Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

Ultrafiltration as a feasible technology to drinking water production: three case studies in Brazil

Fabio Pereira de Carvalho, Renato Ramos, Andre Lima & Andre Belarmino Sousa
Dow Water & Process Solutions

Abstract

Recent advances in membrane technology and increasing requirements on water quality have stimulated
ultrafiltration (UF) for drinking water treatment. Countries like Brazil, where conventional treatment still is
first option, are experiencing some problems due quality problems and operational costs. This paper
introduces UF, with a brief description of technology and important concepts to be learned. To demonstrate
feasibility of membrane application, three pilot studies were performed in different types of surface water in
São Paulo, Brazil. Superior water quality, lower operational costs and compact automated system are some
keywords that can be used to increase interest and potential of application of UF in municipal water
treatment market.

Keywords

Ultrafiltration, surface water, water quality, pilot trials, operational parameters

1. Introduction UF technology is expected to spread worldwide


Ultrafiltration (UF) technology is an attractive (Figure 1), especially in large-scale water
alternative to conventional treatment for drinking purification plants (8)
water production, removing numerous
contaminants (virus, Giardia, Cryptosporidium,
coliforms, suspended solids) and will reduce
turbidity from different water sources, including
surface waters. Membrane technology is
becoming popular for water applications, based
on advances in membrane technology and
increasing water quality requirements and
regulatory compliance (1, 2, 3); in Brazil, for
example, a recent regulation (“Regulation
2.914/11”) requires public water supply systems
to have a superior quality water. The use of Figure 1. Ultrafiltration and Microfiltration Global Capacity
surface water as water supply requires efficient (m³/d) 1995-2012
processes in order to fulfill these requirements
(4). Ultrafiltration membranes may be applied to
Decades ago in 1990’s, UF was not cost variable feed water sources to produce
competitive compared with traditional disinfected and superior quality water, sometimes
technologies (coagulation, flocculation, without the use of pretreatment, generating
decantation and filtration) but in recent years this substantial operational costs reduction and being
situation has changed: lower costs of membranes competitive with most conventional alternatives
(representing 15-20% of plant cost), (3, 11, 17). Critical points to be highlighted for
standardization of the membrane, lower these applications are:
manufacturing costs, higher production volumes  Capital expenditure (CAPEX) – with rapid
and optimization of the process. UF low-pressure development of new technologies, capital
membranes are sufficiently competitive with and operational costs of UF membrane
conventional processes and are becoming technology are still expected to decrease
gradually accepted compared to other (9). In addition, the implementation of
technologies (5, 6, 7). In this century, the trend of innovative cleaning strategies, like air

DOW RESTRICTED
backwash system, and low operation  Inorganic – ceramic (ex.: alumina,
pressures contributes to reduce costs glass, zirconia)
(11). Considering these types of materials and
 Operational expenditure (OPEX) – UF different permeability characteristics, membrane
plants are automated and have low providers use some traditional membrane
operational labor requirements (10), formation techniques (22, 23, 24):
reduced energy consumption and  DIPS (Diffusion Induced Phase
chemicals dosages and solids residual Separation) – this process involves the
streams are generally considered less conversion of a liquid polymer
problematic. The technology is highly solution (based in two or more
flexible and can be used coupled with components) into a two-phase
other treatment processes to achieve system, using a precipitation fluid.
additional removal (11). During the phase separation process,
a solid phase (high content of
2. Technology description polymer) forms the membrane
In membrane filtration, solid particles to be structure, and a liquid phase (low
separated are usually larger than pore size content of polymer) forms the
characteristic of the membrane, with irregular membrane pores. Using this
shapes. The separation depends of different technique it is possible to produce
factors as size, charge or affinity or a combination asymmetric membranes with a graded
of these properties (12). In ultrafiltration, mass pore size, as well as completely
transport is controlled by the convective flux symmetric membranes.
through pores and the driving force is a pressure  TIPS (Thermally Induced Phase
gradient (Figure 2) between feed solution and Separation) – is an alternative
filtrate (13). technique used for producing porous
structures, driven by temperature
change. In this technique a polymer is
dissolved in a poor solvent at high
temperature. Phase separation occurs
upon cooling the homogeneous
solutions, resulting in a porous
material after removal of the solvent.
It can be applied to a wide range of
polymers, generating dense and
porous films, with different types of
Figure 2. Schematic illustration of principle of ultrafiltration microstructures.
These membranes are commonly
The technology is used to remove particulate packaged in different types of modules; some
and microbial contaminants, but not ions and important properties include: mechanical
small molecules. UF membranes are generally robustness, sealing, flow distribution, cleanability,
characterized by pore size (nominal diameter, chemical resistance, low cost, scalability and high
expressed in microns – generally 0.01–0.05 µm) product recovery. The most common modules are
and molecular weight cut-off (MWCO – molecular (14):
weight of solute for >90% of rejection, expressed  cassettes,
in Daltons) (10, 11). Membranes materials  spiral wound,
primarily consist in two types (14):  hollow fibers,
 Organic - polymeric structures (ex.:  tubes,
polyethersulfone, cellulose  flat sheets
derivatives, polysulfone, polyamide, Other important modes of ultrafiltration
polyacrylonitrile, fluoropolymers); process are:

DOW RESTRICTED
 Flow direction: in inside-out mode the effective method to maintain TMP low (4,
feed water flows inside the fibers and the 10);
permeate flows outside. In outside-in  The membrane nominal pore size
mode occurs the opposite, feed water (0.03μm) provides high filtration
flows outside the fibers, while permeate performance and high flux, also stabile
flows inside (16); long term filtration performance
 Flow configuration: in dead-end filtration compared to microfiltration; hollow fiber
all the solution is forced through the has large membrane surface per module
membrane, retaining particles on surface. volume, giving higher performances (10,
In cross-flow membrane filtration, the 20);
feed liquid flows tangentially to the  Pressurized vertical shell-and-tube design
membrane surface. (12) allows the possibility of using an air flush
Three studies (application of UF for producing cleaning, facilitates integrity tests and
drinking water for different surface water provides protection for the membranes.
sources) were carried with an automated pilot Also, it operates at higher fluxes than the
plant using Dow Ultrafiltration module submerged systems (22).
(pressurized vertical) (Table 1) (10). In order to demonstrate the viability of the UF
technology for the production drinking water, a
Table 1. UF module information
Item UF module
pilot plant was installed in three different
Fiber type Hollow fiber locations in São Paulo, Brazil and different surface
Fiber material Hydrophilic PVDF water sources. The study evaluated different
Flow configuration Dead-end operational parameters of UF treatment process,
Filtration configuration Outside-in
Pore size (µm) 0.03
aiming to achieve regulatory goals for drinking
Temperature (°C) 1 - 40 water and optimal operational conditions to each
Max. operation pressure (bar) 6.25 type of feed water.
2
Filtrate flux (25°C) (L/m /hr) 40 - 120
Max. Transmembrane Pressure (bar) 2.1 3. Sites description
pH range for continuous operation 2–11
In this study, three different feed water
(surface water) were evaluated in Brazilian water
Pilot plant configuration was defined, since it
treatment plants (WTP):
specifications can provide many advantages in
 Alto Cotia (28 days of operation): part of
application of UF technology in surface water
Alto Cotia System that supplies the
treatment:
Metropolitan Region of São Paulo using
 Use of polymeric hydrophilic membrane feed water from Graça Reservoir.
PVDF (Polyvinylidene fluoride) hollow
Although historically considered to be one
fibers increase mechanical strength,
of the best sources of water, the recent
chemical resistance (different acids, bases trend is degradation of raw water due to
solvents and oxidants) and longevity human action, mainly color problems.
(generally 5-10 years for water Capacity: 1.2 m³/s; Type of treatment:
applications). Also, surface modification pre-chlorination, coagulation,
provides fouling reduction, easy cleaning, flocculation, sedimentation, filtration,
wettability and increasing flux (10, 12, disinfection and fluoridation (19, 25).
15);
 Alto da Boa Vista (58 days of operation):
 The use of dead-end mode provides is part of the System
higher system recovery (typical recoveries
Guarapiranga/Billings also supplying the
of 90-97%), energy consumption
Metropolitan Region of São Paulo. The
reduction compared with cross-flow water quality of the source (Guarapiranga
mode (10, 18);
Reservoir) is compromised by diffuse
 Use of outside-in configuration is tolerant pollution (human occupation). Problems
of wide ranging feed water qualities and with algal blooms, caused by
solids. Allows air scour cleaning, providing eutrophication, create physicochemical
less consumption of chemicals and an

DOW RESTRICTED
and organoleptic problems in water.  Alto da Boa Vista
Capacity: 14 m³/s; Type of treatment:
Grating, sandbox, pre-chlorination,
coagulation, flocculation, sedimentation,
filtration, disinfection and fluoridation
(19, 26).
 Indaiá (49 days of operation): supplies
part of Bertioga City, São Paulo coast. The
source of water is based on Itapanhaú
River and WTP problems mainly are
related with inefficient control during
color peaks. Capacity: 130 L/s; Type of
treatment: grating, sandbox, pre- Figure 4. Alto da Boa Vista – Feed and filtrate water turbidity
chlorination, coagulation, flocculation, values during tests
sedimentation, filtration, disinfection, pH
correction and fluoridation (27). Again, 99.2% of all samples had values
Table 2 shows several parameters to assess below 0.05 NTU and 0.8% had values between
the feed water quality: 0.05 and 0.1 NTU. Results variations in outlet
turbidity during tests are related with technical
Table 2. Feed water common characteristics for three sites (19)
problems in turbidity sensor (water flow back).
Parameter Alto Alto da Indaiá
Cotia Boa
Vista  Indaiá
Turbidity (NTU) 6.7 5.3 3.8
Apparent color (UC) 49 32.4 256
pH 6 6.9 6.5-6.8
Total suspended solids - <5 20
(mg/L)

4. Results and discussion


4.1. Removal objectives
4.1.1. Turbidity
For three sites, results of turbidity of feed
and filtrate water during tests are plotted below:
 Alto Cotia
Figure 5. Indaiá – Feed and filtrate water turbidity values
during tests

In this case, UF reduced turbidity to less


than 0.1 NTU in all of samples during tests.
These results are in compliance with
standard limit (0.5 NTU in 95% of samples).
Actually, it’s common that conventional treatment
have difficult to reach this goal, which must be
achieved in four years after Regulation publication
(until 2015). In other words, ultrafiltration
Figure 3. Alto Cotia – Feed and filtrate water turbidity values demonstrates feasibility to be applied in different
during tests types of surface waters without problems to reach
turbidity goals with stable operation.
UF reduced turbidity to less than 0.05 NTU
at 99.2% of samples and 0.8% generated 4.1.2. Apparent color
ultrafiltered water with values between 0.05 and  Alto Cotia
0.1 NTU.

DOW RESTRICTED
In the beginning of tests, without addition water and permeate are performed to evaluate
of coagulant, results of apparent color of feed algae removal by UF:
water in six different samples were inside the  Alto Cotia
range of 49-57 uC. After UF treatment color has
reduced to a range of 10-19 uC (Regulation Table 3. Alto Cotia - Hydrobiological analysis results
Sampling 1 Sampling 2
2.914/11 limit – 15 uC). To reduce color until
Content Content Content
acceptable values (target 5 uC), different Content in
in Raw in Raw in
Microorganism permeate
concentrations of coagulant were added in feed water
(ASU/ml)
water permeate
(ASU*/ml) (ASU/ml) (ASU/ml)
water with starting concentration of 25 ppm of
Cyanobacteria 1.53 - 1 -
aluminum sulfate (dosage of local conventional Chlorophyta 307.08 - 210.2 0.2
treatment). After sequential reductions in dosage Bacillariophyta 179.31 - 284.6 0.1
and looking for combined solution between low Others algae 0.31 0.05 18.4 0.1
Others
TMP and low color, 5 ppm of Al2(SO4)3.18H2O microorganisms
14.09 - 10.6 1.2
achieved color values around 5 uC; in six samples Total 502.32 0.05 524.8 1.6
of feed water with results between 40-55 uC, the Removal 99.99% 99.70%
ultrafiltered water achieve values between 2-5 uC, *Algal Areal Standard Units (ASU)

consistent with the regulatory goals.


Predominant organisms: Staurastrum
 Indaiá (Cyanobacteria); Cosmarium, Closteriopsis and
Using combined alternatives with Chlorococcales (Chlorophyta); Melosira and
coagulant and hypochlorite, color removal were Pennales (Bacillariophyta); Dinoflagellate, rotifers,
evaluated: ciliate and filamentous bacteria (others
o in three samples with only dosage of 5 microorganisms).
ppm of FeCl3, color values were
reduced from 22.5 uC to 2.5 uC;
 Alto da Boa Vista:
o in two samples with dosage of 5 ppm
of FeCl3 plus hypochlorite (2 ppm of Table 4. Alto da Boa Vista - Hydrobiological analysis results
residual chlorine), color values were Hydrobiological analysis - Total of microorganisms (ASU/mL)
reduced from 15 uC and 22.5 uC to 2
Date Raw Water Permeate % Removal
uC and 1 uC, respectively;
o in one sample with addition of
11/30/2012 3301.5 0.22 99.99%
hypochlorite, color value were
reduced from 15 uC to 1.5 uC 12/06/2012 1758.5 0.20 99.99%

12/13/2012 2398.9 0.10 100.00%


4.1.3. Algae
12/17/2012 796.6 0.08 99.99%
In drinking water treatment, one of most
common problem to be solved is algae bloom. The 01/07/2013 1342.1 0.02 100.00%

removal of algae content is extremely important, 01/10/2013 2435.8 0.06 100.00%


mainly for harmful cyanobacterias. However it has
01/16/2013 2401.3 0.20 99.99%
been proved that coagulation, flocculation and
filtration (conventional treatment) are partially 01/31/2013 662.4 0.00 100.00%
effective for algae removal, but generating stress 02/14/2013 2762.8 0.00 100.00%
and dissolved toxins. Pre-oxidation has been
02/20/2013 3217.5 0.00 100.00%
shown to be effective, but high dosage required
may induce cell lysis, releasing undesirable Predominant organisms: Cyanobacteria
compounds. Moreover, there is a risk for Chlorophyta and Bacillariophyta.
formation of harmful by-products. Membrane These results demonstrate ultrafiltration
filtration technology is very effective and safe for capacity in algae and microorganism removal
the removal of algae after bloom, since it offers (>99.7%) at those sites, providing safe and
an absolute removal of the algal cells and avoids superior water quality and reducing problems
releasing of harmful toxin by cell lysis (28). Due to related with presence of these organism in
this problem, hydrobiological analysis of raw drinking water.

DOW RESTRICTED
as chlorine; because of their toxicity, they are
4.1.4. Aluminum and iron regulated by drinking water regulations (21).
One of the problems of conventional Based on Brazilian legislation, the limit value is
treatment, the excess of coagulant, can contribute 0.08 mg/L and the sample result was adequate,
to increasing the value of metals (generally iron or with a concentration below 0.01 mg/L.
aluminum) content in the treated water. Also, iron Total organic carbon (TOC) removal also is
and aluminum concentration in feed water can an important parameter of UF performance.
contribute to exceed regulatory limits of these According to US EPA Stage 1
metals. Due to this problem, studying the content Disinfectants/Disinfection Byproducts Rule is
of metals in the treated water was carried out. important to monitor and reduces NOM, and for
 Alto Cotia measuring NOM, total organic carbon (TOC)
With an optimal coagulant (aluminum analysis was chosen (29). So, it’s extremely
sulfate) concentration of 5 ppm, all samples of important analyzing TOC as a parameter of
ultrafiltered water had values below 0.2 mg/L content of DBPs precursors. In Alto da Boa vista
(standard limit based on Regulatory 2914/11). pilot study, TOC content of raw water and
 Alto da Boa Vista permeate were evaluated: reduction of 5.7 mg/L
With 5 ppm of coagulant (ferric chloride), in raw water to 1.0 mg/L in permeate,
a study of presence of total iron in filtered water demonstrating a considerable reduction of TOC
was performed. 64 samples were analyzed and content and reduction of precursors of DBPs,
none exceeded the regulatory limit (0.3 mg/L) and providing a superior water quality, even though
with a mean value of 0.007 mg/L. most of the organic matter was soluble.
 Indaiá
Removal analysis of Fe: this part of the 4.2. Operational efficiency objectives
study was divided into three phases. In first 4.2.1. Optimal conditions and system recovery
phase, the study was conducted by adding 5 ppm Recovery is the volume fraction of feed
of ferric chloride in raw water (samples 1 and 2). water that is converted to permeate a given time
During the second phase (samples 4 and 5), study interval. It should be deducted from the volume
was performed with hypochlorite (guaranteeing of permeated water used for chemical cleanings
residual chlorine of 2 ppm) and 5 ppm of ferric and backwashes. Systems recoveries were
chloride. The third phase (sample 6) was made calculated, after several attempts it was possible
with addition of the hypochlorite, (residual to find optimal parameters for operation in three
chlorine of 2 ppm) and no coagulant. Results different sites:
below (Table 5):  Alto Cotia
Table 5. Iron removal study results
Iron in Table 6. Alto Cotia – Optimal operational conditions
Iron in feed Parameter Optimal conditions
Chemicals dosage ultrafiltered
water
water Backwash frequency 20 minutes
Sample 1 0.32 mg/L <0.05 mg/L Backwash duration 60 seconds
5 ppm ferric chloride 2
Sample 2 0.42 mg/L <0.05 mg/L Operational Flux 50 L/m .h
2
5 ppm ferric chloride Backwash flux 120 L/m .h
+ hypochlorite Sample 3 0.39 mg/L <0.05 mg/L Air Duration 60 seconds
(residual chlorine 2 Sample 4 0.38 mg/L <0.05 mg/L cleaning Output
3
12 Nm /h
ppm) CEB Frequency Every 6 hours
Only hypochlorite Duration Reagents injection: 90 sec.
(residual chlorine 2 Sample 5 0.37 mg/L <0.05 mg/L Soak: 10min
ppm)
Reagents Alkaline: - 650ppm NaOH
- 750ppm NaClO
4.1.5. Haloacetic acids e TOC
 Alto Cotia Acid: -500ppm Oxalic acid
CIP Frequency Every 30 days
A sample of permeate was analyzed to Duration 4 hours
determine total haloacetic acids concentration. Reagents Alkaline: - 1000ppm NaOH
These substances are disinfection byproducts - 2000ppm NaClO
(DBPs), formed by reaction between NOM Acid: -2000 ppm H2C2O 4
o
(natural organic matter) with disinfectants, such Temperature 40 C
Recovery 90%

DOW RESTRICTED
 Alto da Boa Vista 4.2.2. Coagulant dosage
 Alto Cotia
Table 7. Alto da Boa Vista – Optimal operational conditions
Parameter Optimal conditions
It has been established the optimum
Backwash frequency 60 minutes dosage of coagulant (5 ppm) to ensure the water
Backwash duration 60 seconds with color values less than 5 uC (conservative
2
Operational Flux 62.3 L/m .h limit). With this dosage of aluminum sulfate,
2
Backwash flux 120 L/m .h
Air Duration 60 seconds
permeate reached the regulatory goals, while the
cleaning Output
3
12 Nm /h conventional treatment used to consider roughly
CEB Frequency Alkaline: Every 48 hours 20-25 ppm of coagulant to be compliant with
Acid: Every 50 hours
limits.
Duration Reagents injection: 90 sec.
Soak: 8 min  Alto da Boa Vista
Reagents Alkaline: - 750ppm NaOH
In that study, no coagulant was used to
- 650ppm NaClO provide good quality water, in accordance with
Acid: -500ppm Oxalic acid
regulatory standards and optimal parameters of
CIP Frequency Estimated: Every 90 days UF treatment.
Duration 4 hours  Indaiá
Reagents Alkaline: - 1000ppm NaOH It has been established that a dosage of 5
- 2000ppm NaClO
ppm of ferric chloride is sufficient to guarantee
Acid: -2000 ppm H2C2O 4
o
color removal (below 5 uC) and provide good
Temperature 40 C
Recovery 97,1%
water quality parameters in compliance with
regulation.
 Indaiá
4.2.3. Membrane permeability
Table 8. Indaiá – Optimal operational conditions Membrane permeability measures how
Parameter Optimal conditions
the water flows through the membrane and is
Backwash frequency 40 minutes
Backwash duration 60 seconds related to amount of permeate passing through
Operational Flux
2
50 L/m .h the membrane per unit time, unit area and
2
Backwash flux 110 L/m .h transmembrane pressure. The trend in
Air Duration 60 seconds
cleaning 3 permeability values (normalized) for the duration
Output 12 Nm /h
CEB Frequency Alkaline: Every 48 hours of the pilot studies can be seen in following
Acid: Every 24 hours graphs:
Duration Reagents injection: 90 sec.  Alto Cotia
Soak: 10min

Reagents Alkaline: - 650ppm NaOH


- 500ppm NaClO

Acid: -500ppm Oxalic acid


CIP Frequency Estimated: Every 60 days
Duration 4 hours
Reagents Alkaline: - 1000ppm NaOH
- 2000ppm NaClO

Acid: -2000 ppm H2C2O 4


o
Temperature 40 C
Recovery 93,4%

Considering typical recoveries of conventional


treatment (around 95%), it’s important to improve the Figure 6. Alto Cotia - Normalized Permeability and trendline
UF conditions to reach high recovery levels. In all of
the case studies, the recovery rate expected for a UF In figure 6, it has been considered optimal
system was achieved (90-97%). Those recoveries were operational parameters and a coagulant dosage
calculated not taking into consideration a possible (ferric chloride) of 5 ppm. Permeability decreasing
recirculation of part of the backwash water discharge,
can be predicted by a trendline:
which can contribute to an even higher recovery rate.
P = 97.924 – 0.0043T

DOW RESTRICTED
Where: P = permeability (LMH/bar) and T = time In figure 8, it has been considered coagulant
(minutes) dosage (ferric chloride) of 5 ppm. It’s possible to
It’s possible to realize a sustainable operation observe two different stages of performance and
over the time, considering permeability reducing different trendlines:
(TMP increasing) and consistent behavior of - First stage (P1) trendline:
recovery of permeability after regular cleanings. P1 = 99.867 - 5.3809T
- Second stage (P2) trendline:
 Alto da Boa Vista
P2 = 69.195 - 0.00093T
Even with regular cleanings, it’s possible
to observe substantial reducing of permeability,
caused mainly by coagulant use which may
contribute to formation of bigger particles that
cause membrane fouling. Second stage
permeability values, considering optimal
parameters, indicate more stable operation, but
with lower permeability values than initial ones.

4.2.4. Transmembrane Pressure


Transmembrane pressure refers to the
pressure required for water passes through the
Figure 7. Alto da Boa Vista - Normalized Permeability and
trendline membrane pores (inlet pressure minus outlet
pressure). As the filtered water is being produced,
In the study on figure 7, no coagulant was residues will be deposited on the outer surface of
added to treatment, providing higher permeability the membrane, causing pressure increasing.
values.  Alto Cotia
Again, permeability decreasing can be followed by
a trendline:
P = 127.23 – 0.00008T
It was part of the study to operate in different CEB
frequencies and it is noticed a substantial
decrease, suggesting a TMP increase caused
exactly by reducing of CEB frequencies (48 hours –
alkaline/oxidant and 50 hours – acid).

 Indaiá

Figure 9. Alto Cotia - Normalized TMP and Flux, TMP


trendline and pressure limits

Flux values show a very stable operation


during test period (5 ppm of coagulant), even with
TMP increasing. During this period, TMP values
were below CIP limit but considering raising rate,
it’s possible to predict a CIP frequency of 23 days
(trendline: TMP = 0,4988 + 0,00003T), which is
conservative due to use of coagulant for color
removal.
Figure 8. Indaiá - Normalized Permeability and two different
trendlines
 Alto da Boa Vista

DOW RESTRICTED
Several empirical formulas have been
developed which allow obtaining an
approximation of sludge generation in water
treatment plants. All equations are conceptually
similar and generally assumed that sludge
produced by the water treatment have two
components: the first consists of the solid fraction
of the raw water (generally based as solids
concentration or turbidity) and the second
component comes from chemicals used in
treatment, which precipitate, are insoluble or are
Figure 10. Alto da Boa Vista - Normalized TMP, trendline and
adsorbed (examples: polymers, coagulants) (19).
pressure limits Using one of these formulas, it’s possible to have
an idea of sludge production of UF as water
In this graph, it’s possible to observe three treatment technology.
different flux conditions: initial (54.5 lmh),
intermediate (59.7 lmh) and final (62.3 lmh). Formula: Association Francaise Pour L'Etude Des
Therefore, TMP values are directly correlated with Eaux (AFEE)
flux conditions, remarkably TMP peaks appear P = (1,2 x Tu + 0,07 x C + 0,17 x D + A) x 10 -3
during higher flux conditions. Also, it’s observed Where: P = solids production (kg dry matter/m³
stable operation, with values below CIP limit and water treated)
predicted frequency above 231 days (trendline: Tu = raw water turbidity (NTU)
TMP = 0.4551 + 0,000003T). C = apparent color of the raw water (uC)
D = dosage of aluminum sulfate (mg/L)
 Indaiá A = Additives such as polymer (mg/L)

Sludge generation prediction (considering


5 ppm of aluminum sulfate when coagulant is
present):
 Alto Cotia: 0.012 (kg dry
matter/m³ water treated)
 Alto da Boa Vista: 0.009 (kg dry
matter/m³ water treated)
 Indaiá: 0.023 (kg dry matter/m³
water treated)

Figure 11. Indaiá - Normalized TMP, trendline and pressure


Considering formula components, it’s clear
limits that conventional treatment will generate more
sludge compared with UF technology, related to
In this case study, two flux conditions the reduction of coagulant dosage (or even none
were observed. With values around 60 lmh, TMP dosage) and chemicals. This is an important
values indicated an unstable operation (peaks parameter to be considered in a water treatment
near CIP limit) due to high flux for this condition plant, since a reduction in sludge generation, will
(use of coagulant). With flux reduction to 50 lmh, result in less expenses with waste disposal.
UF pilot showed better conditions and
consequently more stable operation over the time
and with a CIP frequency (predicted) of 23 days 5. Conclusions
(trendline: TMP = 0,5 + 0,00003T). In conclusion, outside-in pressurized UF has been
shown to be a viable process for production of
4.2.5. Sludge production drinking water. The UF pilot, in three pilot studies
with different source waters, allowed sustainable

DOW RESTRICTED
operation when optimal parameters were applied, (10) Dow Ultrafiltration – Product Manual
even with variability of feed water quality. In (11) AWWA (2005), Microfiltration and
these cases, good performance and superior Ultrafiltration Membranes for Drinking Water -
water quality (remarkably very low turbidity, color Manual of Water Supply Practices, M53 (1st
removal with coagulant dosing and algae removal) Edition), American Water Works Association
demonstrate feasibility of technology as a good (AWWA)
alternative for conventional water treatments. (12) R. Sahai (2000), MEMBRANE SEPARATIONS |
Additionally, sludge reduction and lower OPEX Filtration, In: Encyclopedia of Separation Science,
contribute to grow interest in membrane Academic Press, Oxford, Pages 1717-1724
technology and help to increase options of (13) H. Strathmann (2012), Membranes and
application of UF. Membrane Separation Processes. In: Ullmann's
For future studies, more pilot trials can be Encyclopedia of Industrial Chemistry, Wiley-VCH
performed to increase knowledge and create
Verlag GmbH & Co. KGaA, Weinheim,Vol. 22,
guidelines for different sources of water and
optimal operational parameters for each one. pages 413-456
(14) H. Lutz (2010), 2.06 - Ultrafiltration:
6. Bibliography Fundamentals and Engineering, In:
(1) Chiu, K. P., Gramith, K., & Oppenheimer, J. Comprehensive Membrane Science and
(2005). Development of a microfiltration and Engineering, Elsevier, Oxford, Pages 115-139
ultrafiltration knowledge base. S. Adham (Ed.). (15) Kubota, N., Hashimoto, T. and Mori, Y. (2008)
American Water Works Association.
Microfiltration and Ultrafiltration, in Advanced
(2) Shengji, X., Xing, L., Ji, Y., Bingzhi, D., &
Juanjuan, Y. (2008). Application of membrane Membrane Technology and Applications , John
techniques to produce drinking water in Wiley & Sons, Inc., Hoboken, NJ, USA
China. Desalination,222(1), 497-501. (16) Xu, J., Ruan, G., Gao, X., Pan, X., Su, B., & Gao,
(3) Herschell Green, J., & Tylla, M. (1998). A C. (2008). Pilot study of inside-out and outside-in
comparison of ultrafiltration on various river hollow fiber UF modules as direct pretreatment of
waters. Desalination, 119(1), 79-83.
seawater at low temperature for reverse
(4) Lipp, P., & Baldauf, G. (2002). Application of
osmosis. Desalination, 219(1), 179-189.
out—in MF/UF-systems for drinking water
treatment with air supported backwash—three (17) Kennedy, M. D., Kamanyi, J., Rodríguez, S. G.,
case studies. Desalination, 147(1), 63-68. Lee, N. H., Schippers, J. C., & Amy, G. (2008).
(5) P. Huck, M. Sozański (2011), 3.16 - Chemical Water treatment by microfiltration and
Basis for Water Technology, In: Treatise on Water ultrafiltration.Advanced Membrane Technology
Science, Elsevier, Oxford, Pages 429-469 and Applications, 131-170.
(6) Laîné, J. M., Vial, D., & Moulart, P. (2000). (18) AWWA Subcommittee On Periodical
Status after 10 years of operation—overview of Publications Of The Membrane Process
UF technology today. Desalination, 131(1), 17-25. Committee (2008), Microfiltration and
(7) Gao, W., Liang, H., Ma, J., Han, M., Chen, Z. L., ultrafiltration membranes for drinking water,
Han, Z. S., & Li, G. B. (2011). Membrane fouling Journal (American Water Works Association), Vol.
control in ultrafiltration technology for drinking 100, No. 12, pp. 84-97
water production: A review. Desalination, 272(1), (19) Katayama, V. T. (2012). Quantificação da
1-8. produção de lodo de estações de tratamento de
(8) Y. Watanabe, K. Kimura (2011), 4.02 - água de ciclo completo: uma análise crítica.
Membrane Filtration in Water and Wastewater Master's Dissertation, Escola Politécnica,
Treatment, In: Treatise on Water Science, Elsevier, University of São Paulo, São Paulo
Oxford, Pages 23-61 (20) Bentama, J., Schmitz, P., Destrac, P., &
(9) Guo, X., Zhang, Z., Fang, L., & Su, L. (2009). Espenan, J. M. (2004). Technological innovation
Study on ultrafiltration for surface water by a for the production of drinking water by
polyvinylchloride hollow fiber membrane processes. Desalination, 168, 283-286.
membrane. Desalination,238(1), 183-191. (21) Karnik, B. S., Davies, S. H., Baumann, M. J., &
Masten, S. J. (2005). Fabrication of catalytic

DOW RESTRICTED
membranes for the treatment of drinking water
using combined ozonation and
ultrafiltration. Environmental science &
technology,39(19), 7656-7661.
(22) Berghmans, S., Berghmans, H., & Meijer, H. E.
H. (1996). Spinning of hollow porous fibres via the
TIPS mechanism. Journal of membrane
science, 116(2), 171-189.
(23) Kimmerle, K., & Strathmann, H. (1990).
Analysis of the structure-determining process of
phase inversion membranes. Desalination, 79(2),
283-302.
(24) Li, Dongmei, William B. Krantz, Alan R.
Greenberg, and Robert L. Sani. "Membrane
formation via thermally induced phase separation
(TIPS): Model development and
validation." Journal of membrane science 279, no.
1 (2006): 50-60.
(25) SABESP (2012), Relatório Anual de Qualidade
da Água – Embu-Guaçu
(26) SABESP (2012), Relatório Anual de Qualidade
da Água – São Paulo
(27) SABESP (2012), Relatório Anual de Qualidade
da Água – Bertioga
(28) Zhang, Y., Tian, J., Nan, J., Gao, S., Liang, H.,
Wang, M., & Li, G. (2011). Effect of PAC addition
on immersed ultrafiltration for the treatment of
algal-rich water. Journal of hazardous
materials, 186(2), 1415-1424.
(29) Wallace, B., Purcell, M., & Furlong, J. (2002).
Total organic carbon analysis as a precursor to
disinfection byproducts in potable water:
Oxidation technique considerations. Journal of
Environmental Monitoring, 4(1), 35-42.

DOW RESTRICTED

You might also like