CH 12

MWH’S WATER TREATMENT
PRINCIPLES AND DESIGN

3rd Edition
by Crittenden, Trussell, Hand, Howe, and Tchobanoglous
HOMEWORK SOLUTION MANUAL
FOR
Chapter 12
Membrane Filtration
Note: If any errors are noted in this solution manual or in the textbook, please notify
Kerry Howe at howe@unm.edu
PROBLEM 12-1
Problem Statement - Discuss the differences between MF and UF membranes. What
impact do these differences have on their use in water treatment?
Solution
1. Differences between MF and UF membranes include the following:
Microfiltration membranes have a larger retention rating (retain smaller particles)
than UF membranes. Microfiltraion membranes have a homogeneous structure
and UF membranes have an asymmetric structure. Different rating systems are
used for each type of membrane. Microfiltration membranes are rated by pore
size and typically have a rating between 0.1 µm and 1 µm. Ultrafiltration
membranes are rated by MWCO and typically have a rating roughly between
1,000 and 500,000 Daltons, which means particle retention capabilities are
between 1 and 30 nm.
2. The difference in retention rating has an impact on the objectives that each
membrane can achieve in water treatment. Microfiltration membranes cannot
reliably remove viruses, but UF membranes with a low enough retention rating can
achieve absolute virus removal.
PROBLEM 12-2
Problem Statement - Discuss the similarities, differences, advantages, and
disadvantages between membrane filtration and rapid granular filtration. This is an
essay question.
Solution
Rapid granular filtration and membrane filtration are both used for the removal of
particles from water. Both can reliably provide safe drinking water when operated
correctly. Both are typically able to achieve product water recovery above 95 percent.
Some of the differences between rapid granular filtration and membrane filtration are
listed in Table 12-4. The filtration rate for rapid granular filtration is about 100 times
greater than the filtration rate typical in membrane filtration. Rapid granular filtration can
operate at a lower pressure and has longer filter run times than membrane filtration, but
Homework Solution Manual Page 2 of 29

MWH’s Water Treatment: Principles and Design, 3rd ed. Version 1
Chapter 12 - Membrane Filtration
the backwash cycles are also longer. The most important difference between rapid
granular filtration and membrane filtration is the filtration mechanism. Membrane
filtration removes particles by straining and rapid granular filtration removes particles by
depth filtration. Depth filtration requires that particles be destabilized before they
adhere to the filter grains, which requires the use of a coagulant upstream of the
filtration process. Depth filtration has a ripening period, which does not exist in
membrane filtration. Straining is a more robust filter mechanism than depth filtration.
Membrane filtration has several advantages over rapid granular filtration. The
coagulation/filtration process of rapid granular filtration is impacted by changes in water
quality and operator error, both of which can adversely impact the water quality of the
filter effluent. Membrane filtration plants require less pre-treatment steps and are less
likely to release improperly treated water to the distribution system when operator errors
occur. The effluent particle concentration from membrane filtration plants is
independent of raw water quality, which is not the case for rapid granular filtration
plants. The lack of coagulation and flocculation facilities can make membrane filtration
easier to operate than rapid granular filtration. The density of filter area, along with the
lack of pretreatment processes often results in a membrane filtration plant that is more
compact than a rapid granular filtration plant would be for treating the same water
source.
The primary advantage of rapid granular filtration over membrane filtration is that rapid
granular filtration has a long and rich operating history, whereas membrane filtration is a
relatively new technology. Because of the long operating history of rapid granular
filtration the process is well understood, and it is possible to design and construct an
efficient rapid granular filtration plant without having to conduct pilot studies. Pilot
studies are necessary for membrane filtration plants to determine flux, backwashing
requirements, and cleaning requirements. Permitting of rapid granular filtration plants is
easier currently than membrane filtration plants, although this situation will likely change
the longer membrane plants are operated successfully and as regulators become more
familiar and comfortable with membrane technology.

PROBLEM 12-3
Problem Statement - How effective do you think membrane filtration is for each of the
following treatment issues? Explain your reasoning.
a. Arsenic
b. Anthrax spores
c. Hardness
d. Taste and odor
Solution
a. Membrane filters primarily remove particles, and because the filters are frequently
operated without coagulation or other pretreatment processes, they are typically
ineffective for removing dissolved species. Arsenic is a dissolved constituent in
natural water and thus would be poorly removed by membrane filters (although
some research has shown that a fraction of the arsenic present in water can
adsorb to naturally occurring particulate matter, which would then be removed by
membrane filters).
b. Anthrax spores are particles and would be removed effectively by membrane
filters.
c. Hardness is present as dissolved species and would not be removed by
membrane filters without some pretreatment process.
d. Taste and odor are present as dissolved species and would not be removed by
membrane filters without some pretreatment process.
PROBLEM 12-4
Problem Statement - Calculate rejection and log removal value for the following
filtration process (to be selected by instructor). Use the number of significant figures
necessary to correctly illustrate the removal being obtained.
A B C D E
Influent concentration (#/mL) 106 6.85 × 105 7.1 × 105 1.65 × 107 2.8 × 106
Effluent concentration (#/mL) 10 136 0.16 65 96

Instructor’s Notes: This problem is similar to Example 12-2. The solution is worked
out for Problem B and the answers for the remaining problems are included in a table at
the end of the solution.
Solution
1. Calculate removal using Eq. 12-3:
Cp 136 mL−1
R =−
1 =−
1 =0.999801
Cf 6.85 × 105 mL−1
2. Calculate the log removal value using Eq. 12-4:
 Cf   6.85 × 105 mL−1 
= =
LRV log   log  = −1  3.70
 Cp   136 mL 
Answers to all problems:
A B C D E
Rejection 0.99999 0.999801 0.99999977 0.99999606 0.9999657
Log rejection value 5.00 3.70 6.65 5.40 4.46
PROBLEM 12-5
Problem Statement - An inside-out hollow-fiber membrane system is operated with a
cross-flow configuration. Each module contains 10,200 fibers that have an inside
diameter of 0.9 mm and a length of 1.75 m. Calculate the following for one module:
a. Feed flow necessary to achieve a cross-flow velocity of 1 m/s at the entrance to
the module.
b. Permeate flow rate if the system maintains an average permeate flux of 80
L/m2 · h.
c. Cross-flow velocity at the exit to the module.
d. Ratio of the cross-flow velocity at the entrance of the module to the flow velocity
toward the membrane surface. Given the magnitude of this ratio, what effect would
you expect cross-flow velocity to have on fouling in cross-flow versus dead-end
filtration?

e. Ratio of permeate flow rate to feed flow rate (known as the single-pass recovery).
What impact does this ratio have on operational costs in cross-flow versus dead-
end filtration?
Solution: Part a
1. Calculate the cross-sectional area of one fiber lumen.
π 2 π
( )
2
a XC = d =0.9 × 10 −3 m 6.362 × 10 −7 m2
=
4 4
2. Calculate feed flow to 1 fiber.
QF,1 =
va XC = (
(1 m/s ) 6.362 × 10−7 m2 = )
6.362 × 10 −7 m3 /s
3. Determine the total flow by multiplying QF,1, the result of the step 2, by the number
of fibers in the module.
QF = (10,200 fibers ) ( 6.362 × 10 −7 m3 /s ) ( 3,600 s/h ) = 23.4 m3 /h
Solution: Part b
1. Calculate the inside surface area of one fiber.
aS = π ( 0.9 × 10 −3 m ) (1.75 m ) =
πdL = 4.948 × 10 −3 m2
2. Calculate permeate flow from 1 fiber.

JaS ( 80 L/m2 h )( 4.948 × 10 −3 m2 ) =
QP,1 == 0.3958 L/h
3. Determine the total flow by multiplying QF,1, the result of the step 2, by the number
of fibers in the module.
QP 10,200 fibers )( 0.3958 L/h ) (10-3 m3 /L )
(= 4.04 m3 /h
Solution: Part c
1. Calculate the retentate flow by mass balance.
Q=
F QP + QR
QR = QF − QP = 23.4 m3 /h − 4.04 m3 /h = 19.36 m3 /h

2. Calculate the retentate velocity.
QR 19.36 m3 /h
=
vR = = 0.829 m/s
NF a XC (10,200 fibers ) ( 6.362 × 10−7 m2 ) (3,600 s/h)
Solution: Part d
1. Calculate the ratio of the feed velocity to the permeate velocity.
v (1 m/s ) (103 L/m3 ) ( 3,600 s/h )

Ratio= = = 45,000
J (80 L/m2 h)
3. Impact:
The water velocity parallel to the membrane surface is 45,000 times greater than
the velocity toward the membrane. This high cross-flow velocity causes strong
shear forces that are not present in dead-end filtration, which reduces the
accumulation of particles at the membrane surface. As a result, cross-flow filters
foul at a slower rate than dead-end filters and should be able to operate at a
higher average flux.
Solution: Part e
1. Calculate the ratio of the permeate flow rate to the feed flow rate.
QP 4.04 m3 /h
=
Ratio = = 0.173
QF 23.4 m3 /h
2. Impact:
The permeate flow is only 17.3 percent of the feed flow, and the remainder of the
feed flow becomes retentate and is recycled to the head of the plant. A 20,000
m3/d (5.3 mgd) water treatment plant would have feed pumps and piping sized to
handle 180,000 m3/d (47.6 mgd) and would be recycling 82.7 percent of the flow,
adding a significant amount to the cost of operating the plant.
PROBLEM 12-6
Instructors Note: The procedures for the solution of this problem are demonstrated in
Examples 12-3, 12-4, and 12-5. The complete solution for Problem A is worked out in
this solution, and the solution to the remaining problems is shown in a table at the end
of the solution.
Problem Statement - Hollow fiber membranes with a membrane area of 23.3 cm2 were
tested in a laboratory and found to have the clean water flow shown in the table
below, at the given temperature and pressure.

A B C D E
Flow (mL/min) 4.47 4.22 2.87 6.05 1.22
Temperature (°C) 16 22 23 25 22
Pressure (bar) 0.67 0.80 0.71 1.25 0.21
For the data set selected by your professor,
a. Calculate the specific flux at 20º C.
b. Calculate the membrane resistance coefficient.
c. Does membrane resistance coefficient depend on the pressure and temperature

used for the tests? Why or why not?
Solution
1. Calculate the flux at a standard temperature of 20°C using Eq. 12-8. Since flow =
4.47 mL/min or 0.268 L/h, J=Q/α or 115 L/m2•h:
Q 4.47 mL/min ( 60 min/h ) (L/1000 mL)
J= = =115.1 L/m2 ×h
A 23.3 cm2 (1 × 10 −4 )
(T )
(115.1 L/m h) (1.03 )( ) 20 C −16 C
M (1.03 )
−T
=JS J= S M
=
2
129.6 L/m2 h
2. Calculate the specific flux using Eq. 12-9.
JSP,20=
JS
=
(129.6 L/m =
h)
2
193.4 L/m2 hbar


C
∆P 0.67 bar
3. Rearrange Eq. 12-6 to solve for the membrane resistance coefficient. The
viscosity of water at 20 °C is 1.00 x 10-3 kg/m·s (see Appendix C-1). Recall that 1
bar = 100 kPa = 105 N/m2 = 105 kg/s2·m.
∆P (1 bar ) (105 )
kg/s2 mbar ( 3,600 s/h ) 103 L/m3 ( )=
κM= = 1.89 × 1012 m-1
µJ (1.00 × 10 -3
)(
kg/ms 193.4 L/m h 2
)
4. The membrane resistance coefficient can be calculated using either the actual flow
and pressure (115.1 L/m2·h at 16 °C and 0.67 bar) or the standard flow and
pressure (193.4 L/m2·h at 20 °C and 1 bar) and the results are the same (within the

accuracy of the temperature correction formula). Thus, the membrane resistance
coefficient is independent of pressure and temperature.
Answers to all problems:

A B C D E
Measured flux L/m2∙h 115.1 108.7 73.9 155.8 31.4
Flux correction Eq. 12-8 L/m2∙h 129.6 102.4 67.6 134.4 29.6
Specific flux at std temp L/m2∙h∙bar 193.4 128.0 95.3 107.5 141.0
Flux correction Eq. 12-7 L/m2∙h 127.4 103.6 68.8 138.5 29.9
Specific flux at std temp L/m2∙h∙bar 190.2 129.4 96.9 110.8 142.6
Membrane resist. coeff.

m-1 1.89E+12 2.78E+12 3.71E+12 3.24E+12 2.52E+12
w/ Eq. 12-6
Membrane resist. coeff.

m-1 1.89E+12 2.78E+12 3.71E+12 3.24E+12 2.52E+12
w/ spec flux
PROBLEM 12-7
Problem Statement - The 0.2-µm polyethersulfone microfiltration membrane shown on
Fig. 12-3 was tested in the laboratory and found to have a clean-water flux of 6500
L/m2 · h at 23º C and 0.69 bar. Assume that the flow through a microfiltration
membrane can be modeled using the Kozeny equation for flow through porous
media (Eq. 11-11 in Chap. 11).
a. Calculate the specific surface area of the membrane assuming a porosity of 0.6,
thickness of 0.10 mm, and Kozeny coefficient of 5.0.
b. What would the theoretical grain diameter be if the membrane were composed of
spherical granular media with the same specific surface area (see Eq. 11-6)?
c. How does the theoretical grain diameter compare to the retention rating for the
membrane? Using concepts of particle retention through granular media from
Chap. 11, what does this comparison suggest about the mechanisms for particle
removal in microfiltration?
d. Using the theoretical grain diameter as the characteristic dimension, calculate the
Reynolds number for flow through a microfiltration membrane. Is the flow laminar
or turbulent?

Solution: Part a
1. Rearrange Eq. 11-11 to solve for the specific surface area.
0.5
 h ρ gε 3 
S= L W 
 κk µvL 
Note that head can be expressed in units of pressure by multiplying by ρW g, thus
∆P = hLρW g.
2. Water density and viscosity at 23 °C are 997.5 kg/m3 and 9.326 x 10-4 kg/m·s,
respectively (by extrapolation from Appendix C-1).
3. Convert flux to units of m/s (velocity).
Note that membrane flux is equivalent to superficial velocity or filtration rate.
6,500 L/m2 h
=v = 1.805 × 10 −3 m/s
( )
103 L/m3 ( 3,600 s/h )
4. Calculate specific surface area
( ) ( )
0.5
 ( 0.69 bar ) 105 kg/s2 mbar ( 0.6 )3 103 mm/m 
S 
= = 4.21× 106 m-1
 ( −4
)(
 ( 5 ) 9.326 × 10 kg/ms 1.805 × 10 m/s ( 0.1 mm ) 
−3
)
Solution: Part b
Rearrange Eq. 11-6 and calculate the theoretical grain diameter, assuming the
membrane is constructed of spherical grains.
6 (1 − ε ) 6 (1 − 0.6 )
d= = 5.70 × 10 −7 m =
= 0.570 µm
S 4.21× 106 m-1
Solution: Part c
The nominal retention rating is given as 0.2 µm and the calculated theoretical grain
diameter is 0.57 µm (see Part b). Theoretically, granular media can strain particles
larger than about 0.15 times the grain diameter (see Fig. 11-8). Thus, a media
composed of spherical grains with a diameter of 0.57 µm would be able to strain
particles 0.086 µm and larger. The similarities between the size of the particles retained
by a granular media (using the calculated theoretical grain size) and the corresponding
nominal retention rating of the microfiltration membrane, means there are similarities
between the flow-pressure relationship in granular media and membrane filters.
Solution: Part d
1. Calculate the Reynolds number.
=
ρvd
Re =
( 997.5 kg/m )(1.805 × 10
3 −3
)(
m/s 5.70 × 10 −7 m )
= 1.10 × 10 −3
−4
µ 9.326 × 10 kg/ms
2. A Reynolds number of 1.10 x 10-3 in porous media corresponds to laminar flow.
Laminar flow means that the head loss through the media varies linearly with flow
velocity, which is commonly observed for membrane filters (if flow were turbulent,
head loss would vary by the square of the velocity). In addition, the Kozeny
equation (which applies to laminar flow) would be an appropriate way to calculate
flow through porous media, assuming the parameters in the equation could be
measured with some degree of accuracy.
PROBLEM 12-8
Instructors Note: The solution of this problem is similar to Example 12-4.
Problem Statement - A membrane plant is operated at a volumetric flux of 75 L/m2 · h

at 17º C and 0.85 bar. Calculate the specific flux at 20º C.
Solution
1. Calculate the flux at a standard temperature of 20 °C using Eq. 12-8.
M (1.03 )
(T )
( 75 L/m h) (1.03 )( )
−T 20 C −17 C
=JS J= S M 2
= 82 L/m2 h
JSP,20=
JS
=
( 82 L/m =
h)
2
96 L/m2 hbar

C
∆P 0.85 bar
PROBLEM 12-9
Problem Statement – Feed water pressure and temperature and permeate flux at a
membrane filtration plant are reported on two dates below. For the plant selected by
your instructor, calculate the specific flux on each date, and indicate whether fouling has
occurred between the first and second dates.
A B C D E
Day 1

Flux (L/m2·h) 72 26 31 86 112
Temperature (°C) 21 17 17 22 19
Pressure (bar) 0.62 0.24 0.24 0.72 0.66
Day 2
Flux (L/m2·h) 56 26 27 90 120
Temperature (°C) 4 15 10 25 11
Pressure (bar) 0.80 0.29 0.26 0.77 1.05
Instructors Notes: This problem is similar to Example 12-4. The solution is solved
completely for Problem A and the solution to the remaining problems are shown in a
table at the end of this solution.
Solution
1. Calculate the flux at a standard temperature of 20 °C using Eq. 12-8 for Day 1.
(T )
( 72 L/m h) (1.03 )( ) 20 C − 21 C
M (1.03 )
−T
=JS J= S M 2
= 69.9 L/m2 h
JSP,20=
JS
=
( 69.9 L/m =
h)2
112.7 L/m2 hbar


C
∆P 0.62 bar
3. Calculate the flux at a standard temperature of 20 °C using Eq. 12-8 for Day 2.
( )
= M (1.03 )
JS J= S M (T −T )
(56 L/m h) (=
2
1.03 )
20 C − 4 C
89.9 L/m2 h
JSP,20=
JS
=
(89.9 L/m =
h) 2
112.3 L/m2 hbar


C
∆P 0.80 bar
5. Solutions for remaining filtration plants:
A B C D E
DAY 1
Flux temp corrected
with Eq. 12-8 (L/m2·h) 69.9 28.4 33.9 81.1 115.4
Specific flux
(L/m2·h·bar) 112.7 118.4 141.1 112.6 174.8
DAY 2
Flux temp corrected 89.9 30.1 36.3 77.6 156.6

with Eq. 12-8 (L/m2·h)
Specific flux
(L/m2·h·bar) 112.3 103.9 139.6 100.8 149.1
Not
Fouling occurred? No Yes much Yes Yes
PROBLEM 12-10
Instructors Note: Q = Ja, so Eq. 12-9 can be used to calculate the standard flow rate
the same way standard flux would be calculated.
Problem Statement - A new membrane plant is being designed. Pilot testing indicates
that the membrane will be able to operate at a specific flux of 120 L/m2·h·bar at
20ºC. Water demand projections predict a summer peak-day demand of 90,000
m3/d and a winter peak-day demand of 60,000 m3/d. Historical records indicate
that the source water has a minimum temperature of 3º C in winter and 18º C in
summer.
a. Which season will govern the size of the plant?
b. What is the required membrane area, assuming the plant will operate at 0.8
bar, the online production factor is 95 percent, and the recovery is 97
percent?
Solution
1. Flux at standard temperature, JS = 120 L/m2•h•bar x 0.8 bar = 96 L/m2
a. Calculate the flux in summer at a standard temperature of 20°C using Eq.12-8.
=
Jm, Summer =
JS
=
(
96 L/m2 h ) 90.5 L/m2 h
(
1.03 )
( TS − TM )  1.03(


20
C −18
C )


b. Calculate the flux in winter at a standard temperature of 20°C using Eq. 12-8.
=
Jm, W int er =
JS
=
(
96 L/m2 h ) 58.1 L/m2 h
(
1.03 ( TS −TM )
) 1.03(


20
C − 3
C )


2. Calculate the membrane area for summer conditions using Eq. 12-29.

=
Qf
a =
(90,000 m 0.97 )(10 L/m=)
3 3 3
45,000 m2
Jη (90.5 L/m h) ( 24 h/d)(0.95 )
2
Calculate the membrane area for winter conditions using Eq. 12-29.
=
Qf
a =
( 60,000 m 0.97 )(10 L/m=)
3 3 3
46,700 m2
Jη (58.1 L/m h) ( 24 h/d)(0.95 )
2
3. Compare the two values.

46,700 m2 > 45,000 m2, so winter > summer
The winter operating conditions require greater capacity than summer, when they
are compared at the same temperature. Therefore, the winter operating condition
controls the capacity, and the plant will have excess capacity during the summer
months.
PROBLEM 12-11
Problem Statement - An ultrafiltration membrane with a membrane resistance
coefficient of 2.7 × 1012 m–1 is used to filter a 150 mg/L suspension of 0.5-µm latex
particles in a laboratory unstirred dead-end filtration cell. The experiment is
operated at a constant flux of 120 L/m2 · h and temperature of 20º C, and the
membrane has an area of 28.2 cm2. Assume that fouling is due to cake formation,
the particle density is 1050 kg/m3, the cake porosity is 0.38, and the Kozeny
coefficient is 5. Neglecting the backmigration of particles due to diffusion, calculate
and plot the transmembrane pressure and specific flux over the first 90 min of the
filter run.
Solution
The basic solution strategy is to create a spreadsheet with columns for (1) time, (2)
volume of water filtered, (3) cake thickness, (4) cake resistance coefficient, (5) pressure,
and (6) specific flux, and then plot pressure and specific flux as a function of time. The
solution shown below uses time steps of 5 minutes. The calculation steps are shown
for the first time step, and the remaining values are shown in the following table.
1. Convert the given values to SI units.

28.2 cm2
Membrane area: =a = 2.28 × 10 −3 m2
(100 cm/m )
2
120 L/m2 h
Flux: =J = 3.33 × 10 −5 m/s
( )
10 L/m ( 3,600 s/h )
3 3
=
Concentration:
150 mg/L 103 L/m3
C = 0.15 kg/m3
( )
106 mg/kg
0.5 µm
Particle diameter: dP= = 5 × 10 −7 m
106 µm/m
2. Calculate the volume of water filtered at 5 minutes.
V = Jat
( )(
3.33 × 10 −5 m/s 2.28 × 10 −3 m2 ( 5 min )( 60 s/min ) =
V= )
2.28 × 10 −5 m3
3. Calculate the cake thickness at 5 minutes using Eq. 12-13.
δ=
CV
=
( 0.15 kg/m )( 2.28 × 10 m ) =
3 −5 3
2.30 × 10 −6 m
C
ρPa (1 − ε ) (1050 kg/m )( 2.28 × 10 m ) (1 − 0.38 )
3 −3 2
4. Calculate the cake resistance coefficient at 5 minutes using Eq. 12-12.
κ=
36 κK (1 − ε ) δC 36 ( 5 )(1 − 0.38 ) 2.30 × 10 m
2
=
−6 2
(
= 1.16 × 1010 m-1
)
ε3 dP2 ( 0.38 ) 5 × 10 m ( )
C 3 2
−7
5. Calculate the pressure using Eq. 12-11, noting the κA = 0. The viscosity at 20 °C is
1.0 x 10-3 kg/m·s (see Appendix C-1).
∆P = Jµ ( κM + κC )
= ( 3.33 × 10 −5
)( )(
m/s 1.0 × 10 −3 kg/ms 2.7 × 1012 m-1 + 1.16 × 1010 m-1 )
= 9.04 × 10 4 kg/ms2
= 0.904 bar
J 120 L/m2 h
J= = = 132.76 L/m2 hbar
∆P 0.904 bar
SP
7. Repeat steps 2 through 6 for the remaining time steps, and plot pressure and
specific flux as a function of time. The results are shown in the following table and
graphs.

Cake
Cake
Volume resistance
thickness, Pressure,
Time, filtered, coefficient, Specific flux,
3 -1 2
min m m m bar L/m ·h·bar
0 0.00 0.00 0.00 0.900 133.33
-5 -6 10
5 2.82 x 10 2.30 x 10 1.16 x 10 0.904 132.76
-5 -6 10
10 5.64 x 10 4.61 x 10 2.32 x 10 0.908 132.20
-5 -6 10
15 8.46 x 10 6.91 x 10 3.49 x 10 0.912 131.63
-4 -6 10
20 1.13 x 10 9.22 x 10 4.65 x 10 0.915 131.08
-4 -5 10
25 1.41 x 10 1.15 x 10 5.81 x 10 0.919 130.52
-4 -5 10
30 1.69 x 10 1.38 x 10 6.97 x 10 0.923 129.98
-4 -5 10
35 1.97 x 10 1.61 x 10 8.14 x 10 0.927 129.43
-4 -5 10
40 2.26 x 10 1.84 x 10 9.30 x 10 0.931 128.89
-4 -5 11
45 2. x 10 2.07 x 10 1.05 x 10 0.935 128.36
-4 -5 11
50 2.82 x 10 2.30 x 10 1.16 x 10 0.939 127.83
-4 -5 11
55 3.10 x 10 2.53 x 10 1.28 x 10 0.943 127.31
-4 -5 11
60 3.38 x 10 2.76 x 10 1.39 x 10 0.946 126.78
-4 -5 11
65 3.67 x 10 3.00 x 10 1.51 x 10 0.950 126.27
-4 -5 11
70 3.95 x 10 3.23 x 10 1.63 x 10 0.954 125.76
-4 -5 11
75 4.23 x 10 3.46 x 10 1.74 x 10 0.958 125.25
-4 -5 11
80 4.51 x 10 3.69 x 10 1.86 x 10 0.962 124.74
-4 -5 11
85 4.79 x 10 3.92 x 10 1.98 x 10 0.966 124.24
-4 -5 11
90 5.08 x 10 4.15 x 10 2.09 x 10 0.970 123.75

1.00
0.98
0.96
0.94
Pressure (bar)
0.92
0.90
0.88
0.86
0.84
0.82
0.80
0 10 20 30 40 50 60 70 80 90 100
Time (min)
140
135
130
Specifc flux (L/m2•h•bar)
125
120
115
110
105
100
0 10 20 30 40 50 60 70 80 90 100
Time (min)

PROBLEM 12-12
Problem Statement - Show how the cake layer resistance coefficient (Eq. 12-11) can
be derived from the Kozeny equation (Eq. 11-11 in Chap. 11) when the membrane
feed water is a suspension of monodisperse, well-characterized particles.
Solution
Derive the cake resistance coefficient from the Kozeny equation.
a. If no adsorptive fouling is occurring (κA = 0), Eq. 12-11 can be written as:
∆P
J=
µ ( κM + κ C )
b. Rearrange the equation in step 1a to solve for DP as follows:
∆P = Jµ ( κM + κC ) = JµκM + JµκC
c. The pressure drop through the membrane and cake layer consists of two pressure
terms, one for the membrane and one for the cake layer. The pressure drop
through just the cake layer can be written as follows:
∆PC = JµκC
d. The Kozeny equation (Eq. 11-11) is written as follows:
hL κK µS2 v
=
L ρ W gε 3
e. To eliminate the specific area term in the Kozeny equation, substitute Eq. 11-6 into
Eq. 11-11.
hL κK µ6 (1 − ε ) v
2 2
=
L ρW gε3 dP2
f. Rearrange the equation in step 1e to solve for velocity.
ρW ghL ε3 dP2
v=
36 κK µ (1 − ε ) L
2
g. Substitute the following equalities into the equation developed in step 1f.
1. ∆PC =
ρW ghL (converting pressure to head)
2. v = J (flux is the same as superficial velocity)
3. δC = L (cake thickness is equal to the granular bed depth)

∆PC ε3 dP2
J=
36 κK µ (1 − ε ) δC
2
h. Rearrange the equation in step g3 to solve for DPc.
 36 κK (1 − ε )2 δC 
∆PC = Jµ  
 ε3 dP2 
i. Combining the equations developed in steps 1c and 1h yields the cake resistance
coefficient, which is Eq. 12-12.
36 κ (1 − ε ) δC
2
κC = K 3 2
ε dP
PROBLEM 12-13
Problem Statement - A membrane plant containing 1200 m2 of membrane area
operates at a constant permeate flux of 45 L/m2 · h at a temperature of 15º C and
pressure of 0.25 bar immediately after backwash. The feed water contains 12 mg/L
of suspended solids. After 40 min of operation, the pressure rises to 0.30 bar.
a. Assuming that pressure rise between backwashes is due to formation of a cake
layer, calculate the specific cake resistance.
b. If permeate flux is increased to 50 L/m2 · h, calculate the pressure immediately
after backwash and the pressure after 40 min of operation.
Solution: Part a
1. Convert the flux given in the problem statement to SI units.
45 L/m2 h
=J = 1.25 × 10 −5 m/s
( )
10 L/m ( 3,600 s/h )
3 3
2. Calculate the membrane resistance using Eq. 12-7, assuming that no cake exists
immediately after backwash. From Appendix C-1, µ = 1.14 x 10-3 kg/m·s and ∆P =
0.25 bar = 0.25 x 105 kg/m·s2.
∆P 0.25 × 105 kg/ms2
κM= = = 1.75 × 1012 m-1
µJ ( −3
)( −5
1.14 × 10 kg/ms 1.25 × 10 m/s )
3. Calculate the cake resistance using Eq. 12-11 with only two resistances
(membrane and cake) after 40 minutes of operation.
∆P
J=
µ ( κM + κ C )
∆P 0.3 × 105 kg/ms2

κ= − κ= − 1.75 × 1012 m-1
C
µJ
M
( −3 −5
1.14 × 10 kg/ms 1.25 × 10 m/s)( )
= 3.51× 10 m 11 -1
4. Calculate the volume of water filtered in 40 minutes.
V
(
45 L/m2 h 1,200 m2 ( 40 min )
= 36,000 L
)
60 min/h
5. Calculate the specific cake resistance using Eq. 12-14.
α=
κCa
=
( 3.51× 10 11
)(
m-1 1,200 m2 103 mg/g )( )
= 9.75 × 1011 m/g
C
CV (12 mg/L )( 36,000 L )
Solution: Part b
1. Convert the new flux given in the problem statement to SI units.
50 L/m2 h
=J = 1.39 × 10 −5 m/s
( )
10 L/m ( 3,600 s/h )
3 3
2. Calculate the new pressure immediately after backwash, using Eq. 12-6 and the
membrane resistance calculated in Part a step 2.
∆P = JµκM = (1.39 × 10 −5
)( )(
m/s 1.14 × 10 −3 kg/ms 1.75 × 1012 m-1 )
= 2.77 × 10 4 kg/ms2
= 0.277 bar
3. Calculate the volume of water filtered in 40 minutes at the higher flux.
V
(
50 L/m2 h 1,200 m2 ( 40 min )
= 40,000 L
)
60 min/h
4. Calculate the cake resistance coefficient after 40 minutes using Eq. 12-14.
κ= αC
CV
= 9.75 × 1011 m/g
(12 mg/L )( 40,000 L=
) 3.90 × 1011 m-1
C
a 2 3
1,200 m 10 mg/g ( )( )
5. Calculate the pressure after 40 minutes of operation by re-arranging the equation
used in step a3 (Eq. 12-11 with only two resistances; membrane and cake.)
∆P = Jµ ( κM + κC )
= (1.39 × 10 −5
)( )(
m/s 1.14 × 10 −3 kg/ms 1.75 × 1012 + 3.90 × 1011 m-1 )
= 3.40 × 10 4 kg/ms2
= 0.34 bar
PROBLEM 12-14
Instructor’s Notes: The solution is worked out for Problem A and the answers for the
remaining problems are shown at the end of the solution.
Problem Statement - Calculate the membrane fouling index for the following data, for
the data set specified by your instructor.
A. Experimental flat-sheet laboratory filter, membrane area = 30 cm2, initial flux = 3,560
L/m2·h·bar, test pressure = 0.69 bar, test temperature = 23.9 °C.
Permeate Permeate
Time, Volume, Time, Volume,
min mL min mL
0 0 6 458.3
1 108.8 7 506.8
2 199.8 8 552.1
3 277.4 9 594.1
4 345.0 10 634.1
5 404.2 11 670.8
B. Full scale plant operating at constant permeate flow of 15,000 m3/day, temperature
= 20 °C, 5800 m2 of membrane area, pressure each day as shown below. Use Day
0 as the initial flux.
Transmemb. Transmemb. Transmemb.
Time, Pressure, Time, Pressure, Time, Pressure,
Day Bar Day Bar Day Bar
0 0.704
2 0.712 12 0.747 22 0.786
4 0.721 14 0.754 24 0.794
6 0.726 16 0.765 26 0.801
8 0.735 18 0.770 28 0.812

10 0.740 20 0.777 30 0.812
C. Data from a 30-minute filter run in the middle of a day of laboratory testing of
coagulated feed water, membrane area = 23 cm2, initial flux = 238 L/m2·h·bar, test
pressure = 2.07 bar, test temperature = 21.5 °C.
Permeate Permeate Permeate
Time, Volume, Time, Volume, Time, Volume,
min mL min mL min mL
0 2276.64
2 2292.62 12 2370.17 22 2444.76
4 2308.41 14 2385.31 24 2459.35
6 2324.05 16 2400.33 26 2473.88
8 2339.53 18 2415.24 28 2488.26
10 2354.92 20 2430.04
Solution for Problem A

1. Divide the volume filtered by the membrane area to determine the specific
throughput. Results are in the third column in table below. For the third row,
(199.8 mL ) (104
cm 2 /m 2 )
Vsp = 66.6 L/m 2
( 30.0 cm )(10 mL/L )
2 3
2. Calculate the volume filtered in each time increment by subtracting the previous
volume. Results are in the fourth column in the table. For the third row:
=
∆V 199.8 mL − 108.8
= mL 91 mL
3. Divide the volume filtered in each increment by membrane area and time to
determine flux. Then correct for temperature and pressure using Eqs. 12-8 and 12-9
to determine specific flux. Results are in the fifth column in the table. For the third
row,
( 91 mL ) (104 cm 2 /m 2 ) ( 60 min/h )
Jm = 1820 L/m 2 h
( 30 cm2 ) (1 min ) (103 mL/L )
J m (1.03) s m 1820 L/m 2 h (1.03)
T −T 20 − 23.9
=J sp = = 2350.5 L/m 2 h bar

∆P 0.69 bar

4. Divide the specific flux (Jsp) by the initial specific flux (Jsp0). Results are in the sixth
column in the table. For the third row:
2350.5
=J sp' = 0.66
3560
5. Invert the normalized flux from column 6. Results are in the seventh column.
6. The results of the above calculations are shown in the table below for all rows of
data.
(1) (2) (3) (4) (5)

(6) (7)
Normalized Inverse normalized
Filtration Volume Specific Delta Specific specific flux,
specific flux,
time, filtered, throughput, volume, flux, '
min mL L/m2 mL L/m2·h J sp 1/ J sp'
0 0
1 108.80 36.3 108.80 2810.2 0.79 1.27
2 199.80 66.6 91.00 2350.5 0.66 1.51
3 277.40 92.5 77.60 2004.4 0.56 1.78
4 345.00 115.0 67.60 1746.1 0.49 2.04
5 404.20 134.7 59.20 1529.1 0.43 2.33
6 458.30 152.8 54.10 1397.4 0.39 2.55
7 506.80 168.9 48.50 1252.7 0.35 2.84
8 552.10 184.0 45.30 1170.1 0.33 3.04
9 594.10 198.0 42.00 1084.8 0.30 3.28
10 634.10 211.4 40.00 1033.2 0.29 3.45
11 670.80 223.6 36.70 947.9 0.27 3.76
7. Plot the inverse of the normalized specific flux (1/ J ´sp) as a function of the specific
throughput (Vsp), as shown in the following figure:

4.00
3.50
3.00
y = 0.013x + 0.607
2.50 R² = 0.987
1/Jsp
2.00
1.50
1.00
0.50
0.00
0 50 100 150 200 250
Specific throughput (L/m2)
The slope of the line is the membrane fouling index for the filter run is 0.013 L/m2 = 13
m-1.
Solution for Problem B
(1) (2) (3) (4) (5) (6) (7)

Filtration Specific Delta Specific specific flux,
specific flux,
time, TMP, throughput, volume, flux, '
d bar L/m2 m3 L/m2·h J sp 1/ J sp'
0 0.704 0 30,000 153.1
2 0.712 5172.4 30,000 151.3 0.99 1.011
4 0.721 10344.8 30,000 149.5 0.98 1.024
6 0.726 15517.2 30,000 148.4 0.97 1.031
8 0.735 20689.7 30,000 146.6 0.96 1.044
10 0.740 25862.1 30,000 145.6 0.95 1.051
12 0.747 31034.5 30,000 144.3 0.94 1.061
14 0.754 36206.9 30,000 142.9 0.93 1.071

16 0.765 41379.3 30,000 140.9 0.92 1.087
18 0.770 46551.7 30,000 139.9 0.91 1.094
20 0.777 51724.1 30,000 138.7 0.91 1.104
22 0.786 56896.6 30,000 137.0 0.90 1.117
24 0.794 62069.0 30,000 135.7 0.89 1.128
26 0.801 67241.4 30,000 134.5 0.88 1.138
28 0.812 72413.8 30,000 132.7 0.87 1.153
30 0.815 77586.2 30,000 132.2 0.86 1.158
1.18
1.16
1.14
1.12
1.10 y = 2.05E-06x + 1.00E+00

1/Jsp
R² = 9.98E-01
1.08
1.06
1.04
1.02
1.00
0 10000 20000 30000 40000 50000 60000 70000 80000 90000
Specific throughput (L/m2)
The slope of the line is the membrane fouling index is 2.05 × 10-6 L/m2 = 2.05 × 10-3 m-1.
Solution for Problem c
(1) (2) (3) (4) (5)

(6) (7)
Filtration Volume Specific Delta Specific specific flux,
specific flux,
time, filtered, throughput, volume, flux, '
d mL L/m 2
m 3 2
L/m ·h J sp 1/ J sp'

0 2276.64
2 2292.62 996.8 15.98 96.3 0.40 2.47
4 2308.41 1003.7 15.79 95.2 0.40 2.50
6 2324.05 1010.5 15.64 94.3 0.40 2.52
8 2339.53 1017.2 15.48 93.3 0.39 2.55
10 2354.92 1023.9 15.39 92.8 0.39 2.57
12 2370.17 1030.5 15.25 91.9 0.39 2.59
14 2385.31 1037.1 15.14 91.3 0.38 2.61
16 2400.33 1043.6 15.02 90.5 0.38 2.63
18 2415.24 1050.1 14.91 89.9 0.38 2.65
20 2430.04 1056.5 14.80 89.2 0.37 2.67
22 2444.76 1062.9 14.72 88.7 0.37 2.68
24 2459.35 1069.3 14.59 87.9 0.37 2.71
26 2473.88 1075.6 14.53 87.6 0.37 2.72
28 2488.26 1081.9 14.38 86.7 0.36 2.75
The slope of the line is the membrane fouling index is 3.102 × 10-3 L/m2 = 3.102 m-1.

PROBLEM 12-15
Instructor’s Notes: The solution is worked out for Problem A and the answers for the
remaining problems are included in a table at the end of the solution.
Problem Statement - A membrane filtration plant is to be designed using results from a

pilot study. Treatment plant requirements and pilot results are given in the table
below. For the selected system (to be specified by the instructor), determine (a)
the online production factor, (b) system recovery, (c) feed flowrate, (d) total
membrane area, (e) number of skids, and (f) number of modules per skid. The
pilot system contained 2 membrane elements that had 45 m2 of membrane area
each. In the full-scale plant, integrity testing will be required by regulations once
per day and will take 15 minutes. Chemical cleaning (CIP) will take 4 hours.
A B C D E
Design capacity (m3/d) 56,000 115,000 38,000 76,000 227,000

Membrane area in full-scale
45 55 45 45 80
modules (m2)
Max modules in skid 80 90 80 80 100
Pilot results
Flux (L/m2·h) 80 125 40 80 110
Backwash frequency (min) 30 25 25 22 30
Backwash duration (min) 1.5 0.5 1 2 1
Backwash volume (L) 270 100 200 240 240
Cleaning frequency (day) 45 30 60 30 30
Solution
1. Determine the fraction of time the system is producing permeate using Eq. 12-27:
 
1440 min/d
tbw (=
1.5 min )   72 min/d
 30 min 
tdit = 15 min/d
=tcip
(=
4 h )( 60 min/h )
8 min/d
30 d

1440 − tbw − tdit − tcip 1440 − 72 − 15 − 8 min/d
=η = = 0.934
1440 1440 min/d
2. Determine the system recovery. The system recovery is the same for one element
as for all elements and can be calculated using Eq. 12-26. For one element that
filters for 28.5 minutes per cycle (1.5 minutes out of every cycle is backwash), the
volume from Eq. 12-28 is,
= =
V f JAt
(80 L/m h )( 45.0 m ) ( 28.5 min
2 2
=
)
1710 L
f
60 min/h
270 L
=Vbw = 135 L
2 modules
V f − Vbw 1710 L − 135 L
=r = = 0.921
Vf 1710 L
3. Calculate required feed flow by solving Eq. 12-26 for Qf.

Q p 56, 000 m3 /d
=
Q f = = 60,804 m3 /d
r 0.921
4. Calculate the total membrane area required using Eq. 12-29.
=
Qf
A =
( 60,804 m /d )(10 L/m =
3 3
) 3
34, 385 m 2
J η (80 L/m h ) ( 24 h/d )( 0.921)
2
5. Calculate the total number of modules required.

area required 34,385 m 2
=
N MOD = = 764
surface area per module 45 m 2
6. Determine the number of skids and modules/skid. Since the skids can accommodate
up to 80 modules, at least 10 skids will be required. Dividing the required modules
evenly among skids is preferred. In addition, leaving space in the skids is
recommended as an inexpensive way to provide flexibility to reduce flux or increase
capacity by adding additional modules in the future.
N Racks = 10
764
N MOD /=
Rack = 77
10
The system will have 10 skids that each have 80 modules.

Solution to the other problem statements:
A B C D E
A. Online production factor 0.936 0.964 0.947 0.893 0.951
B. Recovery 0.921 0.978 0.861 0.9 0.95
C. Feed flow rate 60800 117559 44129 84444 238987
D. Required membrane area 33836 40649 48550 49245 95220
E. Number of skids 11 10 14 15 14
F. Modules/skid 69 74 78 73 86


CH 12

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CH 12

Uploaded by

Copyright:

Available Formats

MWH’S WATER TREATMENT

PRINCIPLES AND DESIGN

by Crittenden, Trussell, Hand, Howe, and Tchobanoglous

HOMEWORK SOLUTION MANUAL

Homework Solution Manual Page 2 of 29

Homework Solution Manual Page 3 of 29

Homework Solution Manual Page 4 of 29

Homework Solution Manual Page 5 of 29

2. Calculate permeate flow from 1 fiber.

QR = QF − QP = 23.4 m3 /h − 4.04 m3 /h = 19.36 m3 /h

v (1 m/s ) (103 L/m3 ) ( 3,600 s/h )

Homework Solution Manual Page 7 of 29

Flow (mL/min) 4.47 4.22 2.87 6.05 1.22

Pressure (bar) 0.67 0.80 0.71 1.25 0.21

For the data set selected by your professor,

a. Calculate the specific flux at 20º C.

b. Calculate the membrane resistance coefficient.

c. Does membrane resistance coefficient depend on the pressure and temperature

2. Calculate the specific flux using Eq. 12-9.

193.4 L/m2 hbar

Homework Solution Manual Page 8 of 29

Answers to all problems:

Membrane resist. coeff.

Membrane resist. coeff.

Homework Solution Manual Page 9 of 29

4. Calculate specific surface area

Instructors Note: The solution of this problem is similar to Example 12-4.

Problem Statement - A membrane plant is operated at a volumetric flux of 75 L/m2 · h

2. Calculate the specific flux using Eq. 12-9.

Homework Solution Manual Page 11 of 29

2. Calculate the specific flux using Eq. 12-9.

112.7 L/m2 hbar

4. Calculate the specific flux using Eq. 12-9.

112.3 L/m2 hbar

Homework Solution Manual Page 12 of 29

a. Which season will govern the size of the plant?

Homework Solution Manual Page 13 of 29

3. Compare the two values.

Homework Solution Manual Page 14 of 29

3. Calculate the cake thickness at 5 minutes using Eq. 12-13.

4. Calculate the cake resistance coefficient at 5 minutes using Eq. 12-12.

Homework Solution Manual Page 15 of 29

Homework Solution Manual Page 16 of 29

Homework Solution Manual Page 17 of 29

b. Rearrange the equation in step 1a to solve for DP as follows:

Homework Solution Manual Page 18 of 29

h. Rearrange the equation in step g3 to solve for DPc.

∆P 0.3 × 105 kg/ms2

4. Calculate the volume of water filtered in 40 minutes.

Homework Solution Manual Page 21 of 29

Solution for Problem A

=J sp = = 2350.5 L/m 2 h bar

Homework Solution Manual Page 22 of 29

(1) (2) (3) (4) (5)

Homework Solution Manual Page 23 of 29

Solution for Problem B

(1) (2) (3) (4) (5) (6) (7)

Homework Solution Manual Page 24 of 29

1.10 y = 2.05E-06x + 1.00E+00

Solution for Problem c

(1) (2) (3) (4) (5)

Homework Solution Manual Page 25 of 29

Homework Solution Manual Page 26 of 29