Professional Documents
Culture Documents
Waste Management
journal homepage: www.elsevier.com/locate/wasman
a r t i c l e i n f o a b s t r a c t
Article history: A high content of refractory organic matter, ammonia and toxic compounds is characteristic of landfill
Received 31 October 2011 leachate. Advanced oxidative processes (AOPs) are an attractive alternative for landfill leachate treat-
Accepted 14 August 2012 ment. However, when applied as a unique process treatment, they do not provide a complete solution
Available online 9 October 2012
for the effluent treatment. Combining AOP with a membrane separation process (MSP) presents a number
of benefits and provides an adequate solution for this problem. With this in mind, the present work aims
Keywords: to evaluate, using a bench scale, leachate treatability through AOP by Fenton’s reagent (AOP/Fenton)
Landfill leachate
combined with microfiltration (MF) and nanofiltration (NF). A high efficient removal of COD (63%), true
Fenton’s reagent
Microfiltration
color (76%) and humic substances (50%) was observed during AOP/Fenton under optimized conditions
Nanofiltration (1.7 g H2O2/g CODraw leachate; FeSO47H2O:H2O2 = 1:5.3; pH = 3.8; reaction conditions = 115 rpm/28 min).
According to the evaluated parameters, MSP presented an efficient complementary treatment, in which
the integrity of the stages was sufficient for reaching regulatory levels in the effluent (Deliberação Norma-
tiva Conjunta COPAM/CERH-MG No. 1, May 5, 2008).
Ó 2012 Elsevier Ltd. All rights reserved.
0956-053X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.wasman.2012.08.009
90 W.G. Moravia et al. / Waste Management 33 (2013) 89–101
is applied as an oxidizing agent, however, the addition of activators and characteristics. The sanitary landfill of the CTRS-BR-040 oper-
such as iron, to produce a high rate of required radicals, is neces- ated for 33 years, since it was created in 1975. Received in recent
sary (Li et al., 2010; Bull and Zeff, 1991). Reaction occurs in a sim- years approximately 4200 t/day of waste until its closure in
ple reactor, without the need for more complex equipment and 2008. Currently generates 300 m3/day of leachate and serves as
extreme conditions of temperature and pressure. Another feature transshipment area. The leachate generated is transported and
of the AOP/Fenton is its ease of operation when compared with treated along with domestic wastewater using activated sludge
other AOP techniques, such as processes based on ozone and ultra- system.
violet radiation in which respectively an O3 generator and a UV For this research, the raw leachate was collected from the
source are needed. equalization tank which receives leachate from a 9-year-old
Combining the AOP/Fenton with membrane separation process AC05 cell.
(MSP) provides a range of advantages. The use of microfiltration
(MF) can eliminate the decantation stage used for separation of 2.2. Physicochemical characterization of raw leachate
mud produced during AOP/Fenton, drastically reducing the time
of treatment. The combination of AOP/Fenton – MF with nanofil- Raw leachate characterization lasted for a period of four months
tration also allows for greater efficiency in pollutant removal, since over different seasons (water surplus period – January/February
it offers a complementary treatment enabling the guidelines for 2009; water deficit period – May/June 2009), in which five
legislation standards of effluent to be followed, besides reducing monthly samples were collected composing a total of 20 samples.
costs in area requirements for the implementation of a treatment The following physicochemical parameters were taken into
plant. account: true and apparent color, chemical oxygen demand –
MSP can be understood to be a group of methods or processes COD (method 5220 C), biochemical oxygen demand – BOD (meth-
that involve the use of synthetic membranes intended for separa- od 5210 B), pH, alkalinity (method 2320 B), series solids – total and
tion substances. Such operations can be used for effluent treatment dissolved (method 2540 B and 2540 D), total and ammoniacal
and are based on the selective permeability of one or more compo- nitrogen – NKT and N–NH3 (method 4500 – N B and method
nents through the membrane. 4500 – NH3 B, C), phosphorus total (method 4500 P C), chlorides
Habert et al. (2006) define membranes as a barrier that sepa- (method 4500 Cl C) and metals. Analyses were performed in accor-
rates two phases and restricts, under action of a driving force, total dance with Standard Methods for the Examination of Water and
or partial transportation of one or more chemical species present. Wastewater (APHA, 2005).
Molecules smaller than the membrane pores pass through it, while
those which are greater in size are rejected by the retention effect 2.3. Experimental unit
or by the repulsive force acting on the membrane surface. The fluid
that traverse the membrane is called permeate and that which re- The AOP/Fenton–MF–NF experimental unit consists of the reac-
mains in the other side of the membrane containing the solute is tor (V = 10 L), agitator of blades, feeding reservoir, diaphragm
referred to as the concentrate (Giraldo-Zuñiga et al., 2004; Ordó- pump (Qmáx = 456 L h1), MF hollow-fiber membrane modules
nez, 2005). (manufacturer: PAM Membranas Seletivas) for optimization
Membrane fouling has, in some cases, been a limiting factor for (0.1 m2 by module/0.1–0.4 lm) and operation (1.0 m2 by mod-
MSP application. Over time, such a phenomenon leads to a decline ule/0.1–0.4 lm), NF spiral-wound modules (NF90-2540 manufac-
in permeate flux, causing the flux to become independent of the tured by DOW Water & Process Solutions) for operation (2.6 m2
pressure across the membrane. Appropriate selection of feeding by module/retention of 97% stabilized salts), regulator valves,
characteristics, the hydrodynamic conditions of the system and manometers (permeate and backwashing), rotameters (feeding
cleaning techniques can minimize fouling (Judd, 2006). Several and permeate), thermometer and reservoirs for the collection of
techniques exist to control fouling but what use and how to use permeate and concentrate. The unit also features a permeate back-
will depend on the characteristics of the system and feed that will wash cleaning (system) for MF membranes. Fig. 2.1 shows the dia-
influence the mechanism of fouling. Thus knowledge of fouling gram schematic of AOP/Fenton–MF–NF experimental unit.
mechanisms is necessary to adequately select methods for mitiga- MF and NF main characteristics are presented in Tables 2.1 and
tion of fouling. 2.2, respectively. Table 2.1 – MF Module main characteristics.
Regarding the benefits of combining the AOP/Fenton with MSP
in synergy, AOP/Fenton also serves to minimize MSP fouling, since 2.4. AOP/Fenton optimization and operation
during oxidation reaction, colloidal particle coagulation occurs,
which can reduce the penetration of such particles within the mem- The landfill leachate used in the AOP/Fenton optimization and
brane pores, thus increasing sustainable flux by modifying the operation was collected during water deficit period – May/June
characteristics of the deposited material (Chiu and James, 2006). 2009, for to represent the most critical situation, since the pollu-
Therefore, the objective of this research was to evaluate landfill tants concentration in the leachate in this period is the highest that
leachate treatment in Belo Horizonte/MG through the combination can be found over the year.
of the AOP/Fenton and microfiltration (MF), for removal of mud For each test, the oxidation stage consisted of pH adjustment/
generated by AOP/Fenton, and nanofiltration (NF) as a form of final reaction of the slow addition of sulfuric acid to 1-L of raw leachate,
polishing to bring effluent levels into legislation standards. followed by the addition of solid iron sulfate heptahydrate and a
hydrogen peroxide solution. The system was agitated throughout
the reaction. Aliquots of the sample, pre and post-treatment were
2. Materials and methods collected and analyzed with regard to COD and TS parameters. Jar
Test equipment was used for the tests. The reagents hydrogen
2.1. Sampling peroxide solution (H2O2, 30% v/v – VETEC) and ferrous sulfate
heptahydrate (FeSO47H2O – SYNTH) were also employed for the
The leachate utilized for the experiments was provided by the AOP tests.
landfill of Central de Tratamento de Resíduos Sólidos – CTRS (Treat- The optimization of AOP/Fenton operational conditions empha-
ment Center for Solid Waste) in Belo Horizonte/MG – Brazil. Occu- sized the percentage removal of COD organic matter and mud
pying an area of 133 ha, the landfill contains cells of different ages generation measured in TS. A 25–1 fractional factorial design was
W.G. Moravia et al. / Waste Management 33 (2013) 89–101 91
C
CONCEN NTRA
ATE
DIISPO
OSAL
L
V
VAL
LVE VA
ALV
VE
(CH
HARA
ACT
TERIZ
ZAT
TION
N)
L
LAN
NDFFILL
L RE
ETAIINED
D
AC
CIDIIFICAATIO
ON
L
LEAC
CHAATE
E and
REAAGEENT MEM
MBRRANEE
DOOSAG GE MO
ODUULE PER
RMEA
ATE
(H
H2O2 and F
FeSO
O4) VAL
LVE (M
MF/N
NF)
MA
ANOM
MET
TER
P
R
REACCTOR
(AO
OP)
P
PUM
MP VA
ALV
VE
Table 2.5
Variation levels of 23 central composite rotational design factors in AOP/Fenton
Table 2.2 testing for critical value (maximum) for COD removal.
NF module main characteristics.
Condition H2O2 Fe:H2O2 pH
Characteristics NF Factora Value (g L1) Proportionb Value (g L1)
Nomenclature NF90-2540 +1681 1.5 30 1:1 6.0 5
Selective layer material Polyamide +1 1.3 26 1:2 3.0 4.6
Module length (mm) 1016 0 1.0 20 1:5 1.2 4
Module diameter (mm) 61 1 0.7 14 1:8 0.8 3.4
Selective ability (% stabilized salt retention) 97 1.681 0.5 10 1:10 0.6 3
Filtration effective area (m2) 2.6
a
Multiplying factor 2.125 g for each COD gram of raw leachate
(CODraw leachate = 3.141 mg L1).
b
FeSO47H2O:H2O2 (30% v/v) proportion considering O2 stoichiometric quantities
Table 2.3 necessary for total COD stabilization.
Variation levels of 251 fractional factorial design factors to AOP/Fenton testing.
Table 3.1
Raw leachate characterization.
COD – chemical oxygen demand; BOD – biochemical oxygen demand); TOC – total organic carbon; NKT – nitrogên Kjeldahl total; N–NH3 – ammoniacal nitrogen; TS – totais
solids; FTS – fixed total solids; VTS – volatile total solids; TSS – total suspended solids; FSS – fixed suspended solids Fixos; VSS – volatile suspended solids.
corresponded to a 30 min time span. SVI was determined by the directly of the hydrodynamics of the system, and other factors that
concentration of suspended solids in the sample. SVI values were are also important, the evaluation of the frequency of cleaning in
calculated according to Eq. (1), as established by von Sperling (2001). bench scale is difficult to be transfer to the real scale. Moreover
all tests performed in this work were developed in batch mode
Z 030 106 and an assessment of frequency of cleaning would have to be
SVI ¼ ð1Þ developed continuously.
Z 00 TSS
where SVI is the Sludge Volume Index (mL g1); Z 030 the interface 2.6.2. Investigation of fouling
height 30 min after test onset (m); Z 00 the interface height at test on- The evaluation of fouling of the MF process was carried out
set (m); and TSS is the total suspended solids concentration through the evaluation of the profile of the effluent permeate flux
(mg L1). (Jv) in relation to the pure water permeate flux for the new mem-
brane (Ji). In addition total fouling resistance and the resistance of
2.6. Evaluation of MF in AOP/Fenton generated mud removal each part constituting the total resistance (membrane resistance –
Rm, static adsorption – Ra, pore blockage – Ra and gel formation
2.6.1. Optimization layer – Rg) were evaluated. To calculate these resistances, it was
MF operating conditions were optimized by monitoring the per- necessary to determine the fluxes Ji, Ja, Jf e Jv as follows: the Ji flux
meate flux profile versus time throughout the different evaluated was determined for pure water permeation to the new membrane,
conditions of operation, except for operating pressure, which was such a step being performed to characterize membrane permeabil-
assessed through critical pressure. The operating conditions evalu- ity. Ja flux was determined for pure water permeation after a 2-h
ated were operating pressure and crossflow velocity. To determine static adsorption of the effluent (unpressurized). Jf flux was deter-
critical pressure, tests were performed in which the pressure was mined by pure water permeation in the encrusted membrane and
fixed and the permeate flux was monitored for 18 min, time re- Jv was determined by permeation of the effluent under conditions
quired for compressing the membrane pores and stabilization of to be established based on the determination of the critical pres-
the flow. Pressure was increased every 3 min following the perme- sure for 2 h. Resistance values were calculated using Eqs. (2)–(5),
ate flux. Evaluated pressure values were 0.25, 0.5, 0.75, 1.0, corresponding to the Resistances in Series Model.
1.25 bar (0.025, 0.05, 0.075, 0.1, 0.125 MPa). The effect of flow DP
velocity was evaluated by varying the feed flow rate corresponding
Rm ¼ ð2Þ
lJi
to the variation of hydraulic regimes related to a Reynolds (Re)
number range of 2600, 4400 and 5700 (operational limitation of where Ji is the pure water permeate flux to the new membrane
the experimental unit). Tests were carried out with effluent at (m3 m2 s1); DP the pressure difference applied through the mem-
room temperature. Upon completion of each test, the module brane (Pa); l the fluid viscosity (Pa s); and Rm is the membrane
was subjected to physical and chemical cleaning processes using resistance (m1).
solution of citric acid at pH 3 for 20 min, sodium hypochlorite Ji
(NaClO – 2% solution) for 60 min and water recirculation until Ra ¼ 1 Rm ð3Þ
Ja
complete recovery of the membrane.
The cleaning procedure and frequency of cleaning was not
!
Ji
evaluated in this article, although it is an important aspect to be Rb ¼ 1 Rm Ra ð4Þ
Jf
evaluated and optimized. As the frequency of cleaning will depend
W.G. Moravia et al. / Waste Management 33 (2013) 89–101 93
Table 3.2
Factor condition and response values obtained in AOP tests using 25–1 fractional factorial design.
where H2O2 = H2O2 solution concentration corresponding to the H2O2 fraction in relation to the stoichiometric quantity of O2 required for COD total stabilization; Fe:H2-
O2 = ferrous sulfate heptahydrate concentration, with reference to the FeSO47H2O:H2O2 proportion; pH = hydrogenionic potential; Vm = mixing velocity; T = reaction time.
Fig. 3.1. (a) Main effects and (b) interactions among effects on response efficiency of COD removal during determination of the significant factors.
Ji carried out according to Standard Methods for the Examination of
Rg ¼ 1 Rm Ra Rb ð5Þ
Jv Water and Wastewater (APHA, 2005).
where RT is the total resistance (m1); Rm the transport resistance 2.7. Evaluation of complete treatment cycle
through membrane (m1); Ra the additional resistance due to static
adsorption (m1); Rb the additional resistance due to physical After the optimum conditions were established, MF and NF
blockage of pores (m1); Rg the additional resistance due to gel layer processes were operated under such conditions in order to jointly
formation (cake) (m1); Ji the pure water permeate flux to the new assess efficiency of AOP mud removal by MF and final polishing of
membrane (m3 m2 s1); Ja the pure water permeate flux after the the effluent through the NF process. Samples from each stage of
effluent static adsorption (m3 m2 s1); Jf the pure water permeate treatment were collected and analyzed according to the parame-
flux into fouled membrane (m3 m2 s1); and Jv is the permeate flux ters: COD, BOD, total carbon (TC), inorganic carbon (IC), TOC, pH,
with effluent (m3 m2 s1). true and apparent color, total and ammoniacal nitrogen, alkalinity,
chloride, total phosphorus, series solids (total and dissolved) and
metals.
2.6.3. Evaluation of the removal of mud generated in the AOP/Fenton
The evaluation of MF performance in the removal of mud from 3. Results and discussion
the effluent post-AOP/Fenton at different pH (3, 5 and 7) took into
consideration the hydraulic permeability and selectivity through 3.1. Physicochemical characterization of raw leachate
the effectiveness in color, total iron, TS and COD removal, where
total iron and TS were the prioritized parameters and the main Table 3.1 shows the maximum, minimum and average values
objectives of this stage. Adjustment of pH was performed with and their respective deviations, obtained from the analyzed param-
sodium hydroxide avoiding damage to membranes. Analyses were eters of raw leachate in different seasons. It is noteworthy that the
94 W.G. Moravia et al. / Waste Management 33 (2013) 89–101
Fig. 3.2. (a) Main effects and (b) interactions between effects for response of mud generation in terms of TS during significant factors.
average monthly rainfall was 270 mm in the water surplus period development and stability of community microbiological, since
in contrast to 30 mm during the water deficit period, showing sig- VSS can be directly associated to the existing biomass in the
nificant climatic differences between the monitored periods. leachate.
As expected, it can be observed that the organic matter in terms Metals concentration is low and shows no evidence of toxicity
of COD of the generated leachate is strongly affected by rainfall that can be attributed to these elements. The concentration of iron
patterns, being approximately 66% lower during high rainfall peri- superior to other metals can be explained by the abundant pres-
ods due to dilution. Nevertheless, the high values of the color pre- ence of iron oxides in the soil of the region.
sented confirm the pollution potential of the leachate in relation of
this parameter. The small difference between the real and apparent 3.2. AOP/Fenton optimization and operation
color values means that the color of the effluent is associated with
the dissolved substances in the leachate; a statement that is rein- 3.2.1. Determination of main factors of process
forced by the low solids concentration and significant presence of Table 3.2 shows coded levels of variation, actual values of each
humic substances. factor involved in the process as well as analytical responses
A high concentration of organic matter in terms of COD obtained by the 25–1 fractional factorial design.
throughout the characterization can be observed. The low BOD/ It is interesting to note that all COD tests at this and later stages
COD ratio indicates low leachate biodegradability, confirming the were corrected for the presence of H2O2 residual interference in the
advanced stabilization stage, and also demonstrating the high analysis (obtained results <5.0 mg COD L1). The analyses for resid-
leachate oxidation condition, i.e., the microbiological reactions of ual hydrogen peroxide were performed through the iodometric
BOD degradation are limited. This suggesting that the physical– method (Kolthoff, 1920), which is less susceptible to organic com-
chemical mechanisms are most suitable for the treatment of this pound interference.
effluent.
It is observed that the pH encountered was elevated in compar- 3.2.1.1. Determination of the significant factors involved in COD
ison to similar leachate found in literature with similar character- removal efficiency. Fig. 3.1 shows the estimated main effects of fac-
istics (Gomes et al., 2009), substantiated by the high values of tors involved in the AOP, when change occurs from the lowest level
alkalinity, derived from the large amount of construction
waste (–) to the highest level (+) of each factor, and Pareto chart of
which was disposed, where the carbonate ions CO2 and bicar- Standardized effects for the COD removal efficiency response.
3
bonate ions HCO 3 can be leaching from ceramic materials. According to the Main Effects charts, a marked curvature in the
The BOD/NKT ratio is the quantity of organic matter related to model can be observed, which means, the intermediate values of
the amount of the macronutrient nitrogen, similar to the C/N rela- the chosen limits produced a favorable response. This fact also
tionship. The values for this ratio are very low (1/16) indicating indicates that the fractional factorial design (linear fit) is not rec-
that excess nitrogen is present mainly in the form of ammonia, ommended for optimization of this model.
probably generated by the degradation of protein material consti- Analyzing the effect of H2O2 concentration at extreme limits, its
tuting the major fraction of biodegradable organic matter in leach- increase from the lowest to the highest level reduced the COD re-
ate. As BOD5/COD ratio is very low (0.03 < 0.4), microbial moval efficiency by 8.4%. This reduction can be explained by the
degradation in the landfill aerobic cell may still be limited due to fact that better results of oxidation are obtained for hydrogen per-
lack of easily biodegradable organic matter or the presence of toxic oxide excess in the reaction (>25 mg L1 H2O2 solution) with a fer-
compounds in the effluent. Additionally, the limitation of phospho- ric ion in the media instead of ferrous ions. H2O2 excess also favors
rus concentration is not observed when compared to the ratio indi- self-decomposition reactions of this reagent and sequestration of
cated as favorable for the development of microorganisms hydroxyl radical, generating radicals with lower oxidation power.
(C:N:P = 100:5:2) (Metcalf and Eddy, 2003). Nevertheless, the lack of this reagent may result in failure of OH
It can also be seen that the content of suspended and total solids evolved in the oxidation of organic compounds.
is low, also explaining the small difference between true and The addition of Fe in the form of solid ferrous sulfate heptahy-
apparent color of the effluent (low turbidity). VSS/FSS represents drate improved COD removal efficiency by 14.8% when considering
the ratio between volatile solids amount (organic fraction) and the variation limits of this factor. The increase of Fe concentration
fixed suspended solids (inorganic fraction). Besides the low solids accelerates Fenton reaction, causing H2O2 concentration to rapidly
concentration, a VSS/FSS relationship ranging from 3 to 6 is decrease, minimizing OH sequestration (Villa et al., 2007). How-
observed in high and low rainfall periods respectively, indicating ever, for the values of this factor close to the center points, higher
W.G. Moravia et al. / Waste Management 33 (2013) 89–101 95
Table 3.3
Factor conditions and chosen response values for the applied 22 central composite rotational design in AOP for determination of critical values during the optimization phase of
the AOP auxiliary factors.
Fig. 3.3. (a) Response surface and (b) contours curves for mud generation in terms of TS as a function of temperature and mixing velocity during optimization stage of AOP
auxiliary factors.
Table 3.4
Factor conditions and chosen response values for 23 central composite rotational design applied in AOP tests to determine critical values during optimization phase of AOP main
factors.
where H2O2 = H2O2 solution concentration corresponding to the fraction of the quantity of H2O2 in relation to the quantity of stoichiometric O2 required for COD total
stabilization; Fe:H2O2 = concentration of ferrous sulfate heptahydrate, with reference to the FeSO47H2O:H2O2 proportion; pH = hydrogenionic potential.
96 W.G. Moravia et al. / Waste Management 33 (2013) 89–101
Fig. 3.4. (a) Response surface and (b) contour curves for COD removal efficiency as a function of H2O2 and Fe during optimization phase of AOP/Fenton main factors.
Fig. 3.5. (a) Response surface and (b) contour curves for COD removal efficiency as a function of H2O2 and pH during optimization phase of AOP/Fenton main factors.
Fig. 3.6. (a) Response surface and (b) contour curves for COD removal efficiency as a function of Fe concentration and pH during optimization phase of AOP main factors.
performance was obtained, suggesting an optimal concentration of and by hydrogen peroxide decomposition. It is also noteworthy
this element in the process. that for very acidic pH values (<2), degradation efficiency also
The pH effect on Fenton’s reaction was the most significant decreases in spite of iron species remaining soluble, since high
within the factors studied. Efficiency of COD removal decreased concentrations of H+ can sequester hydroxyl radicals.
by 15.8% when operated at this factor upper level in relation to For an a = 0.05 significance level, Fig. 3.1b shows that for the
its lowest level. The influence of pH on oxidation is a consequence COD removal efficiency, all evaluated factors (except interactions)
of Fe precipitation at pHs above 4, drastically reducing its interac- were significant in the process. Thus, pH factors, Fe and H2O2
tion with hydrogen peroxide and consequently the OH production, concentration, in this order, presented the greater degree of
W.G. Moravia et al. / Waste Management 33 (2013) 89–101 97
Fig. 3.7. Evaluation of AOP/Fenton generated mud sedimentability: (a) sedimentation curve and (b) velocity of sedimentation curve.
Fig. 3.8. (a) Relationship between permeate flux and operating pressure to determine critical pressure in MF proccess; (b) profile of MF permeate flux profile for different flow
velocities during the operation.
significance for the response under study. These factors were con- For an a = 0.05 significance level, it is observed that by the Par-
sidered as the main process, taking into account the principle eto charts, only mixing speed and reaction time were significant in
objective of this treatment (AOP). the process for mud generation in terms of TS. These factors were
considered as process auxiliaries and were later optimized, in order
3.2.1.2. Determination of significant factors involved in mud genera- to determine the critical operating condition (optimal).
tion. Fig. 3.2 shows the estimated main effects of factors involved
in the AOP, in which changes occur from the lowest () to the high- 3.2.2. Optimization of process auxiliary factors
est level (+) of each factor, and also the Pareto chart of effects With the objective of determining the critical values (which en-
according to the standardized response for mud generation in able the minimum generation of mud) of the factors considered to
terms of TS. be auxiliary to AOP, a model was optimized using a 22 central
For the mud generation response in terms of TS, a low effect of composite rotational design. Table 3.3 shows the coded levels of
curvature was observed, which indicates a linear pattern of behav- variation and actual values of each factor involved in the process,
ior regarding the response when varying each factor values be- as well as analytical responses obtained from the 22 CCRD.
tween established maximum and minimum levels. Note also that Fig. 3.3 shows the surface response and contour lines for mud
mixing velocity and reaction time factors have impacted more inci- generation response in terms of TS as a function of reaction time
sively on the process for mud generation response. The other fac- and mixing velocity.
tors interfered less severely in the response process of the study. The identification of the critical point coordinates of the re-
At the fastest mixing velocity (150 rpm), mud production was sponse surface were determined by the resolution of a system with
reduced by 6.2%. This factor mainly affects the AOP on mud two linear equations and two unknowns (Eqs. (6) and (7)), result-
generation and separation, since it influences flake formation and ing from the application of derivative equations obtained for the
characteristics. At very low mixing velocities (slow agitation), flake surfaces.
formation decreases due to a smaller number of collisions between @YðXÞ
solid particles in suspension. With regards to mixing at very high ¼ 0:2455 þ 0:0024X 1 0:0011X 2 ð6Þ
@X 1
velocities (violent agitation), there is formation of flakes followed
by their breaking into smaller particles due to the high kinetic
energy involved. Table 3.5
Prolonged reaction time contributed to a 3.1% lower mud gener- SVI for post-AOP/Fenton generated sludge.
ation. A longer reaction time creates more opportunities for more Sample TSS (mg L1) Z00 (m) Z300 (m) SVI (mL g1)
flakes to break during agitation, leaving released particles more
Post-AOP/Fenton + sludge 1067 0.400 0.137 321
prone to oxidation.
98 W.G. Moravia et al. / Waste Management 33 (2013) 89–101
Fig. 3.9. (a) Ji, Ja, Jf and Jv flux values; (b) relationship between Ja/Ji, Jf/Ji and Jv/Ji; (c) Rm, Ra, Rb, Rg and Rtotal resistance values, (d) resistance fraction in relation to Rtotal in MF
being fed at pH 7, 0.25 bar (0.025 mPa) operating pressure and flow Re equal to 5700.
Critical pressure is defined as that in which the relationship be- high when compared to that found in systems that operate with
tween pressure and permeate flux becomes nonlinear, i.e., from solids concentration similar to this system, reinforcing the viability
which membrane fouling is observed. Fig. 3.8a and b shows the of applying this process.
relationship between permeate flux and operating pressure used Using techniques to minimize fouling, the increased flux can be
to determine critical pressure and the MF permeate flux profile used in the design stage to reduce the demand for membrane area,
when operated at different Re. in the other way, the minimum flux may be considered to design
The critical flux test could have been carried out with constant and increased flux be used to make the process more sustainable
pressure or constant flux. In this study, the choice of setting pres- by providing lower energy requirement, lower cleaning frequency
sure and not the flux is due to the fact that system has more sen- and therefore longer life of the membrane since the frequency of
sibility to measure the flux than the pressure. The loss of flux cleaning directly influences the lifetime of the membrane and
indicates the occurrence of fouling, and then the indication of the increased productivity.
critical pressure or critical flux. According to the results, if compared with Re of 5700, the stabi-
It is observed that the flux declines from the lower values of lized permeate flux of the MF system increased from 16.3 to
applied pressure, thus indicating considerable potential for mem- 32.7 L m2 h1 when operated with Re of 2600 compared to Re of
brane fouling and that is not possible to determine the critical flux. 5700. These results can be explained by greater turbulence and
It is also observed that applied pressure higher than 0.75 bar did shearing next to the membrane surface, resulting in stronger drag,
not produce a higher permeate flux indicating the occurrence of and consequently, lower particle concentration next to the mem-
limit flux. This fouling is most likely due to the characteristics of brane selective layer, thus increasing operational permeability.
feeding and represents the actual process conditions of this appli- The results also demonstrate a permeate recovery coefficient next
cation. Probably due to the fact that colloidal solids, the convective to the flow velocity evaluated. Re equal to 5700 (pressure = 0.25 -
forces toward the membrane surface brought by the pressure gra- bar–0.025 MPa/Qa = 5.2 L min1) was adopted as the system opti-
dient, albeit low, are sufficient to raise the solids concentration mal flow velocity.
next to the membrane selective layer, and thus promote the forma-
tion of cake, possibly due to chemical precipitation of soluble com- 3.4.2. Investigation of fouling
pounds. These results suggest that improved performance of this The evaluation of the MF fouling process was performed through
process can be achieved through the operation of the system at assessment of the total fouling resistance (Rtotal) and the resistance
low pressure. of each part that constitutes Rtotal (membrane resistance – Rm, static
These results suggest that improved performance of this pro- adsorption – Ra, pore blockage – Rb and cake – Rg), by determining
cess can be achieved through the operation of the system at low the pure water permeation flux towards the new membrane (Ji),
pressure and using fouling mitigation strategies. However it is the pure water permeation after static adsorption of effluent (Ja),
noteworthy that this test was realized in the worst situation that pure water permeation in the fouled membrane (Jf) and the effluent
can be found in real scale since it does not use any technique to permeation under the established conditions (Jv).
minimize fouling (e.g. backwash, relaxation and pulse). Even in The operational conditions of the MF process during the evalu-
these conditions, the results show that although there is a great ation of fouling mechanisms were fed at pH 7, a 0.25 bar
loss of initial permeability, the permeability after stabilization is (0.025 MPa) operating pressure and Re equal to 5700. Fig. 3.9a–d
Fig. 3.10. Pollutant concentration in the feed and permeate in MF process during different times of operation and their respective removal efficiencies (a) apparent color; (b)
total iron; (c) total solids and (d) total COD.
100 W.G. Moravia et al. / Waste Management 33 (2013) 89–101
Table 3.6
Specific collective characterization of raw leachate, post-AOP/Fenton effluent and permeate produced by MF and NF.
COD – chemical oxygen demand; BOD – biochemical oxygen demand; TOC – total organic carbon; TKN – Total Kjeldahl nitrogên; N–NH3 – ammoniacal nitrogen; TS – totals
solids; FTS – fixed total solids; VTS – volatile total solids; TSS – total suspended solids; FSS – fixed suspended solids; VSS – volatile suspended solids.
a
Supernatant + sludge – after neutralization.
b
Deliberação Normativa Conjunta COPAM/CERH-MG No. 1, May 5, 2008 (Brazil).
the total resistance, while 20% of the resistance was due to cake
formation. Despite being very low, fouling due to static adsorption
was also observed (3.5% of the membrane total resistance). The
most significant resistance was the membrane resistance repre-
senting more than 50% of the total.
3.5. Evaluation of full cycle treatment of water, characterized as a superior solution for fast, compact
and efficient removal of pollutants when compared to conven-
The NF process was employed as the final polishing stage of the tional systems.
effluent. The manner of NF operation occurred as a single stage and
employed the optimized operational conditions previously estab- Acknowledgements
lished, where operating pressure was 12.5 bar (1.25 MPa) and flow
rate equivalent to a Re of 1800 (Qa = 4.0 L min1). For an integrated The authors gratefully aknowledge the Conselho Nacional de
evaluation of NF performance, samples were characterized in trip- Desenvolvimento Científico e Tecnológico – CNPq (Brazilian National
licate in all stages of treatment. Table 3.6 shows the obtained aver- Council for Scientific and Technological Development) and Fun-
age values and respective deviations of parameters. Fig. 3.11 dação de Amparo a Pesquisa do Estado de Minas Gerais – FAPEMIG
illustrates the appearance of raw leachate samples after each stage (Research and Support Foundation of Minas Gerais) for granting
of treatment. financial resources, and Superintendência de Limpeza Urbana da Pre-
Note that the AOP has significantly contributed to the removal feitura Municipal de Belo Horizonte – SLU/PBH (Cleaning and Landfill
of true color from raw leachate, reaching an average efficiency of Management Department of the Belo Horizonte’s City Council) for
76.4%. The remaining percentage was almost completely removed the supply of leachate and operational support.
by NF (99.4%). In contrast, the MF process removed 91.2% of solids
in the post-AOP/Fenton effluent. This effluent, when considered as References
mud and supernatant, presented a 92%-increase in TS concentra-
tion due to the oxidation reaction, which also played an important APHA, AWWA, WEF, 2005. Standard Methods for the Examination of Water and
Wastewater, 21st ed. American Public Health Association, Washington, DC.
role in conditioning the effluent for the later treatment stage (NF). Bacchin, P. et al., 2006. Critical and sustainable fluxes: theory, experiments and
Regarding the removal of organic pollutants, average removal applications. Journal of Membrane Science 15, 42–69.
efficiencies of 62.5% for COD and 71.0% for TOC respectively were Bull, R.A, Zeff, J.D., 1991. Hydrogen peroxide in advanced oxidation process for
treatment of industrial process and contaminated groundwater. In: Eckenfelder,
obtained in the AOP/Fenton. The MF process had little significance W.W., Bowers, A.R., Roth, J.A. (Eds.), Chemical Oxidation: Technologies for the
for this type of effluent, as expected by the results of the molecular Nineties. Lancaster, Technomic, pp. 26–36.
weight distribution. Nonetheless, the NF process was effective as Chiu, T.Y., James, A.E., 2006. Sustainable flux enhancement in non-circular ceramic
membranes on wastewater using the Fenton process. Journal of Membrane
effluent polishing causing an average removal efficiency of 98.4% Science 279, 347–353.
and 98.1% for COD and TOC, respectively in a complete treatment Christensen, T.H., Kjeldsen, P., 1991. Basic biochemical processes in landfills. In:
cycle. Sanitary Landfilling: Process, Technology and Environmental Impact, Academic
Press.
The membrane separation processes were effective in removing
Cortez et al., 2011. Evaluation of Fenton and ozone-based advanced oxidation
nutrients. The nitrogen, due to a neutral pH (<8.0) presented as vir- processes as mature landfill leachate pre-treatments. Journal of Environmental
tually ionized ammonia NHþ 4 , was significantly removed by the
Management 92, 749–755.
El-Fadel, M. et al., 2002. Temporal variation of leachate quality from pre-sorted and
NF process. This was expected, as the literature already asserts
baled municipal solid waste with high organic and moisture content. Waste
the fact that the NF membrane is effective in rejection ions (Habert Management 22, 269–282.
et al., 2006). Regarding phosphorus, the removal of approximately Environmental Protection Agency (EPA), 1998. Handbook on Advanced
52% during AOP/Fenton, is probably due to its adsorption by the Photochemical Oxidation Processes. EPA/625/R-98/004, Washington, DC.
Giraldo-Zuñiga, A.D. et al., 2004. Tecnologias aplicadas ao processamento do soro de
mud generated in the process, which was retained by MF, explain- queijo. Revista do Instituto de Laticínios Cândido Tostes 59, 340–341.
ing the average removal of phosphorus of 92.3% by MF. Yet, consid- Gomes, L.P., et al., 2009. Estudos de caracterização e tratabilidade de lixiviados de
ering the effect of effluent polishing, the NF stage was able to aterros sanitários para as condições brasileiras. In: Gomes, L.P. (Ed.), Resíduos
Sólidos Urbanos: Aterro Sustentável Para Municípios De Pequeno Porte. Rima
remove almost this nutrient. ABES. Rio de Janeiro, Brasil, 360p.
The alkalinity, probably due to bicarbonates present in the Habert, A.C., et al., 2006. Escola Piloto em Engenharia Química: Processos de
leachate (pH < 8.3), was also significantly removed during the separação com Membranas. Programa de Engenharia Química, COPPE/UFRJ, Rio
de Janeiro, Brasil, 180 p.
AOP/Fenton (94.1%). According to EPA (1998), oxidative processes Judd, S., 2006. The MBR Book Principles and Applications of Membrane Bioreactors
tend to consume alkalinity as carbonate and bicarbonate ions are in Water and Wastewater Treatment. Elsevier, Oxford, England, 325 p.
Kjeldsen, P.I. et al., 2002. Present and long-term composition of MSW landfill
OH sequestering agents. The NF process was enough to remove
leachate: a review. Environmental Science and Technology 32, 297–336.
the remaining AOP alkalinity which had not been previously re-
Kolthoff, I.M., 1920. Chemical. Weekblad, 17–197.
moved by MF. With regards to chloride parameters, similar to Kurniawan, T.A. et al., 2006. Physico-chemical treatments for removal of
the removal of ammonia in the ionized form, the NF process was recalcitrant contaminants from landfill leachate. Journal of Hazardous
Materials 129, 80–100.
effective in retaining these ions, enabling an average removal effi-
Li, W. et al., 2010. International Journal of Chemical Engineering. http://dx.doi.org/
ciency close to 80%. 10.1155/2010/270532.
Machado, C.F., 2004. Avaliação da presença de microrganismos indicadores de
contaminação e patogênicos em líquidos lixiviados do aterro sanitário de Belo
4. Conclusion Horizonte. 2004. 125 p. Dissertação (Mestrado em Saneamento, Meio Ambiente
e Recursos Hídricos). Escola de Engenharia da Universidade Federal de Minas
Regarding the properties of the leachate from the Belo Horizon- Gerais (EE/UFMG), Belo Horizonte, Brasil.
Metcalf, Eddy, 2003. Wastewater Engineering: Treatment, Disposal and Reuse,
te landfill, the results showed that the leachate in the study pre- fourth ed. McGraw-Hill, New York, USA, 1334 p.
sented characteristics of stabilized landfill. The treatment of the Ordónez, J.A., 2005. Tecnología de alimentos: componentes dos alimentos e
leachate using all of the proposed stages was sufficient for reach- processos. Artmed Editora, Porto Alegre, Brasil, 294 p.
US Peroxide, 2010. Introduction to Hydrogen Peroxide. Available at US Peroxide site
ing, with the exception of nitrogen, all parameters according to
<http://h2o2.com> (accessed June 2010).
current legislation (Deliberação Normativa Conjunta COPAM/CERH- Villa, R.D., et al., 2007. Potencial de aplicação do processo foto-fenton/solar como
MG No. 1, May 5, 2008). Despite nitrogen concentration above the pré-tratamento de efluente da indústria de laticínios. Química Nova 30, 1799–
limits, the removal of this pollutant was significant. Thus, the 1803.
von Sperling, M., 2001. Dimensionamento de lodos ativados por batelada utilizando
AOP system in conjunction with MPS proved to be a potential os principios da teoria do fluxo de sólidos. Engenharia Sanitária e Ambiental 6,
alternative for landfill leachate treatment with a focus on reuse 149.