Water Treatment Processes

ENVR 890
Mark D. Sobsey
Spring, 2007

Water Sources and Water Treatment

Drinking water should be essentially free of disease-causing microbes,
but often this is not the case.
A large proportion of the worlds population drinks microbially contaminated water,
especially in developing countries

Using the best possible so

source
rce of water
ater for potable water
ater ssupply
ppl and
protecting it from microbial and chemical contamination is the goal
In many places an adequate supply of pristine water or water that can be protected
from contamination is not available

The burden of providing microbially safe drinking water supplies from

contaminated natural waters rests upon water treatment processes
The efficiency of removal or inactivation of enteric microbes and other pathogenic
microbes in specific water treatment processes has been determined for some
microbes but not others.
The ability of water treatment processes and systems to reduce waterborne
disease has been determined in epidemiological studies

Summary of Mainline Water Treatment Processes

Storage
Disinfection
Physical: UV radiation, heat, membrane filters
Chemical: Chlorine,
Chlorine ozone
ozone, chlorine dioxide
dioxide, iodine
iodine, other
antimicrobial chemicals

Filtration
Rapid granular media
Slow sand and other biological
g filters
Membrane filters: micro-, ultra-, nano- and reverse osmosis

Other pphysical-chemical
y
removal pprocesses
Chemical coagulation, precipitation and complexation
Adsorption:
p
e.g.,
g , activated carbon,, bone char,, etc,,
Ion exchange: synthetic ion exchange resins, zeolites, etc.

Water Treatment Processes: Storage

Reservoirs, aquifers & other systems:
store water
protect it from contamination
Factors influencing microbe reductions (site-specific)
detention
d t ti ti
time
temperature
microbial activity
water quality: particulates, dissolved solids, salinity
sunlight
sedimentation
land use
precipitation
runoff or infiltration

Water Storage and Microbial Reductions

Microbe levels reduced over time by natural
processes and microbial death/die-off
antimicrobial p
Human enteric viruses in surface water reduced 4001,000-fold when stored 6-7 months (The Netherlands)
Indicator bacteria reductions were less extensive,
probably due to recontamination by waterfowl.
Protozoan cyst reductions (log10) by storage were 1.6
for Cryptosporidium and 1.9 for Giardia after about 5
months
th (The
(Th Netherlands;
N th l d G.J
G J Medema,
M d
Ph.D.
Ph D di
diss.))
Recent ICR data indicates lower protozoan levels in
reservoir or lake sources than in river sources;
suggests declines in Giardia & Cryptosporidium by
storage

Typical Surface Water Treatment Plant

Chemical Coagulation-Flocculation
g
Removes suspended particulate and colloidal substances
from water
water, including microorganisms
microorganisms.
Coagulation: colloidal destabilization
Typically,
T i ll add
dd alum
l ((aluminum
l i
sulfate)
lf ) or fferric
i chloride
hl id
or sulfate to the water with rapid mixing and controlled
pH
H conditions
di i
Insoluble aluminum or ferric hydroxide and aluminum
or iron hydroxo complexes form
These complexes entrap and adsorb suspended
particulate and colloidal material.

Coagulation Flocculation Continued

Coagulation-Flocculation,
Flocculation:

Slow mixing (flocculation) that provides for for a period

off time to promote the aggregation and growth off the
insoluble particles (flocs).
The particles collide, stick together abd grow larger
The resulting large floc particles are subsequently
removed by gravity sedimentation (or direct filtration)
particles are too small to settle and are
Smaller floc p
removed by filtration

Microbe Reductions by Chemical CoagulationFlocculation

Considerable reductions of enteric microbe concentrations.

Reductions In laboratory and pilot scale field studies:
>99 percent using alum or ferric salts as coagulants
Some studies report much lower removal efficiencies (<90%)
Conflicting information may be related to process control
coagulant concentration, pH and mixing speed during
flocculation.
Expected microbe reductions bof 90-99%, if critical process
variables are adequately controlled
No microbe inactivation by alum or iron coagulation
Infectious microbes remain in the chemical floc
The floc removed by settling and/or filtration must be properly
managed to prevent pathogen exposure.
Recycling back through the plant is undesirable
Filter backwash must be disinfected/disposed of properly.

Cryptosporidium Removals by Coagulation

(Jar Test Studies)

C
Coagulant
l t

Dose
D
(mg/L)

O
Oocyst
t Removal,
R
l % (log
(l 10)

Alum

5
1

99.8 (2.7)
87 (0.9)

Iron

6
5

99.5 (2.3)
97 (1
(1.5)
5)

Granular Media Filtration

Used to remove suspended
p
particles
p
(turbidity)
(
y) incl. microbes.
Historically, two types of granular media filters:
Slow sand filters: uniform bed of sand;
low flow rate <0.1 GPM/ft2
biological process: 1-2 cm slime layer (schmutzdecke)
Rapid sand filters: 1, 2 or 3 layers of sand/other media;
>1 GPM/ft2
physical-chemical process; depth filtration

Diatomaceous earth filters

fossilized skeletons of diatoms (crystalline silicate);
powdery deposit; few 10s of micrometers; porous

Slow Sand Filters

Less widelyy used for large
g US municipal
p water supplies
pp
Effective; widely used in Europe; small water supplies;
developing
p g countries
Filter through a 3- to 5-foot deep bed of unstratified sand
flow rate ~00.05
05 gallons per minute per square foot
foot.
Biological growth develops in the upper surface of the sand is
primarily responsible for particle and microbe removal.
removal
Effective without pretreatment of the water by
coagulation-flocculation
Periodically clean by removing, cleaning and replacing the
upper few inches of biologically active sand

Microbial Reductions byy Slow Sand Filtration

Effective in removing enteric microbes from water.
Virus removals >99% in lab models of slow sand filters.
Up to 4 log10; no infectious viruses recovered from filter effluents

Field studies:
naturallyy occurring
g enteric viruses removals
97 to >99.8 percent; average 98% overall;
Comparable removals of E.
E coli bacteria.
bacteria
Virus removals=99-99.9%;
high bacteria removals (UK study)
Parasite removals: Giardia lamblia cysts effectively removed
Expected
E
t d removals
l  99%

Roughing Filter
Used in developing
countries
inexpensive
low
low
maintenance
local materials
Remove large solids
Remove microbes
1-2 log10
bacterial
reduction
d i
90% turbidity
reduction

Microbe Reductions by Rapid Granular Media Filters

Ineffective to remove enteric microbes unless preceded by
chemical coagulation-flocculation.
g
Preceded chemical coagulation-flocculation & sedimentation
Enteric microbe removals of 90->99 % achieved.
Field (pilot) studies: rapid sand filtration preceded by iron
coagulation-flocculation: virus removal <50% (poor control?).
Giardia
Gi di llamblia:
bli removals
l not always
l
hi
high;
h related
l d to turbidity
bidi
removal; >99% removals reported when optimized.
Removal not high unless turbidity is reduced to 0.2
0 2
NTU.
Lowest removals shortly after filter backwashing
Microbes primarily removed in filter by entrapped floc
particles.
Overall, can achieve 90% microbial removals from water
when preceded by chemical coagulation-flocculation.

Microbe Reductions by Chemical Coagulation-Flocculation and

Filtration of River Water by Three Rx Plants in The Netherlands
Organisms

Plant 1

Plant 2

Plant 3

Log10 Reductions of Microbes

Enteric
1.0
1.7
>2
Viruses
F+
No data
0.4
1.7
Coliphages
Fecal
2.0
>2
0.2
Coliforms
F
Fecal
l
21
2.1
>2
2
06
0.6
Streptococci
Clostridium 0.6
2.1
>2
spores
Plant 1 used two stages of iron coagulation-flocculation-sedimentation.
Plant 2 used iron coagulation-flocculation-sedimentation
coagulation flocculation sedimentation and rapid filtration
Plant 3 used iron coagulation-flotation-rapid filtration.

Cryptosporidium Removals by Sand

Filtration

Type

Reduction
Rate (M/hr) Coagulation % (log10)

Rapid, shallow
Rapid
Rapid, shallow

5
5

No
Yes

65
90

(0
(0.5)
5)
(1.0)

Rapid, deep

Yes

99.999 (5.0)

Slow

02
0.2

No

99 8
99.8

(2
(2.7)
7)

Cryptosporidium Removal by Coagulation and Direct

Filtration

Run No.

Log10 Reduction of
Cryptosporidium

Turbidity

1
2
3
4

3.1
28
2.8
2.7
1.5

1.3
12
1.2
0.7
0.2*

Mean

2.5

0.85

Raw water
R
t tturbidity
bidit = 0.0
0 0 - 5.0
5 0 NTU
Alum coagulation-flocculation;
Anthracite-sand-sand filtration; 5 GPM/ft2
*Suboptimum alum dose
Ongerth & Pecoraro. JAWWA, Dec., 1995

Reported Removals of Cryptosporidium

Oocysts by Physical
Physical-Chemical
Chemical Water
Treatment Processes (Bench, Pilot and
Field Studies)
Process

Log10 Reduction

Clarification by:
Coagulation flocculation
flocculation-sedimentation
sedimentation
or Flotation

<1
1 - 2.6
26

Rapid Filtration (pre-coagulated)

1.5 - >4.0

Both Processes

<2.5 - >6.6

Slow Sand Filtration

>3.7

Diatomaceous Earth Filtration

>4.0

Coagulation + Microfiltration

>6.0
6.0

Ultrafiltration

>6.0

Cryptosporidium Reductions by Coagulation

and Filtration
Laboratory studies on oocyst removal:
- Jar test coagulation with 1 hr. setting =
2.0 - 2.7 log
g10
- Sand filtration, no coagulant, 10 cm bed
depth = 0
0.45
45 log10
- Sand filtration, plus coagulation, 10 cm bed
d th = 1.0
depth
1 0 log
l 10
Gregory et al.,
al 1991.
1991 Final Report.
Report Dept.
Dept of the Environ.,
Environ UK

Membrane Filters
More recent development
p
and use in drinking
g water
Microfilters: several tenths of M to M diameter pore size
nano- & ultra-filters: retention byy molecular weight
g cutoff
Typically 1,000-100,000 MWCO
Reverse osmosis filters: p
pore size small enough
g to remove
dissolved salts; used to desalinate (desalt) water as well as
particle removal
High >99.99% removal of cellular microbes
Virus removals high >9.99% in ultra-, nano- and RO filters
Virus removals lower (99%) by microfilters
Membrane and membrane seal integrity
g y critical to effective
performance

Cryptosporidium Reductions by Membrane

Filtration
Log10
Membrane, Pore Size Cryptosporidium
Reduction
Type
A, MF
0.2 m
>4.4
B MF
B,
0 2 m
0.2
>4 4
>4.4
C, MF
0.1 m
4.2->4.8
D UF
D,
500 KD
>4 8
>4.8
E, UF
300 KD
>4.8
F UF
F,
100 KD
>4.4
44
MF = microfilter filter; UF = ultrafilter
Jacangelo et al., JAWWA, Sept., 1995

Adsorbers and Filter-Adsorbers

Adsorbers:
Granular
G
activated carbon adsorption

remove dissolved organics

poor retention of pathogens, esp. viruses
biologically active; develops a biofilm
can shed microbes into water

Filter-adsorbers
Sand plus granular activated carbon
reduces particles and organics
biologically active
microbial
i bi l retention
t ti is
i possible
ibl

Cryptosporidium
C
t
idi
Removals
R
l by
b Sand
S d
Filtration

Type

Rate (M/hr)

Coagulation

Reduction
% (log10)

R id shallow
Rapid,
h ll

N
No

65

(0
(0.5)
5)

Rapid, shallow

Yes

90

(1.0)

Rapid, deep

Yes

99.999 (5.0)

02
0.2

No

99 8
99.8

Slow

(2
(2.7)
7)

Cryptosporidium Reductions by Membrane

Filtration
Membrane,
Type
yp

A, MF
B MF
B,
C, MF
D UF
D,
E, UF
F UF
F,

Pore Size

0.2 m
0 2 m
0.2
0.1 m
500 KD
300 KD
100 KD

Log10 Cryptosporidium Reduction

>4.4
>4 4
>4.4
4.2->4.8
>4 8
>4.8
>4.8
>4.4
44

MF = microfilter filter; UF = ultrafilter

Jacangelo et al., JAWWA, Sept., 1995

W t Softening
Water
S ft i and
d Microbe
Mi
b Reductions
R d ti
Hard" Water: contains excessive amounts of calcium
and magnesium ions
iron and manganese can also contribute to hardness.
hardness

Hardness ions are removed by adding lime (CaO) and

sometimes soda ash ((Na2CO3) to precipitate
p
p
them as
carbonates, hydroxides and oxides.
This process, called softening, is basically a type of
coagulation-flocculation process.
Microbe reductions similar to alum and iron
coagulation when pH is <10
Microbe reductions >99.99% possible when pH is >11
microbial inactivation + physical removal

Microbial Reductions by Softening Treatment

Softening with lime only (straight lime softening); moderate
high pH
ineffective
i ff ti enteric
t i microbe
i b reductions:
d ti
about
b t 75%.
75%

Lime-soda ash softening

results in the removal of magnesium as well as calcium
hardness at higher pH levels (pH >11)
enteric microbe reductions >99%.
Lime-soda
Li
d ash
h softening
ft i att pH
H 10
10.4,
4 10
10.88 and
d 11.2
11 2 has
h produced
d d
virus reductions of 99.6, 99.9 and 99.993 percent, respectively.

At lower pH levels (pH <11),

11), microbe removal is mainly a
physical process
infectious microbes accumulate in the floc particles and the
resulting
lti chemical
h i l sludge.
l d

At pH levels above 11, enteric microbes are physically

removed and infectivity is also destroyed
more rapid and extensive microbe inactivation at higher pH
levels.

Disinfection of Microbes in Water:

Conventional Methods used in the Developed
p World
Historically, the essential barrier to prevention and control of waterborne
microbial transmission and waterborne disease.
disease
Free chlorine: HOCl (hypochlorous) acid and OCl- (hypochlorite ion)
HOCl at lower ppH and OCl- at higher
g ppH; HOCl a more ppotent ggermicide than OCl strong oxidant and relatively stable in water (provides a disinfectant residual)

Chloramines: mostly NH3Cl: weak oxidant; provides a stable residual

ozone, O3 , strong oxidant; provides no residual (too volatile and
reactive)
Chlorine dioxide,
dioxide ClO2,, string oxidant but not very stable residual
Concerns due to health risks of chemical disinfectants and their
by-products
by
products (DBPs)
(DBPs), especially free chlorine and its DBPs
UV radiation
low pressure mercury lamp: low intensity; monochromatic at 254 nm
medium pressure mercury lamp: higher intensity; polychromatic 220-280 nm)
reacts primarily with nucleic acids: pyrimidine dimers and other alterations

Disinfection Kinetics
Disinfection is a kinetic process
Increased inactivation with increased exposure or contact time.
Chick's Law: disinfection is a first-order reaction. (NOT!)
Multihit-hit or concave up kinetics: initial slow rate; multiple targets to be
hit
hit
Concave down or retardant kinetics: initial fast rate; decreases over time
Different susceptibilities
p
of microbes to inactivation;; heterogeneous
g
population
Decline of of disinfectant concentration over time
CT Concept:
C
t Disinfection
Di i f ti can be
b expressedd att the
th product
d t off disinfectant
di i f t t
concentration X contact time
Applies best when disinfection kinetics are first order
Disinfectant concentration and contact time have an equal effect on
CT products
Applies less well when either time ofrconcentration is more important.

DISINFECTION AND MICROBIAL INACTIVATION KINETICS

Log Survvivors

First
Order

Multihit

Retardant

Contact Time

Factors Influencing Disinfection of Microbes

Microbe type: disinfection resistance from least to most:
vegetative bacteria viruses protozoan cysts
cysts, spores and eggs
Type of disinfectant: order of efficacy against Giardia from best to worst
O3 ClO2 iodine/free chlorine chloramines
BUT, order of effectiveness varies with type of microbe

Microbial aggregation:
protects microbes from inactivation
microbes within aggregates can not be readily reached by the disinfectant

Effects of Water Quality on Disinfection

Particulates: protect microbes from inactivation
microbes shielded or embedded in particles

Dissolved organics: protects

consumes or absorbs (UV radiation) disinfectant
coats microbes

Inorganic compounds and ions: effects vary with disinfectant

pH: effects depend on disinfectant.
Free chlorine more biocidal at low pH where HOCl predominates.
Chlorine dioxide more microbiocidal at high pH

Reactor design, mixing and hydraulic conditions; better activity in

"plug flow" than in "batch-mixed" reactors.

Inactivation of Cryptosporidium Oocysts in

Water by Chemical Disinfectants
Disinfectant

CT99 (mg-min/L)

Reference

Free Chlorine

7,200+

Korich et al., 1990

Monochloramine

7,200+

Korich et al., 1990

Chlorine Dioxide

>78

Korich et al., 1990

Mixed oxidants

<120
120

Venczel et al
al., 1997

Ozone

~3-18

Finch et al., 1994

K i h ett al.,
Korich
l 1990
Owens et al., 1994

C. parvum oocysts inactivated by low doses of UV radiation: <10 mJoules/cm2

Disinfection:
A Key Barrier Against Microbes in Water
Free chlorine still the most commonly used disinfectant
Maintaining disinfectant residual during treated water storage
and distribution is essential.
A problem for O3 and ClO2, which do not remain in water for very long.
A secondary disinfectant must be used to provide a stable residual

UV radiation is a promising disinfectant because it inactivates

Cryptosporidium at low doses
UV may have to be used with a chemical disinfectant to protect the
water with a residual through distribution and storage