CHAPTER 16

WATER TREATMENT PLANT
RESIDUALS MANAGEMENT
David A. Cornwell, Ph.D., P.E.
President
Environmental Engineering & Technology, Inc.

Water treatment plants typically produce some type of waste stream. The quality
and characteristics of these waste streams are related to the main treatment process.
Furthermore, waste streams can impact the finished water quality of the treatment
process itself. This is especially true when the waste is stored internal to the process
or recycled.
Despite the strong linkage between the treatment process and its waste streams,
however, water treatment plant waste management has historically been treated as
a stand-alone management issue. Whatever the treatment process produced was
dealt with in a technically appropriate manner. With increasing costs associated with
managing waste streams, it has become prudent to consider the waste stream quality
and characteristics as part of the overall evaluation and design of the main water
treatment process.Water treatment processes produce unique waste streams, each of
which has different associated waste handling costs. The waste streams must be
viewed as part of the overall process to be optimized when determining the most
economical method for meeting a specific set of finished water quality goals. As an
example, it is now recognized that storing solids in a sedimentation basin is not desirable from a water quality perspective. Including solids storage considerations in the
overall design of the water treatment process will then influence the decision
regarding the type of sedimentation basin to install. Similarly, some filter media
combinations produce more spent filter backwash waste than do others. In determining the main treatment process components, waste streams should be considered
in the overall decision tree, not viewed as an issue that is handled separately.
It is interesting to note the goals of waste treatment as described in the last edition of this book (1990)—“What must be removed? Where will it be disposed? What
treatment is necessary to prepare it for disposal?” Recently, however, a new set of
issues other than just disposal of the waste has become important. Instead of disposal, the first approach to end use is now beneficial utilization and solids treatment.

Systems are often geared to preparing a material that can be used. Minimization of
the liquid volumes of waste produced is also increasingly important. Water quality
issues associated with storing residuals in the process train or associated with recy16.1


16.2

CHAPTER SIXTEEN

cling the water back to the treatment plant have become important in planning a
waste management system.
The original term used to describe all water treatment plant wastes was sludge. In
fact, sludge is really only the solid or liquid-solid component of some types of waste
streams. The term residuals is now used to describe all water treatment plant process
wastes, either liquid, solid, or gaseous.
Hydrolyzing metal salts or synthetic organic polymers are added in the water
treatment process to coagulate suspended and dissolved contaminants and yield relatively clean water suitable for filtration. Most of these coagulants and the impurities they remove settle to the bottom of the settling basin where they become part of
the waste stream. These residuals are referred to as alum, iron, or polymeric sludge
(even though they may be made up largely of water), being named after the primary
coagulant used. These residuals account for approximately 70 percent of the water
plant solids generated. Similar solids, called lime sludge, are produced in treatment
plants where water softening is practiced, and these lime or lime/soda ash plants
account for an additional 25 percent of the industry’s solids production. It is therefore apparent that most of the waste generation where solids are produced involves
water treatment plants using coagulation or softening processes. The above wastes
are solid/liquid wastes in that the liquid waste (water) contains suspended solids
(and, as indicated above, are referred to as sludges). Other solid/liquid wastes produced in the water industry include wastes from iron or manganese removal plants,
spent GAC, spent precoat filter media, wastes from slow sand filter plants, and spent
filter backwash water (SFBW).
The water industry also produces liquid-phase residuals, referred to as such
because the liquid phase (water) contains primarily dissolved solids which are within

the liquid phase itself. These wastes are often called brines or concentrates and
include spent brine from ion exchange regeneration, reject water from membrane
systems (although microfilter and ultrafilter membranes will produce a concentrate
containing suspended solids, they are included in this category), reject water from
electrodialysis plants, and spent regenerant from specific adsorption media such as
activated alumina. Gas-phase residuals are produced as off-gases from air stripping
systems, and off-gas from ozone contactors. The major types of treatment plant
residuals are shown in Table 16.1. This chapter primarily addresses characterization,
handling, and ultimate utilization of sludges. Some introduction to the handling of
liquid-phase residuals is included. More information on handling other residuals can
be found in AWWA (1996) and Cornwell, Bishop, Gould, and Vandermeyden
(1987). A good presentation on softening pellets is in Cornwell and Koppers (1990).

QUANTITY OF SOLID/LIQUID
RESIDUALS GENERATED
Most conventional coagulation plants produce two major residuals—residuals from
the sedimentation basin (commonly referred to as sludge) and residuals from backwashing a filter (referred to as spent filter backwash water).
The quantity of these solid/liquid residuals generated from water treatment
plants depends upon the raw water quality, dosage of chemicals used, performance
of the treatment process, method of sludge removal, efficiency of sedimentation, and
backwash frequency.
One of the most difficult tasks facing the utility or engineer in planning and
designing a residuals treatment process is determining the amount of material (volume and solids) to be handled. The solids quantity is usually determined as an


16.3

WATER TREATMENT PLANT RESIDUALS MANAGEMENT

TABLE 16.1 Major Water Treatment Plant Wastes

Solid/liquid residuals
Alum sludges
Iron sludges
Polymeric sludges
Softening sludges
Spent filter backwash water
Spent GAC or discharge from carbon systems
Slow sand filter wastes
Wastes from iron and manganese removal plants
Spent precoat filter media
Softening pellets
Liquid-phase residuals
Ion-exchange regenerant brine
Waste regenerant from activated alumina
Membrane concentrate
GAC transport water
Gas-phase residuals
Air stripping off-gases
Ozone off-gases

annual average for a given design year and is a function of flow demand projections.
Sometimes overlooked, but very important, is information on seasonal or monthly
variations. It is not unusual for order of magnitude differences in sludge production
to exist for different months of the year.
The amount of alum (or iron) sludge generated can be calculated fairly closely by
considering the reactions of alum or iron in the coagulation process. Using an empirical relation to account for the sludge contribution from turbidity will improve the
estimate, and the contribution from other sources can be added as required.
When alum is added to water as aluminum sulfate, the reaction with respect to
sludge production is typically represented by the simplified equation that includes
three waters of hydration in the product (Cornwell et al., 1987; Cornwell and Koppers, 1990; AWWA, 1996).

Al2(SO4)3 ⋅ 14H2O + 6HCO3− = 2Al(OH)3 ⋅ 3H2O + 6CO2 + 11H2O + 3SO4−2
(16.1)
If inadequate alkalinity is present, lime or sodium hydroxide is normally added to
maintain the proper pH. The three waters of hydration satisfy the covalent bonding
number of six for aluminum. Not including the waters of hydration in the reaction
will tend to underestimate the amount of solids that are produced. This chemically
bound water increases the sludge quantity, increases the sludge volume, and also
makes it more difficult to dewater because the chemically bound water cannot be
removed by normal mechanical methods. Commercial alum has a molecular weight
of 594 and contains two moles of aluminum, each with a molecular weight of 27.
Therefore, alum is about 9.1 percent aluminum (54/594). The resulting aluminum
hydroxide species (Al(OH)3 ⋅ 3H2O) has a molecular weight of 132, and therefore,
1 mg/L of aluminum will produce 4.89 mg of solids (132/27), or 1 mg/L of alum added
to water will produce approximately 0.44 mg/L of inorganic aluminum solids
(0.091 × 4.89). Suspended solids present in the raw water produce an equivalent
weight of sludge solids because they are nonreactive. It can be assumed that other
additives, such as polymer and powdered activated carbon, produce sludge on a one-


16.4

CHAPTER SIXTEEN

to-one basis. The amount of sludge produced in an alum coagulation plant for the
removal of suspended solids is then:
S = 8.34 Q (0.44Al + SS + A)
where

(16.2)


S = sludge produced (lb/day)
Q = plant flow (mgd)
Al = dry alum dose (mg/L) (as 17.1 percent Al2O3)
SS = raw water suspended solids (mg/L)
A = additional chemicals added, such as polymer, clay, or powdered activated carbon (mg/L)
Note: To convert from lb/day to kg/day, multiply by 0.45.

If iron is used as the coagulant, then the equivalent product of equation 16.1 is
Fe(OH)3 ⋅ 3 H2O with a molecular weight of 161. The solids production equation
becomes:
S = 8.34 Q (2.9 Fe + SS + A)
3+

(16.3)
2+

where the iron dose is expressed as mg/L of Fe added or produced via Fe oxidation
(note that significant Fe2+ in the raw water will also produce sludge at a factor of 2.9
if it is oxidized). For iron coagulants, the solids production is best expressed as a function of iron because iron coagulant is purchased in may different forms. It should not
be interpreted from equations 16.2 and 16.3 that iron produces several times the
amount of sludge that alum produces.The units for the coagulant are significantly different for the two equations. In reality, 1 mole of coagulating equivalent of iron produces about 20 to 25 percent more dry-weight sludge than 1 mole of aluminum, based
on the ratio of molecular weights of the product. When iron is purchased as ferric
chloride (FeCl3), the coagulant dose is usually reported as equivalent dry weight of
chemical without waters of hydration (although this should be confirmed with the
manufacturer) and, thus, the coagulant has a molecular weight of 162.3.This results in
the production of 1.0 mg of solids produced for each milligram of ferric chloride
(FeCl3) added. Ferric sulfate is reported by different manufacturers with different
waters of hydration and the individual products need to be referenced.
A treatment plant in the mid-Atlantic area of the United States had
an average raw water turbidity of 4.5 ntu for the period 1991 through 1994 and used

an average ferric chloride dose of 11.5 mg/L (as FeCl3). After conducting a correlation study between SS and turbidity, they found the b value, to convert from turbidity (TU) to suspended solids (SS), to be 1.4 at an average flow of 198 mgd. What was
the annual sludge production?
EXAMPLE 16.1

The solution can be found using equation 16.3. In this equation, the
coagulant dose is expressed as Fe. Therefore, convert FeCl3 dose to Fe dose by:

SOLUTION

MW Fe
56
ᎏᎏ = ᎏ (11.5 mg/L) = 4 mg/L as Fe
MW FeCl3 161
and the solids production
S = 8.34 Q (2.9 Fe + bTU)
S = 8.34 (198) [2.9(4) + 1.4 (4.5)]
= 29,558 lb/day
or 149 lb/MG (million gallons)


WATER TREATMENT PLANT RESIDUALS MANAGEMENT

16.5

Polyaluminum chloride (PACl) is the third major coagulant used. Care especially
needs to be used when converting PACl dose to solids production, because each
manufacturer may use different strengths and utilities report these doses differently.
Some utilities report PACl dose as a neat solution, some as Al2O3, and some as PACl
product.A “typical” manufactured PACl liquid contains about 30 to 35 percent PACl
and around 10 percent Al2O3. One manufactured product contains 33.3 percent

PACl and 10.3 percent Al2O3 or, in this case, the PACl itself contains 30.9 percent
Al2O3. This is equivalent to 16.4 percent aluminum, and therefore 1 mg of PACl (as
PACl) will produce 0.8 mg of solid product (0.164 × 4.89).
The above equations can then be used to track yearly or even daily variation
changes in sludge dry weight produced. One difficulty in applying the relationships
is that most plants do not routinely analyze raw water suspended solids concentrations. The logical correlation is to equate a turbidity unit to a suspended solids unit.
Unfortunately, the relationship is generally not 1 to 1:
SS (mg/L) = b ⋅ TU

(16.4)

The value of b for low-color, predominately turbidity removal plants can vary from
0.7 to 2.2 (Cornwell et al., 1987). It may vary seasonally for the same raw water supply. A utility can therefore either continually measure suspended solids, or it may be
possible to develop a correlation between turbidity and suspended solids. Figure
16.1 shows one such correlation for a low-color raw water source (Cornwell, 1981).
Ideally, this correlation should be done weekly until information is learned as to seasonal variations in the suspended solids/turbidity relationship. Afterward, a monthly
correlation may be sufficient.
Another complication exists for raw water sources that contain a significant
amount of total organic carbon (TOC). Total organic carbon can be a large contributor to the sludge production. Values of b for low-turbidity, high-TOC raw waters
can be as high as 20, but unless turbidity and TOC vary together, a correlation
between suspended solids and turbidity will not exist.
Figure 16.2 shows the relationship between calculated and measured solids production done by the City of Philadelphia (EE&T, 1996). The City used iron as the
coagulant during this time period. A correlation was developed that showed the
ratio of suspended solids to turbidity was 1.4. The calculated quantities using equation 16.3 were within 5 percent of the measured quantities. Through careful calibration and measurements, a complete solids mass balance can be prepared, as was
done by the City of Philadelphia and shown in Figure 16.3.
Through similar theoretical considerations, a general equation has been developed (Cornwell et al., 1987; AWWA, 1981) for plants that use a lime softening process for carbonate hardness removal with or without the use of alum, iron, or
polymer. The equation is:
S = 8.34 Q [2.0 Ca + 2.6 Mg + 0.44 Al + 2.9 Fe + SS + A]
where


(16.5)

S = sludge production (lb/day)
Ca = calcium hardness removed as CaCO3 (mg/L)
Mg = magnesium hardness removed as Mg(OH)2 (mg/L)
Fe = iron dose as Fe (mg/L)
Al = dry alum dose (mg/L) (as 17.1 percent Al2O3)
Q = plant flow (mgd)
SS = raw water suspended solids (mg/L)
A = other additives (mg/L)

The preceding equations or prediction procedures allow estimation of the dry
weight of sludge produced. For sludge productions from noncarbonate hardness


16.6

CHAPTER SIXTEEN

FIGURE 16.1 Suspended solids versus turbidity. (Source: Cornwell, 1981.)

removal or when sodium hydroxide is used, see AWWA (1981). These equations do
not predict the volume of sludge that will be produced.
Volumes and suspended solids concentrations of sludges leaving the sedimentation basins or clarifiers are a function of raw water quality, treatment, and the
sludge removal method. When basins are cleaned only periodically by manual procedures accumulating sludges tend to compact and thicken at the bottom. There is
often a stratification of solids with the heavier particles settling to the bottom and
the hydroxide, or lighter, particles at the top. However, the actual volume produced
will depend largely on the amount of water used to flush the solids out of the basin
during cleaning. With increasing finished water quality standards, there will be a
trend to remove the solids as quickly as possible, generally with continuous collection equipment. In this case, the solids concentrations will be lower because compaction height and time have been less. Solids concentrations using continuous

collection equipment for sludges produced with alum or iron coagulants and for
low- to moderate-turbidity raw waters will be about 0.1 to 1.0 percent leaving
the sedimentation basin. Some of the upflow clarifier devices will produce sludge at
a concentration below 0.1 percent, whereas some of the sludge blanket clarifiers
can produce sludge at over 2 percent solids concentration. The higher the ratio of
coagulant-to-raw-water-solids, the lower the solids concentration and the higher


WATER TREATMENT PLANT RESIDUALS MANAGEMENT

16.7

FIGURE 16.2 Baxter WTP theoretical versus measured residuals quantities. (Source: EE&T, 1996.)

the sludge volume. Coagulant sludges from highly turbid raw waters may be in the
2 to 4 percent solids concentration range and occasionally higher. Sludge volumes
from sedimentation basins tend to be 0.1 to 3 percent of the raw water flow, with
one survey (Cornwell and Susan, 1979) finding an average of 0.6 percent. Softening
sludges will concentrate higher, usually as a function of the CaCO3:Mg(OH)2 ratio
and the type of clarifier. Conventional sedimentation basins may only produce
solids concentrations of 2 to 4 percent, whereas sludge blanket clarifiers can produce solids concentrations of up to 30 percent. Sludge volumes will correspondingly
vary considerably, from 0.5 to 5 percent of the water plant flow.
Spent filter backwash water is characterized by its large water volume, high
instantaneous flow rate, and low solids concentration. Filters can be backwashed at
anywhere from 15 to 30 gpm/ft2, depending upon the media size and water temperature, and the backwash time may be 15 to 20 min. Backwash water volumes are in the
range of 3 to 10 percent of plant production. Accurate plant records often exist on
the amount of backwash water used. The percentage of plant production used for
backwashing can be computed from the ratios of the unit run volumes. For example,
a filter producing water at 4 gpm/ft2 with a 24-h run time has a unit production of
5760 gal/(run ft2 ). If it is backwashed at 20 gpm/ft2 for 20 min, the unit volume of

backwash water is 400 gal/(run ft2 ), for a ratio of about 7 percent backwash water
compared with production water. Spent filter backwash water will typically contain
10 to 20 percent of the total solids production and have suspended solids concentrations of 30 to 300 mg/L depending upon the applied turbidity to the filters and the
ratio of backwash water to production.

PHYSICAL AND CHEMICAL CHARACTERISTICS
OF SOLID/LIQUID RESIDUALS
Physical characterization of water plant wastes is primarily directed at solid/liquid
waste streams of various percent suspended solids concentrations. Solid/liquid
wastes are terms used to describe free-flowing liquids that are predominantly water
all the way up to mixtures that are predominantly solids and behave like a soil texture. Therefore, whenever referring to the physical properties of sludge, it is important to know the suspended solids concentration of the solid/liquid mixture to assess


FIGURE 16.3

Baxter WTP baseline residuals distribution. (Source: EE&T, 1996.)

16.8


WATER TREATMENT PLANT RESIDUALS MANAGEMENT

16.9

the physical state. Cornwell and Wang (Cornwell et al., 1992) used the Atterberg
limit test to classify a sludge’s physical state. The Atterberg test was originally developed to describe quantitatively the effect of varying the water content on the consistency of fine-ground soils. The test consists of measuring five limits; however, the
liquid limit and the plastic limit are the most applicable to sludges. Figure 16.4 shows
the relative location of the liquid and plastic limits. The plastic limit identifies the
solids concentration at which a sludge transitions from a semisolid to a plastic stage
(the plastic state ranges from soft butter to stiff putty). The liquid limit is the solids

concentration below which the sludge exhibits viscous behavior; the consistency
could be described as ranging from soft butter to a pea soup–type slurry. Coagulant
sludges that were tested had liquid limits in the 15 to 20 percent solids concentrations range. Solids concentrations below but near this range would result in a material that still had free water associated with it but may not flow. Generally, a
coagulant sludge is still free flowing up to an 8 to 10 percent solids concentration.
The plastic limit for coagulant sludges was found to be anywhere from 40 to 60 percent solids concentration.
Knocke and Wakeland (1983) divided the physical properties of sludge into
macroproperties and microproperties. Macroproperties include parameters such as
specific resistance, settling rates, and solids concentrations. The indices described
above could be considered macroproperties. Microproperties included particle size
distribution and density. Vandermeyden et al. (1997) studied the micro- and macroproperties of about 80 water plant sludges. Coagulant sludges had a median particle

FIGURE 16.4

Relative location of liquid and plastic limits.


16.10

CHAPTER SIXTEEN

diameter of 0.005 mm with a range of approximately 0.001 to 0.03 mm, as shown in
Figure 16.5. Lime sludge had a similar range, but the median diameter was 0.012 mm.
They also measured the specific gravity of the solid material in the sludge mixtures,
as shown in Figure 16.6. The coagulant residuals had an average specific gravity of
2.32 and the lime residuals averaged 2.50. Koppers (Cornwell and Koppers, 1990)
reported the dry density of coagulant sludges to be about 2.5. Knocke et al. (1993)
found densities for coagulant sludges to range from 2.45 to 2.86, lime sludges to be
2.47, and polymer sludges to be 1.60.
Vandermeyden et al. (1997) also evaluated drainage properties of the 80 residuals using the capillary suction time (CST), specific resistance (SR), and time to filter
(TTF) tests. The CST test is a fast and relatively simple test that is performed to


FIGURE 16.5 Average particle diameter for coagulant and lime residuals. (Source: Vandermeyden et al., 1997.)

FIGURE 16.6 Specific gravity distribution for coagulant and lime residuals. (Source: Vandermeyden et al., 1997.)


WATER TREATMENT PLANT RESIDUALS MANAGEMENT

16.11

determine the rate of free water release from a residual sample. The test is especially
useful for comparing the drainage characteristics of different residuals and for optimizing polymer conditioning of residuals. The test consists of measuring the time in
seconds for free water to travel 1 cm when a 5- to 7-mL sample of residuals is placed
in a special cylinder on top of Whatman No. 17 chromatography paper. As the free
water drains from the residuals through the chromatography paper, it passes by an
electronic sensor that activates a timer. The timer stops when the free water reaches
a second electronic sensor, 1 cm away. The time, in seconds, recorded by the instrument is the CST time.
The resistance to fluid flow exerted by a cake of unit weight of dry solids per unit
area is defined as the specific resistance (SR). To evaluate SR, a sample of residuals
is subjected to a vacuum using a Buchner funnel apparatus. Typically, 100-mL portions of residuals are added to the Buchner funnel, which is lined with a paper filter,
and a vacuum is applied to the filter apparatus. The volume of filtrate generated at
various times is recorded. This procedure is continued until enough water has been
drawn out to produce cracking in the cake on the filter paper, and subsequent loss of
vacuum.
A simplification of the SR test is the TTF test. This test is set up with the same
Buchner funnel apparatus as the SR test, but is much simpler to run. The only data
collected is the amount of time it takes for one-half of the sample volume to filter.
The result is expressed in seconds. Details of the procedures for all these tests are in
Vandermeyden et al. (1997).
Vesilind (1988) presented a model for what occurs during the performance of a

CST test. He proposed that the rate at which water is released from the sludge material into the chromatography paper is a function of two distinct and separate processes. The first is absorption associated with the test instrument, and the second is
water release associated with the sludge material. The absorption associated with
the test instrument can be quantified as a function of the test apparatus and the
chromatography paper. In terms of the test apparatus, the flow of free water from
the solids is a function of the bottom diameter area of the stainless steel reservoir,
the permeability of the chromatography paper used, and viscosity. The values and
effects of each of these parameters can be evaluated and determined through simple
measurements conducted on the test instrument. Viscosity is a function of temperature, so its value must be determined for each test conducted. Because in all likelihood these tests will all be conducted on the same equipment, an instrument
constant can be evaluated.
This instrument constant accounts for the change in diameter between the first
and second sets of electrodes used to measure the CST. It also quantifies the permeability of the chromatography paper and the effects on dewatering associated with
the reservoir.
Vesilind further proposed that the water released from the sludge material is a
function of solids concentration and viscosity. It has long been recognized that solids
concentration has an effect on CST. The sludge concentration is directly proportional
to the filterability constant. The filterability constant can be determined as follows:
µSS
χ = 10−6 Φ [ᎏ]
CST
where

χ = filterability constant [kg2/(s2 m4)]
Φ = dimensionless instrument constant
µ = viscosity [centipoise (cP)]
SS = solids concentration (mg/L)

(16.6)


16.12


CHAPTER SIXTEEN

Equation 16.6 suggests that the filterability constant is a fundamental measure of
sludge dewaterability. Due to the cumbersome nature of the SR test, it is often
preferable to use an easily determined value, such as CST, which allows calculation
of a measure of dewaterability that is independent of solids concentration.
If the filterability index and specific resistance are both measures of dewaterability, then they should plot linearly. Figure 16.7 (Vandermeyden et al., 1997) shows a
plot of the inverse of the filterability constant and SR. As predicted by Vesilind, a
strong correlation exists between the filterability index and the specific resistance.
Both the CST and the TTF tests measure the rate of water release from the sludge,
and therefore one would expect a relationship to exist between the two tests. A plot
of such a relationship is shown in Figure 16.8. The strong correlation suggests that
either test could be successfully used to evaluate sludge drainage characteristics.

FIGURE 16.7 Inverse filterability constant versus specific resistance. (Source: Vandermeyden et al., 1997.)

FIGURE 16.8 CST versus TTF for coagulant residuals. (Source: Vandermeyden et al., 1997.)


16.13

WATER TREATMENT PLANT RESIDUALS MANAGEMENT

Table 16.2 shows the summary statistics for each of the measured drainage tests
(Vandermeyden et al., 1997). Interestingly, the general trend from all the tests is similar. For example, lime sludges had the lowest CST,TTF, and SR.The tests would predict the ease of dewatering order to be lime, ferric, alum, and PACl.
Two additional physical factors, compaction density and the shear strength, are
important when a dry sludge is to be disposed of in a landfill or monofill, or utilized
as backfill or a soil substitute. Compaction tests can be used to determine the achievable dry density for a particular sludge.This value can be useful when estimating volumes required for a landfill or the truck volumes required to haul a certain weight of
sludge. The first work completed on using compaction density values for water plant

sludge characterization was done in Europe in the mid-1980s, as reported by Koppers (Cornwell and Koppers, 1990). Koppers used moisture-density curves to show
optimum compaction for a particular sludge. At low solids concentrations, the density is essentially the same as water (62 lb/ft3, 1,000 kg/m3), whereas he reported
highly dewatered coagulant sludges to have a density of 100 lb/ft3 (1700 kg/m3). In
work by Cornwell (Cornwell et al., 1992), coagulant sludges at about an 80 percent
solids concentration (highly dewatered) were reported to have a density of 110 to
125 lb/ft3 (1850 to 2100 kg/m3) and, up to a 20 percent solids concentration, the densities were about 60 lb/ft3 (1000 kg/m3), essentially the same as water.
The shear strength of sludges relates to the overall ability of the sludge to support
itself and external loadings, such as vehicle traffic or earth-moving equipment. Measuring shear strengths of water plant sludges is complicated by the fact that the shear
strength will vary with sludge age and with disturbance. Novak and Calkins (1973)
first used the shear value of sludges as a way to measure their handleability. They
reported that to achieve a shear value of 0.7 psi (4.7 kN/m2), where they believed the
sludge could be “handled,” required a solids concentration for an alum, iron, and
lime sludge of 30, 40, and 55 percent, respectively. Koppers reported shear values for
coagulant sludges at 20 to 30 percent solids concentration to be 0.3 to 0.6 psi (2 to
4 kN/m2). They felt that appropriate landfilling would require a shear strength of
1.4 psi (10 kN/m2) or above, which would require above a 35 percent solids concentration. Figure 16.9 shows an example of shear strength as a function of solids concentrations for three different water plant sludges (Cornwell et al., 1992). The
techniques used by Cornwell and Wang (Cornwell et al., 1992) resulted in similar
conclusions as Koppers, although they predicted that a shear value of over 4 psi
(28 kN/m2) would be needed for monofilling sludge to support earth-moving equipment, which would require a solids concentration for coagulant sludges of between
30 and 50 percent. There was a wide variation in results for the three sludges tested.
Chemical characterization of solid/liquid water plant residuals is primarily concerned with determining total metal concentrations, leachable metals, and nutrient
TABLE 16.2 Summary Statistics for Drainage Parameters
CST(s)

SR(1013 m/kg)

TTF(s)

Sludge
type


n

Mean

Std.
dev.

n

Mean

Std.
dev.

n

Mean

Std.
dev.

Alum
Ferric
PACl
Lime

38
9
5

9

194.1
103.0
289.8
70.0

195.4
64.5
258.8
34.5

38
9
5
9

319.5
104.7
410.9
34.3

412.6
79.5
562.5
20.4

38
9
5

9

15.8
6.4
13.8
0.5

21.3
8.1
11.0
0.82

Source: Vandermeyden et al., 1997.


16.14

CHAPTER SIXTEEN

FIGURE 16.9 Shear strength versus solids concentration, cone penetration, and triaxial compression tests. (Source: Cornwell et al., 1992.)

levels. Several publications (Schmitt and Hall, 1975; Cornwell et al., 1987; AWWA,
1996; Cornwell et al., 1992; Cornwell and Koppers, 1990) have presented values for
total metal concentrations of coagulant sludges; one such example is shown in Table
16.3. The source of the metals can be the raw water as well as the coagulant itself.
Metals that are often found in coagulant sludges include aluminum, arsenic, occasionally cadmium, chromium, copper, iron, lead, manganese, nickel, and zinc.
Extraction tests are designated by USEPA to be one of the methods to determine
if a waste is hazardous as per Subtitle C of RCRA. The toxicity characteristic leach
procedure (TCLP) is used to determine if a waste is toxic, and therefore classified as
hazardous. The presence in the extract from the TCLP test of any one of a number

of contaminants above a specified regulatory level constitutes failure of the test and
results in the waste being classified as hazardous. There are no reported coagulant or
lime sludges that have failed the TCLP test. In fact, it is rare to even find detectable
levels of the regulated contaminants in the extract from a TCLP test on a coagulant


16.15

WATER TREATMENT PLANT RESIDUALS MANAGEMENT

TABLE 16.3 Example Total Metal Analysis for Coagulant Sludges

Metal

Alum sludge 1
(mg/kg dry weight)

Alum sludge 2
(mg/kg dry weight)

Alum sludge 3
(mg/kg dry weight)

Aluminum
Arsenic
Barium
Cadmium
Chromium
Copper
Iron

Lead
Manganese
Mercury
Nickel
Selenium
Silver
Zinc

107,000
25.0
30
1
120
168
48,500
11
1,180
0.1
24
<2
<2
91.7

123,000
32.0
<30
1
130
16
15,200

9
233
<0.1
23
<2
<2
393

28,600
9.2
230
2
50
52
79,500
40
4,800
0.2
131
<2
<2
781

Source: Cornwell et al., 1992.

or lime sludge. Due to the pH of lime sludges, the TCLP test is not used, and the
sludges are classified as nonhazardous according to this procedure.
Leaching tests provide another method that can characterize the release of metals from a sludge. Cornwell et al. (1992) conducted leaching tests in a lysimeter, in
which the solids were subjected to rainfall of volume equivalent to about 12 years of
normal precipitation and the leachate was analyzed for metal concentrations. Figure

16.10 shows an example of the leaching of arsenic from three sludges and compares
the leach levels with in-stream water quality standards and to the drinking water
MCL. Only arsenic, copper, iron, manganese, and zinc leached from all three sludges.
A slight degree of leaching of nickel and cadmium was exhibited by the ferric sludge
only. Selenium leached from two sludges, but only during the first week. Aluminum

FIGURE 16.10 Arsenic leaching in relation to drinking water and in-stream water quality
from an alum sludge. (Source: Cornwell et al., 1992.)


16.16

CHAPTER SIXTEEN

did not leach from any of the sludges, and none of the primary drinking water MCLs
for metals monitored were exceeded in any of the leachate samples analyzed.
Water plant coagulant sludges are generally considered to be lower in nutrient
content than biosolids, particularly nitrogen and phosphorus. However, some residuals can have as much as 3 percent total nitrogen or phosphate content, which can
have some fertilizer value. In fact, in some cases nitrogen can be the limiting constituent in determining the maximum amount of residuals that can be applied to a
specific land (EE&T, 1992).
Although specific inorganic concentrations are not reported in the literature for
lime sludges, it is certain that they will be present to the extent that they are removed
from the raw water. Some removal from the raw water of most of inorganic contaminants will take place during lime softening, with high removals for compounds such
as lead, cadmium, and arsenic (at a high softening pH).
Many groundwaters being treated for hardness removal also contain background
concentrations of naturally occurring radium. Radium is a naturally occurring
daughter product of U238. Decay of U238 over millions of years passes through a series
of elements eventually producing radium. The parent elements of radium are generally insoluble in water so that radium is often the first radioactive element which is
found in drinking water supplies. The major threat to human health from radium
daughters comes from breathing air containing radon and its very short lived daughters, which can accumulate as solids in the lungs. This exposes the lungs and other

internal organs to continuous radiation. In addition, the sludge can directly expose
humans to gamma radiation from the decay of radium. However, for the most part,
the safe handling and disposal of lime sludges containing radium involves the prevention of radon exposure.
Measurement of radioactive components is expressed in curies or picocuries, pCi
(10−12 Ci).A curie is the official unit of radioactivity, defined as exactly 3.70 × 1010 disintegrations per second. Snoeyink (1984) reported data for radium concentrations in
various lime softening sludges. The sedimentation basin sludge concentrations of
Ra226 ranged from 1,000 to 11,000 pCi/L of sludge, Ra228 varied from 200 to 12,000
pCi/L. Because the radium is associated with the sludge solids, its concentration in
the liquid stream is a function of the solids concentration. The concentration per
gram of solids is 10 to 20 pCi/g for Ra226 and 1 to 11 pCi/g for Ra228. Backwash water
concentrations for Ra226 ranged from 6 to 50 pCi/L.

THICKENING SOLID/LIQUID RESIDUALS
After removal from a clarifier or sedimentation basin, sludges can be thickened in a
gravity concentration tank. Thickening can be economically attractive in that it
reduces the sludge volume and produces a more concentrated sludge for further
treatment in the dewatering process, or for perhaps hauling to a land application
site. Some dewatering systems will perform more efficiently with higher solids concentrations. Thickening tanks can also serve as equalization facilities to provide a
uniform feed to the dewatering step. Although there are a few types of thickeners
available on the market, the water industry almost exclusively uses gravitational
thickening.
Sludge thickening is performed primarily for reduction in the volume of sludge.
The relationship between the volume of sludge and the solids concentration is
expressed as:


WATER TREATMENT PLANT RESIDUALS MANAGEMENT

16.17


M
V=ᎏ
rsP

(16.7)

where V = volume of sludge (m3)
M = mass of dry solids (kg)
r = density of water = 103 kg/m3 (at 5°C)
s = specific gravity of the sludge mixture
P = percent solids expressed as a decimal (weight/weight)
An approximation, assuming the specific gravity of the sludge does not change and
that there is 100 percent solids capture, for determining volume reduction based on
percent solids is expressed as:
V2 P1
ᎏ=ᎏ
V1 P2

(16.8)

This is a quick and useful equation because the specific gravity of the sludge is not
always known, but generally does not change within the limits of most thickening
operations. Therefore, for thickening a 1 percent solids concentration sludge to 10
percent solids concentration a volume reduction of approximately 90 percent is
achieved.
EXAMPLE 16.2 An alum sludge is produced from a sedimentation basin at a 0.3 percent solids concentration. The plant produces 100 kg of sludge per day. What is the
volume of sludge produced? What would be the volume if a thickener were used to
increase the solids concentration to 2 percent?
To determine the volume from Eq. 16.7, the specific gravity of the sludge must be
known. It was reported earlier that the specific gravity for low solids concentrations

is essentially 1.0. For alum sludges, this seems to hold up to about a 20 percent solids
concentration, well above the performance of thickeners. Therefore, the specific
gravity of 1.0 can be used in Eq. 16.7 and Eq. 16.8 applies.

M
100
V = ᎏ = ᎏᎏ
= 33.3 m3
rsP 103 (1) (0.003)
The volume after thickening would be:
0.3
P1
V2 = ᎏ (V1) = ᎏ (33.3)
2.0
P2
V2 = 5 m3
Gravity sludge thickeners are generally circular settling basins with either a
scraper mechanism in the bottom (see Figure 16.11), or equipped with sludge hoppers (Figure 16.12). They may be operated as continuous flow or as batch “fill-anddraw” thickeners. For continuous flow thickeners, the sludge normally enters the
thickener near the center of the basin and is distributed radially. The settled water
exits the thickener over a peripheral weir, or trough, and the thickened sludge is
drawn from the basin. For tanks equipped with a scraper mechanism, the scraper is
located at the thickener bottom and rotates slowly.This movement directs the sludge
to the draw-off pipe near the bottom, center of the basin. The slow rotation of the
scraper mechanism also prevents bridging of the sludge solids. The basin’s bottom is
sloped to the center to facilitate collection of the thickened sludge.


FIGURE 16.11

Continuous flow gravity thickener.


16.18


WATER TREATMENT PLANT RESIDUALS MANAGEMENT

FIGURE 16.12

16.19

Batch-thickening tank schematic.

Batch fill-and-draw thickening tanks are often equipped with bottom hoppers as
was shown in Figure 16.12. In these tanks, sludge flows into the tank, from either a
periodic or continuous removal of sludge from the sedimentation basin, until the
thickening tank is full. The sludge is allowed to quiescently settle and a telescoping
decant pipe is used to remove supernatant. The decant pipe may be continually lowered as the solids settle until the desired solids concentration is reached or the
sludge will not thicken further. The thickened sludge is then pumped out of the bottom hoppers to further treatment or disposal.
Design of batch or continuous flow thickeners is usually accomplished based on
previous experience of similar full-scale installations or on laboratory or pilot settling
tests. Thickeners are used not only to increase the concentration of the sludge, but
also to remove solids from the liquid phase so that the supernatant can be discharged
to a receiving stream or recycled to the head of the treatment plant. Quality requirements for discharge are primarily concerned with meeting a set turbidity or suspended solids level. Considerations for quality suitable for recycle include the
effective removal of parasitic cysts (see the section on recycle). For designs based
on thickening the residuals, the thickener surface area is often solids limited, and a
flux rate is used for determining the thickening area in units such as kg/(m2 h) or
lb/(day ft3). However, units designed to meet a given supernatant quality are often
hydraulically limited such that sizing is based on the solids settling velocity, m/h or
gpm/ft2. Although there is some available literature to set flux rates for coagulant or
calcium carbonate sludges, site-specific study is needed to determine thickening

requirements and appropriate settling velocities when supernatant quality is the criterion. Settling velocity studies can be conducted similarly to those for designing sedimentation basins (see Chapter 7).AWWA (1996) reports that a typical loading rate for
alum sludge thickening is 4 lb/(ft2 day) [20 kg/(m2 day)], whereas later in this chapter, it is reported that for cyst removal from sludge streams, a hydraulic loading of


16.20

CHAPTER SIXTEEN

0.05 gpm/ft2 (0.12 m/hr) may be needed if polymer is not used. For these hydraulic and
flux values, the thickener would be hydraulically limited rather than solids limited.
The common settling test used in the laboratory is conducted in a transparent
cylinder filled with sludge and mixed to evenly distribute the solids. At time zero, the
mixing is stopped and the solids are allowed to settle. Water plant sludges from clarifiers and sedimentation basins will generally settle as a blanket with a well-defined
interface. By recording the height of the interface with time, a plot such as Figure
16.13 can be created. The free settling velocity is then determined as the slope of the
straight-line portion of the plot.
In considering the size of the test system, the cylinder diameter is probably the
most critical factor.Vesilind (1979) has evaluated the effects of various cylinder diameters. At low suspended solids concentrations (<0.4 percent), the smaller cylinders
tended to underestimate the settling velocity, which would result in a more conservative design. However, at suspended solids concentrations over 0.5 percent, the smaller
cylinders overestimated the settling velocity.The results are probably site specific, but
their work does show the importance of selecting the proper size cylinder for pilot
studies. Vesilind made four recommendations for conducting thickening tests:
1. The cylinder diameter should be as large as possible; 8 in is a practical compromise.
2. The initial height should be the same as the prototype thickener depth. When this
is not practical, 3 ft should be considered minimum.
3. The cylinder should be filled from the bottom.
4. The sample should be stirred throughout the test, but very slowly—0.5 rpm is a
reasonable speed for an 8 in cylinder. This slow stirring will help the test results
of small cylinders better approach that of full-scale.
After completing a settling test as described by Figure 16.13, the test is repeated

with several different initial suspended solids concentrations, resulting in plot A of
Figure 16.14. The solids flux, F, is then computed as:
F = vCi

FIGURE 16.13 Thickening test in a cylinder with resulting interfaceheight-versus-time curve.

(16.9)


WATER TREATMENT PLANT RESIDUALS MANAGEMENT

16.21

(a)

(b)

FIGURE 16.14 Thickening test in a cylinder with resulting
batch-thickening test plots. Note: The results of several batchthickening tests ploted as (a) interface velocity versus initial
solids concentration, and (b) solids flux versus initial solids
concentration.

where F = solids flux [kg/(m2 h)]
v = settling velocity (m/h)
Ci = initial suspended solids concentration (kg/m3)
The flux curve will generally take the shape as shown in plot B of Figure 16.14.
The flux is zero at zero suspended solids concentration, and the settling velocity
approaches zero as the solids concentration increases, thereby driving the flux to
zero and representing the maximum possible solids concentration to be achieved.
For batch fill-and-draw tanks, the curves of plot A of Figure 16.14 can be directly

used to estimate the settling time required and predict the thickened solids concentrations. Similarly, the curves of plot B of Figure 16.14 could be used to develop possible flux rates for the anticipated range of influent solids concentrations.


16.22

CHAPTER SIXTEEN

For continuous flow thickeners, the solids move to the bottom of the tank not
only due to the batch sedimentation discussed above, but also due to the velocity
created by the underflow of sludge being removed from the thickener. The flux due
to the sludge withdrawal is:
Fu = vu Ci

(16.10)

where Fu = flux due to underdrain withdrawal of sludge
vu = downward velocity caused by sludge removal
Ci = solids concentration at a given layer in the thickener
The flux due to settling has already been defined and can be labeled for the continuous flow thickener as:
FB = viCi

(16.11)

where FB = flux due to solids settling
vi = settling velocity of solids concentration, Ci
Ci = solids concentration at a given layer in the thickener
Therefore, the total flux, F, is:
F = viCi + vuCi

(16.12)


The total flux is plotted in Figure 16.15. The minimum point of this flux curve
occurs at the solids concentration layer which restricts performance, and thus the
flux cannot be higher. Several methods are available to find this maximum flux, most
of the methods having been described by Vesilind (1979). A widely used method was
developed by Coe and Clevenger (1916).
Coe and Clevenger used mass and liquid balance equations to develop the following expression for thickener area:

FIGURE 16.15 Solids flux curve for continuous thickener with Yoshioko solution. (Source: Visiland, 1979.)


WATER TREATMENT PLANT RESIDUALS MANAGEMENT

΂

Q0 C0 1
1
A=ᎏ ᎏ−ᎏ
vi
Ci Cu

΃

16.23

(16.13)

where A = thickener area (m2)
Q0 = feed flow rate (m3/sec)
C0 = influent sludge solids concentration (kg/m3)

vi = settling velocity of solids, Ci (m/sec)
Ci = design solids concentration of given layer within the thickener (kg/m3)
Cu = desired underflow concentration (kg/m3)
It is necessary to find a number of values of vi and Ci, and conduct a series of area
calculations. The largest area calculated should be used as the minimum design
thickener area. A graphical solution was introduced by Yoshioko et al. (1957), as
shown in Figure 16.15. A line is drawn tangent to the minimum point in the flux
curve beginning at the desired underflow solids concentration. This line will intersect the limiting flux and allow calculations of the required area.
A thickener is to be sized to accomplish both meeting required discharge limits for the supernatant of 10 mg/L suspended solids and to thicken the
underflow solids to 2 percent suspended solids from an initial suspended solids concentration of 0.1 percent. The initial sludge flow is 500 gpm.
To properly size this thickener, two types of pilot study are required. The first
study would determine the effective solids settling velocity to achieve a supernatant
suspended solids of 10 mg/L. This testing could be done using techniques described
in AWWA M37 (1992) and would result in determining a settling velocity, for example, of 0.4 cm/min. The area required for appropriate settling to meet the discharge
limit would be found from this value.

EXAMPLE 16.3

΂

ft
(0.4 cm/min) 0.03 ᎏ
cm

(7.5 gal/ft )
΃ ΂ᎏ
ft ΃
ft2

3


2

= 0.10 gpm/ft2
The area is found as
500 gpm
ᎏᎏ2 = 5000 ft2
0.10 gpm/ft
The area must also be found to accomplish the required thickening. A series of settling tests would be conducted resulting in the construction of a curve such as Figure
16.15, resulting in the determination of the allowed flux, for instance, of 4 lb/(ft2 day).
The pounds of solids loaded to the thickener on an equivalent day basis is:
ᎏ ᎏ ᎏᎏ
΂ᎏ
100 ΃ ΂ min ΃ ΂ gal ΃ ΂ day ΃
0.1

500 gal

8.34 lb

1440 min

= 6004 lb/day
and the required area is
6004 lb/day
ᎏᎏ
= 1500 ft2
4 lb/(ft2 day)



16.24

CHAPTER SIXTEEN

In this case, the limiting area would be controlled by the discharge requirements
for supernatant quality and would be 5000 ft2.
To avoid the need to conduct several settling tests at different solids concentrations as the procedure above requires, researchers (Kynch, 1952; Talmage and Fitch,
1955) have developed simpler test procedures to determine the flux rates. They have
formulated techniques that make it possible to develop a complete velocityconcentration profile from only one settling test. Unfortunately, experiments have
proven that these methods do not apply to highly compressible materials and thus
are not applicable in designing water plant sludge thickeners (Vesilind, 1979).

NONMECHANICAL DEWATERING
OF SOLID/LIQUID RESIDUALS
Nonmechanical dewatering, as the name implies, is the dewatering of water treatment plant residuals through means that do not require the use of mechanical
devices such as centrifuges or filter presses. Nonmechanical dewatering is used in
locations where land is available and where it can be both economical and efficient
for dewatering water treatment plant residuals.
A variety of means are employed to accomplish nonmechanical dewatering. The
most basic of these is separation of solids and free water through sedimentation followed by natural air drying of the residuals.A second method allows free water to be
percolated through sand and into an underdrain system, while additional solids concentration increases are achieved through evaporation. In northern climates, a third
system is utilized whereby water treatment plant residuals are subjected to freezing
and thawing, which dramatically reduces the residuals volume and correspondingly
increases the solids concentration.

Sand Drying Beds
Sand drying beds were initially developed for dewatering municipal wastewater
biosolids, but they have since been used to dewater residuals from water treatment
plants. Drainage (percolation), decanting, and evaporation are the primary mechanisms for dewatering. Following residuals application to the drying bed, free
water is allowed to drain from the residuals into a sand bottom from which it is

transported via an underdrain system consisting of a series of lateral collection
pipes. This process continues until the sand is clogged with fine particles or until all
the free water has been drained, which may require several days. Secondary free
water can be removed by decanting once a supernatant layer has formed. Decanting can also be utilized to remove rainwater that would otherwise hinder the overall drying process. Water remaining after initial drainage and decanting is removed
by evaporation over a period of time necessary to achieve the desired final solids
concentration.
Several variations on sand drying beds are currently in use, and Rolan (1980) proposed the following classification categories:
1. Sand drying beds. These are conventional rectangular beds with sidewalls and a
layer of sand or gravel with underdrain piping.These are built with or without the
provisions for mechanical removal of the dried residuals and with or without a
roof or greenhouse-type covering.


WATER TREATMENT PLANT RESIDUALS MANAGEMENT

16.25

2. Paved rectangular beds. These have a center sand drainage strip, with or without
heating pipes, buried in the paved section and with or without covering to prevent rain incursion. The paved bottom beds are referred to as solar drying beds.
3. Drying beds with a wedge-wire septum. These incorporate provisions for an initial flood with a thin layer of water followed by introduction of liquid residuals on
top of the water layer, controlled formation of cake, and provisions for mechanical cleaning.
The layout and construction of sand drying beds is very site specific—topography, available land, and operational constraints must all be considered. Topography
plays a key role in how beds are laid out on a site, and operational constraints (such
as residuals pumping distance) must also be considered when siting a bed location.
Materials used in construction are typically cast-in-place concrete or concrete block
when the beds are constructed at grade, or earthen sides with a liner when the beds
are constructed below grade.
Underdrain systems for sand drying beds are used to collect water that has percolated down through the sand and gravel and transmit it to a point of discharge.
Some states require an impervious clay or liner below the underdrain piping. When
the beds are equipped for decanting, the flow from the decant mechanism is often

tied to the underdrains so that a combined effluent is produced. Underdrains are
typically constructed of vitrified clay or plastic piping and many configurations exist,
but the most common is collecting drainage with laterals and conveying the flow to
a header pipe. Figure 16.16 shows a section view of a typical sand drying bed.
Several sand drying beds are typically used at a given site, which offers some
advantages from an operations point of view. Chief among these is the ability to
rotate bed use, so that as one sand drying bed is loaded and the residuals begin to
dry, another bed is cleaned and readied for a new application of residuals. Cleaning of the sand drying bed can be accomplished with mechanical equipment if concrete support runners are properly installed in the bed. Front-end loaders and
Vac-haul trucks have been used successfully by utilities operating sand drying
beds.
The time required to evaporate the water remaining after percolation and
decanting is the controlling factor in determining the bed size required. Therefore,
maximizing the removal of water prior to evaporation will maximize the bed yield.
The solids that remain after drainage and decanting are referred to as the drained
solids. The value of the drained solids concentration is dependent upon the initial
solids concentration, the use of conditioning agents such as polymer, the applied
depth and the efficiency of the water removal system. A series of pilot tests can be
conducted to determine the combination of loading depth, initial solids concentration, and polymer use that maximizes the drained solids concentration. Vandermeyden et al. (1997) developed the following unit specific equation to determine the bed
yield once the drained solids concentration is known:
SSf SSd E
Y = (0.624) ᎏᎏ
SSf − SSd

(16.14)

where 0.624 = unit conversion assuming a specific gravity of 1 (the conversion factor is 1.2 for a yield in kg/(m2 year) and evaporation in cm/month)
Y = sand bed yield [lb/(ft2 year)]
SSf = desired final solids concentration (percent)
SSd = drained solids concentration (percent)
E = net pan evaporation (in/month)