Selecting model for treatmentof oily wastewater by MF-PAC hybrid process using mullite-alumina ceramic membranes
Hermia’s models for cross flow filtration were used to investigate the fouling mechanisms of mullite-alumina ceramic membranes in treatment of oily wastewaters in a hybrid microfiltration-powdered activated carbon process (MF-PAC). Results show that cake filtration model can be applied for predictio...
Збережено в:
| Дата: | 2016 |
|---|---|
| Автори: | , |
| Формат: | Стаття |
| Мова: | English |
| Опубліковано: |
Інститут колоїдної хімії та хімії води ім. А.В. Думанського НАН України
2016
|
| Назва видання: | Химия и технология воды |
| Теми: | |
| Онлайн доступ: | http://dspace.nbuv.gov.ua/handle/123456789/160801 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Цитувати: | Selecting model for treatmentof oily wastewater by MF-PAC hybrid process using mullite-alumina ceramic membranes / Mohsen Abbasi, Aboozar Taheri // Химия и технология воды. — 2016. — Т. 38, № 3. — С. 311-323. — Бібліогр.: 18 назв. — рос. |
Репозитарії
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
irk-123456789-160801 |
|---|---|
| record_format |
dspace |
| spelling |
irk-123456789-1608012019-11-21T01:26:10Z Selecting model for treatmentof oily wastewater by MF-PAC hybrid process using mullite-alumina ceramic membranes Mohsen Abbasi Aboozar Taheri Технология водоподготовки и деминерализация вод Hermia’s models for cross flow filtration were used to investigate the fouling mechanisms of mullite-alumina ceramic membranes in treatment of oily wastewaters in a hybrid microfiltration-powdered activated carbon process (MF-PAC). Results show that cake filtration model can be applied for prediction of permeation flux decline for MF and MF-PAC process up to 400 ppm PAC. The complete pore blocking model and the intermediate pore blocking model can predict permeation flux decline with time for MF-PAC with 800 and 1200 ppm PAC respectively. Average error for prediction of permeation flux with cake filtration model is 2.19% for MF process and 2.16; 2.06 and 1.31% for MF-PAC process with 100; 200 and 400 ppm PAC respectively. Also for MF-PAC process with 800 and 1200 ppm PAC, average error for prediction of permeation flux with complete pore blocking model and intermediate pore blocking model was 6.11 and 6% respectively. 2016 Article Selecting model for treatmentof oily wastewater by MF-PAC hybrid process using mullite-alumina ceramic membranes / Mohsen Abbasi, Aboozar Taheri // Химия и технология воды. — 2016. — Т. 38, № 3. — С. 311-323. — Бібліогр.: 18 назв. — рос. 0204-3556 http://dspace.nbuv.gov.ua/handle/123456789/160801 en Химия и технология воды Інститут колоїдної хімії та хімії води ім. А.В. Думанського НАН України |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| language |
English |
| topic |
Технология водоподготовки и деминерализация вод Технология водоподготовки и деминерализация вод |
| spellingShingle |
Технология водоподготовки и деминерализация вод Технология водоподготовки и деминерализация вод Mohsen Abbasi Aboozar Taheri Selecting model for treatmentof oily wastewater by MF-PAC hybrid process using mullite-alumina ceramic membranes Химия и технология воды |
| description |
Hermia’s models for cross flow filtration were used to investigate the fouling mechanisms of mullite-alumina ceramic membranes in treatment of oily wastewaters in a hybrid microfiltration-powdered activated carbon process (MF-PAC). Results show that cake filtration model can be applied for prediction of permeation flux decline for MF and MF-PAC process up to 400 ppm PAC. The complete pore blocking model and the intermediate pore blocking model can predict permeation flux decline with time for MF-PAC with 800 and 1200 ppm PAC respectively. Average error for prediction of permeation flux with cake filtration model is 2.19% for MF process and 2.16; 2.06 and 1.31% for MF-PAC process with 100; 200 and 400 ppm PAC respectively. Also for MF-PAC process with 800 and 1200 ppm PAC, average error for prediction of permeation flux with complete pore blocking model and intermediate pore blocking model was 6.11 and 6% respectively. |
| format |
Article |
| author |
Mohsen Abbasi Aboozar Taheri |
| author_facet |
Mohsen Abbasi Aboozar Taheri |
| author_sort |
Mohsen Abbasi |
| title |
Selecting model for treatmentof oily wastewater by MF-PAC hybrid process using mullite-alumina ceramic membranes |
| title_short |
Selecting model for treatmentof oily wastewater by MF-PAC hybrid process using mullite-alumina ceramic membranes |
| title_full |
Selecting model for treatmentof oily wastewater by MF-PAC hybrid process using mullite-alumina ceramic membranes |
| title_fullStr |
Selecting model for treatmentof oily wastewater by MF-PAC hybrid process using mullite-alumina ceramic membranes |
| title_full_unstemmed |
Selecting model for treatmentof oily wastewater by MF-PAC hybrid process using mullite-alumina ceramic membranes |
| title_sort |
selecting model for treatmentof oily wastewater by mf-pac hybrid process using mullite-alumina ceramic membranes |
| publisher |
Інститут колоїдної хімії та хімії води ім. А.В. Думанського НАН України |
| publishDate |
2016 |
| topic_facet |
Технология водоподготовки и деминерализация вод |
| url |
http://dspace.nbuv.gov.ua/handle/123456789/160801 |
| citation_txt |
Selecting model for treatmentof oily wastewater by MF-PAC hybrid process using mullite-alumina ceramic membranes / Mohsen Abbasi, Aboozar Taheri // Химия и технология воды. — 2016. — Т. 38, № 3. — С. 311-323. — Бібліогр.: 18 назв. — рос. |
| series |
Химия и технология воды |
| work_keys_str_mv |
AT mohsenabbasi selectingmodelfortreatmentofoilywastewaterbymfpachybridprocessusingmullitealuminaceramicmembranes AT aboozartaheri selectingmodelfortreatmentofoilywastewaterbymfpachybridprocessusingmullitealuminaceramicmembranes |
| first_indexed |
2025-07-14T13:26:07Z |
| last_indexed |
2025-07-14T13:26:07Z |
| _version_ |
1837628987117928448 |
| fulltext |
ISSN 0204–3556. Химия и технология воды, 2016, т.38, №3 311
© Mohsen Abbasi, Aboozar Taheri, 2016
Mohsen Abbasi1, Aboozar Taheri2
SelecTing Model for TreATMenT of oily
wASTewATer by Mf-PAc hybrid ProceSS uSing
MulliTe-AluMinA cerAMic MeMbrAneS
1Department of Chemical Engineering,
School of Cemical and Petroleum Engineering,
Persian Gulf University, Bushehr, Iran,
2Department of Chemistry, Islamic Azad University, Lamerd, Iran
m.abbasi@pgu.ac.ir
Hermia’s models for cross flow filtration were used to investigate the fouling
mechanisms of mullite-alumina ceramic membranes in treatment of oily
wastewaters in a hybrid microfiltration-powdered activated carbon process
(MF-PAC). Results show that cake filtration model can be applied for prediction
of permeation flux decline for MF and MF-PAC process up to 400 ppm PAC.
The complete pore blocking model and the intermediate pore blocking model
can predict permeation flux decline with time for MF-PAC with 800 and 1200
ppm PAC respectively. Average error for prediction of permeation flux with cake
filtration model is 2.19% for MF process and 2.16; 2.06 and 1.31% for MF-PAC
process with 100; 200 and 400 ppm PAC respectively. Also for MF-PAC process
with 800 and 1200 ppm PAC, average error for prediction of permeation flux with
complete pore blocking model and intermediate pore blocking model was 6.11 and
6% respectively.
Keywords: oily wastewater treatment, microfiltration, powdered activated
carbon, mullite-alumina membranes, membrane Fouling.
1. introduction
Oily wastewaters are one of the major pollutants of the aquatic environment
and removing oil from these oil-in-water emulsions is an important aspect
of pollution control. This is due to the emission of a variety of industrial
oily wastewaters from sources such as refineries, petrochemical plants and
transportation [1 – 3].
Low pressure driven membrane separation techniques such as microfilt-
ration (MF) have been considered as indispensable treatment methods in water
ISSN 0204–3556. Химия и технология воды, 2016, т.38, №3312
and wastewater treatment applications to remove specific pollutants which are
not normally removed by conventional processes [4].
Adsorption using powdered activated carbon (PAC) in combination with
membrane MF process can be used as a hybrid system for removing organic
materials and improving permeate flux. PAC is generally used as a pretreatment
step prior to the membrane operation or in combination with the membrane
in feed tank. PAC was used as membrane pretreatment for both water and
wastewater treatment [5 – 9].
Ceramic membranes have been known for years and used in many different
applications and they have numerous advantages: stability at high temperature
and pressure resistance, good chemical stability, high mechanical resistance,
long life and good antifouling properties. Mullite-alumina ceramic membranes
have very high chemical and thermal stability and are very cheap because they
can be prepared by extruding and calcining kaolin clay [1].
One of the major inhibiting factors for successful commercialization of the
membrane processes is membrane fouling. Membrane fouling is characterized
in general as the reduction of permeate flux through the membrane, and
hence leads to an irreversible loss of system productivity over time, caused by
interactions between the membrane and the various components in the process
stream [10 – 13].
In order to enhance economy and efficiency of MF membranes,
understanding the membrane fouling mechanisms is necessary for the further
development.
In the last two decades there have been a large number of studies focused
on effects of operating parameters on flux decline and membrane fouling
mechanisms. In these studies, membrane filtration testes under different
experimental conditions were preformed to obtain data on permeate flux
variation with time [10, 14]. Although some advances in fundamental MF
membrane fouling mechanisms have been achieved, further researches are
needed to better understand the fouling mechanisms. From the analyses of
permeation loss and resistance coefficient of fouling, the filtration flux can
then be predicted by using the blocking models [15 – 17].
The behavior of permeation flux decline with time of ceramic membranes
for treatment of oily wastewater in MF-PAC process has not been demonstrated
in literature. Therefore for knowledge of fouling mechanisms, Hermia’s models
for cross flow filtration [18] were used to investigate the fouling mechanisms
involved in MF-PAC process of oily wastewater at different time intervals
((0 – 2.5 min), (0 – 5 min), (5 – 20 min), (20 – 60 min) and (0 – 60 min))
with mullite-alumina ceramic membranes. The fitted results of the models for
ISSN 0204–3556. Химия и технология воды, 2016, т.38, №3 313
cross flow filtration were presented and compared with the experimental data
in this novel research. Also, more detailed study of the models was provided for
cross flow filtration to explain the fouling mechanisms in MF-PAC of the oily
wastewaters.
2. experimental
2.1. Theory. Permeation flux, flux reduction and total organic carbon
(TOC) rejection are important parameters in design and construction of MF
separation units.
Permeation flux (J) is volume of permeate (V ) collected per unit membrane
area (A) per unit time (t):
. (1)
Flux reduction (FR,%) is calculated as follows [1]:
, (2)
where J
wi
is water flux of clean membranes and J
ww
is water flux of fouled membra-
nes (at the end of filtration) were measure in operating condition with a pressure of
1 bar, temperature of 25°C and cross flow velocity (CFV) of 1 (m · s-1).
TOC rejection (R,%) is calculated as follows [1]:
, (3)
where C
p
represents concentration of a particular component (i.e. TOC) in
permeate, while C
f
is its feed concentration.
2.2. Membranes. In this research, mullite-alumina (50% alumina content)
MF membranes were synthesized from kaolin clay and α-alumina powder.
Commercial grade of α-alumina with 99.6% purity was used to prepare the
mullite-alumina membranes. The powder had an average particle size of 75 μm.
Chemical analysis of the kaolin clay is listed in Table 1. Fig. 1 shows surface and
cross section of a synthetic mullite-alumina ceramic membrane. Preparation and
characterization of membranes has been illustrated in previous research [1].
ISSN 0204–3556. Химия и технология воды, 2016, т.38, №3314
Fig. 1. SEM micrograhs of the mullite-alumina ceramic membrane: a – surface
15000 х, b – cross section 30000 х.
Table 1. Chemical analysis of the kaolin clay
Component Percent Phases Percent
SiO
2
61.62 Kaolinite 64
TiO
2
0.4 – –
Al
2
O
3
24 –25 Illite 2.4
Fe
2
O
3
0.45 – 0.65 – –
K
2
O 0.4 Quartz 27
Na
2
O 0.5 – –
L.O.I 9.5 – 10 Feldspar 6.6
Total 100 – 100
2.3. Process feed. Oil-in-water emulsions (synthetic oily wastewaters) with
1000 ppm oil were prepared by mixing condensate gas from Seraje, Ghom,
Iran, (C
8
– C
12
) and distillated water. Droplet size distribution of the emulsion
(1000 ppm oil in water) is presented in Fig. 2. As observed, mean droplet size
of oil droplets is 1.09 μm. Detail of information has been illustrated in previous
paper [1].
2.4. Setup. Fig. 3, a shows the experimental setup used in all the experiments.
The laboratory scale setup was operated in cross flow mode. Also Fig. 3, b gives
structure of MF membranes module. More information has been presented in
previous paper [1].
ISSN 0204–3556. Химия и технология воды, 2016, т.38, №3 315
Fig. 2. Droplet size distribution of the synthetic oily wastewater with 1000 ppm oil
in water.
Fig. 3. Microfiltration setup (a) and structures of membranes module (b).
2.5. experimental procedure. In order to determine the best operating
conditions, 1000 ppm condensate gas in water emulsions were employed as synthetic
oily wastewaters using mullite-alumina membrane. The effects of different operating
parameters such as pressure (0.5 – 4 bar), cross flow velocity (0 – 2 m · s-1), temperature
ISSN 0204–3556. Химия и технология воды, 2016, т.38, №3316
(15 – 55°C), on permeation flux, FR, and TOC R of mullite-alumina membranes for
treatment of synthetic wastewaters were investigated [1].
Table 2 shows performance of MF and MF-PAC process for treatment of
synthetic wastewaters using mullite-alumina membranes at the best operating
conditions (pressure 3 bar, cross flow velocity 1.5 (m · s-1) and temperature
35°C). Results indicate that addition of PAC in best concentration (400 ppm)
is effective to increase permeation flux and TOC rejection and decreasing
membranes fouling [5].
Table 2. Summary results of MF and MF-PAC system using mullite-alumina
ceramic membranes
PAC concentration,
ppm
Permeation flux,
Lm-2 h-1 FR, % TOC rejection, %
0 118.32 58.5 89.6
100 157.53 42.13 89.8
200 178.1 35.32 89.9
400 190.47 31.22 90.2
800 95.82 64.39 91.9
1200 88.23 65.61 92.4
3. Modeling
Hermia’s models for cross flow filtration are the most useful and applicable
models for microfiltration flux decline prediction. The general equation is as
follows [17, 18]:
, (4)
where n = 2 for complete blocking; n = 1.5 for standard blocking; n = 1 for in
complete pore blocking (intermediate fouling) and n = 0 for cake filtration. K is
a constant and depends on the pressure, the dynamic viscosity of permeate, the
blocked area and the membrane resistance, also J
ss
is steady state permeation
flux. If the models can predict permeation flux decline of membranes, by
linearization of this models and with selection of largest the best coefficient of
determination (R2), slope shows constant of models (K). Therefore with theses
fitting parameters, permeation flux at each time during filtration and fouling
mechanism can be predicted.
ISSN 0204–3556. Химия и технология воды, 2016, т.38, №3 317
3.1. cake formation model. Cake/gel formation usually occurs when
particles/oil droplets larger than the average pore size accumulate on the
membrane surface, forming a "cake/gel". Permeation flux can be predicted as
follows [17, 18]:
;
;
.
J
0
in the initial permeation flux (J = J
0
at t = 0).
3.2. Standard pore blocking model. Standard pore block is the most
dominant phenomenon when retained particles/oil droplets are dimensionally
smaller than the average pore size of the membrane. Permeate flux can be
obtained by the following equation [17, 18]:
(6)
3.3. complete pore blocking model. This process typically occurs when the
particles/oil droplets are dimensionally similar to the mean pore size of the
membrane. In this model, particles/oil droplets plug individual pores.
Permeate flux can be simply represented by the following equation [17,
18]:
(7)
3.4. intermediate pore blocking model. This model assumes each particle/
oil droplet can block some membrane pores or settle on other particles/oil
droplets previously blocked some other pores with superposition of particles/oil
droplets. Permeate flux can be obtained by the following equation [17, 18]:
(8)
For modeling, firstly, the relationship between time (t) and permeate flux
(J) was drawn for all mullite-alumina membranes in MF and MF-PAC process.
(5)
ISSN 0204–3556. Химия и технология воды, 2016, т.38, №3318
In all cases, the permeate volume decreased with time. A linear relationship of
M versus t, 1/J 0.5 versus t, ln[(J – J
ss
)/(J
0
– J
ss
)] versus t and ln[J(J
0
– J
ss
)/J
0
(J –
J
ss
)] versus t was determined experimentally for cake filtration model, standard
pore blocking model, complete pore blocking model and intermediate pore
blocking model to calculate constants (K) in models respectively:
(9)
To determine whether the data agree with any of the considered models,
the coefficient of determination (R2) of each plot for one model was compared
with the others. For better comparison of the models, average prediction errors
of models are calculated. For determination of average prediction errors of
models, by using the experimental data, average value of models constant
(K) are calculated and replaced in equations (2) – (5) to calculate predicted
permeation flux. Therefore average error at different times for predicted flux
and actual flux are determined:
(10)
It is possible that fouling mechanisms has been changed during filtration
and transitions of fouling mechanisms were occurred [16, 17]. Therefore
models were used to investigate the fouling mechanisms of membranes at
different time intervals ((0 – 2.5 min), (0 – 5 min), (5 – 20 min), (20 – 60 min)
and (0 – 60 min)) for MF and MF-PAC process.
4. results and discussions
4.1. Prediction of permeation flux decline by pore blocking models for Mf
process. Tables 3, 4 for all models, indicate that the cake filtration model with
average error of 2.19% coincidence better relative to the intermediate pore
blocking and complete pore blocking models (average error of 3.56 and 7.43%
respectively). Large deviations between experimental and predicted flux decline
are observed for the standard pore blocking model with average error of 14.16%.
Results of Table 3 show that cake filtration model can predict flux of
permeate better than other model at first times of filtration ((0 – 2.5 min) and
ISSN 0204–3556. Химия и технология воды, 2016, т.38, №3 319
(0 – 5 min)). By increasing time to 60 min, results indicate that prediction
of permeation flux with cake filtration model can be applied for prediction of
permeation flux for other intervals. Therefore it can be conclude that pores of
mullite-alumina membranes becomes fill and cake layer formed and it become
thicker by increasing time at the begin of filtration [6, 16]. It must be noted
that by comparing particle size distribution of oil droplet (see Fig. 2) and mean
average pore diameter of mullite-alumina membranes (0.728 μm), it can be
found that mean diameter of oil droplets is larger than average pore diameter
of mullite-alumina membranes and a large percent of oil droplets cannot inter
into mullite-alumina pores. After cake filtration model, intermediate pore
blocking model, can predict filtration flux well.
Table 3. (R2) of models for prediction of permeation flux with time at different time
intervals without PAC and with different PAC concentrations
Models
0 – 2.5 0 – 5 5 – 20 20 – 120 Total time
(0 – 120 min)min
Without PAC
Cake filtration model 0.999 0.999 0.999 0.999 0.999
Intermediate pore
blocking model
0.997 0.997 0.998 0.999 0.992
Standard pore
blocking model
0.995 0.993 0.99 0.983 0.917
Complete pore
blocking model
0.995 0.994 0.994 0.994 0.996
100 ppm
Cake filtration model 0.997 0.998 0.991 0.998 0.998
Intermediate pore
blocking model
0.995 0.996 0.977 0.995 0.988
Standard pore
blocking model
0.993 0.994 0.958 0.974 0.924
Complete pore
blocking model
0.994 0.994 0.965 0.989 0.958
200 ppm
Cake filtration model 0.995 0.998 0.996 0.984 0.994
Intermediate pore
blocking model
0.991 0.996 0.979 0.977 0.988
Standard pore
blocking model
0.993 0.994 0.958 0.974 0.924
Complete pore
blocking model
0.996 0.979 0.999 0.973 0.977
ISSN 0204–3556. Химия и технология воды, 2016, т.38, №3320
Complete pore
blocking model
0.994 0.994 0.965 0.989 0.958
400 ppm
Cake filtration model 0.999 0.988 0.987 0.98 0.991
Intermediate pore
blocking model
0.999 0.977 0.995 0.973 0.98
Standard pore
blocking model
0.998 0.968 0.998 0.934 0.912
Complete pore
blocking model
0.998 0.97 0.997 0.962 0.956
800 ppm
Cake filtration model 0.997 0.995 0.993 0.934 0.904
Intermediate pore
blocking model
0.999 0.999 0.998 0.964 0.979
Standard pore
blocking model
0.999 0.998 0.99 0.998 0.982
Complete pore
blocking model
0.999 0.998 0.989 0.99 0.988
1200 ppm
Cake filtration model 0.906 0.962 0.983 0.97 0.977
Intermediate pore
blocking model
0.997 0.983 0.998 0.981 0.99
Standard pore
blocking model
0.996 0.977 0.998 0.941 0.936
Complete pore
blocking model
0.996 0.979 0.999 0.973 0.977
Table 4. Average error of models for prediction of permeation flux and constant of
models (K) at total time (0 – 60 min) for MF process without PAC and with different
PAC concentrations
Models
K
Average error for prediction
of permeation flux, %
Without PAC
Cake filtration model 1.84 · 10-6 2.19
Intermediate pore blocking model 1.98 · 10-4 3.56
Standard pore blocking model 5.17 · 10-4 14.16
Complete pore blocking model 1.78 · 10-4 7.43
100 ppm
Cake filtration model 9.31 · 10-7 2.16
Intermediate pore blocking model 1.34 · 10-4 2.72
Table 3. (Cont.)
ISSN 0204–3556. Химия и технология воды, 2016, т.38, №3 321
Standard pore blocking model 4.14 · 10-4 10.08
Complete pore blocking model 8.67 · 10-5 6.97
200 ppm
Cake filtration model 8.10 · 10-7 2.06
Intermediate pore blocking model 1.32 · 10-4 2. 03
Standard pore blocking model 3.65 · 10-4 8.81
Complete pore blocking model 8.98 · 10-5 5.18
400 ppm
Cake filtration model 7.19 · 10-7 1.31
Intermediate pore blocking model 1.23 · 10-4 4.35
Standard pore blocking model 3.47 · 10-4 12.95
Complete pore blocking model 8.13 · 10-5 8.35
800 ppm
Cake filtration model 2.57 · 10-6 15.87
Intermediate pore blocking model 2.40 · 10-4 28.37
Standard pore blocking model 8.49 · 10-4 8.00
Complete pore blocking model 1.06 · 10-4 6.11
1200 ppm
Cake filtration model 3.02 · 10-6 15.78
Intermediate pore blocking model 2.58 · 10-4 6.00
Standard pore blocking model 9.28 · 10-4 12.35
Complete pore blocking model 1.04 · 10-4 11.32
4.2. Prediction of permeation flux decline by pore blocking models for Mf-PAc
process. Tables 3, 4 shows (R2) of models, average of predicted permeation flux
and constant of models (K) at different time intervals. The results show that the
fitting of models with experimental data for MF-PAC hybrid process is as good
as MF process. Largest deviations between experimental and predicted flux
decline were observed for the standard pore blocking model up to 400 ppm PAC
concentration and intermediate pore blocking model and cake filtration models
for 800 and 1200 ppm PAC concentration respectively. As shown in Table 3 for
MF-PAC process with 100 ppm PAC, cake filtration model with average error
of 2.16% is best model for prediction of flux decline. By employing models for
different time intervals, permeation flux can predicted with cake filtration model
for all time intervals similar to MF process (see Table 3). Also for concentration of
200 and 400 ppm PAC, cake filtration model is the best model at total time interval
with average error of 2.06 and 1.31% respectively. After this model, intermediate
pore blocking model, can predict flux decline well. For high dosage of PAC (800
and 1200 ppm), complete pore blocking model and intermediate pore blocking
Table 4. (Cont.)
ISSN 0204–3556. Химия и технология воды, 2016, т.38, №3322
model with average error of 6.11 and 6% is best model respectively. The reason for
this phenomenon is that PAC particles adsorb some of the oil droplets and also
detach the layer formed by the oil droplets at begins of filtration [6]. Of course at
the beginning of filtration, membrane surface is clean and pores of membrane are
empty, therefore small oil droplets enters into membranes pore and complete the
pores [17]. With 800 and 1200 ppm PAC in experiments, wastewater and fouled
membranes were become dark. This is due to filling of membranes pore with PAC
particles. Also high dosage of PAC can well reduce fouling layer of oil and adsorb
oil droplets but PAC particles cover membranes surface and fill membrane pores.
Results of Table 3 indicate that for all time intervals with 200 ppm PAC, cake
filtration models is best model for all time intervals because R2 of it is larger than
other models. But by increasing PAC concentration to 400 ppm, for first time
intervals (0 – 2.5 min) and (0 – 5 min), cake filtration models has largest (R2)
(see Table 3) but for (5 – 20 min) complete pore blocking model is best model.
Results in Table 3 indicate that in for different time intervals (0 – 2.5 min), (0 –
5 min), (5 – 20 min) intermediate pore blocking models can predict permeation
flux decline with 800 ppm PAC. In addition, for (20 – 60 min), standard pore
blocking models has largest (R2). According to results of Table 3, for first time
intervals (0 – 2.5 min) and (0 – 5 min), complete pore blocking model has largest
(R2) but for (5 – 20 min), standard pore blocking model is best model.
5. conclusions
In this novel research, mechanisms of flux decline for treatment of oily
wastewaters in MF-PAC hybrid process using homemade mullite-alumina
ceramic membranes have been investigated. For this purpose Hermia’s models
for cross flow filtration were used in different time intervals with different PAC
concentration. The coefficient of determination (R2) of each case and average
error of models for prediction of permeation flux, were compared between the
fouling models. According to the obtained results, it can be concluded that the
best fit to experimental data is for the cake layer formation model for MF and
MF-PAC process with PAC concentration up to 400 ppm with maximum and
minimum average error of 2.19 and 1.31%. But for MF-PAC hybrid process
with 800 and 1200 ppm PAC concentration, complete pore blocking model
intermediate pore blocking model with average error equal to 6.11 and 6%
are best models for prediction of flux decline. Average error for prediction
of permeation flux with cake filtration model is 2.19% for MF process and
2.16; 2.06 and 1.31% for MF-PAC process with 100; 200 and 400 ppm PAC
ISSN 0204–3556. Химия и технология воды, 2016, т.38, №3 323
concentration respectively. Results of modeling show that pore blocking
behavior of membrane during filtration is changed. Finally it can be concluded
that modeling results for total time is practical and result of short time intervals
is useful for knowledge of fouling mechanisms.
references
[1] Abbasi M., Mirfendereski M., Nikbakht M., Golshenas M., Mohammadi T.
// Desalination. – 2010. – 259. – P. 169 – 178.
[2] Otadi N., Hassani A.H., Javid A.H., Khiabani F. F. // J. Water. Chem. and Technol. –
2010. – 32, N 6. – P. 370 – 377.
[3] Mamchenko A.V., Gerasimenko N.G., Deshko I.I., Pakhar T.A. // Ibid. – 2010. –
32, N 3. – P. 167 – 175.
[4.] Goncharuk V.V., Kucheruk D.D., Balakina M.N., Dul’neva T.Yu. // Ibid. – 2009. –
31, N 6. – P. 396 – 404.
[5] Abbasi M., Sebzari M.R., Mohammadi T. // Ibid. – 2011. – 34. – P. 1252 – 1258.
[6] Abbasi M., Sebzari M.R., Salahi A., Abbasi S., Mohammadi T. // Desalin. Water.
Treat. – 2011. – 28. – P. 1 – 7.
[7] Lesage N., Sperandio M., Cabassud C. // Chem. Eng. Proc. – 2008. – 47. – P. 303 – 307.
[8] Satyawali Y., Balakrishnan M. // J. Hazard. Materials. – 2009. – 170. – P. 457 –
465.
[9] Konieczny K., Klomfas G. // Desalination. – 2002. – 147. – P. 109 – 116.
[10] Remy M., Marel P., Zwijnenburg A., Rulkens W., Temmink H. // Water Res. –
2009. – 43. – P. 345 – 350.
[11] Salahi A., Abbasi M., Mohammadi T. // Desalination. – 2010. – 251. – P. 153 – 160.
[12] Kosvintsev S., Holdich R.G., Icumming W., Starov V.M. // J. Membr. Sci. – 2002. –
208. – P. 181 – 192.
[13] Abbasi M., Sebzari M.R., Salahi A., Mirza B. // Chem. Eng. Comm. – 2012. –
199. – P. 78 – 93.
[14] Hermia J. // Trans. Inst. Chem. Eng. – 1982. – 60. – P. 183 – 187.
[15] Vela M.C.V., Blanco S.A., Garcia J.L., Rodriguez E.B. // Desalination. – 2008. –
222. – P. 451 – 456.
[16] Vela M.C.V., Blanco S.A., Garcia J.L. // Separ. Purif. Technol. – 2008. – 62. –
P. 489 – 498.
[17] Vela M.C.V., Blanco S.A., Garcia J.L., Rodriguez E.B. // Chem. Eng. J. – 2008. –
149. – P. 232 – 241.
[18] Field R.W., Wu D., Howell J.A., Gupta B.B. // J. Membr. Sci. – 1995. – 100. –
P. 259 – 272.
Received 11.03.2013
|