Development for its support of the rehabilitation efforts at the Zahle landfill facility.

Additional information

Notes on contributors

M. El-Fadel

M. El-Fadel is a professor and chairperson, Department of Civil and Environmental Engineering, Faculty of Engineering and Architecture, American University of Beirut. He holds the Dar Al-Handasah (Shair & Partners) Chair in Engineering.

J. Hashisho

J. Hashisho is a graduate student at the Department of Civil and Environmental Engineering, Faculty of Engineering and Architecture, American University of Beirut.