EPA/600/2-78/201
tvaluation of
Physical Chemical
Treatment at
Rosemount
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development. U S Environmental
Protection Agency have been grouped into nine series These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields
The nine series are
1 Environmental Health Effects Research
2 Environmental Protection Technology
3 Ecological Research
4 Environmental Monitoring
5 Socioeconomic Environmental Studies
6 Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8 Special1 Reports
9 Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161
-------
EPA-600/2-78-201
December 1978
EVALUATION OF PHYSICAL CHEMICAL TREATMENT AT ROSEMOUNT
AGENCY
by 1445 ROSS AVENUE
DALIAS, TEXAS 7520?
R. C. Polta, R. W. DeFore and W. K. Johnson '
Metropolitan Waste Control Commission
St. Paul, Minnesota 55101
Grant No. S802666
Project Officer
S. A. Hannah
Wastewater Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
UNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publi-
cation. Approval does not signify that the contents necessarily reflect
the views and policies of the U.S. Environmental Protection Agency, nor
does mention of trade names or commercial products constitute endorsement
or recommendation for use.
11
-------
FOREWORD
The Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health
and welfare of the American people. Noxious air, foul water, and spoiled
land are tragic testimony to the deterioration of our natural environment.
The complexity of that environment and the interplay between its components
require a concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem solu-
tion and it involves defining the problem, measuring its impact, and search-
ing for solutions. The Municipal Environmental Research Laboratory develops
new and improved technology and systems for the prevention, treatment, and
management of wastewater and solid and hazardous waste pollutant discharges
from municipal and community sources, for the preservation and treatment of
public drinking water supplies, and to minimize the adverse economic, social,
health, and aesthetic effects of pollution. This publication is one of the
products of that research; a most vital communications link between the
researcher and the user community.
A physical-chemical treatment plant designed to produce a very high
quality effluent was placed in operation in 1973 in Rosemount, Minnesota,
to treat the municipal wastewater. The U.S. Environmental Protection
Agency and the Metropolitan Sewer Board agreed to jointly fund an evaluation
of the new Rosemount facility. This publication reports the performance
and cost data for this new technology.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
-------
ABSTRACT
This research program was conducted to demonstrate the effectiveness of
physical chemical treatment of raw municipal wastewater at the Rosemount
Advanced Wastewater Treatment Plant and more specifically to evaluate: the
performance of the system as a whole, the performance of the individual treat-
ment processes, and the costs associated with operation and maintenance. During
the two year demonstration period the facility treated an average flow of
approximately 0.25 mgd by means of chemical clarification, filtration, carbon
adsorption and ion exchange; both the activated carbon and ion exchange media
were regenerated onsite.
The performance data are summarized according to the five process flow
schemes used in addition to discussion of the individual treatment processes.
Cost data are presented for each treatment process. These data are used to
construct an estimate of operating and maintenance costs for a 10 mgd facility.
The operating and maintenance problems encountered during approximately
four years of operation are described along with their solutions when they
were determined. Design recommendations are presented.
This report was submitted in fulfillment of Grant No. S802666 by the
Metropolitan Waste Control Commission under the partial sponsorship of the
U.S. Environmental Protection Agency. This report covers the period from
June 6, 1975 to August 13, 1977, and the work was completed as of January 12,
1978.
-------
CONTENTS
Foreword iii
Abstract iv
Figures viii
Tables x
Abbreviations xii
Conversion Factors xiii
I. INTRODUCTION 1
II. CONCLUSIONS 3
III. RECOMMENDATIONS 5
IV. FACILITY DESCRIPTION 7
A. GENERAL 7
B. TREATMENT PROCESSES 10
1. Clarification 10
2. Filtration 10
3. Carbon Adsorption 14
4. Ion Exchange 16
C. PUMPING 16
D. FLOW MEASUREMENT 17
E. CHEMICAL FEED SYSTEMS 17
1. Lime 17
2. Ferric Chloride 17
3. Polymer & Nitrate 19
4. Sulfuric Acid 19
5. Chlorine 19
6. Caustic 19
F. ACTIVATED CARBON REGENERATION 19
G. ZEOLITE REGENERATION AND BRINE RECOVERY 21
1. General 21
2. Original System 23
3. Final System 24
H. PROCESS CONTROL 26
1. General 26
2. Clarifier 28
3. Filters 28
4. Carbon Regeneration 29
5. Zeolite Regeneration and Brine Recovery 30
V. ANALYTICAL PROGRAM 31
A. SAMPLING AND ANALYSIS SCHEDULE 31
1. Treatment Processes 31
2. Activated Carbon Regeneration 32
3. Zeolite Regeneration and Brine Recovery 36
B. PROCEDURES 37
C. CONTINUOUS MONITORING 37
-------
VI. STAFF 41
VII. COST ACCOUNTING 43
VIII. WASTEWATER CHARACTERISTICS 46
IX. PLANT OPERATION CHRONOLOGY 50
A. GENERAL 50
B. MODE I 50
C. MODE II 54
D. MODE III 60
E. MODE IV 60
F. MODE V 64
G. AMMONIA REMOVAL SYSTEM 64
X. OVERALL PLANT PERFORMANCE 70
A. PRIOR TO DEMONSTRATION PROJECT 70
B. DURING DEMONSTRATION PROJECT 72
1. General 72
2. Mode I 76
3. Mode II 76
4. Mode III 82
5. Mode IV 82
6. Mode V 87
XI. UNIT PROCESS PERFORMANCE 94
A. CLARIFICATION 94
1. Operating Procedures 94
a. Sludge Blanket Control 94
b. Mixer Speed 94
c. Sludge Wasting 95
2. Chemical Conditioning Alternatives 96
3. Removal Efficiencies 97
a. Overall 97
b. pH 10.5 100
c. pH 9.5 106
4. Mixing Modifications 110
5. Costs 110
a. Chemicals 110
b. Labor Ill
c. Supplies 112
d. Power 112
B. FILTRATION 112
1. Operating Procedures 112
2. Removal Efficiencies 113
3. Costs 114
C. CARBON ADSORPTION 115
1. Operating Procedures 115
a. Service Time 115
b. Backwash 115
c. H2S Control 117
2. Removal Efficiencies 119
a. General 119
b. Upflow 120
c. Downflow 122
3. Costs 126
4. Corrosion Problems 127
vi
-------
5.
ION
1.
2.
3.
4.
Carbon Losses
EXCHANGE AND
General . .
NH3-N Removal
a. Scheme
b. Scheme
c. Scheme
d. Scheme
e. Scheme
f. Scheme
g. Scheme
h. Scheme
i . Scheme
Summary of
Costs . .
1,
2,
3,
4,
5,
6,
7,
8
9
ZEOLITE REGENERATION
as
pH
pH
pH
pH
PH
pH
pH
a Functi
11
12.5 -
13 -
13 -
12 -
12 -
12 -
4
4
4
4
4
8
on
.6
.6
.6
.6
.6
.2
14
of
BV
BV
BV
BV
BV
BV
BV
Regeneration Conditions. .
- IN
- IN
- IN
- 0.
- 0.
- 0.
- 0.
Na
Na
Na
5N
5N
5N
5N
Na
Na
Na
Na
Performance
. . 130
. . 130
. . 130
. . 132
. . 132
. . 134
. . 136
. . 137
. . 139
. . 139
. . 141
. . 142
. . 142
. . 143
. . 143
E. CARBON REGENERATION 145
1. Efficiency 145
2. Losses 147
3. Cost 148
XII. COSTS 150
A. OBSERVED - 0.25 mgd 150
B. PREDICTED - 10 mgd 151
XIII. DESIGN RECOMMENDATIONS AND CONSIDERATIONS 156
A. GENERAL 156
B. SAMPLING AND MONITORING 156
C. CLARIFICATION 156
D. FILTRATION 157
E. CARBON ADSORPTION 157
F. CARBON REGENERATION 158
G. AMMONIA REMOVAL 158
H. VALVES 160
XIV. OTHER OBSERVATIONS 162
A. DATA CORRELATIONS 162
B. PERFORMANCE MONITORING 162
1. General 162
2. Temperature 162
3. pH 167
4. Conductivity and chloride 167
5. Total Organic Carbon 168
6. Turbidity 168
7. Ammonia and Phosphorus 170
References 171
VI 1
-------
FIGURES
Number
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
Regeneration
Regeneration
System
and
Rosemount AWTP
Plant Layout
Profile of Solids Contact Clarifier
Schematic of Second Stage Dual Media Filter
Underdrain Strainers
Schematic of Activated Carbon Column
Ferric Chloride Feed System
Schematic of Activated Carbon
Schematic of Original Zeolite
Brine Recovery System
Brine Pumping Schedule for Zeolite Regeneration -
Final Conditions
Main Process Control Panel
Typical Compartment and Control of Sample
Collection System
Typical Dry Weather Influent Hydrograph
Frequency Distributions for Raw Wastewater Parameters
Flow Scheme During Mode I Operation
Flow Scheme During Mode II Operation
Flow Scheme During Mode IV Operation
Flow Scheme During Mode V Operation
Frequency Distributions for Effluent Parameters
During Demonstration Project
Frequency Distributions for Effluent Parameters
During Mode I
Performance Characteristics of Zeolite Columns when
Regenerated at pH 11
Frequency Distributions for Effluent Parameters
During Mode II
Frequency Distributions for Effluent Parameters
During Mode III
Frequency Distributions for Effluent Parameters
During Mode IV
Frequency Distributions for Effluent Parameters
During Mode V
Effect of Sample Transport on Concentration of TOC
Weekly Average Suspended Solids Concentrations
Weekly Average COD Concentrations
Average of Hourly pH Values for Clarifier Reaction Zone
8
9
11
12
13
15
18
20
22
25
27
34
47
49
52
55
62
65
74
78
79
81
84
86
89
92
98
99
101
VTM
-------
FIGURES
(Continued)
Number
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
Rates
and 6
Ammonia
4 and 5
Breakthrough Curves for Zeolite Columns
- Regenerated at pH 11
Ammonia Breakthrough Curves for Zeolite Columns
5 and 6 - Regenerated at pH 12.5
Cracks on Outside Surface of Steam Stripping
Column
Teflon Lined Brine Transfer Pump
Relationship Between COD and TOC for Plant Effluent
103
Frequency Distributions for Suspended Solids Data -
6/7/75 through 12/5/75
Frequency Distributions for BOD Data -
6/7/75 through 12/5/75
Frequency Distributions for Total P Data -
6/7/75 through 12/5/75
Frequency Distributions for Suspended Solids Data - 107
12/6/75 through 6/6/77
Frequency Distributions for BOD Data -
12/6/75 through 6/6/77
Frequency Distributions for Total P Data -
12/6/75 through 6/6/77
Hydrogen Sulfide Concentration of GCC Sample
Summary of Soluble COD Application and Removal
Soluble COD Removal Across Carbon Columns 4, 5
During Upflow Operation
Soluble COD Removal Across Carbon Columns 4, 5 and 6 124
During Downflow Operation
Soluble COD Removal Across Carbon Columns 1, 2 and 3 125
During Downflow Operation
Corroded Valve Disc
Corroded Flowmeter
Corrosion in Carbon Column No. 4
104
105
108
109
118
121
123
128
129
131
133
135
159
161
169
IX
-------
TABLES
Number
1.
2.
3.
4.
5.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
Design Criteria
Filter Backwash Sequence
Sample Identification Scheme
Sample Location and Flow Signal
Routine Sampling and Analysis Schedule for
Treatment Processes
Sample Location - Zeolite Regeneration and
Brine Recovery
Analytical Schedule - Zeolite Regeneration and
Brine Recovery
Summary of Analytical Procedures
Monitoring Equipment
Continuous Monitoring Schedule
Permanent Staff Positions
Cost Centers
Chemical Inventory System
Raw Wastewater Characteristics
Clarifier and Carbon Column Status
Clarifier Operating Conditions - Mode I
Zeolite Regeneration Sequence - Mode I
Clarifier Operating Conditions - Mode II
Clarifier Chemical Feed - Mode II
Filter Backwash Frequency - Mode II
Clarifier Operating Conditions - Mode III
Clarifier Operating Conditions - Mode IV
Carbon Column Status - Mode IV
Clarifier Operating Conditions - Mode V
Carbon Column Status - Mode V
Service and Regeneration Conditions for
Ammonia Removal System
Plant Performance Data for Period Prior to
Demonstration Project
Discharge Permit Conditions
Summary of Analytical Data
Performance
Performance
Performance
Performance
Performance
Data
Data
Data
Data
Data
Summary -
Summary -
Summary -
Summary -
Summary -
Mode
Mode
Mode
Mode
Mode
I
II
III
IV
V
14
31
33
35
35
35
33
40
40
41
44
45
48
51
53
55
57
57
53
61
53
63
66
66
67
71
72
75
77
80
83
85
88
-------
TABLES
(Continued)
Number
35. Comparison of Population Means - Eff vs DMF2, GCC 91
36. Effect of Additional Slow Mixing on Solids and 95
Phosphorus Removals
37. Variability of Raw and Clar Parameters TOO
38. Clarifier Performance Characteristics at pH 10.5 102
39. Clarifier Performance Characteristics at pH 9.5 106
40. Range of Clarifier Effluent Parameters - no
5th through 95th Percentile
41. Clarification Chemical Costs m
42. Clarification Power Costs 112
43. First Stage Filter Performance Characteristics 113
44. Filtration Costs 114
45. Carbon Service Cycles 116
46. Summary of Removal Efficiencies Across Activated 120
Carbon Columns
47. Carbon Adsorption Costs 126
48. Valve Disc Analyses 127
49. Average Brine Conditions - Scheme 2 134
50. Average Brine Conditions - Scheme 3 136
51. Average Brine Conditions - Scheme 4 137
52. Zeolite Losses for pH 13 Regenerations 138
53. Average Brine Conditions - Scheme 6 140
54. Average Brine Conditions - Scheme 7 141
55. Ammonia Removal System Performance Summary 144
56. Ammonia Removal Costs 145
57. Activated Carbon Regeneration Data Summary 146
58. Carbon Regeneration Costs 149
59. Operating and Maintenance Cost Summary 150
60. Labor Requirement for Cost Centers 152
61. Estimated Full Time Staff for 10 mgd Facility 153
62. Estimated Chemical Costs - 10 mgd Facility 154
63. Cost Summary - 10 mgd Facility 155
64. Summary of Linear Regression Calculations 163
XI
-------
LIST OF ABBREVIATIONS
cfm - cubic feet per minute
fps - feet per second
ft - feet
gm/cc - grams per cubic centimeter
gm. eq. wt - gram equivalent weight
gpd/sq ft - gallons per day per square foot
gph - gallons per hour
gpm - gallons per minute
hp - horsepower
hr - hour
in - inch
KWH - Kilowatt hour
Ib - pound
Ibs/cut ft - pounds per cubic foot - Ibs/cu ft
ma - milliampere, direct current
me/gm - gram milliequivalent weights per gram of material
mgd - million gallons per day
mil gal - million gallons
min - minute
mm - millimeter
N - gram milliequivalent weights per liter of solution
psf - pounds per square foot
psi - pounds per square inch
sq ft - square foot
xii
-------
CONVERSION FACTORS
English Unit
Cubic feet per minute
Feet
Gallons
Gallons per day per square foot
Horse power
Kilowatt - hours
Million gallons
Pound
Pounds per cubic foot
Pounds per- square foot
Pounds per square inch
Square foot
multiply by
0.472
0.3048
3.785
0.352
1.014
860.5
3785
454
16.02
4.88
0.0703
0.0929
to yield Metric Unit
Liters per second
Meters
Liters
Liters per day per square meter
Horse power
Kilogram - calories
Cubic meters
Grams
Kilograms per cubic meter
Kilograms per square meter
Kilograms per square centimeter
Square meters
XTM
-------
I. INTRODUCTION
The Metropolitan Sewer Board, precursor of the current Metropolitan
Waste Control Commission (MWCC) was formed in 1969. This agency is responsible
for the operation of water pollution control facilities in the seven county
metropolitan area in and around Minneapolis-St. Paul. In 1970 the agency
acquired 33 wastewater treatment facilities along with 320 miles of inter-
ceptor sewer. Only four of the 33 plants produced treated effluents in
compliance with applicable water quality standards at that time. Fifteen of
the original plants have been closed to date and others have been upgraded.
Three new treatment facilities including the Rosemount Advanced Wastewater
Treatment Plant have been put into service since 1972.
Tha City of Rosemount is located 15 miles south of St. Paul and has a
current population of approximately 4500. Prior to December 1973, the
wastewater collected was treated at a high rate trickling filter plant and
subsequently discharged to an effluent seepage pond. Both treatment plant
and seepage pond were operating at their hydraulic capacity in 1970. Because
there was no acceptable outlet for the plant effluent at the existing site,
expansion of the existing facilities was not feasible. The decision was made
to discharge the effluent from the new plant to the Mississippi River through
Spring Lake. Spring Lake consists of the northwestern part of the impounded
reach of the Mississippi River immediately upstream of Lock and Dam No. 2,
located at Hastings, Minnesota. Proposals had been advanced to isolate the
lake from the main river channel and develop its potential for water-based
recreational use. The evaluation of the water quality standards and their
applicability to receiving waters of limited dilution capacity led to the
adoption of the effluent quality limits presented in TABLE 1 for design
criteria.
TABLE 1. DESIGN CRITERIA
BOD - 10 mg/1
Suspended solids - 10 mg/1
Total phosphorus - 1 mg/1
Ammonia nitrogen - 1 mg/1
Total coliform organisms - 100 MPN/100 ml
-------
Several treatment alternatives were evaluated for a design flow of
0.6 MGD. Estimated capital and operating costs indicated that a physical-
chemical treatment facility would be most cost efficient. (1) It was also
recognized that the use of these new treatment processes would serve to fully
evaluate the feasibility, economics and application of the individual unit
processes for future treatment plant construction elsewhere.
The construction contract for the Rosemount Advanced Wastewater Treatment
Plant (AWTP) was awarded December 15, 1971. In June of 1973 the U.S.
Environmental Protection Agency and the Metropolitan Sewer Board agreed to
jointly fund an evaluation and demonstration of the Rosemount facility. The
specific objectives were to evaluate:
(a) performance of the system as a whole
(b) performance of individual unit operations and unit processes
(c) costs associated with operation and maintenance.
The major portion of the facility was put into service November 19, 1973.
The zeolite regeneration and brine recovery system was not operational at
that time however. The project agreement stipulated that the facility must
be formally accepted by MWCC prior to the initiation of the demonstration
program. The formal acceptance tests were concluded May 15, 1975 and the
demonstration project was initiated June 6, 1975.
The purpose of this report is to summarize the information obtained by
operating the Rosemount AWTP for the two year period June 6, 1975 through
June 5, 1977. Additional data from the ammonia removal system are included
for the period ending August 12, 1977 because they are significant in terms
of evaluating that system.
-------
II. SUMMARY AND CONCLUSIONS
A two year performance and cost evaluation program was conducted at the
Rosemount AWTP. This physical chemical facility treated an average domestic
wastewater flow of approximately 0.25 mgd by means of chemical clarification,
filtration, carbon adsorption and ion exchange. Each process was evaluated
in terms of both treatment characteristics and actual operating and main-
tenance costs. In addition total plant performance was evaluated during each
of five flow schemes or modes of operation.
Although the average raw wastewater characteristics changed considerably
during the treatment modes, the treatment efficiency in terms of percent
removal of suspended solids, BOD and total phosphorus did not vary signif-
cantly and for the two year period removals averaged 99%, 93% and 94% re-
spectively. The ammonia removal efficiency did vary substantially throughout
the project as a result of modifications to the zeolite regeneration system.
The process modifications were made to determine the effect of several parameters
on the useful ammonia exchange capacity and associated operating costs.
The clarification process was aided by additions of lime and ferric
chloride to provide for coagulation of finely divided solids and precipitation
of soluble phosphorus. For a six month period the lime addition was regulated
to maintain a pH of 10.5 in the clarifier. The lime dose was lowered during
the following 18 month period to maintain a pH of 9.5. The ferric chloride
dose was adjusted to maintain an effluent total phosphorus^1 mg/1. During
the high pH operation the removal efficiencies averaged 94%, 81% and 95%
for suspended solids, BOD and total phosphorus respectively. During this period
the quality of the clarifier effluent approached that of a conventional
secondary treatment facility. The performance degraded measureably when the
pH was reduced to 9.5 and the respective removal efficiencies dropped to 88%,
75% and 88%. During both periods the clarifier exhibited a considerable
capacity to accept shock loadings both in terms of flow and concentration and
as a result significantly dampened possible suspended solids and BOD loading
fluctuations on the downstream treatment units.
The dual media filters were located directly downstream of the clarifier
(first stage) and carbon contractors (second stage). The suspended solids
loading of the first stage filters averaged 0.47 psf per day and removals
averaged 47%. The suspended solids loading of the second stage filters was
considerably lower and removals averaged 30%.
During that period when the carbon columns were operated in the downflow
mode the overall plant performance did not suffer when the second stage filters
were removed from the flow scheme. However several attempts to operate the
plant without the first stage filters failed because of rapid fouling of the
-------
second stage filters. Cleansing techniques using salt and/or chlorine along
with normal air scour and backwash were developed.
The removal of organics across the carbon columns averaged 55% in terms
of soluble COD. The removals were found to decrease as the cumulative loading
(Ib COD removed per Ib carbon) increased. For the period during which the
carbon columns were operated in the upflow mode the loading averaged 0.20 Ib
soluble COD removed per Ib carbon which is considerably lower than several
investigators have reported. The useful carbon loading rate increased to
0.4 Ib per Ib when the columns were operated in the downflow mode.
Daily air scour and backwash did not eliminate or control the production
of hydrogen sulfide in the carbon beds and both odor and corrosion problems
were encountered. It was found that hydrogen sulfide generation could be
controlled when sodium nitrate was added to the column influent stream at a
concentration of 5 rng/1 N03-N and the columns were air scoured and backwashed
daily.
Excessive carbon losses were encountered and appeared to be primarily
related to the lack of screens at the backwash water discharge points and
inadequate screening at the inlet to the screw conveyer serving the carbon
regeneration furnace.
Carbon regenerations were conducted onsite with a four hearth gas fired
furnace. Although carbon losses during regeneration were not documented it
is estimated that they were minor compared to losses described above.
The ion exchange system used to remove NH--N was evaluated for nine
operating schemes. It was found that the regenerant volume and pH were the
most significant variables. The practical NH3~N exchange capacity was found
to fall in the range of 0.12 to 0.37 me/gm depending on regeneration pro-
cedure.
The actual unit cost of treatment was extremely high, $5.11 per 1000 gal,
because the plant capacity and flow were low enough to preclude any economy
of scale yet the plant complexity required full staffing. If the plant had
operated at its design capacity during the two year period, the unit cost of
treatment would have been reduced by approximately 50%. Although treatment
facilities of this type may not be cost effective for small flows, operating
costs for larger facilities are estimated to be considerably lower. The
operating and maintenance cost for a 10 mgd facility similar to the Rosemount
facility was estimated to fall in the range of $0.56 to $0.83 per 1000 gal.
-------
III. RECOMMENDATIONS
It is important that the coagulants added in the chemical clarification
process be completely dissolved and rapidly mixed with the incoming flow.
Subsequent to this, gentle mixing is required to ensure a reasonable degree of
flocculation which will, in turn, enhance the removals obtained by sedimenta-
tion. Four years of experience at the Rosemount AWTP indicate that mechanical
mixers are extremely effective for removing rags and stringy debris from the
wastewater flow. It is essential that nonfouling or self cleaning mixing
devices be developed so that maintenance effort can be reduced and mixing
conditions maintained near optimum.
Although the performance characteristics of the standard automated
chemistry modules used to monitor phosphorus concentrations were acceptable,
the equipment was costly to operate and maintain. Currently available phos-
phorus monitors should be evaluated on raw and clarified wastewater because
their successful use in control schemes could yield significant saving in
terms of coagulant chemicals.
Based on the results at Rosemount and more recent experience at other
facilities it appears that there is a need to develop and/or demonstrate a
reliable method of protecting mild steel if that material is to be used in
contact with activated carbon.
The addition of sodium nitrate can be used to limit the generation of
hydrogen sulfide in carbon contactors however the cost of this method of
control is significant. Some considerable effort to develop and demonstrate
more cost effective means of hydrogen sulfide control is thus justified by
the potential savings.
Although the measured organic loading of the carbon columns was relatively
constant on a day to day basis the removal efficiency demonstrated significant
fluctuations. Work should be continued at the Rosemount facility to determine
the cause of these variations in treatment efficiency and thus provide a
better understanding of the limitations of activated carbon in treating
municipal wastewaters.
Because of the significant costs (capital, operating, and maintenance)
involved in carbon transfer and regeneration it would be desirable to develop
a method of in situ regeneration to eliminate or at least prolong the interval
between thermal regenerations. It was observed by accident at Rosemount that
a significant mass of soluble organic material is released from a carbon bed
when it is exposed to pH values in the range of 9.5 to 10.5. Leaching with
high pH solutions should be investigated.
-------
The data presented in this report and those obtained by other invest-
tigators demonstrate that the working ammonia exchange capacity of the
zeolite clinoptilolite is directly related to the regeneration conditions as
well as the chemical characteristics of the flow during the service cycle. It
is important that the relationships that do exist among the several parameters
be better defined so as to allow near optimum design for future treatment
facilities.
During the four plus years of operation many engineers have called and
visited the Rosemount AWTP in search of information on the performance char-
acteristics and operating problems associated with the unit processes used.
Because the operating problems are often related to specific design deficiencies
and performance characteristics are affected by many parameters it is not easy
to generalize on the applicability of physical chemical treatment or the
associated costs. Although several P-C facilities are now operating in the
United States there is a dearth of operating information in the technical
literature. It is thus recommended that periodic seminars be held to allow
design engineers and representatives from all operating P-C facilities to
meet and discuss the practical problems encountered by operating personnel.
It is anticipated that both groups would benefit from a free and informal
exchange of information and possibly lead to more efficient and cost effective
treatment.
-------
IV. FACILITY DESCRIPTION
A. GENERAL
The treatment plant is located in the City of Rosemount, Minnesota,
approximately 6 miles from the center of the sewered area. The raw sewage
flow is transported to the plant through an intercepting sewer which ranges
from 27 to 48 inches in diameter and is approximately 5.5 miles long.
The treatment process units are totally enclosed in a 15,000 sq ft
steel building, Figure 1. After metering and coarse screening the flow is
pumped to one or two parallel process trains each with a design capacity of
300,000 gpd. The flow from the two trains is mixed prior to chlorination and
discharged to the Mississippi River (Spring Lake) through a force main
approximately one and one half miles long.
The layout of the treatment plant is presented in Figure 2. Each process
train consists of a clarifier, two first stage dual media filters, three
granular activated carbon columns, two second stage dual media filters, and
three zeolite columns. Because of its experimental nature the facility design
incorporated flexibility within each process train. One or both sets of
filters can be bypassed if desired and the carbon columns can be operated in
several modes. In addition the process trains can be coupled in filtered
water storage tank No. 2.
Both of the process trains are served by common facilities for sludge
storage and hauling, chemical storage, activated carbon regeneration, zeolite
regeneration and brine recovery. The major portion of the return flows
generated in the plant by backwashing and rinsing are piped to the backwash
pit which has a capacity of approximately 15,700 gal and is located between
the train 1 and 2 filters below grade. The plant and effluent chlorinators
along with two air blowers are located in a separate building approximately
25 feet north of the main building. The boiler, which supplies heat and
process steam, and a natural gas powered standby generator are located in a
masonry structure directly adjacent to the main building. The maintenance
shop is located directly east of the cTarifiers.
The laboratory and staff facilities are located in the single story
masonry structure south of the main building. Office space for the plant
superintendent, maintenance staff and staff associated with the demonstration
project is provided in three trailers located near the main building.
-------
c:
CL.
4->
C
-a
o
-------
eg
o o o
m 10
O O O
ID
h- o:
i-i m
O i
I
<: <
I CD
> UJ
o
>- o
LU (
: h- c£ o LU i
O I
I _j :
: =t o >
_ i 3 s: z i o i
O O UJ I ^ ' CO ^
UJ _l 31 et O < O
M D3 CJ S! CQ I CJ
oo oo o o o
u_ o s: o
HH I <=C => I CC
C£ i*v j oi
e£ GO O O t/) O
_j en t t_> o; y
O UJ LO UJ (/)
\ ^L\ 00
h- _J C£ O 1 C£ ^
O i UJ CQ ' UJ El
cCLi_| C£Li_( rDCO
:^<:^o<=t:3oo
JO11 ' c_) |
J C_J Q Q fV f) (i ej;
: uj uj » Ql
«_ll ^II _IO
:^jci;_jcc<_iO-J
:o^3>i CE: n> i
-------
B. TREATMENT PROCESSES
1. Clarification
A suspended solids contact clarifier is the first treatment unit in each
of the process trains. The units were supplied by the Graver Water Conditioning
Company and are identified by the trademark Reactivator. The clarifiers are
25 ft in diameter and have a side water depth of 12.67 ft, Figure 3. The
clarifiers are constructed of steel and are mounted on concrete pads to
elevate the units above grade. The normal water surface elevation is approxi-
mately 22 ft above the floor of the plant. The maximum settling area is
424 sq ft. This yields a surface settling rate of 0.5 gpm/sq ft at the
design flow of 0.3 mgd per process train.
A variable speed, turbine type impeller is used to mix the treatment
chemicals with the sewage flow. The impeller also effects some sludge re-
cycle up through the draft tube. Each unit is equipped with a skimmer and
grease box as well as rotating sludge plow. Sludge is removed from the
clarifier on a flow proportional basis and discharged to a holding tank.
The clarifier overflow is collected along a peripheral launder and is
discharged to a baffle chamber where sulfuric acid is added for pH control
as required. The neutralized flow is, in turn, discharged to the filter
influent pipes through two triangular weirs.
2. Filtration
Both the first and second stage filters are of the Monoscour design as
supplied by Graver Water Conditioning Company. Each filter vessel is 8 ft
in diameter, Figure 4. The hydraulic loading rate at design flow is 2 gpm/sq
ft. The filter media consist of approximately 24 in. of anthracite coal with
an effective size of 1.1 mm and 12 in. of silica sand with an effective size
of 0.45 mm. The filter media are supported on the underdrain structure with-
out the aid of any gravel subfill. The underdrain system consists of a steel
false bottom fitted with 126 molded plastic strainers. Each of the strainers
is covered with a plastic funnel, Figure 5.
A backwash water storage compartment is located above each filter and is
integral with the filter vessel. Each of the storage compartments has a
capacity of approximately 3000 gal. The second stage filter effluent is used
for filter backwash. The backwash compartments of the first stage filters
are connected to the compartments of the second stage filters and are filled
by gravity flow. With the exception of the flow through the backwash com-
partment in the second stage filters as illustrated in Figure 4 the first and
second stage filters are identical.
During normal operation the filter influent enters through valve A, flows
around the baffle plate, through the media and underdrains, through valve W,
up through the backwash water storage compartment and over the discharge weir.
The backwash procedure is initiated manually although a timer is available
for automatic initiation. The backwash sequence is automatically controlled,
10
-------
rO
O
o
o
O
O
-------
EFFLUENT
BACKWASH WATER
STORAGE
FEED
WASTE
xj-
COAL
SAND
44444444
X
o^
WASTE
AIR
Figure 4. Schematic of second stage dual media filter.
12
-------
re
S_
ro
S-
TD
LO
-------
once initiated, by a multicam timer which energizes the appropriate pneumatic
valve operators, TABLE 2.
TABLE 2. FILTER BACKHASH SEQUENCE
Step Valves Open Time-min
Draindown U 5
Air scour M,C 5
Settle bed C 6
Backwash W,C 5
Settle bed - 3.5
Service A,W Variable
The air scour rate is approximately 5 cfm/sq ft and the backwash rate
is approximately 16 gpm/sq ft. The filter controls are interlocked to allow
only one filter backwash to be in progress per train at any time.
3. Carbon Adsorption
The carbon columns are steel vessels eight ft in diameter and approxi-
mately 27 ft high - Figure 6. A nominal depth of 12 ft of granular activated
carbon is maintained in each column. The carbon, Westvaco NuChar WV-G,
12 x 40 mesh, is supported on an underdrain system identical to that used in
the filters. The system was designed to allow the utilization of several
operating modes. Any two of the three columns, per train, may be operated
in the following modes: series upflow, series downflow, parallel downflow
and primary upflow secondary downflow. In the upflow mode the feed enters
the upflow header and is distributed to eight laterals. The laterals are
constructed of 3 in. stainless steel pipe with 3/8 in. diameter holes 2 1/2 in.
on center along the bottom. At design flow the average carbon contact time,
based on empty bed volume, falls between 30 and 40 min depending on the volume
of plant recycle flow.
The carbon beds can be backwashed to remove debris and limit the accumula-
tion of biomass. The backwash procedure is initiated manually. Once started
however the following sequence is completed automatically: drain to air scour
level, stop drain and vent, air scour, settle bed, and backwash. The air scour
rate is fixed at approximately 5 cfm/sq ft however the backwash rate and the
times for each step in the sequence are controlled by the plant operator.
When the desired organic removal cannot be maintained across the carbon
columns, one of the columns is taken out of service and replaced with a column
14
-------
INFLUENT
DOWNFLOW MODE
11
EFFLUENT
UP FLOW MODE
UP FLOW
LATERAL-
ACTIVATED CARBON
UPFLOW HEADER
EFFLUENT DOWNFLOW
INFLUENT BACKWASH
CARBON
TRANSFER
Fiqure 6. Schematic of activated carbon column.
15
-------
containing regenerated carbon. The spent carbon is transported to the regen-
eration facility by an air aided hydraulic transfer. The column is put in the
backwash mode to partially fluidize the bed and air is supplied at the top of
the column at 15 to 20 psi as the transfer valve is opened. The carbon is
forced through the 3 in stainless steel transfer line and is discharged to the
spent carbon holding tank.
4. Ion Exchange
The ion exchange columns are cylindrical steel pressure vessels 8 ft in
diameter and approximately 15 ft high. Each column contains a 6 ft depth of
the natural zeolite clinoptilolite (20 x 50 mesh) supported on gravel subfill.
The underdrain structure consists of a central collection hub and eight pipe
laterals with 1 in. holes cut a 45° on both sides of the bottom line. One
or two, of the column per train, can be maintained in the service cycle
but only in the series downflow mode. Zeolite regeneration is conducted in
the upflow mode.
C. PUMPING
The raw sewage enters the plant 17 ft below grade, passes through a
Parshall flume, mechanically cleaned bar screen with 1 in. clear spacing
and discharges to the wet well approximately 25 ft below grade. All backwash
water and other wastewaters generated in the plant are discharged to the wet
well via the backwash pit. The raw sewage and recycle flows are pumped to
one or both clarifiers. The raw pumping station consists of 3 variable
speed pumps each with a discharge capacity of 575 gpm against a total dynamic
head of 48 ft. The pump speed and number of pumps in service are automatically
controlled based on the water elevation in the wet well.
The carbon column pumps, two per train, are located on the east side of
filtered water storage tank No. 1, Figure 2. These pumps draw from the storage
tank and discharge to the carbon column inlet manifold. Each pump has a dis-
charge capacity of 250 gpm against a total dynamic head of 55 ft. The carbon
column pumps operate a constant speed. A flow control valve, located down-
stream of the carbon columns, is driven by a level controller in filtered
water storage tank No. 1.
The effluent pumps, two per train, draw from filtered water storage tank
No. 2 and discharge to the inlet of the zeolite columns. These pumps operate
at constant speed and have a capacity of 250 gpm each against a total dynamic
head of 85 ft. The rate of flow to the zeolite column is controlled by a
float operated recycle valve which maintains a fixed level in filtered water
storage tank No. 2.
Water for backwash of the carbon beds and rinse of the zeolite beds is
pumped from filtered water storage tank No. 2 with one of the two backwash
pumps. Each of the constant speed pumps has a capacity of 750 gpm against
a total dynamic head of 51 ft. A flow control loop consisting of a control
valve, two mode controller and a flow meter allows the operator to set point
16
-------
and control the backwash rates.
D. FLOW MEASUREMENT
The raw sewage flow is metered with a six in. Parshall flume equipped
with a bubbler. The pneumatic signal is used to operate a local readout and
is converted to a current signal (4-20ma) for transmission to the plant control
room. Here the flow signal is integrated, totalized and recorded.
The process flows are metered at three points in each train; the discharge
from the clarifier, pumped discharge to the carbon columns and the pumped
effluent flow. After acid addition and mixing in the baffle chamber the
clarifier discharge flows through a weir box fitted with two triangular weirs.
The head on the weirs is determined with a bubbler. A local readout is pro-
vided for the pneumatic signal and the output from a pressure to current
transducer (4-20ma) is conducted to the plant control room. The flow signals
from both clarifiers are integrated, totalized and recorded. Since all back-
wash and wastewaters generated in the plant are returned to the plant wet
well the recycle rate within the plant is equal to the difference of the raw
and clarifier flows for a given time period.
The flow metering system for the discharge of both the carbon column
and effluent pumps consists of; flow tube, differential pressure transmitter,
integrator and totalizer.
E. CHEMICAL FEED SYSTEMS
1. Lime
Lime slurry is used to elevate the pH in the clarifier as required. The
lime delivery system consists of a lime storage silo with a capacity of
approximately 30 tons hydrated lime, a slurry mixing tank with a capacity of
1650 gal, slurry transfer pump, a slurry storage tank with 8000 gal capacity
and four, variable stroke, diaphragm chemical feed pumps. The lime slurry is
made up in the range of 2.5 to 5% on a weight basis. The chemical feed pumps
have a capacity of 5 gpm each and are controlled to maintain a flow proportional
lime feed rate.
2. Ferric Chloride
The ferric chloride feed system is illustrated in Figure 7. Liquid
ferric chloride (approximately 40% FeC^) is delivered to the plant and
stored in a 6500 gal fiberglass tank. The solution is transferred to day
tanks, consisting of 55 gal polyethylene drums, on a daily basis. Four,
variable stroke, diaphragm chemical feed pumps are used to deliver the ferric
chloride to the clarifiers on a flow proportional basis. The pumps have a
capacity of 15 gph each. All of the chemical feed pumps used at the plant
are similar to those illustrated in Figure 7.
17
-------
CD
-(->
CO
OJ
-------
3. Polymer and Nitrate
An anionic polymer was used as a coagulant aid in the clarifier during
the early months of the evaluation. The polymer was delivered in liquid form
and diluted to a concentration of 0.25% or 0.75% in one of two, 300 gal,
fiberglass mixing tanks. Four, variable stroke, diaphragm chemical feed
pumps were used to feed the solution to the clarifier on a flow proportional
basis. The pumps have a capacity of 10 gph each. Polymer usage was terminatec
November 16, 1975.
In early January 1976 this system was modified to feed sodium nitrate
solution to the suction side of the carbon column pumps. Sodium nitrate is
delivered in 100 Ib bags and the solution is made in the fiberglass tanks.
4. Sulfuric Acid
Concentrated sulfuric acid is delivered to the plant site by tank truck
and stored in a 4400 gal mild steel tank. The acid is pumped directly from
the storage tank to the clarifier as required for pH control. Four, variable
stroke, chemical feed pumps are used to deliver the acid based on the output
of a pH controller. The capacity of each pump is 10 gph.
5. Chlorine
Chlorine is supplied to the plant in one ton cylinders. Two standard
vacuum operated chlorinators with manual rate control are used to deliver
chlorine solution to the effluent and at several points in the plant. The
plant effluent is served by a unit with a capacity of 100 Ib/day. The in-
plant chlorine is supplied by a unit with a capacity of 400 Ib/day.
6. Caustic
Sodium hydroxide solution, approximately 50% NaOH, is used to adjust the
pH of the brine used for zeolite regeneration. The solution is delivered to
the plant by tank truck and stored in a heated 800 gal mild steel tank. Caustic
is discharged to the brine storage tanks through manually operated valves.
F. ACTIVATED CARBON REGENERATION
The carbon regeneration system is illustrated in Figure 8. Spent carbon
is discharged to the dewatering tank where the moisture content is reduced to
approximately 50% by gravity drainage. A variable speed screw conveyor de-
livers the moist carbon to the 54 in diameter refractory lined furnace. The
four hearth furnace is equipped with a wet scrubber and afterburner to con-
trol stack emissions. Natural gas is used to fuel four burners - two on hearth
No. 2 and one each on hearth No. 3 and 4. Steam is injected into the furnace
on hearth No. 4 at a controlled rate during thermal regeneration.
The quenched carbon slurry is transported to a storage tank, similar
to the on-line carbon columns, with an educator. The regenerated carbon is
transferred from the storage tank to one of the on-line columns by the transfer
19
-------
Q
UiJ
1
c^
ct.
LU
r^
LU
W^ en
W LU
a:
LU
Z CD
0 ,
(S)
6
S-
Q)
C
OJ
£T.
O
JD
S_
re
u
a
QJ
U
ro
o
o
fO
E
O)
_sz
o
00
oo
O)
20
-------
technique previously described.
G. ZEOLITE REGENERATION AND BRINE RECOVERY
1. General
The zeolite regeneration system consists of facilities for contacting
the spent zeolite with a sodium chloride brine at a controlled rate and pH.
Brine recovery involves calcium and magnesium removal by soda ash and caustic
additions, steam stripping the volatile NH3 and condensing and collecting the
ammonia-water vapor mixture. These two operations combine to form, by far,
the most complex system in the plant in terms of operation and physical de-
scription. The system consists of the following major items, as illustrated
in Figure 9.
Brine storage tanks (2)
Heat exchangers (2)
Heat exchanger/cooler (1)
Stripping column (1)
Condenser (1)
Ammonia receiver (1)
Brine transfer pump (1)
Stripper product pump (1)
- below grade concrete construction,
fiberglass lined, equipped with 5 hp
turbine type agitator and steam heat-
ing coils, total volume 15,400 gal,
normal working volume of approximately
10,300 gal.
- plate and frame type, titanium plates,
293 sq ft heating surface, design at
150 psi and 250°F.
- plate and frame type, titanium plates,
65 sq ft heating surface, design at
150 psi and 150°F.
- fiberglass reinforced plastic construc-
tion, 3 ft diameter by 18 ft high, 12
trays design at 14 psi and 250°F.
- finned tube construction, 860 ft, 1 in.
diameter, two 5 hp fixed pitch fans,
design at 100 psi and 250°F.
- steel tank, 6 ft diameter by 8 ft long.
- transfer brine to zeolite columns and
stripping column, discharge 100 gpm
against 285 ft head, FEP teflon lined
ductile iron construction, 25 hp drive.
- transfer brine from bottom of stripper
to reclaimed brine storage, discharge
50 gpm against 165 ft head, FEP teflon
lined ductile iron construction, 7.5 hp
drive.
21
-------
J.N3dS
O
LJ C£
_) UJ
O CO
0 Z
tj UJ
a
oc: z:
< 8
1 AMMONIA 1
1 SOLUTION
OJ
+->
oo
>>
oo
2
OJ
>
o
u
O)
O)
-a
c
fO
o
(O
s_
OJ
c
(1)
Cn
OJ
OJ
O
01
N
S-
o
o
o
)
na
CJ
CO
QJ
S-
cn
22
-------
Caustic storage (1) - steel tank 4 ft diameter by 9 ft high
with steam heating coil.
Acid storage (1) - steel tank 2.5 ft diameter by 5 ft high
Fiberglass pipe and rubber lined valves are used throughout this system.
The process control panel for the regeneration and recovery system is located
in the plant adjacent to the brine storage tanks. The outputs for the require
instrumentation - pH, temperature, pressure and conductivity - are located in
the panel along with the flow controllers and switches.
The operating procedures for this system changed considerably during the
two year demonstration period. The system layout and operating procedures for
the initial and final conditions will be presented in the following two sectio
All other modifications are described in later sections of this report.
2. Original System
Regeneration is initiated shortly after the zeolite column is taken out
of service and consists of the following automated sequence - Figure 9.
(a) Backwash, upflow, at a rate of 8 gpm/sq ft for 10 min.
(b) Settle bed for 5 min.
(c) Pump hot brine through heat exchangers No. 1 and 2, through the
cooler (protects zeolite against temperature over 80°F) and up
through zeolite bed at a rate of 100 gpm. The process water in
the column flows through valve Jl to filtered water storage tank
number 2. When the conductivity of this flow exceeds the set
point value the conductivity monitor output causes valve Jl to
close and J2 to open and the brine is discharged to the waste
brine storage tank.
(d) The brine transfer pump shuts down when the brine level in the
reclaimed brine tank reaches low level.
(e) Pressurize the zeolite column with air to discharge the brine to
the waste brine tank through appropriate valve and pipe arrange-
ment.
(f) Rinse the zeolite bed at a rate of 3 gpm/sq ft, downflow, for
45 min. Discharge the rinse water to filtered water storage tank
No. 1.
The spent brine was recovered as follows:
(g) Determine the calcium concentration of the brine.
(h) Adjust the pH to 11.5 with sodium hydroxide solution, add the
appropriate weight of sodium carbonate and mix. Allow the CaC03
and Mg(OH)2 formed to concentrate in the tank bottom and pump to
23
-------
the sludge storage tank.
(i) Pump the desludge brine through heat exchanger 1 or 2 (automatically
cycled each regeneration) and discharge to the stripping column at
a rate of 50 gpm.
(j) Inject steam into the stripping column at a rate of 1 Ib per gal
of brine.
(k) Discharge the volatile ammonia and water vapor from the top of the
stripper to the air cooled condenser. Collect the condensate in
a receiver at a concentration of approximately 1% NFU-N.
(1) Pump the brine from the bottom of the stripper through the haat
exchanger (No. 1 or 2) and discharge to the reclaimed brine tank.
(m) Adjust the pH to 11.0 with caustic or sulfuric acid as required
and adjust the sodium concentration to 1.0 N by the addition of
NaCl.
3. Final System
The final modification of the regeneration procedure, used during the
last 2 months of the study, eliminated the need for heated brine and heat
exchangers.
The brine pumping schedule is described in Figure 10 and as follows:
(a) Initial conditions - reclaimed brine tank at level (A)
(b) Pump (?) discharges brine to zeolite column, upflow, at a rate
of 100 gpm. The process water in the column exits through
valve Jl - Figure 9.
(c) When the reclaimed brine level drops to (F) , valve Jl and valve
J2 are manually activated and the spent brine discharges to the
waste brine tank.
(d) Pump (?) is automatically shut down when reclaimed brine level
reaches level Q§) - waste brine now at level(T).
(e) Pump (Z) discharges contents of waste brine tank to reclaimed
tank adjusting levels to (3) and (£) respectively - manually
controlled.
(f) The pH of the brine in the reclaimed tank is adjusted to 12.0
and solids allowed to settle for 35 min - manually controlled.
(g) Pump (7) discharges brine to zeolite column, upflow, at a rate
of 100 gpm.
24
-------
c/1
C
O
o
o
01
c
11)
Cr
CD
o
N
S-
o
O)
Z3
T3
-------
(h) Repeat steps d through g.
(i) Repeat steps d through g.
(j) Repeat step d.
(k) Pressurize the zeolite column with ajr to discharge brine to
waste tank - levels now at (B) and (K).
(1) Rinse the zeolite bed downflow at 100 gpm for 60 min or until
the manually determined pH of rinse water is^-8.5. Discharge
spent rinse water to headworks via backwash pit.
The spent brine was recovered as follows:
(m) Determine the calcium concentration of the waste brine.
(n) Adjust the pH to 12.0, and the appropriate weight of sodium
carbonate added and mixed. Allow the CaCOs and Mg(OH)2 formed to
concentrate in the tank bottom and pump to the sludge storage tank.
(o) Pump waste brine through heat exchanger number 1 and discharge to
the stripper at approximately 45 gpm.
(p) Inject steam into the stripper at a rate of 1 Ib per gal brine.
(q) Discharge the volatile ammonia and water vapor from the top of the
stripper to the air cooled condenser. Collect the condensate in a
receiver at a concentration of approximately 1%
(r) Pump the brine from the bottom of the stripper, through heat
exchanger number 1 and discharge to the reclaimed brine tank.
(s) Automatically terminate stripping operation when the waste brine
level falls to
(t) Adjust the pH of the brine to pH 12 and the sodium concentration
to 0.5 N by the addition of NaCl.
H. PROCESS CONTROL
1. General
Process control is concentrated at three locations in the plant. The
main control panel illustrated in Figure 'il is located in the control room
and consists of a graphic panel along with the analog controllers for flow
and pH and their associated indicators and recorders. Lights on the graphic
panel indicate the status of all treatment process elements - clarifier
filters, carbon columns and zeolite columns - along with the status of all
pumps and drives and the position of all on-off valves. All of the processes
26
-------
OJ
re
Q.
o
o
CO
CO
O)
O
o
Q.
QJ
S-
27
-------
with the exception of carbon regeneration and brine recovery are controlled
from this main panel. The carbon regeneration and brine recovery operations
are controlled at smaller local panels.
2. Clarifier
The flow rate through the clarifier is determined with two triangular
weirs located just downstream of the point of acid addition, Figure 3. The
flow signal, 4-20 ma, is used to control the addition of lime and ferric
chloride and the sludge blowdown rate. The stroke adjusting mechanisms of
the chemical feed pumps are operated directly from the flow signal with 4
to 20 ma yielding 0% to 100% maximum stroke respectively. The maximum stroke
or capacity of each pump is field adjustable to allow operation over a wide
range of concentrations.
The flow integrator outputs a 24 VDC pulse for every 10 gal discharged.
This pulse activates the flow totalizer and an automatic reset count controller.
When the operator entered count value is reached the counter resets and activates
the sludge blowdown cycle. The sludge blowdown is preceded by a water back-
flush to clear the blowdown line. Both the rinse and the blowdown times are
operator entered.
A sample from the draft tube is transported by gravity to a pH monitor
located outside the clarifier. The signal generated is transmitted to the
pH receiver and recorder on the main control panel. The plant operator
records the pH value every hour and adjusts the lime dosage as required. Al-
though the alkalinity of the raw wastewater is relatively constant the varying
rates of recycle affect the alkalinity of the clarifier influent and, in turn,
the pH at constant lime dose. Precise control is difficult because of the
long response time. For this reason the lime dose is not changed unless the
pH deviates by more than 0.5 from the set point value.
A sample of the acid neutralized clarifier effluent is transported by
gravity to a second pH monitor. The signal is transmitted to the pH receiver
and, in turn, to a recorder and PID controller. The controller outputs to
the stroke adjusting mechanism of the sulfuric acid feed pump based on the
deviation from the pH set point value.
3. Filters
Each filter is equipped with a differential pressure transmitter to monitor
the head loss across the filter media. The head loss is indicated on the main
control panel. The filters are generally backwashed once each day and backwash
is initiated on a time schedule rather than on a head loss basis. This pro-
cedure is followed to ensure that most of the recycle flow is returned to the
headworks during periods of low raw wastewater flow and thus equalize somewhat
the hydraulic loading on the plant. The head loss indicators are used to evalu-
ate the efficiency of the backwash procedure and to determine when special
cleansing procedures are required.
28
-------
4. Carbon Regeneration
The carbon regeneration furnace controls are mounted in a panel adjacent
to the furnace. The panel contains the following devices:
(a) on/off/fail switch/indicators for all drives
(b) small alarm status panel
(c) indicating temperature controllers for hearths 2, 3 and 4
and the afterburner
(d) furnace damper controllers
(e) furnace pressure indicator
(f) scrubber differential pressure indicator
(g) rabble arm speed control
Although the furnace was the smallest size available at the time of plant
construction its capacity is greater than required. For this reason, along
with the fact that the plant is being operated at approximately 40" of de-
sign capacity, the furnace is operated for 5-10 day periods only two or three
times per year. The furnace is brought up to operating temperature slowly
during a 24 to 30 hour period to minimize thermal stresses in the refractory
material. This is accomplished by manually adjusting the set points on the
temperature controllers.
Carbon starts to discharge from hearth number 4 approximately one hour
after initiation of carbon feed. Throughout the regeneration the hearth
temperatures are maintained in the following ranges:
Hearth °F
1 900 to 1050
2 1400 to 1500
3 1550 to 1650
4 1600 to 1700
Steam is added on hearth number 4. The steam addition rate is maintained
and regulated by a combination of a pressure regulating valve, rotometer and
control valve.
The rates of carbon feed and steam addition are determined by the operator
based on the apparent density of the regenerated carbon. The procedure and a
description of the apparatus required to determine apparently density are
found elsewhere.(2). The apparent density of the virgin carbon loaded in the
columns were 0.48 gm/cc. The plant operator samples the regenerated carbon,
prior to quench, each hour and determines the apparent density. If the value
falls outside the range 0.46 to 0.50 the operator adjusts the carbon feed
29
-------
rate and/or the steam feed rate accordingly. The apparent density is decreased
by reducing the carbon feed rate or increasing the steam rate. Composite samples
of both the regenerated and spent carbon are constructed from the hourly grab
samples. The composite samples are analyzed at a later date to determine the
efficiency of the regeneration.
5. Zeolite Regeneration and Brine Recovery
The zeolite beds are put into service at midnight and taken out of service
24 or 48 hours later depending on the mode of operation. Because of time
constraints and the limited storage capacity for spent backwash and rinse water,
the length of the service cycle is not varied to reflect ammonia loading and
breakthrough but rather maintained constant. The regeneration is controlled
from the main panel. The regeneration cycle is essentially completely auto-
mated however the operator must set the sequence times and flow controllers
prior to initiation.
The brine recovery process is controlled at a local panel adjacent to
the brine storage tanks. The following instruments are available for
control.
(a) brine pH monitors
(b) conductivity monitor
(c) temperature readouts for brine and stripper
(d) flow controllers for brine transfer and steam
(e) stripper liquid level control
The brine recovery process is manually controlled with the aid of two automated
flow control loops (brine and steam feed rates) and automatic control of the
product level in the stripper bottom. The set point values for the controlled
variables are maintained constant from day to day and changed only when problems
are encountered such as high head loss in the heat exchanger or high temperature
in the ammonia receiver.
30
-------
V. ANALYTICAL PROGRAM
A. SAMPLING AND ANALYSIS SCHEDULE
1. Treatment Processes
The original demonstration project agreement specified the location and
frequency of sampling along with an analysis schedule. The sampling frequency
and analysis schedule was modified when the Minnesota Pollution Control Agency
issued the NPDES discharge permit for the plant in June 1975. During the
project it became obvious that additional sampling and analyses were required
to better evaluate the unit processes. The sample identification scheme used
at the plant and throughout this report is presented in TABLE 3.
TABLE 3. SAMPLE IDENTIFICATION SCHEME
Sample
Label
Plant influent
Clarifier influent
Clarifier effluent
First stage filter effluent
Primary carbon column effluent
Carbon column system effluent
Second stage filter effluent
Influent to primary zeolite column
Primary zeolite column effluent
Zeolite column system effluent
Plant effluent
Raw
Clarin
Clar
DMF
GCd
GCC
INAEC
AEC,
AEC
Eff
'1
31
-------
All of the above samples were collected on a flow proportional basis. The
flow signals used to activate the samplers and the location of the sample taps
are summarized in TABLE 4. The raw sewage sample was transported from the wet
well area to the sampler on the main floor through iron pipe using a Moyno
model WA400 grinder pump. The DMF and DMF2 samples were delivered to the
sampling apparatus through Moyno moael FM22T pumps. All other samples described
in TABLES 3 and 4 were delivered to the sampling apparatus through PVC pipe
and polyethylene tubing by the head available at the sample tap.
The Raw and Clarin samples were collected in a Sanford Products Corp.
sampler model TC-2. The samples were collected in one gal polypropylene
bottles under refrigation. All other flow composite samples were transported
to the 8 ft x 11 ft sampling and monitoring enclosure, Figure 2, and collected
in the custom built apparatus described in Figure 12. Each sample flowed
through a constant head box constructed of acrylic sheet and discharged through
polyethylene tubing to the sample box also constructed of acrylic sheet. A
double acting pneumatic cylinder was used to locate the sample discharge above
the waste and sample compartments. The sample compartment discharged to a
refrigerated 0.5 gal polyethylene bottle through Tygon tubing. The pneumatic
cylinder was actuated by a four-way pneumatic solenoid valve. The sample
volume collected during a day was controlled by the reset count controller
and the adjustable time delay relay. The relays were set to provide sample
times of 1 to 2 seconds. The count controllers were set to actuate the
sampler once every 1750 to 4000 gal. These combinations yielded sample volumes
of slightly less than one gal per day. Clean sample bottles were placed in
the refrigerator each day at approximately 0000 hr and 1350 hr. The two sample
bottles were stored in the laboratory refrigerator prior to mixing. The routine
analysis schedule utilized at the end of the project period is presented in
TABLE 5. For the most common sampling frequency, 3 per week, samples were
collected on Tuesday, Wednesday and Thursday and analyzed on Wednesday,
Thursday and Friday. This schedule was adopted to eliminate laboratory hours
on Saturday and Sunday, i.e. 5 day BOD's.
Grab samples of the clarifier blowdown were collected twice each weekday
during periods of high and low flow. The total solids concentrations of the
two samples are averaged to estimate the solids concentration for each day.
2. Activated Carbon Regeneration
The shift operator collected a sample of the regenerated carbon, prior
to quench, every hour the furnace was discharging carbon. After the sample
was cooled in a closed metal container a 50 ml volume was placed in the
composite bucket. After each furnace run the composite sample, approximately
4000 ml volume, was reduced to approximately 250 ml in volume with a riffle
type sample splitter (Humbolt model 3985). Samples of the moist spent carbon
were collected every two hours and placed in individual plastic bags. The
samples were dried overnight at 103°C and a composite constructed using equal
volumes from each grab sample. The following analyses were conducted on
the composite samples: apparent density, iodine number, ash and volatiles.
32
-------
. r J
^£
2^
CD
l-H
00
*^g
o
1
u_
Q
y
3
O
CU
CL
i
CL
40
cu
r"
g3
r~
^_
QJ
r
4-
r
s-
(O
^~
S-
a>
i
4-
r
5-
fO
,
o
c
r-
^_
fO
p >
o
s*.
r
a;
^
4-)
C
OJ
3
^
4-
4-
QJ
4-
O
E
(O
QJ
i_
4-)
CO
CL
X
O
.a
s-
r
QJ
3
i_
CU
r-
4-
r-~
$~-
fO
i
S-
*T3
r
CJ
r
i-
OJ
o
E
£3
sx
c~
fO
4_>
QJ
cn
n3
S-
O
4-2
CO
i-
QJ
4-)
(O
2
a
QJ
S-
QJ
4->
r-
4-
O
4J
CU
CL
CL
I *
QJ
r-~
c~
1 i
S-
QJ
r
4-
r
S-
(13
^~
O
f -
u_
s:
Q
OJ
Q.
r-
CL
4-)
c
QJ
3
^
4-
4-
OJ
C
E
3
n~
O
O
E
3
i
O
0
c
o
_Q
t.
rt3
0
i
o
o
C3
CO
S-
QJ
QJ
en
(13
40
co
~d
c
o
CJ
QJ
CO
S-
o
4-
co
QJ
>
^~
to
>
co
co
^
JO
T3
(13
4.)
CU
^~
c
r-
4-
O
E
(13
QJ
S-
4J
co
CL
E
3
i^
O
c~
O
JO
£_
(T3
CJ)
O
O
O
CO
s_
QJ
4->
n
r
4-
-C
4-^
0
_Q
1
to
4-*
^-
CU
E
s_
(O
CL
E
O
o
CU
cn
(D
^_
O
4->
CO
i_
QJ
(O
2
-C
co
(13
2
O
(13
CO
E
3
i
0
c~
0
C
QJ
3
i
4-
4-
QJ
4-
0
a
'o
4-
t
c
n3
P^
QJ
cn
s_
(13
-C
O
CO
r
Q
4_)
C
QJ
3
^-«
4-
M-
LU
CJ)
UJ
=1
1 1
c~
0
.,
4_>
i
-o
-o
(D
C\J
r
0
4-
0
E
(13
QJ
40
CO
CL
3
T3
^ .
0
4-
QJ -i-
CL C
r- (T3
Q- E
4-J 4->
£^ ^
QJ QJ
3 3
1 ^-~
4- 4-
4- 4-
QJ QJ
c c:
E E
3 3
^ ^
0 0
-j_j 4_>
c" ci
CJ>
UJ UJ
=t ==c
CL
E
3
CL
o
QJ
QJ
4-
^_
0
4->
ns
c~
r
S-
o
n
^~
0
4-
0
QJ
cn
S-
n3
c~
o
CO
«r
Q
4->
C"
QJ
3
,
4-
4-
UJ
4-
4-
LlJ
33
-------
O)
+->
CO
>>
co
-P
O
O>
O
O
O)
'n.
to
CO
O
O
O
u
T3
C
03
0)
S-
fO
CL
O
O
fO
O
O)
en
34
-------
B. PROCEDURES
The total phosphorus, Kjeldahl nitrogen and heavy metals analyses were
conducted at the MWCC central laboratory. The samples for P and N analyses
were acidified, refrigerated and transported to the laboratory once each week.
The samples for metals analyses were acidified and transported to the central
lab where they were refrigerated prior to analysis. All other analyses were
conducted in the plant laboratory by the laboratory staff except as previously
noted.
Laboratory methods and equipment are summarized in TABLE 8. The soluble
fractions for the BOD, COD and TOC analyses are defined as that portion passing
through a Reeve Angel 934 AH glass fiber filter. Soluble phosphorus is defined
as phosphorus passing a 0.45u, distilled water washed, membrane filter.
All flow composite samples were collected on a calendar day basis. The
analyses conducted at the Rosemount laboratory were generally completed the
day following sample collection.
C. CONTINUOUS MONITORING
Prior to the initiation of the two year study it was anticipated that a
limited amount of on line monitoring would be required to better define process
performance characteristics. The description of the monitoring equipment used
is summarized in TABLE 9. All of the monitoring equipment was located in the
sample room - Figure 2. The temperature, pH, conductivity and chloride ion
measurements were made on a continuous basis. The sample streams monitored
are presented in TABLE 10.
Two sample manifolds were constructed to allow semi-continuous monitoring,
for TOC & turbidity and phosphorus & ammonia, of up to 6 sample streams each.
Sample was delivered from the head box, Figure 12, to a polyethylene header
through air operated pinch tube valves. These valves were closed and opened
by applying and relieving pressure (air @ 30 psi) to the rubber membranes.
The air flow was controlled by 3-way solenoid valves which were, in turn,
controlled by a step programmer (ATC model 1800) and timer (ATC model 333).
Samples were cycled to the TOC and turbidity monitors at approximately 20 min
intervals and to the phosphorus and ammonia monitor at 10 min intervals. The
numbers and identities of the samples cycled to the monitors varied considerably
and were dependent on the mode of operation and plant or process operating
problems encountered during a given time period.
37
-------
oo
LU
cy*
1
Q
LU
CJ
O
o:
PI
_i
<
O
i i
H-
>-
1
<
"^
^£
LJU
O
>_
Di
s:
:>
rD
oo
.
CO
LU
ai
C
Ol
E
Q.
i
13
CT
LU
-o
£1
ro
T3
O
-C
_l_)
CU
:>
1
CO
>3
1
OO
03
C
o
4-J
CO
CU
3
Cl
c:
0
i- X
4-> 3
13 i
^ CU
-o s-
4_J f .
O ftJ
QJ 13
S- C
i (^
Q s:
Q Q
0 0
CQ O
UD r-.
OO OO
c\j LO
i i
*d- oo
> CTi
en
O r
i CU
o en
o c
LU =3;
O CU
i- >
4-> CU
CO CU
^
4-J
r-
-o
1
_Q
i_
13
1
O1
OO
LO
1
OO
CU
0
0
4-
^
0
-C
E
i i
co
~e>
iZ
o
co
CU
r
JD
ro
CU
i
1 *
4_)
CU
C/1
oo o o r-- LO
<^f o r~-- CTI oo
OO LO OO LO
t i i i i
oo oo oo oo oo
S-
cu s-
4-> CU
CU 4-1
E CU
0 E
E
i- nr
CU Q- CU
.c en
4-> o c:
OO ro
CU CM S- CU
-O O 4->
ro 4-> ro
i- tt) i S-
CO E >> 4->
3 -C !-
>> O 4-> S= C
i- O CU CU
o <: E o >
4-> - O
ro S- S-
S- cu i rs c_j
O -C ro O o
J3 co 4-J S- OO
ra !- O CU O
_i u. i s: ^
co
CU "O
i- >1 "1
^ 4* r
4-J -I CU O
ra c: "O co
s_ - -^
CU 1 S- r
Q- ro O ro
E ^ i 4J
cu n: t -c o
1 CX < 0 (
^o
OO l
LO
1
00 1
S-
CU
N
ro
O
^
LO 3
LO =£
l/l
T3
'o
CO CO
CU S-
r- O
r- -C
4-J Q_
ro CO
i O
0 -C
> a.
f^ 1
ro ra
4-J 4-J
0 0
h- 1
O
oo
LO
1
oo
o
r^^
o
i
c:
O
c_
4_>
U
CU
Q.
OO
-Q
E
O
_J
03
-C
O
CO
a
CQ
rt
cu
-a
r
S-
o
-C
o
CO
^
o
c~
c:
ro
4_>
00
o_
0
-in
-i >
s^
0
1
ra
4-J
o
1
CO
CNJ
CNJ
1
oo
o
^
o
1
c~
o
S-
40
o
CU
Q.
00
fl
E
o
1
o?5
c~
o
CO
^
ra
CQ
f<
£^
o
1
4_)
ra
N
*l
S-
cu
CO
co
CU
4-J
(_)
CU
^_
1
Q
^^"
ro
r
c
o
E
E
<
1
1
S-
01
N
ra
o
3
<
c
ro
o
r-
c~
f
0
CU
1
1
c
o
If
4-)
CO
CU
en
1
-a
-o
o
ra
o
1
s-
0
r
c~
o
S-
a>
CI-
TS
CU
ro
E
o
4->
^
<
~z*"
r
c~
ro
T3
O)
^
38
-------
-o
0)
3
C
i
41
C
O
C T
00
LU
CO
cC
1
O)
en
03
Q-
1
0)
O
cu
S-
01
o>
4-3
d
O)
E
d
i
7?
cr
LU
~o
c
ft3
~o
o
c~
4_)
CD
*>"
00
i
{/)
>s
f
ftf
c:
^£
^i-
00
f,
cn
r*~" i
i
i
OO LO
^
r-^
0
^j-
S-
0)
cu
E
c
o
-l~
(_)
p-
C|
.,
U
cu
Q.
to
c
o
i-
o
**
en
i
i
oo
en
i
OJ
T3
O
E
CU
c ~o
0 O
i S-
4-1 4-1
(O U
S- OJ
*i CD
4_>
d
E
c~
0
r
.^
l*~
o
O)
Q-
CO
c~
O
*i
£_,
0
^
o
CM
1
cn
oo
LO
1
oo
r
!
CU
-o
0
E
OJ
"O
o
&
4-1
(J
OJ
OJ
s_
O)
CO
r
UL.
E
^
1
T5
O
OO
1
1
to
1
oo
IT
1
OJ
-o
o
E
4_>
i
jy
4->
00
OJ
4_>
0)
-o
.,
H-
r
3
CO
-C
o
03
3:
cu
-a
i
4
I*
13
00
LO
LO
1
^~
t_
O)
4-J
CU
E
O
Q
O
OO
OO
XX
U
03
4_)
OO
°s
£Z
O
4-1
CO
0)
0)
cn
>^
X
o
T3
OJ
>
i
0
CO
to
1
a
CM
t-~
r
1
OO
^_
0
4-^
03
4-1
r
41
CJ
r
5-
4_)
O)
E
0
cu
CL
E
03
C"
03
c~
£_
O)
i
1
o?S
OJ
u
03
^Z
03
3
~o
1
00
Q\
c?
CD
c:
i
s-
o
r
c~
O
^~
CT>
|*O
1
OO
J^~
4->
(T3
_Q
^_
01
4-J
03
2
CO
LO
*^-
<^J-
^
S-
O)
4->
r
r
4-
O)
c~
03
S-
JD
E
O)
E
3
LO
CD
E
S-
O
4-
r-
i
0
(_)
(
03
(J
01
U-
1
OO
LO
O
OO
~\
00
o
1
s-
0)
E
UJ
c~
^
^^
|L
01
a.
i
c~
0
i
4_>
Q.
i_
O
CO
-Q
03
0
E
O
4-J
CO
r
O3
40
fll
^>
>>
>
03
O)
m
oo cn
CM *3-
CO OvJ
i i
CM LO
i
r
r
^>
'
03
ca
o
C"
1 1
ri
CO LO
OJ 1*^
i
S- i
4-1 OJ
CO T3
3 O
"0 E
c
1 1 i-
OJ
X -i«i
O) 03
f^l C~
OO OO
A ^
^ r«.
i- OJ
OJ S-
jz s-
4J 3
O CO
to
tf n
t
4-J i
C 1
- 0
O 0
0-0
00
LO
s- 'oJ
0 T3
LO E
4-3
«i
CO S-
C 0)
OJ -Q
/*^> f
3
4-3 ^~
c
O) O)
S- C
03 !-
0- "O
Q- 0
OO
r^
oo
i
r~-
OJ
o
r
U
^
S-
o
t-
0
o
r
>
«
OJ
u
03
C
^_
3
t)
O
0
LO
r~-
c~
to
e^
C
o
_Q
C
03
O
LO
r^
oo
i
00
O)
^
o
1
0
13
s~
u
1
cu
o
1
c
*1
OJ
o
03
c
S-
^3
4-
O
fo
LO
cn
to
o>
r
[ ^
03
f
O
^>
C
O
£_
03
0
39
-------
TABLE 9. MONITORING EQUIPMENT
Parameter
Hardware
Tmperature
PH
Type T thermocouple, Leeds and Northup (L & N)
Speedomax W multipoint recorder
L & N 7779 flow through transmitter mounting with
electrodes, L & N 7073 multi-channel pH receiver,
L & N Speedomax W multipoint recorder
Specific conductance Uni-Loc flow through cell No. 400176, Uni-Loc
conductivity transmitter No. 711, Soltec recorder
No. B261A
Chloride ion
TOC
Turbidity
Phosphorus
&
Ammon i a
Uni-Loc flow through sensor No. 320, Orion chloride
electrode No. 94-17, Uni-Loc selective ion trans-
mitter No. 1032, Soltec recorder No. B261A
Astro Ecology No. 1000 total organic carbon analyzer
Hach No. 2411 Surface Scatter turbidimeter
Soltec recorder No. B161
Technician Autoanalyzer II - porportioning pump II,
2 single channel colorimeters, dual pen recorder,
2 continuous type filters, ammonia cartridge, total
inorganic phosphate cartridge, high temperature bath
TABLE 10. CONTINUOUS MONITORING SCHEDULE
Sample Stream
Clar
Temperature x
pH *
Specific conductance -
Chloride ion
DMF] GCC DMF? AEC EFF
X X X X X
X X X X X
- - X
- - - - x
* recorded at main process control panel
40
-------
VI. STAFF
The permanent staff involved in operating, maintaining and evaluating the
Rosemount AWTP during the two year study consisted of 17 persons as presented
in TABLE 11. Two additional operating attendants were used during the summer
months.
TABLE 11. PERMANENT STAFF POSITIONS
Position
Operator
Assistant operator
Operating attendant
Electrician
Pipe fitter
Secretary
Chemist
Laboratory technician
Superintendent
Project manager
number
4
4
2
1
1
1
1
1
1
1
shift
all
all
day
day
day
day
day
day
day
day
The staff during the afternoon (3 pm - 11 pm) and midnight (11 pm - 7 pm)
shifts consisted of an operator and an assistant operator. The day shift staff
during the normal work week consisted of an operator, assistant operator, two
operating attendants and the other permanent staff. On weekends the plant
was staffed with two-man shifts only. Essentially all of the maintenance work
was conducted by the plant staff.
During the winter months the sludge generated at the plant was hauled to
another MWCC facility for disposal by a private contractor. During the summer
months of 1975 and 1976 the sludge was injected on land adjacent to the plant.
All sludge generated during 1977 was hauled to other disposal facilities. The
injection work during the 1975 season was conducted by MWCC personnel not listed
above. During the 1976 season the plant staff carried out approximately
41
-------
one half of the injection tasks.
Janitorial services for the control room, locker room, laboratory
and offices were provided by a private contractor.
42
-------
VII. COST ACCOUNTING
A separate accounting system was established at the Rosemount facility in
June, 1974 to develop cost data for each of the processes utilized. The plant
was divided into 17 cost centers as illustrated in TABLE 12. Physical
boundaries define 15 of the cost centers. One cost center, Laboratory
Services, is defined by function and the remaining cost center, Indirect
Services, is used to identify charges that cannot reasonably be identified
with direct plant operation.
During normal operation, the three major costs for a plant of this type
are labor, chemicals, and power. All MWCC employees at the plant completed
two time cards on a biweekly basis, the regular time card for pay purposes
and a supplemental time report to identify hours in the various cost centers.
Time estimates were made in half hour increments.
Chemical usage was recorded on a daily or biweekly basis throughout the
study as summarized in TABLE 13. A report of chemical usage was prepared on
a biweekly basis to facilitate the calculation of average dosage rates for
the two week periods.
Electric power and natural gas were metered at the plant entrance. The
volume of fuel oil in storage was monitored. Fuel usage was not metered at
the appropriate cost centers but it was readily estimated. The average line
power consumption of each of the 20 largest motors was determined with the
combination of a watthour meter (GE no 700X1365) and digital clocks to record
time-on. The time-on was recorded for each of the 20 drives on a biweekly
basis.
_A11 materials, supplies, and services purchased for the Rosemount facility
during the two year study were charged to the appropriate cost center/s.
43
-------
TABLE 12. COST CENTERS
Cost Center Title and Description
10 PRELIMINARY TREATMENT - plant inlet to discharge to wet well,
includes flow measurement, bar screens, raw sampling pump
20 INFLUENT PUMPING STATION - wet well to discharge to clarifier
30 CLARIFICATION & PHOSPHORUS REMOVAL - inlet to clarifier to weir
box on clarifier
40 FILTRATION-! - weir box on clarifier to discharge to filtered
water storage tank 1
50 CARBON ADSORPTION - filtered water storage tank 1 to weir box
feeding 2nd stage filters
60 FILTRATION-2 - weir box to and including filtered water
storage tank - 2
70 AMMONIA REMOVAL - ion exchange pumps to discharge from
building
80 CHLORINATION - chlorine system for disinfection and odor
control
90 EFFLUENT DISPOSAL - building to outfall
100 SLUDGE HANDLING & DISPOSAL - blowdown line (outside clarifier)
to discharge from truck
110 CARBON REGENERATION - transfer and regeneration
120 CLINOPTILOLITE REGENERATION
130 AMMONIA RECOVERY SYSTEM - condenser and storage tank
140 AMMONIA DISPOSAL - disposition of stored ammonia solution
150 LABORATORY SERVICES - sampling, analysis, monitoring
160 BUILDING & GROUNDS - general systems including water, drain,
compressed air, heating, ventilation, cleaning, etc.
170 INDIRECT SERVICES - nondirect labor, plant supervision, project
supervision, clerical labor, data processing and other indirect
services
44
-------
TABLE 13. CHEMICAL INVENTORY SYSTEM
Chemical
Record
Ferric chloride
Hydrated lime
Sulfuric acid
Polymer
Nitrate
Sodium chloride
Soda ash
Sodium hydroxide
Chlorine
Activated Carbon
level measurement in day tank daily and storage tank,
biweekly
measure concentration of each batch slurry elevated to
storage tank, daily
level measurement in storage tank, biweekly
volume of liquid polymer added to dilution tank, daily
weight of sodium nitrate added to dilution tank, daily
number of 100 Ib bags used, daily
weight used, daily
volume of 50% solution used, daily
rotometer settings on chlorinators, daily
number of 50 Ib bags used, daily
45
-------
VIII. WASTEWATER CHARACTERISTICS
During the two year reporting period the metered raw wastewater flow
averaged 242,100 gpd. The sewered population in the City of Rosemount is
estimated to have been approximately 3000 during the study. The recorded
sewer services were as follows: residential - 800, institutional - 10,
commercial - 35, and industrial - 2. The records of the Metropolitan Waste
Control Commission indicate the industrial flows averaged 13,700 gpd and
consisted primarily of sanitary sewage.
Flow variations throughout the day and from day to day were substantial
as is typical for small sewer systems. The average dry weather influent
hydrograph for the period October 27 through November 2, 1975 is presented in
Figure 13. These data illustrate dry weather flow variations in excess of
4:1 (max:min). The variation of daily flows is characterized as follows:
minimum - 85,000 gpd, maximum 642,000 gpd, and standard deviation - 51,000 gpd.
The range of values is considerable because of infiltration and inflow during
wet weather and because the main gate was periodically closed to allow for
repairs. After the repairs were completed the wastewater accumulated in the
interceptor was treated at higher than normal rates.
The chemical characteristics of the raw wastewater are summarized in
TABLE 14 and Figure 14. The average values presented are typical of domestic
sewage with the exception of the value of chloride and the relationship be-
tween ammonia and Kjeldahl nitrogen. The chloride and hardness concentrations
in the Rosemount municipal water supply were determined on several occasions
and found to be approximately 5 mg/1 and 300 mg/1 respectively.
The elevated chloride level of the raw wastewater was most likely due to
the discharge of industrial and commercial wastes and the discharge of spent
brine during the regeneration of individual water softeners of the ion ex-
change type. The relatively low value for organic N and corresponding high
value for NhU-N are most likely caused by the biochemical degradation of
organic nitrogen compounds, and associated release of ammonia, in the inter-
cepting sewer. Approximately 25,000 ft of pipe varying from 27 to 48 in.
diameter are used to transport the wastewater from the edge of the sewered
area to the treatment plant. The calculated velocity of flow is 1 fps for
a flow rate of 0.25 mgd. During periods of low flow scour velocity was most
likely not maintained and organic solids deposited.
46
-------
250
200
150
s:
CL.
O
100
50
\-r
0600 1200
TIME OF DAY
1800 2400
Figure 13. Typical dry weather influent hydrograph.
47
-------
TABLE 14 RAW WASTEWATER CHARACTERISTICS
Parameter
Temperature °F
pH
Alkalinity, mg/1 CaC03
Hardness, mg/1 CaCO^
Calcium, mg/1 CaC03
Chloride, mg/1
Total Solids, mg/1
Total Volatile Solids, mg/1
Suspended Solids, mg/1
BOD - total, mg/1
BOD - soluble, mg/1
COD - total , mg/1
COD - soluble, mg/1
TOC - total , mg/1
TOC - soluble, mg/1
Total Phosphorus, mg/1
Total Soluble Phosphorus, mg/1
Total Ortho Phosphorus, mg/1
Soluble Ortho Phosphorus, mg/1
Ammonia-nitrogen, mg/1
Kjeldahl-nitrogen, mg/1
Copper, mg/1
Nickel , mg/1
Lead, mg/1
Zinc, mg/1
Cadmium, mg/1
Chromium, mg/1
Iron
Number of
Analyses
703
697
309
97
31
300
93
91
319
316
305
317
301
313
232
299
283
305
302
440
258
14
14
14
14
7
14
13
Min
44
6.7
212
182
128
106
996
293
29
71
19
150
51
35
11
4
2.6
3.1
2.4
11
19
0.05
-
<0.05
0.17
<0.02
<0.05
0.93
Max
66
9.2
622
452
274
826
2338
999
1530
999
200
2435
350
496
200
52
25
26
24
99
270
0.15
-
0.20
0.57
<0.02
1.00
3.35
Ave.
55.5
7.6
427
296
177
311
1426
556
237
211
60
505
143
104
42
11.7
8.5
9.2
8.2
35
42
0.08
<0.05
-
0.27
<0.02
-
1.43
Std.
Deviation
4.5
0.2
38
43
33
103
272
149
152
84
24
234
36
47
19
3.7
2.3
2.4
2.3
7
17
0.04
-
-
0.10
-
-
0.62
48
-------
1000
CONCENTRATION - mg/1
i O
0 0
^^
---"*
/
/
/
/
-f *""
Q-
^ 1
COD
,*
""s^
1
^^*'
rvii
... *^
j 2
^"'
f^"^
~~~
i-i
~~&
^*
-.
^-J
0 4
i -
"4^
.4--
1 '
5**
*
0-
r<**"
^C
. ' '
ss
+^
^4
3
J 60 8
^^**"
^
'BOD
-
-er
TOT;
^* ~
^,^***
,x
^
jS
tTo
*'
&''
^-
NH^-N
-------
IX. PLANT OPERATING CHRONOLOGY
A. GENERAL
The Rosemount AWTP was put into service in mid-November 1973 and since
that time has been the only wastewater treatment facility serving the City
of Rosemount. Although the on-line processes were all operational at that
time, the carbon and zeolite regeneration systems were not available for
service. The initial carbon regenerations were conducted in November 1974
and the zeolite regeneration system was put into service in May 1975. The
original plant start up which was supervised by the contractor's representative
lasted approximately 6 weeks. The contractor however, through his representa-
tive, retained the ultimate responsibility for plant operation through May
1975 although the plant was staffed by Metropolitan Waste Control Commission
personnel. Although the facility was not formally accepted by the Metropolitan
Waste Control Commission until November 4, 1975 the contractor was relieved
of the direct responsibility for plant operations subsequent to the formal
performance test which was conducted May 11 through May 15, 1975.
During the approximately 18 months the contractor was responsible for
operating the plant a number of operating modes were used. For several
months both treatment trains were used. During the entire period however the
overall flow scheme was constant and consisted of: clarifier, dual media
filters, one or two carbon columns, and dual media filters followed by chlor-
ination. The operating modes for the clarifier/s and carbon columns are
summarized in TABLE 15.
Prior to initiating the study it was planned to operate the plant utilizing
six defined process configurations. Each configuration or mode of operation
was to be used for an eight week period. The most successful mode was then to
be used for approximately one year of operation. During the study however the
schedule was modified to reduce the number of operating modes fo the five de-
scribed in the following sections.
B. MODE I
The process configuration used during the eight week period June 7, 1975
through August 1, 1975 is illustrated in Figure 15. All the treatment units
of train number 2 were utilized during this period. The clarifier was operated
with a chemical feed combination of lime, ferric chloride and polymer. The
carbon columns were operated in the series upflow mode. The zeolite columns
were operated in the series downflow mode. The average raw sewage flow to
the plant was approximately 0.22 mgd.
50
-------
TABLE 15. CLARIFIER AND CARBON COLUMN STATUS
Month
Clarifier
Chemicals
Carbon
Column
1973 Nov.
Dec.
1974 Jan.
Feb.
March
April
May
June
July
August
Sept.
Oct.
Nov.
Dec.
1975 Jan.
Feb.
March
April
May
Fed3
Polymer
Lime
Fed 3
Polymer
Lime + Polymer
Lime
FeCl3
Polymer
several comb.
2 col. series downflow
several comb.
2 col. series upflow
Col. 4,5
series upflow
Col. 5,6
series upflow
Col . 4 downflow
several
combinations
2 trains,
2 columns
series
downf 1 ow
51
-------
?io i
~7
SECOND
STAGE
FILTER
Figure 15. Flow scheme during Mode I operation.
52
-------
The operating conditions of the clarifier are summarized in TABLE 16.
The lime and ferric chloride dosages were relatively constant throughout
the eight week period. The polymer dose averaged 1.5 mg/1 during the first
four weeks and 4.2 mg/1 during the second four weeks. The increased polymer
dose was required by the operators to control the sludge bed level.
First stage filters number 3 and 4 and second stage filters number 7
and 8 were in service during this reporting period. The filters were taken
out of service for backwashing only.
TABLE 16. CLARIFIER OPERATING CONDITIONS - MODE I
Average flows - gpd
raw 222,000
clarifier 277,000
Mixing zone pH
average of 1290 hourly values 10.5
Effluent pH
set point 7.5
average observed 7.5
Average chemical additions
lime - mg/1 CaO 424
ferric chloride - mg/1 Fe 14.1
polymer - mg/1 Nalco 677 2.9
sulfuric acid - mg/1 SO^-S 51
The backwashing frequencies were as follows:
filter no. no backwash cycles
3 77
4 77
7 59
8 58
53
-------
The backwash water flow amounted to approximately 14,500 gpd.
Carbon columns 4 and 5 were operated in the series upflow mode. Column
number 4 was regenerated and put into service on March 20, 1975. It was
used as a primary column in the downflow mode for 21 days and as a primary
upflow column for 39 days prior to June 7, 1975. Column under 5 was re-
generated and put into service on May 5, 1975. It was used as a secondary
upflow column for 33 days prior to June 7, 1975. The carbon regeneration
data are as follows:
Apparent Density 12 Number
Djite Column Spent Regen Spent Regen
3/20/75 4 0.53 0.52 710 795
4/28/75 5 0.49 0.46 850 1020
The regeneration of column 4 was not complete because of scrubber, induced
draft fan, and burner outages during the run.
A control system was provided to recirculate flow around the carbon
columns tomaintain the upflow rate equal to, or greater than, a manually
entered set point flow. In the automatic mode the control valve cycled.
In order to avoid mechanical damage to the piping system the recycle system
was operated in the manual mode. However the system received little operator
attention and the recycle valve remained open during periods when the forward
flow was in excess of the minimum flow set point. The flow through the carbon
columns was thus diluted and the contact time needlessly reduced. Recycle
was terminated on July 28, 1975.
The carbon columns were backwashed at approximately 5 day intervals and
at a rate of 12 gpm/sq ft. Columns 4 and 5 were backwashed 11 and 10 times
respectively during the reporting period. The backwash water requirement was
approximately 3400 gpd.
The ion exchange columns were in operation for 22 days during the re-
porting period. It was found that because of the constraints of time and
limited storage capacity for spent backwash water only one complete regenera-
tion-strip cycle could be routinely accomplished per day. The regenerations
were conducted with a IN sodium solution at a pH of 11.0. The regeneration
sequence utilized is summarized in TABLE 17. Approximately 17,700 gal were
recycled to the headworks via the backwash pit during each regeneration
cycle. Because the columns were in service only 22 days the recycle flow
averaged only 6950 gpd.
C. MODE II
The mode II configuration illustrated in Figure 16 was used a total of
34 days for the periods August 2-21, September 5-12 and September 25-30, 1975.
Treatment train number 2 was used with the first stage filters out of service
- bypassed. The zeolite columns were out of service during the entire period
54
-------
/ /
77 7 7 7 /
/7 /
/ / " / 7
SECOND
STAGL
FILTFR
Figure 16. Flow scheme during Mode II operation.
55
-------
because of problems with the regeneration system.
TABLE 17. ZEOLITE REGENERATION SEQUENCE - MODE I
Function
Control
Column backwash - 420 gpm
Settle bed
Brine upflow - water to waste
Brine upflow to recovery - 100 gpm
Air displacement of brine to recovery
Refill upflow - 100 gpm
Settle bed
Rinse to waste - 150 gpm
10 min - timer
5 min - timer
conductivity control
low level in brine tank
120 min - timer
45 min - timer
5 min - timer
60 min - timer
The average clarifier conditions are summarized in TABLE 18. The lime
dose was relatively constant but the ferric chloride and polymer concentrations
varied substantially as illustrated in TABLE 19. The polymer dose was reduced
intentionally to eliminate the possibility of polymer discharge from the
clarifier. At that time polymer coating or blinding of the activated carbon
granules was considered a possibility.
The drastic reduction in ferric chloride dose occurred when the feed
pump was adjusted to give a dose of 7 mg/1 as Fe. It was found however that
the pump calibration was not accurate in the extreme low range. At that
time the feeders pumped directly from the ferric chloride storage tank to the
clarifier. Because the tank is 8 feet in diameter, it was difficult to de-
termine consumption for periods of less than 10-14 days. Two 55 gallon day
tanks were put into service to eliminate this problem - Figure 7.
The first stage filters (number 3 and 4) were taken out of service at
2400 hr August 1, 1975. After about one week of operation, the backwash
requirements of the second stage filters increased substantially as illustrated
in TABLE 20. On August 13 both filters were chlorinated prior to backwash to
eliminate the problem of rapid head loss increases. This procedure had been
used successfully during the previous year to remove the scum formed on the
surface of the second stage filters. During the week of August 14-21 the
head box feeding filters 7 and 8 overflowed on numerous occasions. These
overflows discharged to the backwash pit and substantially increased the
plant recycle flow.
56
-------
TABLE 18. CLARIFIER OPERATING CONDITIONS - MODE II
Average flows - gpd
raw 226,000
clarifier 285,000
Mixing zone pH
average of 782 hourly values 10.6
Effluent pH
set point 7.5
average observed 7.6
Average chemical additions
lime - mg/1 CaO 474
ferric chloride - mg/1 Fe 10.5
polymer - mg/1 Nalco 677 2.8
sul^uric acid - mg/1 SO^-S 56
TABLE 19. CLARIFIER CHEMICAL FEED - MODE II
Period
August 2-21, 1975
September 5-12, 1975
September 25-30, 1975
FeCl3-mg/l Fe
14.5
2.5
10.5
Polymer-mg/1 Nalco
4.1
1.0
0.4
677
57
-------
TABLE 20. FILTER BACKWASH FREQUENCY - HOPE II
Date
August 2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
September 5
6
7
8
9
10
11
12
25
26
27
28
29
30
Total
Filter
7
2
1
1
1
2
2
2
5
3
2
3
6
3
3
3
2
3
2
2
3
2
1
1
3
2
2
3
1
1
5
1
2
1
1
77
Number
8
2
1
1
1
2
2
1
4
2
2
2
5
2
3
3
2
3
1
2
3
2
1
1
3
3
3
3
3
3
4
1
2
1
2
76
58
-------
The first stage filters (number 3 and 4) were put back into service on
August 22. On August 27, filter number 8 was drained down and the media
surface exposed. The surface mat had a consistency of wet clay and could
easily be formed by hand. The surface mat, approximately 1/2 inch deep,
was raked off and the filter put back into service.
On September 3, the headbox serving the second stage filters was
examined. It was observed that when one filter was being backwashed or
had reached its hydraulic capacity and flooded its headbox inlet the headbox
would overflow before the second filter reached its hydraulic capacity. The
headbox design was modified to eliminate this problem.
Based on suggestions from the plant operators and laboratory testing,
it was found that the filter media cleansing was expedited by a presoak in
a solution of 0.2N sodium chloride. This cleaning technique was used on
filters 7 and 8 on September 2. The media was soaked in the salt solution
for approximately 30 minutes. The media was then air scoured for 15-20
minutes prior to normal backwash.
The first stage filters were taken out of service early on September 5.
The backwash requirements for the second stage filters again gradually in-
creased. On September 12 during an attempt to treat the carbon columns with
a salt solution a considerable quantity of scum, grease and other debris was
released from the walls of filtered water storage tank number 1. All of the
material had been deposited below the water line. The tank was taken out of
service and thoroughly cleaned.
A carbon regeneration was initiated on September 13. The spent scrubber
water from the carbon furnace discharges to the clarifier effluent line. With
the first stage filters out of service, any solids in the spent scrubber
water would have been deposited directly in the clean storage tank. For this
reason the first stage filters were put back into operation on September 13.
When the regeneration was terminated the first stage filters were again
taken out of service - September 25.
On October 1, storage tank number 1 was drawn down and inspected.
Grease deposits and scum were observed on the water surface and tank walls.
The first stage filters were put back into service to reduce or eliminate
the transport of more of this material to the storage tank and eventually
to the carbon columns. The attempt to operate the plant with the first
stage filters out of service was terminated.
For the reporting period the filter backwash requirements averaged
approximately 18,000 gpd. The backwash requirement was not substantially
reduced during the period the first stage filters were out of service.
Carbon columns 4 and 5 were operated in the upflow series mode until
August 11. At that time column 4 was taken out of service, column 5 was
moved to the primary position and column 6, which was freshly regenerated,
was brought into service in the secondary position - again in the upflow
mode. For the reporting period, the backwash requirement averaged approxi-
mately 9500 gpd.
59
-------
D. MODE III
The process configuration during mode III was identical to that of mode I
as illustrated in Figure 15. Data were collected for the 78 day period
August 22 - September 4, September 13-24, and October 1 - November 21, 1975.
The average operating conditions for the clarifier are summarized in
TABLE 21. For the 7 day period August 23-29 no lime was available for use in
the clarifier because of a late delivery. During the interim period FeClj was
used as the prime coagulant at an average dose of 54 mg/1 Fe along with polymer
at an average dose of 4.1 mg/1. All polymer use was terminated on November 16,
1975.
Both the first and second stage filters were in service (number 3, 4, 7,
8) during the reporting period. The backwash requirement averaged approxi-
mately 23,200 gpd. This was significantly higher than the backwash require-
ments in Modes I and II. During the 4 day period August 26-29 the head loss
buildup in the first stage filters was more rapid than usual and frequent
backwash was required. During this same period difficulty was encountered
in the laboratory in the suspended solids analysis. It was not possible to
filter more than 50 ml of raw sewage through an 11 cm glass fiber filter although
the solids concentrations were normal. Under normal conditions 250 ml was
filtered easily. On August 30 and 31 the two first stage filters were back-
washed a total of 24 times, once with a salt solution. Backwash requirements
for the primary filters were normal after August 31. After those periods
during which the primary filters were out of service the secondary filters
required frequent backwash and special cleansing to remove the scum and debris
collected.
Carbon columns 5 and 6 were operated in the series upflow mode for the
entire reporting period.
The ion exchange columns were in service for 15 days near the end of
the reporting period. Columns 5 and 6 were used on alternate days for 24 hour
periods. The effects of multiple regenerations and regeneration at pH values
greater than 11 on zeolite exchange capacity were determined.
E. MODE IV
The general process configuration illustrated in Figure 17 was utilized
for a period of 420 days, November 22, 1975 - January 14, 1977. During the
first 9 days of the period only one carbon column was in service in the down-
flow mode. Two columns in the series downflow mode were used for the following
411 days. The zeolite columns were used for a total of 123 days with one or
two columns in service.
The clarifier conditions are summarized in TABLE 22. For the first
15 days of mode IV the internal pH of the clarifier was maintained near 10.9.
For the periods December 6, 1975 through 1200 hr June 16, 1976 and 1200 hr
July 9, 1976 through January 14, 1977 lime was fed at a reduced rate to maintain
60
-------
TABLE 21. CLARIFIER OPERATING CONDITIONS - MODE III
Average flows - gpd
raw 209,000
clarifier 292,000
Mixing zone pH
average of 1753 hourly values* 10.5
Effluent pH
set point * 7.5
average observed * 7.5
Average chemical additions *
lime - mg/1 CaO 492
ferric chloride - mg/1 Fe 9.1
polymer - mg/1 Nalco 677 - terminated
November 16, 1975 0.95
Sulfuric acid - mg/1 S04-S 64
* excluding data for August 23-29, 1975
61
-------
Figure 17. Flow scheme during Mode IV operation.
62
-------
TABLE 22. CLARIFIER OPERATING CONDITIONS - MODE IV
Average flows - gpd
raw 238,600
clarifier 293,000
Mixing Zone pH
Nov. 21 - Dec. 5, 1975 - Period I 10.9
average of 336 hourly values
Dec. 6, 1975 - June 16, 1976, 1200 hr) .
July 9, 1976, 1200 hr - Jan. 14, 1977) Ferlod u 9.5
average of 8714 hourly values
Effluent pH
set point, Periods I & II 6.5 to 7.5
average observed Periods I & II 7.5
June 16, 1976, 1200 hr - July 9, 1976, 1200 hr - Period III
average observed 7.5
Average chemical additions
lime - tng/1 CaO
Period I 467
Period II 338
Period III 0
ferric chloride - mg/1 Fe
Periods I & II 18.5
Periods III 54
polymer - mg/1 Nalco 677
Periods I & II 0
Periods III 5.7
sulfuric acid - mg/1 S04-S
Period I 76
Period II 63
TABLE 23. CARBON COLUMN STATUS - MODE IV
Dates
From
11/22/75
12/31/75
2/17/76
2/26/76
3/2/76
5/18/76
8/25/76
11/23/76
12/10/76
12/18/76
To
12/30/75
2/17/76
2/26/76
3/1/76
5/17/76
8/24/76
11/22/76
12/9/76
12/17/76
1/14/77
Treatment
Train
2
2
1 & 2
1
2
2
1
1
2
1
Carbon Columns
in service
*
i
4,5 series
2,3 series & 4,5 series
2,3 series
4,5 series
5,6 series
1 ,2 series
2,3 series
5,6 series
3,1 series
63
-------
a pH of approximately 9.5. For the 23 day period, 1200 hr June 16 through
1200 hr July 9, 1976, ferric chloride was used as the primary coagulant along
with a substantial quantity of polymer. This change was necessitated by a
prolonged labor dispute involving the regular lime supplier.
The status of the treatment trains and carbon columns are presented in
TABLE 23. Because of maintenance requirements both treatment trains were
used during this reporting period. However both trains were used simultaneously
for only 9 days during a period of extremely high raw wastewater flow which
was caused by inflow/infiltration during an early spring storm. The treatment
trains were alternated to accomplish required maintenance such as cleaning rags
from the clarifier mixing impeller and replacing valves on the filters and
carbon columns. Both the primary and secondary dual media filters were used
throughout the period.
A number of combinations of brine pH, brine volume and sodium concentration
were used for zeolite regeneration. For this reason the changes dealing with
the zeolite regeneration process are presented in a separate section.
F. MODE V
The process configuration illustrated in Figure 18 was utilized for 143
days, January 15, 1977 to June 6, 1977. Process train No. 1 was used throughout
the period. For a period of 5 days in March both treatment trains were used to
accommodate the high flows associated with snow melt. The second stage filters
were not in service during this operating mode.
The clarifier conditions are summarized in TABLE 24 and the status of
the carbon columns is presented in TABLE 25. As in mode IV several zeolite
regeneration procedures were used and are summarized in the following section.
G. AMMONIA REMOVAL SYSTEM
Although the treatment plant was put into service in November of 1973 the
ammonia removal system was not completed and tested until May of 1975. For
this reason the plant personnel did not have a great deal of operating experience
with this system when the demonstration project was initiated. During the
period June 7, 1975 through August 12, 1977 the zeolite columns were in service
a total of 277 days and a total of nine operating modes were evaluated as
summarized in TABLE 26. For the period prior to March 1977 the ammonia removal
system was in service on only a sporadic basis because of a series of problems
related to design, construction and installation and because of the energy
crisis which prevailed during the winter of 1976-77.
During period number 1 - TABLE 26 - the system was operated with two
zeolite columns in series. At midnight each day the primary column was taken
out of service for regeneration, the secondary column was moved to the primary
position and a freshly regenerated column was put into service in the secondary
position. The changes in column status were made on a time basis so as not to
64
-------
~7 /7 7 // / 7 7
FIRST STORAGE
STAGE TANK
FILTER NO 1
/ / 7 7
Figure 18. Flow scheme during Mode V operati
on.
65
-------
TABLE 24. CLARIFIER OPERATING CONDITIONS - MODE V
Average flows - gpd
raw
clarifier
Mixing zone pH
average of 3225 hourly values
Effluent pH
set point
averaged observed
Averaged chemical additions
lime - mg/1 CaO
ferric chloride - mg/1 Fe
sulfuric acid - mg/1 $04-8
280,900
327,600
9.75
6.5 to 7.5
7.3
327
19.6
52
TABLE 25. CARBON COLUMN STATUS - MODE V
Dates
From
To
1/15/77 2/28/77
3/1/77 3/8/77
3/9/77 3/13/77
3/14/77 6/6/77
Treatment
Train
1
1
1 & 2
1
Carbon Columns
in Series
3,1 series
1 ,2 series
1,2 series, 5,6 series
1,2 series
66
-------
B
i
oo
>-
OO
1
e£
^^
o
B
C£
e£
i i
2^
O
21
^£
O
|_l_
00
O
1 1
[
1 1
-y
O
o
1 1
1
Ct^
LU
p=^
UJ
o
1 1 1
o:
Q
=c
UJ
0
o:
UJ
oo
10
CM
UJ
1
CQ
1
CO
o
+-> O>
ro r
S- U
CO O) >,
ceo
O Ol
r- Ol O-
i- C£ i-
"O -l->
c 01 oo
o +->
0 01 S-
, cu
>! 0- Q.
S- E
Ol O
> 0
o
0
o>
OL
CD
c
r- E
S_ <0
CQ QJ
1 »
CO
_Q
Ol
c
1^
i.
ro
01
*
CO
E
o
4J
>n-
-a
c
o
0
c
o
1 <
CO
Ol
c
Ol
O"
0)
a;
K
^»
23
1
O)
C
J_
03
K
S^
1
03
Q_
CO CU
E -r^
3 C >
1 !- S-
0 0)
O 00
c
4- 0
O -r-
co ro
ro O)
Q 0.
O
o
o
'£
0)
Q.
CO
0)
o
S- -C ^"
O) O
Q.OO
o
LD LT>
O O
<^O ^O ^D *^ t*O C\J
>*
^- ^- ^- ^ ^J- OO ^~ ^d" ^"
r^ ^ i
ooounLOLnLOLnLO
r r- ( OOOOOO
(/) CO t/» CO C/l (/) CO
a> i
r*** ^ r^
en CM
LO t~-
*° CM * r^
CTi co vo co I"***- crj
r^- ^ cu 01 01 01 ^.
CJl CU -Q I r I I O) "
i J3 E *- i CM
E Ol " » S- :
OCJCUi OOet CO4-J
CM O) Q Id) CO
Q 1 Ol Ol CO C >> 3
i COr 3 3 Ol "-D "3 3
3 -o "-D i i --3 <:
-3 S- S- 1 1 Oil
1 CUO1CMCO "OOCMOO
r-~ JD J2 i oo r CM
E E -E .
C>OS^Cr^>sIcr^>
3OCU>
E ex
O uti
II
o
r^ CQ
*
* *
67
-------
disrupt the scheduling for the remainder of the plant. Because of the severe
raw flow variations (Figure 13) a serious attempt was made to return all recycle
flows to the headwords during periods of low raw flow and thus minimize fluctua-
tions in hydraulic loading throughout the plant. This required that all recycle
flows be generated on a scheduled basis because of the limited storage capacity
for spent backwash water.
The teflon lining of the brine transfer pump casing failed on July 10, 1975.
This failure allowed hot brine to leak from a weep hole at a rate of approxi-
mately 5 gpm. The regeneration system was out of service for approximately
3 weeks while the pump was repaired. After 23 days of operation it was apparent
that the zeolite was not being thoroughly regenerated because the ammonia removal
efficiency was well below design values. It was also apparent that the steam
stripping operation was not functioning properly because the observed condensate
volumes were approximately three times the design volume.
After several test runs it was determined that the stripper product level
control system had malfunctioned and allowed the stripping column to fill with
brine and overflow to the ammonia receiver. The control system piping was
modified, all instruments calibrated and a sight glass installed in the column
to allow visual determination of the liquid level if required. The regeneration
system was put back into service in the 4th week of September 1975.
Several plant scale experiments were conducted during October 1975 to
identify the ammonia elution characteristics as a function of brine volume
and pH. As a result of these tests the regenerations during period number 2
were conducted at pH 12.5. For this period of 28 days zeolite columns number
5 and 6 were used on alternate days. Only one column was used at a time be-
cause the sampling system had not yet been modified to allow collection of
the AEC] sample and it was desirable to identify ammonia removals per column.
On December 11, 1975 the regeneration procedure was again modified and
for a period of 13 days the regenerations were conducted at pH 13. The
ammonia removal system was taken out of service at midnight December 25, 1975
because substantial leaks had developed in both heat exchangers - Figure 9.
The heat exchangers were dismantled and the 170 plates cleaned individually.
All of the gaskets were replaced and several of the plates which had ruptured
were replaced. After the heat exchangers were repaired a system was constructed
to allow periodic in situ cleaning with dilute nitric acid.
The zeolite columns were put back into service on March 2, 1976 with two
columns operated in series as in period number 1. The regenerations were
conducted at pH 13 with a 0.5N NaCl solution. Several tests were made to
determine the need for heated brine storage. Based on the results obtained
the heat input (steam-heat exchanger) was terminated and both of the heat ex-
changers (Figure 9) were removed from the system.
Because of the extremely high cost for caustic and rinsing requirements
the regeneration pH was brought down to 12. The system was operated in this
manner for 14 days - period number 5. Starting in July 1976 the regeneration
procedure was modified to yield a 78% increase in brine volume for regeneration.
68
-------
The initial 4.6 bed volumes, BV, were passed through the zeolite following
the normal procedure, however, the spent brine was softened and passed through
the zeolite a second time. The pH was adjusted as required prior to the second
pass. During periods number 5 and number 6 it was observed that the ammonia
removal efficiency in the stripping column had decreased because of the lower
brine inlet temperature. One of the original heat exchangers was installed to
preheat the brine to the stripper and in turn cool the brine discharged from
the stripper. This modification was completed August 10, 1976. For the period
August 15 to October 21, 1976 the zeolite columns were out of service to allow
for repair of all the system valves some of which did not seal properly. The
system was out of service for the period December 2, 1976 through March 21, 1977
because of the failure of the teflon lining of the brine transfer pump casing
and because of the energy crisis which existed in the State of Minnesota during
the months of January and February 1977.
In late April 1977 the regeneration procedure was modified to increase the
brine throughput to 14 BV during each regeneration cycle. The steam rate used
for ammonia stripping was reduced by 25% in mid-June 1977. The final modifica-
tion during the reporting period was made on July 23, 1977 and consisted of
eliminating 50% of the stripping cycles by stripping on alternate days. The
data obtained during the period June 5 through August 12, 1977, periods number
8 and number 9 are included because they are significant in terms of system
evaluation.
69
-------
X. OVERALL PLANT PERFORMANCE
A. PRIOR TO DEMONSTRATION PROJECT
Despite the frequent discontinuities associated with plant startup and
shakedown the Rosemount AWTP produced a high, although somewhat variable,
quality effluent during its first year and one half of operation. The data
summary presented in TABLE 27 illustrate this fact. Some of the variability
in treatment efficiencies during the first year of operation was also due,
no doubt, to the fact that the operating personnel, although experienced in
conventional wastewater treatment, had no experience with the unit processes
utilized at Rosemount. The effluent suspended solids data demonstrate that
the experience of the plant operators, although not readily quantified, is
an important parameter when making realistic estimates for future plant
performance.
Several combinations of ferric chloride, lime and polymer were used to
remove phosphorus and aid solids removal. The treatment goal was an effluent
total phosphorus concentration of 1 mg/1. Experience in laboratory jar tests
and in the clarifier indicated that any combination of ferric chloride, lime
and polymer that effected the required phosphorus removal would also yield
high removals of suspended solids. The operator attention required to assure
high solids removals was found to vary substantially. During the 7 1/2 month
period, November 1973 - June 1974, that ferric chloride was used as the primary
coagulant (average approximately 45 mg/1 Fe) the clarifier was difficult to
control and subject to upset on a daily basis because of the unsteady hydraulic
loading and the characteristics of the ferric hydroxide floe. In order to pre-
vent significant floe carry over to the filters the operators would have to
increase the polymer dosage during periods of increasing and high flow. On at
least two occasions the sludge bed discharged to the launder and plugged the
first stage dual media filters. On both occasions the filters were out of
service for several hours because the standard backwash procedure did not
provide sufficient cleansing. Based on this experience, the operators would
often dump several pounds of polymer directly into the raw sewage wet well in
order to control a rising sludge blanket.
During the 11 month period that lime was used, along with ferric chloride
and polymer or with polymer alone, the problem of floe carry over was greatly
reduced. The lime-ferric chloride combinations varied in the pH range of 9 to
10.5 and from 4.5 to 45 mg/1 Fe. For the 30 day period November 7 through
December 6, 1974 the clarifier was maintained at a pH of approximately 11.4 and
polymer used as a coagulant aid. Solids removal in the clarifier and through
the plant was consistently high; however, it was not always possible to main-
tain the effluent total phosphorus below 1 mg/1.
70
-------
i CO
r CO
i
71
-------
The effluent BOD values demonstrated a high degree of variability because
of the number of operating modes that were tried without regard for the past
history of the carbon columns. It was necessary to alternate between upflow
and downflow operation to diagnose more completely problems with the upflow
distributors. Because of problems with the carbon furnace the first success-
ful carbon regeneration was not made until November of 1974 and, thus, partially
spent carbon was utilized in all columns for most of 1974.
No significant ammonia removals were accomplished until April and May 1975
when the contractor started to test the zeolite columns and the regeneration
system.
B. DURING DEMONSTRATION PROJECT
1. General
Throughout the period of the demonstration project the Rosemount AWTP was
operating under the authorization of NPDES permit MN0025488 which was issued
by the Minnesota Pollution Control Agency on June 2, 1975. The permit con-
ditions are summarized in TABLE 28.
TABLE 28. DISCHARGE PERMIT CONDITIONS
7 day
Effluent Characteristic average
30 day average
Normal Experimental
Notes
BOD - mg/1
SS - mg/1
Fecal Coliform per 100 ml
pH
400
15
15
200
25
30
200
arithmetic mean
arithmetic mean
MPN
range of 6.5 to 8.5 for instantaneous values, not
subject to averaging.
no discharge of floating solids or visible foam in
other than trace amounts.
discharge shall not contain oil or other substances
in amounts sufficient to create a visible color film
on the surface of the receiving waters.
72
-------
The two sets of BOD and SS standards were provided to accommodate the
experimentation during the demonstration project. In terms of effluent quality
the plant operations staff had three goals when the project was initiated as
fo11ows:
(a) to meet the discharge permit conditions;
(b) to maintain the monthly average total phosphorus concentration
of 1.0 mg/1;
(c) to maintain the monthly average ammonia nitrogen concentration
of 1.0 mg/1.
The overall plant performance during the two year period is characterized
by the data summaries presented in Figure 19 and TABLE 29. Individual BOD
values were less than or equal to 15 mg/1 and 25 mg/1 54 and 89 percent of
the time respectively. Individual suspended solids values were less than or
equal to 15 mg/1 99.7 percent of the time. In like manner the effluent total
phosphorus goal of 1 mg/1 was exceeded only approximately 17 percent of the
time.
The BOD removal efficiency could have been improved by regenerating carbon
at more frequent intervals; however, one of the purposes of the evaluation was
to document the effect of service time and/or organic loading on carbon
performance.
After several months of operation it became obvious that the goal of
1 mg/1 NH3-N was not attainable. The new endeavor was to keep the ammonia
removal system in service as much as possible and to optimize removal
efficiencies with the existing facilities.
Most of the laboratory analytical data that relate to process performance
are summarized in TABLE 29. Although these data are difficult to interpret
because the sampling frequencies were not identical throughout the two year
period several observations can be made as follows:
[a] Overall removal efficiencies averaged approximately 93%, 99%, and
94% for BOD, suspended solids and total phosphorus respectively.
[b] Based on average raw and clarifier flows of 0.242 mgd and 0.298 mgd
respectively the removals accomplished in the clarifier averaged
approximately 75%, 90%, and 90% for BOD, suspended solids and
total phosphorus respectively.
[c] With the exception of iron all measurable amounts of the heavy metals
analyzed were removed in the clarifier.
[d] The two stages of filtration accounted for approximately 5% of the
BOD and suspended solids removal.
73
-------
en
£
i
O
0.1 _______ __^_____
2 TO ZD TO 50 80" ~90
PERCENT OF TIME VALUES LESS THAN OR EQUAL TO
Figure 19. Frequency distributions for effluent paramaters
during demonstration project.
98
74
-------
TABLE 29. SUMMARY OF ANALYTICAL DATA
Parameter
BOD - total , mi/1
BOD - soluble, mn/1
COD - total , mq/1
COD - soluble, mg/1
TOC - total , mg/1
TOC - soluble, my/1
Suspended Solids, mq/1
Turbidity, NTU
Settleable Solids , ml/1
Temperature, "F
PH
Alkalinity, mg/1 CaC03
Ch 1 on de , mq/1
Total Solids, mq/1
Total Volatile Solids, mq/1
Total P, mg/l
Total Soluble P, mq/1
Total Ortho P, mg/1
Soluble Ortho P, mq/1
Ammon i a-N , mq/1
Kjeldahl-N, mg/1
Hardness , mg/1 CaC03
Calcium, mq/1 CaC03
Potassium, mg/1
Sulfide, mg/1 H2S-S
Dissolved Oxygen, mg/1
Chlorine Residual, mg/1
Fecal Coliform per 100 ml -
Copper, mg/1
Nickel , mg/1
Lead, mg/1
Zinc, mg/1
Cadmium, mg/1
Chronn urn , mq/1
I ron , mg/1
Average Concentrations
RAW
211
60
505
143
104
42
237
-
55.5
7.6
427
312
1426
556
11.7
8.5
9.2
8.2
35
42
296
177
17
_
-
-
geometr
0 08
<0.05
<0.1
0.27
<0.02
0.12
1 .43
Clar in
175
47
439
118
89
33
255
-
-
_
7.8
404
352
-
-
9.1
5.8
7.2
5.7
37
38
-
-
18
_
-
-
ic mean
0.07
<0.05
<0.1
0 18
<0.02
0.08
2.3
Clar
41
24
91
62
24
18
19
9
-
_
7 5
160
368
-
-
1.0
0.4
0.5
0.3
33
33
255
206
16
_
-
-
<0.02
<0.05
<0.1
<0 05
<0.02
<0.05
0.81
DMF1
31
26
74
58
21
17
10
7
-
_
7.5
-
_
-
-
_
-
-
_
-
-
_
-
-
trace
-
-
<0.02
<0.05
<0.1
<0 05
<0.02
0 08
0.45
GCC1
18
14
48
40
13
12
5
4
-
_
7.5
-
_
-
-
_
-
-
_
-
-
_
-
-
0.7
-
-
-
-
_
-
-
_
GCC
18
12
38
26
10
8
7
7
-
_
7.5
-
_
-
-
_
-
-
_
32
34
_
-
17
2.2
-
-
<0.02
<0.05
<0.1
<0.05
<0.02
0.06
0.63
DMF2
15
-
36
-
9
-
6
8
-
_
7.6
-
_
-
-
_
-
-
_
31
29
_
190
-
0.6
_
-
<0.02
<0.05
<0.1
<0.05
<0.02
<0.05
0.58
In AEC
_
-
38
-
10
-
2
6
-
.
7.5
-
_
-
-
_
-
-
_
33
33
295
193
14
.
_
-
-
-
_
-
-
.
AEC1
.
-
-
_
-
-
_
3
-
_
7.5
-
_
-
-
_
_
-
.
24
-
298
196
19
_
-
_
-
.
_
-
AEC
_
-
-
_
9
-
3
5
-
_
7.7
221
_
-
-
_
_
-
.
9
11
255
155
12
_
-
<0.02
<0.05
<0.1
<0.05
<0.02
<0.05
0.32
Eff
15
14
40
33
11
10
3
7
0
59.8
7.6
169
397
1327
268
0.7
0.4
0.4
0.3
23
25
260
159
13
6.0
2.2
1.7
<0.02
<0.05
<0.1
<0.05
<0.02
< 0.05
0.49
75
-------
[e] Approximately 10% of the BOD entering the plant was removed by
contacting with the activated carbon.
[f] For those days that the ion exchange system was in service ammonia
removals were substantial and averaged approximately 73%.
Because of the diameter and length of the intercepting sewer, a con-
siderable quantity of raw sewage could be stored upstream of the plant without
adversely affecting sewer users. This storage capacity was used on 55 separate
occasions for a total of 388 hours when the influent gate was closed. Although
this operating advantage may have been used more than absolutely necessary
the above values are at least a semi quantitative evaluation of overall plant
reliability.
2. Mode I
The analytical data generated during this eight week period are summarized
in TABLE 30 and Figure 20. The treatment goals for BOD, suspended solids, and
total phosphorus were met and removal efficiencies averaged 92%, 98%, and 95%
respectively. The removal of total BOD through the carbon columns was quite
low. This was not unexpected because the columns were operated in the upflow
mode and thus discharged suspended solids to the filters. However, the data
indicate an average removal of only 6 mg/1 of BOD through the carbon columns
and second stage filters. This apparent low removal is due to at least two
factors. The concentration driving force for adsorption was low and, perhaps
more important in terms of average values, a plant upset occurred which caused
the BOD concentration of the carbon column effluent to increase substantially.
On June 29, 1975 at approximately 1800 hr the pH meter servicing the pH con-
troller for clarifier effluent was inadvertently left in the standby position
after the pH and reference electrodes were cleaned. This error was not dis-
covered until June 30 at 1100 hr. During the interim the pH downstream of the
clarifier was approximately 10.5. The BOD of the GCC sample for June 30 was
44 mg/1. This extremely high value was no doubt caused by desorption of or-
ganics from the activated carbon under the high pH conditions.
No significant NH^-N removals were accomplished because the zeolite columns
were in service for only 22 days and an efficient method of regenerating the
zeolite had not yet been devised. The data presented in Figure 21 illustrate
the performance of the zeolite columns during period number 1 as described
in TABLE 26. Both columns 4 and 5 were freshly regenerated and operated in
series mode for approximately 52 hours and samples were collected at approxi-
mately 2 hr intervals.
3. Mode II
The Mode II performance data are summarized in TABLE 31 and Figure 22.
The BOD, suspended solids and total phosphorus removals averaged 93%, 98% and
97% respectively. Plant performance did not suffer when the first stage dual
media filters were taken out of service however operational problems were
compounded. For this reason the Mode II configuration was used for only 34 days,
76
-------
TABLE 30. PERFORMANCE DATA SUMMARY - MODE I
Parameter
BOD - total , mq/1
BOD - soluble, mq/1
COD - total , mq/1
COD - soluble, mq/1
TOC - total , mq/1
TOC - soluble, mq/1
Suspended Solids, mq/1
Turbidity, NTU
Settleable Solids, ml/1
Temperature, °F
PH
Alkalinity, mq/1 CaC03
Chloride, mq/1
Total Solids , mg/1
Total Volatile Solids, mq/1
Total P, mq/1
Total Soluble P, inq/1
Total Ortho P, mq/1
Soluble Ortho P, mq/1
Ammonia-N, mq/1
Kjpldahl-N, mq/1
Hardness, mq/1 CaC03
Calcium, mq/1 CaC03
Potassium, mq/1
Sulfide, mq/1 H?S-S
Dissolved Oxygen , ma/1
Chlorine Residual, mq/1
Fecal Coll form per 100/ml -
Average Concentrations
Raw
151
39
333
111
78
'I
193
-
6
55.1
7 7
397
2^
1219
409
11.8
1 1
9.2
o n
2P
34
295
-
-
-
-
-
Geometr
Clar in
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
ic Mean
Clar
21
18
67
47
19
15
9
6
-
-
7 5
79
314
-
-
0.6
0.4
0 3
0.2
24
26
-
-
-
-
-
-
-
DMF1
16
-
57
44
16
K
5
4
-
-
7 r.
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
GCC1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
GCC
14
10
42
26
10
8
7
6
-
-
7.5
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
~
DMF2
10
-
44
-
8
-
5
5
-
-
7.4
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
In AEC
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
~
AEC1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
~
AEC
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
~
Eff
12
8
36
25
10
9
4
5
0
62.0
7.7
75
351
943
127
0.6
0.4
0.3
0.2
22
23
187
-
-
-
6.3
1.2
1.5
77
-------
100
10
o
o
O
o
1.0
0.1
98
10 20 40 60 80 90
PERCENT OF TIME VALUES LESS THAN OR EQUAL TO
Figure 20. Frequency distributions for effluent parameters during Mode I
78
-------
CD
S-
OJ
O)
CD
s_
CD
.c:
2
co
c:
o
o
O)
O
Ol
N
O
-I-J
O
fO
i-
(T3
-C
O
cu
(J
O
O)
O-
CD
L/6ui -
79
-------
TABLE 31. PERFORMANCE DATA SUMMARY - MODE II
Parameter
BOD - total , mq/1
BOD - soluble, mg/1
COD - total , mg/1
COD - soluble, mq/1
TOC - total , mg/1
TOC - soluble, mg/1
Suspended Solids, mg/1
Turbidity, NTU
Settleable Solids, ml/1
Temperature, °F
pH
Alkalinity, mg/1 CaC03
Chloride , mg/1
Total Solids , mg/1
Total Volatile Solids, my/1
Total P, mg/1
Total Soluble P, mq/1
Total Ortho P, mg/1
Soluble Ortho P, mq/1
Ammonia-N, mq/1
Kjeldahl-N, mq/1
Hardness , mg/1 CaC03
Calcium, mg/1 CaC03
Potassium, mq/1
Sulfide, mg/1 H2S-S
Dissolved Oxyqen , mq/1
Chlorine Residual, mg/1
Fecal Coliform per 100/ml
Average Concentrations
Raw
194
53
481
126
98
37
240
-
8
60.4
7.7
413
296
1285
373
12.2
8.8
9/j
8.5
32
40
311
-
-
-
-
-
Goori"t
Clar in
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
ic Mean
Clar
25
20
70
51
17
16
8
4
-
-
7.6
79
305
-
0.4
0.2
0.2
0.1
33
30
-
-
-
-
-
-
DMF1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
'
GCC1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
"
GCC
22
11
42
22
8
8
9
7
-
-
7.5
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
"
DMF2
12
-
32
-
9
-
6
5
-
-
7.5
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
"
n AEC
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
AEC1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
"
AEC
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Eff
13
11
37
28
12
10
5
6
0
65.2
7.7
81
316
1155
128
0.4
0.2
0.3
0.2
26
26
191
-
-
-
6.8
0.8
2.6
80
-------
TOO
10
en
£
O
1I
I
eC
C_J
o
O
1.0
0.1
BOD
SS
98
2 10 20 40 60 80 90
PERCENT OF TIME VALUES LESS THAN OR EQUAL TO
Figure 22. Frequency distributions for effluent parameters during Mode II
81
-------
The data in TABLE 31 indicate significant removals of COD and TOC through
the carbon columns, however, essentially no total BOD removal. Most of the
observed removal of total BOD occurred across the second stage filters. The
data present a second anomaly; the BOD, COD, and TOC values in Eff were all
greater than in DMF2- The sampling system was examined several times during
this period and no malfunctions or cross connections were found.
Although the zeolite columns were not in service during this reporting
period approximately 20% of the influent NH3-N was removed. The removal occurred
downstream of the clarifier and was most likely related to the production of
biomass in the carbon columns and filters. It is possible that some nitrifica-
tion took place in filtered water storage tank No. 2 (Figure 2); however, the
NOo-N concentration of the plant effluent was determined on an irregular basis
ana no significant values were found.
4. Mode III
The concentration of oxidizable organics (BOD, COD, TOC) in the plant
effluent was somewhat higher during this reporting period than in the two
previous periods, however, the raw waste strength was substantially higher
also - see TABLE 32 and Figure 23. The removal efficiencies did not change
substantially compared to modes I and II and were 91%, 98%, and 96% for BOD,
suspended solids and total phosphorus respectively. Substantial removals of
COD and TOC were accomplished across the carbon columns but, as in mode II,
no BOD removal was observed. The observed BOD removal took place in the
second stage filters. The effluent values for BOD, COD, and TOC were again
all higher than the corresponding values for the DMF2 sample.
The zeolite columns were in service for 15 days during the 78 day reporting
period. For those 15 days of operation the ammonia removal efficiency at the
plant averaged 71% - 37 mg/1 in and 10.5 mg/1 out.
5. Mode IV
Plant operation during mode IV differed from previous periods in two
respects; the clarifier was operated at a pH of 9.5 for most of the period and
the carbon columns were operated in the downflow mode. These changes did
not substantially affect plant removal efficiencies as illustrated by the
data in TABLE 33 and Figure 24.
Several modifications were made to the clarifier to improve mixing in the
flocculation zone, however, no performance improvements were observed and, as
a result, the modifications were eliminated.
Several plant upsets occurred because of modifications in the sludge dis-
posal procedures. During the latter part of the summer of 1975 and during the
summer of 1976 the sludge generated at the Rosemount plant was disposed of by
shallow sub-surface injection over an area of several acres adjacent to the
plant. The sludge generated in the clarifier and zeolite regeneration area
was pumped from the sludge storage tank once or twice per week as required.
82
-------
TABLE 32. PERFORMANCE DATA SUMMARY - MODE III
Parameter
BOD - total , mq/1
BOD - soluble, mq/1
COD - total , mg/1
COD - soluble, mq/1
TOC - total , mg/1
TOC - soluble, mq/1
Suspended Solids, mq/1
Turbidity, NTU
Settleable Solids, ml/1
Temperature, "F
pH
Alkalinity, mq/1 CaC03
Chloride , mq/1
Total Sol ids , rnq/l
Total Volatile Solids, niq/1
Total P, nq/1
Total Soluble D, mq/1
Total Ort^o P , mq/1
Soluble Ortho p, mq/1
Ammonia-N, .-la/I
Kjeldahl-N, mq/1
Hardness, mq/1 CaCO-}
Calcium, mq/1 CaCO-j
Potassium, mq/1
Sulfide, mg/1 ll;,S-S
Dissolved Oxygen, mq/1
Chlorine Residual, nq/1
Fecal Coli^orm per 100/ml
Averaqe Concentration
Raw
222
60
f <-j T
W
105
41
325
-
9
', (.
~ £
1?8
?1 "
1293
dpi
1 3.0
Q f.
K.:
Q _K
")A
/>8
31A
-
-
-
-
G"omc tr
Clar in
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
ic Mean
Clar
34
20
7°
r,l
?0
16
16
9
-
7.6
109
580
-
-
0.4
0 2
.1
0 1
34
-
-
-
-
DMF1
22
-
6"
49
I '
15
q
q
-
-
-
-
-
-
-
-
-
-
0 1
-
-
GCC1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
GCC
21
14
5C
30
11
10
9
14
-
7.5
-
-
-
-
-
-
-
-
-
-
-
3 1
-
-
DMF2
14
-
44
-
10
-
7
10
-
7.6
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
In AEC
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
.-
-
-
-
-
-
-
-
-
-
-
-
AEC1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
AEC
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Eff
19
18
47
38
14
13
5
10
0
63.5
7.7
116
412
1150
170
0.5
0.2
0.2
0.1
25
27
223
-
-
-
5.6
1.3
2 8
83
-------
C7>
E
I
O
*t
-------
TABLE 33. PERFORMANCE DATA SUMMARY - MODE IV
Parameter
BOD - total , mg/1
BOD - soluble, mq/1
COD - total , mg/1
COD - soluble, mg/1
TOC - total , mg/1
TOC - soluble, mg/1
Suspended Solids, mg/1
Turbidity, NTU
Settleable Solids, ml/1
Temperature, "F
PH
Alkalinity, mg/1 CaC03
Chloride, mg/1
Total Sol ids , mg/1
Total Volatile Solids, mg/1
Total P, mg/1
Total Soluble P, mg/1
Total Ortho P, mg/1
Soluble Ortho P, mg/1
Ammonia-N, mq/1
Kjeldahl-N, mg/1
Hardness, mg/1 CaC03
Calcium, mg/1 CaCOj
Potassium, mg/1
Sulfide, mg/1 H?S-S
Dissolved Oxygen, mg/1
Chlorine Residual, mg/1
Fees, Coliforn per 100/ml -
Average Concentration
Raw
221
61
521
145
104
42
237
-
7
55.8
7.6
134
315
1464
5R3
12.7
9.2
9.0
8.9
37
45
290
161
18
-
-
-
jo ony* t r
Clar in
175
47
454
117
89
31
283
-
-
-
7.8
411
3bl
-
-
10.2
6.6
8.0
6.4
37
40
-
-
18
-
-
-
c Moan
Clar
46
25
95
64
26
19
22
10
-
-
7.5
184
3/4
-
-
1.1
0 5
0.6
0.5
34
-
-
16
-
-
DMF1
34
25
77
60
21
17
11
7
-
-
~! C
-
-
-
-
-
-
-
-
trac?
-
-
GCC1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
_
-
-
GCC
21
12
38
27
10
8
7
7
-
-
7.5
-
-
-
-
-
-
-
-
32
-
-
-
17
2.4
-
-
DMF2
15
-
34
-
9
-
5
;
-
-
7.6
-
-
-
-
-
-
-
-
32
30
-
196
16
0.6
-
-
In AEC
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
_
-
-
AEC1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
_
-
-
AEC
_
-
-
-
-
-
3
6
-
-
7.8
-
-
-
-
-
-
-
-
10
12
-
166
12
.
-
Eff
15
14
40
33
10
10
3
8
0
59.5
7.7
196
405
1384
293
0.8
0.5
0.5
0.4
25
27
281
181
16
.
5.8
2.4
2.0
85
-------
100
10
CD
I
O
I1
-------
Starting August 1, 1976 the sludge was allowed to settle in the storage tank
(mixer off) and the supernatant recycled to the treatment units. The decanting
procedure substantially decreased the volume of the sludge removed from the
plant.
Because the sludge injection equipment was needed at another plant, the
last injection was accomplished August 24, 1976. After that time the sludge
was hauled to another MWCC facility for ultimate disposal. Between August
24 and October 7, 32,000 gal of supernatant were returned to the plant and
40,500 gal of sludge hauled. Up to that time no difficulty was encountered
because of the additional recycle flow.
On October 10 approximately 7000 gal of supernatant were recycled. An
upset condition was observed in the clarifier starting at 2300 hr on October 10
and continuing through the next day. The sludge bed level in the settling
zone of the clarifier rose dramatically. Approximately 8000 gal of supernatant
were recycled on October 12 and the upset occurred again on the following day.
Similar reactions occurred after supernatant recycle on October 24 and 25.
It was observed that the pH of the stored sludge was in the range of 6 to 7
although the clarifier was operated at pH 9.5 and the sludge from the zeolite
regeneration area had a pH of 11.5 to 12, Evidently the stored sludge was
undergoing anaerobic digestion aid the end or by-product of this reaction caused
the clarifier upsets. The upsets did not occur during the period sludge was
injected; however, the sludge storage tank was completely pumped out during
each injection. During the period that sludge was hauled the sludge storage
tank was never completely emptied and it appears that sufficient seed organisms
remained to ensure a continuous digestion process. For this reason the attempt
to thicken the Rosemount sludge prior to disposal was terminated on October 25,
1976.
During this reporting period no apparent increase in BOD or TOC between
the GCC and Eff samples was observed.
Although the zeolite columns were in service for a total of 123 days or
approximately 30% of the time the removal efficiency for NH3-N averaged only
32%. For the days the columns were in service the removals averaged approxi-
mately 73%. During that period however several regeneration procedures were
evaluated.
6. _MQde_V
The performance data obtained during mode V, TABLE 34 and Figure 25,
demonstrate that the removal of the second stage filters from the process
configuration had little effect on the overall plant performance. The
performance goals were met in terms of suspended solids, BOD, and total
phosphorus removals. The zeolite columns were in service a total of 62 days
or approximately 43% of the time. The zeolite regeneration system was kept
out of service during January and February 1977 because of the energy crisis
in the State of Minnesota. During that two-month period the temperature in
the plant was maintained at 55°F and all nonessential fuel consumption was
eliminated. For those days the zeolite columns were in service the ammonia
removals averaged 79%.
87
-------
TABLE 34. PERFORMANCE DATA SUMMARY - MODE V
Parameter
BOD - total , mg/1
BOD - soluble, mq/1
COD - total , mq/1
COD - soluble, mq/1
TOC - total , mq/1
TOC - soluble, mq/1
Suspended Solids, mq/1
Turbidity, MTU
Settleable Solids , ml/1
Temperature, °F
PH
Alkalinity, ma/1 CaC03
Chloride, mq/1
Total Solids , mq/1
Total Volatile Solids, mq/1
Total P, mq/1
Total Soluble P, mo/1
Total Ortho P, mg/1
Soluble Ortho P, mq/1
Ammonia-N, mq/1
Kjeldahl-N, mq/1
Hardness , mq/1 CaC03
Calcium, mq/1 CaCO^
Potassium, mq/1
Sulfide, mq/1 H2S-S
Dissolved Oxyqen, mq/1
Chlorine Residual, mg/1
Fecal Coliform per 100/ml -
Average Concentration
Raw
204
65
459
151
107
47
214
-
7
50.9
7.6
417
318
1405
547
S.I
5.7
6.4
5.4
32
38
?" o
187
-
-
-
-
Geo"iet
Clar in
175
49
409
118
88
36
199
-
-
-
7.9
391
353
-
-
7.2
4.5
5.4
4.1
-
35
-
-
-
-
-
-
ic Mea
Clar
40
28
100
71
28
21
20
9
-
-
7.3
170
379
-
-
1 .0
0.3
0.4
0.2
-
33
DMF1
32
27
80
65
23
19
9
5
-
-
7.3
-
-
-
-
-
-
-
-
-
-
-
-
-
trace
-
-
-
GCC1
18
14
48
40
13
12
5
4
-
-
7.5
-
-
-
-
-
-
-
-
-
-
-
-
-
0.7
-
-
-
GCC
13
9
31
24
10
8
5
4
-
-
7.4
-
-
-
-
0.4
-
-
-
32
33
260
184
-
1.4
-
-
-
DMF2
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
In AEC
-
-
38
-
10
-
2
6
-
-
7.5
-
-
-
-
-
-
-
-
33
33
295
193
14
-
-
-
-
AEC1
-
-
-
-
-
-
1
3
-
-
7.5
-
-
-
-
-
-
-
-
24
-
298
196
19
-
-
-
-
AEC
-
-
-
-
9
-
2
4
-
-
7.7
-
-
-
-
0.4
-
-
-
7
9
255
147
11
-
-
-
-
Eff
15
14
38
33
12
11
2
6
0
56.8
7.5
184
401
1352
276
0.5
0.3
0.4
0.2
15
17
267
164
-
-
6.6
2.8
1.4
88
-------
100
10
-------
The average concentration of organics in the effluent (measured by BOD,
COD, and TOC) was again greater than the average concentration discharged
from the carbon columns. Because zero is the lower unit and the upper range
is generally not constrained the concentrations of pollutants in wastewater are
usually not normally distributed. The logarithms of the concentration values
are however often normally distributed. (9)(10)(11) The plant effluent data
presented in Figures 19, 20, 22, 23, 24, and 25 indicate that the log normal
distribution is an appropriate description for the concentration data.
The student's t test (12) was used to compare the means of the two sample
populations. The hypothesis ux = uy was tested against the alternative
ux / Uy where Uy is the average value for an effluent parameter and ux is the
average value of the parameter for the DMF? or GCC samples. The type one
error (the probability that Ux = Uy but rejected) was fixed at 5%.
The tests were conducted for the data collected by mode assuming both
normal and lognormal distributions. Because the number of observations for
the Eff sample was usually greater than for the DMF~ or GCC sample the tests
were duplicated for paired data only. The results summarized in TABLE 35 in-
dicate that for modes III, IV, and V, which account for 88% of the demonstration
period, the organic content of the plant flow increased significantly between
the last treatment process and the plant effluent sampling point.
On February 22 and 23, 1977 grab samples were collected to determine if
the composite samples were affected by transport from the point of sampling
to the sample compositing apparatus. The following samples were collected.
[1] effluent - collected at discharge of chlorinator feed pump, see TABLE 4
[2] effluent - collected in sampling room, see Figure 2
[3] GCC - sampled from bottom tap on secondary carbon column
[4] GCC - sampled in sampling room, see Figure 2 and TABLE 4
Sample [2] was transported through approximately 60 ft of 1/2 in polyethylene
tubing and 70 ft of 3/4 in PVC pipe. Sample [4] was transported through
approximately 80 ft of 3/4 in PVC pipe. Grab samples were collected on an
hourly basis and TOC analyses conducted. The results of this test, presented
in Figure 26, indicate that sample transport had no significant effect on the
measured organic content of the samples.
Further linear regression analysis of the data presented in Figure 26
indicate the the TOC values for the GCC and Eff samples are related as
follows:
For samples [2] and [4] Eff - 0.55 + 1.03 GCC
For samples [1] and [3] Eff = 0.71 + 0.99 GCC
90
-------
TABLE 35. COMPARISON OF POPULATION MEANS - Eff vs DMF0, GCC
Hypothesis: ux = uy; Type 1 error = 5%
Uy = average value of parameter for Eff sample
ux = average value of parameter for DMFg sample except
for mode V where GCC sample is usea
Status of Hypothesis
Mode
Distribution Assumed Parameter
I
II
III
IV
Lognormal
Lognormal - Pairs
Normal
Normal - Pairs
BOD accept accept reject accept reject
COD accept accept accept reject reject
TOC no data accept reject accept reject
BOD accept accept reject reject reject
COD accept reject reject reject reject
TOC no data reject reject reject reject
BOD accept accept reject accept reject
COD accept accept accept reject reject
TOC no data accept reject accept reject
BOD accept accept reject accept reject
COD accept accept accept reject reject
TOC no data accept reject accept reject
91
-------
o
o
CIS
z.
II
I
Q-
s:
oo
o
CJ>
O-
^1
oo
u,
o
o
H-
CT)
I
25
20
15
10
GCC Sample Y - 0.95 + 0.96 X
Coefficient of correlation = 0.98
Eff Sample Y = 0.81 + 1.00 X
Coefficient of correlation =0.98
0 5 10 15 20
X - MG/L TOC OF SAMPLE COLLECTED AT PROCESS
Figure 26. Effect of sample transport on concentration of TOC
92
-------
These data demonstrate that the sample transport system did not have a
significant influence on the measured organic content of the sample.
The data, however, do demonstrate that the organic content of both the GCC
and Eff samples varied in the range of 4:1 during a days time. Variations
of this magnitude were observed during several sampling programs and were
apparently related to the carbon column backwash procedure. Prior to March 1,
1977 when one of the carbon columns was taken out of service for backwashing
only one column remained in service and the organic content of the GCC and
downstream samples increased. The apparent increase in concentration of
organics between the GCC and Eff samples may have been due to the variations
described along with some bias in the flow measuring/sample activation systems
93
-------
XI. UNIT PROCESS PERFORMANCE
A. CLARIFICATION
1. Operating Procedures
a. S_1_u_dge Blanket Control
The clarification system was designed with no direct mechanical energy
input to the flocculation zone - Figure 2. However since the number of
collisions between particles is a function of the average velocity gradient,
G, the volume concentration of particles, C, and the basin detention time,
T, the efficiency of flocculation can be enhanced by maintaining a high con-
centration of solids in the flocculation zone and/or maintaining a sludge
blanket in the upflow clarification zone. During the first several months of
operation it was observed that if the flow passed up through an expanded
sludge blanket of approximately 30 inches the solids capture was enhanced and
the effluent turbidity reduced substantially. Experience during that period
also indicated however that control of the sludge blanket was extremely diffi-
cult because of the flow variations experienced at the plant - Figure 13. On
several occasions prior to initiation of the demonstration project the sludge
blanket discharged over the weirs and plugged the first stage filters which
required the entire plant to be shut down for approximately one half of a day.
Because of the above experience and that of others (6)(13)(14) the clarifier
was normally operated without a sludge blanket in the upflow zone.
The bottom of the skirt dividing the flocculation and settling zones -
Figure 2 - is approximately 110 in. below the normal water surface elevation.
A sample tap along with a hose and cable device was used to locate the sludge-
water interface to a depth of 120 in. The plant operators' normal routine in-
cluded locating the interface level on an hourly basis. The distance between
the water surface and the interface was normally maintained at 120 in. If for
some reason the sludge level started to rise some remedial action was taken
which would include adjusting the mixer speed, the sludge blowdown rate or the
clarifier flow rate depending on the rate of change of the sludge level.
b. Mixer Speed
The contractor's representative initially recommended that the mixer speed
be adjusted according to the results of what was identified as the V-|-V2 test.
The V value was defined as the volume of sludge below the solids-liquid inter-
face after 10 min settling in a 100 ml graduated cylinder. It was recommended
that samples be taken at two locations, the bottom of the draft tube at the
level of the mixer and at the top of the draft tube. The recommendations were
94
-------
to maintain the V] value at the bottom of the draft tube at ^ 15 and to maintain
the V2 value at the top of the draft tube at Vi-(3 to 5) by varying the mixer
speed. The V-V tests and speed adjustments were to be made on a semi-hourly
basis. For the lime-ferric chloride combinations used during the demonstra-
tion period this method of control did not appear to affect clarifier perform-
ance but did require continuous adjustment of the mixer speed.
During the period when the lime-ferric chloride combinations were used
the mixer was adjusted to yield visible movement at the water surface near
the center of the flocculation zone. The speed was adjusted down when the
sludge bed moved up for more than a day and when the turbidity of the clarifier
effluent increased. During the two-year demonstration period the mixer speed
was operated in the range of 70 to 155 RPM and averaged 97 RPM. Because of
fouling with rags and other debris the mixer RPM was not always a good indication
of energy input or mixing intensity. The clarifier was drawn down and partially
cleaned 14 times during the two-year period. During each cleaning operation
the mixer was found to be heavily fouled with rags. Although the raw wastewater
was coarse screened at 1/4 in. clear openings a sufficient quantity of rags and
other stringy debris passed through to the clarifier. On several such occasions
it was observed that rags and other debris completely filled the mixing zone
to a depth of approximately 2 ft above the top of the mixer. Similar experiences
are reported in the literature (15).
After a period of approximately 10-15 days after cleaning the mixing in-
tensity and ability to recirculate settled solids was significantly reduced
based on visual observations at the water surface.
c. Sludge Wasting
During the last half of 1974 when lime-ferric chloride combinations were
used in the clarifier it was determined that the sludge wasting rate could be
used to control the solids content of the waste sludge. When the sludge wasting
rate was decreased the solids inventory in the clarifier increased and the blow-
down concentration also increased. In September 1974 blowdown concentrations
as high as 20% solids we*~e obtained. It was also observed however that as the
sludge thickened in the clarifier the rate of solids recycle decreased sub-
stantially causing increased turbidity in the discharge. The thickened sludge
also caused the sludge plow drive to fail on several occasions because of over-
load conditions.
In order to minimize problems with solids recirculation and the possibility
of upsets related to increased microbial activity the sludge wasting schedule
was adjusted to maintain the blowdown concentration in the range of 5 to 10%
solids. During the two-year period the waste sludge averaged 7.44% solids.
This was obtained by wasting an average of 4780 gpd or approximately 12,250 Ibs
solids per million gal raw wastewater treated. The average blowdown rate was
one cycle per 1630 gal and the average blowdown time was 9 sec.
95
-------
2. Chemical Conditioning Alternatives
As previously discussed a number of combinations of lime, ferric chloride
and polymer were used prior to the initiation of the demonstration project.
The ferric chloride and polymer combinations were eliminated from consideration
because of the carryover associated with the normal diurnal flow variations
and because the sludge produced did not thicken beyond 1% in the clarifier or
sludge storage tank.
Jar test data indicated that the low lime process - lime to pH 10.1 plus
polymer - could be used to achieve required phosphorus removals. This
scheme was used several times in the clarifier for short periods (2-3 days)
during the period July 1974 to May 1975. During each attempt the suspended
solids and phosphorus concentrations increased substantially and the low lime
process was thus eliminated from further consideration. The failure to duplicate
the treatment efficiencies obtained in the jar test was most likely related to
the relative mixing regimes and to the fact that sludge was not recycled in
the jars. The jar tests were conducted with apparatus as manufactured by
Coffman Industries, Inc. One liter samples of raw wastewater were dosed with
the lime slurry, rapid mixed at 100 RPM for 5 min., slow mixed at 50 RPM for
15 min., and settled for 15 min. The slow mix speed was sufficient to keep
all solids in suspension. Based on visual observations of the jars and the
liquid surface in the flocculation zone of the clarifier the mixing intensity
was much greater in the jars. The mean velocity gradient, G, for similar
stirring devices has been reported to be approximately 70 fps/ft at room
temperature. (16)(17) It is difficult to estimate the mixing intensity in the
flocculation zone because there is no input of mechanical energy. In February
of 1975 samples were taken at the exit of the mixing and flocculation zones.
One liter aliquots were subjected to further slow mixing at 20 RPM in the jar
test apparatus, G~25 fps/ft (16). The results presented in TABLE 36 demon-
strate that the mixing intensity in the flocculation zone was not sufficient
to provide for efficient flocculation.
TABLE 36. EFFECT OF ADDITIONAL SLOW MIXING ON SOLIDS AND PHOSPHORUS REMOVAL
Slow mix time *
Exit
Exit
Sample
mixing zone
flocculation zone
min
0
5
10
20
0
5
10
20
Turbidity
8.9
4.0
3.0
2.5
20
8.5
6.7
5.1
Ortho P
0.37
0.35
0.28
0.28
0.34
0.28
0.24
0.24
All jars settled 20 min. prior to sampling
96
-------
Jar tests indicated that floe, assumed to be Mg(OH)2> formed at a pH
value of approximately 11.1 to 11.3 and that good clarification could be ob-
tained with high lime treatment. The jar tests indicated that a lime dose
of approximately 510 mg/1 CaO would be required to obtain a pH value of
11.1. For a 30-day period during November and December 1974 the high lime
process was used. The average of approximately 700 hourly pH values was
11.4 (the average pH was somewhat higher). In order to maintain a high
degree of solids capture approximately 2.7 mg/1 of Nalco 677 was added in
the mixing zone. The clarifier performance data for that period were as
follows: suspended solids 3.9 mg/1, turbidity 4.0 NTU, BOD 44 mg/1 and total
ortho phosphorus 0.7 mg/1. Although the solids removal efficiency was more
than adequate the removals of phosphorus and BOD were less than anticipated.
The process was eliminated from further consideration because of the extremely
high lime dose required - approximately 830 mg/1 CaO. This was significantly
higher than anticipated from the jar test data and may have been related to
the solids inventory in the clarifier (14) and to incomplete dissolution of
the lime in the clarifier mixing zone. As indicated in Figure 2, the lime
slurry addition point was slightly below the plane of the mixing impeller.
It is possible that a significant fraction of the lime slurry was not mixed
with the incoming flow but rather settled into the sludge zone directly.
Menar and Jenkins (18) recommended that ferric chloride or alum be used
to coagulate precipitated calcium phosphate when the pH is less than 10.0.
Stamberg et al (19) found that total phosphorus residuals exceeded 2 mg/1 at
pH values less than 11.5 unless a supplemental flocculant such as ferric
chloride was used. Other investigations (15) (20) have demonstrated high
removal efficiencies for phosphorus, suspended solids and organics when using
lime and ferric chloride combinations. Based on these results and jar tests
at the Rosemount facility only lime and ferric chloride combinations were
considered for use after December 1974.
3. Removal Efficiencies
a. Overall
The data presented in TABLE 29 and Figures 27 and 28 demonstrate the
average performance characteristics of the chemical clarification process in
terms of removing suspended solids, phosphorus and organics. During this
period the average hydraulic loading to the clarifier was 0.5 gpm/sq ft
and varied in the range of approximately 0.1 to 0.9. Although the character-
istics of the raw wastewater varied substantially even on a weekly average
basis, the characteristics of the clarifier effluent were relatively constant.
This damping effect is more pronounced when daily data are considered as
in TABLE 37 . The removal efficiencies for suspended solids, phosphorus and
organics obtained at Rosemount compare favorably with data reported for
pilot facilities (15)(21).
It is interesting to note that significant fractions of what are herein
termed soluble materials were also removed in the clarification process.
These materials were most likely not soluble but suspensions of colloidal and
supra-colloidal particles. The removal of these finely divided particles was
97
-------
o <
r©~ T
o
^
o
o
Ik-
-a
<<-
! <
<
t
t
^,-4
CO
!Z
O
(13
S-
-I-J
C
01
o
c:
o
o
to
-a
o
CO
O)
-a
c
O)
CL
co
3
LO
OJ
cn
ro
S-
OJ
>
rt)
O)
O)
OJ
i/9w - sanos
98
-------
o <
)
o
c
o
_/
o
o
(
o
(
-tt-
c
o
c
o
c
o
)c
°c
,°
c
0
c
o
o
°0
0
)
b
D
Do
o
D
/I
o
D
c
p,
c-
0
8
0°
o
q
o8
°d
o
c
i
8
o
nc
1
o
D w
o
o
In
0
o
o
0
o
o
°o'
(
3
< <
<
<1
<
<
< "
^
<1
<]
<
<3
<
D «<
<
0 <
<
>
^
<1
<
c:
o
o
a
8
a>
01
(O
5-
O)
S-
1/9W - 003
99
-------
most likely effected by the coagulants through the two mechanisms - adsorption
and entrapment. If the proper mixing conditions could have been established
in the rapid mix and flocculation zones of the clarifier the removal effi-
ciencies of soluble constituents could no doubt have been increased and the
observed variability decreased further. Several investigators have demon-
strated the importance of rapid and complete initial mixing of the coagulant
in terms of achieving efficient use of the applied chemicals. (22)(23).
TABLE 37. VARIABILITY OF RAH AND CLAR PARAMETERS
Raw Clar
mean std. deviation mean std. deviation
ss
COD -
COD -
Total
total
soluble
P
237
505
143
11.7
152
234
36
3.7
19
91
62
1.0
13.5
26
20
0.6
The data presented in TABLE 29 indicate that the heavy metals which were
found to have concentrations greater than detection limits were efficiently
removed in the clarifier. The removals were most likely accomplished through
formation of the sparingly soluble metal hydroxides. Only iron was found to
carry over in significant concentrations and this was most likely related to
the discharge of pinpoint ferric hydroxide floe.
b. pH 10.5
Clarifier operation during the demonstration period can be divided into
two operating modes based on the pH maintained in the mixing and flocculation
zones. During the first six months the set point pH was 10.5. For the follow-
ing 18-month period the set point pH was lowered to 9.5. As previously de-
scribed the lime slurry feed pumps were controlled to maintain constant dosage.
Because of the variable alkalinity of the clarifier influent - due primarily
to recycle flows - constant dosage did not yield good pH control. Better pH
control was obtained by manually adjusting the set point of the pump controller
on an hourly basis as required. The averages of the 24 hourly pH values were
calculated for each day and are presented in Figure 29. Although these are
not average pH values (on a flow proportional basis) the values demonstrate
to a limited degree the consistency of pH control which was one of the im-
portant process parameters.
For the first six month period the average of 4161 hourly pH values was
10.5. During this period the average hydraulic loading of the clarifier was
0.47 gpm/sq ft and the average chemical feed rates were 448 mg/1 lime [CaO]
and 12.4 mg/1 Fe. During the first 22 weeks of the period polymer (Nalco 677)
100
-------
o
NJ
o
ro
oi
cu
i
q-
sl
o
4-
oo
O)
M-
O
QJ
cn
S-
CL)
cn
CM
O)
O CTi OD
o
CTi
CO
re
Q-
Q_
101
-------
was also used at a rate of approximately 2.5 mg/1. Polymer use with the lime-
ferric combinations was terminated on November 16, 1975. The lime addition
point was moved to the influent pipe (Figure 3) during the last week in
September 1975. It was anticipated that the somewhat better mixing condition
would yield more efficient lime utilization and lead to a reduction in dose.
Although no significant change was observed the alternate addition point was
used for the remainder of the study.
The average ferric chloride feed rates, determined on a 28 day basis
varied from 2.6 to 15.1 mg/1 Fe. The clarifier performance, based on the
analyses of the 24 hr composite samples, appeared to be independent of the
dosage over the range observed. It was observed, however, that during periods
of low flow the clarifier was extremely turbid when the average ferric chloride
dose was less than 7 mg/1 Fe. Based on visual observations treatment efficiency
increased as the flow through the unit increased. This apparent contradiction
was most likely caused by increased mixing intensity in both the rapid mix
and slow mix zones due to the increased input of energy.
The average performance characteristics are summarized in TABLE 38. In
the State of Minnesota the normal secondary treatment standards for 30 day
averages are 25 mg/1 BOD and 30 mg/1 suspended solids. Based on this defini-
tion secondary treatment can be nearly obtained by chemical clarification
alone. The range and variability of several clarifier effluent parameters
are presented in Figures 30, 31 and 32. The sludge wasting rate averaged
approximxately 6500 gpd at an average total solids concentration of 5.3%. An
average of 13,200 Ibs solids were wasted per million gallons raw wastewater
treated.
TABLE 38. CLARIFIER PERFORMANCE CHARACTERISTICS AT pH 10.5
Average Concentrations - mg/1
Parameter
Suspended solids
Turbidity (NTU)
BOD - total
BOD - soluble
COD - total
COD - soluble
Total P
Total soluble P
Kjeldahl-N
Raw
270
-
198
53
539
131
12.5
9.3
42
Clar
12
7
28
19
73
50
0.5
0.2
30
% Removed *
94
-
81
56
82
50
95
97
6.5
* based on mass flow rates in raw and clarifier flows
102
-------
1000
I
c/o
o
t/1
Q
LU
Q
UJ
O-
00
t/0
TOO
10
RAW
n= 75
2 10 20 ~ 40""' 60 80 90
PERCENT OF TIME VALUES LESS THAN OR EQUAL TO
Figure 30. Frequency distributions for Suspended Solids data
6/7/75 through 12/5/75.
98
103
-------
en
. CVI
<:
a: c
*
N
ii
en
en
LT>
en
in
CNJ
en
O
s_
o
en
O
Lf>
cr
o;
o
oo
oo
LU
o;
LU
D_
IQ
(O
"3
TD
O
03
S-
o
to
c:
O
S-
LO
-a
o
rs
CT
O)
QJ
S-
cn
O
O
LT>
O
O
o
tn
L/6oi - Q09
104
-------
100
10
en
£E
i
Q-
1.0
0.1
CLAR
n=76
98
2 10 20 40 60 80 90
PERCENT OF TIME VALUES LESS THAN OR EQUAL TO
Figure 32. Frequency distributions for total P data - 6/7/75 through 12/5/75.
105
-------
c. pH 9.5
For the 18-month period December 1975 through May 1977 the average of
11,939 hourly pH values was 9.64. The clarifier hydraulic loading averaged
0.49 gpm/sq ft and chemical feed rates averaged 336 mg/1 lime [CaO] and 19
mg/1 Fe. Based on chemical inventories conducted every 4 weeks the chemical
feed rates varied from 275 to 391 mg/1 for CaO and 14.5 to 22 mg/1 for Fe.
The Fed3 dose was varied as necessary to maintain the total phosphorus
concentration at or below 1 mg/1. The average performance characteristics
are presented in TABLE 39. These data demonstrate a measurable performance
degradation when compared to the data obtained at pH 10.5. The frequency
distributions presented in Figures 30 to 35 illustrate that the treatment
variability was also much greater during the period the pH set point was 9.6.
The ranges for 90% of the observations (5th through 95th percentile) for the
three parameters are presented in TABLE 40.
TABLE 39. CLARIFIER PERFORMANCE CHARACTERISTICS AT pH 9.6
Average Concentrations - mg/1
Parameter
Suspended solids
Turbidity (NTU)
BOD - total
BOD - soluble
COD - total
COD - soluble
Total P
Total soluble P
Kjeldahl-N
Raw
227
-
215
62
494
164
11.4
8.2
43
Clar
22
10
44
26
96
66
1.1
0.5
34
% Removed *
88
-
75
50
76
51
88
92
4.5
* based on mass flow rates in raw and clarifier flows
During this period the sludge wasting rate averaged approximately 4200
gpd at an average total solids concentration of 8.15%. Based on these
vlaues the solids wasting rate averaged approximately 11,400 Ibs per
million gallons raw wastewater treated.
106
-------
oo
Q
O
oo
a
UJ
a
1000
100
10
1
RAW
n=244
CLAR
n=229
2 10 20 40 60 80 90
PERCENT OF TIME VALUES LESS THAN OR EQUAL TO
Figure 33. Frequency distributions for Suspended Solids data
12/6/75 through 6/6/77.
98
107
-------
CTi
L/Bui - QOa
108
-------
I
D-
100
10
1.0
0.1
RAW
n= 225
8
2 10 20 " 40
PERCENT OF TIME VALUES LESS THAN OR EQUAL TO
Figure 35. Frequency distributions for total P data - 12/6/75 through 6/6/77.
109
-------
TABLE 40. RANGE OF CLARIFIER EFFLUENT PARAMETERS 5TH THROUGH 95TH PERCENTILE
Clarifier ph
Parameter
Suspended solids
BOD -
Total
total
P
3
12
0.2
10.5
to
to
to
22
46
0.75
mg/1
mg/1
mg/1
8
24
0.4
to
to
to
9.6
47
78
2.2
mg/1
mg/1
mg/1
4. Mixing Modifications
Several attempts were made to increase the average velocity gradient in
the flocculation zone and thereby increase removals of colloidal and supra-
colloidal particles. In December 1975 a small centrifugal pump was mounted
just below the weir box and clarifier effluent was recycled to the floccula-
tion zone. The flow discharged through a 1/2 in. horizontal pipe tangentially
with respect to the skirt dividing the flocculation and settling zones. The
Q vs H data supplied with the pump indicated a discharge in the range of 30 to
40 gpm however visual observations indicated the actual discharge was consider-
ably less. The discharge did cause a slight rotational motion in the floccula-
tion zone but had no measurable effect on clarifier performance.
Starting in January, 1976 air was injected into the flocculation zone
along with the pumped recirculation. The air was discharged through three
1/2 in. pipes located just inside the skirt and approximately 4 ft below the
water surface. Air mixing did not have an observable effect on clarifier
performance however the concentration of the blowdown sludge increased by over
2% solids within two days after mixing was initiated.
This mixing scheme was used for approximately four months. During that
time the air flow rate was adjusted frequently because of what appeared to
be a gradual plugging of the discharge pipes. On two occasions the turbulence
generated inside the skirt was sufficient to generate a flow through the scum
collector box and to the sludge holding tank. On one of these occasions the
sludge holding tank overflowed. Because of this problem and the fact that
the increased mixing did not measurably affect clarifier performance the mix-
ing modifications were terminated.
5. Costs
a. Chemicals
Clarification was one of two unit processes in the treatment train which
required significant expenditures for chemicals. During the summer of 1977
the delivered unit costs for the chemicals required were as follows:
110
-------
Lime - $0.0329 per Ib CaO
Ferric Chloride - $0.269 per Ib Fe
Sulfuric Acid - $0.079 per Ib S
The unit chemical costs for treatment are summarized in TABLE 41. The chemical
cost during operation at pH 10.5 was approximately 4.5% greater than during
operation at pH 9.6. Polymer costs are not included because it was found that
polymer addition did not improve clarifier performance and was therefore
terminated after 5 months use.
TABLE 41. CLARIFICATION CHEMICAL COSTS - $/106 GAL
Clarifier pH
Lime
Ferric chloride
Sulfuric acid
Total
9.6
92.09
42.38
38.36
172.83
10.6
122.79
27.74
40.04
190.57
b. Labor
The cost accounting system that was utilized divided labor into two
categories, direct and maintenance. Direct labor was defined as time that
the operating staff spend in the appropriate cost center. Typical tasks for
cost center 30 - (Clarification and Phosphorus Removal - TABLE 12) included:
locating the sludge/water interface and checking flow to pH monitors on a
semi-hourly basis, checking chemical feed pumps and making visual inspections
of clarifier at least once per 8 hr shift, making batches of lime slurry and
transferring to slurry storage tank as required, lubricating all drives as
per suppliers' recommendations and draining and cleaning the clarifier as re-
quired. Maintenance labor was defined as time spent by the pipefitter and
the electrician performing the major nonroutine maintenance tasks such as
repair of chemical feed pumps and controllers, modification of chemical feed
lines, repair of motors and drives and cleaning the clarifier influent piping
of debris.
The direct labor requirements for clarification totaled 5816 hours and
averaged 26.8 hr per million gal. The maintenance labor totaled 936 hours
yielding an average of 4.3 hr per million gal. Total labor cost, including
fringe benefits, averaged $265 per million gal.
Ill
-------
c. Supplies
The total expenditures for materials and supplies used for maintaining
equipment plus the cost of contract maintenance totaled $8831. The unit
treatment cost was approximately $40.71 per million gal. This cost includes
repair and replacement of several circuit boards in the chemical feed pump
controllers located in the plant's main control panel.
d. Power
The power consumption data for the drives associated with the clarifier
are summarized in TABLE 42. The values presented are based on wattmeter
readings for those motors which ran 24 hr per day and on watt-hour meter and
on-time clock readings for those which ran intermittently. The average unit
cost of power during the 2 year demonstration period was $0.022 per KWH.
TABLE 42. CLARIFICATION POWER COSTS
Unit
Lime slurry mixer
Lime slurry transfer pump
Lime slurry storage tank mixer
Lime slurry feed pump
Clarifier mixer
KWH/106 gal
40
3
225
61
25
$/106 gal
0.88
0.07
4.95
1.34
0.55
B. FILTRATION
1. Operating Procedures
Shortly after the plant was put into service it became obvious that the
backwash storage pit was much too small to allow for storage of all recycle
flows and discharge to maximize flow equalization. Starting early in 1974
all routine backwashing of filters and carbon columns was accomplished on a
time basis. The normal backwash procedure consisted of a 5-min air scour at
a rate of approximately 5 cfm/sq ft. Each filter in service was backwashed
once per day according to the schedule and more often as required. During
the entire two year period the number of standard backwash cycles averaged
1.25 per day for the first stage filters. The routine backwash requirements
averaged 2.5% and 2.3% of the treated flow for the first and second stage
filters respectively.
On several occasions it was found that the standard backwash procedure did
112
-------
not completely cleanse the filter as evidenced by only a partial reduction of
head loss. On these occasions the filter was taken out of service and a
special backwash procedure used. This procedure consisted of soaking the filter
media in a 0.2N NaCl solution or a chlorine solution (50 Ib HTH) and subjecting
the media to a multiple air scour. The length of the air scour was based on
the operator's judgement. The media was then backwashed at a rate of 16 gpm/sq
ft for 5 to 10 min or until the backwash water was clean based on the operator's
visual inspection. A total of 37 special backwash cycles were conducted, 16
on the first stage filters and 21 on the second stage filters.
The total plant recycle flow associated with filter backwashing averaged
14,800 gpd or approximately 6.1% of the plant influent flow.
2. Removal Efficiencies
The suspended solids and BOD loadings and removal efficiencies for the
first stage filters are summarized in TABLE 43. The filters substantially
reduced both the suspended solids and BOD loading to the carbon columns.
TABLE 43. FIRST STAGE FILTER PERFORWNCE CHARACTERISTICS
Period
Parameter Clar pH = 10.5 Clar pH = 9.5
Suspended solids applied - mg/1
- psf/day
Suspended solids discharge - mg/1
Suspended solids - % R
BOD applied - mg/1
- psf/day
BOD discharged - mg/1
BOD - % R
Soluble BOD applied - mg/1
Soluble BOD discharged - mg/1
12
0.28
6
50
28
.66
19
32
19
20
22
0.55
10
54
44
1.11
34
23
26
26
2 Year
19
0.47
10
47
41
1.02
31
24
24
26
During the six month period that the clarifier was operated at pH 10.5,
the suspended solids and BOD concentrations discharged from the first stage
filters (when in service) were well within the limits defined as secondary
treatment. For this period all of the insoluble BOD was removed by filtration.
113
-------
The suspended solids removal efficiency did not appear to be a function
of the solids loading rate; however, the removal of insoluble BOD decreased
from 100% to 55% when the loading rate was increased.
The suspended solids loading of the second stage filters was extremely
low even during treatment Modes I, II, and III when the carbon columns were
operated in the upfl ow mode. Because of the position in the treatment scheme
only relatively small diameter solids were applied to the second stage filters
and the removal efficiency averaged approximately 30% - TABLES 30, 31, 32,
33. During treatment Modes I through IV the removal of insoluble BOD varied
from 67% to 100% and averaged approximately 75%.
3. Costs
The costs associated with operating and maintaining the two sets of dual
media filters are summarized in TABLE 44. The operator's tasks associated
with the filters included: recording head loss of filter at end of service
cycle, initiating backwash sequence, monitoring position of valves from main
control panel and manually assisting malfunctioning valve positioners as
required, recording head loss of clean filter at start of service cycle,
adjusting sequence controller to provide multiple air scour as required. The
operators also conducted the special backwashing procedures with salt or
chlorine. The maintenance labor (pipefitter and electrician) and materials
and supplies costs were primarily related to the repair and replacement of valves
and operators. The power cost was limited to the cost of electricity to operate
the blowers which provided the air for air scour. Although both sets of filters
operated under conditions of gravity flow the necessary head was provided by
the raw sewage pumps and the carbon column pumps. These pumping costs were
absorbed by cost centers 20 and 50 respectively - TABLE 12.
TABLE 44. FILTRATION COSTS
First Stage Filters Second Stage Filters
Category
Operating labor - hr
Maintenance labor - hr
Benefits
Materials and supplies
Electricity - KWH
Quantity
1,486 10
340 2
2
1
450
$
,187
,976
,632
,033
10
Quantity
1,184
346
-
-
325
$
8,004
2,940
2,189
869
7
Based on total treated
costs averaged $81 per mil
stage filters respectively.
flows of 208 mi
gal and $82 per
1 gal and 1
mil gal for
71 mil gal
the first
the treatment
and second
114
-------
C. CARBON ADSORPTION
1. Operating Procedures
a. Service Time
During the 24-month demonstration project the carbon columns in each of
the two treatment trains were used for extended periods. The length of time
in service for an individual column was primarily based on plant effluent
quality considerations. As the adsorption capacity of the carbon beds was
utilized the organic content (BOD, COD, TOC) of the effluent gradually increased.
When the treatment goal of 15 mg/1 BOD was exceeded for 1 to 2 weeks the carbon
column configuration was generally changed. The column configurations and
their service times are summarized in TABLE 45.
The columns were operated in the upflow mode for approximately 6 months.
During that time and previous upflow periods plugging problems were encountered
with the upflow distribution header and laterals - see Figure 6. Carbon would
migrate into and through the laterals and header. On November 26, 1975 column
number 4 would not take flow. The header and laterals were found to be com-
pletely filled with compacted carbon. No further attempts at upflow were made.
b. Backwash
During the 6-month period the columns were operated in the upflow mode
each column was backwashed once per week. The backwash procedure was as
follows:
function time
drain to air scour level 8 min
stop drain and vent 0.5 min
air scour 30 min
settle bed 5 min
backwash 15 min
The air scour rate was approximately 5 cfm/sq ft and the backwash rate
approximately 12 gpm/sq ft. During the Mode II configuration, first stage
filters out of service, slime accumulated in filtered water storage tank no. 1.
The storage tank was readily cleaned with a salt solution. On September 17, 1975
both columns 5 and 6 were given a special backwash with 800 Ibs NaCl added to
each column just prior to air scour. Visual inspection of the backwash water
indicated that this technique helped to remove debris which had accumulated
in the carbon beds. The removal of soluble orgam'cs did increase after the
special backwash but it was not repeated because the Mode II configuration
was terminated shortly thereafter.
115
-------
00
LU
I
O
o
LU
o
p 1
^>
a:
LU
00
2^
0
CQ
§
.
LO
«^J-
LU
_l
03
h-
C/l
-p->
C
CU
E
E
O
o
3; CU
r^ O
LJU SI
>,
re
*O
C
o
o
0 CU
z oo
f
c
"o to
0 E
o
z:
XX
cu
cu
s-
cu
o
S-
O) CU
00 4->
to
a
LO
r\
en
t~
ft
LO
>^
03
s:
-a
c
£
Q}
CO
C
i
£
cu
2
to
c
E
r
o
o
QL
3
LO
"*
,
LO
1^-
^1
n
^^
CU
c
a
cu
rO
s-
cu
E
CU
CD
eu
s-
>^
r
_C**
CO
s~
4
CD
C
E
3
p
O
o
Q.
CD
LO
O
p~_
LO
^x.
en
n
r
f
,
a>
-i
T3
CU
-M
fO
S-
CU
c~
eu
CD
cu
S-
>^
1
.c:
to
CU
14
"*
c
E
r
o
C_5
Q.
^
1
*
,3-
CM
LO
p^.
en
ft
P
CM
>
O
c
o
JD
i_
rO
O
-C
O
IO J2
i S-
fO
o o
c~
to c
>,
S- CD
CU S-
tO >
CU
-C LO
2 c
0 E
i 3
4 r-~
Q. O
~~^ CJ
C C
2 2
O O
T3 -O
1 LO
3- ^r
LO O
CM CO
LO LO
r^ ^^
en en
r\ n
V.O ^^
CM CO
, .
> O
o cu
Z Q
2
to
s>_
JT
CD
P
-C
tf^
o
cu
CO
3
rO
(J
CU
.a
cu
o
r
>
C
CU
to
c:
i
to
c
r- tO
rO 2
S- O
4_> |
q-
c t-
cu eu
^^
rO 2
eu cu
S- +J
4-J to
to
CM 2
E
2
O
-a
co
A
LO
CM
r^
co
C£>
^^
en
*
px^
1 ,
_Q
cu
u_
CM
C
p
rO
S-
+->
C
0
cu
o
c
to
c
cu
c
p
fO
s;
c
2
o
-a
CO
CM
00
CO
CD
[*">
a>
rt
^iO
CM
f*>
eu
LU
-o
cu
p
rO
S-
cu
c
cu
CD
cu
S-
>)
r
C~
CO
OJ
*+-*
CD
C
E
3
i
O
O
C E
2 2
O O
a -a
LO CD
* LO
en o
CO LO
CD CD
^*^ ^^
en en
ft f.
CM CO
P
c~
U
s- >>
£ £
-o
cu
+->
to
£_
cu
E
CU T3
CD CU
CU -P
S- fO
S-
>> cu
r C
-c: cu
to CD
CU O)
s- s-
M-
>^
CM r
JC
"O tO
c cu
to >
Cp
r-^
co
to
c c
E E
3 3
p (
O O
0 CJ
c c
2 2
0 O
-a -a
CM CO
r CM
3- r^
CD r-»
CD CD
i^x r^>
en en
^ ft
un co
CM CM
CD >
3 0
>^
P
c~
to
CU
s-
tf
CD
T3
C
10
LO
to
£1
E
3
P
O
O
ft
r
C
ra
S-
pj
c
0
cu ~a
o eu
C -4->
fO ra
C i-
eu cu
-i-> c:
c cu
i- CD
re CU
S s_
c
2
o
o
CD
LO
en
fv^
CD
r--
en
^
0
,
o
cu
Q
"O
CU
-pj
tO
i_
cu
c:
eu T3
CD CU
CU -t-5
S- rO
S-
>-> cu
r C
-C CU
to CD
cu cu
S- S-
M-
>^
p p
-C
TD tO
c cu
fO S-
4-
co
CM
to
c c
E E
3 3
p p
o o
CJ O
c: E
2 2
o o
~o -a
r CM
CO r
, 1
co en
co r^
r^^ r*^
en en
"~~ "~
ft ft
CO r
p
_c
(J
CJ S-
CU to
Q s:
116
-------
During most of the approximately 18-month period the columns were operated
in the downflow mode they were backwashed on a daily basis. Even with daily
backwash, hydrogen sulfide generation was encountered until sodium nitrate was
fed to the primary column influent line. Soon after daily backwashing was
initiated observable quantities of activated carbon were found in the backwash
pit. It was determined that the drain down step was not repeatable and that
carbon could be blown out of the column and into the piping system during
the air scour. A manual drain-down scheme was instituted to eliminate this
possibility; however, carbon still accumulated in the backwash pit. The
backwash rate was subsequently lowered to 8 gpm/sq ft. No significant quantities
of carbon were found in samples of backwash water; however, significant
quantities of carbon accumulated in the backwash pit during the remainder of
the project.
c. Hydrogen Sulfide Control
During the first two years of operation (1974 and 1975) H2S generation
in the carbon beds caused odor and corrosion problems in the plant. A
sampling program conducted in October 1974 indicated that the H2$ concentration
in the plant ambient air ranged from 0.15 to 0.5 ppm. The minimum concentra-
tion at which the characteristic odor is preceptible is reported to be 0.13
ppm. (24) During the cold weather months when the plant doors were closed
and outside air did not circulate freely the concentration were considerably
higher based on odor levels. The sulfate reducing bacteria which are
anaerobes appeared to thrive in the carbon beds.
Directo et al. (25) demonstrated that hydrogen sulfide generation can
be controlled by providing N03 as an alternate hydrogen acceptor. They found
that 5.4 mg/1 N03-N in the carbon column feed eliminated HpS production. The
polymer feed system (polymer feed terminated November 16, T975) was modified
to allow the addition of sodium nitrate solution to the carbon column feed
stream on a flow proportional basis. Nitrate feed was initiated on January
20, 1976 and remained in service throughout the remainder of the demonstration
period. The hydrogen sulfide concentration of the GCC sample was determined
on a daily basis when its characteristic odor was detected in the plant and
these data are summarized in Figure 36.
Column 4 was put into service late in November 1975 (TABLE 45) and by
mid January 1976 the HgS level was as high as 12 mg/1. For the period
January 20, 1976 through March 13, 1976 the carbon columns were backwashed
on a daily basis and a sodium nitrate dose of approximately 7 mg/1 N was
maintained in the carbon column feed stream. The HgS-S level averaged 0.5
mg/1. For the two week period starting March 14 the nitrate dose was reduced
to 3.5 mg/1 N and the average FLS-S level increased to 4 mg/1. For the
period April 23 through May 13, 1976 backwashing was conducted on alternate
days (primary column one day, secondary column the following day) and the
nitrate addition averaged 4.2 mg/1 N. The I^S-S level increased substantially
and averaged 6.5 mg/1 during the first two weeks of May. This average is based
on only four analyses; however, odor levels in the plant were significantly
higher than in nrevious weeks.
117
-------
i
,
*
.
.
.
<
,
.
,
*
.
*
s
.
1
.
A ^
»*
*
4
1
,
» .
,
i
'\
. <
4
.
«
r
1
1
i
O)
5: Q-
E
CO
C_3
JANUARY FEBRUARY MARC
1977
concentration of GC
, i -s
° £
^
CO
i C
1 OJ
S en
0
S-
1 .>>
a
1 w
a
g
§ .H
ss
118
-------
During the 12-month period June 1976 through May 1977 the carbon columns
were backwashed on a daily basis and the set point for N03-N addition was
5 mg/1. The actual dose averaged 4.8 mg/1 NC^-N based on metered carbon column
flows and recorded quantities of NaNOs used. During that period the H2S-S
concentration in the GCC sample stream varied from 0 to 16 mg/1 and averaged
1.9 mg/1.
The variable efficiency of sulfide control demonstrated in Figure 36
was related to three factors: failure of the chemical feed pumps, increased
organic load, and operation of the carbon regeneration furnace. The most
common purnp failures were caused by plugging of the filter screen and air
accumulation in the diaphragm and check valves. As soon as a pump failure
was identified the bleed valve was opened and the filter screen cleaned.
Because of its location it was not convenient to check the liquid level in
the sodium nitrate solution tank. As a result pump failures could have
occurred for up to 12 hours prior to detection. On several occasions when
the organic content of the clarifier effluent was significantly higher than
normal the H2$ levels increased. This was most likely related to an increased
oxygen demand exerted by the biornass in the carbon beds.
During November of 1976 it was observed that the h^S level in the GCC
sample stream increased substantially during the period the carbon regeneration
furnace was operated- The organic content of the carbon column influent stream
was normal and the nitrate feed pumps were functioning. When the carbon
column influent was sampled the NC^-N concentration was found to be zero.
The measured N03~N concentration of the sodium nitrate feed solution was found
to be 4800 mg/1^ When one volume of the feed solution was mixed with 99
volumes of the carbon column influent water the measured F^-N concentration
was found to be 0.8 mg/1.
When the carbon furnace was operated, spent scrubber water and quench
water were discharged to the first stage filter influent line. These re-
cycled flows apparently contained a significant quantity of reducing agents
that reacted with the nitrate ion. It was found that this recycle stream
had a five minute chlorine demand of 41 mg/1. In order to eliminate the
reducing agents the clarifier effluent was chlorinated heavily during sub-
sequent carbon regenerations.
2 . Removal Ef f i ci enci es
a. General
The removal efficiencies obtained with activated carbon treatment are
summarized in TABLE 46. For the entire two year period the overall average
removals of soluble organics were in the range of 53 to 55% depending on
the parameter. Likewise overall average removals of total organics were in
the range of 42 to 52%. The data presented in TABLES 29 through 34 indicate
that removals of insoluble organics were substantial when the carbon columns
were operated downflow - Modes IV and V.
119
-------
TABLE 46. SUMMARY OF REMOVAL EFFICIENCIES ACROSS ACTIVATED CARBON COLUMNS
Constituent
BOD - total
BOD - soluble
COD - total
COD - soluble
TOC - total
TOC - soluble
Suspended solids
Mode I
38*
44
23*
41
50*
47
-40
Mode II
52*
45
54*
57
47*
50
-12
Average %
Mode III
36*
30
34*
39
44*
33
-12
Removal
Mode IV
38
52
51
55
52
53
36
Mode V
59
67
61
63
56
58
44
Total
42
54
49
55
52
53
30
* Removal obtained in second stage filters included when carbon columns
operated upflow.
The average soluble COD loading was calculated for each of the 104 weeks
of study period. The mass of soluble COD in a particular flow stream was
defined as the product of the total weekly mass flow (mil Ibs) and the average
of the weekly analyses (usually 3). The loading calculations are summarized
in Figure 37. Although there were no dramatic changes in carbon efficiency
throughout the study the data can be divided into 3 groups. During the first
6 months, when the clarifier was operated at pH 10.5 and the carbon columns
were operated upflow, both the rates of soluble COD application and removal
were relatively low. During the following 12 month period both rates were
slightly higher and during the final 6 months slightly higher still. The
columns were operated in the downflow mode during the latter two periods.
Although no attempt was made to regulate the service times or ultimate
loadings of the individual columns it does appear that both the mode of
operation (upflow vs downflow) and the application rate (concentration) did
affect the removal efficiency of soluble COD.
b. Upflow
Columns No. 4 and 5 were put into service in the upflow mode early in May
1975 - TABLE 45. The upflow mode was utilized until late November 1975. During
that period approximately 10,200 Ibs soluble COD were removed in the carbon beds.
Columns 4, 5, and 6 contained an estimated total of approximately 51,000 Ibs
freshly regenerated activated carbon (the bulk density of the carbon in place was
estimated to be 26 Ib/cu ft) yielding a carbon loading of 0.20 Ib soluble COD per
Ib carbon. Directo et al. (25) reported soluble COD removal efficiencies on the
120
-------
I
\
\
00
O)
CD
OO
O
CL
Q.
fC
Q
O
O
O)
o
CO
ro
o>
N08HV3 Q3iVAIiDV NI Q3AOW3d 003 319mOS JO
121
-------
order of 65% when the loading approached 1.5 lb soluble COD per Ib carbon and
the concentration applied averaged approximately 50 mg/1. The data presented
in Figure 38 illustrate that the removal efficiency decreased significantly as
the carbon loading increased. The average concentration of soluble COD applied
to the carbon was slightly less than 50 mg/1. The extremely low removal effi-
ciency of 11% for the fourth week was related to the pH control problem dis-
cussed in the previous section. Based on the measured parameters plant operation
was normal for the remainder of the period. Because the GCC] sample stream was
not collected until late in the study no attempt was made to determine the
loadings and removals for the individual columns.
c. Downflow
Column 4 was put into service on November 21, 1975 in the upflow mode
and switched to downflow five days later. The soluble COD removals obtained
with columns 4, 5, and 6 in the downflow mode are summarized in Figure 39.
The low removals obtained during the first two weeks were related to a valve
malfunction. After the switch to downflow operation the DMFi and GCC samples
indicated no organic removals; however, samples taken directly from the column
sample taps indicated almost complete COD removal. This anomaly was resolved
on December 9, 1975 when it was discovered that a valve on the carbon column
inlet header (downflow mode) was open allowing essentially all of the flow to
bypass the carbon column. The valve malfunction was not found earlier because
the valve indicator lights on the main control panel showed the valve in the
closed position. The valve actuator was in the closed position but the key
in the mechanical connection sheared and the valve remained open. Unfortunately
the limit switch used to indicate position was mounted on the actuator. In
addition, this valve, along with most others on the carbon columns, was mounted
above the column approximately 30 ft above the floor and was not readily
accessible for inspection.
The removal efficiency dropped below 30% on two other occasions. On both
occasions the pH of the carbon column feed stream was elevated by the return
of zeolite column rinse water to filtered water storage tank No. 2. This
problem is discussed more fully in the following section.
During the 40-week period that columns 4, 5, and 6 were in service the
loading averaged 0.38 Ib soluble COD per Ib carbon. The data presented in
Figure 39 again indicate that the removal efficiency was a function of loading.
Sometime after column 4 was taken out of service it was discovered that
approximately 85% of the column's initial charge of carbon had been lost. It
was never determined over what time period the losses took place but it is
assumed that they occurred during column backwash. Because of this loss the
actual loading was somewhat higher than the calculated value of 0.38 Ib per Ib.
Columns 1, 2, and 3 were utilized for the last 39 weeks of the study. The
soluble COD removal data for this period are summarized in Figure 40. The rate
at which soluble COD was applied to the carbon columns was substantially greater
during this last period because of increased flow and concentration. When
compared to the initial 6-month period the flow increased by approximately 9%
122
-------
QJ
CL
O
Q.
13
CD
s-
rs
ID
c
fO
LO
c:
o
u
o
00
l/l
o
S_
u
ro
(O
>
O
E
O)
O
o
o
O
oo
oo
CO
s_
Z3
cn
123
-------
o
'H
ro
S-
0)
Q-
O
O
CT
C
LT>
r*
<&
i/l
E
o
o
c
o
-Q
(O
O
CA
I/I
O
o
res
fO
>
O
E
OJ
o
o
o
OJ
3
O
CTl
ro
o;
GOO 3iamos -
124
-------
OJ
Q.
O
O
O
T3
cn
13
-a
-a
c:
O
O
S-
ro
O
p
O
O
a>
0
o
ai
-Q
o
GO
a>
S-
125
-------
and the soluble COD concentration by 40%. The higher removal efficiencies
illustrated in Figures 39 and 40 were most likely related to the higher COD
concentrations applied. For the period that carbon columns 1, 2, and 3 were
in service the average carbon loading was 0.40 Ib soluble COD per Ib carbon.
The carbon loadings obtained during this study, 0.20 to 0.40 Ib soluble
COD/lb carbon were substantially lower than those obtained at pilot facilities
in recent years where values in the range of 0.36 to 1.5 have been reported
(25)(26)(27)(37). Because the COD test is not specific it is difficult to
determine why such a wide variation exists. Although it has been demonstrated
that the type of pretreatment can affect carbon performance (30) the variations
are most likely related to the chemical nature of the waste stream and the
physical/chemical nature of the carbon used. The degree to which pure organic
compounds are adsorbed varies substantially. (31) A wastewater system is further
complicated by the synergistic and inhibitory effects that may be produced by
some compounds on others. (32) It is thus not unexpected to find variations
in treatment efficiency from plant to plant but also from day to day within
a plant. This is especially true for small plants such as Rosemount where
relatively small discharges of specific classes of compounds may have a sig-
nificant effect on the performance of the carbon adsorption system.
It is also possible that significant portions of the organic removals
reported here and elsewhere were related to the production of biomass which
further complicates attempts at data comparisons.
3. Costs
The costs associated with operating and maintaining the carbon columns
of trains 1 and 2 are summarized in TABLE 47.
TABLE 47. CARBON ADSORPTION COSTS
Category
Director labor - hr
Maintenance labor - hr
Benefits
Carbon - Ib
Sodium nitrate - Ib
Materials and supplies
Contracted services
Electricity - KWH
Quantity
1,957
120
-
24,000
41,200
-
-
140,500
$
13,452
1,081
2,898
11,520
5,460
1,813
700
3,091
126
-------
The operators' routine tasks consisted of monitoring and recording head
loss across the carbon beds at least once a shift and monitoring the backwash
operation. Nonroutine tasks consisted of removing carbon from the upflow dis-
tributors and headers (during that period when the columns were operated in
the upflow mode) and sampling backwash water for carbon, cleaning carbon and
other debris from the plant backwash pit and loading makeup carbon into the
columns. Maintenance labor consisted primarily of repairing valves and
operators that malfunctioned. The contracted services consisted of installation
of an additional manway in column 4 to facilitate inspection of the underdrain
system. The electricity cost was estimated from the run time for the carbon
column pumps and the blowers used for air scour prior to backwash.
As of November 1977 at least one of the carbon columns requires considerable
repair before it could be put back into service. A reliable cost estimate for
the required repairs was not available at the time this report was prepared.
4. Corros i on Prob1 ems
A previous section alluded to some of the problems associated with FLS
generation including aesthetic considerations and severe corrosion of exposed
copper and brass fittings and electrical gear. The failure of several copper
tubing air lines was attributed to this type of corrosion. (They were re-
placed with polyethylene tubing. ) In November 1976 the process water inlet
valve on one of the second stage filters (valve A, Figure 4) failed to seat
properly. The valve was removed and the disc was found to be covered with a
scale which appeared to be iron oxide. In the process of removing the scale
the pipefitter observed that the disc metal (cast iron) was soft and could be
removed with the fingernail - Figure 41A. A commercial laboratory examined
the disc and found the soft black corrosion product to be approximately 1/8
in. thick - Figure 41B. The results of a chemical examination of the core
material and the corrosion product are presented in TABLE 58. It appears the
disc was undergoing graphitization.(28) No attempt was made to examine other
unprotected metal surfaces that were contacted by process flows containing
hydrogen sulfide; however, in January, 1977 the flow tube used to meter train
No. 1 effluent flow was removed because of obvious malfunction. The unit,
Figure 42, was made up with a steel flange, steel equalizing ring on the
upstream face (with two pressure taps), brass throat with two pressure taps
and a fiberglass diffuser section. Both of the pressure taps on the upstream
face were covered with heavy scale and several additional holes had developed.
In addition the seal between the high and low pressure chambers was corroded
through.
TABLE 48. VALVE DISC ANALYSES
% By Weight
Parameter Core Material Corrosion Product
Carbon
Silicon
Manganese
Phosphorus
Sulfur
3.21
2.15
0.87
0.12
0.09
10.20
5.92
1.26
0.27
1.44
127
-------
(a) Surface
(b) Section
Figure 41. Corroded valve disc.
128
-------
CL
O
01
o
T3
OJ
o
S-
CD
CD
a
CM
O)
en
129
-------
In mid September 1976 the carbon was transferred from column No. 4 to
the dewatering tank - Figure 8. The operators noted that the transfer took
less than half the normal time period. Based on the level in the dewatering
tank it was estimated that only 2 feet of carbon were transferred from column
4 whereas its initial charge was 12.5 ft. When the lower hatch on the
column was opened approximately 1 gal of carbon was found. (The lower hatch
it located below the underdrain system but is positioned in such a manner
that the underdrains cannot be inspected.) It was assumed that the carbon
was lost through one or more underdrains that may have failed. An additional
manway was installed in the column just above the underdrain level. All of
the funnels were removed and the strainers inspected - Figure 43a. All were
intact. It is assumed that the carbon was lost during the backwash procedure.
The initial inspection of the column interior revealed considerable
corrosion damage to the steel. The tank interior was initially painted with
one coat of epoxy primer and two coats of coal tar epoxy as per the construction
specifications. This protective coating had been almost completely removed
in the lower 12 ft of the column. The tank wall was constructed of 3/8 in.
mild steel and some of the pitting reached a depth of 1/8 in. - Figure 43b, c.
The interior welds were also severly corroded. As of November 1977 no decision
has been made on how to proceed with the necessary repair.
5. Carbon Losses
The initial carbon inventory when the plant was put into service was
approximately 109,000 Ibs. The carbon inventory as of October 10, 1977 was
approximately 85,000 Ibs. Approximately 24,000 Ib carbon was purchased and
added to the columns during the approximately four years of operation. Total
carbon losses were approximately 48,000 Ibs or 146 Ibs per mil gal raw waste-
water treated. During the period November 19, 1973 through October 10, 1977,
27 carbon regnerations were accomplished. A literature review indicates
expected carbon losses fall in the range of 5% to 10% per transfer/regeneration
cycle. (2)(6)(13)(29). The average observed loss was 1,777 Ib per regeneration
cycle or approximately 11.4%.
Because of the difficulties encountered, the carbon inventory was not
determined on a regular basis; however, during the demonstration period the
volume of each regenerated carbon batch was determined in the regenerated
carbon storage column (Figure 8) prior to transfer. The losses were thus
documented; however, it was not determined if the losses were related to the
regeneration cycle or to the service cycle. In the case of column 4 previously
discussed it was demonstrated that a major loss was related to the service
cycle. It is estimated that more than 50% of the carbon loss was associated
with the service cycle and the backwashing procedure specifically.
D. ION EXCHANGE AND ZEOLITE REGENERATION
1. General
Early in the investigation it was determined that the treatment goal
of 1 mg/1 NH3-N would not be attainable unless significant changes were made.
130
-------
(a) floor
(b) wall
(c) wall
Figure 43. Interior views of carbon column number 4.
131
-------
The plant design was based on an average NH3~N concentration of 15 mg/1 which
was obtained at the old treatment facility. The two-year average Nh^-N at
the new facility was 35 mg/1 (TABLE 14) and the zeolite columns were over-
loaded when the plant was put into service. At that time two alternatives
were available: reduce the length of the column service cycle or modify the
regeneration procedure. If the service cycle time was reduced more than one
regeneration cycle would be required per day. If the columns were operated
to breakthrough and then regenerated the backwashing schedule for the entire
plant would have been disrupted. If two or more complete regeneration cycles
were required per day the schedule would not be disrupted but the labor re-
quirement, both operating and maintenance, for the regeneration process would
have increased substantially.
Koon and Kaufman (33) demonstrated that the efficiency of regeneration
with a constant brine volume was a function of the sodium concentration and
the pH. Because of the ease of implementation the early modifications to
the regeneration scheme at Rosemount were centered on these two parameters.
The process data however indicated that the regenerant volume would have to
be increased substantially in order to obtain more efficient regeneration. The
latter modifications allowed the regenerant volume to be increased by a factor
of up to 3.3.
2. NH.q-N Removal as a Function of Regeneration Conditions
a. Scheme 1, pH 11 - 4.6 BV - 1 N Na
For 23 days the zeolite regeneration system was operated according to
the operating instructions supplied by the contractor. Two zeolite columns
were in service - one primary and one secondary. At midnight the secondary
column was moved to primary position, the primary column was taken off line
for regeneration and a freshly regenerated column put into service in the
secondary position. The regeneration conditions were as follows: 4.6 BV
brine at pH 11.0 and 1.0 N Na - see TABLE 26. During the first three days
of operation the removals averaged only 10% and it is assumed that some
errors were made in the regeneration procedure. For the remaining operating
days the column influent and effluent concentrations averaged 24 mg/1 NHs-N
and 12 mg/1 NH3-N respectively.
Columns 4 and 5 were both regenerated and put into service at 0800 hr
on June 26, 1975. The influent and effluents were sampled every two hours
until 1200 hr on June 28, 1975. The data obtained are summarized in Figure
44. Initial leakage from both columns was high. The primary column equili-
brated after 118 BV throughput and the secondary column after approximately
237 BV. A total of 73 Ibs NH3-N was removed. The two columns combined con-
tained approximately 27,600 Ibs clinoptilolite, assuming a bulk density of
46 Ibs/cu ft (34). Under the test conditions the effective dynamic capacity
of the zeolite was approximately 0.19 me/gm.
The data obtained on regeneration chemical requirements were not valid
because of the atypical brine losses that occurred - see Section IX G.
132
-------
rtJ
"O
O)
OJ
c:
OJ
en
O)
£-
I
LD
-o
C
(13
O
O
cu
4-)
O
OJ
i-
o
(/I
O)
cn
13
O
i,
it)
2!
JD
o
-------
b. Scheme 2. ph 12.5 - 4.6 BV - 1 N Na
During September and October 1975 a number of full scale regeneration
experiments were conducted. Ammonia elution data were collected during the
regeneration of columns that had equilibrated with a process stream con-
taining between 30 and 40 mg/1 NH,-N. These data demonstrated that regeneration
at pH 12.5 (4.6 BV, 1 N Na) was mdch more efficient than at pH 11. Columns
5 and 6 were in service on alternate days for a total of 28 days. They were
regenerated with 4.6 BV at a nominal pH of 12.5 and 1 N Na. However, prior
to going into service both columns were subjected to multiple regenerations
and the elution data indicated both columns were completely regenerated. For
the first six days of service the column effluents were sampled on a semi
hourly basis. The data for the first four days are presented in Figure 45.
(Heavy precipitation on days 4 and 5 diluted the raw wastewater on days 5
and 6 and the NHo-N concentrations were much lower than normal.) With the
columns completely regenerated the initial leakage was much lower than
previously observed and breakthrough did not occur until approximately 100 BV
passed through the column. After one standard regeneration cycle at pH 12.5
the leakage increased and breakthrough occurred sooner although the average
influent concentration did not change substantially.
Column performance continued to degrade somewhat with successive re-
generations. After the fourth cycle however it appeared as if the changes
in performance were more related to the influent NH^-N concentration and flow.
For the 28-day period the column influent averaged 38 mg/1 NH3-N and the
effluent 12 mg/1 NH3-N. An average of 47 Ibs NH.-N was removed by the zeolite
during each service cycle yielding an average dynamic exchange capacity of
0.24 me/gm.
An average of 800 Ibs NaCl and 138 Ibs Na2C03 were added to the brine
during each regeneration and brine recovery cycle. The weight of NaOH added
for pH control was not quantified because the sight glass was not yet installed
on the caustic storage tank. The average brine conditions are summarized in
TABLE 49.
TABLE 49. AVERAGE BRINE CONDITIONS - SCHEME 2
Brine
Reclaimed
Waste
Desludged
Stripped
PH
12.35
10.3
12.1
12.1
mg/1
Sodium Calcium - CaCQ-\
0.92 N
0.77 N 1,401
54
-
NH3-N
-
520
-
68
134
-------
LT>
CM
IE
O.
-M
fO
-o
cu
-t->
"3
i_
CL)
C
O)
CD
-------
The average sodium concentration of the reclaimed brine was slightly
less than the target value of 1.0 N and the average of the pH values was less
than the goal of 12.5.
Both the reclaimed and waste brine tanks were constructed with hopper
bottoms to facilitate the collection and storage of sludge which was primarily
Mg(OH)2 and CaCOo. The suction lines of the brine pumps were located off the
floor of the tanks and as a result, a minimum of approximately 1700 gal of
brine and sludge remained in the tanks at all times. Based on measured
depths the average brine volumes were as follows: reclaimed - 12,255 gal,
waste - 11,985 gal, desludged - 11,332 gal, and stripped - 11,392 gal. These
volumes and the data of TABLE 49 indicate that approximately 1,225 gm eq wt
of calcium were exchanged during each complete service/regeneration cycle
along with 1,524 gm eq wt NH3-N (48 Ibs). An average of 7,745 mg eq wt of
sodium was lost from the brine during each column regeneration. Thus only
20% of the sodium used was exchanged for NH3-N as follows
ZNH4+ + Na+ ^±r ZNa+ + NH4+
Another 16% of the sodium is accounted for by exchange for calcium. Although
no other cations were determined on a regular basis it appears that somewhat
over 50% of the sodium lost between the reclaimed brine and waste brine states
was wasted due to housekeeping procedures. A significant portion of this
loss was most likely related to the operation of the J-| and Jo valves illus-
trated on Figure 9. The conductivity bridge originally supplied by the con-
tractor did not have the proper range and as a result the valves were activated
manually based on the volume of brine pumped. It is possible that the operators
did not always activate the valves at the proper time.
c. Scheme 3, pH 13 - 4.6 BV - 1 N Na
For 13 days in December, 1975 columns 5 and 6 were in service on alternate
days and regenerated at pH 13. During this period the column influent averaged
35 mg/1 NHo-N and the effluent 6.3 mg/1 NH3-N. An average of 54 Ibs NH3-N
was removed each day yielding a working exchange capacity of 0.28 me/gm.
The average brine conditions are summarized in TABLE 50.
TABLE 50. AVERAGE BRINE CONDITIONS - SCHEME 3
Brine
Reclaimed
Waste
Desludge
Stripped
PH
13.0
12.4
12.3
12.4
mg/1
Sodium Calcium - CaCOs
1.00 N
0.85 N 1,186
83
-
NH3-N
-
888
-
263
136
-------
The average chemical additions per regeneration cycle to maintain these
conditions were as follows: 755 Ibs NaCl, 122 Ibs Na2C03 and 508 Ib NaOH.
Approximately 630 Ibs sodium were added in the three forms. The average
brine volumes were as follows: reclaimed - 12,232 gal, waste - 11,962 gal,
desludged - 11,152 gal, and stripped - 11,130 gal.
The relatively high concentration of NH^-N in the stripped brine was
caused by a leaking pressure relief valve that allowed a significant portion
of the brine to bypass the stripping column.
The removal of Ca++ averaged 2,510 gm eq wt per cycle and the NH^-N
removal was 1,751 gm eq wt. The sodium addition averaged 12,500 gm eq wt;
however, approximately 2,610 gm eq wt were lost in the desludging operation.
Approximately 45% of the sodium added to the system was unaccounted for.
d. Scheme 4, pH 13 - 4.6 BV - 0.5 N Na
The system was operated for a period of 32 days with two columns in
series with one regenerated every 24 hours at pH 13 with 0.5 N NaCl. The
brine strength was decreased in an attempt to reduce sodium losses. Koon
and Kaufman (33) reported that increasing sodium concentrations above 0.2 N
had no effect on regeneration performance. The average brine conditions are
presented in TABLE 51. The average of the reclaimed brine pH value was slightly
higher than the target value of 13.0.
TABLE 51. AVERAGE BRINE CONDITIONS - SCHEME 4
Brine
Reclaimed
Waste
Desludged
Stripped
mg/1
pH Sodium Calcium - CaCO^
13.2 0.54 N
12.8 0.38 N 327
12.8 - 20
12.7
NH-^-N
-
735
-
63
During this period the average NH3-N concentrations for the column
influent and effluent were 32 and 7.1 mg/1 respectively. An average of 52 Ibs
NH3-N was removed each day for an average exchange capacity of 0.27 me/gm.
However, because of problems encountered with the main zeolite column bypass
valve a significant portion of the flow bypassed the columns and it is possible
that the calculated exchange capacity would have increased if all of the flow
had passed through the columns.
The average chemical requirements for brine makeup were as follows:
300 Ib NaCl, 30 Ib Na2CO , and 500 Ib NaOH. A total of approximately 420 Ib
sodium was added per regeneration cycle compared to 630 Ib per cycle when
1.0 N NaCl brine was used.
137
-------
The average volumes for the brine states were as follows: reclaimed -
12,230 gal, waste - 11,350 gal, desludged - 10,670 gal, and stripped -
10,860 gal. Based on these values and the data in TABLE 51 approximately
681 gm eq wt calcium were removed during each cycle. The NH3-N removal
averaged 1,686 gm eq wt. The measured sodium addition amounted to approxi-
mately 8,300 gm eq wt; however, approximately 980 gm eq wt were lost during
desludging. Thus approximately 60% of the sodium was unaccounted for.
The data indicate that the zeolite capacity for NH3-N was not signif-
icantly affected by the reduction of sodium concentration; however, there was
a significant reduction in the observed calcium removed based on the brine
analyses.
Attrition losses of clinoptilolite are reported to increase as the
regeneration pH is increased and reach approximately 0.5% weight loss per
regeneration cycle at pH 12.5 (33). Prior to regenerating the zeolite beds
at pH 13, columns 4, 5, and 6 were drained and the zeolite surface located.
The same measurements were made after a significant number of regenerations
at pH 13 were made. The results are presented in TABLE 52.
On a volume basis it did not appear that significant losses had occurred,
It is possible however, that some weight loss did take place.
TABLE 52. ZEOLITE LOSSES FOR pH 13 REGENERATIONS
Column No.
4
5
6
Number of Regenerations at pH 13
7
14
14
Zeolite Loss-in
- 4
2.5
- 0.5
When the ammonia removal system was first put into service in May 1975
a measurable pH increase was observed across the zeolite columns. When the
regeneration pH was increased to 13, the pH increase was dramatic. The efflu-
ent composite pH was a high as 9.0 and the instantaneous effluent pH was
as high as 12.0. The column effluent pH would reach this high value immediately
after a freshly regenerated column went on line and then gradually decrease.
For the period December 10, 1975 to March 24, 1976 the average of the effluent
pH values (obtained from flow composite samples) was 8.2.
On March 24 an acid neutralization system was put into service to preclude
pH violations (NPDES limit - 8.5) and to eliminate the return of rinse water
to the treatment processes. After a column was regenerated and refilled the
rinse cycle was started. Water was pumped from filtered water storage tank
number 2 through the column, downflow, at a rate of 150 gal/min and dis-
charged in the train 1 side of filtered water storage tank 1 where acid was
added to reduce the pH to approximately 7.5. The pH control equipment on
138
-------
clarifier No. 1 was modified to achieve this task. The neutralized flow was
then pumped through train 1 piping, bypassing the treatment units, and dis-
charged to the train 1 side of filtered water storage tank No. 2. From this
point the flow was pumped directly to the outfall line. It was found that
rinse times on the order of 360 minutes were required to bring the pH of the
column effluent to a value below 8.5.
The above rinse procedure was terminated on July 3, 1976 when the pH con-
trollers serving both clarifiers were modified by the contractor to utilize
the clarifier flow signals as well as pH inputs. This modification was an
attempt to improve pH control in the clarifier (see Figure 29). For reasons
not yet known the contractor and instrument supplier were not capable of
successfully executing the modifications; however, no further attempt was
made to acid neutralize the zeolite rinse waters.
e. Scheme 5, pH 12 - 4.6 BV - 0.5 N Na
The pH 13 regenerations were terminated because it was difficult to main-
tain the effluent pH below the required 8.5 and the cost of the sodium hy-
droxide used for pH control made the high pH regenerations impractical. For
a 14-day period the regeneration pH was lowered to 12 but no other changes were
made in the operating procedures. During this period the NH3-N concentrations
for the zeolite column influent and effluent averaged 26 mg/1 and 11 mg/1 re-
spectively. The NHo-N concentration of the plant effluent however, was sub-
stantially higher than the column effluent for 5 of the 14 days and averaged
13 mg/1. It was obvious that several valves were not seating properly at all
times and as a result, a significant portion of the flow bypassed the columns.
On several occasions both the chloride concentration and conductivity of the
plant effluent increased dramatically during the zeolite regeneration process.
This indicated that brine was leaking to the effluent through supposedly closed
valves.
Because of the problems encountered with the valves it is not possible to
interpret the performance data. At that time, however, it was not possible to
repair or replace the defective valves. The elevation of a considerable
portion of the plant outfall is higher than the plant floor and that portion
contains approximately 450,000 gal. At that time there was no positive shut
off valve to isolate the plant from the outfall and it was not known if the
installed check valves would seal properly. The outfall was drained and a
positive shut off valve was installed June 29, 1976.
f. Scheme 6, pH 12 - 8.2 BV - 0.5 N Na
The brine transfer piping was changed to allow direct transfer from the
waste to the reclaimed tank. This modification facilitated regenerations
with more than the previously fixed 4.6 BV of brine. The columns were operated
for a period of 84 days during which time the average regenerant volume was
8.2 BV. The data obtained during the first 18 operating days are suspect be-
cause of the problem associated with the leaking valves; however, at least one
potential problem was identified. During these 16 days of operation the columns
were rinsed at a rate of 200 gpm (4 gpm/sq ft) in the upflow mode for
139
-------
approximately 2 hrs. The rinse water was discharged to the backwash pit and
recycled to the headworks. On July 22, 1976 a substantial amount of material
that appeared to be zeolite was found in the backwash pit. On the following
day columns No. 4, 5, and 6, which had been used during that period, were
drained and opened. Depth measurements indicated that a total of approximately
115 cu ft of zeolite was lost. Subsequent upflow rinsing was accomplished
at rates<100 gpm (2 gpm/sq ft).
The valve repair program was completed in mid October 1976. The average
column influent and effluent NH3-N concentrations for the remaining 68 days
where 44 mg/1 and 12 mg/1 respectively. Removals ranged from 31 to 89 Ibs
NH3-N per day and averaged 60.4 Ib/day yielding an average exchange capacity
of 0.31 me/gm. The average brine conditions are summarized in TABLE 53 and
the average chemical additions per regeneration were as follows: 480 Ibs
NaCl, 137 Ibs Na2C03, and 127 Ibs NaOH. A total of approximately 320 Ibs
sodium was added per regeneration cycle.
The average volumes recorded for the brine conditions were as follows:
reclaimed - 12,320 gal, waste - 12,090 gal, desludged - 11,610 gal and
stripped - 11,640 gal. Based on these volumes approximately 2,900 gm eq wt
calcium were removed during each cycle and 710 gm eq wt sodium were lost by
desludging. The sodium addition averaged 6,300 gm eq wt per cycle and the
NH^-N removal averaged 1,960 gm eq wt. Thus approximately 88% of the sodium
addition was accounted for.
TABLE 53. AVERAGE BRINE CONDITIONS - SCHEME 6
Brine
Reclaimed
Waste
Desludged
Stripped
pH
12.0
10.0
11.8
11.8
Sodium
0.48 N
0.39 N
-
-
mg/1
Calcium - CaCCh Nhh-N
-
1 ,357 544
76
90
It is reported that for cations normally encountered in domestic waste-
waters the preference of clinoptilolite for ammonium ion is exceeded only by
its preference for potassium ion. (36) If the column influent contains sig-
nificant levels of potassium the capacity for ammonium ion may be reduced
substantially. In such a situation the potassium level in the reclaimed brine
may build up to such a level as to preclude complete sodium regeneration.
During March and April 1977 the potassium concentrations of the process
and brine samples were determined. Prior to initiating the potassium analyses
the brine storage tanks were drained and cleaned and fresh brine made up. The
potassium concentration of the reclaimed brine increased from approximately
140
-------
10 to 250 rng/1 after eight regeneration cycles. After that time the con-
centration varied between 100 and 250 mg/1. This variation was most likely
related to the volume of makeup water required on any particular day. For
a 20-day period after the potassium level increased the concentration varied
from 85 to 250 mg/1 and averaged 169 mg/1.
During this same period the potassium concentration for the INAEC, AEC]
and AEC samples averaged 14.5 mg/1, 19.5 mg/1 and 13 mg/1 respectively. The
net potassium removal averaged 3.6 Ib/day or 42 gm eq wt/day which is
equivalent to approximately 2% of the NH3-N removed. It does not appear that
the nitrogen removals were substantially reduced because of the net potassium
removals.
g. Scheme 7, pH 12 - 14 BV - 0.5 N Na
The regeneration procedure was modified to include a 3rd and 4th pass
of the waste brine after pH adjustment - see Section IV. G.3. The columns
were operated for a total of 35 days using this procedure. The sampling
system was modified to provide a composite sample of the primary column
effluent. The NHg-N concentrations of the primary column influent and effluent
and secondary column effluent averaged 35 mg/1, 23 mg/1 and 4.6 mg/1 respec-
tively. The average flow treated was 123 BV (based on one column) and the
average removal was 71 Ibs/day. The zeolite capacity for these conditions
averaged 0.37 me/gm.
During this period the average regenerant volume was 14.0 BV. The
brine characteristics are summarized in TABLE 54 and the average volumes
for the brine conditions were as follows: reclaimed - 12,234 gal, waste -
12,276, desludged - 11,804 gal and stripped - 11,640 gal. The chemical
additions averaged 450 Ib NaCl, 112 Ib H&2^3 and 310 Ib NaOH per regeneration.
Approximately 68% of the total sodium addition of 400 Ibs was accounted for
in terms of ion exchange for NH4+ and Ca++ and loss during the desludging
operation.
TABLE 54. AVERAGE BRINE CONDITIONS - SCHEME 7
Brine
Reclaimed
Waste
Desludged
Stripped
pH Sodium
12.0 0.49 N
10.9 0.39 N
-
-
mg/1
Calcium - CaCOs NH3-N
-
1,077 730
49
116
141
-------
h. Scheme 8
The ammonia stripping procedure was modified slightly during 27 days of
operation; however, during this period the regeneration and operating procedures
for the zeolite columns were as described in the previous section. The metered
steam application rate for all previous stripping operations was maintained at
1 Ib steam per gal brine. During this period the steam rate was reduced to
0.75 Ib/gal. As expected the NH3-N concentration of the stripped brine in-
creased and averaged approximately 270 rng/1. It was not anticipated that the
increased NH^-N level would have a measurable effect on the regeneration
efficiency because the brine pH was maintained at 12.0 for all four passes.
The NH3-N concentrations averaged 27 rug/1, 15 mg/1, and 3 mg/1 for the
primary column influent and effluent and the secondary column effluent re-
spectively. The flow averaged approximately 112 BV (based on one column).
The calculated zeolite capacity for these conditions is 0.27 me/gm. Unfortunately
both the plant flow and NH^-N concentration were significantly lower than during
the previous period and the data are not easily interpreted.
Based on an isotherm developed in the laboratory the exchange capacity at
equilibrium conditions is decreased by approximately 13% when the NH--N con-
centration is decreased from 35 to 27 mg/1. In addition, at the lower flow
rates the columns were most likely further from a state of complete exhaustion
than during the previous period (Scheme 7). Based on these results it does
not appear that reducing the steam rate by 25% produced a significant effect.
i. Scheme 9
One additional modification was made to the NHg-N stripping procedure.
For a 21-day period during which the zeolite columns were operated as in the
previous section the NH^-N was stripped from the waste brine on alternate
days. The NH3-N concentration of the stripped brine averaged 420 mg/1. The
NH3-N concentration in the reclaimed brine ranged from 350 to 1,070 mg/1 and
averaged 700 mg/1. The pH of the reclaimed brine was again maintained at
12.0 and the ammonia was thus quite volatile. The dissociation constant for
aqueous ammonia solutions is approximately 1.7 x 10~5 at room temperature. (35)
At pH 12 at equilibrium conditions no measurable portion of the NF^-N is in
the form of ammonium ion (NH +) to interfere with the ion exchange process.
During this period the N^-N concentrations of the primary column influent
and effluent and the secondary column effluent averaged 30 mg/1, 23 mg/1, and
6.6 mg/1 respectively. The treated flow averaged 121 BV and the calculated
exchange capacity was 0.27 me/gm. When these values are compared to the results
obtained during operating schemes 7 and 8 it appears that the zeolite performance
was measurably degraded by the high residual NH3-N in the regnerant. Severe
odor problems were also associated with this modification because of the high
concentration of volatile NHs formed in the reclaimed brine.
142
-------
3. Summary of Performance
The data collected during 263 days of operation of the ammonia removal
system are summarized in TABLE 55. For the eight operating schemes considered,
the NH^-N removal efficiency varied from 41% to 89% and averaged approximately
72% while treating approximately 122 BV flow. The exchange capacity and re-
moval efficiency were significantly affected by both the regenerant volume and
pH. It did not appear that the sodium content of the brine had a significant
effect when the concentration was maintained above 0.5N.
The chemical requirements for brine makeup and conditioning varied sub-
stantially for several of the regeneration schemes. Chemical consumption
was reduced substantially by reducing the sodium strength from l.ON to 0.5N
and by conducting multiple pass regenerations at pH 12.0.
Although significant zeolite losses were experienced as a result of
prolonged upflow rinsing at a rate of 4 gpm/sq ft, no zeolite attrition was
directly attributable to high pH regenerations. However, when the regenerations
were conducted at pH values above 11 a significant volume of water was required
to rinse the zeolite to ensure that the pH of the column effluent was less
than 8.5 which was the plant's upper limit according to the discharge permit.
Notwithstanding the permit conditions, the rinse was required to ensure that
the major portion of the ammonia was in the form of NH4+ and thus available
for exchange and removal from the wastewater.
4. Cost
The cost data summarized in TABLE 56 include the charges against the four
cost centers (070, 120, 130 and 140) associated with the ammonia removal system
(see TABLE 12). For the 226 days of operation that fell within the grant
period the total operating cost was approximately $1,425 per mil gal.
Approximately 74% of the costs were related to operating labor plus main-
tenance labor and materials. The operating labor requirement was high be-
cause of the poor reliability of several of the installed automated control
loops and because the modifications to the regeneration procedure required
manual control. The maintenance costs are associated primarily with: piping
modifications, valve repair and replacement and repair and cleaning of the
heat exchangers. The major portion of these costs would not have been incurred
if the facility design was based on information now available. The operating
costs reported are thus not representative of a well design facility treating
the same flow.
The chemical costs represent 14.5% of the recorded operating cost. Both
the sodium chloride and soda ash were purchased by the bag because bulk storage
facilities were not provided at the plant. The average delivered unit prices
for salt and soda ash were $0.0215/lb and $0.069/lb respectively. The unit
cost for caustic was also relatively high, $0.128/lb, because the small storage
tank required delivery of partial loads by tank truck. In addition the total
power cost for the effluent pumps was charged against the ammonia removal
process and further inflated the cost.
143
-------
2>_
r>^
e^
S
ZD
CO
LU
0
^
c£
o
LI-
o;
LU
OL-
^*
LU
1
>
C/)
5
>i
^^
j^
i_i_
fV
^
o
2!
c:
O
1
4-J
rO
CD
E
CD
en
CD
CXL
S-
CD
Q.
CO
_Q
"~~
"^
\
CO
n: CD >>
^ CT>4->
c: !-
CD rO O
CT) f~ rrj
rO O Q-
S- X rO
CD LU O
>
=£
^
O
1 1
L-l_
^~
"^s^
cr
E
*
"^
i
c^
n:
^
4->
C"
(^
^_
CD
CD
O
CD
o:
rO
W
O
n3
^y
co
o
o
c\
rd
,
O
rC
E
CD
^ ^
CD
E
>^
03
-o
^^.^
fv*
Li.
4-^
~"
O
C
^
\
a:
~^_
^^^
CQ
a
c
c
t~
o
CO
o t o o co r^
i i o o o oo r^ CM
UD ^J- CO CO CO CO
CO 0 r- 0 0 «=1-
1 1 O O CM i CM LO
LO LO i CO CO CM
co CM o r-~ CM *d- i
1 CO CM CO CO r O O
1 !~ " ^^ ^^ 1 1
O LO O O O CM O
1 O LO O CO LO CM O
co r^ co ^f ^t1 *d~ ^
CM ^J~ co r*--. i r*^ f r^
i CMCMCMCOCOCMCM
.
o ooooooo
-)c
00 CO CM ^J~ CZ5 CO CM i~~
CTl CTlOOLOCMi CM
CO i VO O CD
*
r* CM CD i CM ^i~ co CD
r r ^~
CTlOOLOCMCOLOr^O
CMCOCOCOCOCOCMCO
LO LO LO LO LO
* .
i i r OOOOO
CD CD CO CO CM
LO
i CMCOCOCMCMCMCVI
* *
* *
i CMCO«3-<£>! ^COCTl
(/I
E CD
3 .(->
i rO
0 S-
O
E
$ £
i E CO
O -r-
O> S- T3
N JO CD
O
a 3
CD rO ~O
CO Ol CD
CO ' . c
rO -Q
Q- i -C
^> "P
J2 LO -I
r*^ ^
> .
O O CO
"4! O rd
4-> ~o
M-
O "O CD
CD 4->
>> O rO
-P 3 C
r- -O S-
4-> CD CD
E S- 4->
rrt ,
*u ^^
3 CD ra
CT 4->
rC E
E S- 0
0 E Q-
5 CD T
C 4-> 4->
3 CO CO
*
X *
* * -X
144
-------
TABLE 56. AMMONIA REMOVAL COSTS
Category
Direct labor -
Maintenance labor -
Benefits
Acid -
Sodium chloride -
Soda ash -
Sodium hydroxide -
Electricity -
Gas -
Materials & supplies
Total
hr
hr
Ib
Ib
Ib
Ib
KWH
MCF
Quantity
4,231
1,853
-
-
163,350
25,690
54,196
209,170
5,315
-
-
$
28,787
15,804
8,917
280
4,327
1,763
6,937
4,603
6,263
16,335
94,016
%
30.6
16.8
9.5
0.3
4.6
1.9
7.4
4.9
6.6
17.4
100.0
E. CARBON REGENERATION
1. Efficiency
Successful carbon regenerations were made as early as October 1974 as
illustrated by the data summary in TABLE 57. Although the temperature con-
trol system functioned in an erratic manner and difficulty was encountered in
maintaining the proper air to gas ratio in the burners, the system was operated
and manually adjusted to regenerate the spend carbon. The values for apparent
density and iodine number of the regenerated carbon indicate that 2 of the 22
regenerations can be rated poor. The regeneration completed December 4, 1974
was conducted without the aid of the apparent density apparatus which was
delivered to the plant the first week in January 1975. During the first 5
regenerations the operators had no means to evaluate the process during
operation. The second poor regeneration, ending March 20, 1975, was caused
by several failures of the burners, induced draft fan and scrubber.
Prior to August 1975 when the variable speed drive was installed on the
screw conveyor (Figure 8) the operator had little control of the regeneration
process except for varying the steam rate and temperature set points. The
temperature control system did not function properly until it was rewired
145
-------
CD
en «3- ro co
-
CM o
LO o
LO IO
en r o CM r-- o
CD en LO CM CO VD LT> LO
CM «d" i O i
O O CTi O i~~
OJ
en
&
en o co o CM ** cocn
i CMi CMCMCM CMCM
o r^. 10 i LO
»* CTi CO CO CM CD CO
«3- ^1-
O in ^ r-* *y
en co o co en
O
Q
LO co co en
i r-- LO ro r--
LO LO LO LO
CD co r^ i co
r--. U3 r-- r--. CM
co co r-. r-. LO
g:_
CD CD CD CD CD
o o o o o
o o o o o
O O O CD CD
CD CD CD O O
O CD O CD CD
CD
CfL
'***'*'^- LOLOLOLOLO LOLOLOLOLD LO^l-O'kO >^)V
I cr> en en cn en cri en en CTI en en en cr» en CTI en C
_
UOOCflD
Q- > U CJ O
146
-------
in January 1976. Prior to that time large temperature excursions were common-
place. For the regeneration ending May 2, 1975 the standard deviations of
the 80 hourly temperature readings were 143°F, 26°F, 209°F, and 196°F for
hearths 1, 2, 3, and 4 respectively. The standard deviations fell in the
range of 20°F to 60°F after the wiring modifications were completed.
Because of the difficulty in identifying carbon batches it was not
possible to make this type of determination. Although it is possible to trace
the movement of carbon batches from column to column through each transfer and
regeneration sequence their integrity was not maintained. Two of the initial
carbon batches were eliminated when they were used to top off other columns
and significant quantities of virgin carbon were also used to top off columns.
When the replacement carbon is taken into consideration the carbon in inventory
as of December 23, 1976 had been regenerated an average of only 3 times. The
COD removal data previously presented do not indicate that carbon age was
significant.
The iodine number is an indication of the capability of an activated
carbon to adsorb low molecular weight substances.(2) Several investigations
indicate that although the iodine number of the carbon decreases with successive
regenerations its capacity to adsorb wastewater organics may increase. (25)(37)
Experience at Rosemount indicates that the iodine number of the regenerated
carbon is more closely related to the state of the spent carbon and the re-
generation conditions than the number of previous regenerations.
2. Losses
When the total carbon loss is divided by the number of regenerations the
average loss is approximately 11.4% per regeneration cycle. As previously
stated however, it is estimated that at least 50% of the loss was associated
with the routine carbon backwash procedures. During the initial regenerations,
significant quantities of carbon were discharged to the backwash pit because
of a design deficiency. The spent carbon storage tank discharged through a
slide gate to a small hopper (volume approximately 1 cu ft) which fed the
screw conveyer. The hopper was partially flooded to maintain the furnace seal.
The water level was controlled by a short section of weir. The dewatered
carbon did not discharge from the storage tank by gravity because of interpar-
ticle bridging. Water was added in the lower section of the storage tank to
maintain a carbon discharge; however, the discharge was not steady. As a
result the small hopper was overloaded and carbon overflowed the weir and
discharged to the backwash pit. Although the drain was covered with a high-
hat screen the losses were significant until the operators developed a feel
for balancing the discharge from the storage hopper and the screw transport
rate.
Because the carbon furnace was used infrequently and was subjected to a
corrosive atmosphere of both H2S and moist activated carbon the startup pro-
cedure involved repair or recalibration of the gas and air control devices.
During one of the initial regenerations the carbon bed caught fire and the
hearth temperature went wild. After the furnace was shut down and the charge
147
-------
rabbled out it was determined that several of the gas/air metering devices
had malfunctioned and allowed considerable excess oxygen in the furnace.
After this experience the startup procedure included the determination of the
oxygen content on the hearths prior to initiating carbon feed and adjusting
the metering devices when the oxygen content increased beyond 2% by volume.
Because of the location of the metering devices the calibration and repair
procedures were extremely time consuming. Fortunately it was determined that
firing the furnace once each month eleiminated the need for the recalibration
procedure and as a result the oxygen levels were no longer determined. It is
possible that a measurable quantity of carbon was burned in the furnace since
no direct column measurements were made before and after regeneration.
During recent regenerations (September 1977) significant quantities of
carbon were found to accumulate in the backwash pit toward the end of the
regeneration. The regenerated carbon is hydraulically transported to an
off-line storage column with an eductor. The carbon is deposited in the
column and water is removed through the underdrain system. This column is
also equipped with an unscreened overflow near the top. It appears that as
the column fills with carbon the water level rises to the overflow level and
carbon which is discharged to the top of the column floats out the overflow.
When this problem is rectified the carbon losses will be minimized.
Although no direct measurements were made it was not felt that significant
losses were related to mechanical attrition. The regenerated carbon when back-
washed did not appear to contain a significant weight of carbon fines. An
attrition loss of 0.3% per cycle is reported by Von Dreusche for movement
through an eight hearth furnace.(39)
3. Cost
A total of approximately 140,000 Ibs carbon were reactivated in the 9
regenerations that took place during the demonstration period. The total
recorded cost was approximately $21,000 as summarized in TABLE 58. The
costs associated with the evaluation of the regeneration in terms of apparent
densities, iodine numbers, etc. and the cost of make up carbon were not in-
cluded in this cost center.
148
-------
TABLE 58. CARBON REGENERATION COSTS
Category Quantity $
Direct labor - mhr 1,311 9,684
Maintenance labor - mhr 565 4,730
Benefits - 2,722
Gas - MCF 1,286 2,060
Materials and supplies - 1,486
Contract maintenance - 298
Total - 20,980
149
-------
XII. COSTS
A. OBSERVED - .25 mgd
The total cost for operating the Rosemount AWTP during the two-year demon-
stration project was approximately $1,005,000. Of this total approximately
$100,000 was directly attributable to the evaluation of performance. The
actual operating and maintenance costs are summarized in TABLE 59. The costs
associated with process modifications are included although they might not
be considered normal 0 & M items. Although sludge hauling costs are included
no charges were allocated for sludge disposal.
TABLE 59. OPERATING AND MAINTENANCE COST SUMMARY
Category
Labor
Chemicals
Materials & supplies
Energy & fuel
Contract maintenance
Sludge hauling
Miscellaneous
Total
$
587,721
60,452
80,296
70,092
64,022
32,325
10,103
905,011
% total cost
64.9
6.7
8.9
7.7
7.1
3.6
1.1
100.0
The total raw wastewater flow during the two-year period was approxi-
mately 177 mil gal and the estimated cost for treatment is $5.11 per 1,000
gal. This unit cost is extremely high because the plant capacity and flow
are low enough to preclude any economy of scale and yet the plant complexity
required full staffing. Assume, for the sake of simplicity, that the costs
for chemicals, sludge hauling and energy and fuel are related to treated flow
on a 1:1 basis. If the plant had operated at its design capacity, 0.6 mgd,
for the two-year period the average unit cost of treatment would decrease by
approximately 49% to $2.62 per 1,000 gal.
150
-------
Salaries, wages, fringe benefits and payment for vacation and sick
leave amounted to approximately 65% of the total 0 & M cost. The recorded
distribution of hours amoung the 17 cost centers is presented in TABLE 60.
Cost centers 10 through 140, which are related to actual treatment units,
accounted for approximately 42% of the direct and 65% of the maintenance
labor times.
More direct labor was expended on clarification than any of the other
treatment processes; however, Lne cos+ centers associated with carbon re-
generation and the ammonia removal systems were much more labor intensive.
The carbon regeneration system, cost center 110, was operated for 103 days
including startup times and required approximately 12.7 mhr per day. The
entire ammonia removal system, cost centers 70, 120, 130, and 140 was in
operation for 226 days during the grant period and averaged 18.7 mhr per day.
The maintenance labor requirements for the carbon regeneration system and
the entire ammonia removed system were significant and averaged 5.5 mhr and
8.2 mhr respectively.
B. PREDICTED - 10 rngd
Estimates for the operating and maintenance costs for a 10 mgd facility
are presented ir. ',-nr following paragraphs. It is assumed that this new in-
dependent physical chemical treatment facility would include the design
modifications discussed in the following sections and would thus be somewhat
more efficient than the Rosemount facility. The following assumptions were
used to construct the cost estimates which were based on December 1977 costs
at Minneapolis-?!. Pau">.
Effluent quality
Clarification
Carbon Adsorption
Ion exchange
BOD - 10, SS - 5, Total P - 1, NH3-N - 3.
lime to pH 10.5, 20 mg/1 Fed 3, neutral ize to pH 7.5
remove 25 mg/1 soluble COD, capacity = 0.4 Ib COD/lb
carbon, add minimum of 5 mg/1 N03-N, carbon losses
limited to 5% per regeneration cycle
remove 15 mg/1 N, zeolite capacity = 0.2 me/gm
regenerate with 20 BV brine
(a) reclaim brine by steam stripping
(b) reclaim brine by biological nitrification as
described by Semmens (39) with return of a
portion of nitrified brine to carbon columns
to satisfy NOg requirement for sulfide control.
151
-------
TABLE 60. LABOR REQUIREMENT FOR COST CENTERS
No.
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
Cost Center
Title
Preliminary Treatment
Influent Pumping Station
Clarification & Phosphorus Removal
Filtration 1
Carbon Adsorption
Filtration 2
Ammonia Removal
Chlorination
Effluent Disposal
Sludge Handling & Disposal
Carbon Regeneration
Clinoptilolite Regeneration
Ammonia Recovery System
Ammonia Disposal
Laboratory Services
Buildings & Grounds
Indirect Services
Labor
Direct
1,666
963
5,816
1,486
1,957
1,184
160
454
178
327
1,311
2,679
1,102
290
6,310
11,945
9,087
- mhr *
Maintenance
149
284
936
340
120
346
168
88
203
341
565
1,609
38
50
492
1,947
396
* does not include vacation or leave time
152
-------
The labor requirements presented in TABLE 60 should not be considered
to be representative of the requirements for a well designed and constructed
treatment facility. For the 10 mgd facility it is estimated that 21 full
time positions would be required as summarized in TABLE 61.
TABLi 61._ESTIMATETJ FULL TJME STAFF FOR 10 MGD FACILITY
Position
Operator
Assistant operator
Operating attendant
Electrician
Pipefitter
Machinist
Chemist
Laboratory technician
Secretary
Superintendent
Number
4
4
4
2
2
1
1
1
1
1
Shift
all
all
all
day
day
day
day
day
day
day
Based on wage and salary rate schedules available December 1977 the total
labor cost would be $453,300 including benefits and leave. The unit cost
would be approximately $0.124 per 1,000 gal.
The observed chemical costs averaged approximately $0.34 per 1,000 gal.
If the ammonia removal system had been in service 100% of the time this cost
would have increased substantially. The observed chemical costs, however,
should not be considered representative because purchases were made in small
lots and no economy of scale was obtained. The estimated chemical costs for
the 10 mgd plant are summarized in TABLE 62 for the two methods of brine
recovery considered.
All of the chemical costs are based on bulk delivery rates available
to the MWCC in December 1977.
It is assumed that materials and supplies, contract maintenance and
miscellaneous expenses would total $200,000 per year or approximately $0.055
per 1,000 gal. The fuel and power cost estimates are dependent on the method
of brine recovery. For the method of steam stripping total annual fuel and
153
-------
and power costs are $989,400 and $391,400 respectively or a total of $0.378
per 1,000 gal. For biological regeneration the estimates are $78,600 and
$409,000 or $0.133 per 1,000 gal.
TABLE 62. ESTIMATED CHEMICAL COSTS - 10 MGD FACILITY
Chemical
Lime
Ferric chloride
Sulfuric acid
Sodium nitrate
Activated carbon
Chlorine
Soda ash
Sodium hydroxide
Total cost $
$/1000 gal
Method of Brine
Steam Stripping
424,900
45,600
136,500
67,200
58,000
34,500
105,100
117,900
989,700
0.271
Recovery
Biological
424,900
45,600
136,500
0
58,000
34,500
198,000
0
897,500
0.246
The cost summary of TABLE 63 indicates that the treatment can be accom-
plished in the range of approximately $0.56 to $0.83 per 1000 gal when the
cost of sludge disposal is not included. Because no sludge disposal cost
data were generated at the Rosemount facility no estimate is included in this
report; however, it is assumed that land disposal of liquid or cake would
be utilized.
154
-------
TABLE 63. COST SUMMARY - 10 MGD FACILITY
Category
Labor
Chemicals
Materials & supplies
Energy and fuel
Contract maintenance
Miscellaneous
Total cost - $
- $/l,000 gal
Method of Brine
Steam Stripping
453,300
989,700
165,000
1,380,800
25,000
10,000
3,023,800
0.83
Recovery
Biological
453,300
897,500
165,000
487,600
25,000
10,000
2,038,400
0.56
155
-------
XIII. DESIGN RECOMMENDATIONS AND CONSIDERATIONS
A. GENERAL
Although the Rosemount AWTP has generally produced a high quality effluent
during its four years of operation the costs associated with several of the
unit processes, in terms of manhours and chemical consumption, have been
relatively high. It does not appear that independent physical chemical treat-
ment will be cost effective in the 0.2-0.6 mgd range; however, larger plants
of this type may be effective both in terms of cost and pollution abatement
if properly designed and constructed. The following recommendations are
based on the four years operating and maintenance experience and thus are
based on relatively long-term observations.
B. SAMPLING AND MONITORING
The sampling system should be included with the process design and not as
an afterthought. Sample valves should be installed both upstream and down-
stream of every process element so that the performance of each carbon column,
for example, can be evaluated and malfunctions identified without delay. All
sample streams should be conducted to a centrally located laboratory where
24 hr flow composite samples can be constructed. Grab samples or continuous
monitors can be used to document diurnal variations and the effects of flow
and loading fluctuations. If continuous monitors are used on a routine basis
they should output to a digital computer system for data storage and retrieval,
rather than to strip chart recorders. Analog data stored on paper are not
^eadily accessible because the charts are cumbersome and the data reduction
process is labor intensive. If the output of monitors is not to be used on
a routine basis it would be better to use a grab sampling program to obtain
time related data as required.
C. CLARIFICATION
The efficiency of the chemical clarification process in terms of chemical
and labor requirements as well as effluent quality should improve by using
the conventional water treatment flow scheme of rapid mix, flocculation
followed by sedimentation rather than a solids contact type device. If lime
is used, the rapid mix chamber must provide sufficient mixing intensity and
detention time to ensure complete dissolution of the sparingly soluble
Ca(OH)2- Mechanical mixers are not recommended as experience indicates that
they foul rapidly with rags and other debris. If grit removal is to be pro-
vided the lime could be added to and mixing accomplished in an erated grit
chamber/s. The ferric chloride addition point should be near the discharge
156
-------
to the flocculation basin.
The flocculation basins should be equipped with variable speed flocculators
and a solids recycle stream from the settling basin to allow for continuous
optimization as the wastewater characteristics change and/or chemical usage
is modified. Solids recycle may help to stabilize the flow with respect to
calcium carbonate precipitation. Calcium carbonate precipitation on the
clarifier launder, which required regular cleaning at Rosemount, and other
structural surfaces should be significantly reduced.
D. FILTRATION
On numerous occasions multiple air scour and water backwash procedures
along with the addition of chlorine or sodium chloride were required to
clean the filter media. On one occasion the surface of the media was hand
raked to remove scum that had developed and was not removed by the backwashing
procedure. Based on these results it is recommended that a surface wash be
incorporated in the filter design and that the rates of both the air scour
and water backwash be controllable along with their durations. It would also
be advisable to provide for the addition of a cleaning agent to the filter
bed as required. Samples of the spent backwash water should also be readily
available and perhaps monitored for turbidity to control backwash volumes.
E. CARBON ADSORPTION
If the carbon column feed stream contains a significant concentration
of sulfate, problems will most likely be caused by the generation of hydrogen
sulfide. This problem must be addressed by controlling the H2S generation,
with NO^ addition for example, or by oxidizing the H2S generated. Cost analyses
will dictate the method of approach for any particular facility.
The cost of replacing activated carbon can be significant and it thus
appears that the design of the contractors should eliminate the possibility
of losses of specification size carbon particles during the service and back-
wash cycles. This would require that both the inlet/s and outlet/s be properly
screened. As with the filters both the rates of air scour and backwash along
with duration should be readily controllable and the backwash water monitored
for solids.
Because moist activated carbon corrodes mild steel rapidly the vessels,
if constructed of mild steel, must be coated. Based on the experience at
Rosemount the reliability of coal tar epoxy paint which has been recommended
(2) is suspect.
157
-------
F. CARBON REGENERATION
The carbon transport, storage, and furnace feed systems must be designed
to eliminate the loss of specification size carbon particles. This can be
accomplished by properly screening all drains and overflows.
Combustion losses in the furnace can be mimized by careful control of
excess oxygen. This can perhaps be best accomplished by direct metering of
both the gas and air streams rather than an aspirator meter-mix device used on
the furnace at Rosemount. If the furnace is not operated on a continuous basis
the start up procedure should include a determination of oxygen levels on all
hearths at operating temperature prior to carbon feed. This requires sampling
ports and an oxygen analyzer that is reliable in the 0 to 2% range.
The combustion air intake should be removed from the carbon handling area
and filtered to preclude the entry of carbon fines to the metering system which
may in turn cause corrosion problems. If hydrogen sulfide is generated in the
plant building it may be advisable to draw combustion air from outside the
building, even if it must be preheated in the winter months. Corrosion problems
were encountered with the fuel and air metering valves at Rosemount prior to
moving the air intake from the floor near the furnace to near the ceiling
of the building.
All of the carbon vessels should be equipped with a device that can
be used to measure carbon depth and thus allow routine inventory determinations.
The devices may vary from a plate and staff moved through a packing gland to
a non contact level monitor. If routine inventory determinations are made,
atypical losses can be identified and corrective action taken before large
quantities of relatively expensive carbon are lost.
G. AMMONIA REMOVAL
Because of the corrosive nature of the brine used to regenerate the zeolite
and the scale forming characteristics of the spent brine the design must faci-
litate scale removal operations and utilize corrosion resistant materials. If
the steam stripping procedure is used to remove ammonia from the brine an acid
cleaning system is required for the heat exchangers. The pressure drop across
the heat exchangers should be continuously monitored and the service cycle
terminated at an inlet pressure somewhat below the rated capacity to preclude
gasket failures.
The interior of the stripping column must be accessible to facilitate in-
spection and cleaning. The fiberglass column at Rosemount was used in a total
of approximately 400 stripping operations or approximately 1600 hr. After that
period of operation the column was severely cracked, Figure 46, and the interior
heavily coated with calcium carbonate scale. The white streaks visible in
Figure 46a were related to leaks that developed near the top of the column.
The brine transfer equipment must be constructed of chemically resistant
materials. They must also, however, be abraision resistant. The brine transfer
158
-------
(a)
(b)
Figure 46. Cracks on outside surface of steam stripping column.
159
-------
pump supplied by the contractor (Gould model 3196) was teflon lined ductile
iron. The lining of the pump casing failed in July 1975 after only minimal
service time and was replaced at a cost of $1,643. The lining on both the
casing and the impeller failed in December 1976 - Figure 47. The pump was
replaced with the identical model constructed of cast Gould-A-Loy 20 and no
further problems were identified.
H. VALVES
Most of the process on/off valves at Rosemount are of the butterfly type.
The valves in the brine area were supplied with stainless steel discs, all
others were cast iron. Neoprene rubber was used for the seat material for
all these valves. After approximately 1 year of operation many of the valves
on the carbon column influent and effluent manifolds started to malfunction.
When the valves were disassembled for repair it was observed that the rubber
seats appeared to be slightly swollen. Because of this the valve operator
did not have sufficient torque to open the valve once completely closed. All
of the carbon column valves were repaired with Hi-Car rubber seats. During
1976 the same problems were encountered with the valves serving the ion ex-
change columns and the neoprene seats were again replaced with Hi-Car rubber.
No problems have been encountered with the new seat material. Because of the
nature of the physical-chemical processes it is not unusual to have several
flow streams carried alternately in the same conduit. It is essential that
the valves maintain drip tight integrity. If this is to be accomplished
the question of material compatibility must be thoroughly addressed.
Because of the frequent backwashing and column sequencing procedures a
treatment facility of this type will utilize a large number of remotely operated
on/off valves. It is advisable that the limit switches, used to provide remote
indication of valve position, be attached to the valve stem or shaft rather
than to the operator. In the event that the coupling between the operator and
the valve fails, this would insure that the remote indicator would provide a
true indication of valve position.
In order to enhance plant reliability and the safety of the operating
and maintenance staff it is essential that all valves be readily accessible.
From an operating viewpoint it would be best to locate all valves within easy
reach from the floor. However, in the event that some valves are located above
structures (filters, carbon contractors, etc.), they must be accessible by
means of properly protected ladders and catwalks.
160
-------
(a)
Impeller
(b)
Casing
Figure 47. Teflon lined brine transfer pump.
161
-------
XIV. OTHER OBSERVATIONS
A. DATA CORRELATIONS
Prior to and during the early months of the demonstration project it
appeared that several parameters, which might be monitored on a continuous
basis, could be used to indicate the performance level of the unit processes
and the plant overall. Linear regression and correlation coefficients were
calculated for parameters which appeared to be related or where good correla-
tion was anticipated. These coefficients, based on two years of laboratory
data are summarized in TABLE 64.
A number of parameter pairs which are often assumed to be related did not
exhibit strong correlation. Several of the raw wastewater characteristics
as well as the performance characteristics of the clarification and carbon
adsorption processes were found to exhibit little dependence on the daily
flow. Based on these calculations it does not appear that the treatment
efficiencies of the processes used were significantly affected by loading
in the ranges observed.
The results of the regression calculations were not entirely negative.
They did indicate that both turbidity and TOC could be used to characterize
the quality of several process streams. If reliable instruments are available
both parameters could be used to monitor performance on a continuous basis.
This might provide potential to improve process control.
B. PERFORMANCE MONITORING
1. General
The original funding agreement for the demonstration project specified
that several parameters be monitored on a continuous basis. The hardware
and sample streams are described in section V of this report. Although the
instruments were not formally evaluated some considerable operating experience
was obtained. The following paragraphs discuss the general performance
characteristics of the hardware used and their applicability at the Rosemount
AWTP and other facilities of the same type.
2. Temperature
No problems were encountered with the temperature monitor during approxi-
mately three years of operation. The total operating and maintenance cost for
the system averaged approximately $200.00 per year including the cost of the
chart paper, periodic calibration checks and semi-annual maintenance of the
162
-------
TABLE 64. SUMMARY OF LINEAR REGRESSION CALCULATIONS
Line
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
! dai
X
Sample Parameter Sample
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Clar
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
ly flow
Q:
Q:
Q:
Q:
Total P
TOC
TOC
COD
COD
COD
COD
SS
Total P
SS
COD
SCOD
Total P
Total SP
SS
SS
SS
SS
in thousand
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Clar
Clar
Clar
Clar
Clar
Clar
Clar
Clar
Clar
Clar
gallons
Y
Parameter
NH3-N
SS
COD
Total P
Total SP
BOD
COD
SCOD
Total P
SS
BOD
Total P
Total SP
SS
COD
SCOD
Total P
Total SP
COD
SCOD
Total P
Total SP
Coefficient of
Line of Best Fit Correlation
Y -
Y =
Y =
Y =
Y -
Y =
Y =
Y -
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y -
Y =
Y =
Y =
Y =
48 - 0
281 -
622 -
18.5 -
3.2 +
34 + 1
65 + 4
119 +
7.1 +
- 40 +
59 + 0
8.6 +
- 0.08
18 + 0
86 + 0
29 + 0
0.9 +
0.3 +
90 + 0
64 - 0
1 - 0.
0.5 -
.05X
0.18X
0.47X
0.02X
0.46X
.71X
.21X
0.047X
0.009X
0.55X
.30X
0.013X
+ 0.53X
.007X
.01X
.23X
0.002X
0.02X
.002X
.008X
0002X
0.0002X
- 0.
- 0.
- 0.
- 0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
- 0.
- 0.
- 0.
29
05
08
30
72
85
75
30
59
84
84
55
82
08
09
42
01
13
01
06
06
10
163
-------
TABLE 64 (Continued)
Line
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
*
**
Sample
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
ar*
ar*
ar*
ar*
ar**
ar**
ar**
ar**
ar
Clar
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
ar
ar
ar
ar
ar
ar
ar
ar
DMF1
DMF]
Cl
Clari
ar
fier
Clarifier
X
Parameter
COD
Total P
SS
SCOD
COD
Total P
SS
SCOD
q:
Q:
Q:
Q:
Q:
Q:
Q:
Q:
Q:
Q:
Turb
Turb
SS
pH = 10.5
pH = 9.5
Sample
Clar*
Clar*
Clar*
Clar*
Clar**
Clar**
Clar**
Clar**
Clar
Clar
Clar
Clar
Clar
Clar
Clar
Clar
Clar
Clar
DMF1
DMF1
DMF
Y
Parameter
Turb
Turb
Turb
Turb
Turb
Turb
Turb
Turb
Turb
SS
COD
SCOD
BOD
SBOD
TOC
STOC
Total P
Total SP
SS
BOD
SS
Li
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Coefficient o
ne of Best Fit Correlation
= 0.6 +
= 3.2 +
= 2.9 +
- 1.5 +
= - 1.3
= 4.0 +
= 4.8 +
= 6.1 +
= 11.4
= 19 +
= 91 +
= 59 +
= 47 -
= 23 +
= 23 +
= 16 +
= 1 .2 -
= 0.7 -
0.87X
8. IX
0.33X
0.11X
+ 0.11X
4.2X
0.22X
0.05X
- 0.008X
0.002
O.OOOX
0.013X
0.02X
0.003X
0.004X
0.009X
0.0008X
0.0009X
= 5.6 + 0.59X
= 28 +
0.40X
- 5 + 0.21X
0.
0.
0.
0.
0.
0.
0.
0.
- 0.
0.
0.
0.
- 0.
0.
0.
0.
- 0.
- 0.
0.
0.
0.
20
25
26
18
47
47
52
18
06
01
00
03
05
01
03
06
07
12
52
18
42
164
-------
TABLE 64 (Continued)
Line
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
Sample
DMF1 *
DMF]*
DMF-j**
DMF]**
GCC
DMF-j
DMF]
DMF]
DMF]
DMF]
DMF]
DMF]
Clar
Clar
Clar
Clar
Clar
Clar
Raw
Raw
Eff
Eff
GCC
X
Parameter
Turb
Turb
Turb
Turb
SS
SBOD
BOD
SCOD
COD
STOC
TOC
SS
Q:
Q:
Q:
Q:
Q:
Q:
Temp, °F
Temp, °F
Temps, °F
Temp, °F
H2S-S
Sample
DMF]*
DMF]*
DMF]**
DMF] **
DMF2
GCC
GCC
GCC
GCC
GCC
GCC
GCC
GCC
GCC
GCC
GCC
GCC
GCC
GCC
% RCOD
GCC
% RCOD
GCC
Y
Parameter
SS
BOD
SS
BOD
SS
SBOD
BOD
SCOD
COD
STOC
TOC
SS
BOD
SBOD
COD
SCOD
TOC
STOC
COD
in GCC
COD
in GCC
Turb
Li
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
ne of Best Fit
= 4 + 0.22X
= 17 + 0.02X
= 3 + 1.03X
- 27 + 0.87X
- 3 + 0.31X
= 8 + 0.086X
= 10 + 0.28X
= 16 + 0.17X
= 23 + 0.19X
= 3 + 0.28X
- 4 + 0.30X
= 5 + 0.14X
= 28 - 0.03X
= 16 - 0.015X
= 54 - 0.05X
= 30 - 0.01X
=11- 0.005X
- 8 - 0.00025X
= - 11 + 0.89X
= 112 - 1.2X
- 12 + 0.43X
- 92 - 0.76X
- 6.3 + 0.4X
Coefficient o
Correlation
0.68
0.03
0.65
0.29
0.55
0.20
0.36
0.31
0.27
0.48
0.49
0.20
- 0.14
- 0.13
- 0.16
- 0.04
- 0.05
0.00
0.26
- 0.26
0.14
- 0.17
0.21
165
-------
TABLE 64 (Continued)
Line
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
Sample
GCC
GCC
Eff
Eff
Eff
Eff
Eff
Eff
Eff
Eff
Eff
Eff
Eff
Raw
All
All
All
All
All
All
All
All
All
X
Parameter
H2S-S
H2S-S
Turb
COD
SS
ss
SS
Turb
Turb
Turb
TOC
TOC
Temp, °F
Temp,°F
SS
SS
SS
Turb
Turb
Turb
Turb
TOC
TOC
Sample
GCC
GCC
Eff
Eff
Eff
Eff
Eff
Eff
Eff
Eff
Eff
Eff
Eff
Eff
All
All
All
All
All
All
All
All
All
Y
Parameter Line of
BOD
COD
TOC
BOD
BOD
COD
TOC
BOD
COD
SS
BOD
COD
COD
COD
COD
BOD
TOC
SS
COD
BOD
TOC
BOD
COD
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
= 19
= 33
= 10
= - 3
= 13
= 36
= 11
= 11
= 32
= 0.4
- 2.7
= 15
= 42
- 14
= 93
= 43
- 27
- 4.1
= 49
= 19
= 13
= - 2
= - 9
Best Fit
+ 0.
+ 1.
+ 0.
+ 0
+ 0.
+ 0.
+ 0.
+ 0.
+ 0.
+ 0
+ 1
+ 2.
- 0.
+ 0.
+ 1 .
+ 0.
+ 0.
+ 0
+ 1.
+ 0.
+ 0.
+ 1
+ 4
14X
56X
17X
.46X
52X
9X
15X
58X
91X
.42X
.15X
21X
04X
47X
03X
37X
15X
.68X
14X
68X
22X
.95X
.8X
Coefficient o
Correlation
0
0
0
0.
0.
0.
0.
0.
0.
0.
0.
0.
- 0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
.05
.38
.15
83
19
18
08
33
29
66
76
78
01
13
79
71
62
43
23
26
16
95
91
166
-------
TABLE 64 (Continued)
Line
90
91
92
93
X
y
Coefficient of
Sample Parameter Sample Parameter Line of Best Fit Correlation
All
All
All
All
TOC
COD
ss
COD
All
All
All
All
SS
BOD
BOD
SS
Y
Y
Y
Y
= - 26 + 2.6X
= 8 + 0.37X
= 47 + 0.33X
- - 35 + O.C3X
0,6;;
0. j4
0.68
0 . 70
recorder by the supplier on a contract basis.
Temperature effects may be significant when heated flows are recycled to
various points in the plant on a regular basis. If the data are to be used
on a regular basis however, they must be transcribed by hand and daily or other
time based averages calculated. Excursions from the average must also be
characterized. Data stored on paper charts are not readily accessible. At.
standard recorder speeds approximately 60 feet of inked charts are generated
each month from each recorder. Based on the experience at Rosemount and other
treatment facilities these data are not fully utilized because of the ex-
cessive cost of transcription. If data are collected with continuous monitor:.
with the intent that they be used on a routine basis a computer based data
acquisition system would be appropriate.
3. JDJH
The pH monitors functioned for approximately three years without
The multipoint recorder was serviced by the supplier on a semi-annual
problem.
basis at
a cost of approximately $100 per year. The cost of chart paper averaged $7C
per year. After an initial period of frequent calibration checks, during which
time it was determined that the instruments were stable, the checks were"made
at 3 to 4 week intervals and required approximately 12 man hours per ve,>sr.
The calibration adjustments fell in the range of 0 to 1.9 pH units and averaged
0.17 pH units.
It appears that continuous pH monitoring of at least several process streams
is essential._ The data produced with strategically located pH monitors can bo
used to identify malfunctions which would allow unwanted streams of acid or
base access to the process flow streams. However, if the data are to be used
they must be readily accessible and the comments of the previous section apply.
4. Conductivity and Chloride
Both the specific conductance and the chloride ion concentration along with
the pH of the effluent stream were used to identify equipment malfunctions and
167
-------
operating procedures which allowed portions of the brine, used for zeolite
regeneration, to enter the effluent stream. The conductivity monitor required
no maintenance other than monthly calibration. The chloride monitor was
calibrated every 2 to 3 weeks and the flow cell was cleaned annually.
The calibration adjustment for the conductivity monitor fell in the range
of 0 to 19% full scale (0-5000 umho) and averaged 3.2% full scale. The re-
quired adjustments for the chloride monitor were substantially greater but
were difficult to quantify because the instrument's logarithmic output (two
decade scale 10-100-1000) was recorded on a linear scale. Because the two
instruments were used to identify the time at which brine entered the effluent
flow stream calibration status was not critical.
Specific conductance can be used only as a semi quantitative measure of
effluent brine contamination because of the relatively high specific conductance
of the OH" ion. If more than a semi quantitative measure of brine loss is
required a chloride ion monitor can be used if calibrated on a daily basis.
5. Total Organic Carbon
The performance characteristics of the process TOC analyzer were not
adequate to allow or justify the use as a routine monitoring device. The
original analyzer was plagued by failures of all major components. The
manufacturer supplied a new instrument in October of 1976; however, the original
infrared analyzer was used to make the C02 measurements. Although the new
instrument was superior to the original device the span drift was significant.
During a six week period of continuous operation, in the 0-50 mg/1 TOC range,
the instrument was calibrated on a daily basis and it was found that, while
the zero drift was -6 4 mg/1 95% of the time, the span drift was 2: 4 mg/1 40%
of the tire. After several unsuccessful attempts were made to have the infrared
analyzer repaired a new infrared analyzer was purchased and installed. Based
on 2 weeks of operation since that time the average daily zero drift and
span drift were 1.5 mg/1 and 0 mg/1 respectively. It thus appears that,
for at least the short term, reliable TOC measurements can be made on a
continuous basis.
The data presented in TABLE 64 and Figure 48 demonstrate that TOC analyses
may be used in lieu of the more time consuming analyses of organics when time
is critical. Although the data presented in Figure 48 were generated with a
laboratory TOC analyzer most process analyzers utilize the same measurement
principle and should yield similar correlations with BOD and COD. MWCC ex-
perience with process TOC analyzers at the Rosemount AWTP and another facility
however indicate that their maintenance costs may be significant and their
use should be justified in terms of process control and/or required monitoring.
Based on the experience obtained during the demonstration period the use of a
process TOC analyzer is not justified for the Rosemount and similar facilities.
6. Turbidity
The turbidimeter functioned adequately during the project however problems
encountered in the sample delivery system kept the instrument out of service
168
-------
c:
cu
S-
o
o
o
"O
c:
"3
Q
O
c:
O)
OJ
O)
-Q
Q.
l/l
c:
o
-------
for prolonged periods. The direct maintenance cost associated with the in-
strument during the demonstration was approximately $300 for the re-
placement of several circuit boards.
Laboratory turbidity measurements provided fair estimates of the suspended
solids in several of the process streams (TABLE 64, lines 27, 29, 41, 44, 46,
and 76). If the same relationships can be obtained with a flow through tur-
bidimeter, use for monitoring process flow streams for estimates of suspended
solids is justified, Because of the rapid response time of instruments of
this type a number of sample streams can be monitored at 3 to 10 min. intervals
as described in section V.C.
7. Ammonia and Phosphorus
Both the ammonia and phosphorus monitors functioned adequately; however,
the instruments did require a significant maintenance effort. Because daily
maintenance was required the monitors were generally operated only 5 days
per week and the total operating time was approximately 6500 hrs. During
that period the cost for consumables and repair parts was approximately $2500
and the labor requirement was approximately 550 man hours.
For a period of 100 operating days both monitors were calibrated on a
daily basis. The zero and span drift for the ammonia monitor averaged approxi-
mately 1% and 10% of full scale respectively when operated in the range of
0-50 mg/1 NH3-N. The zero and span drift for the phosphorus monitor averaged
approximately 3% and 6% of full scale respectively when operated in the 0 to 5
mg/1 P04-P range.
Although the maintenance costs are substantial both ammonia and phosphorus
monitoring may be justified when these compounds are being removed by physical-
chemical means. Both the ammonia and phosphorus data could be used in process
control schemes to ensure compliance with applicable standards and to minimize
operating costs.
The equipment used for monitoring NH3-N and P were standard laboratory
automated chemistry modules (TABLE 9) and were not designed for continuous
monitoring applications. At the time the above modules were purchased monitors
designed for continuous service were available but their performance character-
istics had not been sufficiently demonstrated and the combined cost of separate
NH3-N and P monitors was considerably higher than for the equipment purchased.
It is possible that the performance characteristics and operating cost of
monitoring equipment currently available are superior to those reported above.
170
-------
REFERENCES
1. Dye, L.E., Feasibility Report on The Rosemount Advanced Wastewater
Treatment Plant, Metropolitan Sewer Board, St. Paul, Minnesota, April 1972.
2. Process Design Manual for Carbon Adsorption, U.S. Environmental Protection
Agency Technology Transfer, October 1973.
3. Standard Methods for the Examination of Water and Wastewater, 13th ed.,
American Public Health Association, New York, 1971.
4. Methods for Chemical Analysis of Water and Wastes, U.S.Environmental
Protection Agency, Cincinnati, 1974.
5. Instruction Manual - Potassium Electrode, Orion Research, Inc., Cambridge,
1976.
6. Gulp, R.L. and Gulp, G.L., Advanced Wastewater Treatment. Van Norstrand
Reinhold Co., New York, 1971.
7. Standard Method of Test for Ash in the Analysis Sample of Coal and Coke.
1975 Annual Book of ASTM Standards, Part 26, American Society for Testing
of Materials, Philadelphia, 1975.
8. Standard Method of Test for Volatile Matter in the Analysis Sample of
Coal and Coke, 1975 Annual Book of ASTM Standards, Part 26, American
Society for Testing Materials, Philadelphia, 1975.
9. Viessman, W. Jr., Harbaugh, T.E. and Knapp, J.W., Introduction to Hydrology,
Intext Educational Publishers, New York, 1972.
10. Dean, R.B. and Forsythe, S.L., Estimating the Reliability of Advanced
Waste Treatment - Part 1. Water and Sewage Works, 123 (6):87-89, 1976.
11. Dean, R.B. and Forsythe, S.L., Estimateing the Reliability of Advanced
Waste Treatment - Part 2. Water and Sewage Works 123 (7):57-60, 1976.
12. Beyer, W.H., Handbook of Tables for Probability and Statistics, 2nd ed.
The Chemical Rubber Co., Cleveland, 1976.
13. Physical-Chemical Wastewater Treatment Plant Design. U.S. Environmental
Protection Agency, Washington, D.C., 1974, 41 pp.
171
-------
14. Tofflemire, T.J. and Hetling, L.J., Treatment of a Combined Wastewater
by the low-lime process. Journal Water Pollution Control Federation
45(2), 210, 1973.
15. Parker, D.S., de la Fuente, E., Britt, L.O., Spealman, M.L., Stenquist,
R.J., and Zadick, F.J., Lime Use in Wastewater Treatment: Design and
Cost Data. EPA-600/2-75-038, U.S. Environmental Protection Agency,
Cincinnati, Ohio, 1975, 298 pp.
16. McClellan, T.J., Coagulation and Flocculation of Supra-Colloidal Particles
Contained in a Tailings Basin Overflow. M.S. Thesis, Michigan Technological
University, Houghton, Michigan, 1971.
17. Camp, T.R., Floe Volume Concentration. J. American Water Works Association
60:656, 1968.
18. Menar, A.B. and Jenkins, D., Calcium Phosphate Precipitation in Wastewater
Treatment. EPA-R2-72-054, U.S. Environmental Protection Agency, Washington
D.C., 1972. 96 pp.
19. Stamberg, J.B., Bishop, D.F., Warner, H.P., and Griggs, S.H., Lime Precipita-
tion in Municipal Wastewaters. Federal Water Pollution Control Administra-
tion, Cincinnati, Ohio, 1969, 19 pp.
20. Wuhrmann, K., Objectives, Technology, and Results of Nitrogen and Phos-
phorus Removal Processes. In: Advances in Water Quality Improvement,
E.F. Gloyna and W.W. Edkenfelder, eds. University of Texas Press, Austin,
Texas, 1968 pp 21-48.
21. Kugelman, I.J. and Cohen, J.M., Physical-Chemical Processes. U.S. En-
vironmental Protection Agency, Cincinnati, Ohio, 1972. 9 pp.
22. Stenquist, R.J. and Kaufman, W.J., Initial Mixing in Coagulation Processes.
EPA-R2-72-053, U.S. Environmental Protection Agency, Cincinnati, Ohio,
1972. 161 pp.
23. Vrale, L. and Jorden, R.M., Rapid Mixing in Water Treatment. Journal
American Water Works Association 63(1) 52-58, 1971.
24. American National Standards Institute, Inc. American National Standard
Acceptable Concentrations of Hydrogen Sulfide. ANSI Z37.2 - 1972,
American National Standards Institute. New York, New York, 1972 8 pp.
25. Directo, L.S., Chen, C.L. and Kugelman, I.R. Pilot Plant Study of Physical-
Chemical Treatment. Presented at Water Pollution Control Federation
Conference, Denver, Colorado, 1974. 30 pp.
26. Weber, W.J., Hopkins, C.B. and Bloom, R., Physicochemical Treatment of
Wastewater. Journal Water Pollution Control Federation 42(1): 83-99
1970.
172
-------
27. Weber, W.J. The Role of Activated Carbon in Physicochemical Treatment.
Presented at the US/USSR Symposium on Physical-Chemical Treatment of
Wastewaters, Environmental Research Center, Cincinnati, Ohio, 1975, 49 pp.
28. Process Design Manual for Sulfide Control in Sanitary Sewerage Systems.
U.S. Environmental Protection Agency Technology Transfer, October 1974.
29. Schimmel, C. and Griffin, D.B., Treatment and Disposal of Complex
Industrial Wastes. EPA-600/2-76-123, U.S. Environmental Protection
Agency, Cincinnati, Ohio, 1976, 181 pp.
30. Westrick, J.J. and Cohen, J.M. Comparative Effects of Chemical Pre-
treatment on Carbon Adsorption. Presented at the Water Pollution
Control Federation Conference, Atlanta, Georgia 1972, 37 pp.
31. Ford, D.L. Putting Activated Carbon in Perspective to 1983 Guidelines.
Industrial Water Engineering 14(3):20-27, 1977.
32. Weber, W.J., Physiocochemical Processes for Water Quality Control. John
Wiley and Sons, Inc., New York, New York, 1972, 640 pp.
33. Koon, J.H., and Kaufman, W.J. Ammonia Removal From Municipal Wastewaters
By Ion Exchange. Journal Water Pollution Control Federation 47(3):448-
465, 1975.
34. Optimization of Ammonia Removal By Ion Exchange Using Clinoptilolite.
Water Pollution Control Research Series 17080 DAR 09/71, U.S. En-
vironmental Protection Agency, Washington, D.C. 1971, 189 pp.
35. Hodgman, C.D., Weast, R.C. , and Selby, S.M. Handbook of Chemistry and
Physics, 39th ed. Chemical Rubber Publishing Co., Cleveland, Ohio,
1958, 3211 pp.
36. Ames, L.L. The Cation Sieve Properties of Clinoptilolite. American
Mineralogist, 45:689-700, 1960.
37. Gardner, F.H., Williamson, A.R., and Skovronek, H.S. Naval Stores Waste-
water Purification and Rinse by Activated Carbon. EPA-600/2-76-227,
U.S. Environmental Protection Agency, Cincinnati, Ohio. 1976, 34 pp.
38. Von Dreusche, C. Jr. Process Aspects of Regeneration in a Multiple
Hearth Furnace. Presented at 78th National Meeting of American Institute
of Chemical Engineers, Salt Lake City, Utah, 1974.
39. Semmens, M.J. The Feasibility of Using Nitrifying Bacteria to Assist
the Regeneration of Clinoptilolite. Presented at 32nd Purdue Industrial
Waste Conference, West Lafayette, Indiana, 1977, 30 pp.
173
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1 REPORT NO
EPA-600/2-78-201
4. TITLE AND SUBTITLE
EVALUATION OF PHYSICAL CHEMICAL TREATMENT AT ROSEMOUNT
5. REPORT DATE
December 1978 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
3 RECIPIENT'S ACCESSI Ol> NO.
7 AUTHOR(S)
R. C. Polta, R. W. DeFore and W. K. Johnson
8. PERFORMING ORGANIZATION REPORT NO.
9 PERFORMING ORGANIZATION NAME AND ADDRESS
Metropolitan Waste Control Commission
350 Metro Square Building
St. Paul, Minnesota 55101
10. PROGRAM ELEMENT NO.
11 CONTRACT/GRANT Nd.
Grant No. S802666
Task RF/02
12 SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory-Cinti,OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final (6/6/75-8/13/77)
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer: Sidney A. Hannah (513/684-7651)
16. ABSTRACT
This research program was conducted to demonstrate the effectiveness of physical
chemical treatment of raw municipal wastewater at the Rosemount Advanced Wastewater
Treatment Plant and more specifically to evaluate: the performance of the system as
a whole, the performance of the individual treatment processes, and the costs associ-
ated with operation and maintenance. During the two year demonstration period the
facility treated an average flow of approximately 0.25 mgd by means of chemical
clarification, filtration, carbon adsorption and ion exchange; both the activated
carbon and ion exchange media were regenerated onsite.
The performance data are summarized according to the five process flow schemes used
in addition to discussion of the individual treatment processes. Cost data are pre-
sented for each treatment process. These data are used to construct an estimate of
operating and maintenance costs for a 10 mgd facility.
The operating and maintenance problems encountered during approximately four years of
operation are described along with their solutions when they were determined. Design
recommendations are presented.
17.
a
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Sewage Treatment*, Chemical Removal*,
Activated Carbon Treatment, Coagulation,
Clarification, Filtration, Water Pollution
Physical-Chemical Treatment
Ammonia Removal
b IDENTIFIERS/OPEN ENDED TERMS
COSATl Held/Group
13B
13 DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19 SECURITY CLASS (This Report)
Unclassified
21 NO. OF PAGES
20 SECURITY CLASS {This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
174
d U S GOVERNMENT PRINTING OFFICE 1979-657-060/1567
------- |