Water and Wastewater Engineering

Подождите немного. Документ загружается.

14-20 WATER AND WASTEWATER ENGINEERING

14-9 PHARMACEUTICALS AND ENDOCRINE-DISRUPTING

CHEMICALS (EDCs)

Pharmaceutical compounds, such as sulfamethoxazole, carbamazepine, and ibuprofen, and

endocrine-disrupting compounds, such as 17b-estradiol, testosterone, and bisphenol-A, have

been identified in the surface waters of the United States. Figure 14-6 shows the most frequently

detected pharmaceuticals and EDCs. Typical concentrations are low. For example, in surface

water supplies they are generally less than 1  g/L (Hullman, 2009).

Although research on removal of pharmaceuticals and EDCs is in its infancy, some early

results are useful in pointing to potential technologies. For a selected set of pharmaceuticals (bezaf-

ibrate, clofibric acid, carbamazepine, and diclofenac), Ternes et al. (2002) reported the following:

• Conventional sand filtration and coagulation with ferric chloride provide no significant

removal,

• Ozonation was effective for some compounds, and

• Ozonation followed by granular ac tivated carbon filtration was very effective for all

compounds investigated but clofibric acid.

For selected sets of pharmaceuticals (sulfamethoxazole, carbamazepine, diclofenac, and

ibuprofen) and EDCs (estradiol, es

trone, testosterone, and progesterone), several authors

Coprostanol(5)

Cholesterol(5)

N-N-diethyltoluamide(4)

Tri(2-chloroethyl) phosphate(4)

4-nonylphenol(4)

4-nonylphenol monoethoxylate(4)

4-nonylphenol diethoxylate(4)

5-methyl-1H-benzotriazole(4)

1,4-dichlorobenzene(4)

Fluoranthene(4)

Pyrene(4)

1,7-dimethylxanthine(3)

Ethanol, 2-butoxy-phosphate(4)

4-octylphenol monoethoxylate(4)

4-octylphenol diethoxylate(4)

Bis

phenol-A(4)

Cotinine(3)

Diazinon(4)

4-methyl phenol(4)

Acetaminophen(3)

Tetrachloroethylene(4)

Trimethoprim(3)

Erythromycin-H2O(1)

Estriol(5)

Lincomycin(1)

Sulfamethoxazole(3)

Phthalic anyhydride(4)

Triclosan(4)

Caffeine(4)

Frequency of detection (%)

FIGURE 14-6

Most frequently detected pharmaceuticals and endocrine-disrupting compounds. ( Source: U.S.G.S., 2002.)

REMOVAL OF SPECIFIC CONSTITUENTS 14-21

(Drewes et al., 2005; Nghiem et al., 2004 and 2005) indicate NF and RO are highly effective in

removing these compounds. Ozonation also appears to be highly effective for the destruction of

these compounds (Westerhoff et al., 2005). UV irradiation of pharmaceuticals appears to be only

effective in the presence of hydrogen peroxide (Rosenfeldt and Lin

den, 2004).

14-10 RADIONUCLIDES

The most common radionuclides of concern in natural waters are radium-226 (

226

Ra) radium-

228 (

228

Ra), radon-222 (

222

Rn), and uranium-234 (

234

U). In groundwater,

226

Ra and

228

Ra exist

primarily as divalent cations . Radon-222 is a volatile gas. Uranium-234 exists as uranyl ion. It

readily complexes with carbonate and hydroxide. The revised Radionuclide Rule outlining the

concentration limits for these compounds was publi

shed on December 8, 2003.

The U.S. EPA has designated the following as best available technologies (BAT) for removal

of the radionuclide shown (U.S. EPA, 2005):

• Radon-222: air stripping.

• Radium-226 and radium-228 together: coagulation/flocculation/sedimentation/filtration.

• Radium-226 and radium-228 separately: ion exchange, RO or lime-soda softening.

• Uraniu

m-234: ion exchange; coagulation/flocculation/sedimentation/filtration.

Table 14-7 summarizes the performance of specific tec hnologies for specific radionuclides.

Sludge and resin disposal alternatives are discussed in Chapter 15.

TABLE 14-7

Performance of specific technologies for specific radionuclides

Removal efficiency, %

Technology Radon Radium Uranium

Activated alumina up to 99

Aeration, diffused bubble to 99 

Aeration, spray 70–95 

Air stripping, packed

tower

to 99 

Coagulation-filtration 85–95

GAC adsorption 62–99 

Greensand

19 to 82

Anion exchange up to 95

Cation ion exchange 81–97

Lime-soda softening 75–90

16 to 97

Reverse osmosis 90–95  90

19 to 63% for radium-226 and 23 to 82% for radium-228.

pH of at least 10.6.

A dapted from Lowery and Lowery, 1988; U.S EPA, 2005.

14-22 WATER AND WASTEWATER ENGINEERING

14-11 SYNTHETIC ORGANIC CHEMICALS (SOCs) AND VOLATILE

ORGANIC CHEMICALS (VOCs)

Synthetic organic chemicals include pesticides, polycyclic aromatic hydrocarbons, and polychlo-

rinated biphenyls. SOC is considered a regulatory term rather than a chemical description. Some

of the SOCs are volatile organic chemicals. The term volatile organic chemical (VOC) refers to

a compound with a high vapor pressure that causes it to evaporate readily. In water treatm ent

practice, they are classified into three broad groups:

• Those found in petroleum products (e.g., benzene, toluene, and xylene).

• Halogenated compounds used as solvents, degreasers, and intermediates

(e.g., tetrachloro-

ethylene and methylene chloride).

• Disinfection byproducts like the trihalomethanes.

With the exception of the THMs, these compounds enter the water from industrial waste dis-

charge, leaking fuel storage tanks, and uncontrolled waste disposal sites.

Treatment Strategies

The U.S. EPA designated air stripping and granular activated c arbon (GAC) as best available

technology (BAT) for treatment of VOCs except for vinyl chloride and methylene chloride. For

these two chemicals, only air stripping is recognized as BAT. As a rule of thumb, chemicals hav-

ing a dimensionless Henry’

s law coefficient  7.5  10



at 20  C can be removed by packed

tower stripping (54 FR 22062). Air stripping is more economical than GAC for removal of VOCs

if the off-gas can be directly discharged without treatment (Ball and Edwards, 1992; Gross and

TerMaath, 1985; Hand et al., 1986). The economic advantage of air stripping over GAC dimin-

ishes when off-gas treatment is requ

ired and other strategies should be investigated (Ball and

Edwards, 1992).

For SOCs that cannot be stripped, GAC is the BAT technology.

Stripping. Although there are a number of air stripping systems available, packed towers

( Figure 14-7 ) serve as a primary method for removal of VOCs. The relevant equations are

( LaGrega et al., 2001):

Z ()()HTU NTU (14-13)

HTU 

AKa

()( )

(14-14)

NTU 



S

CC S

⎛

⎝

⎞

⎠

⎡

⎣

⎢

⎤

⎦

⎥

in eff

()()

(14-15)



()( )

(14-16)

where Z  packing height, m

HTU  height of transfer unit, m

Q  flow rate of water, m

/ s

REMOVAL OF SPECIFIC CONSTITUENTS 14-23

A  cross-sectional area of column, m

 overall transfer rate constant, s

1

NTU  number of transfer units

S  stripping factor

 influent concentration, mg/L

eff

 effluent concentration, mg/L

H  Henry’s constant, dimensionless

 flow rate of air, m

/ s

The overall transfer rate constant ( K

) is a complex function of the diffusion coefficients,

liquid mass loading, liquid viscosity, and packing size. The Sherwood and Holloway equation and

the Onda correlations (see LaGrega et al., 2001) are two techniques for estimating K

. However,

for design purposes, K

should be determined experimentally. A factor of safety is required for

both the pilot scale data and the estimating techniques because they overestimate K

. They are

between 33 percent and 47 percent high (Djebbar and Narbaitz, 1995; MWH, 2005).

Dimensionless Henry’s constants for selected compounds are given in Appendix A.

Air-to-water ratios ( Q

/ Q ) range from 5:1 to 300:1 (Kavanaugh and Trussel, 1980; LaGrega et al.,

2001). For chemicals with dimensionless Henry’s law coefficients from 0.003 to 0.3, the air-to-water

ratio that provides the minimum tower volume and power requirement is approximately 3.5 times the

minimum air-to-water ratio needed to meet a treat

ment objective concentration of C

eff

(Hand et al.,

1986; MWH, 2005). The minimum air-to-water ratio is calculated with the following equation:





in eff

()( )

(14-17)

where Q

/ Q  air-to-water ratio

 influent liquid concentration, mg/L

eff

 effluent liquid concentration, mg/L

H  Henry’s law coefficient, dimensionless

Liquid in

Gas in

Liquid out

Shell

Packing suppor

Liquid

redistributor

Packing

restrainer

Liquid

distributor

Random

packing

Gas out

FIGURE 14-7

Packed tower stripping column.

14-24 WATER AND WASTEWATER ENGINEERING

For VOC stripping, the ratio should be selected to achieve the desired concentration of VOC in

the effluent while maintaining a height-to-diameter ratio greater than 1:1. In general, it is more

economical to provide higher air-to-water ratio because it lowers the tower height. The trade-off

is between the operating cost for the blower to provide air and the capital cost for a taller tower.

The pressure drop per unit of tower height (



P / Z ) in the packed tower is a function of

the superficial gas velocity and the friction factor for dry packing ( F

). The proportionality

e xpression is:





()()

(14-18)

where



P / Z  pressure drop per unit length of packed column, Pa/m

 friction factor or, more commonly, packing factor, dimensionless

 superficial gas velocity, m/s

The packing factor i s specific to the shape and size of the packing. A few examples are shown in

Table 14-8 .

The cross-sectional area of the column (A) is estimated using the Eckert pressure drop curves

shown in Figure 14-8 (Eckert, 1961).

The estimating procedure is

1 . Specify the water temperature, packing factor ( F

), air-to-water ratio ( Q

/ Q ), and gas

pressure drop (



P / Z ).

2. Compute the value of the ratio G

/ L

using the following expression:



⎛

⎝

⎜

⎞

⎠

⎟

⎛

⎝

⎜

⎞

⎠

⎟

(14-19)

TABLE 14-8

Selected packing factors and specific surface area

Type Nominal diameter, mm

Packing factor, F

Plastic saddles

50 20.0

75 16.0

Plastic tripacks

5015.0

90 14.0

Flexring

50 24.0

90 20.0

IMPAC

d TM

85 15.0

140 6.0

LANPAC

d TM

60 21.0

90 14.0

Norton Co., Akron, OH.

Jaeger Co., Houston, TX.

Ceilcote Co., Cleveland, OH.

LANTEC Co., Los Angeles, CA.

REMOVAL OF SPECIFIC CONSTITUENTS 14-25

where G

 air mass loading rate, kg/m

· s

 water mass loading rate, kg/m

· s



 air density, kg/m



 water density, kg/m

3. Compute the value of x on the x -axis of the Eckert curve using

G L



0 5

(

)

⎡

⎣

⎢

⎤

⎦

⎥

⎡

⎣

⎢

⎤

⎦

⎥

–

(14-20)

4. Determine y on Figure 14-8 using the computed value of x and the specified pressure drop.

5. With the value of y, solve the following expression for G



0.5

()( )(r

01.

()()m

⎡

⎣

⎢

⎤

⎦

⎥

–

)

(14-21)

where y  numerical value on y-axis found in step 4

 packing factor, dimensionless



 dynamic viscosity of water, Pa · s

0.4

0.2

0.10

0.08

0.06

0.04

0.02

0.01

0.008

0.006

0.004

0.002

0.001

0.01 0.02 0.04 0.1 0.2 0.4 1.0 2 4 10

p

/ft

N/m

 6.37  10

–3

 1.224  10

–3

inH

Approximate

flooding

Gas pressure drop

N/m

1200

800

400

200

100

1/2

)(

␮

–r

)

0.1

 r

FIGURE 14-8

Flooding and pressure drop in randomly packed towers.

( Source: Treybal, R. E. 1980. Mass Transfer Operations. New York: Chem. Engrg. Series, McGraw-Hill, 3rd ed.)

Reproduced with permission of The McGraw-Hill Companies.

14-26 WATER AND WASTEWATER ENGINEERING

6. Determine the water mass loading rate.

agw



()()//rr

(14-22)

7. Determine the cross-sectional area of the packed tower.



()( )r

(14-23)

T ypical design ranges are given in Table 14-9 .

Prefabricated tower dimensions are limited by highway transportation clearances . Multiple

towers in series are used to meet height requirements when tower heights exceed 9 m. Multi-

ple towers in parallel are used when tower diameters exceed 3.6 m. The tower height is

greater

than the height of the packing. Approximately 1 to 2 m may be added to the packing height for

support structures and internal distribution piping. Prefabricated units are generally built in 0.3 m

increments.

Potential operating problems arise when the water contains appreciable quantities of iron,

manganese, and/or hardness. The aeration will oxidize the iron/manganese, which will then pre-

ipitate on the packing and foul it. Preoxidation and filtration may be used to avoid this problem.

Because CO

i s being stripped, the pH of the water will rise. This may lead to calcium carbonate

precipitation in the packing. Using a larger packing diameter ( 50 mm), careful monitoring, and

periodic cleaning of the packing may be required. A design with capability to periodically acid

wash the packing has also been used

when high iron/manganese/calcium was encountered (Ball

and Edwards, 1992).

Example 14-2. Springfield has a well field that is contaminated with trichloroethylene (TCE).

The design flow rate is 8,200 m

/ d. A packed tower has been selected to remove the TCE. A

treatment objective of 95 percent removal of TCE has been selec ted. Design a packed tower to

strip TCE from the water. The following design data have been provided:

TABLE 14-9

Typical packed column design ranges

Parameter ValueComment

Tower height 1.5 to 9 m Prefab will be sized to fit on flat

bed trailer

Diameter 0.3 to 3.6 m Restriction to 3.6 m for transport

of prefab units

Height:diameter

 1:1

Without liquid redistribution use

 4:1 for proper liquid distribution

Pressure drop 50 to 100 Pa/m of packing Economics favor 50 Pa/m

5:1 to 300:1

Ratio of diameter to packing size 8:1 to 15:1

 15:1 preferred

Sources: Dyksen, 2005; Hand et al., 1999; LaGrega et al., 2001; MWH, 2005; Treybal, 1980.

REMOVAL OF SPECIFIC CONSTITUENTS 14-27

Raw water TCE  72.0  g/L

T e mperature  10  C

 0.0128 s

1

P a cking  plastic tripack with diameter of 50 mm



P / Z  50 Pa/m



 1.2 kg/m



 1,000 kg/m

H at 10  C  0.116

Determine the following to complete the design:

/ Q

Column diameter

Stripping factor

Height of packing

Overall height of packed tower

Solution:

a . Determine the minimum air to water ratio ( Q

/ Q ) from Equation 14-17. At 95 percent

removal, the effluent concentration of TCE is 3.60  g/L.



 





72 0 3 60

0 116 72 0

g/L g/L

g/L()( )

..19

To achieve the air-to-water ratio that provid es the minimum tower volume and power

requirement, multiply the minimum air-to-water ratio by 3.5.

()()819 35 28 66 30.. .or

b. Determine the diameter of the column

(1) Compute the value of the ratio G

/ L

()30

1 000

00360

kg/m

⎛

⎝

⎜

⎞

⎠

⎟

(2) Compute the value of x on the Eckert curve using Equation 14-20.

x 



00360

1 000 1 2

0 5

⎡

⎣

⎢

⎤

⎦

⎥

kg/m

kkg/m

096

⎡

⎣

⎢

⎤

⎦

⎥

 .

(3) Find y on Figure 14-8 using x  0.96 and the



P / Z  50 Pa/m curve:

y  00039.

(4) With the value of y, solve Equation 14-21 for G

. The packing factor of 15.0 for

50 mm tripacks is found in Table 14-8 .

14-28 WATER AND WASTEWATER ENGINEERING





0 5

00039 1 2 1 000 1 2

.. , .()( )(kg/m kg/m kgg/m

kg/

3 01

15 01307 10

0 778

)

()( )..







⎡

⎣

⎢

⎤

⎦

⎥

mms



(5) Determine the water mass loading rate.



0.778 kg/m ·s

(30)(1.2 kg / m /1,000 kg/m

)

21.61kg/ ms

(6) Determine the cross-sectional area of the packed tower.

A 



()( )

(

8 200 1 000

21 61

m /d kg/m

kg/ms

))( )86 400

4 39

s/d

m

(7) Determine the diameter of the column.

D 

0 5

4 394

2 3624

()()m

or m



⎛

⎝

⎜

⎞

⎠

⎟

c. The height of a transfer unit (HTU) is a function of the area of the column. Using the

area of 4.39 m

HTU

m /d





8 200

4 39 0 0128 86 400

.. ,()( )(

ss/d

)

169.

d. With the Henry’s law constant of 0.116, the stripping factor is

S ()()0 116 30 3 48..

e. The number of transfer units (NTU) is

NTU ln

g/L/ g/L





3 48

3 48 1

72 0 3 60.

⎛

⎝

⎞

⎠

()(

3348 1 1

3 48

3 76





) +

⎡

⎣

⎢

⎤

⎦

⎥

f. The height of the packing ( Z ) is then

Z ()()169 3 76 6 35...mm

The height to diameter ratio is greater than 1:1 so this design is acceptable.

g. The estimated overall height of the tower will be 6.35 m  1.0 m for distribution piping,

etc.  7.35 m. Because prefabricated units are

usually constructed in 0.3 m increments,

the actual final dimensions will be about 2.4 m in diameter and 7.5 m in height.

Comments:

1 . The overall tower height meets the  9 m criterion for prefab units. If it d id not, alter-

native designs would be considered including increasing the air-to-water ratio and

/or

splitting the tower into two sections and using them in series.

2. The K

for this problem was selected arbitrarily. It is not based on experimental data for

the water or packing specified and should NOT be used for actual design.

REMOVAL OF SPECIFIC CONSTITUENTS 14-29

3. A spreadsheet was used to perform the computations and explore several variables. Other

solutions provide acceptable answers.

Granular Activated Carbon (GAC). The removal of a chemical from solution by activated

carbon is a mass-transfer proc ess in which the chemical is bonded to the s

olid. This process is

called adsorption. The chemical (the adsorbate ) penetrates into the pores of the solid (the adsor-

bent ), but not into the lattice itself. The bond may be physical or chemical. Electrostatic forces

hold the chemical when physical binding is predominant. Chemical bonding is b

y reaction with

the surface. Activated carbon (the adsorbate) is made from various materials such as wood, coco-

nut shells, coal, and lignite. The manufacturing process is essentially a carbonization of the solid

followed by activation using hot air or steam. Like ion exchange resins, the num

ber of active sites

is finite and the carbon becomes saturated with use over time. It is regenerated by heating with

hot air or steam.

Adsorption isotherms are used to select one of the manufactured activated carbons for remov-

ing the SOCs of concern in the water supply. The isotherms are prepared from experim

ental data.

The selected activated carbon is placed in a solution containing the chemical or chemicals of

interest. The solution is agitated to provide adequate contact between the granules of carbon and

the chemical. The slurry is mixed until equilibrium is a

chieved. In general, this will take about

one to four hours. The initial concentration will decrease to an equilibrium value. By employing

a series of slurry tests, a plot can be made that describes the relationship between the equilibrium

concentration and the mass of SOC ( x ) adsorbed per unit mass of activated carbon (

m ). Because

the adsorption phenomenon is very temperature and pH dependent, the experimental temperature

is controlled, that is, the experiment is isothermal, and the pH must be the same as that used in the

full-scale treatment process. As a consequence, the data are only relevant for the temperature and

pH at which the experiment is conducted.

Freundlich (1906) developed an empirical equation that is used to describe the results of the

adsorption isotherm experiment. A form of the equation is

()

(14-24)

where q

 mass of solute adsorbed per mass of activated carbon, mg/g

K  Freundlich adsorption capacity parameter, (mg/g)(L/mg)

1/ n

 equilibrium concentration of solute, mg/L

1/ n  Freundlich adsorption intensity parameter, dimensionless

The linear form of the Freundlich equation is

log log log()qK



⎛

⎝

⎞

⎠

(14-25)

A log-log plot will yield a straight line with a slope of 1/ n. With C

equal to 1.0, K  q

EPA has published isotherm data for individual chemicals that may be used for preliminary

feasibility studies (Dobbs and Cohen, 1980). For actual selection of a manufactured carbon, iso-

therm experiments with the actu al raw water are necessary be

cause there are usually multiple

chemicals that compete for adsorption sites and many other constituents that are not SOCs will

adsorb.