-------
SECTION 5
SAMPLING AND ANALYTICAL PROCEDURES
SAMPLING PROCEDURES
The sampling procedures described below were utilized for the duration
of this research effort. Samples for routine wet chemistry and metals
analyses were collected by the operators on duty at the Demonstration Plant,
and samples for microbiology were collected by microbiology laboratory
technicians.
Routine Chemistry Samples
Samples for routine wet chemistry analyses were collected by the
plant operators seven days a week at 1:00 am, 5:00am, 9:00am, l:00pm,
5:00pm, and 9:00pm. Wide-mouth, half-pint plastic bottles were used for
sample collection. These sample bottles were placed in a refrigerator until
transported to the laboratory, at which time they were composited by the
staff chemists. Since the Demonstration Plant was operated at hydraulic
steady-state, equal volumes (400 ml) of each of the six grab samples were
used for the 24-hour composite sample.
Metals Samples
Samples for metals determinations were collected by the plant operators
at the same time samples for routine analyses were collected. Four-hundred
milliliter fractions were composited in one-gallon amber bottles to which
redistilled nitric acid (10 ml per liter) had been previously added for
sample preservation.
Microbiological Samples
Either the staff microbiologist or the microbiology laboratory tech-
nicians collected all samples for microbiological evaluation. The samples
were collected in 125 ml, wide-mouth glass bottles with glass stoppers
that had been previously dry sterilized at 177°C for one hour.
Sampling Frequency
Most of the more conventional water quality parameters that had process
control significance were evaluated daily on 24-hour composite samples.
Those parameters necessary for general background information, such as -->'-.
32
-------
chlorides and sulfates, were evaluated on weekly composite samples.
Samples were collected for metals analyses every other day.
ANALYTICAL PROCEDURES
The analytical procedure .used in this research effort followed the
13th Edition of Standard Methods for the Examination of_ Water and Wastewater
in so far as practicable (3).
Flow
The influent flows to all unit processes, with the exception of the
recarbonation basin, were measured by BIF/Brooks magnetic flow meters. This
combination of meters and recorders proved to be unreliable, and instrumen-
tation technicians found it impossible to keep the flow meters properly
calibrated and maintained. These problems resulted in the installation of
several physical flow measuring elements such as orifices, weirs and
venturi sections in order that accurate flow measurements could be obtained.
Chemical Oxygen^ Demand
The following procedures were utilized to determine COD values on the
routine samples. The low-level technique was employed for those samples
where the COD was expected to be less than 50 mg/1.
High-Level Technique—•
The procedure used was as described in Section 220 of Standard Methods.
Low-Level Technique—
The COD of low-level samples was determined by using the procedure
given on page 19 of Methods for Chemical Analyses of_ Water and Wastes 1971
(4). Two modifications were made to the procedure. The amount of mercuric
sulfate was reduced from 1.0 to 0.4 grams, and the ferrous ammonium sulfate
solution was 0.01 N instead of 0.025N.
Total Organic Carbon
All total organic carbon determinations were made using a Beckman Model
915 Total Carbon Analyzer.
Total Residue
Total solids determinations were made by employing the procedure in
Section 224A of Standard Methods.
Nonfiltrable Residue
Total suspended solids determinations were made by employing the
procedure in Section 224C of Standard Methods using 2.4 cm diameter
glass-fiber filters and Gooch crucibles.
33
-------
Total Dissolved Solids
Total dissolved solids were computed by subtracting the nonfiltrable
residue from the total residue.
Total Phosphorus
The single reagent method given in Methods for Chemical Analyses of
Water and Hastes 1971 was used for all total phosphorus determinations.
The amount of ammonium persulfate used was increased from 0.4 to 0.5 grams,
and the amount of combined reagent was increased from 8 ml to 10 ml.
Ammonia Nitrogen
Ammonia nitrogen determinations were made by using an ion-specific
electrode and the Known Addition Method (5). The electrode used was an
Orion Model 95-10.
Total Kjeldahl and Organic Nitrogen
Total Kheldahl nitrogen was determined by using an ion-specific
electrode and the Known Addition Method after completing the digestion
phase of the procedure given in Section 216 of Standard Methods. Organic
nitrogen was determined by subtracting the ammonia nitrogen from the
total Kheldahl nitrogen.
Nitrite Nitrogen
Nitrite nitrogen determinations were made using the procedure described
in Section 134 of Standard Methods.
Nitrate Nitrogen
The phenoldisulfonic acid method, Section 213D of Standard Methods,
was used to determine combined nitrite-nitrate nitrogen.
Sulfate
Sulfate was determined by an indirect atomic absorption spectroscopy
method by adding a known concentration of barium chloride to form a barium
sulfate precipitate. The barium concentration in solution was then
determined by atomic absorption, and the sulfate concentration determined by
subtraction, as outlined in the Perkin-Elmer Applications Manual.
Chloride
Chloride concentrations were determined by the mercuric nitrate method
described in Section 112B of Standard Methods.
34
-------
Alkalinity
Total and phenolphthalein alkalinity was determined by using the
procedures given in Section 102 of Standard Methods using methyl orange
and phenolphthalein.
Turbidity
Turbidity was determined by the nephelometric method described in
Section 163A of Standard Methods with a Hach Model 2100A Turbidimeter.
The standard references were formazin polymer suspensions.
Color
Color determinations were made by plant operators using a Hellige
Aqua Tester and platinum-cobalt color disk.
Metals Determinations
Samples for metals analyses were filtered through a glass fiber filter
and then concentrated by a factor of ten. Concentration was accomplished
by heating (below the boiling point) a 500 ml sample until the volume was
reduced to less than 50 ml, and then making up to volume in a 50 ml
volumetric flask.
Atomic Absorption—
Atomic absorption spectroscopy was utilized to determine the concen-
trations of aluminum, barium, cadmium, calcium, cobalt, copper, chromium,
iron, lead, magnesium, manganese, silver, strontium, and zinc. A Perkin-
Elmer Model 403 was used for these analyses, and standard procedures
given in the Perkin-Elmer Operator's Manual (6) and Standard Methods were
employed.
Flame Emission—
Sodium and potassium concentrations were determined by flame emission
spectrophotometry by operating the PE 403 in that mode, and using methods
given in Standard Methods.
Arseni c—
Arsenic concentrations were determined
thiocarbamate method presented in Section
by using the silver diethyldi -
104A of Standard Methods.
Boron--
The curcumin method given in Section 107A of Standard Methods was used
to determine boron concentrations, with an ion-exchange modification to re-
move cationic interferences.
Beryllium—
The morin
ations.
fluorometric method (7) was employed for beryllium determin-
35
-------
Mercury--
Mercury concentrations were determined by the flame!ess atomic
absorption method with a Perkin-Elmer Model 290B atomic absorption
spectrophotometer.
Molybdenum—
The dithiol method of Brown, et al. (8) was used to determine
molybdenum concentrations.
Selenium--
Selenium concentrations were determined by employing the diaminobenzi-
dine method given in Section 150A of Standard Methods.
Silica—
Silica determinations were made by using the heteropoly blue method
given in Section 151C of Standard Methods.
Vanadium—
The catalytic oxidation method presented by Brown, et. al. (U.S.
Geological Survey) was used to determine vanadium concentrations (8).
VIRUS DETERMINATIONS
When the virus studies were first conceived, it was thought that surplus
stocks of vaccine strains of poliovirus could be purchased for use in
seeding the various unit processes, The grant which funded the studies pro-
vided the assistance of the National Environmental Research Center (NERC)
virology program in Cincinnati. NERC was supposed to help by obtaining the
viruses, preparing and titering the stocks for seeding, and processing the
samples from the experiments, thus performing the key functions. The price
quotations NERC received for the attenuated viruses were prohibitive, and
NERC had to resort to its own resources for growing the viruses.
Stock
The stock culture used for the virus study were the Poliovirus type I
(vaccine strain), f2 Coliphage, and _E_. coli K12 (f+) indicator cells.
Poliovirus Assay
One- to four-liter samples were collected during the experimental runs in
gas-sterilized flexible Cubitainers, capped, and placed in ice. The samples
were shipped by air in insulated boxes containing ice pre-frozen in water-
tight quart-size Cubitainers, and arrived in Cincinnati the same night
(except for the last shipment -- run No. 4, which did not arrive until noon
the following day in spite of all possible efforts to insure prompt handling).
The shipments were picked up at the airport and taken to the NERC
Cincinnati laboratory by EPA personnel. All samples (except sludge) were
Swinny-filtered with 0.45 y Mi Hi pore filter membranes treated with Tween
80 and then inoculated onto BGM (Barren Green Monkey Kidney tissue) cell
36
-------
lines using 1/2 ml in each of 4 bottles for each dilution. For sludge
samples, approximately 200 ml of sample was centrifuged and a 15 gram
portion of the centrate placed in a beaker. To this residue was added 40 ml
of 10 percent buffered beef extract (Oxoid, Lab Lemco Powder, Flow Labora-
tory, Rockville, Maryland) which was mixed for 30 minutes on a magnetic
mixer and then Swinny filtered with a 0.45 u Mi Hi pore membrane. All
of the filtrate was then inoculated onto BGM cell lines using 1 ml per
bottle (approximately 40 bottles).
Coliphage Media
The coliphage media consisted of three substances: Tryptone broth,
tryptone overlay agar, and tryptone plating agar. The tryptone broth
consisted of 10 rag/1 of Tryptone (Difco 0123), 1.0 g/1 of yeast extract
(Difco 0127), 1.0 g/1 of glucose, 8.0 g/1 NaCl, and 0.22 g/1 of CaCl?.
Tryptone overlay agar was the same as tryptone broth with the addition
of 7.0 g/1 agar (Difco 0140). Tryptone plating agar was the same as
tryptone broth with the addition of 15 g/1 of agar. The salt diluent was
8.5 g/1 CaCl and 0.22 g/1 CaCl2. Media and diluent were sterilized by
autoclaving at 15 psi and 121°C for 15 minutes. Glassware was sterilized
in a hot air sterilizer at 170°C for two hours.
Coliphage seed
An overnight culture of IE. coli K12 (f+) was diluted 1:100 in one
liter of tryptone broth. The culture was grown on a shaker at 37°C to an
optical density of 0.2-0.3 which was approximately 108 cells/ml. The
culture was infected with f2 coliphage at a multiplicity of infection (MOI)
of 3, and grown for 4-6 hours longer on the shaker. Twenty to thirty
milliliters of chloroform were added and it was refrigerated overnight.
The following day, the culture was centrifuged at 16,000 G for 20 minutes
at 4°C to remove cellular debris. The supernatant yielded a stock suspension
with a titer of at least 1 x 1011 pfu/ml.
Coliphage Assay
Samples of 10 ml were collected and 1/2 ml chloroform was added
immediately. The samples were stored in a refrigerator overnight. The
next day, the following procedure was utilized. To sterile aluminum-capped
tubes in a 47°C water bath, the following mixture was added: 2.5 ml molten
tryptone overlay agar, 2.0 ml of IE. coli K12 indicator cells diluted
in .tryptone broth to a concentration of lO7 cells/ml, and 0.5 ml of sample
containing the phage or, if nece'ssary, 0.5 ml of a 10-fold serial dilution.
Salt diluent was used to make the sample dilutions. The tube contents were
mixed on a vortex mixer. Contents were poured onto a petri plate containing
20 ml of solidified tryptone plating agar. The plate was swirled to evenly
distribute the overlay agar and then allowed to solidify. The plates were
inverted and incubated 18 hours at 37°C. Plaques were then counted and the
titer calculated.
According to the Health Department requirements, all workers were
immunized against polio.
37
-------
SECTION VI
HIGH-pH LIME COAGULATION
GENERAL
During the first phase in the investigation of metals removals, the
Demonstration Plant process configuration was as shown in Figure 11 .
The upflow clarifier (Infilco Densator) was operated as a high-pH lime
coagulation process in this portion of the project, and the target effluent
pH range was 11.3 to 11.5. Work commenced in June 1972 and terminated
in October 1972, but was resumed briefly in November and December of 1973.
The month of July 1972 was excluded because of a change in process configura-
tion, hence, the exact dates for investigation of high-pH lime coagulation
without recarbonation were: June 1-31, August 1-October 31, 1972 and
November 2-December 9, 1973.
During the first three months, i.e., June through August, the sludge
age in the activated sludge system averaged slightly less than 5 days,
allowing only partial nitrification. However following installation of
the Fiscalin aeration equipment and another air compressor in the first
week of August, the ability to nitrify was greatly enhanced, and almost
complete nitrification was maintained thereafter.
The multimedia filters were operated at an average rate of 6.11 m/hr.
(2.5 gpm/ft ). The No. 1 (Neptune-Microfloc tri-media) and the No. 2
(conventional dual-media) filters were alternated during this project. One
filter processed the activated sludge effluent, while the other filtered
the effluent from the chemical treatment process. The No. 1 filter provided
the higher degree of suspended solids and turbidity removal, although the
performance of either filter was sufficient to produce a product water
wi
fi
Co
mi
th turbidities less than 2 NTU, when a properly coagulated effluent was
Itered. Flow from the filter was then pumped through one of the two
rbon columns, which provided a theoretical empty-bed contact time of 37
nutes.
A water quality summary for the high-pH lime coagulation sequence is
presented on Table 10. In spite of large reductions in the concentrations
of various pollutants, there were net increases in TDS, specific conductance,
alkalinity, N02 and NO--N. As discussed later in this section, increases
were also observed in the concentrations of some metals.
One important aspect of the high-pH lime train was that no effort was
made to recarbonate or otherwise neutralize the Densator effluent. The
38
-------
TABLE 10. SUMMARY OF WATER QUALITY DATA FOR THE HIGH-pH LIME COAGULATION
STUDY
Parameter
COD
BOD5
TSS
SC,umho/cm
TDS
NH3-N
Org-N
N02+N03-N
pH, units
T.-Alk.(as CaC03)
P-Alk.(as CaC03)
Std. Plate Count
(per ml)
Total Coliforms
(per 100 ml )
Fecal Coliforms
(per 100 ml)
White Rock STP
Raw Wastewater
(mg/1 )
574
233
254
820
532
20.9
13.9
0.5
7.3
214
0
--
--
—
Final Product
(mg/1)
16
3
7
1349
642
3.7
1.8
9.4
11.5
253
228
6
<2
<2
Removal
(mg/1 )
97.2
98.7
97.2
N/A
N/A
82.3
87.1
N/A
N/A
N/A
N/A
N/A
N/A
N/A
**
N/A =Not applicable
— = Not available
40
-------
o
. .0
"•p *~
(O S-
r— O)
3 XI
o o
U O)
o
CL) I
£ S-
•r- 0)
"T"
I O
-c z:
o>
' 3C CM
o s-
•r- O)
4J J3
3 O
C7>O
O 3
O CD
to
O) •»
00)
O E
S- 3
D- --3
O)
39
-------
scaling tendency of this water was far less than anticipated, although
ultimate neutralization would undoubtedly be a requirement in any similar
full-scale facility.
The bactericidal effect of the high-pH lime process is evident from the
final product water quality of Table 10. No bacteriological counts were run
on the White Rock raw wastewater; we considered such an exercise to be
unecessarily redundant. In becoming final product waters, the only disin-
fection process to which the wastewaters were subjected was the high-pH
process.
The treatment sequence generally performed very well, and product
water quality was consistently good. The COD values observed in the un-
treated wastewaters entering the White Rock STP and the product water at
the Demonstration Plant are shown in Figure 12.
The improvement in water quality is evident. Product water COD values
were stable and ranged from 0 to 25 mg/1 after the early part of August 1972.
Figure 13 is a time-series plot of the effluent nitrogen concentrations
(except organic nitrogen) observed during the high-pH lime coagulation por-
tion of the research effort. It is important to note the lower and more
stable COD values that result when nitrification is well advanced. The
observed COD concentrations were low during the months of August-September
1972 and October, 1973. At those times nitrification was well established;
however, during June 1972 product water COD concentrations as high as 40 mg/1
were observed when nitrification was erratic.
COMPLETELY-MIXED ACTIVATED SLUDGE SYSTEM
The performance of the No. 1 completely-mixed activated sludge (CMAS)
system is summarized in Table 11. The performance of the process can be
characterized as satisfactory, but not outstanding. BOD5 and TSS concentra-
tions in the effluent averaged 28 and 27 mg/1, respectively; however, the
effluent COD averaged 91 mg/1 which was about twice the desired COD con-
centration of 50 mg/1.
Nitrification was not consistent, as the average of 4.4 mg/1 of NHo-N
in the effluent indicates. Likewise, the effluent nitrite-nitrate nitrogen
concentration was lower than one would obtain from a process that was achiev-
ing complete nitrification. The lack of a stable, nitrifying microbial
population accounts for the high NH3-N concentrations, and indirectly result-
ed in high effluent COD concentrations.
Table 12 summarizes the hydraulic operation of the system and more
significant process control parameters. The hydraulic operation of the
clarifier is very conservative by conventional design criteria, but it is
a small basin that was being operated on a nitrifying system. Under these
conditions conservative design and operation are necessary if adequate
liquid/solids separation is to occur.
Although the combination of a sludge age of 5 days and a temperature of
41
-------
400
300
200
100
WhiteRock
Raw Wastewater
JUNE
AUG SEPT
.1972
OCT
DEC
1973
Figure 12. Raw wastewater and product water COD values observed
during the high-pH lime coagulation study.
42
-------
UJ
o
o
o
J:UNE
Effluent N
c.
Effluent NH3-N
Effluent COD
& N03-N
AU6
SEPT
OCT
1972
NOV
1973
Figure 13. Effluent nitrogen concentrations observed
during the high-pH lime coagulation study.
43
-------
TABLE 11. PERFORMANCE SUMMARY OF THE COMPLETELY MIXED ACTIVATED
SLUDGE SYSTEM HIGH-pH LIME COAGULATION STUDY.
Parameter
COD
TOC, Soluble
BOD5
TSS
TDS
SC,ymho/cm
NH3-N
Org.-N
N02+N03-N
N02-N
Total -P
pH, units
T-Alk.(as CaC03)
P-Alk.(as CaC03)
Std. Plate Count
(per ml )
Total Coliforms
(per/100 ml)
Fecal Coliforms
( per 100 ml)
Activated
Sludge Influent*
(mg/1)
273
42
88
124
532
819
16.1
9.5
0.77
0.06
11.0
7.3
210
0
5.2xl06
6.2xl07
6.3xl06
Activated
Sludge Effluent
(mg/1)
91
19
28
27
541
739
4.4
4.9
8.2
0.59
9.5
7.1
114
0
2.0xl05
1.4xl06
S.OxlO4
Reduction
(percent)
66.7
54.8
68.2
78.2
N/A**
9.8
72.7
48.4
N/A**
N/A**
13.6
N/A**
45.7
N/A**
96.1
97.7
98.7
*White Rock effluent.
**N/A = not applicable.
44
-------
TABLE 12. PROCESS SUMMARY OF THE COMPLETELY-MIXED ACTIVATED SLUDGE
SYSTEM, HIGH-pH LIME COAGULATION STUDY.
HYDRAULIC OPERATION
FLOW (Influent)
FLOW (Return)
FLOW (Waste)
Aeration T
Clarifier overflow rate
Weir loading
Clarifier T
11.7 I/sec
(185 gpm)
12.0 I/sec
(191 gpm)
12.3 m3/day
(3271 gpd)
1.62 hour
15.4 m3/day/m2
(377 gal/day/ft2)
3
36.3 m /day-m
(2927 gal/ft-day)
2.30 hours
PROCESS CONTROLS
MLSS
MLVSS
RAS
SVI
Air supplied
D.O.
D.O. Uptake rate
F/M (COD)
F/M (SOC)
F/M (BOD)
Sludge Age
Temperature
2815 mg/1
2136 mg/1
5533 mg/1
186 mg/1
181. 72 I/sec
(385 cfm)
2.8 mg/1
25.3 mg/l-hr
0.574 day"1
0.088 day"1
0.185 day"1
5.0 days
28.3°C
45
-------
The
This
30°C should have been sufficient for promoting the growth of nitrifying
organisms, ammonia oxidation was mediocre. It is possible that increasing
the mean sludge age to approximately 8 days would have significantly
improved the stability of the nitrification process.
The return sludge flow averaged 103 percent of the influent flow. The
high return rate was used to keep the sludge blanket in the final clarifier
as shallow as possible so that the effects of uncontrolled denitrification
could be minimized. Additionally, many months of operation of the activated
sludge system at the Demonstration Plant led to the conclusion that the
best performance was obtained at a one to one recycle ratio.
UPFLOW CLARIFIER
The up-flow clarifier (Infilco Densator) performed as expected during
the high-pH lime coagulation study. Process control and operating variables
of most interest are presented in Table 13, and the performance of the
process is summarized in Table 14.
An average lime dose of 425 mg/1 (as CaO) was required to raise the
activated sludge effluent to a pH of 11.5, and a ferric chloride dose
of 16 mg/1 was applied as a flocculation aid. The clarifier was operated at
a conventional overflow rate of 29.0 rtrVday-m^ (713 gal/day-ft^)
results were good, except that the effluent TSS averaged 46 mg/1. IT
is higher than desired, although it had no significant impact on the
subsequent filtration process. The effluent solids contained almost no
volatile matter, which indicated that the solids carryover from the activat-
ed sludge system was removed in the up-flow clarifier.
The high effluent TSS concentrations observed during this portion of
the study seemed inconsistent with the apparent quality of the water being
produced by the Densator. The on-line surface scatter turbidimeter during
this time generally indicated turbidity values less than 5 JTU, and visual
examination of the water always indicated a product with good clarity. The
high effluent TSS concentrations can thus be attributed to post precipitation
of calcium carbonate from the high-pH effluent.
In addition to removing most of the particulate organic material, the
high-pH coagulation process reduced the soluble TQC by 53. percent. Further-
more the mean organic nitrogen concentrations were reduced by 53 percent,
and the ammonia nitrogen concentration was reduced from 4.4 to 2.6 mg/1.
As expected, increases were observed in TDS, specific conductance, pH,
and alkalinity. The effluent total phosphorus concentration averaged almost
1 mg/1, which was higher than initially anticipated, and was attributed
mostly to the carryover floe in the effluent.
The microbiological data presented in Table 14 show very clearly the
bactericidal effectiveness of high-pH treatment. Total coliforms were
reduced by almost six logs from 1.4xlOb to 3 per 100 ml.
46
-------
TABLE 13. PROCESS SUMMARY FOR THE UP-FLOW CLARIFIER, HIGH-pH
LIME COAGULATION STUDY
Q (influent)
Q (recycle)
Q (waste)
Mixing T
G
Flocculation T
G
Settling T
Clarifier overflow rate
Weir loading
Lime Dose (as CaO)
FeCl3 Dose
29.1 m /day-m
(713 gal/day-ft2)
32.5 tn3/day-m
(2618~gal/ft-day)
425 mg/1
16 mg/1
6.3 I/sec
(100 gpm)
1.8 I/sec
(29 gpm)
8387 I/day
(2216 gpd)
6.6 min.
_1
72 sec.
37 min.
-1
95 sec.
3.4 hours
2
47
-------
TABLE 14 PERFORMANCE SUMMARY FOR THE UPFLOW CLARIFIER,HIGH-pH LIME
COAGULATION STUDY.
Activated
Sludge Effluent
(mg/1)
COD
TOC, Soluble
BOD5
TSS
TDS
SC, iamho/cm
NH3-N
Org. N
N02+N03-N
N02-N
Total P
pH, units
T-Alk. (as CaC03)
P-Alk.(as CaC03)
Std. Plate Count
(per ml)
Total Col i forms
(per 100 ml)
Fecal Col i forms
(per 100 ml)
91
19
28
27
541
739
4.4
4.9
8.2
0.59
9.5
7.1
114
0
2.0xl05
1.4xl06
S.OxlO4
Densator
Effluent
(mg/I)
27
9 :
4
46
680
. 1355
2.6
2.3
9.7
0.95
0.99
11.5
265
228
13
3
'2 "
, Reduction
(percent)
70.3
52.6
85.7
N/A*
N/A*
N/A*
40.9
53.1
N/A*
N/A*
23.6
N/A*
N/A*
N/A*
99.9935
99.9998
99.9975
N/A = not applicable.
48
-------
MULTIMEDIA FILTERS
The No. 1 and No. 2 filters were alternated during the project approx-
imately monthly. During the high-pH lime coagulation portion of the study
the filters were operated as shown in Table l§ , which is the process
summary. The influent flow averaged 1.96 I/sec (31 gal./min.), which
resulted in a filtration rate of 6.11 m/hr. (2.5 gpm/ft2).
TABLE 15- PROCESS SUMMARY FOR THE MULTIMEDIA FILTER, HIGH-pH LIME
COAGULATION STUDY
Q (influent)
Surface loading
Average run time
Washwater consumption
US_
31 gpm
2.5 gpm/sq.ft.
39 hours
2.28%
METRIC
1.96 I/sec.
6.11 m/hr.
This is a rather conservative filtration rate in terms of conventional
wastewater treatment practice, but it closely approximates American filter
operations in the water supply industry. An average run time of 39 hours
resulted from this operation, which is very good considering the influent
TSS concentration was 46 mg/1. The filters were backwashed when the
headless reached 10 feet.
Performance of the filters is summarized in Table 16 , and these data
reflect anticipated performance. Effluent TSS are higher than the 0-2 mg/1
that one should expect, but some of the solids can be explained by the post
precipitation of calcium carbonate. The mean total phosphorus concentration
was reduced from 0.99 to 0.41 mg/1, almost a 60-percent reduction.
ACTIVATED CARBON ADSORPTION SYSTEM
During this research effort only one of the two adsorption columns at
the Demonstration Plant was in operation during any given phase of the
project. In all cases the carbon was virgin (unused) at the start of each
of the three phases; the spent carbon in the adjacent column was removed
and replaced'With virgin"-carbon in preparation for the next part of the . •
project.
Process operation is summarized in Table 17 , and process performance
is summarized in Table 18. The average flow of 1.6 I/sec (25 gpm)
resulted in a surface loading of 4.89 m/hr. (2.0 gpm/sq.ft.). At this
low surface loading and with an applied water that had already been filtered,
49
-------
TABLE 16. PERFORMANCE SUMMARY FOR THE MULTIMEDIA FILTERS, HIGH-pH
LIME COAGULATION STUDY.
Parameter
COD
TOC, Soluble
BOD5
TSS
TDS
SCsumho/cm
NH3-N
Org.-N
N02+N03-N
N02-N
Total P
pH, units
T-Alk. (as CaC03)
P-Alk. (as CaC03)
Std. Plate Count
(per ml)
Total Col i forms
(per 100 ml)
Fecal Coliforms
(per 100 ml)
Densator
Effluent
(mg/1 )
27
9
4
46
680
1355
2.6
2.3
9.7
0.95
0.99
11.5
265
228
13
3
2
Multimedia
Filter
Effluent
(mg/1 )
28
**
4
12
642
1384
2.2
2.2
9.7
1.09
0.41
11.5
274
237
8
<2
<2
Reduction
(percent)
N/A*
N/A*
0.0
73.9
5.6
N/A*
15.4
4.3
0.0
N/A*
58.6
N/A*
N/A*
N/A*
38
N/A*
N/A*
*N/A = not applicable.
**— = not available
50
-------
headless across the carbon column increased very slowly. The average run
time between backwashes was 92 hours. The average headloss at backwashing
was about 3.5 feet, therefore, the backwash of the carbon columns was
principally a function of convienience for the operating personnel and was
not dictated by process hydraulics.
TABLE 17. PROCESS SUMMARY FOR THE ACTIVATED CARBON ADSORPTION SYSTEM,
HIGH-pH LIME COAGULATION STUDY.
Q (influent)
Surface loading
Empty-bed contact time
Average run time
Washwater consumption
1.61 I/sec
(25 gpm)
4.89 m/hr.
(2.0 gpm/sq.ft.)
37 minutes
92 hours
2.01%
The empty-bed contact time of 37 minutes resulted in a COD reduction of
almost 43 percent, which corresponds to a product water COD of 16 mg/1.
Effluent TSS averaged 7 mg/1, which seemed inordinately high since the water
had been filtered through 91.4 cm (36 inches) of filtering media followed by
3.1 meters (10 feet) of granular activated carbon. Therefore, the presence
of a TSS concentration of 7 mg/1 in the effluent seemed very unlikely.
The average turbidity of the product water was 0.4 NTU which is not indica-
tive of 7 mg/1 TSS. Post precipitation of calcium carbonate in the high-pH
effluent was determined to be the cause of the apparently high TSS values'.
The decrease in alkalinity that was observed across the carbon column
is significant. After this portion of the project was concluded layers
about 0.7 meter (2.5 ft.) thick at both the top and bottom of the column
were found to be cemented together with calcium carbonate. Shovels and
picks were required to break-up the carbon for removal from the column. The
direct filtration or adsorption of high-pH waters was necessary, but very
undesirable, during this phase of the project, but the alkalinity data
did indicate that the problem was developing.
After this problem was discovered the .backwash frequency was increased
to every other day. Additionally, the duration of the air scour was
increased from 2 to 5 minutes. Neither of these actions was sufficient to
break-up the encrusted carbon. No increases in headloss were observed at
the time the encrustation occurred, which indicates that the carbon's
porosity was not affected significantly.
51
-------
TABLE 18. PERFORMANCE SUMMARY FOR THE ACTIVATED CARBON ADSORPTION SYSTEM,
HIGH-pH LIME COAGULATION STUDY.
Parameter
COD
TOC, Soluble
BODC
0
TSS
TDS
SCj ymho/cm
NH3-N
Org.-N
N02+N03+-N
N02-N
Total P
pH, units
T-Alk. (as CaCO
P-Alk. (as CaCO
Standard Plate
(per ml)
Total Coliforms
(per 100 ml)
Fecal Coliforms
(per 100 ml)
Multimedia
Filter
Ef f 1 uent
(mg/D
28
*
4
12
642
1384
2.2
2.2
9.7
1.09
0.41
11.5
3) 274
3) 237
Count 8
<2
<2
Carbon
Col umn
Effluent
(mg/1)
16
6
3
7
642
1349
3.7
1.8
9.4
1.28
0.16
11.5
253
228
6
<2
<2
Reduction
(percent)
42.9
N/A**
25.0
41.7
0.0
2.5
N/A**
18.2
3.1
N/A**
61.0
N/A**
7.7
4.0
25.0
N/A**
N/A**
**
Not available.
Not applicable.
52
-------
METAL REMOVALS
Metals data during the period of operation on lime without recarbonation
can be found in Tables 19 through 23 summarizing the activated sludge
influent, activated sludge effluent, upflow clarifier effluent, filter
effluent, and carbon column effluent samples, respectively.
Silver
Silver is present in Dallas'wastewater only at extremely low levels,
too low in fact to confidently assess its behavior during treatment.
Detectable amounts were measured in only 18 percent of the train influent
samples and 26 percent of the train effluent samples. For an undetermined
reason most of the detectable silver appeared during the month of August
1972. Silver was removed by the activated sludge, filtration, and carbon
adsorption processes, but was apparently increased by the chemical treatment
process. In view of the paucity of samples containing measurable concen-
trations of sjlver, it is difficult to justify the increase as being an
actual occurrence. However, the removals perfectly balance the increase, and
the net change through the train is zero. At the low concentrations there
appears to be no significant removal of silver.
The probability distributions for silver in the activated sludge
influent and the carbon column effluents are presented in Figure 14. The
MCL (maximum contaminant level) of the NIPDWR (National Interim Primary
Drinking Water Regulations) is also shown.
Aluminum
Mean removals of 59 percent and 47 percent were obtained for aluminum
through the activated sludge and chemical treatment processes, respectively.
The apparent increase through the filter and carbon column is based on a
single sample set collected on November 27th. The lack of sufficient data
for aluminum on this particular treatment sequence precludes further comment.
Arsenic
The slight increase in arsenic through biological treatment appears to
be a random pattern. The biggest removals occurred in the upflow clarifier,
44 percent based on the mean and 61 percent based on the median. The
reductions were proportional to the influent concentrations (r=0.86).
A removal of 35 percent (mean) or 28 percent (median) occurred through
the multimedia filter; the reduction was proportional to the influent
concentrations (r=0.90). A removal of only 9 percent (mean) to 14. percent
(median) occurred through carbon filtration; ther reduction exhibiting a,
moderate concentration effect (r=0.53). Overall removal of arsenic through
the treatment train was 66 percent (mean) and 70 percent (median). The
drinking water limit for arsenic of 0.05 mg/1 was exceeded on only 4 percent
of the train influent samples, and none of the effluent samples.
53
-------
TABLE 19. ACTIVATED SLUDGE INFLUENT METALS SUMMARY HIGH-pH LIME COAGULATION
STUDY JUNE, AUGUST-OCTOBER 1972, NOVEMBER-DECEMBER 1973
Ag*
Al
As*
B
Ba
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
Si
Sr
v*
Zn
MEDIAN
0.0
0.73
14.5
0.42
0.130
40.0
10.0
0.040
0.190
0.069
1.00
0.50
14.9
5.21
0.070
2.0
110.0
0.090
0.100
11.0
9.0
0.26
3.6
0.305
GEO
MEAN
0.31
0.84
14.8
0.38
0.126
43.8
9.0
0.040
0.193
0.048
1.00
0.38
14.7
5.23
0.69
2.3
107.2
0.089
0.088
8.0
9.5
0.27
4.1
0.299
ARITH
MEAN
0.62
0.92
18.7
0.39
0.149
45.0
13.3
0.045
0.209
0.138
1.05
0.53
14.7
5.27
0.071
2.5
108.8
0.104
0.100
14.2
10.0
0.27
4.6
0.323
tf
1.92
0.46
13.4
0.091
0.083
11.7
13.2
0.021
0.092
0.205
0.31
0.46
0.97
0.64
0.017
1.9
17.8
0.060
0.057
11.9
3.7
0.044
2.7
0.14
MAX
10.0
1.60
51.5
0.52
0.42
80.0
79.0
0.110
0.62
1.04
1.86
1.95
16.2
7.16
0.13
5.0
148.0
0.28
0.45
40.0
17.0
0.33
8.5
0.88
MIN
0.0
0.54
3.6
0.20
0.01
34.0
0.0
0.010
0.07
0.0
0.43
0.0
12.3
3.09
0.05
0.0
57.0
0.03
0.02
0.0
6.4
0.23
2.8
0.12
N
34
6
28
28
44
66
68
43
68
66
66
23
34
56
66
10
43
43
65
23
6
6
4
66
NoterO = std. deviation
MAX = Maximum
MIN = Minimum
N = Number of samples
GEO MEAN = Geometric Mean
ARITH MEAN = Arithmetic Mean
Concentration in mg/1
*yg/l
54
-------
TABLE 20 , ACTIVATED SLUDGE EFFLUENT METALS SUMMARY HIGH-pH LIME TREATMENT
WITHOUT RECARBONATIQN JUNE, AUGUST-OCTOBER 1972, NOVEMBER-
DECEMBER 1973
Ag*
Al
As*
B
Ba
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
Si
Sr
v*
Zn
MEDIAN
0.0
0.35
18.0
0.43
0.075
38.7
7.0
0.040
0.070
0.050
0.29
0.18
14.3
4.98
0.050
1.5
108.0
0.074
0.030
2.0
9.6
0.24
3.2
0.100
GEO
MEAN
0.27
0.33
12.6
0.41
0.065
41.4
6.0
0.029
0.072
0.050
0.30
0.18
14.3
4.93
0.047
1.7
106.7
0.064
0.033
1.7
9.2
0.24
3.7
0.109
ARITH
MEAN
0.26
0.38
19.5
0.42
0.079
42.5
7.6
0.037
0.085
0.080
0.34
0.29
14.3
4.97
0.051
2.1
108.1
0.082
0.044
2.4
9.6
0.24
4.2
0.123
tf
0.75
0.22
17.2
0.077
0.050
10.7
5.1
0.020
0.068
0.115
0.19
0.42
0.90
0.63
0.017
2.0
17.1
0.047
0.042
2.4
2.9
0.03
2.6
0.079
MAX
3.0
0.78
68.0
0.57
0.29
76.0
22.0
0.10
0.56
0.69
1.31
1.90
16.2
6.60
0.096
5.0
150.0
0.200
0.300
9.0
13.5
0.29
8.0
0.45
MIN
0.0
0.16
0.3
0.17
0.0
26.5
0.0
0.0
0.013
0.01
0.10
0.0
12.4
2.43
0.010
0.0
57.0
0.0
0.0
0.0
4.6
0.20
2.4
0.05
N
34
6
27
28
44
66
68
43
68
66
66
23
34
56
66
10
43
43
65
23
6
6
4
66
Concentration in mg/1
55
-------
TABLE 21- UP-FLOW CLARIFIER EFFLUENT METALS SUMMARY HIGH-pH LIME TREATMENT
WITHOUT RECARBONATION JUNE, AUGUST-OCTOBER 1972, NOVEMBER-
DECEMBER 1973
MEDIAN
GEO
MEAN
ARITH
MEAN
or
MAX
MIN
Ag*
Al
As*
B
Ba
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
Si
Sr
v*
Zn
o.o
0.19
7.0
0.42
0.065
153.0
7.5
0.050
0.010
0.029
0.18
0.13
14.5
0.66
0.010
3.4
110.0
0.040
0.040
0.7
13.1
0.23
3.3
0.030
0.42
0.19
6.6
0.40
0.061
148.5
5.0
0.053
0.006
0.023
0.22
0.13
14.4
0.77
0.003
3.0
114.7
0.031
0.032
1.7
12.1
0.22
3.4
0.028
0.94
0.20
10.8
0.41
0.083
156.1
7.6
0.058
0.012
0.056
0.30
0.14
14.5
0.99
0.010
4.2
117.3
0.052
0.039
1.9
13.4
0.23
3.6
0.063
2.00
0.077
11.3
0.077
0.062
49.3
5.5
0.027
0.013
0.121
0.36
0.15
1.1
0.89
0.011
3.5
27.4
0.042
0.027
2.5
6.6
0.07
1.3
0.121
10.0
0.30
46.5
0.57
0.270
313.0
20.0
0.130
0.060
0.770
2.44
0.62
17.4
4.68
0.050
11.5
200.0
0.163
0.100
10.0
20.0
0.32
5.3
0.68
0.0
0.13
0.0
0.22
0.0
53.0
0.0
0.020
0.0
0.0
0.07
0.0
12.7
0.29
0.0
1.0
82.0
0.0
0.0
0.0
7.2
0.15
2.6
0.0
34
4
24
27
42
64
66
41
66
64
64
23
34
54
64
10
41
41
63
23
4
4
4
64
Concentration in mg/1
*U9/1
56
-------
TABLE 22- FILTER EFFLUENT METALS SUMMARY HIGH-pH LIME TREATMENT WITHOUT
RECARBONATION JUNE, AUGUST-OCTOBER 1972, NOVEMBER-DECEMBER 1973
Ag*
Al
As*
B
Ba
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo
Na
Ni
Pb
Se*
Si
Sr
V*
Zn
MEDIAN
0.0
N/A
5.0
0.43
0.070
145.5
7.0
0.050
0.005
0.037
0.07
0.065
14.3
0.46
0.0
2.0
112.0
0.030
0.030
1.75
N/A
0.22
N/A
0.030
GEO
MEAN
0.41
0.29
3.5
0.41
0.067
147.8
5.0
0.050
0.004
0.032
0.08
0.11
14.6
0.52
0.001
1.8
114.9
0.020
0.031
1.4
6.7
0.21
2.4
0.025
ARITH
MEAN
0.94
0.29
7.0
0.41
0.092
156.1
7.4
0.057
0.009
0.066
0.12
0.094
14.7
0.66
0.0061
1.7
117.1
0.042
0.040
2.04
6.7
0.23
2.4
0.048
cr
2.1
N/A
7.6
0.075
0.067
51.0
5.2
0.030
0.013
0.125
0.14
0112
1.8
0.56
0.0084
1.6
24.6
0.038
0.032
1.97
N/A
0.09
N/A
0.076
MAX
10.0
N/A
30.0
0.55
0.23
308.0
24.0
0.13
0.05
0.77
0.68
0.43
23.0
2.96
0.04
4.6
187.0
0.153
0.20
6.5
N/A
0.32
N/A
0.45
MIN
0.0
N/A
0.0
0.20
0.0
43.0
0.0
0.02
0.0
0.0
0.01
0.0
12.6
0.22
0.0
0.0
80.0
0.0
0.0
0.0
N/A
0;14
N/A
0.0
N
33
1
21
23
38
60
61
37
61
60
60
20
33
51
60
9
37
37
58
20
1
3
1
60
Concentration in mg/1
*yg/l
N/A: Not Applicable
57
-------
TABLE 23. CARBON COLUMN EFFLUENT METALS SUMMARY HIGH-pH LIME TREATMENT
WITHOUT RECARBONATION JUNE, AUGUST-OCTOBER 1972, NOVEMBER-
DECEMBER 1973
Ag*
Al
As*
B
Ba
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
Si
Sr
v*
Zn
MEDIAN
0.0
N/A
4.3
0.38
0.080
138.0
7.0
0.050
0.010
0.026
0.050
0.04
14.3
0.37
0.0040
2.0
109.0
0.020
0.040
0.21
N/A
0.28
N/A
0.030
GEO
MEAN
0.37
0.36
3.1
0.37
0.072
136.2
5.0
0.047
0.005
0.027
0.05
0.10
14.4
0.45
0.002
1.8
115.0
0.013
0.035
1.0
8.8
0.22
2.9
0.018
ARITH
MEAN
0.62
0.36
6.4
0.39
0.091
144.9
7.2
0.054
0.0095
0.071
0.064
0.11
14.5
0.58
0.0067
1.9
117.5
0.029
0.043
1.00
8.8
0.23
2.9
0.041
cr
1.21
N/A
6.8
0.13
0.060
49.4
5.0
0.026
0.0103
0.149
0.045
0.17
1.4
0.53
0.0081
1.7
26.8
0.032
0.025
1.42
N/A
0.10
N/A
0.071
MAX
4.0
N/A
22.14
0.85
0.24
286.0
24.0
0.133
0.030
0.77
0.19
0.70
19.0
2.79
0.040
6.0
195.0
0.155
0.10
4.5
N/A
0.30
N/A
0.45
MIN
0.0
N/A
0.0
0.17
0.01
39.0
0.0
0.010
0.0
0.0
0.01
0.0
12.1
0.19
0.0
0.0
84.0
0.0
0.0
0.0
N/A
0.12
N/A
0.0
N
34
1
21
23
39
61
62
38
62
61
61
20
34
52
61
10
38
38
59
20
1
3
1
61
Concentration in mg/1
*yg/l
N/A : Not Applicable
58
-------
Q
UJ
Q
UJ
LU
O
X
LlJ
a;
o
o
o
-o
CQ ^
i |
g; o>
(LI
=5
O1
59
-------
Probability distributions are presented for arsenic at various points
in the treatment sequence in Figure 15.
Boron
Boron was refractory to the high-pH lime treatment train, and the mean
observed influent and effluent concentrations were the same. The only
sustained removals occurred in the carbon column, and even then the boron
reduction was only 6 percent (mean) or 12 percent (median). However, virgin
carbon removed more boron than partially saturated carbon. The removals were
found to be inversely proportional to the unit COD loading on the carbon
(X/M value).
Bari urn
The greatest barium removal occurred in the activated sludge unit, which
achieved 47 percent (mean) and 42 percent (median) reductions. A slight
increase was noted in the Densator, the most likely source being contamina-
tion within the commercial hydrated lime. Barium, a Group II element, is
often not extracted during the refining of native lime. Although the barium
content of the lime shipments was not determined, it would probably have been
at least 0.05 percent, the average concentration found in the Earth's
crust (9). Using this value, there could have been an increase of 0.21 mg/1
barium through the Densator, yet the observed increase was only 0.004
mg/1; hence, some removal is suggested. There was a concentration-dependent
reduction of barium (r=0.55) in the upflow clarifier at influent concentra-
tions greater than about 0.05 mg/1. Removals by filtration and carbon
adsorption were minor, and no significant patterns could be identified.
Probability distributions for treatment sequence influent and effluent
concentrations are presented in Figure 16.
Calcium
A large amount of calcium was imparted to the wastewater during the
high-pH lime treatment, and no positive removals of any significance were
observed. The average lime dose of 425 mg/1 as CaO represents a gross
addition of approximately 303 mg/1 as calcium. Since the observed increase
through the upflow clarifier was 114 mg/1 or 38 percent of the amount added,
it would appear that the remaining 62 percent precipitated as sludge. A
considerable portion of the calcium in the Densator effluent could have
been present in a divalent ionic state (Ca++) as opposed to CaCOo, since no
change in the calcium concentration occurred after filtration. Oddly enough,
there was very little evidence of scale buildup in the piping, filter
plenums, pumps, etc. downstream of the Densator, with the notable exception
of the granular carbon, where visible scale accumulation on the carbon
surface was observed. The total alkalinity was reduced an average of 21
mg/1 as CaCOo, and the observed decrease in calcium concentration as a result
of carbon adsorption was 11.2 mg/1. During the three-month period from
August through October 1972, starting with virgin carbon, a calculated total
of 236 Ib. of calcium was deposited on the bed. The only parameter which
60
-------
UJ
c/>
CO -Q
-P C S-
O
flj
O
O
O)
Q-
JE:
o>
o
c
oj
S-
(8
g
O
JD
"r—
s-
CQ
CQ
O
cu _
S- -M
U_ CO
s-
3
CJ>
CD
O
O
O
61
-------
E
f-t
CO
0
0
JD
S-
o
0
Q
UJ
Q
LU
O
X
o;
•£.
O>
CD
o
>-
en
<
CQ
O
o;
o
+j^
CTT3
O
o
o
o
o
CD
L/6UJ '
62
-------
correlated with calcium was suspended solids, particularly in the multi-
media filter effluent. TSS reductions through the filter also showed a
negative correlation with calcium. The net effect of the treatment train,
then, was the addition of about 100 mg/1 calcium to the water.
Cadmi urn
The highest removals of cadmium occurred during biological treatment,
43 percent (mean) and 30 percent (median). Subsequent physical/chemical
treatment removed only an additional 5 percent (mean), and zero removal
on the basis of the medians. The reductions resulting from physical/
chemical treatment orocesses exhibited a strong concentration effect (r=0.95)»
and were moderately, proportional to the Time dose, and inversely proportional
to effluent turbidity, TSS and total P. No significant removal patterns
were observed in the multimedia filters, but a moderate concentration effect
was seen on the carbon columns. In spite of the above mentioned removal
patterns, the observed removals were not sufficient for consistently achiev-
ing the maximum contaminant level (MCL) of 0.01 mg/1 as promulgated in the
National Interim Primary Drinking Water Regulations (NIPDWR). Approximately
46 percent of the train influent samples and 21 percent of the train effluent
samples exceeded this limit, the highest observed train effluent concentra-
tion out of 62 composite samples being 0.024 mg/1.
The frequency distributions for observed cadmium concentrations in the
treatment sequence influent and effluent are shown in Figure 17. The
median concentration decreased approximately 0.003 mg/1 as a result of
treatment, which would indicate that cadmium was relatively unaffected by
the treatment processes employed during this phase of the project.
Cobalt
Cobalt was one of the refractory metals in the treatment train.
Although the mean concentration dropped 16 percent during biological
treatment, the influent and effluent median concentrations were the same.
Attention is called to the increase in cobalt concentration resulting from
chemical treatment, most likely originating from trace quantities within
the lime and ferric chloride slurries. There was, in fact, a positive
correlation between the chemical dose and cobalt concentration, but no
removal patterns were observed. The net median increase through the
train was 0.010 mg/1.
The probability distributions of the influent and effluent cobalt
concentrations shown in Figure 18 reflect its very refractory nature.
Chromi urn
The mean and median removals of chromium were 59 and 63 percent through
the activated sludge unit, and 86 percent in the Densator. Reductions in
the latter exhibited a strong concentration effect (r=0.99), and were in-
versely proportional to total phosphorus concentration in the process
effluent. Multi-media filtration removed an additional 24 percent (mean) or
63
-------
»•
.x
O
3
DV
Iff
O
O
0)
3
o
Q
LU
O
O
^
CD-
CD
i—i
UJ
Lt-'
>-
3
O
S-
•p
J2
s-
3,
D*
o
0
o
0
O
L/Brt 'NOIl\/yiN30NOO
64
-------
H-
LU,
o
i— O
oo
*
. J3
•<-> S-
o
C3>
tS
0
O
Q
UJ
0
X
Qi
O
o-
CD
LU
GO
Q-
C7>
ns
O
£-
•!->
J3s
CQ
m
O
3 >>
CTT3
Ll_ CO
00
O)
S-
3
CD
O
O
o
o
o
•
CD
L/6UI *NOIJA/HJ>I33NOO
65
-------
50 percent (median), while subsequent carbon filtration failed to remove any
chromium. All of the train influent samples exceeded the drinking water MCL
of 0.05 mg Cr/1, while only 3 percent of the Densator effluent samples and
none of the train effluent samples exceeded the limit.
As the data in Figure 19 clearly indicate, chromium was effectively
removed by the treatment sequence used. It is significant to note that the
biological process was very effective in removing chromium.
Copper
Most of the copper was removed by the activated sludge and high-pH
lime treatment, with little or no subsequent removal through filtration and
carbon adsorption processes. A removal of 42 percent (mean) or 28 percent
(median) occurred through the activated sludge process, and a 30 percent
(mean) or 41 percent (median) removal occurred in the Densator. Reduction
in copper by high-pH lime clarification exhibited a concentration effect,
and was proportional to both the effluent methyl orange and phenolphthalein
alkalinities. However, removals were poor when effluent turbidity exceeded
about 3 NTU. During filtration, the mean and median concentration increased
by 18 percent and 28 percent, respectively. The reason for the observed
increase in copper concentrations has not been positively identified, how-
ever, corrosion of bronze piping appurtenances seems to be the plausible
explanation.
Iron
Each unit in the high-pH lime treatment train significantly reduced the
concentration of iron. Approximately 72 percent of the total iron removed
by the treatment sequence was effected in the activated sludge unit, with
mean and median removals of 67 percent and 71 percent, respectively. The
Densator removed an additional 12 percent (mean) and 40 percent (median),
the reductions exhibiting a concentration effect. Occasionally turbidity
and suspended solids breakthroughs in the filters caused a decline in iron
removal. On average, the multimedia filters removed about four times as
much iron as the Densator, with mean and median removal efficiencies of
59 percent and 60 percent, respectively. The reductions exhibited a strong
concentration effect (r=0.997), although there was no correlation with
suspended solids or turbidity. Removal by carbon adsorption averaged 48
percent (mean) and 29 percent (median), exhibiting a strong concentration
effect (r=0.95). It would appear that extreme values were damped better
in the carbon column than in the multimedia filter. The Secondary Drinking
Water Regulation recommends a MCL of 0.3 mg/1, and this concentration was
exceeded by all of the train influent samples and none of the train effluent
samples.
Mercury
Mercury was effectively removed by each process in the treatment train.
Removal through the activated sludge unit averaged 45 percent and 64
percent based on the mean and median values, the respective Densator removals
66
-------
1.000
0.100
LU
O
52:
O
0.010
0.001
\
Act. SI. Inf.
Act. SI. Eff.
Dens. Eff.
Carb. Col. Eff.
V
2 5 10 15 20 30 40 50 60 70 80- 85 90 95
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 19. Frequency distributions for chromium, high-pH
lime coagulation study.
98
67
-------
being 51 percent and 31 percent. Reductions through the Densator exhibited
a concentration effect (r=0.86). Mean and median removals through the
multimedia filters were 35 percent and 48 percent, respectively* the
reductions exhibiting a concentration effect (r=0.80). The mean concentra-
tion of mercury increased by 17 percent through the carbon column, but the
median concentration declined by 38 percent. Although the influent and
effluent distribution patterns are very similar, the effluent skews slightly
more toward the high side, which would explain the inequity between the mean
and median removals. Reductions through the carbon column exhibited a non-
linear concentration effect, with a cut-off reduction of around 0.15 yg/1.
The maximum train influent concentration was only 1.95 ug/1, just below the
MCL of 2.0 ug/1 established in the NIPDWR. Consequently, the train effluent
quality was well within the limitation established for drinking water.
Frequency distributions for selected points in the Pilot Plant process-
ing are presented in Figure 20.
Potassium
As anticipated potassium was not significantly removed by any of the
unit processes. The greatest removals occurred in the activated sludge
system, 2.5 percent by means and 4.0 percent by median. The remaining
units in the treatment train effected only minor, and probably random,
changes in the concentration.
Magnesium
Removal of magnesium through the activated sludge unit was only 5 per-
cent (mean) or 4 percent (median), compared to 80 percent (mean) or 87 per-
cent (median) through the Densator. The probability distribution pattern
of magnesium was significantly altered as a result of chemical treatment,
and the reductions through the Densator exhibited a concentration effect
(r=0.56). Effluent turbidity and suspended solids were a fair measure of the
amount of magnesium in the effluent, accompanied, however, by a few larger
variations. Effluent total P concentration correlated well with effluent
magnesium (r=-0.69), while the reductions were found to correlate with both
effluent TSS (r=0.50) and total P (r=0.70).
Removal of magnesium by filtration averaged 34 percent (mean) or 30
percent (median). The reductions exhibited a linear concentration effect
(r=0.78), increasing sharply with suspended solids removals of 90 percent
or more. The reductions also decreased as the filter effluent TSS and
turbidity increased.
Carbon filtration effected removals of 11 percent (mean) or 21 percent
(median), the reductions exhibiting a concentration effect (r=0.48). Prob-
ability distributions on the Densator, filter, and carbon effluents were
quite similar, all skewed toward high concentrations, as opposed to the
log-normal pattern of the activated sludge effluent and influent distribu-
tions.
68
-------
10.00
1.00
CT)
*\
O
1-Tl
£
o
o
o.io
0.01
Act. SI. Eff.
Dens. Eff.
— Carb. Col, Eff
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 20. Frequency distributions for mercury, high-pH
lime coagulation study.
69
-------
Manganese
The activated sludge unit removed 28 percent (mean) or 29 percent
(median) of the influent manganese, but the biggest removals were observed
as a result of chemical treatment, 80 percent of both the mean and median
values. Reductions through the Densator exhibited a strong linear concen-
tration effect (r=0.94), but failed to correlate with other water quality
parameters. The median filter effluent concentration was zero; however,
the removal of manganese, based on mean values, was about 40 percent.
Although non-linear,the reductions through the filter exhibited a concentra-
tion effect (r=0.71), and were inversely proportional to effluent TSS. There
was no removal of manganese across the carbon column. The Densator caused
a major shift in the distribution of concentrations, and singly achieved
about 2/3 of the total train reduction. The MCL of 0.05 mg/1 was exceeded
by 86 percent of the activated sludge influent samples, 44 percent of the
activated sludge effluent samples, but none of the succeeding samples in the
treatment train.
Molybdenum
The activated sludge removals of molybdenum were 16 percent (mean) and
25 percent (median). However, the concentrations increased through the
Densator by about 0.002 mg/1, originating probably from contamination in
the lime. A removal of 59 percent (mean) or 41 percent (median) through the
multimedia filters essentially counteracted the increase. The reductions
through the filters exhibited a concentration effect (r=0.89) and possibly
a correlation with several other parameters; however, the paucity of data
precluded conclusive analysis.
The totally refractory behavior of molybdenum is clearly evident in
Figure 21, which presents the probability distribution functions for train
influent and effluent concentrations.
Sodium
No significant sodium removal was expected, and none was observed. The
largest removals of sodium occurred in the activated sludge unit and carbon
column, about 2 percent in both cases based on median concentrations.
Increases occurred in the Densator, originating most likely from the coagul-
ants. Although the mean increase was about 9 mg/1, the median increase was
only slightly less than 2 mg/1. Because the most frequent and largest
increases occurred during August 1972, the problem may be a result of one
chemical shipment. The highest observed train effluent concentration was
195 mg/1, approximately 67 percent over the mean.
For practical purposes sodium is completely refractory to the unit
processes utilized during this study.
Nickel
Nickel was removed about equally well by both the activated sludge unit
70
-------
0)
Q-
0>
Q
LU
Q
LU
LU
O
X
LU
Q;
o
Q >>
LU r—
_1 O
Cf
LU
CD
LU
CO
vt
o
•r—
4->
O -r- "O
I— (/>(/>
I—I "i—
I "O E
t-H O
§ ^TJ
CQ c ns
o a> «—
o; 3-3
Q_ O" O>
o>
O
O
71
-------
and the Densator. Activated sludge removals were 21 percent (mean) and 18
percent (median), while the Densator removals averaged 36 percent (mean)
to 46 percent (median). Reductions through the Densator exhibited a concen-
tration effect (r=0.51), and were inversely proportional to effluent TSS
(r=0.44). Percent nickel removals also correlated inversely with TSS, and
the concentrations of nickel and TSS exhibited a linear correlation (r=0.52).
Further removal of nickel was accomplished by multimedia and carbon filtra-
tion, with an observed decrease of about 0.01 mg/1 in each unit.
The relatively equal division of nickel removal between the biological
process and the AWT process is indicated in Figure 22.
Lead
The largest portion of lead was removed in the activated sludge process,
57 percent (mean) to 70 percent (median), and there were no significant
changes in the concentration throughout the remainder of the train. No
significant patterns were observed on the multimedia filter, but the carbon
column reductions exhibited a concentration effect (r=0.69). The drinking
water MCL of 0.05 mg/1 was exceeded in 83 percent of the train influent
samples and 34 percent of the train effluent samples. The highest train
effluent concentration was 0.10 mg/1, which occurred in two samples, both
collected during June 1972.
Frequency distributions are shown in Figure 23, and definitely indicate
the importance of the biological process for removing lead.
Selenium
The activated sludge unit removed most of the selenium, 83 percent
(mean) and 82 percent (median). High-pH lime coagulation removed 21 percent
(mean) and 65 percent (median), and the reductions exhibited a concentration
effect (r=0.43). Selenium possesses the unusual property of being soluble
in caustic alkali solution, which explains the linear correlation between
effluent selenium and methyl orange alkalinity (r=0.62). Also, there were
negative correlations between the reductions in selenium and methyl orange
alkalinity (r=0.33), phenolphethalein alkalinity (r=0.41), and lime dose
(r=0.38). Following high-pH lime clarification there was an apparent
increase in selenium through the multimedia filters, then a substantial
removal on the carbon columns. From the Densator through the carbon columns
there was 47 percent removal of selenium (mean), or 71 percent (median)
removal. Of all the metals investigated, selenium had the highest average
removal efficiency on activated carbon: 51 percent (mean) to 88 percent
(median). However, there were no apparent correlations between selenium
removal and the operating/process control parameters. The NIPDWR MCL of
0.01 mg/1 was exceeded in 52 percent of the train influent samples, and
in none of the remaining samples.
Frequency distributions for selected sampling locations are presented
in Figure 24.
72
-------
a
'a:
°
trs
uj
-3
1
0
o
0
£3 S.
o- p
sz
o
•»->
OU
V)
-------
a
LU
O
X
UJ
o
a
o-
LLJ
CD
UJ
CQ
U_
O
CQ
•=C
CO
O
Qi
CL
a>
-o
fO
a;
s-
o
c/>
c
o
s_
VI
•o
a> •
3 >>
cr-o
a» =s
s- •»->
u_
oo
CM
a>
3
D>
L/6ui 6
74
-------
Q
UJ
o
a
LU
•
CJ3
LU
O)
Q.
I
D>
e
-------
Silicon
Silicon had negligible removal by the activated sludge process and a
slight increase through the Densator and carbon column. All the significant
removals occurred in the multimedia filters: 50 percent (mean) and 49 per-
cent (median). Unfortunately, there were insufficient data to develop
confidence in the removal efficiencies or correlations with other parameters.
The same may be said of strontium and vanadium.
Strontium
The activated sludge process removed some strontium, 11 percent (mean)
and 10 percent (median). Little or no removal occurred throughout the
remainder of the treatment sequence.
Vanadium
A slight reduction in vanadium occurred in the activated sludge unit,
but the multimedia filter exacted the greatest removals: 33 percent
(mean) and 26 percent (median). Net mean and median train removals were 37
percent and 18 percent, respectively.
Zinc
Zinc was removed chiefly by biological and chemical treatment, the
filter and carbon column only decreasing a few extreme values. Activated
sludge removals were 62 percent (mean) and 67 percent (median), while
Densator removals averaged 48 percent and 70 percent, respectively.
The Densator reductions exhibited a strong concentration effect
(r=0.67), and generally increased as the effluent alkalinities and lime
dose increased and as effluent turbidity decreased.
A mean removal of 24 percent was, observed across the multimedia filter,
although there was no change in the median concentrations. Reductions
through the filter were due primarily to a concentration effect (r=0.65),
particularly at influent concentrations exceeding about 0.10 mg/1. The
same comments apply to the carbon column where a 14 percent removal of mean
concentrations occurred without any change in the median values. The
highest value of zinc ever observed was 0.88 mg/1 in the train influent, well
below the Secondary Drinking Water Regulation recommendation of 5.0 mg/1.
76
-------
SECTION 7
ALUM COAGULATION
GENERAL
f The second phase in the investigation of metals removals by AWT process-
es involved the study of alum coagulation of an activated sludge effluent
followed by multimedia filtration and activated carbon adsorption. A 3-week
start-up period commenced in early November of 1972 to establish steady-
state conditions in the upflow clarifier prior to collecting samples and
acquiring data. The starting date, was November 20, 1972, and this phase
of the project terminated October 30, 1973 -- nearly a year's duration.
The process configuration utilized at the Demonstration Plant for the
alum coagulation phase of the project is shown in Figure 25 The
fundamental difference between the three phases of the project was the
primary coagulant used for chemical treatment. An average alum dose of 130
mg/1 was used to coagulate the activated sludge effluent; however, the low
alkalinity in the activated sludge effluent, resulting from nitrification,
necessitated the feeding of 50 mg/1 lime (as CaO) to permit the coagulation
reactions to proceed to completion.
u Tab™24 Summar1zes the water quality data for the alum coagulation
phase. Effluent quality was excellent, as was the overall performance
of the treatment sequence. COD removal was slightly better than 97 percent
and BOD5 reductions approached 99 percent. '
The mean soluble total organic carbon (SOC) concentration in the
effluent was 4 mg/1, which was at the lower limit of sensitivity for the
analytical instrument being used.
The effluent total dissolved solids (TDS) concentration averaged 491
mg/1, and was not changed significantly by the treatment employed during
this phase of the study. No significant changes in either TDS or specific
conductance was expected, and none was observed. The treatment sequence
proved to be very effective in controlling the suspended solids concentra-
tions in the product water, which averaged only 2 mg/1.
The reductions in microorganism populations observed during the alum
coagulation phase were good, but not comparable to the dramatic kills
obtained with high-pH lime treatment. The geometric mean fecal coliform
density in the effluent from the activated carbon was 630 organisms per
77
-------
o
CU
in
o.
1C
CTl
CD
03
O
O
E
(O
s-oo
o o
O CO
in s-
c/1 S- (J
+•> M- O
S 01
OJ (J S- 4->
(C O .
CO
; i—
: o_
78
-------
•"
• 1
.-•*
.*
..
I'' f
i
« •"
f «
• •_
t
s\"g
C_J
o
h-
_u
00
C_D
ZD
-
1
Z3
rs
LU
2:
ZD
•-3
=a:
D;
0.
eC
o:
=c
03
Jj
=£
•-3
O
UJ
Q
O
CO
CTi
CXI
CTl
O.
O
0 °, ° 0000 O O
°, CxJ 000 CM
1X1 CM i—
un =d- co CM
CM
L/Sui L/6iu L/6'U L/6iu
39v Beams aoo N-SHNI QOD N-eHN
NWniOO N08HV3 'JJ3 39QniS Q31VAI13V
-a
01
4->
o
CD
'a! •
co >,
T3
"0 13
E 4->
rt5 CO
CO O
CD -r-
3 +->
i — ro
ro i —
> ^s
CD
CD re
CT) O
ro o
01 E
a> 3
~a i —
=3 ro
r—
CO OJ
O
S-
IO O
•M Ct_
O
i — to
a. c
o
CO T-
O) -l->
•i- ro
S- S-
a> -t->
(^ c:
CD
CD o
CO
OJ
CD
79
-------
TABLE 24.
OVERALL WATER TREATMENT SUMMARY ALUM COAGULATION STUDY
Parameter
COD, mg/1
SOC, mg/1
BODg, mg/1
TSS, mg/1
TDS, mg/1
SC, iimho/cm
NH3-N, mg/1
Org. N, mg/1
(N02+N03)-N, mg/1
N02-N, mg/1
T-P, mg/1
pH
T. Alk., mg/1
P. Alk., mg/1
Standard Plate Count,
per ml
Total MPN, per 100 ml
Fecal MPN, per 100 ml
-- Not Available
N/A Not Applicable
Raw
Waste-
Water
mg/1
494
--
185
238
--
750
16.7
12.4
0.5
0.1
7.4
211
0
—
—
--
Carbon
Col umn
Effluent
mg/1
13
4
<2
2
491
711
1.7
1.5
8.5
0.2
1.9
7.5
114
3
6.9xl02
6.5xl03
6.3xl02
Reduction
(percent)
97.4
N/A
>98.9
99.2
N/A
5.2
89.8
,87.9
N/A
N/A
N/A
N/A
46:0
N/A
N/A
- . m ••>'•"•
N/A
80
-------
100 ml, which is about a four-log reduction.
COMPLETELY-MIXED ACTIVATED SLUDGE
The No. 1 completely-mixed activated sludge system operated during*
this portion of the research effort with unsettled effluent from the stage
1p!nc*llng filters at the White Rock STP serving as the influent. Table
1 V, 1S a Performance and water quality summary for the twelve month
study period, and all values presented are arithmetic means. Table 26 is
a summary of the hydraulic operation and process control variables for the
^,7 «tSm •
One objective of this phase was to achieve complete, stable nitrifica-
tion in the activated sludge system. The effluent NH--N concentration
averaged 1.7 mg/1, which represented a considerable improvement over the
3.7 mg/1 mean for the lime coagulation phase. Figure 26 presents time
series plots of the sludge age in the activated sludge process, and
effluent concentrations for COD and NH,,-N in both the activated sludge and
carbon column effluents. .- • •
Operation of the process was not conventional in that the theoretical
residence time in the aeration basin was only 4.4 hours (2.05 hours is
based on both influent and return sludge flows). The residence time was
very low for a nitrifing process when evaluated in terms of conventional
design criteria for plug-flow systems. These data indicate that long
?v hJn^^H1^5^! n0t "ecessary Provided that the sludge age is sufficient-
ly high and that adequate oxygen transfer capacity is available. Figure 37
presents the probability distributions for the various forms of nitrogen
in the activated sludge influent. Median values are the 50-percentile
n^lcf^J • T?? data are ln rather good agreement with the arithmetic means
presented in Table 26 For instance the median concentration for NH--N
during this phase was 13.2 mg/1, while the mean concentration was 13.8 mVl
The close agreement between mean and median, and the fact that the data
&2V5 re atlvely straight lines on log-probability paper indicates that
the data closely approximate a normal distribution.
Figure 28 presents frequency distributions for .the upflow clarifier
effluent nitrogen concentrations, and these data closely approximate the
nitrogen concentrations expected in the activated sludge effluent, with the
obvious exception of the organic nitrogen data. The median NH--N concen-
tration shown is 0,6 mg/1, although the mean concentration Tn3the upflow
£nd iS?r uent.was 2.04-wg/l. The 240 percent discrepancy between mean
and median values is not uncommon when evaluating water quality data for
AWT processes.
Figure 29 presents the probability distribution for COD and TOC »
concentrations observed in the activated sludge influent. These values
k
81
-------
TABLE 25. PERFORMANCE SUMMARY OF THE COMPLETELY-MIXED
SYSTEM, ALUM COAGULATION STUDY
ACTIVATED SLUDGE
Parameter Raw
Waste-
Water
COD 494
TOC, Soluble
BOD5 185
TSS 238
TDS 657
SC, ymho/cm
NH3-N 16.7
Org. N 12.4
N02+N03-N
N02-N
Total P
pH, units 7.4
T. Alk. (as CaC03)
P.Alk. (as CaC03)
Standard Plate
Count, per ml
Total Coliforms
per 100 ml
Fecal Coliforms
per 100 ml
Activated
Sludge
Influent
(mg/1)
237
20
62
142
507
750
13.82
10.06
1.2
0.126
9.8
7.3
206
0
l.SxlO6
1.3xl07
l.SxlO6
Acti yated
Sludge
Effluent
(mg/1)
57
10
28
28
498
691
2.36
3.85
8.4
0.076
7.8
7.2
126
0
5.1xl04
4.2xl05
S.lxlO4
Reduction
by the A.S.
System only
(percent)
75.95
50.00
54.84
80.28
1.78
7.87
82.92
61.73
N/A
39.68
20.41
N/A
N/A
38.8
N/A
N/A
N/A
Not Available
N/A Not Applicable
82
-------
TABLE 26, PROCESS CONTROL SUMMARY FOR THE COMPLETELY-MIXED ACTIVATED
SLUDGE SYSTEM, ALUM COAGULATION STUDY
HYDRAULIC OPERATION
Q (influent)
Q (return)
Q (waste)
Aeration T
Clarifier Overflow Rate
Weir loading
10.7 I/sec
(169 gpm)
10.7 I/sec
(169 gpm)
4994 I/day
(1293 gpd)
2.05 hours
3 7
14.0 m/day*m
(344 gal/ft. -day)
33.2 m3/m-day
(2674 gal/ft-day)
PROCESS CONTROLS
MLSS
MLVSS
RAS
SVI
Air supplied
D.O.
D.O. Uptake rate
F/M (COD)
F/M (SOC)
F/M (BOD)
Sludge Age
Temperature
4127 mg/1
2852 mg/1
8248 mg/1
182 mg/1
154.8 I/sec
(328 cfm)
2.6 mg/1
60.6 mg/l-hr.
0.311 day'1 ,
0.026 day"1
0.081 day"1
10.6 days
21°C
(70°F)
83
-------
100.0
10.0
Di
UJ
O
O
O
> 5 10 15 20 30 40 50 60 70 80 85 90
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 27. Probability distributions for different forms
of nitrogen in the activated sludge influent.
84
-------
100.0,
10.0
O
I—I
g
UJ
O
O
1.1
0.1
Org. N
NH--N
2 5 10 15 20 30 40 50 60 70 80 85 90 95 98
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 28. Probability distributions for different
forms of nitrogen in the upflow clarifier
effluent.
85
-------
1000.0
100.0
1.0
10 15 20 30 40 50 60 70 80 85 90 95
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 29. Probability distributions for COD and TOC
concentrations In the activated sludge influent.
86
-------
UPFLOW CLARIFIIR
The performance of the upflow cUrifier (Infilco Densator) Is summa-
rized in Table 27. The chemical treatment process performed reasonably
well, and the process control criteria are summarized in Table 23.
The operation of the Densator wa$ relatively conventional with the
exception of a few significant points/ The mean value of the dissipation
function, G, was 67 sec~ during this portion of the study, which is sub-
stantially lower than the 500 to 1000 sec'4 normally found in mixing basins.
However due to the construction of the Densator, the theoretical residence
time in mixing zone was 8.4 minutes, about 11 times the normal detention
time of 45 seconds. The resulting Gt value of 34,000 is reasonable for
mixing basins.
The average recycle flow of 0.3 liter/sec (4gpm) is somewhat misleading.
Sludge was recycled during portions of the study, but no improvement in
effluent quality was observed as a function of the sludge recycle; therefore,
during most of the alum coagulation phase sludge recycle was not employed.
Figure 29a shows the observed frequency distributions for total phos-
phorus and turbidity values. The median total phosphorus concentration was
2.3 mg/1, which is much higher than one would expect from an. AWT facility,
but the average alum dose of 130 mg/1 would be expected to yield about that
concentration. Factually, budgetary constraints prohibited feeding alum
doses adequate for acceptable phosphorus removal.
The reduced alum dose probably had a detrimental effect on metals
removals, but the extent of the effect of reduced coagulant dose cannot be
accurately assessed. It should be noted that the turbidity data indicate
that excellent coagulation and liquid/solids separation were obtained during
this part of the project.
MULTIMEDIA FILTERS
The No. 1 mixed-media filter was in operation during most of this phase
of the project, although the No. 2 dual-media filter was occasionally used.
The filters were operated at an average flow of 2.5 liters/sec. (39 gpm)
which resulted in a mean filtration rate of 7.58 m/hr. (3.1 gpm/sq.ft.).
The filters were backwashed when the headloss was 9 to 10 feet. The average
run time between filter washes was 40 hours, and the backwash water consump-
tion was 2.31 percent of the product water.
The data in Table 29 indicate that the filters performed well. The
reduction in suspended solids was almost 60 percent, and the product water
clarity was always very good. No statistically significant changes in the
various forms of nitrogen were observed. The microbiological data indicate
that the filters were not very effective in reducing the bacterial popula-
tions; all reductions were less than one-half log.
87
-------
TABLE 27.
PERFORMANCE SUMMARY FOR THE UPFLOW CLARIFIER.ALUM COAGULATION
STUDY
Parameter
COD
TOC, Soluble
BODC
0
TSS
TDS
SC, pmho/cm
NH3-N
Org. N
N02+N03-N
N02-N
Total P
pH, units
T. Alk as CaC03
P. Alk as CaC03
Acti vated
Sludge
Effluent
mg/1
57
10
28
28
498
691
2.4
3.9
8.4
0.08
7.8
7.2
126
0
Standard Plate Count „
per ml 5.1x10
Total Coliforms
per 100 ml
Fecal Coliforms
per 100 ml
N/A: Not Appl
4.2xl05
S.lxlO4
i cable
Densator
Effluent
mg/1
28
8
6
17
532
723
2.0
2.4
8.6
0.10
3.0
7.6
120
5
4.0xl03
2.3xl04
l.SxlO3
Reduction
(percent)
50.9
20.0
78.6
39.3
N/A
N/A
13.6
37.9
N/A
N/A ,
61.5
N/A
4.8
N/A
92.2
94.5
94.2
88
-------
TABLE 28, PROCESS SUMMARY FOR THE UPFLOW CLARIFIER.ALUM
COAGULATION STUDY
Q (influent)
Q (recycle)
Q (waste)
Mixing T
G
Flocculation T
G
Settling T
Clan'fier overflow rate
Weir loading
Alum Dose
Lime Dose
6.1 I/sec
( 97 gpm)
0.3 I/sec
(4 gpm )
75064 I/day
(19,832 gpd)
8.4 minutes
67 sec."1
48 minutes
80 sec. ~1
4.3 hours
28.2 m3/day-m2
(691 gal/ft 2-day)
31.6 nvVm-day
(2540 gal/ft-day).
130 mg/1
50 mg/1
89
-------
100.0
10.0
1.0
0.1
To
»
•
*
*
*
*
2
tal P
•
«
•
»
«
•
(mg
— •» —
4
•
/I)
»
•
»
•
•
•
t
* '
•
*
*
• «
*
•
. '
*
•
Tur
*
*
*
*
•
bidil
•
•
*
:y
*
•
«
«
[NTU
•
* *
9
•
•
)
(
» "
•
«
5 10 15 20 30 40 50 60 70 80 85 90 95 98
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 29a. Frequency distributions for total P and
turbidity values in the upflow clarifier
effluent.
90
-------
TABLE 29,
PERFORMANCE SUMMARY FOR THE MULTIMEDIA FILTER, ALUM
COAGULATION STUDY
Parameter Densator
Effluent
(mg/i)
COD
TOC, Soluble
BOD5
TSS
TDS
SC, pmho/cm
NH3-N
Org. N
N02+N03-N
N02-N
Total P
pH, units
T. Alk as CaC03
P. Alk as CaC03
Standard Plate Count
per ml 4.
Total Coliforms
per 100 ml 2.
Fecal Coliforms
per 100 ml 1.
28
8
6
17
532
723
2.0
2.4
8.6
0.10
3.0
7.6
120
5
Oxl O3
3x1 04
8x1 03
Multimedia
Filter
Ef f 1 uent
(mq/1 )
26
8
3
7
515
710
2.1
2.1
8.6
0.15
2.4
7.5
117
4
2.6xl03
1.3xl04
1.5xl03
Reduction
(percent)
71
0.0
50.0
58.8
3.2
1.8
2.0*
12.1
0.0
42.3*
20.0
N/A
2.5
20.0
35.0
43.5
16.7
91
-------
ACTIVATED CARBON ADSORPTION COLUMN
The No. 3 column was placed in service with a fresh charge of virgin
carbon in November 1972. During portions of the months of February and
March 1973, the No. 4 column replaced the No. 3 column in the treatment
sequence; however, from March until the end of the alum coagulation study
the No. 3 column was used in the treatment sequence. The final X/M was
0.29 Ib. COD applied/lb. carbon, although the COD removal efficiency had not
begun to decline noticeably.
The carbon contactors were operated at an average flow of 1.6 liters/
sec (25 gpm) which resulted in a surface loading of 4.80 m/hr.
(2.0 gpm/sq.ft.) and an empty-bed contact time of 37 minutes. The average
run time between backwashes was 113 hours, and the backwash water consump-
tion averaged 0.85 percent of the product water.
Performance in the column was excellent, as the data in Table 30
indicate. COD reduction averaged 50 percent, as did the soluble TOC con-
centration (SOC). At a surface loading of 4.89 m/hr. (2.0 gpm/ft.2),
the column functioned as an excellent filter and reduced the TSS concentra-
tion by 71 percent to 2 mg/1.
Frequency distributions for the observed concentrations of different
forms of nitrogen are shown in Figure 30. The median ammonia nitrogen
concentration of 0.4 mg/1 indicates that good nitrification was obtained
during this portion of the project. The median nitrate and organic nitrogen
concentrations were 8.2 and 1.4 mg/1, respectively.
Figure 31 presents frequency distributions for COD, TOC, color, and TSS
values in the carbon column effluent. The product water COD concentrations
were relatively low and had a median concentration of less than 12 mg/1.
The COD concentration was less than 20 mg/1 in 85 percent of the samples
analyzed. The curve for TSS data indicates that the down-flow carbon
columns provided additional filtering, and that 95 percent of the samples
had TSS concentrations less than 10 mg/1.
The carbon adsorption columns were not particularly effective in
reducing the population of microorganisms. In general a half-log reduction
was observed through the carbon column.
92
-------
TABLE 30.
PERFORMANCE SUMMARY FOR THE ACTIVATED
CARBON ADSORPTION COLUMN, ALUM COAGULATION
OIUUI ; • -
Parameter
COD
TOC soluble
BOD5
TSS
TDS
SC, ymho/cm
NH3-N
Org. N
N02+N03_N
N02_N
Total P
pH, units
T. Alk. as CaCOs
P.Alk. as CaC03
Std. Plate Count
per ml
Total Coliforms
per 100 ml
Fecal Coliforms
per 100 ml
Multimedia
Filter
Effluent
(mg/1)
26
8
3
7
515
710
2.08
2.10
8.6
.15
2.4
7.5
117
4
2.6 X 103
1.3 X 104
1.5 X 103
Carbon
Column
Effluent
(mg/1 )
13
4
<2
2
491
711
1.71
1.49
8.5
0.22
1.9
7.5
114
3
6.9 X 102
6.5 X 103
6.3 X 102
Reduction
(percent)
50.0
50.0
<33.3
71.4
4.7
N/A
17.8
, 22.0
1.2
N/A
20.8
N/A
2.6
25.0
73.5
50.0
58.0
N/A: Not Applicable
93
-------
loo.o r
10.0
CONCENTRATION, mg/1
o — '
, •
_i O
,X
X
s
/
Org.
/
N
—,
'•
N0?
^
^"
& NC
+*"
y
r
1
)3-N
^^
Y
^
-
>
/-
,^-^
y
'
...
^
—
y
j
ir
^, *
, — , —
X
-N
<**'
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 30. Frequency distributions for observed nitrogen
concentrations in the final product water, alum
coagulation study.
94
-------
100.1
_
o
d
Color
(Pt-Co Units)
•
mg
RA
C
CO
COO
7
•• TOC
/
b
TSS
•
10 15 20 30 40 50 60 70 80 85 90
95
98
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 31. Frequency distributions for selected water quality
parameters in the final product water, alum coagulation
study.
95
-------
METALS REMOVALS . , A
A summary of the metals data obtained during the alum coagulation study
appears in Tables 31 through 35, in the order of treatment sequence.
Silver
All samples analyzed during November. and December had less than the
observable detection limit of l.Oyg/1 Ag; therefore , analysis for silver
was terminated after the first week of January 1973. ;
Aluminum
Analysis for aluminum was initiated on April 1, 1973, because of the
interest in aluminum addition originating from the alum. A relatively large
removal occurred through the activated sludge process, 34 percent and 61
percent of the mean and median concentrations, respectively. In the up-ftow
clarifier, the mean concentration increased 1.7 mg/1, and the median concen-
tration increased 0.8 mg/1. These represent about 14 percent and 7 percent,
respectively, of the amount of aluminum fed (11.8 mg/1 as aluminum). Both
the influent and effluent probability distributions exhibited parallel, log-
normal patterns, indicating the increase was consistent yet also dependent
upon the influent concentration. The effluent concentration was proportion-
al to the alum dose (r=0.40), but correlation is poor. It appears that the
higher pH values resulting from the use of lime was, a dominant factor in in-
creasing the solubility of aluminum, as effluent aluminum was linearly .
proportional to phenolphthalein alkalinity (r=0.78).
The highest removals of aluminum within the treatment train occurred in
the multimedia filter, both the mean and median removals being about 64 per-
cent. Filter reductions exhibited a concentration effect (r=0.94), and were
inversely proportional to effluent TSS and turbidity. The effluent concen-
tration of aluminum was also proportional to TSS (r=0.77) and turbidity
(r=0.51).
Further removals occurred in the carbon column, 23 percent and 65 per-
cent based on mean and median values. Reductions were.again proportional
to the influent concentration (r=0.59). Removal of,aluminum in bpth the
multimedia filter and carbon column was nearly sufficient to offset the
increase in the Densator. Although the mean train effluent concentration
exceeded the mean activated sludge effluent concentration, there was an over-
all train of removal of aluminum amounting to 5 percent (mean) or 78 percent
(median). The large difference between these removal efficiencies arises
from the extreme concentrations imparted by the alum coagulation. Consequent-
ly, the frequency of extreme values was roughly the same in the Densator,
filter, and carbon column effluents. ,
Arsenic
Arsenic was removed by each unit process in the treatment train. Mean
and median removals were 12 percent and 6'percent through the activated
96 •; '
-------
TABLE 31.
ACTIVATED SLUDGE INFLUENT METALS
SUMMARY ALUM TREATMENT
NOVEMBER 1972 - OCTOBER 1973
Ag*
Al
As*
B
Ba
Be*
Ca
Cd
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
Si
V*
Zn
MEDIAN
0.0
0.60
8.5
0.36
0.125
0.010
56.0
1KO
0.024
0.215
0.30 .
1.17
0.36
12.8 .
4.78
0.080
13.0
90.5
0.100
0.110
3.0
9.8,
5.2
0.320
6EO.
MEAN
0.0
0.55
8.4
0.35
0.116-
0.007
57.7
11.0
0.025
0.191
0.22
1.13
0.36
13.2
4.77
0.082
9.3
90.5
0.101 •
0.109
3.3
.10.1
4.7
0.364
ARITH.
MEAN
0.0
0.63
11,4
0.36
0.132
0.014
59.3
12.2
0.029
0.236
0.33
1.28
0.52
13.3
4.80
0.083
20.9
92.0
0.114
O.T21
4.6
10.3
4.8
0.520
or
0.0
0.32
12.6
0.071
0.072
0.013
14.4
6.1
0.015
0.146
0.25
0.93
0.59
1.7
0.60
0.016
34.4
16.5
0.064
0.053
4.6
2.4
1.2
0.66
MAX.
0.0
1.80
100.0
0.54
0.48
0.040
105.0
34.0
0.080
0.750
1.03
7.80
3.20
16.7
6.52
0.130
170.0
123.0
0.38
0.35
19.5
14.5
7.2
4.10
MIN.
0.0
0.18
0.0
0.19
0.03
0.0
31.0
1.0
0.0
0.027
0.01
0.28
0.0
11.8
3.67
0.054
0.0
55.0
0.01
0.03
0.0
7.0
2.1
0.05
J
N
9
50
87
84
74
23
65
104
90
104
65
65
50
9
23
66
26
52
66
103
66
12
17
66
Concentration in mg/1
'• *
: 97
-------
TABLE 32; ACTIVATED SLUDGE EFFLUENT METALS SUMMARY
ALUM TREATMENT NOVEMBER 1972 -
OCTOBER 1973
MEDIAN
MAX.
MIN.
Ag*
+y
Al
As*
B
Ba
Be*
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
j
K
Mg
*3
Mn
Mo*
Na
N1
Pb
Se*
Si
v*
Zn
0.0
0.24
8.0
0.35
0.050
0.0
52.0
5.0
0.020
0.060
0.050
0.32
0.15
12.4
4.55
0.053
2.7
87.0
0.070
0.04
0.5
10.2
4.4
0.120
0.0
0.29
8.1
0.35
0.065
0.003
53.6
5.0
0.022
0.054
0.05
0.30
0.18
12.6
4.64
0.050
2.6
87.7
0.073
0.046
1.2
9.4
4.0
0.135
0.0
0.42
10.1
0.35
0.065
0.005
55.8
5.6
0.026
0.066
0.054
0.33
0.26
12.7
4.67
0.055
4.2
89.2
0.079
0.054
1.1
9.7
4.2
0.156
0.0
0.46
8.0
0.067
0.051
0.008
16.4
2.3
0.017
0.042
0.020
0.13
0.36
1.1
0.53
0.020
4.3
16.7
0.038
0.032
1.6
2.2
1.2
1.56
0.0
2.20
44.5
0.60
0.36
0.03
105.0
11.0
0.08
0.27
0.113
0.80
2.2
14.8
6.20
0.110
15.4
125.0
0.28
0.17
7.5
12.3
6.2
1.06
0.0
0.09
1.8
0.17
0.01
0.0
27.1
0.0
0.0
0.006
0.012
0.09
0.0
11.2
3.92
0.013
0.0
50.0
0.02
0.01
0.0
5.6
1.4
0.04
9
50
90
85
74
23
65
104
90
104
65
65
49
9
23
66
27
52
66
104
68
12
18
66
Concentration in mg/1
98
-------
TABLE 33. UP-FLOW CLARIFIER METALS SUMMARY
NOVEMBER 1972 - OCTOBER 1973
ALUM TREATMENT
Ag*
Al
As*
B
Ba
Be*
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
Si
v*
Zn
MEDIAN
0.0
1.03
5.8
0.37
0.040
0.0
68.0
4.0
0.018
0.019
0.025
0.12
0.09
12.3
4.50
0.040
2.0
88.5
0.060
0.030
0.0
10.8
4.0
0.070
6EO.
MEAN
0.0
1.15
5.2
0.36
0.041
0.001
67.3
4.0
0.023
0.017
0.02
0.13
0.15
12.5
4.54
0.032
2.4
87.5
0.058
0.035
1.2
10.2
3.6
0.108
ARITH.
MEAN
0.0
2.13
6.8
0.37
0.057
0.0005
69.4
4.1
0.027
0.023
0.034
0.16
0.32
12.5
4.58
0.037
3.1
89.3
0.068
0.042
1.0
10.4
3.9
0.109
(T
0.0
4.03
5.2
0.08
0.058
0.002
17.4
1.8
0.018
0.018
0.029
0.11
0.84
1.1
0.59
0.017
2.8
17.7
0.041
0.030
1.5
1.9
1.4
0.13
MAX.
0.0
25.5
25.0
0.55
0.41
0.01
122.0
11.0
0.08
0.108
0.19
0.63
4.40
14.4
5.91
0.090
9.2
132.0
0.33
0.16
7.0
13.3
6.2
0.88
MIN.
0.0
0.14
0.0
0.12
0.01
0.0
36.9
0.0
0.01
0.0
0.0
0.03
0.0
11.2
3.81
0.003
0.0
49.0
0.0
0.01
0.0
7.4
1.4
0.01
N
9
44
77
74
68
19
61
93
84
94
61
61
46
9
23
62
25
48
62
93
62
9
15
61
Concentration in mg/1
*ug/i
99
-------
TABLE 34. FILTER EFFLUENT METALS SUMMARY ALUM
TREATMENT NOVEMEBER 1972 - OCTOBER 1973
Ag*
A1
As*
B
Ba
Be*
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mn*
Na
Ni
Pb
Se*
Si
V*
Zn
MEDIAN
0.0
0.37
5.0
0.39
0.040
0.0
66.0
4.0
0.019
0.013
0.036
0.10
0.10
12.3
4.65
0.030
2.5
89.0
0.060
0.030
0.6
10.0
4.0
0.090
6EO.
MEAN
0.0
0.37
4.8
0.38
0.039
0.004
66.3
4.0
0.023
0.013
0.03
0.11
0.14
12.2
4.63
0.024
2.4
86.1
0.059
0.033
1.3
9.6
3.6
0.095
ARITH.
MEAN
0.0
0.78
6.1
0.38
0.056
0.012
68.6
4.0
0.027
0.019
0.056
0.13
0.25
12.3
4.65
0.030
3.8
88.9
0.070
0.040
1.2
9.7
4.0
0.106
or
0.0
1.41
5.0
0.08
0.061
0.016
18.9
1.7
0,017
0.016
0.123
0.11
0.52
1.3
0.41
0.016
5.1
24.6
0.044
0.03
1.9
1.9
1.5
0.074
MAX.
0.0
8.20
26.8
0.62
0.38
0.03
148.0
13.0
0.080
0.096
0.94
0.69
3.10
14.3
5.44
0.080
25.2
197.0
0.320
0.18
7.5
12.0
5.8
0.43
MIN.
0.0
0.07
0.0
0.19
0.0
0.0
35.0
1.0
0.011
0.0
0.0
0.03
0.0
10.0
3.89
0.0
0.0
51.0
0.0
0.01
0.0
7.2
1.1
0.02
N
8
41
71
68
62
5
57
86
77
86
57
57
43
8
21
58
23
45
58
86
56
9
15
58
Concentration in mg/1
*yg/l
100
-------
TABLE 35. CARBON COLUMN EFFLUENT METALS SUMMARY
ALUM TREATMENT NOVEME.BER 1972 - OCTOBER 1973
'
Ag*
AT
As*
B
Ba
Be*
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
Si
V*
Zn
MEDIAN
0.0
0.13
4.2
0.35
0.040
0.010
64.0
3.0
0.017
0.012
0.030
0.08
0.11
12.0
4.68
0.020
2.5
89.0
0.050
0.030
0.0
10.8
4.0
0.055
GEO.
MEAN
0.0
0.20
4.0
0.33
0.039
0.004
64.2
3.0
0.020
0.011
0.03
0.07
0.13
11.6
4.51
0.016
2.4
84.5
0.046
0.032
1.1
9.7
3.8
0.058
ARITH.
MEAN
0.0
0.60
5.8
0.35
0.056
0.010
66.2
3.8
0.025
0.016
0.039
0.09
0.22
11.6
4.55
0.023
3.2
86.4
0.058
0.040
0.6
9.9
4.3
0.067
0*
0.0
1.75
5.0
0.11
0.061
i 0.012
16.7
2.9
0.018
0.016
0.035
0.05
0.41
1.3
0.54
0.016
3.0
18.0
0.040
0.03
1.0
2.2
2.0
0.044
MAX.
0.0
10.80
24.1
0.78
0.38
0.02
105.0
25.0
0.080
0.095
0.25
0.25
2.25
13.0
5.32
0.081 '
8.8
125.0
0.27
0.16
5.0
13.0
8.3
0,250
MIN:.
0.0
0.07
0.0
0.12
0.0
0.0
36.3
1.0
0.008
0.0
0.01
0.0
0.0
8.9
3.23
0.0
0.0
49.0
0.0
0.0
0.0
6.8
1.7
0.0
N
7
39
69
65
58
4
53
83
74
83
53
53
38
7
20
54
20
41
54
83
54
7
13
54
Concentration in mg/1
*ug/i
101
-------
sludge unit, and 32 percent and 27 percent through the Densator. Reductions
in the Densator exhibited a concentration effect (r=0.75), and were pro-
portional to effluent TSS (r=0.47) and reductions in phosphorus (r=0.12).
The effluent concentrations of arsenic and phosphorus were also proportional
(r=0.57). The latter correlation was anticipated, since both elements are
adjacent members of Group 5A.
The multimedia filter removed 10 percent (mean) or 14 percent (median)
of the influent arsenic, and the carbon column removed 5 percent (means)
or 10 percent (medians), although no removal patterns or correlations were
observed in either unit. The entire alum treatment train effected a net
arsenic removal of 49 percent (mean) to 51 percent (median). Only 1 per-
cent of the train influent samples exceeded the EPA drinking water MCL of
0.05 mg/1, with none of the succeeding samples ever reaching this value.
Frequency distributions for arsenic concentrations in the activated
sludge influent and the product water are shown in Figure 32. Only about 2
percent of the samples in the activated sludge influent exceeded the National
Interim Primary Drinking Water Regulations (NIPDWR) arsenic Criterion and
the median product water concentration was about one-tenth of the NIPDWR MCL.
Boron
Boron was refractory to the alum coagulation treatment sequence, the
net mean and median removals being only 3 percent. Slight increases occurred
through the Densator and filter, offset by removals in the activated
sludge unit and carbon column. It should be noted that the magnitude of
these variations is well within analytical error.
Barium
As the probability distributions in Figure 33 clearly demonstrate, both
biological and chemical (alum) treatment removed some barium. Mean and
median removals were 51 percent and 60 percent in the up-flow clarifier.
Filtration and carbon adsorption had little, if any, effect on barium,
and no removal patterns in any unit process could be identified. The high-
est observed concentration, found in the train influent, was only
0.48 mg/l—well below the EPA drinking water MCL of 1.0 mg/1.
Beryllium
Analysis for beryllium was performed once a week over a 6-month period
from June through October of 1973. Only 15 out of 23 train influent samples
had detectable concentrations, the highest being 0.04 yg/1. Only one out
of nineteen Densator effluent samples contained beryllium in a detectable
concentration, that being 0.01 yg/1 Absolutely no correlations could be
developed in the case of beryllium. There was a net mean reduction of 0.004
Vig/1 through the treatment train, but no change in the median concentrations,
Calcium
Activated sludge removal of calcium was 6 percent and 7 percent of the
102
-------
UJ
O
UJ
UJ
O
X
UJ
-
ea
•
O >>
:0ertsator ari-sing probably
from cobalt contamination in the lime. ... ' '•' -'• '
*;." .• \
. Relatively little change in cobaltvcpncentmtiQhs :6ccur,red through the
multimedia and carbon filters. RemoyaH;'as,Mg'h^asv6^-percent were ob-
tained on virgin carbon, but fell off rap^d1^Ja/:the-^'/irjqrea'5ed ;
Reductions through the carbon also exhibited^ve^wea^concentration :
effect. Train influent and effluent d i stations --are 'a.lrobst identical in
the upper 30-percentile, but very dissimi'lar.inithe 'tbwsr^^ ZO-percentile, as
shown in Figure 35. ,; -.. •:'• ;,^--j,.;••'•-*.• '--i ••, .^ '
Chromium ;; '-/• 'f•',•'.'"• '5"^':''''.•:''.'";•?••>
' • •::"•,.'••--'', -^._ .'*,'.">;«.'.'. ' ,'• I-*';-*
Probability distributions for chromium'concentrations-are s'hown« in
Figure 36. v •-.'..- '•''-',."'.;• •''".•..''•'- ••'-'}"•"" ,.
Mean and median removals of chromium-were bo%h;-72:percent .throuah
-------
, .''" • •' ,'' ^ ;•''[' '->/." ,'"v • -"
1
1
; r
(
E
—I
*
t->
O
=c
1
1
(]_
LU
: •
,1^
O
I
1
V _ pl',
. •'• •'
«»-
«f-
.i ' •**J
• u
. -O
; ^ ' •'
'« » ' " *
J '.»*
• •'"
* " ' '
___
.•:;', i<5v
:»:• •*.:','
^ . •
[ ;' ca
'•"'*' ^^
i|5f;
•;;> „
(
- '• ' * ,
'*;:"i,^
"v-''1^,:
;--:'; V... '•'
• /::V '•''('•• »',•''
'••'.','• • •'.'.'•
'''••'': /
^'..^-£:
J
;<
/
..
/
'"••••
,,
/
*.,"*''
/
X
. ,
X
y
r^*
/
j
£
\f*
' ' .
9 ^^
- _ ._ _._ -
^
r .«***
.••'•'/'
--.-
CT^
Oi
cn
en
E
o^ ro
Q E
1x1 -^
ID LU
CTl jjj «
o -°
o g S_
CT> L ^^ ^5
tl
° ° ^
co Q C
'^ —t "r~i
r- i •*-*
o S 5
P- TL
Q
LU
Q
O
X
O
o
LU
res
E
td
O
S-
o
CD .^
s-
4->
t/1
•o
LU
03
PROBAB
c -a
cr co
s- c
M- O
T3 *+J
O) (C
4J •—
O 3
a> en
i— re
a> o
co o
CO
CD
C7>
-------
CO
4->
O
i—
o « cn
o •§ §
LU 00
- -.--o
I— T3 E
o i-j ,0
CO
O
Q_
tt) (U
3 3
-------
1.000
0.100
g
LU
o
o
o
0.010
0.001
X^
v^
— ^.
'•.
NS^
— —
^5
•***
*..
\
••• «•••••
• ^^B*^H
-s
\
• Ac
• Ac
De
• Cr
^^
^"^X
^*5i
•.
^
•t S'
:t. S
ms.
irb. (
*^
-.
^
1. Ir
1. El
Iff.
^
***%
if
7f.
Fff
^
S
*
*%'
N
s,
N
«
5
\
x
* ^
S
s
***
N
-NS
i —
^
i —
\
N
S
S
L
'
«.
I
\
\
^^
s
N
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 36. Frequency distributions for chromium,
alum coagulation study.
108
-------
n !"** 65,^d 69 percent throu9h alL™ clarification. Reduc-
t0r exh1bited a strong concentration effect (r=
effluent '
remova1:: of a lesser magnitude were accomplished through
™™i «• • perc?nt ^ me,ans and 30 percent by medians. The chromium
removal efficiency increased with TSS removal efficiency, and the reductions
in chromium through the filter exhibited a concentration effect (r=0 5?)
& s)y ^bo"fij^^^" Averaged 13 percent (means) In 8 percen? '
and fa?? nn n5/ ^"^ons exhibiting a weak concentration effect (r-0.41).
and falling off sharply at NH3-N concentrations greater than about 1 5 ma/1
Approximately 95 percent of the train influent samples exceeded the EPA
drinking water MCL of 0.05 mg/1 for chromium and 2 percent of the train
effluent samples exceeded the limit in spite of the relatively high net train
removal of 93 percent (means) or 94 percent (medians) 9
Copper .
The activated sludge unit was the most effective copper removal orocess
with mean and median removals of 84 and 83 percent respectively A urn coaa
plat on removed 37tPercent (mean) and 50 percent (median) of the amount 9"
c=^naL"^^
.t
fluent but one out of 65 samples slightly exceeded the EPA Secondary
Regulations recommendation of 1.0 mg/1. secondary
Iron
Both mean and median removals through the multimedia filters averaaed
17 percent, effluent iron being proportional to effluent TSS (r=0 6lT and
the reductions exhibiting a concentration effect (r-0 57) Carbon adsorb
tion reduced the median concentration by only 0.20 mg/1 but redSctionsfn
anadVe209oedrLOU?hly V ^ ^ ^»j -Sia^^m v° s^
and 20 percent, carbon removed about 70
109
-------
The frequency distributions were log-normal on all sample sites within
the train and the net mean and median removals were both 93. percent.
Approximately 98 percent of the train influent samples contained iron in
exS of 0.3 mg/1, whereas none of the train effluent samples ever reached
that level.
Mercury
As shown bv the probability distributions in Figure 37, most of the
mercury removal ywUhiS?he train was accomplished by the activated sludge
process, 51 percent (mean) to 59 percent (median). Two extreme values n
the Densator effluent resulted in an apparent mean increase of 0.06 vg/1 ,
contrasted with a 40-percent median removal , the reductions exhibited a
roncentrltion effect (r=0.82). Increases in mercury occurred in 20 percent
of the paired simples! always coinciding with low influent concentrations.
^^
tKgh the train. Approximately 4 percent of the train influent samples
exceeded the EPA drinking water MCL of 0.002 mg/1, as did 3 percent of the
train effluent samples.
Potassium
Small amounts of potassium were removed through the alum train. Mean
removals were 5 percent by activated sludge, 1 percent by alum clarification,
2 percent by filtration, 5 percent by carbon adsorption, and 13 percent over-
all.
Magnesium
p
were
Magnesium like potassium, was partially and
through the alum train. The net mean and median removal
only 5 percent and 2 percent, respectively.
Manganese
Almost equal amounts of manganese were removed by every unit in the
treatment sequence. Mean and median removal efficiencies both averaged 34
percent in the activated sludge unit, and 32 percent and 25 percent in the
Sensator the reductions in the latter exhibited a conce ntration effect
(r=0.58). Removal efficiency was directly proportional to pH (£-0.37), the
lowest removals occurring at PH<7.0, the highest removals at pH >8.0.
Multimedia filtration removed an additional 19 percent (median) to 26
percent median), the reductions. exhibited a concentration effect (r=0.52)
and increased s ghtly with TSS removal efficiency. Carbon filtration re-
mSved 24 and 32 plrcent of its mean and median influent concentrations.
no
-------
10.00
1.00
en
3.
I-H
C£
Lul
o
o
o
0.10
0.01
2
Ir !
*.
^k «
^» .
v^
^
5
^
k^
1
^v-
~Y$
*v
10
^S
•^,
V"«
S,
ib
s
s
i*'
\
2U
^
N
^
30
X,
^
40
•1
|
,
N
•
bO
mmmm mm
\
V
60
i^n mtmm
• ^^mmm
\
y-J
.V
. ^.
• ^
».
70
Act
Act
Den
- Car
k
%
A
^
\
:\
80
«.
<;
s.
h.
\
\
85
1. 'J
1 [
Erf.
Col
\
\
90
Inf.
-ff
• ii*
Eff.
\
95
, r_
98
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Ftgure 37. Frequency distributions for mercury, alum
coagulation study.
m
-------
The net tratn removals of manganese came to 73 percent (mean) and 75
oercent (median). It appears, however, that the reductions through the
tSent sequence were relatively constant at about 0,06 mg/1, regardless
of the Influent concentration. The Secondary Regulations recommendation of
0.05 mg/1 was exceeded by 100 percent of the train Influent samples and 4
percent of the train effluent samples.
Molybdenum
The activated sludge unit removed 94 percent of the total molybdenum
removed by the alum treatment sequence. The removal efficiency th™ugh,the
unit was 80 percent (means) and 79 percent [medians). Alum coagulation
reduced the concentration by about 1 wg/l Which represents a removal effic-
iency of 26 percent (means and medians"). The reductions were linearly de-
pendent upon the Influent concentrations (r«Q,70), and declined sharply
as effluent TSS exceeded about 10 mg/1.
During multimedia filtration there were mean and median ^creases of
22 percent and 25 percent, followed by removals of 16 percent and 2 percent
during carbon filtration. Overall, the train effected a mean reyoval of 85
percent and a median removal of 81 percent. Frequency distributions for
molybdenum are shown in Figure 38.
Sodium
Sodium was neither added nor removed during alum treatment. Overall,
there was a 6 percent mean and a 2 percent median removal through the treat-
ment sequence, the highest removals occurred in the activated s udge and
carbon column units. The maximum train effluent concentration was 125 mg/1,
which exceeded the mean concentration by about 45 percent.
Nickel
Almost 2/3 of the nickel removed in the alum train was accomplished .
during biological treatment. The Densator removed 14 percent of the mean- and
median concentrations. Multimedia filtration failed to change the^median
nickel concentration, but there was a slight (3 percent) Increase in the mean
Carbon adsorption removed an additional 17 percent by both mean and median,
the reductions exhibiting a concentration effect (r=0.50). The frequency
distribution on all effluents appeared to be bimodal and devoid of extreme
values, as indicated in Figure 39.
Lead
As those data in Figure 40 indicate, lead was removed chiefly in the
activated sludge unit, but some removal was observed in the Densator. Mean
and median removals were 56 percent and 64 percent through activated sludge,
22 percent and 25 percent through alum clarification. Reductions in the
latter exhibited a concentration effect (r=0.42)5 and in general appeared to
decline with the occasional turbidity and suspended solids breakthroughs.
Multimedia and carbon filtration effected only minor removals, in many cases
zero or negative.
112
-------
100.0
10.0
en
P.
ce:
LoJ
o
o
1.0
0.1
\
^r
^
X
-A
I
\
X
s
V •
^
^
y
s^
S.
V
S,
N
•.
Si
*
X
•v
s^
\
• %
•-..
\
\
^N..
•„
Ml
V-
X
SN
'••*,
M^HMMM
V
\
\
A
— I
_ __
. -LU I
.... [
(
k
\
\
-..
^**.
• •
•
•
\
\
\
\
^
\et
1^4-
ALL
Jen
>rl
1,
\
\
H"
1
~h
»
\
•
«.
\
. si
C1
. o 1
3. t-
D. C(
-\-
— \
»
*
•
*
*
*
t
\
. Inf
E-C-C
. ttt
ff.
3l. E1
L
\
\
\
^
^
rf.
I
10 15 20 30 40 50 60 70 80 85 90 95
98
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 38. Frequency distributions for molybdenum, alum
coagulation study.
113
-------
o
r- O
c/o
•
• -Q
+J S-
O (O
«C O
/
,'
t
2l
0
4->
CO
Q 0>
LU «O
Q 0
UJ O
X 3
LU i—
fO
O "
Q
O
02
IO-
CS
03 ^
2 %
§ S
D- 3
cr
d)
co
O)
C3
o
«o
o
o
o
o
L/6tu '
114
-------
-TJ
3
4->
(A
C
°
fO
or;
UJ =*
Q 2>
LU tO
m O
O "
UJ E
D; """
o «*
Q *
UJ ^
CD w
^^ c
UJ O
U. =
r-i -Q
««
CO
<
CO
0)
U.
i.
D5
•r-
LU
O
O
O
O
L/&U '
-------
The distribution of lead at all sample sites was log-normal with no
extreme values. The EPA drinking water MCL of 0.05 mg/1 was exceeded in 91
percent of the train influent samples and in 22 percent of the train
effluent samples.
Selenium
Mean and median removals of selenium through the activated sludge unit
were 77 and 83 percent, respectively. Chemical treatment effected a :mean
removal of 11 percent while the median concentration dropped from 0.5 .jig/1
to zero, indicating 100 percent removal. Approximately 48 percent of the
Densator effluent samples had nonzero concentrations of selenium i.e.,
greater than the observable detection limit of 1 yg/1.
The mean selenium concentration appeared to increase during filtration.
and drop during carbon adsorption. The former case is a reflection of
extreme values in the upper ten percentile, whereas the mean removal through
carbon adsorption (52 percent) was sufficient to overcome the apparent addi-
tion in the filter., Reductions in concentration through the carbon column
exhibited a well-defined concentration effect (r=0.79). Eleven percent df
the train influent samples exceeded the NIPDWR standard of 0.10 mg/1,
compared to zero percent in all succeeding samples. Frequency distributions
are shown in Figure 41.
Silicon
Little or no removal of silicon occurred within the alum treatment train,
The multimedia filter, which effected a larger removal than the other unit
processes, removed only 10 percent (mean) or 7 percent (median). Overall,
there was a 3-percent mean removal and a 10-percent median addition through
the train.
Vanadium
Vanadium was slightly removed, but only by biological and chemical
treatment. Mean and median removals were 13 percent and 15 percent through
the activated sludge unit, 8 and 9 percent through the Densator, 12 and 23
percent through the entire train. The removals failed to correlate with
other parameters.
Zinc
During the months of December 1972 through May 1973 there were six
occasions when the train influent contained more than 1.0 mg/1 zinc. These
were the only months in the 20-month period in which this occurred. It was
assumed that these incidents were related to industrial activity. In all
cases the activated sludge unit was able to significantly reduce the effluent
concentration to ambient, or average background levels. . The greatest re-
movals within the treatment train occurred in the activated sludge unit, 70
percent by means and 63 percent by medians. Mean and median removals during
chemical treatment were 30 and 42 percent, respectively. Reduction through
the Densator exhibited a concentration effect (r=0.70), and removal
116
-------
efficiency dropped off sharply when effluent turbidity exceeded about 2,0
r IU •
Multtmedta filtration removed about 3 percent tmean) of the zinc, yet
there was a simultaneous increase of Q.Q2 mg/1 (by median). It appears that
there was not a substantial shift In concentrations, although the filter did
a good job of damping extreme values. The reductions In zinc by filtration
exhibited a strong concentration effect (HJ.87).. Removal of zinc by
carbon adsorption was consistent,.and averaged 37 percent more by carbon
adsorption than by multimedia filtration, and the reductions were concentra-
tion-dependent (r»0.77). Also, the removals and reductions of zinc were
proportional to the organic loading on the carbon (X/M), Indicating a
possible organic ligand effect. All sample sites except the train effluent
had varying degrees of extreme values whose distribution patterns were
difficult to identify. However, the final effluent distribution can be
considered log-normal. The maximum concentration of zinc ever observed
was 4.1 mg/1 (train Influent); hence, all samples were well below the
Secondary Regulations recommendation of 5.0 mg/1.
117
-------
100.0
10,0 -
O
t-l
h:
LU
O
O
O
1,0
0.1
1
•
\
(
V
i-
\
s
"S
^
'V™
\
\
s
r~^
*
14
*
V
*
s.
S
\
•
»
t
\
>
*.
\
v
\
\
«
•
•
i
*
\
\
\
^
\
\
' '
«
t
«
*
\
\
\
••A
\'-
\
\
^ia_!i
— •
• •••
s,
^1
x
\
V-.
— Ai
— A<
"• D(
•"** C<
X
k* U •
:t.
3ns
irb
SL
\
\
\
1
SI.
SI.
. Ef
. Co
,
I
\
\
1
Inf.
Eff.
f.
1. Ef1
,., , ,.,
\
h
F.
5 10 15 20 30 40 50 60 70 80 85 90 95 98
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 41. Frequency distributions for selenium,
alum coagulation study.
118
-------
SECTION 8
HIGH-pH LIME COAGULATION AND
SINGLE-STAGE RECARBONATION
GENERAL
This phase of the research effort was necessitated by the absence of
any effluent pH adjustment in the previous high-pH lime coagulation study.
Since it is neither practical nor desirable to discharge an effluent with
a median pH value of 11.5, and since the possible effects of the high-pH
effluent on the performance of the filtration and adsorption processes were
relatively undefined, the decision was made to repeat the high-pH lime
coagulation studies and incorporate single-stage recarbonation.
The alum coagulation study was terminated on October 30, 1973, at which
time the up-flow clarifier was drained, washed down, and promptly returned
to service for operation in the high-pH lime coagulation mode.
The single-stage recarbonation basin was fabricated from-a galvanized
steel tank and a C02 diffusion grid .ws.made of 1/2-tach PVC drilled with
small holes. Pure C02 was metered from a 5,700 kg (12,500 Ib.) receiver
into the bottom of the tank through a pressure regulator and flowmeter.
Feed pressure was always in excess of 50 psi, and rate was adjusted by the
operator as required in order to maintain an effluent pH of 6.5 to 7.0.
The up-flowc^lfier effluent flowed by gravity into the bottom of the
recarbonation basin and over-flowed out the top through a V-notdr weir which
produced a co-current contacting arrangement. The neutralized effluent was
then pumped to the No. 1 multimedia filter.
The recarbonation basin was operated off-line from November through
December 9 for de-bugging purposes. During this time a portion of the high-
pH effluent went ta the recarbonation basin, arid the other portion was
routed to the filter gallery for filtration and carbon adsorption; On
December 10 the recarbonated effluent was piped into the filter gallery for
the first time, and this operating mode was maintained until the termination
of the project. The inclusive dates for the recarbonated high-pH lime
coagulation study were November 2, 1973 through January 31, 1974, although
the filter and carbon column did not come on line until December 10.
Because this was a relatively short period of investigation, the sampling
frequency for metals analyses was increased from twice per week to daily,
and the individual processes were monitored very closely to assure good
process control.
119
-------
The process configuration and mean flows through the respective unit
processes are shown in Figure 42. As in the two previous.phases of the
project, the most significant process change was in the chemical treatment
utilized.
The performance of the treatment sequence was excellent, and Table 36
presents an abbreviated summary of the means of selected water quality
parameters. COD, 8005, and TSS concentrations all exhibited reductions
greater than 98.4 percent. The reductions in bacterial densities were
very good, although not as dramatic as the kills observed during the high-
pH lime coagulation phase. This particular point will be addressed in the
following sections.
Figure 43 presents frequency distributions for the observed COD con-
centrations in the raw wastewater, the activated sludge effluent, and the
final product water. When presented in this manner, the relative improve-
ment in water quality due to secondary treatment, and due to advance waste-
water treatment can be readily assessed. The median COD concentrations
for the raw wastewater, activated sludge effluent, and carbon column
effluent were 500, 45, and 5 mg/1, respectively. These data indicate ex-
cellent process performance and an overall COD reduction of about 99 per-
cent. Extreme values did exist in the observed COD concentrations. The
ninety-five percentile values indicated a product water COD concentration
of 60 mg/1; however, the corresponding COD in the activated sludge effluent
was 120 mg/1, and the raw wastewater COD was approaching 1000 mg/1. Under
these conditions the AWT processes provided a COD reduction of 60 mg/1, or
a §0~pe£cent reduction in the COD of the activated sludge effluent.
Figure 44 presents the probability distributions for observed TSS
concentrations at selected points in the treatment sequence. These data
indicate that control of product water suspended solids was very effective.
The median TSS concentration in the product water was 3 mg/1 (4 mg/1 mean),
and in 40 percent of the samples there were no detectable suspended solids.
COMPLETELY-MIXED ACTIVATED SLUDGE SYSTEM
During this portion of the research effort the performance of the
completely-mixed activated sludge (CMAS) system was not as good,as one
would have liked in that the effluent TSS averaged 44 mg/1,, which resulted
in effluent BODs and COD concentrations of 38 and 65 mg/1, respectively.
The performance is summarized in Table 37, and the hydraulic operation
and process control parameters are given in. Table 38.
During this portion of the study the mean sludge age was 13.2 days and
the wastewater temperature averaged 21°C. The major operational difficulty
experienced was obtaining good liquid/solids separation in the secondary
clarifier, since the mixed liquor was denitrifying and rising to some
degree during most of this phase of the project. The return sludge flow
was increased in an effort to remove the sludge from the anaerobic conditions
in the clarifier bottom as quickly as possible; this action did significantly
improve the problem of rising sludge. A subjective evaluation of the oper-
ation of the activated sludge system indicates that the sludge age was.
120
-------
C\J
o
o
en
c
o
•i—
•P
(0
c
o
1Q
i. C
re -r-
o »p— Q)
•p -P
00 r-i—
D T-
CD
U
OJ
to
E C
O E
.a rs
S-i—
(0 O
oo
0)
cn
to
•p
10
cn
c
c cn
Or—
•r"
•P «
fO r—
cn >,
to £
O re
O D
C
(L) re
•r—
r-
co
i cn
.c i—
cn
•r- '«
X (M
S-
.. 0)
C JO
O E
•r-
rO O
s-z:
D
D") •"
•r- C
4- O
E T-
O -P
O (0
C
10 o
CO J3
<1J S-
O ro
O O
S- 0)
a. s-
CM
s_
cn
•r-
U.
121
-------
TABLE 36.
SUMMARY OF WATER QUALITY DATA FOR THE
HIGH-pH LIME COAGULATION AND RECARBONATION
STUDY.
Parameter
COD
BOD5
TSS
TDS
NH3-N
Org. N
N02 & N03-N
N02-N
Total P
pH, Units
Std. Plate Count
per ml
Total Coliforms
per 100 ml
Fecal Conforms
Raw
Waste
Water
(mg/1)
526
198
245
742
17.0
11.9
<0.5
--
—
7.5
—
__
--
Final
Product
(mg/1)
8
2
4
637
2.7
1.1
10.8
0.12
0.9
6.8
3.3 x 103
1.3 x 102
59
Removal
(percent)
98.5
99.0
98.4
14.2
84.1
90.8
N/A
N/A
N/A
N/A
N/A
N/A
N/A
per 100 ml
N/A: Not Applicable
- : Not Available
122
-------
100.0
100.0
D)
E
Q
o
u
10.0
1.0
Raw Wastewater
Recarb. Eff.
Carb. Col. Eff.
«
7
10 15 20 30 40 50 60 ' 70 80 85 90 95 98
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 43. Frequency distributions for selected COD data;
high-pH lime coagulation and recarbonation study.
123
-------
1000.0
100,0
CO
CO
10.0
1.0
i i i i r L
Raw Wastewater
Act. SI. Eff.
Carb. Col. Eff.
«
Z
2 5 10 15 20 30 40 50 60 70 80 85 90 95 98
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 44. Frequency distributions for selected TSS data;
high-pH lime coagulation and recarbonation study.
124
-------
TABLE 37. PERFORMANCE SUMMARY OF THE COMPLETELY-MIXED
ACTIVATED SLUDGE SYSTEM, HIGH-pH LIME
COAGULATION AND RECARBONATION STUDY.
— - „- _ .
Parameter
COD
TOC, Soluble
BOD5
TSS
TDS
SC umho/cm
NH3-N
Org. N
N02+N03-N
N02.N
Total P
pH, units
T. AIL, as CaC03
P. AIL, as CaC03
Std. Plate Count
per ml
Total Coliforms
per 100 ml
Activated
Sludge
Influent
•(mg/1)
217
27
69
133
557
851
14.5
9.1
0.9
0.11
10.2
7.2
215
0
1.9 x 106
1.4 x 10?
Fecal Coliforms 4.2 x 106
per 100 ml
N/A : Not Applicable '
Activated
Sludge
Effluent
(mg/1 )
65
, 11
38
44
536
762
1.9
4.4
9.9
0.053
7.4
7.1
129
0
9.3 x 103
3.2 x 105
2.5 x 104
Reduction
(percent)
70.0
59.3
44.9
66.9
3.8
10.5
86.9
5,1.3
N/A
51.8
27.5
N/A
40.0
0
99.5
97.7
99.4
125
-------
TABLE 3a. PROCESS SUMMARY FOR THE COMPLETELY - MIXED
ACTIVATED SLUDGE SYSTEM, HIGK-pH LIME
COAGULATION AND RECARBONATION STUDY
HYDRAULIC OPERATION
Q (influent)
Q (influent)
Q (waste)
Aeration T
Clarifier Overflow Rate(Qj)
Weir loading
Clarifier T (Qi + Qr)
Clarifier T (Qj)
PROCESS CONTROLS
MLSS
MLVSS
RAS
SVI
Air supplied
D. 0.
D. 0. Uptake Rate
F/M (COD)
F/M (SOC)
F/M
Sliidge Age
Temperature
8.9 I/sec
gpnr)
11.2 I/sec
(180 gpm)
2797 I/day
(739 gpd)
1.70 hour
11.7 m3/day-m2
(287 gal/ft2 -day)
27.7 m3/day-m
(2231 gal/ft-day)
3.3 hours
7.5 hours
4287mg/l
2915 mg/1
7649 mg/1
196 mg/1
23.32 I/sec
(494 scfm)
4.7 mg/1
25.8 mg/l-hr.
0.288 day -1
0.028 day
0.073 day
13.2 days
219 C
(70°F)
-1
126
-------
probably 3 to 5 days too old during this time, and increased sludge wastage
would probably have improved system stability.
Nitrification was desired during this phase of the prbject; however
^arithmetic mean of 1.9 mg/1 NH3-N in the effluent indicates that
notification was rather erratic. Figure 45 presents probability distribu-
tions for the NH3-N data at selected points in the treatment sequence. It is
interesting to note the large difference between the mean (1.9 mg/1) and the
median NH3-N concentration of 0.2 mg/1. The agreement between the mean
2 & N03-N concentration of 9.9 mg/1, and the median concentration (about
12 mg/1) is much better. The data in Figure 46 Indicate that these data
closely approximate log-normal probability distributions, and the mean and
median values should be reasonably close.
As stated earlier the average COD and BODs concentrations were high, 65
and 38 mg/1, respectively, and the problem was attibuted to poor liquid-
solids separation in the final clarifier. The data in Figure 47 clearly
indicate the affect that effluent TSS had on the effluent COD
127
-------
100.0
10.0
o>
U>l 10 15 20 3d 40 50 60 70 80 85 90 95
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 45. Frequency distributions for selected NH^-N data;
high-pH lime coagulation and recarbonation study.
128
-------
100.0
O)
E •
10.0
o3
0°"
1.0
Act. SI. Eff.
Recarb. Eff.
Carb. Col. Eff.
/
2 5 10 15 20 30 40 50 60 70 80 85 90 95 98
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 46. Frequency distributions for selected nitrate-nitrite
nitrogen data; high-pH lime coagulation and recarbonation
study.
129
-------
o
o
•
o
00
o
o
o
in
o
o
o
o
o
CTi
o
o
00
00
r-
O
o
o
oo
o
o
o
LO
O
CM
o
•
o
O
VD
O
CO
O
•
O
0)
p
00
P 3
•P
i-
O
M- C
O
00 -P
iH™ fO
•P O
S J-
3 13
i— O
f- i-
(U
co c
3 T-
Q 03
Or-
O 3
D)
P to
C O
-------
UPFLOW CLARIFIER
Table 39 summarizes the performance of the upflow clarifier (Infilco
Densator), and Table 40summarizes the operation of the process. During
this phase of the research effort lime was fed as the primary coagulant,
and 12 mg/1 of ferric chloride (as FeCl3) was fed as a floatation aid.
The process performed about as anticipated, with notable exception of
the very high effluent TSS concentration (75 mg/1). The effluent solids
seemed to be unsettled floe, and the VSS concentration was always less
than 5 mg/1 which indicates that the upflow clarifier effectively captured
the solids carryover from the activated sludge system.
The high effluent TSS concentrations can be attributed to a lime dose
that was inadequate for obtaining proper coagulation/flocculation. The
target pH value for operation of the up-flow clarifier was 11.3, since
previous studies had indicated that excellent coagulation was obtained at
that pH value. The average pH value of 10.5 produced an effluent that was
poorly coagulated and very turbid. A lime shipment was received just prior
to the start of this phase that had a CaO content of only 45 percent .
(76 percent was typical) and contained large quantities of grit. The lime
fed poorly, and these difficulties resulted in consistent underfeeding.
Very significant reductions in bacterial densities were observed
during this phase of the project as a result of the high-pH lime coagula-
tion process. Even though the mean (arithmetic) pH of the Densator
effluent was only 10.5 a total coliform reduction of almost five logs was
observed.
SINGLE-STAGE RECARBONATION
The recarbonation basin was operated at an average flow of 3.2 liters/
sec (50 gpm) which resulted in a theoretical residence time of 31 minutes.
With an average CO? dose of 460 mg/1 the basin performed reasonably well
as the data in Table 41 indicate.
The major difficulty encountered in the operation of this process was
the calcium carbonate scaling of the PVC piping used for the influent and
scale build-up on the impeller of the effluent pump. This was a very
major problem, even though the piping and pump were acid washed every
three days with dilute HC1.
The reduction in TSS can be attributed to both resolubilization of
CaC03 and settling. The settling of TSS in the basin made it necessary
to remove sludge by bucket and rope about once a month.
No meaningful changes were observed in the gross organic water quality
parameters or in the mean concentrations of the various forms of nitrogen.
However, substantial changes were noted in the observed bacterial densities.
Approximately a one and one-half log increase was observed for the
standard plate counts and for coliform organisms.
131
-------
TABLE 39.
PERFORMANCE SUMMARY FOR THE UPFLOW
CLARIFIER, HIGH-pH LIME COAGULATION AND
RECARBONATION STUDY
Parameter
COD
TOC, soluble
BOD5
TSS
TDS
SC, ymho/cm
NH3 -N
Org. N
N02+N03 -N
N02 -N
Total P
pH units
T. Alk. as CaC03
P. Alk. as CaC03<
Std. Plate Count
per ml
Total Col i forms
per 100 ml
Fecal Col i forms
per 100 ml
N/A: Not Applicable
Activated
Sludge
Effluent
N/1 )
65
11
38
44
536
762
1.90
4.43
9.9
0.053
7.4
7.1
129
0
9.3 x 103
3.2 x 105
2.5 x 104
Densator
Effluent
(mg/1 )
28
9
4 '
75
633
1084
2.70
1.92
10.7
0.096
1.3
10.5
204
130
18
1 8
6
Reduction
(percent)
56.9
18.2
89.5
N/A
N/A
N/A
N/A
56.7
N/A
N/A
82.4
N/A
N/A
N/A
99.8
99.997
99.98
132
-------
TABLE 40. PROCESS SUMMARY FOR THE UPFLOW CLARIFIER,
HIGH-pH LIME COAGULATION AND RECARBONATION
STUDY
Q (influent)
Q (recycle) .
Q (waste)
Mixing T
G
Settling T
Clarifier overflow rate
Weir loading
Lime dose
dose
6.9 I/sec
(110 gpm)
0.9 I/sec
(14 gpm)
8248 I/day
(2179 gpd)
6.9 min.
_1
67 sec.
3.5 hours
31.9 m3/day-m2
(784 gal/ft2-day)
35.8 m3/day-m
2880 gal/ft-day)
279 mg/1
12 mg/1
133
-------
TABLE 41. PERFORMANCE SUMMARY FOR THE RECARBONATION
BASIN, HIGH-pH LIME COAGULATION AND
RECARBONATION SYSTEM
Parameter
COD
TOC, Soluble
BOD5
TSS
TDS
SC, ymho/cm
NH3 -N
Org. N
N02 +N03 -N
N02 -N
Total P
pH, units
T. Alk. as CaCOo
0
P. Alk. as CaCOo
0
Std. Plate Count
•*
Densator
Effluent
(mg/D
28
9
4
75
633
1084
2.70
1.92
10.7
0.096
1.3
10.5
204
130
18
Recarb
Effluent
(mg/1)
27
9
5
33
598
919
2.36
2.21
9.5
0.118
1.0
7.0
221
12
2.3 x 103
Reduction
(percent)
3.6
0.0
N/A
56.0
5.5
15.2
12.6
N/A
11.2
22.9
23.1
N/A
N/A
N/A
N/A
per ml
Total Colifroms
per 100 ml
Fecal Coliforms
per 100 ml
N/A: Not Applicable
2.3 x 102 N/A
1.7 x 102 N/A
134
-------
The increased bacterial populations were attributed to recontamina-
tion of the basin during those periods when the lime feed to the up-flow
clarifier was interrupted. When this interruption occurred the pH in the
Densator decreased to neutral values and large numbers of organisms passed
to the recarbonation basin. When lime feed to the Densator was reestab-
lished, the high-pH did not influence the microorganisms in
the recarbonation basin since it was always operated at a neutral pH
value.
MULTIMEDIA FILTER . '
The No. 1 multimedia filter (Neptune Microfloc media) was operated
at an average flow of 1.8 liters/sec (29 gpm), which resulted in a filtra-
tion rate of 5.62 m/hr (2.3 gpm /ft2). The filters were backwashed
when the headless was approximately 3 meters, and the average run time
between backwashes was 74 hours. The backwash water consumption averaged
1.03 percent of the filter's effluent flow. The No. 1 filter performed
very well during this portion of the project. Arithmetic means for the
water quality data are presented in Table 42.
The TSS reduction of almost 85 percent resulted in low product water1
turbidity and an average TSS concentration in the filter effluent of 5
mg/1.- It should be noted that filtration did not reduce the average
total phosphorus concentration, indicating that the phosphorus was soluble.
This problem resulted from the inadequate lime dose.
Only very slight reductions in the .COD, soluble TOC, and BODs con-
centrations were observed. The result was anticipated since virtually all
of the solids in the effluent from the recarbonation basin were inorganic.
The specific conductance and the TDS concentration both increased
slightly. The increase in total alkalinity suggests that solubilization
of carbonates to bicarbonates is the most likely explanation for the TDS
and specific conductance increases.
The geometric means for the observed microbiological parameters in-
creased less than one log. This increase indicated that some bacterial
growth was occurring within the filter, but no water quality changes of
consequence appeared to have resulted from the growth.
ACTIVATED CARBON ADSORPTION
The activated carbon adsorption column was operated at an average
flow of 101 liters/sec. (18 gpm) during the last phase of the project
which resulted in a filtration (surface loading) rate of 2.32 m/hr
(1.4 gpm/sq.ft.). The average run time between backwashes was 68 hours
and the washwater consumption was 0.95 percent of the product water.
The empty-bed contact time of 52 minutes produced COD, TOC, and BOD5
removals that were excellent as the data presented in Table 43 indicate.
Figure 48 presents probability distributions for TOC data, and the
difference in median soluble TOC concentrations in the activated sludge
135
-------
TABLE 42.
PERFORMANCE SUMMARY FOR THE NO. 1
MULTIMEDIA FILTER, HIGH-pH LIME
COAGULATION AND RECARBONATION STUDY
Parameter
COD
TOC, soluble
BOD5 ,
TSS
TDS
SC, ymho/cm
NH3 -N
Org. N
N02+N03 -N
N02 -N
Total P
pH, units
T. Alk. as CaCO
3
P, Alk. as CaC03
Std. Plate Count
per ml
Total Col i forms
per 100 ml
Fecal Coliforms
Recarb
Effluent
(mg/1)
27
9
5
33
598
919
2.36
2.21
9.5
0.118
1.0
7.0
221
12
2.3 x 103
•
2.3 x 102
1.7 x 102
Multimedia
Filter
Effluent
(mg/1)
25
7
4
5
645
974
2.78
2.15
10.4
0.131 ,
1.0
6.6
230
0
1.9 x 104
9.4 x 102
2.9 x 102
Reduction
(percent)
7,4
22.2
20.0
84.8
N/A
N/A
N/A .
2.7
N/A
N/A
0
N/A
N/A
N/A
N/A
N/A
N/A
per 100 ml
N/A: Not Applicable
136
-------
TABLE 43 .
PERFORMANCE SUMMARY FOR THE NO. 4 CARBON
COLUMN,HIGH-pH LIME COAGULATION AND
RECARBONATION
Parameter
s
COD
TOC, soluble
BOD5
TSS
TDS
SC, jj mho/ cm
NH3 -N
Org. N
N02 +N03 -N
N02 -N
Total P
pH, units
T. Alk. as CaC03
P. Alk. as CaC03
. Std. Plate Count
per ml
i Total Coli forms
per 100 ml
Fecal Coli forms
per 100 ml
N/A: Not Applicable
'
Multimedia
Filter
Effluent
(mg/1)
25
7
4
5
645 -
974
2.78
2.15
10.4
0.131 .
1.0
6.6
230
0
1.9 x 104
9.4 x 102
2.9 x 102
'137
Carbon
Column
Effluent
(mg/1)
8
3
2
4
637
981
2.73
1.11
10.8
0.117
0.9
6.8
236
0
3.3 x 103
1.3 x 102
59
-
Reduction
(percent)
68.0
57.1
50.0
20:0
1.2
N/A
1.8
48,4
N/A
10.7
10.0
N/A
N/A
0
82.6
86.2
79.7
-------
1000.0
100.0
cr>
E
o
o
10.0
1.0
Act. SI. Inf.
Act. SI. Eff.
— — Carb. Col. Eff.
2 5 10 15 20 30 40 50 60 70 80 85
95 98
PROBABILITY OF BEING EQUALLED OR EXCEEDED -
Figure 48. Frequency distributions for TOC data; high-pH
lime coagulation and recarbonation study.
138
-------
and activated carbon effluents is very evident. The ability of the carbon
to remove color, and to a lesser extent organic nitrogen is shown in
Figures 49 and 50, respectively.
Figure 51 shows probability distributions for total phosphorus
concentrations, and indicates the capability of the AWT processes for re-
moving phosphorus, although improved pH control and lime feed would have
reduced the total phosphorus concentrations to even lower levels.
The half-log reductions observed in bacterial densities are
probably not significant in terms of either process design or facility
operations.
METALS REMOVALS
Summaries of the metals data for the high-pH lime coagulation and
single-stage recarbonation phase of the research program are presented in
Tables 44 through 49 in order of their location in the treatment sequence.
For the purpose of studying metals removal the Densator and the recarbon-
ation basin have been treated as a single chemical treatment process.
Even though the metals data from the upflow clarifier are presented in
Table 46, they were not used in statistical comparisons with the other
two treatment sequences, instead the recarbonation basin effluent con-
centrations were used.
Aluminum
The mean and median removals of aluminum were 32 percent and 54 per-
cent through the activated sludge unit, and 68 percent and 59 percent
through the chemical treatment processes. The reductions in the latter
exhibited a very strong correlation with the metal concentration (r=0.995),'
the optimum pH ranged between 10.0 to 10.5. Both multimedia filtration
and carbon adsorption failed to reduce the average concentrations; how-
ever, the carbon column did significantly reduce the extreme values.
Arsenic
Very little removal of arsenic was observed across the biological
treatment process, but high-pH lime clarification and recarbonation.
effected removals of 81 percent (mean) and 77 percent (median). The
reductions exhibited a very strong concentration effect (r=0.99), but
other correlations were not discovered.
Filtration further reduced the mean concentration by about 0.4 yg/1,
and little or no removal occurred in the carbon column.
Mean and median train removals of arsenic were 84 percent and 79
percent, respectively. The EPA drinking water MCL of 0.05 mg/1 was
exceeded by 9 percent of the train influent samples and none of the
train effluent samples.
The probability distributions shown in Figure 52 vividly illustrate
139
-------
100.0
to
O
o
•p
a.
Di
3
o
o
10.0
1.0
Recarb. Eff.
Carb. Col. Eff.
2 5 . 10 15 20 30 40 50 60 70 80 85 90 95 98
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 49. Frequency distributions for selected color data;
high-pH lime coagulation and recarbonation study.
140
-------
100.0
r- 10.0
•\
D)
E
' «
z:
UJ
CD
O
a:
h-
i — i
•z.
i— i
2:
CD
O
1.0
0.1
,
f •
•
•
f .»*
MM* •
-
• ••••
/
^^B •
• flMMII
,«»•
...
'
- Ac
... DC
- Cc
-""
.=••'
/
/
, '
t. SI. Eff.
carb. Eff.
irb. Col. Ef
.•'"
/
/
^^ 0
.••'
•
x-x
/
f.
*'
s
• • ' **
•
,
s
•••
^
§ *
(
j»
t , •
>-
^•****
.»•*
***"*
10 15 20 30 40 50 60 .70 80 85 90 '95
93
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 50. Frequency distributions for selected organic nitrogen data;
high-pH lime coagulation and recarbonation study.
141
-------
100.0
10.0
D)
E
D.
= <
o
1.0
0.
- Act. SI. Inf.
- Act. SI. Eff.
•• Carb. Col. Eff.
2 5 10 15 20 30 40 50 60 70 30 85 90 95
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 51. Frequency distributions for selected total P data;
high-pH lime coagulation and recarbonation study.
142
-------
TABLE 44. ACTIVATED SLUDGE INFLUENT METALS SUMMARY
HIGH-pH LIME TREATMENT WITH RECARBONATION
NOVEMBER 1973 - JANUARY 1974
Al
As*
B
Ba
Ca .
Cd*
Co
Cr
Cu
Fe
Hg*
Mg
Mn
Na
Ni
Pb
Se*
Si
Sr
V*
Zn
MEDIAN
0.64
13.5
0.35
0.110
64.5
11.5
0.051
0.090
0.121
0.80
0.26
5.25
0.079
96.5
0.126
0.080
3.8
9.4
0.88
3.1
0.180
'GEO.
MEAN
0.64
15.0
0.35
0.105
64.2
12.6
0.050
0.09'4
0.137
0.79
0.25
5.00
0.077
90.9
0.119
.0.082
. 4.3
9.4
0.71
3.1
0.184
ARITH.
MEAN
0.69
21.2
0.36
0.115
66.2
13.7
0.052
0.114
0.182
0.85
0.28
5.14
0.079
93.8
0.136
0.089
7.8
9.7
0.79
3.4
0.203
cr
0.28
22.9
0.097
0.042
18.1
7.9
0.014
0.090
0.156
0.32
0.16
1.05
0.016
21.7
0.068
0.035
8.4
2.4
0.32
1.3
0.098
MAX.
1.60
119.0
0.65
0.25
144.0
38.0
0.081
0.620
0.820
1.63
0.75
7.33
- 0.110
127.0
0.310
0.20
33.0
17.0
1.25
8.5
0.60
MIN.
i
0.25
2.3
0.16
0.01
32.0
2.0
0.028
0.015
0.021
0.37
O.O-
.I .62
0,043
33.0
0.015
0.02
0.0
3.9
0.23
0.5
0.08
N
41
45
42
45
44
46
44
46
44
44
38
43
44
44
44
45
26
41
41
32
44
Concentration in mg/1 (*yg/l)
143
-------
TABLE 45- ACTIVATED SLUDGE EFFLUENT METALS
SUMMARY HIGH-pH LIME TREATMENT WITH
RECARBONATION NOVEMEBER 1973 - JANUARY
1974
MEDIAN GEO. ARITH. Cf
MEAN MEAN
MAX,
MIN.
Al
As*
B
Ba
Ca
Cd*
Co '
Cr
Cu
Fe
Hg*
Mg
Mn
Na
Ni
Pb
Se*
Si
Sr
v*
Zn
0.30
13.4
0.35
0.070
65.0
6.0
0.048
0.044
0.055
0.50
0.17
5.05
0.060
94.0
0.103
0.04
2.2
9.4
0.83
2.8
0.090
0.33
15.6
0.36
0.067
63.1
6.0
0.046
0.041
0.065
0.73
0.19 .
4.91
0.053
88'. 0
0.101
0.043
2.2
9.0
0.63-
2.7
0.110
0.47
20.6
0.36
0.075
64.9
7.3
0.048
0.052
0.077
1.45
0.26
4.96
0.066
90.3
0.106
0.050
2.4
9.3
0.72
2.9
0.133
0.61
18.0
0.074
0.034
16.2
4.1
0.013
0.044 *
0.052
2.68
0.25
0.64
0.046
19.6
0.036
0.034
2.2
2.2
0.30
1.1
0.097
3.5
85.0
0.56
0.24
118.0
23.0
0.073
0.270
0.290
14.90
1.075
5.94
0.250
127.0
0.20
0.17
8.8
13.5
1.19
8.0
0.48
0.02
2.4
0.25
0.005
39.0
1.0
0.020
0.002
0.012
0.10
0.05
3.21
0.003
45.0
0.044
0.01
0.0
4,6
0.16,
1.3
0.05
42
47
43
46
45
47
45
47
45
45
39
44
45
45
45
46
28
42
42
34
45
Concentration in mg/1 .(* yg/1)
144
-------
TABLE 45. UP-FLOW CLARIFIER EFFLUENT METALS SUMMARY
HIGH-pH LIME TREATMENT WITH RECARBONATION
NOVEMBER 1973 - JANUARY 1974
MEDIAN
MAX,
MIN.
Al
As*
B
Ba
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
Mg
Mn
Na
Ni
Pb
Se*
Si
Sr
V*
Zn
0.15
3.9
0.37
0.085
103.0
4.0
0.064
0.009
0.047
0.56
0.10
2.71
0.015
94.0
0.084
0.030
0.5
7.9
0.68
3.3
0.040
0.16
3.3
0.37
0.078
101.2
3.0
0.063
0.011
0.0045
0.54
0.11
2.20
0.015
88.4
0.079
0.028
1.4
8.2
0.59
3.2
0.052
0.18
5.6
0.38
0.086
105.4
3.6
0.067
0.014
0.050
0.97
0.11
2.58
0.019
90.5
0.087
0.029
1.4
8.7
0.66
3.4
0.084
0.12
7.4
0.09
0.032
29.6
1.3
0.023
0.018
0.026
• 2.42
0.10
1.28
0.016
18.9
0.039
0.010
2.0
3.2
0.26
1.0
0.110
0.55
46.5
0.64
0.140
172.0
6.0
0.116
0.120
0.140
16.3
0.40
5.56
0.102
121.0
0.200
0.06
7.3
20.0
1.11
6.2
0.58
0.02
0.0
0.22
0.010
49.0
1.0
0.019
0.003
0.019
0.09
0.0
0.56
0.002
49.0
0.030
0.01
0.0
3.7
0.15
1.6
0.01
40
44
42
44
43
45
43
45
43
43
39
42
43
•43
43
44
28
40
40
34
43
Concentration in mg/1 (*yg/l)
145
-------
TABLE 47. RECARBONATION BASIN EFFLUENT METALS
SUMMARY HIGH-pH LIME TREATMENT WITH
RECARBONATION NOVEMBER 1973 - JANUARY
1974
Al
As*
B
Ba
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
Mg
Mn
Na
Ni
Pb
Se*
Si
Sr
v*
Zn
MEDIAN
0.12
3.1
0.35
0.090
98.0 .
3.0
0.065
0.008
0.050
0.51
0.08
2.50
0.014
94.0
0.082
0.030
1.0
7.9
0.66
3.2
0.430
GEO.
MEAN
0.13
3.2
0.34
0.080
100.6
3.1
0.063
0.009
0.053
0.46
0.15
2.10
0.014
87.1
0.076
0.026
0.1
7.8
0.572
3.1
0.355
ARITH.
MEAN
0.15
3.9
0.34
0.084
106.1
3.4
0.067
0.011
0.073
0.82
0.17
2.40
0.016
90.0
0.083
0.027
1.5
8.1
0.64
3.2
0.654
0
•
0.09
3.0
0.07
0.032
34.1
1.4
0.023
0.009
0.069
1.77
0.27
1.10
0.011
20.9
0.032
0.007
1.8
2.3
0,27
0.6
0.864
MAX. •
0.43
15.3
0.49
0.15
180.0
7.0
0.115
0.050
0.33
11.50
1.48
4.71
0.073
120.0
0.160
0.04
. 5.8
12.2
1.16
4.8
4,10
MIN.
0.03
0.0
0.20
0.0
39.0
1.0
0.022
0.004
0.013
0.04 .
' 0.0
0.53
0.004
34.0
0.030
0.01
0.0
3.4
0.15
1.5
0.02
N
38
40
38
41
41
41
41
41
41
41
36
41
41
41
41
40
26
38
39
32
41
Concentration in mg/1 (*yg/l)
146
-------
TABLE '48. FILTER EFFLUENT METALS SUMMARY HIGH-pH
LIME TREATMENT WITH RECARBONATION
NOVEMBER 1973 - JANUARY 1974
MEDIAN
GEO.
MEAN
ARITH. CT
MEAN
MAX. MIN. N
Al
As*
B
Ba
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
Mg
Mn
Na
Ni
Pb
Se*
Si
Sr
V*
Zn
0.15
3.0
0.35
0.090
85.0
3.0
0.059
0.007
0.059
0.21
0.08
2.66
0.15
80.0
0.068
0.030
1.3
7.8
0.64
3.5
0.390
0.16
2.3
0:35
0.090
91.4
3.0
0.057
0.007
0.063
0.21
0.11
2.27
0.015
78.9
0.065
0.025
1.7
7.2
0.59
3.5
0.375
0.21
3.5
0.36
0.094
96.5
3.3
0.062
0.008
0.082
0.24
0.12
2.61
0.017
82.5
0.072
0.026
1.7
7.6
0.66
3.6
0.532
0.19
3.5
0.08
0.028
33.4
1.4
0.022
0.004
0.092
0.17
0.16
1.25
0.010
23.3
0.029
0.006
2.5
2.4
0.26
0.8
0.503
1.01
19.3
0.57
0.15
171.0
6.0
0.108
0.027
0.560
1.00
0.63
5.38
0.047
118.0
0.150
0.03
11 .0
11.1
1.12
4.9
2.60
0.02
0.0
0.19
0.03
54.0
0.0
0.016
0.003
0.026
0.07
0.0
0.55
0.002
36.0
0.010
0.01
0.0
2.3
0.11
1.2
0.04
33
33
31
33
33
33
33
33
33
33
32
33
33
33
33
33
20
33
33
27
33
Concentration in mg/1 (*yg/l)
147
-------
TABLE 49. CARBON COLUMN EFFLUENT METALS SUMMARY
HIGH-pH LIME TREATMENT WITH RECARBONATION
NOVEMEBER 1973 - JANUARY 1974
MEDIAN
GEO.
MEAN
ARITH.
MEAN
(T
MAX. MIN. N
Al
As*
B
Ba
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
Mg
Mn
Na
Ni '
Pb
Se*
Si
Sr
v*
Zn
0.16
2.8
0.25
0.080
94.0
3.0
0.059
0.005
0.040
0.20
0.03
2.37
0.016
79.0
0.072
0.030
0.0
7.8
0.65
2.2
0.450
0.14
2.3
0.26
0.078
94.9
3.0
0.058
0.005
0.044
0.18
0.10
2.19
0.014
79.0
0.068
0.020
1.2
7.1
0.59
2.0
0.426
0.16
3.5
0.32
0.086
101.2
3.0
0.063
0.006
0.054
0.20
0.08
2.46
0.017
81.7
0.078
0.027
0.7
7.3
0.64
2.0
0.617
0.08
3.2
0.22
0.031
38.4
1.4
0.022
0.003
0.037
0.07
0.13
1.13
0.009
20.1 .
0.046
0.008
1.0
1.9
. 0.24
1.0
0.541
, 0.31
17.7
1.06
0.16
207.0
8.0
0.109
0.011
0.150
0.42
0.60
5.26
0.045
113.0
0.260
0.05
3.0
11.0
1.15
4.0
2.20
0.03 '*
0.0
0.08
o.oi.
55.0
0.0
0.020
0.001
0.014
0.08
0.0
0.53
0.002
44.0
0.026
0.02
0.0
3.9
0.09
0.0 •
0.03
33
33
31
33
33
33
33
33
33
33
32
33
33
33
33
33
21
33
33
27
33
Concentration in mg/1 (*ug/l)
148
-------
c
(0
c
o
•r"
•P
(0
a
cn
^ QJ
UJ £
o .=
X
u
UJ
^ S- W
iSi
°- "O -r-
u c
c o
rs s-
O" fO
0) O
S- 0)
LL.. S-
CM
s-
3
D)
149
-------
that the only process that was effective in removing arsenic was chemical
treatment (high-pH coagulation and recarbonation); biological treatment,
filtration, and carbon adsorption were virtually ineffective in reducing
arsenic concentrations.
Boron
Boron was a refractory element in the treatment sequence. Only in
the carbon column was there significant removal, 11 percent (mean) to 29
percent (median). The reductions through.the carbon'were proportional to
the influent concentrations (r=0.32), and inversely proportional to X/M
(r=0.24). In the latter case, the initial removals were in the range of
70-80 percent on virgin carbon, decreasing to 0-15 percent at X/M's of
0.06 and greater. The frequency distributions of boron in all of the ef-
fluents sampled,were log-normal and near-identical, except for the
carbon column effluent which had a bimodal distribution.
Barium
Barium was significantly removed in only the activated sludge process;
the reductions were 35 percent (mean) and 36 percent (median). A mean
increase of about 0.010 mg/1 occurred during chemical treatment and the
probable cause was contamination in the commercial lime. Little change
in barium concentrations occurred through multimedia and carbon filtra-
tion, although' in its virgin state, the carbon reduced the concentration
of barium by 40-60 percent, declining to zero as the X/M approached about
0.10. None of the samples ever reached or exceeded the EPA drinking
water MCL of 1.0 mg/1.
Frequency distributions for barium concentrations in the activated
sludge influent and the product water are shown in Figure 53.
Calcium
The activated sludge process reduced the mean calcium concentration
about 2 percent but a mean increase of'40.5 mg/1 occurred in the Densator.
There was no removal during the single-stage recarbonation process, as
much of the calcium was converted directly to the soluble bicarbonate
form.
Filtration removed about 9.6 mg/1 of calcium, which resulted in
a 9-percent reduction in the mean concentration. There was also some
degree of correlation between effluent calcium and TSS. The mean and
median increases in calcium through the train were 35 mg/1 and 30 mg/1,
respectively, the maximum observed train effluent concentration was 207
mg/1.
Cadmium
Cadmium was removed both by biological treatment, 47 percent by
means and 48 percent by medians; and by chemical treatment, 53 percent
(means) and 50 percent (medians). The reductions observed through the
150
-------
Q
LU
LU
O
X
O
LU
Cf
LU
CD
CQ
CQ
<
DO
O
•a
c
(0
c
O
O1
(0
O
O
0)
E
a:
a.
i
£L
D)
•i—
.e
E
Z5
i-
o
M-
00
c
o
•r— •
•P >,
3 -O
J2 3
•r- -P
S- 00
00 C
•r- O
TD T-
•P
>> (0
O C
E O
QJ .Q
CT fO
CU O
i- OJ
U. i-
00
LT)
0)
S-
D1
•p-
U.
151
-------
Densator and recarbonation basin exhibited a strong concentration effect ,
(r=0.94). Negligible cadmium removals were observed through multimedia
filtration and carbon adsorption. Fifty-six percent of the train in-
fluent samples contained cadmium in excess of theNIPDWR MCL of 0.01 mg/1;
however, none of the product water samples exceeded the MCL.
The frequency distributions for cadmium concentrations in Figure 54
show rather clearly how effectively the biological and chemical processes
removed cadmium.
Cobalt
The frequency distributions shown in Figure 55 indicate that cobalt
was not removed in the high-pH lime with recarbonation train, due to its
refractory nature and the fact that both the lime and ferric chloride
coagulants contained trace quantities of the element. Chemical analysis
revealed a cobalt content of 20 mg/1 in the lime slurry and 19 mg/1 in the
ferric chloride solution. Based on applied chemical doses, it appears
that more cobalt was derived from the lime feed than the ferric chloride;
however, the calculated cobalt increase based on these analyses was only
about 40 percent of the observed mean increase. There was not, as ex-
pected, a reasonably good correlation between the increase in cobalt and
the chemical dose. Also, there was more than three times as much varia-
tion in the cobalt increases as in the chemical feed rates. The mean and
median increases in cobalt through the treatment train were only 0.011.
mg/1 and 0.008 mg/1, respectively.
Chromium
Chromium was removed progressively to a lesser degree by each process
in the treatment sequence. Mean and median removals were 54 percent and
51 percent by activated sludge, 79 percent and 82 percent by chemical
treatment. Reductions through the Densator exhibited a very strong
correlation with concentration (r=0.98), but there was not enough varia-
tion in the operating and process control parameters to evaluate their
influence on chromium removed.
Further removals of 25 percent (means) and 13 percent (medians) were
observed through multimedia filtration, and the reductions exhibited a
concentration correlation (r=0.78). There was a linear relationship
between filter effluent turbidity and chromium concentration (r=0.23).
Carbon adsorption removed 33 percent (mean) and 29 percent (median), and
very effectively damped extreme values. The reductions through the
carbon column exhibited a strong correlation with concentration (r=0.83),
and chromium removal appeared to decrease substantially when the effluent
COD's exceeded about 15 mg/1. Ninety-one percent of the train influent
samples contained chromium in excess of the EPA drinking water MCL of ,
0.05 mg/1, while none of the train effluent samples even approached this
high a concentration.
Frequency distributions for chromium concentrations are shown in
figure 56.
*
152
-------
E
o
•r—
•P
(0
3
^O)
O
o
S v
Q E
LU lp-
LU "—
*£
LU Q.
Q
LU
UJ
CO
U.
O'
PROBAB
D)
§1
CD
O
S-
O
O 3
•r- -P
•P CO
3
-Q C
•r- O
i--i-
•P -P
CO fO
•r- E .
•o o
JO
>>s-
-U fO
E O
0) 0)
3 S-
CT
0) t3
S- C
11 (0
LO
0)
i.
3
O)
O
o
153
-------
Q
UJ
Q
LU
LU
O
X
UJ
Q
UJ
<
CO
U.
O
>-
CQ
<
CD
O
on
Q.
C
O
•r—
•P
tC
O)
(0
o
U
(1)
E
.0.
'I
^:
D1
(0
J3
O
U
s-
o
M-
oo
C
O
•r-
•P
Z!
S-
•P
CO
C
O
(0
C
O
u ro
c o
OJ OJ
D S-
CT
-------
1.000
\
0.100
en
E
h-
UJ
o
•ZL
O
o
0.010
0.001
Act. SI. Inf.
Act. SI. Eff.
Recarb. Eff.
Carb. Col. Eff.
• t.
\
*
\
\
*
2 5 10 15 20 30 40 50 60 70
85 90 95 93
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 56. Frequency distributions for chromium, high-pH
lime coagulation and recarbonation study.
155
-------
Copper
Copper was removed during biological treatment and high-pH lime
clarification, with mean and median removals of 58 percent and 54 percent
in the former, and 35 percent and 15 percent in the Tatter. The removal
of copper in the Densator exhibited a weak concentration correlation
(r=0.58). However, there were sustained increases in the mean concen-
tration through the recarbonation basin and multimedia filter of 0.023 mg/1
and 0.009 mg/1 respectively. The source of the increase through recarbona-
tion was attributed to brass valves and miscellaneous fittings in the basin
and on the influent line.
Copper removals by the activated carbon amounted to 34 percent (mean)
to 32 percent (median). Reductions in concentration through the column
exhibited a strong concentration correlation (r=0.93), but failed to
correlate with NHgN. Overall, there was-a net removal of 71 percent
(mean) and 67 percent (median) through the treatment sequence. In spite
of the increases in the recarbonation basin and filter, the maximum
observed concentration of 0.82 mg/1 was found on the train influent;
therefore, all samples were consistently well below the Secondary
Regulation recommendation of 1.0 mg/1.
Iron
Analysis of iron was complicated by the occurrence of extreme values
at all sample sites except the train influent and train effluent. Ferric
chloride was metered into the aeration basin from January 4-31, 1974 in
order to promote better settling in the secondary clarifier, where un-
controlled denitrification was causing a serious rising sludge problem.
On 40 percent of the paired samples there was an increase in iron through
the activated sludge unit, resulting in a mean negative removal (or
addition) of 71 percent. At the same time, the median concentration
decreased 37 percent. On three occasions the FeCl3 feed rate was far in
excess of the required dose, which provided an opportunity to study the
removal of slug doses of iron through the remainder of train.
When ferric chloride was fed to the Densator as a flocculation aid,
there was a mean average removal of 43 percent through high-pH coagula-
tion and recarbonation. The reductions exhibiting a strong concentra-
tion correlation (r=0.85).
Multimedia filtration removed an additional 71 percent (mean) or
59 percent (median), and the reductions exhibited a very strong
concentration correlation (r=0.997). The filter effluent TSS concentra-
tion was also a fair estimator of the iron concentration (r-0.40). Carbon
filtration removed a final 18 percent (mean) or 5 percent (median),
eliminating all extreme values. The reductions through carbon exhibited
a reasonably strong concentration effect (r=0.90).
Through the entire treatment train there was a mean iron removal of
77 percent,.and a median removal of 75 percent. All of the train influent
samples exceeded 0.3 mg/1, the Secondary Regulation recommendation,
156 .
-------
compared to 6 percent of the train effluent samples. Both the train
influent and effluent probability distributions had similar log-normal
patterns, and were devoid of extreme values.
Mercury
Mean and median removals of mercury were 6 percent and 35 percent
through biological treatment, 35 percent and 53 percent through chemical
treatment. The reductions in the latter case exhibited a very weak1
correlation with concentration (r=0.31), and the Densator effluent
mercury concentration was proportional to the total P concentration
(r=0.44).
There was little change in the observed mercury concentrations as a
result of multimedia filtration, but a median removal of 63 percent was
observed through activated carbon. Reductions in mercury through the
carbon column correlated with concentration (r=0.53); also, a linear
correlation between effluent mercury and COD (r=0.53) was noted. The
train removals for mercury were 71 percent (mean) and 89 percent (median).
All samples were well below the EPA drinking water MCLof 0.002 mg/1.
57.
Frequency distributions are shown for mercury concentrations in Figure
Magnesium
Only an average of 0.2 mg/1 of magnesium was removed during biological
treatment compared to approximately 2.6 mg/1 removed during chemical
treatment. The chemical treatment processes produced magnesium removals
of 52 percent (mean) or 50 percent (median). Because the average pH
established in the Densator was less than that required for complete
magnesium removal, approximately 2.5 mg/1 remained in the effluent.
There were, however, a few days when the pH was sufficiently high to effect
removals up to 9.0 percent, which provided an opportunity to study the pH-
alkalinity relationships on magnesium removal. The effluent magnesium
concentration was inversely proportional to methyl orange alkalinity
(r=0.65), phenolphthalein alkalinity (r=0.66), pH (r= 0.40), and directly
proportional to turbidity (r=0.86) and total P (r=0.34).
Multimedia filtration failed to remove magnesium. There was, in fact,
a slight increase in magnesium Across the filter, due probably to floe
breakup. The effluent magnesium concentration was directly proportional
to TSS (r=0.31) and turbidity (r=0.53). Carbon adsorption reduced
magnesium concentrations 6 percent (mean) to 11 percent (median); the
reductions exhibited a concentration correlation (r=0.53). The distribu-
tion of concentrations in all process streams was clearly bimodal.
Manganese
Biological removal of manganese averaged only 17 percent (mean) or
24 percent (median), contrasted with mean and median removals of 76
percent and 77 percent through the chemical treatment processes. The
157
-------
LU
•P
fO
)
(0
8
u
0
X
Di
°
O5
o-
CD
z
u
t.
o
0 C/5 >>
. C T3
>: o 3
3S"
iff
CO i +J
C^ <* .Trf
Q- -o o
.Q
(J (0
C U
(1J
0) "D
S- E
11 fO
QJ
S-
3
O5
158
-------
reductions through the Densator exhibited a strong concentration correla-
tion (r=0.97); the reductions declined slightly with increasing effluent
JSS and total P. Little or no removal occurred through multimedia fil-
tration or carbon adsorption. The Secondary Regulations recommended
drinking water limit of 0.05 mg/1 was exceeded in 95 percent of the train
influent samples, and none of the train effluent samples.
Sodium
Sodium was refractory with the exception of a slight, and unexplained,
removal of 8 percent (mean) to 15 percent (median) during multimedia
filtration. There was no significant contamination of the treatment
train arising from the addition of chemicals. Upon analysis, the ferric
chloride slurry contained about 310 mg/1 Na, which would represent a
calculated increase of only 0.006 mg/1 in the Densator. Since the
observed increase was higher, the lime may have also contained trace
amounts of sodium, but not enough to impart more than about 0.2 mg/1
(the observed increase) into the, water. Overall removals were 13 percent
(means) and 18 percent (medians), the maximum train effluent concentration
being 113 mg/1.
Nickel
Mean and median removals of nickel during biological treatment aver-
aged 22 percent and 18 percent, respectively. Removals through chemical
treatment averaged 22 percent (mean) or 20 percent (median), and decreased
slightly with increasing effluent TSS and total P. The reductions ex-
hibited a slight linear concentration effect (r=0.52). No correlations
in nickel removal were observed as a result of the multimedia filtration,
where the mean and median removals averaged 13 percent and 17 percent,
respectively. Nickel was not removed by carbon. Both the mean and median
train removals averaged 43 percent.
Probability distributions for nickel concentrations are given in
Figure 58.
Lead
Those data used to develop the probability distributions shown in
Figure 59 indicate that most of the lead was removed during biological
treatment,'the mean and median removal efficiencies were 22 percent and 18
percent, respectively. Removals through chemical treatment averaged 45
percent (mean) and 25 percent (median), the only good correlation being a
concentration effect on lead reductions (r=0.98). There was essentially
no change in the lead concentration as a result of filtration and carbon
adsorption. The train effluent mean and median concentrations were both
approximately 0.03 mg/1. Although the train removals were not great,
approximately 82 percent of the train influent samples violated the EPA
drinking water MCL of 0.05 mg/1, while none of the train effluent samples
exceeded the limit.
Selenium
159
-------
c
o
•r-
-P
(0
ZJ.
D5
rtJ
OS
UJ
Q (L)
LU £
LU £
O i—
X
LU X
O
<
=>
•o-.
LU
0)
i-i S-
LU o
CO q_
U- (/> >>
O C "O
O 3
>- -I- -P
h- -P CO
CO
CQ
O
a:
CL
jQ C
•r- O
S- T-
•P -P
00 (0
•r- E
•a o
O fO
c u
C1J 0)
3 S-
cr
ID T3
S- C
U. fO
00
in
0)
CD
160
-------
o
O
O
o
o
o
Q
UJ
LL)
O
X
UJ
o
Q
cr
LU
a
UJ
CO
(0
c
o
13
01
(C
O
o
OJ
E
•I—
Q.
I
.c -
D)
QJ
i-
O
o
>-
1—
1 — 1
1
__!
I — 1
CO
CQ
0
Qi
Q.
t
(f>
C.
0
'£ >>
3T3
•r- -P
S- V)
•P
W C
•i- O
T5 -r-
-P
U C
C O
OJ -Q
3 $-
O" fO
0) U
i- CD
Uu S-
LO
OJ
S-
D
D)
•r~
U_
161
-------
As shown in Figure 60, selenium was removed in the biological,
chemical, and physical (carbon) treatment processes. Mean and median
removals were 69 percent and 42 percent by activated sludge, 40 percent
and 57 percent by high-pH lime clarification, and 58 percent and 100
percent by carbon adsorption. Reductions in the Densator and carbon
column exhibited concentration correlations (r=0.45 and 0.98, respectively)
but no other correlations. There was no removal of selenium in the
multimedia filter. The EPA drinking water standard of 0.01 mg/1 was
exceeded in 35 percent of the train influent samples, while all train
effluent samples were well below the limit.
Silicon
Silicon removal varied between 4 percent and 6 percent through each
unit process in the treatment train. The Densator and recarbonation
basin together removed 13 percent (mean) or 16 percent (median), the
effluent concentration was proportional to TSS (r=0.41) and total P
(r=0.55). Also, the reductions in concentration were directly proportional
to the methyl orange and phenolphthalein alkalinities (r=0.33 and 0.41,
respectively). Multimedia filtration reduced the mean concentration by
only 6 percent, and the filter effluent silicon concentration was propor-
tional to effluent TSS (r=0.32) and turbidity (r=0.40). Little or no
removal occurred through carbon adsorption.
Strontium
Mean and median removals of strontium were 9 percent and 6 percent
in the activated sludge unit, 11 percent and 20 percent in the Densator/
recarbonation system. The reductions through chemical treatment exhibited
a concentration effect (r=0.42), and the recarbonation basin effluent
strontium concentration was proportional to TSS (r=0.38). There was
little, if any, significant change in concentration following multimedia
and carbon filtration.
The treatment sequence influent and effluent frequency distributions
shown in Figure 61 indicate almost no removal, indicating that strontium
is a quite refractory material.
Vanadium
Vanadium was removed only by biological treatment and carbon filtra-
tion, with mean and median removals of 15 percent and 8 percent in the
former, and 43 percent and 37 percent in the latter. There was a mean
increase of 0.5 mg/1 in the Densator, perhaps originating from contamina-
tion in the ferric chloride slurry. The carbon column more than removed
this added amount, the reductions exhibiting a concentration effect
(r=0.54). Mean and median train removals of vanadium were 39 percent
and 28 percent, respectively.
Zinc was efficiently removed by the activated sludge unit and the
; 162
-------
100.0
10.0
0)
3.
o
I-H
<
I-
2:
UJ
o
o
1.0
0.1
2 • 5
10 15 20 '30 40 50 60 70 80, 85 90 95
PROBABILITY OF BEING EQUALLED OR EXCEEDED
Figure 60.Frequency distributions for selenium, high-pH lime
coagulation and recarbonation study.
163
-------
q_
C
•P
U
o
o
s-
(0
o
7
Q
UJ
Q
UJ
Ld
O
X
LU
C£
O
Q
UJ
cr
UJ
CJ3 q_.
LU
03
U.
o
CQ
O
0)
e
Q.
I
JC
cn
3 •
•r~ >,
•P -0
C 3
O -P
S- CO
•P
VI C
o
S--r-
O -P
(0
c
10 O
C J2
O S-
•r- fO
•P O
3 (!)
J2 5-
t..-O
•P C
CO fO
CQ TJ
O -P
C (0
-------
Densator. Mean and median removals averaged 34 percent to 50 percent,
and 36 percent to 56 percent, respectively. However, the galvanized
finish on the recarbonation basin was vigorously attacked, resulting in a
large increase in zinc during recarbonation. The average mean and median
increases came to 0.57 mg/1 (675 percent) and 0.39 mg/1 (975 percent),
respectively, which more than negated the prior removals. Subsequently,
there were only minor removals through filtration and carbon adsorption,
such that the mean concentration of zinc increased by .0.41 mg/1 (205
percent) through the treatment train. Fortunately, not a single sample
violated the Secondary Regulations recommendation of 5.0 mg/1, in spite
of zinc's erratic behavior in the treatment train. It should be noted
that soon after termination of the grant, the amount of zinc emanating
from the recarbonation basin started to decline, finally reaching the
point where no additions were occurring. At this point all of the gal-
vanizing had been removed by corrosion and the bare steel tank rusted
quickly.
165
-------
SECTION 9
RESPONSES OF INDIVIDUAL METALS
BIOLOGICAL TREATMENT
The period of investigation for metals ran from June 1, 1972 through
January 31, 1974, a total of 610 days. The treatment processes of most
concern were the Densator, filter, and carbon column, operated in series.
Each of three basic types of chemical treatment were studied: Lime, only,
alum, and lime with recarbonation. It was determined by the pilot plant
research staff that the only requirement as to the type of wastewater to be
treated was that the feed to the Densator be a well-nitrified wastewater.
The requirement for nitrification resulted from the planning of the
virus removal studies (discussed in a subsequent section of this report), in
which low ammonia nitrogen concentrations were considered essential to,
proper disinfection with chlorine.
Since adequate nitrification did not occur in the White Rock Plant, it
became necessary to operate the Demonstration Plant's No. 1 activated sludge
unit in a nitryfying mode as pre-treatment to the physical/chemical process-
ing. Although the activated sludge unit is considered a basic part of the
treatment train, its function with respect to grant requirements was
strictly pre-treatment.
Due to the limited size and capacity of the aeration equipment it was
not possible to maintain nitrification for protracted periods on primary
effluent feed. The wastewater sources at the White Rock Plant which were
utilized during the investigation included primary effluent, stage 1
trickling filter effluent, stage II trickling filter effluent, and final
effluent. Table 50 identifies the periods of operation on each source (s).
From August 2-8, 1972, the No. 1 aeration basin was drained for modifica-
tions to the aeration equipment and installation of an additional air
compressor, during this time the No. 2 aeration basin was substituted for
the No. 1 unit. (Unless specifically referred to as No. 1 or No. 2, the
activated sludge data in this report include this week-long period in
August.)
Two other periods of particular interest are August 16-17 and August
30 through November 26, 1972, in which the activated sludge influent was a
combination of primary effluent and unsettled stage II effluent. Although
the exact volumetric ratio of each will never be known, estimates at the
time indicated about a 50/50 mixture. In the light of -the grant require-
ments, and since the activated sludge influent stream was always sampled,
166
-------
TABLE 50. ACTIVATED SLUDGE INFLUENT, JUNE 1972 THROUGH JANUARY 1974
DATES
FROM THRU
6-1-72
8-2-72
6-22-73
,6-23-73
9-1-73
10-11-73
10-21-73
11-20-73
'8-1-72
8-8-72
8-9-72 8-15-72
8-16-72 8-17-72
8-18-72 8-29-72
8-30-72 11-26-72
11-27-72 6-8-73
6-9-73 6-20-73
6-21-73 -
8-31-73
10-10-73
10-20-73
11-19-73
1-31-74
SOURCE
Unsettled stage II trickling filter effluent.
Unsettled stage II trickling filter effluent,
No. 2 aeration basin in service in place of
No. 1 aeration basin.
Unsettled stage II trickling filter effluent,
No. 1 aeration basin back in service.
Combination of unsettled stage II and primary
effluents.
Primary effluent.
Combination of unsettled stage II and primary
effluents.
Unsettled stage II trickling filter effluent.
White Rock final effluent.
Unsettled stage II trickling filter effluent.
White Rock final effluent.
Unsettled stage II trickling filter effluent.
White Rock final effluent.
Unsettled stage II trickling filter effluent.
Unsettled stage I trickling filter effluent.
White Rock final effluent.
167
-------
the exact proportion of primary and stage II effluent in the influent is
not germane to the project. One may correctly assume that the metals con-
tent of primary effluent is generally higher than final effluent; and in
this regard, every effort was made to operate on the strongest wastewater
source and yet provide nitrification.
Table 51 presents the average operating and process control parameters
for the No. 1 activated sludge system during the entire 610-day period of
investigation. The associated wastewater'characteristics follow them on
Table 52. There was a considerable variation in the operation of the system,
depending upon the time of year, organic,loading, etc. The activated
sludge effluent quality was relatively constant, because the process
control parameters were adjusted in order to maintain consistency in
quality. During most of the period, effluent NH3 -N remained less than 1.0
mg/1. Occasional temporary loss of complete nitrification, lasting per-
haps a week to ten days at a time, caused the average concentration to
slightly exceed 2 mg/1. Control of effluent BOD5 and TSS was not easy
because of uncontrolled denitrification that occurred in the clarifier.
However, filtration of the activated sludge effluent indicated that about
80 percent of the BOD5, and roughly half of the COD were associated with
the particulate fraction. All the parameters shown on Table 52 were
analyzed on a daily basis except for TDS and BOD5, which were generally
analyzed every fourth day.
Summaries of the metals analyses on activated sludge influent and
effluent covering the entire period of investigation can be found on
Tables 53 and 54, respectively. The tables list the median, mean, and
standard deviations, maximum and minimum concentrations, and number of
samples. Infrequent extreme concentrations on some of the metals^tended to
weight the mean high. Hence, the median values have particular significance,
since in many instances they are more truly representative of ambient_
conditions. In most cases, the mean values do in fact exceed the medians,
the relative difference representing the influence of extreme values.
Because the activated sludge process operated independently of the
downstream processes, the activated sludge metals data covering the entire
period of investigation are presented separately in this subsection. In a
practical sense, the following comments will be applicable to the perform-
ance of the biological process within each of the three treatment
sequences studied.
Silver
11.1 . ?
It became evident early into the investigation that very little silver
was present in Dallas' raw wastewater, in spite of some photographic
industries connected to the sanitary sewer. Measureable amounts were
found in only 28 percent of the activated sludge influent samples and 21
percent of the effluent samples. The maximum concentration ever observed
was five times lower than the current drinking water MCL of 0.05 mg/1.
Although a slight removal is indicated from the mean concentrations, silver
was not present at high enough levels to develop any definitive information
168
-------
TABLE 51 HYDRAULIC AND PROCESS CONTROL FOR THE ACTIVATED
SLUDGE PROCESS
HYDRAULIC OPERATION
Q (influent)
Q (return)
Q (waste)
Aeration T
Clarifier overflow rate(Q^)
Weir loading
Clarifier T (Q^Q^
PROCESS CONTROLS
MLSS
MLVSS
RAS . •
SVI
Air supplied
D. 0.
D. 0. Uptake rate
F/M (COD)
F/M (SOC)
F/M (BOD)
Sludge Age
Temperature
10,79 I/sec
(171 gpm)
10.3 -I/sec
("163 gpm) ,
6964 I/day
(1840 gpd)
1.95 hours
. 14.2 m3/day-m2
(348 gal/ft2-day)
33.6 m3/day-m
2706 gal/ft/day
2.76 hours
3608 mg/1
2619 mg/1
7241 mg/1
186 mg/1
19.1 I/sec.
(405 cfm)
2.8 mg/1
36.7 mg/l-hr.
0.366 day"1
0.039 day"1
0.106 day"1
7.9 days
24°C
(76°F) •
169
-------
TABLE 52. PERFORMANCE SUMMARY, NO. 1 ACTIVATED SLUDGE SYSTEM
PARAMETER
COD
TOC, soluble
BOD
TSS
TDS
SC, ^mho/cm
NH3-N
Org. N
N02+N03-N
N02-N,
Total P
pH, units
T. Alk. as CaCOq
0
P. Alk. as CaC03
Total Count, per ml
Total MPN, per 100 ml
Fecal MPN, per 100 ml
A.S. INFLUENT
(mg/l)
241
26
70
137
522
788
14.0
9.8
1.1
0.12
10.1
7.3
208
0
2.1 x 106
2.0 x 107
2.5 x 10®
A.S. EFFLUENT
(mg/1)
60
11
30
29
511
717
2.3
4.0
8.9
0.2
8.0
7.1
116
0
4.9 x 104
5.5 x 105
3.8 x 104
REMOVAL .
percent
75.1
57.7
57.1
78.8
2.1
9.0
83.6
59.2
--
—
20.8
--
44.2
--
97.7
97.3
98.5
170
-------
TABLE 53. ACTIVATED SLUDGE INFLUENT METALS SUMMARY JUNE 1972 -
JANUARY 1974
MEDIAN
MEAN
Cf
MAX.
MIN.
Ag*
Al
As*
B
Ba
Be*
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se
Si
Sr
V*
Zn
0.0
0.63
9.5
0.36
0.120
0.010
52.0
11.0
0.035
0.180
0.140
0.99
0.34
14.6
5.18
0.075
5.0
102.0
0.099
•0.100
4.2
9.4
0.88
3.6
0.270
0.78
0.66 --
14.8
0.37
0.129
0.014
54.6
14.0
0.039
0.205
0.224
1.07
0.43
14.5
5.18
0.076
15.2
99.5
0.109
0.108 '
7.2
9.8
0.79
3.9
0.366 ,
1.73
0.30
16.6
0.082
0.068
0.013
16.8
12.5
0.020
0.127
0.225
0.63
0.47
1.2
0.86
0.018
29.5
19.8
0.062
0.052
8.0
2.4
0.32
1.4
0.424
10.0
1.80
119.0
,- 0.65
0.48
0.04
144.0
119.0
0.12
0.75
1.04
7.80
3.2
16.8
8.12
0.13
170.0
148.0
0.38
0.45
40.0
17.0
1.25
8.5
4.10
0.0
.0.18
0.0
0.16
0.01
0.0
31.0
0.0
0.0
0.015
0.0
0.28
0.0
11.8
1.62
0.04
0.0
33.0
0.01
0.0
0.0
3.9
0.23
. 0.5
0.05
58
91
153
149
168
23
181
222
183
222
181
181
m
58
129
182
38
145
159
218
115
53
41
49
182
Concentration in mg/1 (*ug/l)
171
-------
TABLE 54. ACTIVATED SLUDGE EFFLUENT METALS SUMMARY JUNE 1972 -
JANUARY 1974
MEDIAN
MEAN
MAX.
MIN.
Ag*
Al
As*
B
Ba
Be*
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
Si
Sr
V*
Zn
0.0
0.29
9.0
0.36
0.060
0.0
47.4
6.0
0.030
0.060
0.051
0.35
0.15
14.1
5.01
0.052
2.4
96.5
0.075
0.040
1.0
9.4
0.83
3.0
0.110
0.38
0.44
13.4
0.37 .
0.070
0.005
51.9
7.1
0.036
0.071
0.072
0.61
0.26
14.1
4.96
0.056
3.6
97.2
0.082
0.050
1.7
9.4
0.72
3.3
0.139
0.85
0.53
12.9
0.075
0.047 ,
0.008
17.1
4.6
0.023
0.053
0.076
1.41
0.33
1.2
0.64
0.028
3.9
19.8
0.040
0.036
2.1
2.2
0.30
1.3
0.115
3.0
3.50
85.0
0.60
0.36
0.03
118.0
30.0
0.13
0.56
0.69
14.90
2.2
16.5
6.60
0.25
15.4
150.0
0.28
0.30
9.0
13.5
1.19
8.0
1.06
0.0
0.02
0.0
0.17
0.0
0.0
26.5
0.0
0.0
0.002
0.01
0.07
0.0
11.2
2.43
0.003
0.0
45.0
0.0
. 0.0
0.0
4.6
0.16
1.3
0.04
58
92
156
151
169
23
182
223
184
223
182
182
111
58
130
183
39
146
160
220
119
54
42
52
183
Concentration in mg/1 (*yg/l)
172
-------
concerning its removal.
Aluminum
Aluminum removals by activated sludge averaged 33 percent (mean) or
55 percent (median). However, the effluent concentrations were much more
variable than the influent, and exceeded the influent concentrations
approximately 10 percent of the time. The influent frequency distribu-
tion was log-normal, whereas the effluent distribution clearly followed
a bimodal pattern. Both the percentage of aluminum removed and the amount
removed exhibited a pronounced concentration effect, i.e., directly
proportional to the influent concentration. Aluminum removal also
correlated somewhat with COD removal, the highest removals occurring at
COD removals of 70 percent or more. High effluent concentrations also
coincided with high effluent COD and TSS.
Arsenic
Arsenic was not significantly removed by biological treatment, al-
though a slight overall decrease of 9 percent was obtained. The activated
sludge unit did appear to smooth the fluctuations in influent concentra-
tions. The drinking water MCL of 0.05 mg arsenic per liter was exceeded
only 3 percent of the time in both the influent and effluent. The
removals exhibited a weak concentration effect, but zero or negative
removals occurred on a 47-percent frequency.
Boron
Boron was refractory to biological treatment, with no significant
removals or increases through the activated sludge process. Based on
mean values, a 9 percent reduction was observed; however, there was no
removal on the basis of median values. The highest influent concentration
observed during the entire period of investigation was only 0.65 mg/1.
Barium
Removals of barium in the activated sludge unit average 46 percent
by means, 50 percent by medians. Reductions in concentration exhibited a
concentration effect at all influent concentrations (r=0.75), but a plot
of effluent versus influent concentrations was also linear(r=0.67). • COD
removal and effluent COD and TSS concentrations behaved in a manner
similar to effluent barium concentrations.
Beryllium
Analyses for beryllium were performed.on a total of 23 days scattered
over a period of 5 months. The concentrations observed in the activated
sludge influent were exceedingly low, the highest being only 0.04 mg/1.
An average reduction of about 64 percent occurred through biological
treatment, but most of the effluent concentrations (61 percent) were below
the detection limit. Beryllium reductions did exhibit a strong concentra-
tion effect at all influent concentrations (r=0.89), but the limited
173
-------
amount of data prohibited any conclusive observations.
Calcium
The activated sludge influent calcium concentration averaged 54.6
mg/1, and varied from 31 mg/1 to 114 mg/1. The overall removal was 5
percent (mean), or 9 percent (median). No significant correlations or
removal patterns were ever observed.
Cadmium
A significant removal of cadmium occurred during activated sludge
treatment, 49 percent by means, 45 percent by medians. A very pronounced
concentration correlation was observed for all influent concentrations
(r=0.94). The only occasions when zero or negative removals occurred was
when the influent concentration was less than about 10 mg/1. The relative
fluctuations in concentration were also lower on the effluent than on the
influent. In general, effluent concentrations increased with COD and TSS.
The highest concentrations of cadmium in the effluent happened to coincide
with low dissolved oxygen in the aeration basin (<1.0 mg/1), but cadmium
removals were not particularly influenced by the residual dissolved oxygen
concentration in the mixed liquor.
Cobalt
Some cobalt roughly 8 percent (by means) to 14 percent (by medians),
was removed through activated sludge. No removal patterns or correlations
were observed.
Chromium
Chromium was significantly removed by the activated sludge process.
The average removals were 66 percent ( mean) to 67 percent (median), and
the removals were very consistent. Chromium reductions exhibited a classi-
cal concentration correlation at all influent concentrations. A graph
showing the reductions versus influent concentrations appears on Figure
62. The linear correlation coefficient for the estimating equation is:
r=0.92 (N=222). The percentage of chromium removed also exhibited a
concentration correlation. In addition, COD removal was proportional to
percent chromium removal (r=0.62) and chromium reductions (r=0.45).
Copper
Copper removals averaged 68 percent by means or 64 percent by medians.
The distribution of concentrations was skewed toward high values in the
influent and low values in the effluent, atypical of the patterns observed
for the other metals. Like chromium, the reductions in concentration
through the activated sludge unit were strongly concentration-dependent
at all influent concentrations (r=0.94). The removals also correlated
directly with COD removal and inversely with TSS. Significant copper
removals were observed, mainly during nitrifying operations.
174
-------
en
E
UJ
_J
u.
ro
O
rd
CD
c
o •
-C P
00 •!-
C
" rs
to
C 0
O Oi
•r- -a
•P =3
(C r—
s- 10
•p
c -o
0)
U T-
p
P U
c ta
••
P
(/) O
C OJ
P CD
O
Zi C
•a o
a> T-
S- P
ta
=5 P
•r- C
E a)
o u
s- c:
JZ O
o u
CO
l£>
0)
^.
rs
O)
co
o
-------
Iron
As expected, iron was significantly removed by the activated sludge
process, 43 percent (mean), 64 percent (median). Probability distributions
reflect a log-normal pattern for the influent and a bimodal relationship
for the effluent, skewed toward the high side. Only one extreme value
was picked up on the influent, compared to at least eight on the effluent.
However, effluent iron concentrations exceeded influent iron only at
influent concentrations less than 1.0 rng/1. Iron reductions exhibited a
strong concentration effect (r=0.80), and the correlation between iron and
TSS was reasonably good (r=0.66). All effluent concentrations of iron
greater than 1.0 mg/1 accompanied TSS concentrations greater than 45 mg/1.
The EPA Secondary Regulations recommendation of 0.3 mg/1 was exceeded in
99 percent of the influent samples and in 59 percent of the effluent
samples.
Mercury
Mercury removals could be characterized as moderate, ranging from 22
percent (means) to 42 percent (medians). The activated sludge influent
concentrations exceeded the EPA maximum contaminant level for drinking
water (0.002 mg/1) on only 2 percent of the samples and the activated
sludge effluent exceeded this concentration in about 1 percent of the
samples.
No extreme values, or radical variations in the distribution, were
observed. Negative removals occurred on 30 out of 110 sample days, zero
removals occurred on 5 days. Reductions in mercury were strongly'
concentration-dependent at all influent concentrations, and inversely
related to effluent NHo-N. The largest reductions in mercury coincided
with low effluent NH -N.
«J
Potassium
Potassium was essentially unaffected by biological treatment, but a
net overall removal of 3 percent (mean and median) was obtained. -Sludge
age and MLSS were the only parameters which demonstrated any correlation
with potassium removal. In general, the removals increased with both
sludge age and MLSS. No concentration effects were observed.
Magnesium
Magnesium was poorly removed by the activated sludge process, 4
percent (mean) and 3 percent (median). A concentration effect was ' ,
evident at influent concentrations greater than about 5.0 mg/1 (r=0.67),
the reductions increasing at higher concentrations. The largest removal,
34 percent, occurred at the highest influent concentration, 8.1 mg/1.
Sludge age also appeared to exert a minor effect on magnesium removal.
Manganese
Manganese removals averaged 26 percent (mean) or 30 percent (median),
176
-------
varying from a minus 156 percent to a positive 96 percent. However,
there were no extreme values in either the influent or effluent samples.
Mn reductions exhibited a concentration correlation (r=0.42), but negative
removals occurred in 8 percent of the paired samples. Correlations between
effluent NH3-N, COD, and TSS concentrations, and manganese reductions
revealed inverse relationships in each case (r=0.46,-0.49, -0.45,
respectively). The EPA Secondary Regulations recommendation of 0.05 mg/1
was exceeded in 91 percent of the influent samples and 51 percent of the
effluent samples.
Molybdenum
Substantial removal of molybdenum occurred through biological treat-
ment, 77 percent .by means and 52 percent by medians. Complete (100 per-
cent) removal occurred on 7 out of 38 paired samples. The activated
sludge system provided considerable buffering for extreme influent values,
the influent and effluent probability distributions being quite dissimilar.
The influent probability distribution was skewed toward high values.
Molybdenum reductions also exhibited a pronounced concentration effect-at
all influent concentrations (r=0.99).
Sodium
Sodturn, like potassium, was poorly removed during biological treat-
ment. However, a mean removal of 2 percent and a median removal of 5
percent was obtained on the 145 paired samples. The greatest reductions
were obtained at low F/M ratios, although the relative effect was minor.
There were no extreme values of sodium, and the highest concentration was
150 mg/1, roughly 50 percent over the mean value.
Nickel
An average 25 percent removal of nickel occurred through biological
treatment, based on both means and medians. Reductions exhibited a
strong concentration effect at all influent concentrations greater than
about 0.06 mg/1 (r=0.78), with negative removals occurring frequently at
lesser values. Nickel reductions varied inversely with effluent NH^-N,
COD, and TSS concentrations. There were no extreme values of nickel in
either the influent; or the effluent samples.
Lead
Lead was significantly removed by the 'activated sludge system, 53
percent (mean) and.60 percent (median). The reductions exhibited a
strong concentration effect at all influent concentrations (r=0.78). A
reasonably good correlation was observed between lead removal and COD
removal (r=0.52). .Likewise, high effluent TSS and COD often accompanied
low lead removals. Out of 218 influent samples there were only two
extreme values (0.35 mg/1 and 0.45 mg/1), which were readily damped by
the unit. The drinking water MCL of 0.05 mg/1 was exceeded in 88 percent
of the influent samples and 31 percent of the effluent samples.
177
-------
Selenium
Selenium removal averaged 76 percent by both means and medians,
ranging from negative removal to 100 percent removal. Reductions in
selenium were strongly concentration-dependent over the entire range of
influent concentrations (r=0.97). The activated sludge system provided
consistent buffering against high concentrations in the effluent.
Removals were higher at sludge ages greater than about 15 days. Also,
correlations in selenium reduction versus effluent COD and NH3-N were
both inversely related. The drinking water limit of 0.01 mg/1 was exceeded
in 24 percent of the influent samples, but none of the effluent samples.
Silicon
Biological removal of soluble silicon averaged only 5 percent by
means, zero by medians. No similarities in removal patterns were observed
between silicon and other chemical species.
Strontium
j
Strontium removal averaged 9 percent (mean) and 6 percent (median),
and appeared to increase with increasing MLSS and decrease with increasing
F/M, COD, and TSS. The frequency distribution of both the influent and
effluent adhered to a bimodal pattern, almost devoid of values in the
0.30-0.65 mg/1 range. However, there were no extreme values in either
case.
Vanadium
Removal of vanadium averaged 14 percent (mean) and 17 percent (median)
varying considerably within the range from a negative 160 percent to a
positive 100 percent. A slight concentration effect appeared to bias the
reduction when the influent exceeded about 3.0 mg/1 (r=0.49). It appears
that higher removals accompanied higher MLSS concentrations in the
aeration basin, but the lack of sufficient data limited any further
analyses. No extreme values were found in either the influent or the
effluent samples.
An averaged removal of 62 percent (mean) or 59 percent (median) was
observed for zinc over the period of investigation. .Reductions through
the activated sludge process clearly exhibited a concentration effect
(r=0.96). There appeared to be a positive correlation between zinc
reduction and COD removal, and negative correlations between zinc reduc-
tion and effluent NH3-N, COD, and TSS. A few extreme values appeared on
the influent samples, but none in the effluent. At no time did the
concentration of zinc in any sample reach the EPA recommended drinking
water MCL of 5.0 mg/1.
178
-------
THE THREE COMBINATIONS OF TREATMENT SEQUENCES
During the 610-day period covered by this report the Demonstration
Plant of the Dallas Water Reclamation Research Center was configured in
the three separate advanced wastewater treatment (AWT) sequences that have
been described in detail. Tables 55 through 57 are the summaries of the
analytical data for each phase of the project, with the exception of the
metals data which are summarized in Tables 58 through 60 .
Although the study of disinfection per se was not a direct part of
this research effort (with the very obvious exception of the virus
studies reported in the following section), it is interesting to
evaluate the microbiological data presented in Tables 55, 56, and 57.
During the high-pH lime coagulation study (Phase I) the upflow clarifier
was very effective as a disinfection process, and since the effluent pH
was not neutralized, regrowth or recontamination was not observed.
The alum coagulation study constituted Phase II of the project, and
the reductions in coliform organisms resulting from the chemical treat-
ment process were only slightly greater than one log.
Phase III, high-pH lime coagulation and recarbonation, resulted in
significant reductions in bacterial populations; however, the neutralized
effluent was susceptible to recontamination and regrowth.
The influent and effluent mean metals concentrations and the
corresponding removals are shown for all three phases in Table 61,while
Table 62 presents similar information based on median concentrations.
Silver
Silver was not usually present in a concentration high enough to
permit evaluation of its response to the various unit processes.
Aluminum
Aluminum had the highest removal (78 percent based on medians) during
the alum coagulation phase; however, the removal decreased to five
percent when evaluated in terms of mean concentrations. The lime
coagulation and recarbonation phase yielded a consistent removal of 77
percent (mean) and 75 percent (median); furthermore, this was the only
instance when the AWT processes reduced the arithmetic mean aluminum
concentration in the effluent.
Arsenic
Arsenic was effectively removed by the AWT processes, and only
slightly affected by the activated sludge process, as shown in Figure
63. 'The lime and recarbonation study constituted the most successful
treatment sequence which resulted in removals of 83 percent (mean) and 79
percent (median).
179
-------
I-
o
O
u
_!
D_
SI
C
rO
rC
T3
C
O
O
0)
•P
C
01
rC
CU
S-
•P
s-
cu
•p
q-
0
•P
rO
0
•P
(/>
re
5
(/)
0
(/) 0
r- O
0 O
> S-
0 Q.
U S-
•i- re
E -i-
0 -P
i/i i-
i- 0
< -P
ro
0
S-
u.
LO
CM
O
CM
IT)
O
LO
180
-------
z:
o
i— t
%
—yr
*i_.
O
CQ
Qi
O
LU
o:
r-
3
O
I
i— i
200
h- CD
Z! P—
LU
h- LU
< CQ
Di LU
HCJ
LU
LU Q
*r i
_J LU
CQ
Q.UJ
1 >
O
in z;
CD
i — i •»
X CM
>- cr>
cc i—
s: LU
oo o
^ o
rf ?
01-
00
< CD
~~ 3
t—
_j LU
< z:
z: r>
< -D
•
LO
LO
LU
CQ
I-
H
z:
z: z: uu
CQ § 3
Qi _J LU
< O Lu
O O LU
e£
i— <
ao; h-
LU LU z:
SI-LU
r-l _J ID
_! LU Lu
r3 u.
2! LU
Oip-
o z:
|_LU
< ID
00 Zl
z: u_
LU LU
Q LU
n
LU r-
I— LU ZT
< CD LU
^> ^^ *"*5
h- —1 LU
O OO LU
< LU
Q
LU h-
h- LU z:
< CD LU
1-1 Z3 —I
h- -1 LU
o oo z:
a:
LU
LU
Qi
CL
CT> 00 VD LO
CM «tf f-s CO <* CM P- •
p-yDOOr>.VDr-OOp-CT> P- OP-
>^- en p- LO
CM CO CM CM 1^ O •* •
CM i st"! — ^DP-CMCMCTI.P- Oi —
LO LO CT) LO
O LO tD 00 P"s CTi ,CT) •
CMCTl<*l=J-tX)r-CMCMCTl O OP-
1
t— CTi "^ CTi CM LO LOr—
aip-CMCMLOhs'Nt-'d-co o cnrv
p~ r*^ ^o o
00 'LO|^O '00
r^CMCOCMOOr-VO'- • P- •
CM^OOp-LOCOp-CT>O O p-fv
CU
p- E
n u
3 v.. z in
r- 0 1 •*->
o ^: oo CL -r-
10 E z: o c
«tn i • + i «8
QOQOOOO «OOD)CM CM-)-1 '
OOOOOQOX i-o o on:
oi— CQI— i— ooz:oz: z: I— D
•
CO
i rt
u t
CM
!
Js*^
1^^
CM
LO
10
Vi**
CM
s*-
p—
O
CM
00
O
O
O
!
\
•
.(-
00
(V|
^ VI
CM
CO
CM
00
CM
CM
0
O
oo
o
o
(0
o
^0
(y
^f
CL
ID
CO
00
[/>
o
!»-*
X
o
•CM
to
o
X
CM
LO
1
&-
CU
ex
ft
•p
c
a
o
o
(0
•p
0
p-
CVl
VNJ
CM
V
00
LO
o"
p™ *
X
r-
O
P—
X
CM
^
1
O
o
p—
!~
0)
Q.
D-
2!
P™
rO
0
H
CM
V
CM
V
CM
t-
^T™
O
CO
D
o
P—
X
oo
vo
1
o
o
p~"
0
Q.
a.
2:
I1 " '
to
u
CU
Lu
-.
&
(0
•p-
(0
"^
^-
V
1
.
•a
CU
+->
o
l/l
(0
•p
CU
u
X
CU
p_
v^
01
c
c
o
•p-
(0
J.
4->
c
CU
u
c
o
181
-------
h-
t
r-
2:
UJ
g
LU
o:
SCO
ID tv.
_j cn
*^ f""
QC LU
- iE.
—I UJ
^ ^*
"Z. O
^ 2:
•
LO
UJ
CO
H
tZ.
2: 2: uj
O S ID
DQ Z5 — 1
rv in
«C O U_
O O UJ
l-iQi
O LU h_
LUH 2:
S — 1 LU
l— t t— ) 13
h-Lu-J
—1 U-
ZD 1 1
S UJ
Q;H
0 Z
h- LU
^C ~"^
CO _J
2: Lu
1 1 1 1 1
Q LU
O
LU h-
HLU2:
erT en 111
> Q ID
I— — 1 Lu
O CO Lu
< LU
Q
LU h-
H-LU2:
< CD LU
I— < 3 — 1
1— — 1 Lu
O CO 2:
cC i— i
C£
LU
LU
S
o;
D.
(V.
r- i— rv «3- LO CM OlLO O
CMCOOCOtO P- Ol£)O
CMCOUDr-LOrvCMCMCO O COrvp-
ID LO tv
00 i— 00 00 "st" O COCMVD
LOr-CMCM^lOCMCOCO O tv.tv.p-
CM {Q tO
CO O CM
rv cMrvo- -cMr- cocoto
COOCM'd-OLOOOO • • • -O
CMCMtO p-LOrvp-p-rp— O CTirvCM
co
o
o
(0
0) o
J— •
-Q E co
3 Q. Q.
f^ « *l
3 2: 2:
O Q- Q.
p— p— p—
rd rd rd
•P -P U
O O QJ
h- 1- Lu
^
TJ
0)
•P
0
CO
rd
•P
O.
0)
o
X
en
c
•p-
co
c
o
'.p
rd
S-
•p
c
U
c
0
o
182
-------
n:
H
i— i
3:
1-
LU
ST
^1.
LU rx
Di CD
Hr-
LU >-
jr iv
_i r>
:n et-
(-)
•
_* i
CD rv"i
h~l rx
^ CD
I*-""
o; oi
< LU
s: CQ
r> LU
co>
o
^«* ^^
h~
O.O
1— 1
o 2:
HH C 13
1 n f^
1 — — * -
~J ^,
<0
2: LU
^ CL
1^*
LO
LU
_J
CQ
H
~y
O
CQ
o:
<
o
<
f—H
o
LU
21
i— i
h-
_J
s:
o
LU
[—
<
>
i— i
<
Q
LU
<
1 — i
H-
O
<
1-
2Z
2: LU
SZD
Z) _J
_J LU
OLu
O LU
H
^^
o; LU
LU ID
(— . i
-I Lu
H-l LU
LU LU
tl-
CQ LU
C£ ID
^ — J
OLU
OL LU
Qil-
(— LU
CO -1
Z Lu
LU LU
Q LU
LUp-
C5 *^.
Q LU
ID O
_1 _l
CO LU
I t
uu
LU
1—
LU -Z.
CJ3 LU
•=> —1
-J Lu
co -z.
1—1
OL
LU
1-
LU
*>T
^f
cf
Q.
t
00
rx r— tx
00 CO •
CO 00 CM ^" ^O CD CM
CO
LO •?}• rx
LO ^j" rx •
CM fx «vt- LO IO CD CM
^o
00 CD 00
1^^ ^O O^ i™" *
CM CD LO 00 LO CD CM
^h O
oo co rx
CO LO 00 O •
CM CD •* [X UD i— CM
o
tQ CM CD
1 n |w« f*f} ^4" FY^ IO .
l-l i "^ WJ VJ \* J VU
to i — co "^ LO rx i —
<3-
rx oo rx r— •
i — IX, CD 00 LO Ln •si'
CM CM l£> i— LO CO i—
CU
•- E
o o
^ x^
r- O
0 -C
00 £ •'
" LO 1
Q O O CO CO « 00
o o o oo o o :r
O (— CQ 1— 1— CO 21
1—
1—
.
r—
LO
r—
.
CM
p—
CM
CM
CM
CD
p—
00
^t
5^-
cr.
o
•
CD
~y
.
D)
i.
0
CO
.
O
r—
^J-
.
O
LO
CD
•
O
r~"
CD
CD
CD
O
2:
1
OO
0
2:
+
0^
2:
f^
r—
r—
.
O
p_
oo
i—
.
0
00
p—
1—
0
CD
O
O
00
LO
0
0
I—-
IK
.
o
2:
1
0^
CD CO ID
• • 00
O «3 CM
O (£> O
. . (T)
r— 103 CM
O Or-
r- IX, CM
LO
00 • *fr
• O O
r- r- CM
•vt" i — CD
rx rx p—
CM
• CM LO
O • i—
p- rx CM
o
o
o
fO
o
to
to fO
4->
CL T- •
C -^
p- 3 p-
(0 <
I-P- -
01 •
1- D-r-
co
O
p—
X
00
o oo
J-
o
X
.
0 p-
"m
O
p—
X
00
f\ I •
CM •
P- CM
O
00 CO
CO
O
i»
X
00
O CD
ID
O
r~™
X
CD
O r-
I™"*
E
0 00 S-
O CU
O Q.
rd
o •>
-p
W C
(0 3
o
• o
_^
p— p—
< (0
•p
0
Q- h-
0-J
o
r—
X
00
« CD
i— LO
in 4-
0 0
'x'x
CM LO
00 CM
[**« (.0
o o
X X
"vl- OJ
r- «*
p— p— •
E E
0 0
oo
r— i —
t- $-
CU 0)
Q. Q-
f\ A
21 "^- •
D- D_
3T ^r
p- P—
ns re
•p o
0 0
I-LU
.
-o
-p
o
E
to
fO
•P
o
1 '
CD
O
CU
P—
CJ)
c'
•I—
w
c
0
-p
fO
S-
•p
c
CU
o
E
O
o
183
-------
TABLE 58. SUMMARY OF MEAN CONCENTRATIONS,
HIGH-pH LIME TREATMENT WITHOUT RECARBONATION
JUNE, AUGUST-OCTOBER 1972, NOVEMBER-DECEMBER 1973
METAL
A.S.
INFLUENT
A.S.
EFFLUENT
DENSATOR
EFFLUENT
FILTER
EFFLUENT
CARBON
EFFLUENT
Ag*
Al
As*
B
Ba
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
Si
Sr
v*
Zn
0.62
0.92
18.7
0.39
0.149
45.0
13.3
0.045
0.209
0.138
1.05
0.53
14.7
5.27
0.071
2.5
108.8
0.104
0.100
14.2
10.0
0.27
4.6
0.323
0.26
0.38
19.5
0.42
0.079
42.5
7.6
0.037
0.085 '
0.080
0.34
0.29
14.3
4.97
0.051
2.1
108.1
0.082
0.044 '
2.4
9.6
0.24
4.2
0.123
0.94
0.20
10.8
0.41
0.083 •'
156.1
7.6
0.058
0.012
0.056
0.30
0.14
14.5
0.99
0.010
4.2
117.3
0.052
0.039
1.9
13.4
0.23
3.6
0.063
0.94
0.29
7.0
0.41
0.092
156.1
7.4
0.057
0.009
0.066
0.12
0.094
14.7 '.
0.66
0.0061
1.7
117.1
, 0.042
0.040
2.04
6.7
0.23
2.4
0.048
0.62.
0.36
6.4
0.39
0.091
144.9
7.2
0.054
0.0095
0.071
0.064
' 0.11
14.5
0.58
0.0067
1.9
117.5
0.029
0.043
1.00
8.8
0.23
2.9
0.041
Concentrations in mg/1 (*yg/l).
184
-------
TABLE 59. SUMMARY OF MEAN CONCENTRATIONS,
ALUM TREATMENT '
NOVEMBER 1972 - OCTOBER 1973
METAL
Ag*
Al
As*
B
Ba
Be*
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
S.i
v*
Zn
A.S.
INFLUENT
0,0
0,63
11.4
0.36
0.132
0.014
59.3
12.2
0.029
0.236
0.33
1.28
. 0.52
13.3
4.80 .
0.083.
20.9
92.0
0.114
0.121
4.6
10.3
4.8
0.520
A.S.
EFFLUENT
0.0
0.42
10.1
0.35
0.065
0.005
55.8
5.6
0.026
0.066
0.054
0.33
0.26
12.7
4.67
0.055
4.2
89.2
0.1079
0.054
1.1
9.7
4.2
0.156
DENSATOR
EFFLUENT
0.0
2.13
6.8
0.37
0.057
0.0005 .
69.4
4.1
0.027
0.023
0.034
0.16
0.32
12.5
4.58
0.037
3.1
89.3
0.068
0.042
1.0
10.4
3.9
0.109
Concentration in mg/l(
FILTER
EFFLUENT
0.0 .
0.78
. 6.1
0.38
0.056
0.012
68.6
4.0
0.027
0.019
0.056
0.13
0.25 '
12.3
4.65
0.030
3.8
88.9
0.070
0.040
1.2
9.7 •
4.0
0.106
*yg/D
CARBON
EFFLUENT
0.0
0.60
5.8
0.35
0.056
0.010
66.2
3.8
0.025
0.016
0.039
0.09
0.22
11.6
4.55
0.023
3.2
86.4
0.058
0.040
0.6 .
.9,9
• 4.3
0.067
185
-------
TABLE 60. SUMMARY OF MEAN CONCENTRATIONS,
HIGH-pH LIME TREATMENT WITH RECARBONATION
NOVEMBER 1973 - JANUARY 1974
A.S. A.S. DENSATOR RECARB. FILTER CARBON
METAL INFLUENT EFFLUENT EFFLUENT EFFLUENT EFFLUENT EFFLUENT
Al
As*
B
Ba
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
Mg
Mn
Na
Ni
Pb
Se*
Si
Sr
v*
Zn
0.69
21.2
0.36
0.115
66.2
13.7
0.052
0.114
0.182
0.85
0.28
5.14 •
0.079
93.8
0.136
0.089
7.8
9.7
0.79
3.4
0.203
0.47
20.6
0.36
0.075
64.9
7.3
0.048
0.052
0.077
1.45
0.26
4.96
0.066
90.3
0.106
0.050
2.4
9.3
0.72
2.9
0.133
0.18
5.6
0.38
. 0.086
105.4
3.6
0.067
0.014
0.050
0.97
0.11
2.58
0.019
90.5
0.087
0.029
1.4
8.7
0.66
3.4
0.084
0.1.5
3.9
0.34
0.084
106.1
3.4
0.067
0.011
0.073
0.82
0.17
2.40
0.016
90.0
0.083
0.027
1.5
8.1
0.64
3.2
0.654
0.21
3.5
0.36
0.094
96.5
3.3
0.062
0.008
0.082
0.24
0.12
2.61,
0.017
82.5
0.072
0.026
1.7
7.6
0.66
3.6
0.532
0.16
3.5
0.32
0.086
101.2
3.0
0.063
0.006
0.054
0.20
0.08
2.46
0.17
81.7
0.078
0.027
0.7
7.3 '
0.64
2.0
0.617
Concentrations in mg/1 (*yg/l)
186
-------
oo
o
UJ •
1— 1
1-
z:
UJ
s:
u.
o
OL
S
UJ
_J
CQ
1-
ID
CQ
O
UJ
UJ
•z.
o
UJ
2:
^ o LO °^ co °0 O^CM cD'st-cocooooocoLnCMLO^CDi^r^rf- ior^
S3 ^J- in CM,—. ID r- CD CO CD LO r— rvCO ^VDCO 1 r— 00
1
(DO LO VO CD CO COO 1^
O LO LO i— CM CMi— OOCDCMIDLOCM "3- LO "N}- (O
U. O <£> 00 CO O O • 00 O O O O CM • LO O CM • O O «D CD COO
UJ OOLOOOOVOCOOOOOOi— •xtOCOCOOOOCD l-xTO
CM1^ CDtD CO "sti— O
(O CD t— ,-j.
11 i
VD
VD CMOO • (DOLOOcO *Di— I^[^.CM >s)-i—
U_ r— LO 00 O i— O O O O CM O "*O 'OOr^OOVOOUD
UJ IOOOOO li— COOOOOO ICMO ICOOOOI^OCMO
LO CM •vl' CM CD tD CD CO
CDCMUDr— CMI^LOi— COLnCO ^J-I^N COCOCO CD O
U_ ID'OOi— ••Or- r— CO CM r-O •r-O.';^CM
I-H IOCMOO 1 O O O CO CM CD O
LU OO^DOO Ir-I^OOOOOi— OOr-i— OOi-OOOCMO
CD LO CD CO r— CO •?!• O CO
CMCMI'^CD'sJ' OOO^OCOLOOOP^r^P^ "OOCMOI^v CM
LL. VD CD • CO i— • • O CM r- O LO • CM O LO CO r- i— • • CM J-r-OOOr—Oi— LOOCMr- OO'— r— O^O
_1
1— **** * * *
UJ O)r— (/) (0
-------
oo
_i
o
LU
•z.
h-
0
LU
SE
Lu
O
on
oo
CM
t.O
LU
CO
^
s:
CO
OL
LU
CC
LU
H- 1
)
O
LU
s:
»— i
_j
Q; CO r- CO CO O •* CO CT) *t O CO CF> ID CM LO i— CM O COO O 1 CO CO
fc3 i IN. LO C7l CF> OOCOCD ILOO
&5 IIN.IN.CMCM 1 "shrNi— OltDhsCO ILOCO lr-«i-lDOr— CMCMLO
ii ' r~ '7
O O1LOO 1-O i— CO O CO 1 CO 1 O
&s i IIN.I— coi<^-cocMcr>iDcr>cy><"cr)CT> rNtocrii ICTI
CM
1 1 1
OOOOtDO O O O O LO
COLOCMCO OO'vt'CntDOOCIi— IN, 'CTlOO ID O
LU O IN. "sl-i— • • O r— O O LO • CM O O O O i— • O CM ID CO
>-" oor-oo i <*i— oooi— Oi—LoocMr-ooi— cnocoo
_J
LxJ O)r™~ C/) fO QJ fO "O O S- 13 QJ O) O} C O fO *r" UQ Qj »r~ ^< *K d
ST "^C ^^ ^C CQ PO pQ t_) (^} CO CJ CO 1 1 ^T", ^*^ *>* ^^ gg^ ^^ ^P Q^ (f) (f) (f) ^^, [xj
0
r—
fO
•1—
to
P
o
1
1
r—
•X
D)
C
•p-
Concentratic
188
-------
Boron
Boron was very refractory to all treatment processes, and no signifi-
cant statement can be made except that it was not removed.
Barium
The removal of barium was quite variable ranging from 25 to 58
percent for means and 27 to 68 percent for medians. As indicated in
Figure 64 the activated sludge process was responsible for most of the
barium reduction, with the AWT processes being rather ineffective.
Beryllium
Beryllium was present at concentrations too low to permit evaluation
during this project.
Calcium
Calcium was added in the form of calcium hydroxide during all three
phases of the project; therefore, the calcium concentration was increased
in all three phases as a result of the lime feed.
Cadmium
Cadmium removal was generally good, with both the activated sludge
and AWT processes contributing to its removal as shown in Figure 65.
Cadmium was removed most effectively by lime and recarbonatibn sequence;
78 percent by medians. These removals were the result of both the
activated sludge and AWT processes.
Cobalt
Cobalt was removed only slightly by the activated sludge process, and
the data indicate slight removal during alum treatment; however, the
change in mean concentration was only 0.004 mg/1. .Cobalt was added by
the treatment chemicals during both sequences involving lime as the
primary coagulant.
Chromium
Chromium removals were excellent as Figure 66 indicates. The lime
coagulation used in Phase I resulted in the greatest removal (95 percent
by medians), but lime and recarbonation resulted in the lowest median
concentration of 0.005 mg/1. It should be noted that most of the removal
occurred in the biological process.
Copper
The data in Figure 67 clearly indicate that the activated sludge
process was effective in removing copper and that the AWT processes
investigated were not effective.
189
-------
fO
i.
H-
i.
(t5
U
fC
S-
h-
C£ ro
O
•P 0
3 o;
>- o c
-Ll .G T- ^
-P n3 -P
•r- J_ -r-
2 h- 3
0 E 0
E 3 E
o
o
o
F—I
uj sg;
Q O
O
a.
s:
<
c/)
LU
00
•P
c
0
P
(0
0)
S-
' P
s-
0)
P
<4-
fO
J.
OJ •
P co
1C 0
5 oo
CD 00
4-> 0
00 U
(0 O
S S-
Q.
C
c to
o -P
•r- 'r—
-P i-
(d 0 -
S- P
P
C "O
0 C
O (0
c
o >,
O 5-
fC
E-O
3 C
•r- O
J- O
-------
•P (0 -P
>- T-S-T-
UJ 3 I— 3
o
o
UJ
o_
00
4-1
E
cu
(0
0
i.
•p
s-
cu
•p
s.
OJ
•p
fO
CU
•P
W
fC
E W
•r- (D
(/)
cn to
c cu
•r- O
E O
•r- S-
fl3 D..
I ^
rd
tO •!-
E +->
O S-
•r- 0)
-P -P
fO
S- TJ
•P E
E (d
0)
O (0
O TD
E
E O
D O
•r- 01
E
T3
(0 >j
o .a
un
ID
s_
3
cn
•r—
U_
O
IT)
191
-------
u
O
UJ
_j
0.
<
00
S-
<1J
-p
s-
a; •
p oo
03 (U
3 oo
03
C •!-
•r- P
C S-
•r- 0)
(C P
0) T3
S- C
03
to
C >,
O i-
•r- (0
p-o
fO C
s- o
P O
C 0)
0) oo
o
§£
u
p
E c
D 0)
O (0
i- (U
J= S-
O 4->
0)
s-
3
cn
o
OJ
o
r™
O
LO
o
o
o
192
-------
(C
S-
rQ (^
s- l-
(0
u
0
o;
i.
(0
u
-p 0
3 C£
O E
-c T- ^:
•p - T- S- T-
•U] - 2 H S
0 E 0
E =3 E
•r— r— *r-
_j < _j
r
I
L
F
UJ
o
•si-
CD
o
CO
o
CM
o
o
LU
U_
UJ O
Q O
UJ
_!
0.
UJ
00
00
<
o
o
-p
E
0
-p
ta
0
S-
0
-p
<4-
(0
S-
0
•p
,
0) i-
S- fC
w -p
E S.
O OJ
•1- -p
-p
S- C
•P fD
c
OJ >,
u s-
E (0
O -D
O E
O
S- U
0) 0
Q- l/l
Q. .
O >>
O JD
sD
S-
13
CD
'I/Bui
193
-------
Iron
All of the treatment processes investigated were effective in re-
moving iron. The peak in the lime and recarbonation curve shown in
Figure 68 resulted from ferric chloride feed to the aeration basin of
the activated sludge process.. The high-pH lime coagulation train was
the most effective treatment sequence for removing iron, achieving 94
percent by means and 95 percent by medians.
Mercury
Figure 69 showns the mean mercury concentrations at different
points in the treatment sequence for all three sequences studied. The
'high-pH lime coagulation sequence resulted in the best removals of 79
percent by means and 92 percent by medians. The lime and recarbonation
train resulted in the lowest concentrations of 0.08 mg/1 by means and
0.03 mg/1 by medians.
Potassium
No significant potassium removal was observed in any process.
Magnesium
Magnesium was removed by the two treatment sequences utilizing
high-pH lime coagulation. Alum treatment was totally ineffective for
removing magnesium. The best magnesium removal (93 percent by medians)
occurred during the lime-only sequence; however, some difficulty was
experienced during the lime and recarbonation phase in maintaining the
pH at or above 11.0 due to a shipment of low quality lime. One would
expect both processes to be equally effective in magnesium removal.
Manganese
As shown in Figure 70 both treatment sequences employing hiqh-pH
lime coagulation were effective in removing manganese, as was the
activated sludge system. The lime coagulation only sequence resulted
in the greatest removal (94 percent by medians) and the lowest concentra-
tion (0.004 mg/1 by medians).
Molybdenum
No significant removal of molybdenum was observed.
Sodium
No significant removal of sodium was observed.
Nickel
Nickel removals were largest in the lime sequence, 72 percent by
means and 78 percent by medians. The activated sludge process removed
194
-------
-P
c
0
to
OJ
s-
•p
S-
C -P
o
•r- T3
•P C
(0 n3
o; to
O TD
E E
O O
o o
C
O
S-
co
>,
00
^O
LO
O
O
•
o
195
-------
o
I—I
<
CO
o
UJ
<
00
•p
C
QJ
fd
OJ
S-
•p
s-
OJ
4-3
q-
rd
0)
•p
(/)
(0
s
C (/)
•r- 0)
CO
a/ w
E 0)
•r- U
C O
•r- S-
to
CO T-
C -P
O S-
•r- CD
•P 4->
fd
S- T3
•P E
E (0
01
o (d
o -a
E
>> O
S- O
3 0)
U CO
s-
0) >,
5T JD
01
10
0)
.$-
3
Q;
•I—
u.
LO
•
o
•vi-
es
00
6
CM
O
o
o
196
-------
o
0
OL
P
3
O C
1C
S-
.0
S-
ra
o
0
a:
>-
LU
P (0 P
•r- S- T-
0 E CU
E 3 E
•
LU
O
O
/
LU
Q <
O
O
LU
_l
D-
oo
LU
<
(0
,
O S-
fO
0 O
E O
tO 0
cn
-------
a significant amount of nickle in all three treatment sequences studied.
Lead
*
Lead removal was good, generally averaging 60 to 70 percent. The
data shown in Figure 71 indicate that all three treatment sequences
performed approximately the same, with the lime and recarbonation phase
having a slight edge. The activated sludge process was an important
contributor to lead removal in all three sequences.
Selenium
j
As indicated in Figure 72, selenium removals were excellent for all
three treatment sequences. When the median concentrations are evaluated
no selenium was present in the effluents from the alum and the lime and
recarbonation sequences.
Silicon
No truly significant changes in silicon concentrations were observed
during this project.
Strontium
The lime and recarbonation sequence made the most significant re-
duction in strontium concentration, 19 percent by means and 26 percent
by medians.
Vanadium
The high-pH sequences were relatively effective in reducing the
vanadium concentration, with 41 percent (by mean) being the best (lime
and recarbonation). The maximum change in mean concentration was 1 4
mg/1.
Zinc
As shown in Figure 73 zinc removals were excellent except for the
recarbonation study. The zinc from the galvanizing was solubilized by
the carbonic acid which resulted in the obvious increases. All unit
processes were effective in removing zinc to some degree, and an over-
all removal of 85 to 90 percent is a reasonable estimate of the treat-
ment sequence performance.
193
-------
•P
c
0
•P
0
i-
-p
i.
01
0)
-p
01
-p •
oo co
(0 0)
oo
c OJ
•r- O
o
en S-
C Q.
•r- S-
rd fO
0 -P
S- S_
0
C -P
o
•r- T3
-P C
(tS CO
S-
c ^
0 fO
U "D
c c
o u
0
"O 00
rC
OJ >,
_l -Q
0
S-
D
en
199
-------
200
-------
o
LiJ
_l
Q.
<
•p
c
0)
(0
0
S-
•p
0)
•p
Ol
•p
(0
2
01
•p
(/) •
fC 10
S a;
w
C t/)
•i- QJ
u
D> O
C S-
•r- Q
C ,
••- >>
m s?
'E re
OJ-r-
S- -P
5-
10 0)
C -P
o
•r- -D
P C
(0^ fO
5
M J2
00
1^
OJ
i-
^
O!
•r™
LU
CO
*
O
o
o
'ONIZ
201
-------
METALS REMOVAL SUMMARY
This research effort generated an" exhaustive, amount of data which are
relevant to the metals removals that can. be. obtained from AWT processes,
and it is quite impossible to present an tndepth data analyst^ tn any single
document. Many factors such as nitrification, the influent concentration
of the metal, pH, coagulant dose, and product water turbidity can have a
pronounced effect on the removal of a given metal in any given treatment
system. Additionally, the data can be presented in terms of arithmetic
means, geometric means, medians, modes, etc. Conventional practice is to
report arithmetic means, since most effluent limitations are based on
arithmetic means, and that practice was followed in this report to a
considerable extent; however, In many cases significant differences exist
between the arithmetic mean, the geometric mean, and the median concentra-
tion observed.
Table 63 presents the observed changes in mean metals concentrations
that resulted from the chemical treatment process utilized during the three
phases of the research effort. The reasons for increases in calcium,
barium, cobalt, and zinc have been previously discussed. In almost every
instance the absolute value of the change in concentration was very small,
while the percent removal (or increase) was often substantial.
Data relevant to percent metals removal by chemical treatment are
shown in Table 64. Even though changes in concentration may be small,
metals removals of about 40 percent and greater can be considered signifi-
cant. These data indicate that the high-pH lime coagulation processes
are generally more effective for removing metals. This is especially true
of the toxic metals such as arsenic, cadmium, chromium, mercury, and lead.
In addition, the removal of both calcium and zinc would have been greatly
enhanced with proper two-stage recarbonation facilities.
Evaluation of metals removals by the filtration processes does provide
some meaningful insight as to whether the metal was in a soluble form or
in the solids carried over from the upflow clarifier. Table 65 presents
the changes in metals concentrations that resulted from filtration, and
Table 66 shows the percent metals removal effected by the filters. The
very slight increase in copper concentration is most probably an indica-
tion of copper pick-up from piping appurtenances such as bronze valve
seats, etc. The other increases are slight. In this respect aluminum
is a good example. The data in Table 66 indicate increases in aluminum
concentrations of 45 percent for Phase 1 and 40 percent for Phase 2. How-
ever, the data in Table 67 indicate increases in the aluminum concentration
of 0.09 mg/1 and 0.06.mg/l for Phases 1 and 3, respectively. These in-
creases are so slight that one must conclude no significant change in
aluminum concentration resulted from filtration.
Tables 67 and 68 present the observed changes in metals concentrations
and corresponding percent removals as a result of both chemical treatment
and filtration. These data have been evaluated with respect to both percent
202
-------
TABLE 63. CHANGE IN MEAN METALS CONCENTRATIONS BY
CHEMICAL TREATMENT
METAL
PHASE 1
(mg/1)
PHASE 2
(mg/1)
PHASE 3
(mg/D
Ag*
Al
As*
B
Ba
Be*
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
Si
Sr
V*
Zn
+0.68
0.18
8.7
0.01
+0.004
+113.6
0.0
+0.021
0.073
0.052
0.04
0.15
+0.2
3.98
0.041
+2.1
+9.2
0.030
0.005
0.5
+3.8 .
0.01
0.6
0.06
0.0
+1.71
3.3
+0.02
0.008
0.0045
+13.6
1.5
+0.001
0.043
0.020
0.17
+0.06
0.2
0.09
0.018
1.1
+0.1
0.011
0.012
0.1
+0.7
0.3
0.047
0.32
16.7
0.02
+0.009
+41.2
3.9
+6.019
0.041
0.004
0.63
0.09
2.56
0.05
0.3
0.023
0.023
0.9
1.2
0.08
+0.3
+0.521
Concentration in micrograms per liter
+ indicates observed concentration increased
-- Not Available
203
-------
TABLE 64. PERCENT METALS REMOVAL BY CHEMICAL TREATMENT
METAL
Percent Removal
PHASE 1 PHASE 2 PHASE 3
Ag
Al
As
B
Ba
Be
Ca
Cd
Co
Cr
Cu
Fe
Hg
K
Mg
Mn
Mo
Na
Ni
Pb
Se
Si
Sr
V
Zn
+261.5
47.4
44.6
2.4
+5.1
+267.3
0
+56.8
85.9
30.0
11.8
51
+1
80.
80.
.7
.4
,1
.4
+100.0
+8.5
36.6
11.4
20.1
+39.6
4.2
14.3
48.8
0.0
+407
32.7
+5.7
12.3
90.0
+24.4
26.8
+3.8
65.2
37.0
51.5
+23.1
1.6
1.9
32.7
26.2
+0.1
13.9
22.2
9.1
+7.2
7.1
30.1
1
68.1
81.1
5.6
+12.0
+63.5
53.4
+39.6
78.8
5.2
43.4
34.6
51.6
75.8
0.3
21.7
46.0
37.5
12.9
11.1
+10.3
+391.7
+ Indicates observed concentration increased
— Not Available
204
-------
TABLE 55. CHANGE IN MEAN METALS CONCENTRATIONS
BY FILTRATION
METAL
PHASE 1
(mg/D
PHASE 2
PHASE 3
(mg/1)
Ag*
AT
As*
B
Ba
Be
Ca
Cd
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
Si
Sr
V*
Zn
0.0
+0.09
3.8
0.0
+0.009
0.0
0.2
0.001
0.003
+0.1
0.18
0.046
+0.2
0.33
0.0039
2.5
0'.2
0.1
+0.001
+0.14
6.7
0.0-
1.2.
0.015
0.0
1.35
0.7
+0.01
0.001
+0.0115
0.8
0.1
0.0
0.004
+0.022
0.03
0.07
0.2.
+0.07
0.007
+0.7
0.4
+0.002
0.002
+0.2
0.7
+0.1
0.003
+0.06
-0.4
+0.02
+0.01
9.6
0.1
0.005
0.003'
+0.009
0.58
0.05
+0.21
+0.001
7.5
0.011
0.001
+0.2
0.5
+0.02
+0.4
0.122
* Concentration in micrograms per liter
+ Indicates the observed metal concentration increased
— Not Available
205
-------
TABLE 66. PERCENT METALS REMOVAL BY FILTRATION
METAL
PAHSE 1
.Percent Removal
PHASE 2
PHASE 3
Ag
Al
As
B
Ba
Be
Ca
Cd
Co
Cr
Cu
Fe
Hg
K
Mg
Mn
Mo
Na
Ni
Pb
Se
Si
Sr
V
Zn
0.0
+45.0
35.2
0.00
+10.8
0.00
2.7
1.72
25.0
+17.9
60.0
32.9
+1.4
33.3
39.0
59.6
0.17
19.2
+2.6
+7.4
50.0
0.00
33.3
23.8
0.0
63.4
10.3
+2.7
1.8
+2300.
1.2
2.4
0.00
17.4
+64.7
18.8
21.9
1.6
+1.5
18.9
+22.6
0.45
+2.9
4.8
+20.
6.7
+2.6
2.8
+40.0
10.3
+5.9
+11.9
9.0
2.9
7.5
27.3
+12.3
70.8
29.4
+8.8
+6.3
8.3
13.3
3.7
+13.3
6.2
+3.1
+12.5
18.7
+ Indicates the observed concentration increased
— Not Available
206
-------
TABLE 67. CHANGE IN MEAN METALS CONCENTRATIONS BY CHEMICAL
TREATMENT AND FILTRATION
METAL
PHASE 1
PHASE 2
PHASE 3
Ag*
Al
As*
B
Ba
Be
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
Si
Sr
V*
Zn
+0.68
0.09
12.5
0.01
+0.015
+113.6
0.2
+0.02
0.076
0.014
0.22
0.196
+0.4
4.31
0.0449
0.4
+9.0
0.04
0.004
0.36
2.9
0.01
1.8
0.075
0.0
+0.36
4.0
+0.03
0.009
+0.007
+12.8
1.6
+0.001
0.047
+0.002
0.20
0.01
0.40
0.02
0.025
0.4
0.3
0.009
0.014
+0.1
0.00
0.2
0.05
0.26
17.1
0.00
+0.019
+31.6
4.0
+0.014
0.044
+0.005
1.21
0.14
2.35
0.049
7.8
0.34
0.024
0.7
1.7
0.06
+0.7
+0.399
* Concentration in micrograms per liter
+ Indicates the observed concentration increased
-- Not Available
207
-------
TABLE 68. PERCENT METALS REMOVAL BY CHEMICAL TREATMENT
AND FILTRATION
.Percent Removal
METAL
PHASE 1
PHASE 2
PHASE 3
Ag
Al
As
B
Ba
Be
Ca
Cd
Co
Cr
Cu
Fe
Hg
K
Mg
Mn
Mo
Na
Ni
Pb
Se
Si
Sr
V
Zn
+261.5
23.7
64.1
2.4
+16.5
+267.3
2.6
+54.1
89.4
17.5
64.7
67.6
+2.8
86.7
88.0
19.0
+8.3
48.8
9.1
15.0
30.2
4.2
42.9
61.0
0.00 •
+85.7
39.6
+8.6
13.8
+140.0
+22.9
28.6
+3.8
71.2
+3.7
60.6
3.8
3.1
0.43
45.4
9.5
0.34
11.4
25.9
+9.1
0.00
4.8
32.1
55.3
83.0
0.00
+25.3
+48.7
54.8
+29.2
84.6
+6.5
83.4
53.8
47.4
74.2
8.6
32.1
48.0
29.2
18.3
8.3
+24.1
+300.0
+ Indicates observed concentration increased
-- Not Available
208
-------
removal and change in concentration, such that Table 69 could be presented.
. Tables 72 and 74 present data for metals removals by the activated
carbon adsorption process. In all .cases the absolute change in concentra-
tion was small. The carbon columns served as good filters; therefore those
metals which formed precipitates as a result of coagulation were removed
to some extent. Iron, mercury, and copper to a lesser extent, are typical
examples of this removal.
The carbon columns were effective in reducing the selenium concentra-
tion by at least fifty percent in all three phases of the study. Signifi-
cant vanadium removals (20 to 44 percent) were observed in the carbon columns
when high-pH treatment was employed.
The changes in the mean metals concentrations, and the removals ob-
tained by the AWT processes, are presented in Table 72. These data
represent differences between the activated sludge effluent and the effluent
from the activated carbon columns. The AWT processes were effective in
removing arsenic, cadmium, chromium, copper, iron, mercury, magnesium,
manganese, nickel, selenium, vanadium, zinc, and to a lesser extent lead.
The other metals studied during this project'were essentially unaffected
by the AWT processes used.
In any facility being used for potable reuse it will probably be
totally impractical to monitor the concentrations of twenty-plus individual
metals on a continuous basis. In many instances the time required for
analysis of a given metal is several tens of hours; therefore, one of the
major objectives of this research effort that was shared by all members of
the staff, was .the development of parametric relationships, that could be
used to predict the removal of specific metals, without actually analyzing .
for the specific metal.
Tables 73 through 75.present the results of some of the efforts at
parametric correlation. In these studies linear regressions for every metal
were computed as a function of the other individual metals. The linear
correlation analyses were performed for each metal at every point in the
treatment sequence. Under each column, which represents a specific sam-
pling point, the individual metal-metal correlations with a correlation
coefficient greater than |o.80|are listed, along with the number of paired •
data points used for the correlation.
When evaluating these tables it is obvious that the more significant
correlations exist in the lower-quality, wastewaters, which could be anti-
cipated due to the improved analytical accuracy.
Tables 73 and 75 indicate that aluminum is a good indicator for a
number of other metals, including mercury, cadmium, strontium, and
chromium. Table 74 strongly suggests that beryllium should be a good
indicator for chromium, iron manganese, nickle, and lead.
209
-------
I
TABLE 69. SUMMARY OF METALS REMOVAL DEMONSTRATED.BY CHEMICAL
TREATMENT AND FILTRATION PROCESSES ;
OBSERVED
PERFORMANCE
HIGH-pH LIME COAGULATION
AND RECARBONAT10N
ALUM
COAGULATION
Significant
Removals
Observed
AT
As
Cd
Cr
Fe
Hg
Mg
Mn
Ni
Pb
Zn
Se
As
Cr
Fe
Mn
Pb
Zn
No Significant
Removal
Observed
B
Ba
Cu
K
Mo
Na
Si
Sr
V
B
Ba
Cu
K
Mo
Na
Ni
Si
Sr
V
210
-------
TABLE 7o. CHANGE'.IN MEAN METAL CONCENTRATION
BY ACTIVATED CARBON ADSORPTION
METAL
PHASE 1
(mg/1)
PHASE 2
(mg/1)
PHASE 3
(mg/1)
Ag*
Al
As*
B
Ba
Be*
Ca
Cd*
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
Si
Sr
V*
Zn
0.32
+0.07
0.6
0.02
0.001
11.2
0.2
0.003
+0.0005
+0.005
0.056
+0.016
0.2
0.08
+0.0006
+0.2
+0.4
0.013
+0.003
1.04
+2.1
0.0
+0.5
0.007
0.0
0.18
0.3
0.03
0.0
0.002
2.4
0.2
0.002
0.003
0.017
0.04
0.03
0.7
0.1
0.007
0.6
2.5
0.012
0.0
0.6
+0.2
+0.3
0.039
0.05
0.0
0.04
0.008
+4.7
0.3
+0.001
0.002
0.028
0.04
0.04
0.15
0.0
0.8
+0.006
+0.001
1.0
0.3
0.02
1.6
+0.085
* micrograms per liter
— no data available
+ increase in concentration
211
-------
TABLE 71. PERCENT METALS REMOVAL BY ACTIVATED
CARBON ADSORPTION
Percent Removal
METAL
PHASE 1
PHASE 2
PHASE 3
Ag
Al
As
B
Ba
Be
Ca
Cd
Co
Cr
Cu
Fe
Hg
K
Mg
Mn
Mo
Na
Ni
Pb
Se
Si
Sr
V
Zn
34.0
+24.1
8.6
4.9
1.1
7.2
2.7
5.3
+ 5.5
+ 7.6
46.7
+17.0
1.4
12.1
+ 9.8
+11.8
+ 0.3
30.1
7.5
51.0
+31.3
0.0
+20.8
14.6
0.0
23.1
4.9
7.9
0.0
16.7
3.5
5.0
7.4
15.8
30.4
30.8
12.0
5.7
2.1
23.3
15.8
2.8
17.1
0.0
50.0
+ 2.1
+ 7.5
36.8
23.8
0.0
11.1
8.5
+ 4.9
9.1
+ 1.6
25.0
34.1
16.7
33.3
5.7
0.0
0.9
+ 8.3
+ 3.8
58.8
3.9
3.0
44.4
+16.0
— no data available
+ increase
212
-------
TABLE 72. CHANGE IN MEAN METAL CONCENTRATION AND PERCENT
REMOVAL FOR THE AWT PROCESSES '
METAL
Ag*
Al
As*
B
,Ba
Be
Ca
Cd
Co
Cr
Cu
Fe
Hg*
K
Mg
Mn
Mo*
Na
Ni
Pb
Se*
Si
Sr
v*
Zn
PHASE
mg/1
+ 0.36
0.02
13.1
0.02
+ 0.012
-
+102.4
0.4
+ 0.017
0.076
0.009
0.28
0.18
+ 0.2
4.39
0.044
0.2
+ 9.4
0.053
0.001
1.4
0.8
0.01
1.3
0.082
1
(°/\
(/o)
(+138)
(5.3)
(67.2)
(4.8)
(+15.2)
f -- }
(+241)
( 5.3)
(+45.9)
( 88.8)
( H.3)
( 81.2)
(62.1)
(+1.4)
( 88.3)
( :863)
( 9.5)
(+8.7)
( 64.6)
( 2.3)
( 5:8.3)
( 8.3)
( 4.2)
( 30.9)
( 66.7)
PHASE 2
mg/1
0.0
+ 0.18
4.3
0.0
0.009
+.0.005
+10.4
1.8
0.001
0.05
0.015
0.24
0.04
1.1
0.12
0.032 .
1.0
2.8
0.021
0.014
0.5
+ 0.2
+ 0.1
0.089
(%}
( o.o)
(+42.9)
( 46.6)
( o.o)
( 13.8)
(+100 )
(+18.6)
( 32.1)
(. 3.8)
( 75.8)
( 27.8)
( 72.7)
( 15.4)
( 8.7)
( 2.6)
( 58.2)
( 23.8)
( 3.1)
( 26.6)
( 25.9)
( 45.4)
(+2.1)
( " )
(+2.4)
( 57.1)
PHASE
mg/1
0.31
17.1
0.04
+ O.OM
—
+36.3
4.3
+ 0.015
0.046
0.023
1.25
0.18
—
2.5
0.049
__
8.6
0.028
0.023
1.7
2.0
0.08
. 0.9
+0.484
3
(*)
(-- )
(66.0)
(83.0)
(11.1)
( +14.7)
(+55.9)
(58.9)
(+31.2)
(88.5)
(29.9)
(88.2)
(69.2)
( — )
(50.4)
(74.2)
( — )
(9.5)
(26.4)
(46.0)
(70.8)
(21.5)
(11.1)
(31.0)
(+370)
*. nricrograms- per liter
+ increase in concentration
— no data available
213
-------
LU
gr
Z»-<
Px
3?-
LU X
a: CD
ct: t-i
ox
0
n
a: a:
< LU
LUH
z<
M LU Z
_io;o
COM
CD h-
Z Qi eC
M O Z
< O CQ
x coo;
CO O O
Z LU
O Lu C£
MO
h- H
t N-I'?
Z h-
LU 00 M
OH-2
Z Z
OLU h-
O M z
O LU
oo MS:
_) LuH
1— LU LU
LU O Oi
SOI-
t
CO
'^
LU
CQ
H
•X
OZ
CQ LU
air)
OLU
u.
LU
•X
fV [—
LUZ
1— LU
__j ^
M — 1
Lull.
Lu
LU
oil-
O Z
f-LU
OO _l
"^ t I
2. LU
1 1 1 1 1
LLJ U_
O LU
h-
LU
• _J
CO LU
• u_
< LU
1-
^y
LU
^3
00 LU
• Z
LU
O
. >
ft
• S-
* OO
O CT>
CO « •>
« JO T-
E D. CO
M
ft n 1
3 O r-
0 S i . <
en
E
'>
•»—N r~~
M 0
^-^ * -X * >
^j* *^f* ^j* . 'xj* ^j* . ^st* ^r **j* ^d" ^3* c
(d fd • "O • O fr- jQ " i. • i. i-'r-
CQOOOZD->COOOOOCO WE
E O
tf> to "o o to n
<;eCca;e£cC
T3fl3D)i-S-i- "E-r-i- •> M-'r-
OOXCOCOCOJDNOOOO> -P
a. id rd
t\ t\ t\ t\ ^ fi r\ t\ t\ • f\ t\ A «f-J ^p)
i— i— r— i— Id O S- Z3 (d'r- •!- Ofd0
<<<
(dS-0)T-rs_O" OWE
caoLuzjgoos> O^EM
ftftftftftftftf. o'1"1-^^
i— i— i— i— i— r-C-r- -XS-H
<<<<<
-------
2:
o
i— i
LU S
on 3
o; -i
o<
o
• ft
< LU
LU J-
Z <
I-H LU
CJ3
CJ3
S 0
§0
I CO
00 CD
2:
Ou_
HH f^
ef^-s
C£. i.
1
2:
LU 00
01-
2: 2:
OLU h-
O LU
oo I-H s:
-J Lu I—
1— LU LU
LUOQ;
2! O I—
^J"*
w
LU
CQ
H
o 2:
CQ LU
c£ ID
OLU
U_
LU
C£ |
LU 2:
1- LU
1 1 1 1
U_
LU
Dil-
0,2:
1— LU
(/) |
2: u.
LU U-
Q LU
h-
2:
LU
ID
OO Lu
• Li_
< LU ,
1-
2:
LU
00 LU
• 2:
< M
*
^v.
• ft
O5
S
rt
CQ
LO
ft
0)
CQ
ft
<
*
00
ft
(U
CQ
ft
<
£
|^s
ft
O
ft
p—
eC
CTi
\^/
ft
CQ
#
n ^>,
s_
U n3
*—JP
•Si-
ft
0)
CQ ^£
* — s 1— (
1— 1 <• —
LO st *± LO
CO ft ft « «
" S- C . T- JD
05 0 S • 2 D.
rt ft ft rt
" 0) 0) 0 0)
CQ CQ CQ CQ CQ
O"> r—
r— i—
t nj ft ft
O Lu S£ (Q
2:
(U QJ T3 ft
CQ CQ O ^
i— CM
r— r—
ft CTi "
•p- ft »r—
OO (8 00
2:
ft ft
Ol « JQ
T* ^ n
O
O
,_
^~^ V
i— i
* — * S-
o * * en
CTl r— CD CO i v
O5 > i^ i^ 2! O
^>^ 0^
' rt ft > ft ft ,
"
C
1 — 1
^^
1 — 1
215
-------
UJ
S
o!H
i— i
< Q.
uj ac
ct: to
f*S iTrrrl
[,Ta r™"l
o n:
o
0:0:
< UJ
UJ H
z<
t-H UJ
_1 Di
to
CO Z
z 0:0
hH Cj *""•
> h-
XCOZ
•O
c/> oca
z ct:
OU-<
1— U)
DiT01
UJ 00 w
OH3:
z z
O UJ H
O t-HZ
O UJ
cows
_ju.i-
1— UJ UJ
ui o ct:
*
Lf)
r**N
UJ
CO
f
H
O UJ
S -1
1"
rt
C
5:
rt
cu
u.
* 10
cvic\iCMC\jc\iLOLn"~i"Ln^2!x}.
•^••^•<^"<^"^f"lNi"<:;l" i>r'vr'srt3"
65£i£i£i£i2!lQ.'«/5
<<<5
-------
The authors must emphasize the fact that the indirect parametric
determination of any water quality parameter is subject to considerable
debate. The metals data developed during this project are not sufficient
to substantiate any definitive parametric, correlations. One. of the more
important facts learned during this project was that it was totally
impossible to predict the concentration of a specific metal at any given
point in time and process.
The percent of the total samples in the influent and effluent to
each of the three treatment sequences that exceeded the drinking water
limits is presented in Table 75. Although when viewed in this manner
the total number of metals which pose treatment problems are small., the
metals that pose the problems all have demonstrated public health hazards.
It should"be noted that a metal that is present at the maximum contaminant
level (MCL) one day per year has a. recurrence percentage greater than
0.3 percent.
A summary of metals removals is presented in Table 77. When
evaluated in terms of maximum effectiveness for removing metals the lime
and recarbonation sequence would appear to be the most effective. These
data were not obtained over the same time period, or over identical
sampling intervals, which makes data evaluation most difficult.
The use of high-pH lime coagulation without recarbonation is very
unlikely; therefore; this treatment process is not included in the
summary. However, certain data from Phase 1 were utilized to predict the
performance of the Phase 3 system. Zinc removals should be good in the
recarbonation system, provided proper materials of construction are used.
Also the magnesium removals observed during Phase 1 of the study should
be typical of a Phase 3 system. During Phase 1 the upflow clarifier was
operated at a pH of 11.5, which was sufficient to precipitate magnesium
as a hydroxide. However, during Phase 3 the operating pH was 10.5;
too low for excellent magnesium precipitation.
From the data presented one must conclude that high-pH lime coagula-
tion and recarbonation is more effective in removing metals than is alum
coagulation. These observations have been made after an extensive inves-
tigation of advanced wastewater treatment processes on wastewater from
the City of Dallas, Texas, where the industrial component of the waste-
water flow is about seventeen percent. Regional and/or geographical
differences could alter the reported results substantially.
Table 78 may be the most significant approach to evaluating metals
removal. In this table the concentration of each metal in the final
product water from each treatment sequence is evaluated as a function of
the maximum contaminant level (MCL).
When the data in the Table 78 are carefully evaluated the following
facts cannot be escaped; arsenic, cadmium, iron, manganese, and lead
represent potential problems of undefined magnitude.'
217
-------
,10
LU
OOOOi—C\lOOOO'3-C\JOO
CVJ
o en
-o
l.
(0
T3
C
(0
•P
GO
LU
O
UJ
LU
I OOOOOOU3OOOOO
0)
4->
(0
LU
u_
in
cr> oo
a:
a
CD
r- C
OJ JC
> O
C1J S-
i— (0
U
(1)
00
UJ
O
C£.
UJ
a.
LO LO
i— LO
OJ
O Lf) LO
i CD
OOOOOOOCOOOOOO
OOi—i—OOi— OOOOOLO
(0 i—
E LO
(C O
O
3 -P
LU
T3
CU
CO
LU
D)
CO M
X en
(0 CU
2:0;
O
(U
218
-------
TABLE 77. COMPARISON OF OBSERVED METALS REMOVALS
Metal Removal
(percent)
Phase 1 Phase 2 Phase 3
Lime Alum Lime and Recarbonation
90
Cr
Fe
Se
Mn
Cr
Fe
Cr
Se
70 - 90
Mg
Zn
Hg
Zn
Cu
Se
Mo
Mn
As
Mn
Cd
Al
Fe
Hg
Pb
Cu
50 - 70
As
Al
Ni
Pb
Cd
Pb
Hg
Ba
Mg
20 - 50
Cu
Cd
Ba
V
As
Ni
Be
Ni
V
Ba
Si
0 - 20
Sr
Si
Ag
B
Co
K
V
Na
Mg
Al
Si
B
Sr
Na
B
Increase
in Cone.
Na
Co
Ca
Ca
Co
Ca
Zn
219
-------
TABLE 78.
METALS CONCENTRATIONS IN AWT EFFLUENTS AS A PERCENT
OF THE MAXIMUM RECOMMENDED CONCENTRATION
Metal
Ag
As
B
Ba
Cd
Cr
Cu
Fe
Hg
Mn
Pb
Se
Zn
MRC*
0.05
0.01
1.0
1.0
0.01
0.05
1.0
0.3
0.002
0.05
0.05
0.01
5.0
Percent
Lime
1.2
64
39
0.9
72
19
7.1
21.3
5.5
13.4
86.0
10
0.8
of Maximum Recommended Concentration
Alum
0
58 .
35
5.6
38
32
3.9
30
11
46
84
6
1.3
Lime and Recarbonation
—
35
32
8.6
30
12
5.4
66.7
4.0
34
52 >'
7
12.3
* Suggested by the most stringent value published
(PEAW Standards).
220
-------
SECTION 10
VIRUS MOTIVATION STUDY
GENERAL
Water supply research studies of virus removals in flocculation and
settling processes indicated removals in the order of 95-99 percent(10, 11,
•12). These studies were based primarily on the use of alum and included
an observation that higher usage of chemicals increased the virus removal
efficiency. Studies at the University of Illinois using bacteriophage
showed 98 to 99.9 percent removals (13). These same studies were expanded
to include wastewater and showed the effects of increasing concentrations
of organics on depressing virus removal efficiencies. These data indicate
that virus removals from wastewater by flocculation and settling processes
can be expected to be in the lower range of the efficiencies cited above.
As far as lime flocculation is concerned, the first reports of the
effects of high-pH lime treatment in destroying bacteria date back to the
1920's (14, 15). More recent and detailed observations of these effects
by South African workers report marked differences in the response of gram-
negative, gram-positive and acid-fast bacteria. The gram-negatives were
the more.susceptible, and the acid-fast were the more resistant, excepting
the spores (16). The survival of only a.relatively few gram-positive rods
was precisely what was observed in Dallas, with the critical pH-contact
time relationship appearing to exist in the range of pH of 11.2-11.3 and
a time of 1.56 to 2.40 hours (17). Berg, et al_. at the National Environ-
mental Research Center (NERC) in Cincinnati extended the studies of lime
flocculation and high pH to poliovirus type 1 and observed removals of 70
to 99.86 percent from innoculated secondary effluent, with higher removals
occurring at higher pH's (18). When coupled with sand filtration, the
removals increased to levels greater than 99.997 percent.
Operation of the Pilot Plant
An activated sludge effluent was processed in the upflow clarifier at
a rate of 6.31 I/sec (100 gpm). The sludge blanket in the upflow clarifier
was dropped to approximately the level of the flocculator outlet at the
start of a run. The fluidizer bar was turned off, and no sludge was with-
drawn or recycled during the run. A residence time distribution function
study at this flow regime (Figure 74) showed that in spite of the 5 hour
and 10 minute theoretical detention time, peak dye concentration in the
effluent occurred two hours after the addition of a slug load. Nevertheless,
221
-------
CU
s:
-p
UJ
•p
u
O •!-
0)
CD S-
O 0)
0)
co rd
OJ t—
Di U
cu
CO
NOIlVyiN30N03 3AQ 1M 3NIN\/QOHy
222
-------
seeding of the poliovirus-phage mixture continued for 10 hours to assure
the achievement of a steady-state condition. The comparison of results -
obtained during the last five hours would then be used in determining
separation efficiency.
The polioviruses were handled as though they were pathogenic organisms.
The effluent from the upflow clarifier was directed to a chlorine contact
basin where it was subjected to a dose of chlorine of 25 rag/1 free residual
for a period of 30 minutes, prior to discharge to the raw sewage of the
White Rock Plant. It was believed that this treatment would destroy the
viruses (19). The chlorine contact basin was sampled at different points
to check on the virological performance of the disinfection process.
Immediately after the last sampling, the upflow clarifier was taken
out of service. Chlorine was admitted to achieve a 40 mg/1 free residual
level, the fluidizer bar was turned on, the sludge recycle pump was turned
on and the chlorine residual was maintained for 12 hours prior to dumping
the contents to the raw wastewater inlet of the White Rock Plant. This
treatment resulted in a bleached sludge, and it was hoped that this would
destroy all the viruses. Sludge samples were taken before and after the
chlorination process and assayed for disinfection performance.
RESULTS AND DISCUSSIONS OF THE VIRUS RUNS
Virus Run Number One
The frozen poliovirus stock was thawed in a water bath at 27°C. The
thawed virus was aseptically divided into three gas-sterilized is 9-liter
Cubitainers each containing 18.5 liters of sterile distilled water. A
quantity of f2 coliphage was added to each container to give a countable
concentration in the effluent, assuming about 99 percent removal. One of
the three Cubitainers was taken to the plant, immersed in an ice water
bath, and fed into the suction of the pump feeding the upflow clarifier
through a Gilmont No. 13 Flowmeter at 100 ml/min. The other two Cubi-
tainers were stored in a refrigerator, and substituted in turn for the
depleted containers with only momentary seeding disruptions. Each
container was well agitated and sampled prior to the beginning of the
seeding to check the diluted titer.
During this run the upflow clarifier was operated at a liquid alum
feed concentration of 103 mg/1 as the hydrated molecule. This was the
lowest concentration that provided a visual appearance of a high quality
water passing over the effluent weir. Turbidity as monitored by the Hach
Continuous Flow Surface Scatter Trubidimeter on the day of the run
averaged 0.3 units. The A1:P ratio was only 0.44.
Chemical tests were run frequently over the 10-hour seeding and
sampling period. Table 79 gives the results of samplings taken at the
beginning of the run, at a midpoint, and again at the conclusion. Some
analyses were also performed on shipped samples by NERC Cincinnati. Where
223
-------
TABLE 79. CHEMICAL-PHYSICAL ANALYTICAL RESULTS, VIRUS RUN NUMBER 1
TIME
5:00 AM
10:00 AM
3:00 PM
INFLUENT* EFFLUENT* INFLUENT* EFFLUENT* INFLUENT* EFFLUENT*
Temp.0 C
pH, units
29
7.1
Alk.as CaCO 124
Cond. umhos 740
D.O.
Hardness
as CaCOa
Color
Turb.FTU
COD
NH3-N
Org. N
N02+N03-N
SS
Total P
3.9
130
50/40
8.5/11.0
56/52
0.4/0.4
4.3/3.8
4.0/4.9
24/27
11.3/10.
29
6.8
120
766
4.8
114
50.35
2.0/2.5
30/32
0.1/0.8
2.7/1.1
3.5/3.7
5/3
6 6.5/7.8
29
7.1
126
874
4.4
126
50.40
9.0/11.0
-/43
0.0/0.1
4.1/2.4
4.0/4.6
30/16
9.3/9.2
29
6.7
no
908
4.5
116
30
1.7
—
0.0
2.4
3.5
6
5.0
29
7.1
124
738
4.1
128
50/40
9.3/12.0
60/52
0.0/<0.1
4.9/3.7
5.0/5.9
32/24
7.5/8.1
29
6.9
110
772
4.4
120
40
2.5
,39
0.0
2.7
5.0
7
5.8
* Influent to and effluent from Densator
Denominator values are NERC analytical results on shipped samples.
Concentration in mg/1 unless otherwise noted.
224
-------
the time of those samples coincide with similar samples run by the Dallas
staff, the NERC results are shown as a denominator value on the table.
Coliphage analyses on the three seeding containers showed:
Container 1 3.4 x 107 pfu/ml
Container 2
Container 3
Average
1.0 x 107 pfu/ml
2.8 x 107pfu/m1
2.4 x 10' pfu/ml
At the 100 mi/min seeding and 6.31 I/sec (100 gpm) flow, an average
influent titer for the/coliphage of 6360 pfu/ml should have been expected.
The results of the virus samplings are shown in Table 80 for coliphage
and Table 81 for poliovirus. Influent and effluent analyses for both are
depicted graphically.in Figure 75. Calculated removals were 46 percent
for coliphage and 63 percent for polioviruses. All virus samples from
the chlorine contact basin were zero. Poliovirus analyses of the non-
chlorinated sludge yielded 114 pfu/10 gm of sample. Chlorinated sludges
were negative for both poliovirus and coliphage. However, the MPN tubes
on the chlorinated sludge yielded positives for coliforms. Since virus
analyses on the chlorinated sludges were negative, a mixing problem was
suspected. The upflow clarifier was drained and a thick layer of sludge
was found to exist under the fluidizer bar. Rubber scapers were then
mounted on the bar and three air-lift pumps were installed to take the
water from just above the fluidizer bar back to the mixing zone. Two '
test disinfection runs were performed and yielded no positive coliform
tubes.
Virus Run Number Two
In the case of water treatment, it has been emphasized that coagulation
and filtration are really one process and must.be studied together (12).
The same logic can be applied to wastewater in a reclamation operation.
In preparing for the second run, therefore, some preliminary work was done
to determine how to disinfect a mixed-media filter. It was decided to add
calcium hypochlorite to the backwash water tank until a 40 mg/1 free
residual was achieved. This was then admitted to the filter and the filter
allowed to soak overnight. The filter was then backwashed.
The upflow clarifier was still operating on an alum feed. For this run,
the control parameter was not "Appearance", but rather the aluminum to
phosphorus ratio. An effort was made to maintain the ratio in excess of
3.7:1 (20).
Much the same chemistry was run as in Virus Run Number One except
that Tablet, showing the results at the beginning, mid-point, and end
of run, also shows an entry for BOD5. The same tests were run on the
225
-------
TABLE 80. COLIPHA6E RESULTS, VIRUS RUN NUMBER ONE
SAMPLE
INFLUENT
pfu/ml
EFFLUENT
pfu/ml
BACKGROUND
6 AM
7 AM
8 AM
9 AM
10JM
11 AM
12 N
1 PM
2 PM
3 PM
160
5240
4000
5400
5500
J400.
5600
4800
6000
5800
2400
156
600
640
460
3200
3000
3000
2600
3200
1800,.
Average last five hours 4920
Percent removal: 46 Percent
2680
TABLE 81. POLIOVIRUS RESULTS, VIRUS RUN NUMBER ONE
SAMPLE
BACKGROUND
6 AM
7 AM
8 AM
9 AM
10 AM
11 AM
12 N
1 PM
2 PM
3 PM
INFLUENT
pfu/ml
0
510
620
550
440
400
660
620
590
610
500
EFFLUENT
pfu/ml
0
11
33
36
34
52
135
208
289
264
192
Average
Percent removal
596
63 Percent
218
226
-------
0)
G
o
i-
0)
&
£
15
c
3
C£.
to
3
J.
CD-
E
•i—
S-
3
TD
0)
C
3 O
O
.C
C
0)
3
0)
C
(0
•p
C
OJ
3
M-
C
LT5
0)
i.
3
O1
2OL x [ui/njd «snaiAOITOd I 3dAi
227
-------
TABLE 82. • CHEMICAL-PHYSICAL ANALYTICAL RESULTS, VIRUS RUN NUMBER TWO
TIME
5:00
AM
INFLUENT* EFFLUENT*
Temp. ° C
pH, units
Alk. as
CaCOo
0
Cond. umhos
D.O.
Hardness
as CaC03
Color
Turb.
**FTU
COD
BOD5
NH3-N
Org. N
N02+N03~N
SS*
Total P **
22
7.1
145
574
4.8
183
40/22
4.5/5.6
28.49
12
0.3<.l
2 .4/2. 7
5.0/6.0
12/7.5
3.5/3.9
22
6.5
60
649
4.6
183
15
0.5(0
20
1
0.3
1.5
3.5
5(3)
0.2(0
10
INFLUENT*
22
7.3
170
597
5.3
253
30/27
.4) 3.5/5,5
36/65
10
O.K.I
2.5/3.1
6.5/5.6
16/10
.2) 3.0/2.7
:00 AM
3
EFFLUENT* INFLUENT*
22
7.2
no
658
• 5.9
243
10
0.6(0.4)
12
1
0.3
1.7
6.0
1 (2)
0.1(0.1)
22
7/4
175
629
7.7
222
30/13
3.3/8.
36/60
7
O.K.I
2.4/3.
7.0/7.
12/16
3.0/2.
:00 PM
EFFLUENT*
22
6.7
no
686
7.9
258
10
0 2.5(0.3)
24
2
0.3
1 1.5
6 6.0
12 (3)
9 0.3(0.1)
* Influent to and effluent from Densator
** Values in parentheses on filter effluent.
Denominator values are NERC analytical results on shipped samples.
Concentrations mg/1 unless otherwise noted.
228
-------
No. 1 mixed media filter effluent. The filter effluent results for tur-
bidity, suspended solids, and total phosphorus are shown in parentheses.
The A1:P ratio was 7:1, due primarily to unexpectedly low influent total
phosphorus values.
. The seeding technique for this run was similar to Virus Run Number
One except that the 5-gallon Cubitainers were made up freshly before each
use. Additionally, assays for both poliovirus and colipnage were conducted
with the following results:
Container
No.
Time of Use
Poliovirus
pfu/ml
Coliphage
pfu/ml
5:30 - 9:00
9:00 -12:35
12:35 - 3:00
Average
2.0 x 106
2.3 x 106
2.3 x IP6
2.2 x 106
1.2 x 107
1.4 x 107
1.0 x IP7
1.2 x 107
At the same 100 ml/min. seeding and 6.31 I/sec flow rates, an. average
influent titer of 580 pfu/ml poliovirus and 3170 pfu/ml coliphage should
have been expected. Cooler weather prevailed during this run and the
seeding containers were therefore not iced.
The results of the virus samplings for coliphage are shown in Table
83 and for polioviruses in Table 84. Figure 76 shows the influent titers
graphically; effluent recoveries were too small to show on a similar scale.
Calculated removals for coliphage were 99.845 percent on the upflow
clarifier alone and 99.985 percent on coagulation plus filtration. No
polioviruses were recovered on the effluents, and based on a sensitivity
of assay of 0.5 pfu/ml, a greater-than 99.7 percent removal was calculated.
Again, all virus samples from the chlorine contact basin were zero.
Poliovirus titers on the non-chlorinated sludge yielded 223 pfu/gm of
sample. Chlorinated sludges were negative for poliovirus, coliphage, and
coliforms. Filter backwash water was similarly^negative.
Virus Run Number Three
In order to get as high a titer as possible, some staff members
wanted to slug dose the upflow clarifier with the two viruses. Others,
who were more acute to the relatively higher pH-resistance of the
polioviruses, opted for the same type of seeding technique used in the
alum studies. A compromise was effected in which it was agreed to add
the viruses in a steady stream over about a five-minute period. Samples of
the seeded influent were taken midway through the seeding with the
following results:
229
-------
TABLE 83. COLIPHAGE RESULTS, VIRUS RUN NUMBER TWO
SAMPLE
BACKGROUND
1
2
3
4
5
6
7
8
9
10
INFLUENT
pfu/ml
54
3200
2800
2400
3600
3400
2800
2700
2600
2200
2600
EFFLUENT
pfu/ml
0
0
0
0
2
2
4
4
6
6
0
FILTERED
pfu/ml
0
0
0
0
0
0
0
0
0
2
0
Average last 2580
five hours
Percent removal:
Densator: 99.845
Densator plus filter: 99.985
0.4
TABLE 84. F
SAMPLE
BACKGROUND
1
2
3
4
5
6
7
8
9
10
Average last
five hours
(Sensitivity
'OLIOVIRUS F
INFLUENT
pfu/ml
0
no
118
156
109
109
132
108
101
86 .
139
113
of assay =
lESULTS.VIRUS RUN NUMB
EFFLUENT
pfu/ml
0
0
0
0
0
0
0
0
0
0
0
0.5 pfu/ml)
ER TWO
FILTERED
pfu/ml
0
0
0
0
0
0
0
0
0
o.
0
Percent removal: >99.7
230
-------
|W/rud 'SdNIAOnOd I 3dAl
o
LO
o
o
o
to
CO
o
5
s-
0
JD
E
2
2:
E
13
Qi
CO
3
i.
CT;
E
T3
00 XI
S- 0)
rs E
O 'I-'
-E to
LO
r J2
LU O
S
I— I CO
I- -P
in E
' 15
O
O
CO
•p
E
CO CD
0)
S-
3
CD
00
C\J
s39VHdOiy310V8 ZJ
231
-------
TABLE 85. CHEMICAL-PHYSICAL ANALYTICAL RESULTS, VIRUS RUN NUMBER THREE
TIME -5: 15 AM
INFLUENT1 EFFLUENT2 INFLUENT1
Temp0 C 20 20 19
pH3, units 7.0 11.8-6.3 7.5
Alk. as
CaC03 150 195 142
Cond.
umhos 745 833 778
Color 40/20 0/11 50
Turb.
FTU 11.0/14 1.5/6.4 8.0
COD 40/34 28/15 ' 481
NH3-N 0.0/0.2 0.0/0.1 0.0
Org. N 2.7/2.2 1.0/0.8 2.7
N02+N03-N 12.0/10.7 9.0/10.4 11.5
SS 20/14 11/19 32
Total P 6*5/4.9 0.26/0.3 7.5
1 1nfluent to Densator
^ Effluent from Recarbonation
10:15 AM
EFFLUENT2
19
11.9-9
152
800
0
2.2
28
0.0
1.3
5.0
142
0.2
3:15
AM
INFLUENT1 EFFLUENT2
19 '
.6 7.2
136
760
50
8.5
48
• o.o
3.4
12.5
32
'7.5
19
11.7-8.3
332
980
5
1.0
16
0.0
1.8
11.5
61
0.33
\
3
Effluent pH values before and after recarbonation
Denominator values are NERC analytical
Concentration in mg/1 unless otherwise
results on shipped samples .
noted.
232
-------
TABLE 85. COLIPHAGE RESULTS, VIRUS RUN NUMBER THREE
SAMPLE
INFLUENT
pfu/ml
EFFLUENT
pfu/ml
BACKGROUND
89
Seeding 7.2 x 105
(4 min. 20 sees. 4400 mis mixed viruses)
All effluent samples
(Taken each 30 minutes for 10 hours)
Stock titer:
7.2 x 107 pfu/ml
TABLE 87. POLIOVIRUS RESULTS, VIRUS RUN NUMBER THREE
SAMPLE
INFLUENT
pfu/ml
EFFLUENT
pfu/ml
BACKGROUND
0
Seeding 8 x 10
(4 min. 20 sees., 4400 mis mixed viruses)
All effluent samples
(Taken each 30 minutes 'for 10 hours)
Stock titer:
7.0 x 107 pfu/ml
233
-------
oo
H
_J
r)
00
LU
o;
Q
^^
H
co
3C
I—
5
o
a:
CD
LU
a:
00
Q C£
LU 3
h-0
•z.
Ost
CQ CM
Qi
0=)
LU _
>X o
h-
2Z
LU
~~i
P— J
LU
LU
LU
z.
o
1— 1
1—
z.
o
m
a:
o
LU
a:
H
LU
LU
LU
LU
oi
o
c^
00
LU
O
1
E
CD
^s
•p
C
O
o
r~
r— g:
(0
•P i.
O 0
I- Q.
1
E
CD
53
•P
C
^
O
o
1—
r- E
•P S-
O 0
h- Q-
Q.
1
CD
o
•P
3
O
Or-
E
"(0 i-
•P 0
o a
—
X
a.
c
O 1 1 LO O 1
1 1 IV hx |-
•5t O 0 O 00 O
O|— O st- O
O Z. LO CO O
I— | — n r-
A cn A
r—
LO 1 O O O O
1 LO VD *£> ID
LO st- O sj- CO 00
CM CO ID CT> ID
CM CM
Sj- CM Si". O CO LO
rv iv & tv U3 rv
o p o o o o
.
O «D CO i—
CM r— LO CM CM VD
r— r— .
O 00 SJ- ^ CO CO
CM i— O i— i— i—
r— i— r- r— r- r—
oo
oo
LU
_J
QQ
•P
(0
a
0
r—
Q.
(0
00
CO CO CO CO CO 00
CM co sf- LO vo rx
r- r— p— i— i— i—
\ \ ^x \ \ \
CM CM CM CM CM CM
234
-------
Coliphage liter
Poliovirus liter
7.2 x TO7 pfu/ml
8.0 x 10* pfu/ml
The Densator was operated on a lime feed calculated as 391 mg/1 as CaO
and a supplemental feed of 18.6 mg/1 ferric chloride as Fe. The results
of the chemical analyses at three different times during this run are
shown in Table85. The effluent samples were taken after recarbonation
— as were the virus samples. The results of the virus analyses are shown
in Tables86 and87 ; all effluent samples (one sample for each virus every
30 minutes over a 10-hour period following the seeding) were zero. The
results of all the control samples on the terminal chlorination process
and on the sludges were also zero. In this high-pH lime run, the Densator
itself was not terminally chlorinated. Instead, the unit was isolated and
the sludges allowed to mix for a number of days until results of the
poliovirus assay were received from Cincinnati indicating there were no
polioviruses left.
Although plant control, especially with the recarbonation process, was
not as good as desired, the virus results obtained substantiated earlier
observations of effectiveness against coliphage f2. At this point in time
a new problem developed. Routine bacterial tests for total coliforms,
fecal coliforms and total plate counts were showing a possible regrowth
of these organisms in the recarbonated effluent (see Table 88). In
another check of regrowth, a one-liter grab sample of Densator effluent
was taken to the lab and a six-replicate total plant count was performed.
The sample was recarbonated using aseptic technique and another six-plate
count was made. Lastly, the sample was held at room temperature for 24
hours and a third six-plate count run. The results are shown in Table
89.
TABLE 89.
LABORATORY RECARBONATION AND REGROWTH
STUDY RESULTS
SAMPLE
p'H
Total Count/ml
Avg. of 6 plates
Range of
Counts
Densator 11.1
Effluent
Recarbonated 6.8
24 Hours
6.6
11.2
10.5
6,000
5-20
6-17
5700-7400
Virus Run Number Four
The fourth and last virus run duplicated Virus Run Number Three'except
for some slight changes. Better control over recarbonation was effected.
Quicker feeding of the seed virus was accomplished, Soluble TOC's were
performed rather than COD's. Finally, virus sampling was "limited to 14
samples collected over an 8-hour period (0,
2,2%, 3, 3^, 4, 4%, 5,
235
-------
53s, 6, 7, and 8 hours after seeding) instead of every half hour over a
10-hour period. Samples of the seeded influent water coming to the
Densator taken half-way through the seeding showed the following titers:
Coliphage liter
Poliovirus liter
2 x 101* pfu/ml
4 x 10" pfu/ml
Chemicals used in the Densator for this run were 383 mg/1 lime as CaO and
9.3 mg/1 of Fe(Cl)3 as Fe. The results of the chemical analyses at three
different times during the run are shown in Table 90. The results of the
virus samplings are shown in Tables91 and 92. All effluent samples were
zero. The results of all the chlorination controls were again zero.
Samples of sludge were taken 1, 2, and 3 hours after the seeding and again
24 hours later. No polioviruses were found in any of these samples. The
1-and 2-hour samples were not run for coliphage but the 3-hour and 24-hour
samples were. One-half ml of the 3-hour sample plated directly as if it
were water yielded 9 plaques. The 24-hour sample was plated directly and
also blended, centrifuged in a clinical centrifuge, and the supernatant
plated; both yielded no coliphages. Since the pipe and tap from which
the sludge samples were collected could not be sterilized, the integrity
of the samples are questionable. Thus, negative assays would be meaningful
but positive samples could hardly be conclusive.
Summary
The rather large-scale pilot-studies conducted herein demonstrated
that virus removals from secondary effluents by alum coagulation-
sedimentation and coagulation-sedimentation-filtration processes are
essentially as described in "the literature using smaller scale processes.
Removal of bacterial virus as high as 99.845 percent for coagulation-
sedimentation and 99.985 percent for coagulation-sedimentation-filtration
processes were observed at an A1:P ratio of 7:1.
A marked decrease in virus removals was observed at a lower alum dose.
At an A1:P ratio of 0.44:1, removals of only 46 percent of f2 coliphage
and 63 percent of poliovirus by the coagulation-sedimentation process per ,
se_ were observed.
High-pH treatment of secondary effluents achieved very high degrees of
virus removal.
Bacteriological tests (total plate count procedure) of the recarbona-
tion process (used for pH neutralization) indicate that some bacteria
which gain entry to this process can reproduce in the recarbonation basin.
However, no viruses were found in the recarbonated effluents.
Since the viruses that were added in these tests were probably not
imbedded in particulates that could protect them from the adverse high-pH
environment, these results must be interpreted with caution.
236
-------
TABLE 90. CHEMICAL-PHYSICAL ANALYTICAL RESULTS, VIRUS RUN NUMBER FOUR
TIME
3:00
AM
INFLUENT1 EFFLUENT2
Temp °C
pH 3, units
Alk. as
CaC03
Cond. umhos
Hardness
as CaC03
Color
Turb.
COD
TOC •
sol
NH3-N
Org. N
N02+NQrN
SS
Total P
18
7.0
135
864
156
45.52
3.8/4.6
34/35
9/11.4
c
0.07 <0.1
3.35/2.3
9.5/9.7
2/10 . . '
10.3/10.3
19
11.0-6
400
1190
389
15/14
1.5/3.
21/27
9/9.8
0.15/0
1.67/1
11.5/8
1/7 ,
0.5/0.
7:
INFLUENT1
18
.0 7.2
135
828
133
35
2 4.0
29
n
.2 0.14
.9 2.70
.2 8.5
4
2 10.3
00 AM
EFFLUENT2
18
11.0-6
435
1260
403
15
1.5
17 -
6
0.13
1.61
10.5
0
0.4
INFLUENT1
18
.5 7.1
135
824
151
35
4.5
21
11
0.04
2.80
8.5
. 4
10.3
11:00 AM
EFFLUENT2
18
11.3-6.1
450
1270
415
15
2.0
8
0
0.11
1.99
10.0
8
0.5
1
2
3
Influent to Densator
Effluent from Recarbonation
Effluent pH value before and after recarbonation
.Denominator values are NERC analytical results on shipped samples.
Concentration in mg/1 unless otherwise noted.
237
-------
TABLE 91. COLIPHAGE RESULTS, VIRUS RUN NUMBER FOUR
SAMPLE
INFLUENT
pfu/ml
EFFLUENT
pfu/ml
BACKGROUND
28
,n4
Seeding 2 x
(4 nrin., 6 sees. 4000 mis mixed viruses)
All effluent samples
(14 samples collected over 8-hour period)
Stock titer:
9.3 x 107 Pfu/ml
TABLE 92. POLIOVIRUS RESULTS, VIRUS RUN NUMBER FOUR
SAMPLE
INFLUENT
pfu/ml
EFFLUENT
pfu/ml
BACKGROUND
0
Seeding 4 x 104
(4 min., 6 sees. 4000 mis mixed viruses)
All effluent samples
(14 samples over 8-hour period)
Stock titer:
5.7 x 107 pfu/ml
238
-------
TECHNICAL REPORT DATA
(Please read InOnictions on the reverse before completing)
1. REPORT NO.
EPA-600/2-80-149
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
THE REMOVAL OF METALS AND VIRUSES IN ADVANCED
WASTEWATER TREATMENT SEQUENCES
5. REPORT DATE
August 1980 (Issuing Date)
6. PERFORMING-ORGANIZATION CODE
7. AUTHOR(S)
Steven E. Esmond, Albert C. Petrasek, Jr., Harold
Wolf, D., Craig Andrews
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
1BC611
Texas A&M University
Dallas, Texas 75201
11. CONTRACT/GRANT NO.
S-801026
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory—Cin.,OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final June 1972-Dec. 1973
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer: Irwin J. Kugelman (513) 684-7633
16. ABSTRACT
An extensive study of metals and virus removals by advanced wastewater treat-
ment processes was conducted in Dallas, Texas from June 1972 through December 1973.
Processes applied to a biologically nitrified effluent included chemical coagulation
with alum and/or lime, high-pH lime treatment with and without recarbonation, filtra-
tion through multi- and dual-media filters, and carbon adsorption. The high-pH lime
treatment with recarbonation provided a most effective treatment for both metals
removals and disinfection. Boron surfaced as a material that may require other means
of control. Although high-pH, lime treatment was indicated to be extremely effective
for virus removal (or inactivation), metals removal were not of the same order of
magnitude. Thus, efforts to control metals at points of discharge are strongly
supported. \ The removal of some metals by biological processes appeared to be in-
fluenced by' their concentration. Median values were observed to be more indicative
of the plant processes than mean values. Coliphages were observed to provide essen-
tially the same virus removals values as polioviruses. The suggestion is made that
all wastes should be subject to biological treatment, and if such treatment is
found ineffective, then other means of control are warranted.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Water Reclamation
Water Resources, Heavy Metals
Advanced Waste Treatment
Disinfection
Metals Removal
Virus Removal
13B
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)'
Unclassified
21. NO. OF PAGES
259
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
239
•fr U.S. GOVERNMENT PRINTING OFFICE: 1980--657-165/0148
-------
-------