600281092
Technical Report No. 251
October 1980
CHLORINE DIOXIDE FOR WASTEWATER DISINFECTION:
A FEASIBILITY EVALUATION
by
Paul V. Roberts, E. Marco Aieta, James D. Berg, and Bruce M. Chow
Supported by
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Grant R-805426
Department of CIVIL ENGIOSrEJEJiRIlsrG
STANFORD UNIVERSITY
-------
Department of Civil Engineering
Stanford University
Stanford, California 94305
CHLORINE DIOXIDE FOR WASTEWATER DISINFECTION:
A FEASIBILITY EVALUATION
by
Paul V. Roberts, E. Marco Aieta, James D. Berg, and Bruce M. Chow
Technical Report No. 251
October 1980
This research was supported by
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
Research Grant R-805426
Project Officer
Mark C. Meckes
-------
ABSTRACT
Chlorine dioxide was compared with chlorine for the disinfection of
wastewater in laboratory experiments. Disinfection with chlorine dioxide was
also demonstrated at a full-scale wastewater treatment plant. Criteria for
comparison included coliform kill, inactivation of poliovirus and other indi-
cators, and formation of halogenated organic byproducts.
Laboratory experiments were conducted using a 4-liter batch reactor
characterized by intense mixing. The experiments were conducted according to
a full factorial design, with mass dose of disinfectant (3 levels) and contact
time (3 levels) as independent variables. The fractional survival of coliform
bacteria was correlated with the product of disinfectant residual times con-
tact time.
In general, chlorine dioxide accomplished a given fractional kill of
total coliforms with a smaller product (residual x time) than chlorine. For a
given contact time, the residual required to achieve a given fractional kill
of coliforms was 2 to 70 times smaller for chlorine dioxide than for chlorine.
The disinfectant demand is greater for chlorine dioxide than for chlorine in a
conventional activated sludge effluent, but less than for chlorine in a nitri-
fied, filtered effluent, when residuals are expressed on a mass concentration
basis and conditions are chosen to result in equal disinfection performance.
Considering both required residual and demand, the required doses of the
disinfectants were estimated to satisfy three assumed coliform disinfection
levels with two types of effluents: conventional activated sludge and fil-
tered, nitrified activated sludge. The required mass doses of the disinfec-
tants were approximately equal for treating conventional activated-sludge
effluent—approximately 2 mg/1 to satisfy a standard of 1000 total coliforms/
100 ml, 2.5 mg/1 to satisfy a 200 coliforms per 100 ml standard, and 8 mg/1 to
satisfy a standard of 2.2 coliforms per 100 ml. The required dose of chlorine
was approximately 2 to 10 times greater than that of chlorine dioxide for
treating filtered, nitrified effluent, depending on the coliform standard.
The results of studies conducted at a full-scale plant generally agreed within
a factor of two with the predictions from laboratory studies, when compared on
the basis of the product (residual x time) required to accomplish a given
fractional kill.
For the case likely to be most typical in practice—the disinfection of
conventional secondary effluent to meet a total coliform standard of 1000 per
100 ml—disinfection with chlorine is estimated to cost 0.3 to 1.7 cents per
m (1.1 to 6.4 cents per 1000 gallons), depending on plant size, compared with
1.5 to 3.5 cents per m (5.8 to 13.4 cents per 1000 gallons) for disinfection
with chlorine dioxide (1979 price levels).
ii
-------
Chlorine dioxide was found to be more effective for inactivating Polio-
virus I and natural populations of coliphage in both non-nitrified and fil-
tered, nitrified wastewater effluents. Chlorine dioxide treatment formed no
measurable amounts of trihalomethane byproducts, whereas chlorine treatment
formed 0.5 to 5 yMol per liter of trihlomethanes, chiefly chloroform, in
experiments using wastewater effluents. The measured amount of total organic
halogen (TOX) formed by chlorine dioxide disinfection was so small as to be on
the margin of statistical significance; chlorination formed at least 10 to 20
times more undesirable halogenated byproducts measured as TOX than did chlo-
rine dioxide treatment. These advantages of chlorine dioxide should be con-
sidered, along with the cost-effectiveness comparison based on coliform kill,
to reach decisions as to when and where to employ chlorine dioxide as a disin-
fectant in wastewater treatment.
This report was submitted in fulfillment of Grant No. R-805426 by Stan-
ford University under the sponsorship of the U.S. Environmental Protection
Agency. This report covers the period September 5, 1977 to June 5, 1980, and
work was completed as of June 5, 1980.
iii
-------
CONTENTS
Abstract ii
Figures vi
Tables ix
Acknowledgments xii
1. Introduction 1
Background 1
Objectives 2
2. Conclusions 4
3. Recommendations 6
4. Materials and Methods 7
Wastewater characterization 7
Microbiological analyses 9
Disinfectant chemistry 10
Halogenated organics analyses 12
5. Chlorine and Chlorine Dioxide Chemistry and Generation .... 16
Measurements of chlorine dioxide and chlorine species . . 16
Measurement of chlorine dioxide generator product
composition 20
6. Models of Disinfection Kinetics 30
First-order kinetics 30
Decreasing rate and lag time 30
Retardant reaction approach 32
7. Disinfection Experiments with Conventional Coliform Indicators 37
Experimental design 37
Statistical analysis of experimental results 46
The effect of process sequence on bacterial numbers ... 54
Comparisons among wastewaters 57
Effects of treatment level on the disinfection process . . 63
Comparison of disinfection with chlorine and chlorine
dioxide 68
Predicting full-scale performance from laboratory
experiments 71
8. Virus Inactivation 76
Results 76
9. Recovery Following Disinfection 83
Experimental 83
Results 83
10. Halogenated Organics Produced During Disinfection 86
Chlorine dioxide versus chlorine 86
Mixtures of chlorine dioxide and chlorine 95
iv
-------
11. Cost Comparison Between Chlorine Dioxide and Chlorine for
Wastewater Disinfection 100
Basis of comparison 100
Capital costs 101
Operation and maintenance costs—excluding chemical
costs 102
Chemical costs 104
Cost summaries 105
Discussion 112
References 114
Appendices 122
-------
FIGURES
Number Page
1 Chlorine dioxide generator using acid activation of a sodium
chlorite solution 10
2 Diagram of TOX analysis 14
3 Reduction of chlorine dioxide as a function of pH 16
4 Chlorine dioxide generator, acid-chlorite process 26
5 Types of bacterial survival curves 31
6 Multi-media filtration column 38
7 Wastewater treatment plant flow schemes 40
8 Experimental dose-time matrix for chlorine-chlorine dioxide
comparison 42
9 A rapid-mix, rapid-sampling, 4-liter batch reactor 43
10 Disinfection contact tank at the Dublin-San Ramon wastewater
treatment plant ............. 44
11 Results of tracer study of disinfectant contact tank 45
12 Dye tracer concentration measured at overflow weir 47
13 Comparison of chlorine and chlorine dioxide bactericidal
effectiveness at 30-minute contact time 53
14 Chlorine and chlorine dioxide residual die-away in Palo Alto
secondary effluent 54
15 Comparison of chlorine and chlorine dioxide bactericidal
effectiveness at 10 mg/1 dose 55
16 Coliform inactivation by free and combined chlorine in San Jose
nitrified effluent 61
VI
-------
Number
Page
17 Coliform inactivation by free and combined chlorine in San Jose
filtered effluent 62
18 Coliform inactivation by combined chlorine in San Jose
wastewater at three sampling points 62
19 Coliform inactivation by chlorine in San Jose wastewater ... 64
20 Coliform inactivation by chlorine in filtered and unfiltered
secondary effluent, Palo Alto 65
21 Coliform inactivation by chlorine dioxide in San Jose
wastewater effluents 66
22 Coliform inactivation by chlorine dioxide in filtered and
unfiltered secondary effluent, Palo Alto 67
23 Coliform inactivation by chlorine and chlorine dioxide in
1978 experiments with unfiltered secondary effluent from
Palo Alto 68
24 Coliform inactivation by chlorine and chlorine dioxide in
1979 experiments with unfiltered secondary effluent from
Palo Alto 69
25 Coliform inactivation by chlorine and chlorine dioxide in
1979 experiments with filtered secondary effluent 70
26 Coliform inactivation by chlorine and chlorine dioxide in
San Jose secondary effluent 70
27 Coliform inactivation by chlorine and chlorine dioxide in
San Jose nitrified effluent 72
28 Coliform inactivation by chlorine and chlorine dioxide in
San Jose filtered effluent 72
29 Coliform inactivation by chlorine and chlorine dioxide in
laboratory experiments with Dublin effluent 73
30 Coliform inactivation by chlorine and chlorine dioxide in
field experiments with Dublin effluent 73
31 Coliform inactivation by chlorine in laboratory and field
experiments with Dublin effluent 74
32 Coliform inactivation by chlorine dioxide in laboratory and
field experiments with Dublin effluent 74
vii
-------
Number Page
33 Comparison of in situ coliphage and an inoculum of Poliovirus I
in non-nitrified secondary effluent 78
34 Total coliform survival in non-nitrified secondary effluent
comparing chlorine dioxide and chlorine at three doses and
three contact times 79
35 In situ coliphage (E_. coli B host) survival in non-nitrified
secondary effluent, comparing chlorine dioxide and chlorine
at three doses and three contact times 80
36 Responses of several in situ organisms to chlorine dioxide
and chlorine at very short contact times in non-nitrified
secondary effluent 81
37 Total coliform and Poliovirus I survival in non-nitrified
activated-sludge effluent (Palo Alto) and in nitrified,
,• filtered effluent (Dublin) 82
38 Five-day recovery following disinfection 84
39 Procedure for bench-scale determinations of halogenated
organics formation 88
40 TOX and THM production in non-nitrified effluent
(Experiment 4—Palo Alto wastewater) 90
41 TOX and THM production in filtered, nitrified effluent
(Experiment 6—Dublin-San Ramon wastewater) 91
42 Disinfectant residuals in Experiment 8 (Palo Alto wastewater) . 93
43 Disinfectant residuals in Experiment 6 (Dublin-San Ramon
wastewater) 93
44 THM production resulting from various mixtures of chlorine
and chlorine dioxide in Experiment 2 (Dublin-San Ramon
wastewater 97
45 TOX and THM production resulting from various mixtures of
chlorine and chlorine dioxide in Experiment 5 (Dublin-
San Ramon wastewater) 98
viii
-------
TABLES
Number Page
1 Estimated Standard Deviation of THM Analysis 13
2 Precision of TOX Measurement 15
3 Blank Correction for TOX Calculation 15
4 Comparison of Amperometric and lodometric Methods for
Determining Concentration of ClO^ 17
5 Comparison of Amperometric and lodometric Methods by Analysis
of Variance 17
6 Amperometric Titration Evaluation 18
7 Analyses of Variance for Amperometric Titration 19
8 The Effect of Chlorite on the Determination of Chlorine Dioxide
by Amperometric Titration 20
9 Equivalent Weights for Calculating Concentrations on a Mass
Basis 23
10 Composition of Reactants and Products in Laboratory Generation of
Chlorine Dioxide by Sulfuric Acid Activation of Sodium Chlorite 24
11 Composition of Reactants and Products from Chlorine Dioxide
Generation in a Continuous, Full-Sized Acid-Chlorite Reactor
Using H2S04 27
12 Composition of Reactants and Products from Chlorine Dioxide
Generation in a Continuous, Full-Sized Acid-Chlorite Reactor
Using HC1 28
13 Yield of Chlorine Dioxide from Acid Activation in Field
Experiments 28
14 Characterization of Wastewater Effluents Used in Disinfection
Experiments 41
ix
-------
Number Page
15 Mean Log^Q Surviving Total Coliform Counts in 1978 Batch
Experiments with Palo Alto Effluent ............. 49
16 ANOVA for Bacterial Analysis Before Disinfection ....... 50
17 ANOVA for Examining Experimental Reproducibility ....... 50
18 ANOVA for Comparison of Chlorine and Chlorine Dioxide as
Bactericides in Palo Alto Effluent ............. 52
19 Mean Log [Survival Ratio] of Total Coliforms: Comparison of
Disinfectants at Different Combinations of Dose and Contact
Time ............................ 52
20 Initial Total Coliforms in Undisinfected Effluents from the
San Jose Wastewater Treatment Plant ............. 56
21 ANOVA of San Jose Initial Bacteria Numbers .......... 56
22 Logs of Initial Bacterial Numbers at the Palo Alto Wastewater
Treatment Plant ....................... 57
23 ANOVA of Palo Alto Initial Bacterial Numbers ......... 57
24 Values of Fitting Constants for Chlorine and Chlorine Dioxide
in the Disinfection Model .................. 59
25 Chlorine to Ammonia Ratios for San Jose Experiments ...... 60
26 Comparison of Model Coefficients, Segregating Disinfection
with Free vs Combined Chlorine ............... 61
27 One-Hour Chlorine Demand ................... 64
28 One-Hour Chlorine Dioxide Demand ............... 67
29 Typical Characteristics of Wastewater Effluents Used to
Measure Formation of Halogenated Organics .......... 87
30 Wastewater Characteristics in Field Experiments at the
Dublin-San Ramon Chlorine Contactor ............. 88
31 THM and TOX Formation by Chlorine and Chlorine Dioxide in
Bench-Scale Experiments ................... 89
32 Summary of THM and TOX Field Experiments ........... 91
33 Mean Values and Standard Deviations of 24-Hour Halogenated
Organic Byproduct Formation with Chlorine Dioxide and
Chlorine .......................... 92
-------
Number Page
34 Disinfectant Demand 95
35 A Summary of the Results and Conditions of Experiments Comparing
THM and TOX Formation by Mixtures of Chlorine and Chlorine
Dioxide 99
36 Cases for Evaluation of the Relative Costs of Disinfection
with Chlorine and Chlorine Dioxide 101
37 Summary of Estimates of Required Disinfectant Dosages Used in
Cost Evaluation Cases 103
38 Amounts of Disinfectants Required to Achieve Total Coliform
Standards 104
39 Unit Costs of Chemicals Required in Wastewater Disinfection . . 105
40 Disinfection Cost Summary for Case A in Thousand $ per Year . . 106
41 Disinfection Cost Summary for Case B 107
42 Disinfection Cost Summary for Case C 108
43 Disinfection Cost Summary for Case D 109
44 Disinfection Cost Summary for Case E 110
45 Disinfection Cost Summary for Case F Ill
xi
-------
ACKNOWLEDGMENTS
The active assistance of the staff of the Dublin-San Ramon Services
District, particularly Robert Swanson, Plant Superintendent, was essential in
carrying out the full-scale field tests of chlorine dioxide. The Rio Linda
Chemical Co. made available chlorine dioxide generating facilities for the
full-scale demonstration.
The staff of the Palo Alto Regional Water Quality Control Plant and the
San Jose-Santa Clara Water Pollution Control Plant provided samples for the
laboratory studies. The Fischer-Porter Company generously loaned an ampero-
metric titrator. Dr. Gary Stevenson gave useful advice on residual measure-
ment techniques.
Gary D. Hopkins assisted in the full-scale demonstration experiments and
determined total organic halogen concentrations. Annie M. Godfrey, Christoph
Munz, Paul Dandliker, John Gonzales, Laurie Lapat, Mette Horn, Niambi Loud, and
Christian Drozier performed many of the microbiological and chemical analyses.
We are grateful to Professor Robert Cooper, School of Public Health,
University of California, Berkeley, and Professor Carleton Schwerdt, Depart-
ment of Medical Microbiology, Stanford University, for the virus assays per-
formed in their laboratories.
xii
-------
SECTION 1
INTRODUCTION
BACKGROUND
Disinfection of wastewaters has long been practiced in areas where the
use of the receiving waters for public water supply, contact recreation,
and/or shellfish habitat is to be protected. Many wastewater-effluent dis-
charge permits allow a maximum geometric mean of 1000 total coliforms per 100
ml, or 200 fecal coliforms per 100 ml, in a 30-day period. The State of
California permits only 2.2 total coliforms per 100 ml if an effluent is to be
used for groundwater recharge.
Chlorine has been the favored disinfectant in both water and wastewater
treatment by virtue of its bactericidal effectiveness, low cost, convenience,
and relatively long-lived residual. However, chlorination as normally prac-
ticed in water treatment recently has been found to result in the formation of
trihalomethanes and other chlorinated organics that are undesirable from the
viewpoint of public health and water pollution control in general. Although
the formation of chlorinated organics during disinfection of wastewater by
chlorine previously has not been investigated thoroughly, there is reason to
believe that such hazardous byproducts are formed. In view of the residual
organic contamination of even highly treated wastewater effluents and the re-
sulting high doses of chlorine needed to provide a chlorine residual in waste-
waters, it can be expected that undesired, chlorinated organic byproducts will
be formed to a greater extent during wastewater chlorination than has been
reported for water chlorination.
Thus there is a need to find a substitute for chlorine as a disinfectant
in wastewater treatment that fulfills the following criteria:
1. Cost effectiveness as measured by the relationship of the
total cost (amortization plus operating costs) to the effec-
tiveness as a bactericide under the conditions encountered
in practice.
2. Absence of participation in side reactions that yield unde-
sirable byproducts, particularly chlorinated organic com-
pounds .
3. Safety and convenience of use and ease of installation in
existing wastewater treatment plants.
4. No residual toxicity to aquatic organisms in receiving wa-
ters.
-------
Of the possible substitutes for chlorine, ozone has attracted the most
attention. Ozone is a powerful disinfectant but is more expensive than chlo-
rine and does not provide a stable residual.
Chlorine dioxide (C102), although more expensive than chlorine, may offer
an attractive alternative because it:
1. Is a powerful disinfectant over a broad pH range,
2. Provides a residual that is easily measurable for purposes
of control,
3. Does not react with ammonia to form chloramines,
4. Does not enter into reactions with organic material to form
some classes of chlorinated organic compounds considered
hazardous to health (e.g., trihalomethane compounds), and
5. Does not react to form chlorinated phenols to the extent
that chlorine does.
The second factor mentioned above gives chlorine dioxide an inherent ad-
vantage over ozone from the viewpoint of process control. All the other
factors are advantages compared with chlorine. Factor (3), the absence of
reactivity with ammonia, constitutes an especially important advantage for
chlorine dioxide over chlorine in wastewater disinfection, if a free chlorine
residual is required; the chlorine demand to reach the breakpoint may exceed
100 mg C12 per liter.
OBJECTIVES
The purpose of this project is to evaluate chlorine dioxide as an altern-
ative to conventional chlorination for the disinfection of wastewater.
The specific objectives are:
1. To assemble and evaluate the available information concern-
ing the chemistry of chlorine dioxide generation and of its
behavior in aqueous solution, the technology and costs of
manufacture, its effectiveness as a disinfectant, and the
possible side effects of its use.
2. To establish the dose-effectiveness relationship for chlo-
rine dioxide as a disinfectant of wastewater after secondary
treatment and after various stages of advanced treatment,
using as a criterion the survival of coliform bacteria.
3. To compare the effectiveness of chlorine dioxide with that
of chlorine as a disinfectant.
4. To develop a concept for a continuous reactor system to
fulfill coliform requirements corresponding to standard
secondary treatment, and advanced treatment for the purpose
of wastewater reclamation and reuse.
5. To prepare a preliminary design for treatment plants of 1-,
5-, 10-, 50-, and 100-mgd capacity and to estimate the costs
of construction and operation.
-------
6. To obtain preliminary evidence as to whether the formation
of chlorinated organic byproducts during wastewater disin-
fection conforms to the results from studies of water disin-
fection.
-------
SECTION 2
CONCLUSIONS
The use of chlorine dioxide as a disinfectant in wastewater treatment is
technically feasible. This was demonstrated in short-term, continuous opera-
tion at a full-scale wastewater treatment plant, which served to confirm
laboratory results.
Results of chlorine dioxide disinfection can be conveniently correlated
using the logarithm of the surviving fraction of organisms as the dependent
variable and the logarithm of the product of the disinfectant residual times
the contact time as the independent variable. The form of the correlations so
obtained for chlorine dioxide is similar to that for disinfection with chlo-
rine. Compared with chlorine, however, a given degree of disinfection was
achieved with a smaller value of the product (residual x time) when chlorine
dioxide was used. Based on this general finding, it is concluded that, in a
technical sense, chlorine dioxide is superior to chlorine for the disinfection
of wastewater.
The chlorine dioxide disinfection process can be controlled conveniently
and with adequate precision by adjusting the dose to provide a desired resid-
ual. Proven residual measurement techniques, such as automated amperometric
titration, are suitable for measuring the chlorine dioxide residual for con-
trol purposes.
To satisfy a given coliform standard at a given contact time, the re-
quired chlorine dioxide residual is only one-third to one-tenth the corre-
sponding chlorine residual (mass concentration basis). The ratio of required
doses (mass ClO^imass Cl~) ranges from approximately 0.05 to 1.0. Hence, in
general less chlorine dioxide than chlorine must be used to satisfy the coli-
form standards usually applied in wastewater disinfection.
The unit cost of chlorine dioxide is approximately twenty times that of
chlorine. This difference is attributed almost entirely to the high cost of
the sodium chlorite raw material.
The cost of disinfection with chlorine dioxide to achieve a given coli-
form standard is greater than or equal to that with chlorine in all cases
evaluated; the higher unit cost of chlorine dioxide generally more than off-
sets its lower dose requirement. For the most typical case—disinfection of
conventional activated-sludge effluent to satisfy a total coliform standard of
1000 per 100 ml—chlorine dioxide is two to five times more expensive than
chlorine.
-------
Chlorine dioxide forms no significant amounts of undesirable chlorinated
organic byproducts under the conditions of wastewater disinfection. No mea-
surable amounts of trihalomethane compounds (THMs) were formed when secondary
effluents were dosed with 20 mg/1 or 40 mg/1 ClO^- Total organic halogen
(TOX) was formed to a small (but statistically significant) extent in one of
four experiments with ClO^- Chlorine at equal mass doses formed copious
amounts of THMs (0.5 to 5 yMol per liter) under the same conditions. The
average TOX formation from chlorine was 10 to 20 times that from chlorine
dioxide at the same mass dose of disinfectant.
-------
SECTION 3
RECOMMENDATIONS
The evaluation of chlorine dioxide as an alternative to chlorine for the
disinfection of wastewater has resulted in a decision-maker's dilemma. On the
one hand, chlorine dioxide is found to be superior to chlorine according to
technical criteria. On the other hand, chlorine dioxide is not cost-effective
when the customary performance criterion (total coliform standard) is applied.
Two important classes of questions remain: Are the unquantified benefits
of chlorine dioxide compared with chlorine—namely, negligible formation of
potentially hazardous chlorinated organic byproducts and superior virus inac-
tivation—worth the additional cost? Can the unit cost of chlorine dioxide be
reduced to a level at which chlorine dioxide would be economically attractive
in the customary sense, even without considering its additional, unquantified
advantages? Both types of questions deserve further study.
The chemical composition of disinfection byproducts and endproducts of
all disinfectants should be studied systematically. There is an urgent need
for better understanding of the halogenated organic byproducts of wastewater
chlorination.
The apparent advantage of chlorine dioxide with regard to virus inactiva-
tion needs to be documented more convincingly. Comparison of the kill of
coliforms with that of other indicators, using alternative disinfectants, is
worthy of further investigation. Also of potential significance is the devia-
tion of disinfection data from the simple first-order model usually assumed to
apply. The causes underlying these deviations should be studied with a view
toward re-evaluating the adequacy of the total coliform count as an indicator
of disinfection performance.
Development of processes to produce chlorine dioxide more cheaply at
treatment plant sites should be encouraged by supporting demonstration of
candidate concepts.
-------
SECTION 4
MATERIALS AND METHODS
WASTEWATER CHARACTERIZATION
2H
The pH was determined using a pH meter (Corning Model 610A), with a pre-
cision of ± 0.05 pH unit. The pH meter was standardized using a pH-buffered
standard at a pH near the value expected in the sample.
Alkalinity
Alkalinity was determined by potentiometric titration to pH 4.3.
COD
COD was determined using the low-level COD procedure outlined in Standard
Methods, Section 508 (1). A potassium acid phthalate standard was run with
each set of determinations to assure accuracy. Samples were stored as for the
ammonia nitrogen test.
Non-Filterable Residue
Non-filterable residue dried at 103-105°C was determined by the procedure
outlined in Standard Methods (1), Section 208A, using Gooch crucibles. Sam-
ples were tested on the day of the experiment where possible. If storage was
required, the storage procedure used was the same as for the ammonia nitrogen
test.
Ammonia Nitrogen
Two methods were used to determine ammonia nitrogen. From October 1977
to January 1979 a specific ion probe made by Orion Co. was used. The proce-
dure was as follows:
1. A standard curve (semi-log) was prepared at the beginning of
each set of tests and every two hours afterwards using 100-,
10-, and 1-mg/l ammonium chloride standards. The standardi-
zation procedure was the same as for measuring ammonia in
samples as well. The meter was zeroed using the 10-mg/l
standard.
-------
2. Readings were made as follows: (a) 100-ml samples were pre-
pared; (b) 1 ml of 10N_ NaOH was added and the sample was
stirred continuously until a stable meter reading was at-
tained, which took approximately 15 min.
All samples were allowed to temperature equalize for at least 1/2 hr before
the NaOH was added.
After January 1979 samples became too numerous, and a change was made to
the distillation followed by acidimetric titration procedure described in
Standard Methods (1), Sections 418A and B. In addition, total Kjeldahl nitro-
gen was measured by the method listed under "Nitrogen (Organic)" in Section
421 in Standard Methods (1). An ammonium chloride standard and a glycine
standard were used during each set of determinations to assure accuracy.
Samples were acidified to below pH 2 using IkSO, and refrigerated at 4°C.
Samples were analyzed within two weeks.
TOG
TOG analysis was done using a Dohrmann Envirotech DC-52 Organic Carbon
Analyzer with an Oceanographic International Ampule Crushing Unit adapted to
it. Sample preparation was as follows, using 10-ml glass ampules:
1. 1 ml of saturated potassium persulfate was added to each
ampule.
2. Sample water was added (0.5 to 1.5 ml for secondary efflu-
ent).
3. Distilled water was added to fill the ampules to no more
than 10 ml total volume.
4. 0.2 ml of 10% phosphoric acid solution was added.
5. The ampules were purged for 7-9 min using pure oxygen
cleaned by passing the 02 over a 475°C cupric oxide
catalyst.
6. The ampules were sealed and autoclaved for 4 hrs.
7. The special crushing unit was used to open the ampule and
purge the ampule with a helium carrier gas which passed into
a flame-ionization detector.
Note that the initial purging in step 5 to remove inorganic carbon also re-
moved volatile organics. Therefore the TOG value is actually a non-purgeable
TOG. A calibration curve was made by running sucrose standards in the same
manner as the samples. Reagent blanks were also run in this manner.
Samples awaiting sealing in ampules were stored in teflon sealed hypo-
vials with aluminum seals. These samples were also acidified to pH 2 using
HC1 and stored at 4°C.
The detection limit of the TOG procedure is 0.1 mg/1 C. The precision is
approximately ± 5% at TOG concentrations greater than 1 mg/1 C.
-------
MICROBIOLOGICAL ANALYSES
Bacteria
Samples for microbiological analysis were collected in sterile bottles
containing sodium thiosulfate to neutralize any residual oxidant and were
analyzed immediately. Three field experiments were run at the Dublin-San
Ramon Wastewater Treatment Plant. During these experiments, samples were
collected, stored on ice, and analyzed after return to the lab, approximately
one hr after the completion of a 6-hr set of replicate runs. A set of samples
that had been disinfected by chlorine and chlorine dioxide, as well as an
untreated influent sample, were analyzed immediately for total coliform bacte-
ria at the field site and compared with the iced and transported samples.
There were no significant differences between the analyses conducted in the
field and in the laboratory-analyzed samples. Therefore, it is assumed that
the microbiological integrity of the samples was maintained during the course
of a long experiment.
Most probable number (MPN) and membrane filter (MF) methods of bacterial
enumeration were done according to Standard Methods (1). MPN media used were
lactose broth as a presumptive medium and brilliant green bile broth for con-
firmation of total coliforms and EC medium for confirmation of fecal coli-
forms. Gelman GN-6 0.45-micron filters were used for the MF tests with M-Endo
Agar for total coliforms using the single-step direct technique and M-FC
medium for fecal coliforms. KF streptococcal agar was used to assay for the
fecal streptococcus group.
MF tests and MPN presumptive tests were incubated at 35°C ± 0.5°C for 24
± 2 hrs. Confirmed MPN tests and fecal streptococcus tests were incubated at
35°C ± 0.5°C for 48 ± 3 hrs. Fecal coliform analyses, either by MPN or MF
method, utilized an incubation period of 24 ± 2 hrs and a temperature of
44.5°C ± 0.2°C in a circulating water bath.
Viruses
Two bacterial virus methods were evaluated to find a suitable in situ
model for virus inactivation by C102- One method, that of Kott (2), also has
been used in earlier studies (3). The method uses Escherichia coli B as the
host. Samples were enriched overnight in nutrient broth at 35°C and plated
for observation of plaques as a positive test. The procedure employs MPN
tables; low numbers of coliphage, < 2 plaque-forming units/100 ml, are
detectable. The procedure is time-consuming, however. A second method, the
Reverse Phage Titer Rise Reaction (RPTRR), as developed by Atlantic Research
(1978) (4), was also used. The RPTRR method uses _E_. coli C as the host bac-
teria, and yields results in 4 to 6 hrs. The host E_. coli C is thought to be
susceptible to a broader range of coliphage than is J^. coli B.
The animal virus used in the study was Poliovirus I LSC strain that had
been grown in Buffalo Green Monkey Kidney cells (BGM). The virus were harves-
ted from the cells by freezing, thawing, and centrifugation to a titer of ap-
proximately 109 plaque-forming units per milliliter (PFU/ml). The virus were
recovered after experimentation using a standard plaque assay. The animal
-------
virus assays were performed in the laboratories of either Dr. Robert Cooper,
Sanitary Engineering Research Laboratory, University of California at Berke-
ley, or Dr. Carleton Schwerdt, Department of Medical Microbiology, Stanford
University.
DISINFECTANT CHEMISTRY
Chlorine Dioxide Generation
Chlorine dioxide was generated according to the method described in Stan-
dard Methods (1), Section 411A, diagrammed in Fig. 1, except that the acid and
sodium chlorite concentrations were doubled to obtain higher chlorine dioxide
concentrations in the stock solution. The concentration so obtained was 1000
to 1500 mg/1 as CIO™. Spectrophotometric analysis at 360 nm was used to
verify that the chlorine species in solution was indeed chlorine dioxide.
Chlorine Generation
Aqueous chlorine stock solutions were made by bubbling 99.99% chlorine
gas through chilled water. The solutions so prepared contained chlorine at a
concentration of 2000 to 4000 mg/1 as Cl.
AIR
SUPPLY
REACTION
VESSEL
VENT
TO
HOOD
NaCIO,
SALT
TOWER
C102 STOCK SOLN
Figure 1. Chlorine dioxide generator using acid activation of a sodium
chlorite solution.
10
-------
Standardizing of Disinfectant Stock Solutions
Disinfectant stock solutions were standardized on the day of each experi-
ment by the iodometric method described in Standard Methods (1), Sections 409A
and 411A. Standard sodium thiosulfate titrant at O.OlN^was used, standardized
by the dichromate method. Blank titrations were omitted because it was found
that blank contributions were negligible.
Residual Measurement for Routine Analyses
Both chlorine and chlorine dioxide residuals were measured by amperomet-
ric titration (1). The amperometric titrating unit used was a prototype model
made by Fischer and Porter, similar in function to the Fischer and Porter
Model 17T1010 amperometric titrator. The back-titration procedure was used
for both disinfectants. It consists of the following steps:
1. A known volume of 0.00564N^ phenylarsine oxide solution was
added to a 250-ml beaker in excess of the expected chlorine
or chlorine dioxide residual.
2. 2 ml of the appropriate pH buffer was added. A pH-4 buffer
(acetate) was used for chlorine analysis and a pH-7 buffer
(phosphate) was used for chlorine dioxide analysis.
3. A 1-ml aliquot of 5% weight-per-volume potassium iodide
solution was added.
4. A known volume of sample was carefully poured into the
beaker, approximately 150-175 ml.
5. The titrator was turned to the "Total" position and the
above solution was titrated with 0.00564N^ iodine solution to
an amperometric endpoint. For this particular amperometric
titrator, the endpoint was denoted by sudden needle deflec-
tion.
Residual was calculated by the following formulas:
0.00564N. x 35453 mg Cl/equiv
mg/1 Cl = [PAO - (I2 - blank) x CF]
mg/1 C102 = [PAO - (I2 - blank) x CF]
vol sample (ml)
0.00564N x 67500 mg Cl/equiv
vol sample (ml)
where PAO = volume of 0.00564^ phenylarsine oxide used in ml
I2 = volume of 0.00564N_ I2 titrant used in ml
blank = volume of I2 found for blank (see below)
CF = correction factor for I2 titrant (see below).
The correction factor was found as follows:
11
-------
1. A beaker was prepared with 2 ml of the phenylarsine oxide solu-
tion, the appropriate pH buffer, and potassium iodide solution.
2. 150 ml of deionized water was added.
3. The sample was titrated as in Step 5 above. This procedure was
repeated for each pH.
_ vol phenylarsine oxide
volume !„ titrant
The blank was found by the same procedure as for the correction factor except
that the appropriate amounts of undisinfected sample water were added:
Blank = I~ volume - I^ volume from CF determination
A blank was determined for each pH and for each different sample volume used.
HALOGENATED ORGANICS ANALYSES
Trihalomethane (THM) Analysis
THM analysis was accomplished by liquid-liquid extraction to concentrate
the organics, and subsequent gas-liquid chromatographic analysis according to
the procedure of Henderson et al. (5). Liquid-liquid extraction consisted of
the following steps:
1. 1 ml of specially cleaned pentane [purchased as Basic
Alumina (Activity 1) grade] was injected into each of the
sealed 60-ml hypovials containing the samples (displaced
water was discarded).
2. A prescribed amount of 1,2-dibromoethane internal standard
was injected so that its concentration in the hypovial was
near that expected for THMs (between 10 and 30 ug/1).
3. The samples were agitated at 250 to 400 rpm for about 30 min
on a gyrorotatory agitator.
4. A 5-yl aliquot from the pentane phase was injected into the
gas-liquid chromatograph.
The gas-liquid chromatograph was equipped with a linearized electron-
capture detector (Ni6 beta source) with a 20-pg detection limit. The column
was made up of 10% squalane on 80/100-mesh Chromosorb W. Column temperature
was held constant at 75°C while the detector temperature was held at 275°C.
Carrier gas and makeup gas were argon/methane mixtures. Column gas flow was
60 ml/min; makeup gas flow was 40 ml/min making a total 100 ml/min gas flow.
Concentrations were determined by integrating the peak area and comparing with
the peak area of the internal standard. Unknowns were identified by matching
their retention times to those of known standards.
12
-------
The following definition of total THM (TTHM) is used:
[TTHM] = 3[CHC13] + 3[CHBrCl2] + 3[CHBr2Cl] + 3[CHBr3]
with results presented in ymol per liter as halogen atoms. The standard
deviation of this procedure is given in Table 1.
TABLE 1. ESTIMATED STANDARD DEVIATION OF THM ANALYSIS
Sample
1
2
3
4
5
6
7
8
9
10
Average
Std. Dev.
TTHM
ymol/l-x
0.351
0.330
0.301
0.288
0.307
0.310 TTHM produced = (TTHM of sample)
0.311
0.316 - (TTHM of unchlorinated sample)
0.337
0.241
0.309
0.03
'TTHM produced
'TTHM sample TTHM unchlor. sample
/2 x 0.032 = 0.04 nmol/1
TOX Analysis
TOX analysis was performed on a prototype instrument made available by
Dohrmann Envirotech. The procedure is diagrammed in Fig. 2 and is detailed as
follows:
Organics recovery was achieved by two activated-carbon filters
connected in series. Each filter contained 40 milligrams of
powdered activated carbon. Secondary effluent samples filtered
were 25 milliliters in volume.
The carbon filters were washed with a potassium nitrate solu-
tion containing 5000 milligram per liter as nitrate to remove
inorganic halide. Initially, a four-milliliter aliquot of this
solution was used for each set of filters. Subsequently, the
volume of nitrate used was reduced to 2 milliliters.
The carbon was extruded from the filters and placed into a
quartz boat. Under a C02 atmosphere the boat was introduced
13
-------
CO,
ACTIVATED
CARBON
FILTRATION
STEP 1. \^^\
NITRATE
WASH
STEP
ORGAN ICS
RECOVERY
STEP 3.
PYROLYSIS
FILTER
T
STEP 2.
INORGANIC
CHLORIDE REMOVAL
STEP 4.
MICROCOULOMETRIC
TITRATION
4.
Figure 2. Diagram of TOX analysis.
into a 200°C zone to vaporize the volatile organics. Next, the
boat was pushed into the 800°C zone. The C02 atmosphere al-
lowed more complete recovery of bromides. Finally the atmo-
sphere was changed to a CO^/O^ mixture for completing the
combustion.
A microcoulometric titration unit served as the detector to
measure the amount of chloride and bromide released in the py-
rolysis process. In the microcoulometric titration method,
silver ion was titrated in a 70% acetic acid electrolyte solu-
tion.
Generally, 25-ml samples were filtered.
clogging due to particulates.
Higher volumes often resulted in
The precision of the TOX method was investigated by determining TOX in
ten replicate samples of secondary effluent (Table 2). The average TOX was
4.25 jjMol/1-X; the standard deviation of the individual measurements was 0.69
yMol/l-X, or 16 percent of the mean. The calculation of the amount of TOX
formed during disinfection entails subtracting the TOX concentration in the
unchlorinated sample from that in the chlorinated sample; both quantities in-
corporate a blank correction (Table 3). The variance of estimated quantity of
TOX formed by chlorination is given as:
14
-------
TABLE 2. PRECISION OF TOX MEASUREMENT
Sample
1
2
3
4
5
6
7
8
9
10
Average
Std. Dev.
TOX in
Umol/l-X
5.92
4.76
3.85
3.89
4.16
4.27
3.85
4.43
3.87
3.45
4.25
0.69
TABLE 3. BLANK CORRECTION FOR TOX
CALCULATION
TOX in
Sample yMol/l-X
1 1755
2 1.54
3 1.46
Average 1.52
Std. Dev. 0.05
_ / 2 2 2
STOX formed ~ '/ STOX sample STOX unchlor. sample TOX blank
In the case of our measurements, the standard deviation was taken to be 0.69
uMol/l-X for both the treated and untreated samples (Table 2), whereas that of
the blank was 0.05 joMol/1-X (Table 3). Hence, the standard error of estimat-
ing the amount of TOX formed was calculated as:
STVW f A " x °-69 + 4 x 0-05 " °-98
TOX rormed
This value was used in judging whether observed amounts of TOX formed should
be considered statistically significant.
15
-------
SECTION 5
CHLORINE AND CHLORINE DIOXIDE CHEMISTRY AND GENERATION
MEASUREMENTS OF CHLORINE DIOXIDE AND CHLORINE SPECIES
Several methods are available for the measurement of chlorine residuals
in wastewater. These methods have been extensively developed over many years
and the strengths and weaknesses of each are reasonably well understood. Many
of these same methods are also applicable to the measurement of chlorine diox-
ide residuals, with some modifications. Two methods for residual measurement
of chlorine dioxide were chosen for investigation in this study: (1) Diethyl-
p-phenylene diamine, ferrous ammonium sulfate (DPD-FAS) Titrimetric Method,
and (2) Amperometric Method.
Comparison of Residual Measurement Methods
In attempting to measure residual chlorine dioxide by the DPD-FAS tech-
nique (6,7,8), some difficulties were encountered. The residual as measured
by DPD was consistently higher than the chlorine dioxide dose. The stock
solution was standardized immediately before use by the iodometric method, and
the dose calculated on this basis . The discrepancy between the DPD-FAS and
the iodometric methods was observed not only in wastewater, but also in deion-
ized water, distilled water and tap water. This discrepancy was not seen when
chlorine was the oxidant of interest.
In the iodometric method, chlorine dioxide is reduced to chloride (Cl~).
However, several researchers (8,9,10) have reported that the pH used in the
iodometric analysis (pH ~ 1.8) is not low enough to assure complete reduction
of the chlorine dioxide to Cl~ (see Fig. 3). If, indeed, this is the case,
the iodometric results would give a low value for the concentration of chlo-
rine dioxide in the stock solution. The DPD method, which is performed at pH
6.2 to 6.4, is assumed to measure the reduction of chlorine dioxide to chlo-
rite, and would therefore indicate a higher value of chlorine dioxide than
that given by the iodometric analysis. It was hypothesized that the iodomet-
ric analysis was not measuring all the chlorine dioxide and hence the discrep-
ancy between the two methods.
CIO2 + e~ - - CIO2~ pH 7
Cr + 20 — pH<2
Figure 3. Reduction of chlorine dioxide as a function of pH.
16
-------
To test this hypothesis, two experiments were performed. In the first
experiment, the iodometric analysis was performed at various pH values ranging
from 2.0 to 0.8. No difference in titration value was found. In the second
experiment, the iodometric analysis performed at pH 1.8 was compared to the
amperometric analysis at pH 7. The amperometric method is analogous to the
iodometric method with these exceptions:
1. In the amperometric analysis the endpoint is detected by ob-
serving the change in current flow in the sample while adding a
strong reducing agent as titrant.
2. In the amperometric analysis, phenylarsine oxide (PAO) is the
titrant instead of sodium thiosulfate as used in the iodometric
method.
3. In the amperometric analysis the determination is performed at
pH 7. At this pH only one electron transfer is possible for
chlorine dioxide.
Since at pH 7, chlorine dioxide is reduced only to chlorite, a higher
concentration of chlorine dioxide as measured by the amperometric analysis
would indicate that chlorine dioxide was not being fully reduced to chloride
as assumed in the iodometric method. The analyses were run in distilled
water. The results appear in Table 4, and a statistical analysis appears in
Table 5.
TABLE 4. COMPARISON OF AMPEROMETRIC AND IODOMETRIC METHODS
FOR DETERMINING CONCENTRATION OF C100
Trial
1
2
3
4
5
Mean
Std.Dev.
s2
Iodometric
563.7
563.7
566.4
561.0
560.3
563.0
2.18
4.77
Amperometric
547.2
554.8
571.9
560.5
558.6
558.6
8.06
64.98
mg/1 as C102'
TABLE 5. COMPARISON OF AMPEROMETRIC AND IODOMETRIC METHODS BY
ANALYSIS OF VARIANCE
Source of
Variance
Between Methods
Error
Degrees of
Freedom
1
8
Sum of
Squares
48.8
348.7
Mean
Square
48.8
43.6
F-Ratio
1.12*
TOTAL
397.5
Not significant (p < 0.05).
17
-------
As seen in Table 5, the difference between the values obtained by the two
methods is not significant. It can be inferred from these data that the iodo-
metric method accurately measures the concentration of chlorine dioxide in
stock solutions. The inconsistency must then be in the assumed reactions of
the DPD-FAS method. It is hypothesized (although untested at this time) that
the chlorite from the reduction of chlorine dioxide reacts with the DPD rea-
gent to a slight extent. Due to the continuing problems with the DPD-FAS
method, further investigations into utilizing the amperometric method for
measurement of residual chlorine dioxide were made.
Amperometric Back-Titration
The amperometric titration measures iodine (I~) released into solution
when iodide is oxidized by any powerful oxidant such as chlorine dioxide or
chlorine. If this released I~ comes into contact with organic constituents in
the sample (as in the case or wastewater), the 1^ will react with the organ-
ics. To prevent this, a back-titration procedure is recommended, in which the
liberated Io is immediately reacted with phenylarsine oxide (PAO), which has
been added to the sample prior to the addition of iodide (as KI). The PAO is
added in excess and the unreacted PAO is back-titrated with an I~ solution of
known concentration. The concentration of chlorine dioxide (or chlorine) is
then determined by difference. To evaluate the need for this back-titration
procedure, chlorine dioxide residual measurements were made in Palo Alto
secondary effluent, comparing the amperometric method with and without back-
titration.
The results of this experiment are reported in Table 6. The experiment
entailed dosing 2 liters of Palo Alto secondary effluent to 4.50 mg/1 with
pure chlorine dioxide. The sample was then contacted for 2 minutes. After
Sample
1
2
3
4
5
6
7
8
Mean
Std.Dev., s
s2
TABLE 6.
Secondary
Forward
Titration
1.90
1.86
1.90
1.84
1.81
1.73
1.67
1.91
1.80
±0.083
0.0069
AMPEROMETRIC TITRATION EVALUATION
Filtered
Back
Titration
2.12
2.31
2.13
2.22
2.22
2.12
2.12
2.12
2.17
±0.067
0.0045
Secondary
Forward
Titration
1.62
1.62
1.52
1.52
1.52
1.39
1.39
1.27
1.49
±0.109
0.0119
Unfiltered
Back
Titration
1.88
1.97
2.07
1.88
1.88
1.88
1.88
1.79
1.90
±0.077
0.0060
Results are given as mg/1 C102 at 2-minute contact time; dose = 4.50
mg/1 CIO?- Titrations were performed at pH 7.
18
-------
the contact period, sufficient KI and pH 7 buffer were added to the sample to
react with all the chlorine dioxide present. In the case of the ^ back-
titration, sufficient PAO was added to swamp the released !«• Eight repli-
cates were run at approximately one-minute intervals. The experiment was
repeated for filtered Palo Alto secondary effluent. The values reported have
been corrected for blank titrations. In the case of the forward titration, no
correction was necessary, but for the I~ back-titration more 1^ solution was
required in the blank than was indicated by the amount of PAO added. (The 1^
solution was standardized immediately prior to use.) Although not shown in
Table 6, the blank titrations (done in triplicate) showed the same variation
as the forward or backward titration. The standard deviations shown in Table
6 are not statistically different from one another as measured by the F-test.
An analysis of variance (ANOVA) comparing the forward titration with the
I~ back-titration showed that there is a significant difference between the
values measured by the two titrations (Table 7). Based on the ANOVA results,
the amperometric back-titration method was chosen for the measurement of chlo-
rine dioxide residuals in wastewater.
TABLE 7. ANALYSES OF VARIANCE FOR AMPEROMETRIC TITRATION
Source of Variation
Degrees of
Freedom
Sum of
Squares
Mean
Square
F-Ratio
ANOVA for Filtered Secondary
Effluent:
Between forward and back-
titrations
Error
TOTAL
ANOVA for Unfiltered Secondary
Effluent:
Between forward and back-
titrations
Error
TOTAL
1
14
15
1
14
15
0.54023
0.09135
0.63158
0.67651
0.14274
0.81924
0.54023
0.006525
82.79
0.67651
0.01020
66.35
P < 0.001 that F is exceeded.
Chlorite Interference
To investigate the possible interference of chlorite in the determination
of chlorine dioxide by the amperometric method, a solution of sodium chlorite
was added to a sample containing chlorine dioxide in distilled water. Since
the sodium chlorite used was only ~ 95% pure, glycine (aminoacetic acid) was
added to the sample to react any chlorine or bromine compounds present which
would give an apparent positive result from the chlorite (7). The results of
this experiment (Table 8) showed no significant difference caused by the
presence of chlorite.
19
-------
TABLE 8. THE EFFECT OF CHLORITE ON THE DETERMINATION OF CHLORINE
DIOXIDE BY AMPEROMETRIC TITRATION
C102 + 4.33 mg/1 C102"
Sample
I
2
3
4
5
cio2
5.47
5.55
5.72
5.61
5.59
+ Glycine
5.55
5.51
5.55
5.61
5.61
Mean 5.59 5.57
Two ml of 100 g/1 solution of glycine added to 200 ml of sample.
The amperometric back-titration method for measurement of chlorine resid-
uals in wastewater is a widely used and well-accepted method (11). Since this
method was found acceptable for chlorine dioxide residual measurements, it was
also chosen for the measurement of chlorine residuals. The analyses for chlo-
rine residuals (as Total Available Chlorine) are performed at pH 4 to ensure
quantitative response from all forms of combined chlorine (11). Results from
our laboratory indicate that at pH 4 there is no need for a blank correction.
The reaction of PAO with I~ must be somewhat slower at pH 7 than at pH 4, so
that some of the added I2 titrant would be available for reaction with other
sample constituents; hence the need for blank correction in determining chlo-
rine dioxide residuals at pH 7.
MEASUREMENT OF CHLORINE DIOXIDE GENERATOR PRODUCT COMPOSITION
Chlorine dioxide for use in wastewater disinfection is generated by
either a chlorine-sodium chlorite reaction or by a mineral acid-sodium chlo-
rite reaction. Impurities of the reactants and incomplete reaction necessi-
tate a methodology to identify all possible chlorine species. The possible
chlorine species in the final products are:
1. Chlorine - C12, HOC1
2. Chlorine Dioxide - C102
3. Chlorite - C102~
4. Chlorate - C103~
5. Chloride - Cl~
Analysis of Chlorine Species
The analytical methodology developed for this study to identify reactor
product composition is essentially an iterative process in which successive
measurements include one more chlorine species than the previous measurement.
The method is as follows:
20
-------
1. Measurement of chlorine dioxide: Chlorine dioxide absorbs ultravio-
let radiation in the 240-440 nm wavelength. A wavelength of 360 nm was chosen
for chlorine dioxide determinations because this corresponds to the peak of
the absorbance curve; also, there is no interference from other chlorine
species at this wavelength.
A calibration curve was prepared by appropriate dilutions of a standar-
dized stock solution to yield concentrations in the 5 mg/1 to 60 mg/1 range.
The response of C102 is linear in this concentration range and an equation can
be derived relating concentration to absorbance. A typical concentration
equation with correlation coefficient r is given below:
mg/1 C102 = 58.26 x (absorbance) - 0.255
r2 = 0.998
The calibration equation will depend on the instrument and path length
used and should be checked often.
A sample of reaction product was diluted so that the concentration of
C10? to be analyzed was in the linear range (5-60 mg/1). The instrument was
set to zero absorbance with a blank of the dilution water just prior to sample
analysis. The concentration of C102 in the diluted solution was calculated
using the calibration equation and the appropriate dilution factor was ap-
plied. This was "Reading A."
2. Measurement of chlorine dioxide and chlorine: Chlorine dioxide and
chlorine were measured by forward amperometric titration at pH 7. An appro-
priate sample size was chosen so that no more than 5 ml of phenylarsine oxide
were used in the titration. To approximately 200 ml of distilled water, were
added 2 ml pH 7 buffer, the sample, and 2 ml of 5% KI solution, in that order.
This mixture was titrated with 0.00564N_ PAO and the resulting volume (ml) of
PAO solution was recorded as "Reading B."
3. Measurement of chlorine dioxide, chlorine, and chlorite: Chlorine
dioxide, chlorine, and chlorite were measured by iodometric titration at pH 2.
To 50 ml of distilled water were added 5 ml of concentrated acetic acid, 5 ml
of sample and approximately 1 g of KI. This mixture was allowed to react in
the dark for 5 min and was then titrated to a starch endpoint with 0.1N_ (stan-
dardized) sodium thiosulfate. This was "Reading C."
4. Measurement of chlorine dioxide, chlorine, chlorite, and chlorate:
This procedure was essentially as in Step 3 of this enumeration, with the fol-
lowing exceptions. To a 50-ml erlenmeyer flask was added 5 ml concentrated
hydrochloric acid and 0.5 g KI. To this mixture was added 5 ml of sample. The
mixture was sealed and allowed to react in the dark in a closed flask for 10
min. A blank was also carried through the procedure since KI is oxidized by
oxygen in this low pH environment (pH < 0.1). After 10 min, the mixture was
diluted to ~ 50 ml and titrated with 0.1N_ sodium thiosulfate. This reading
minus the blank value was "Reading D."
21
-------
5. Measurement of chloride: This procedure was essentially the Mercuric
Nitrate Method (408B) as given in Standard Methods (1) with the following ex-
ception: when measuring the chloride concentration in a reaction mixture con-
taining high levels of chlorine dioxide, the sample was diluted to yield a
final chlorine dioxide concentration of about 10 mg/1. To 100 ml of this
solution was added 4.0 ml of indicator-acidif ier reagent instead of 1.0 ml as
given in (1). The sample was then titrated as specified. This reading minus
the blank value was "Reading E."
All calculations, with the exception of the chlorine dioxide/ spectro-
photometric measurement and the chloride measurement were based on the equiva-
lents of reducing titrant required to react with the equivalents of oxidants
present. The equivalent weight of a compound was defined as that weight of
the compound which contains one gram atom of available electrons, i.e., equi-
valent weight equals molecular weight divided by number of electrons trans-
ferred. For chlorine dioxide, the equivalent weight is pH-dependent , because
at pH 7 chlorine dioxide is reduced to chlorite, while at pH < 2 chlorine di-
oxide reduces to chloride. Chlorite and chlorate also exhibit an equivalent
weight pH-dependency . A listing of the equivalent weights used in the calcu-
lations is shown in Table 9.
Evaluation of Chlorine Dioxide Reactor Yields and Reactor Stoichiometry
Acid Activation of Sodium Chlorite —
When chlorine dioxide is generated from sodium chlorite by acid acti-
vation, chlorine dioxide is not the only end product. The final composition
is dependent on several variables, among which are: sodium chlorite concen-
tration; purity of sodium chlorite used; acid concentration and pH of reaction
mixture; reaction time; and temperature of reaction.
Several stoichiometries have been reported in the literature for acid ac-
tivation of a chlorite solution (12). The two stoichiometries which are most
widely accepted are (shown as sulfuric acid activations):
4NaC102 + 2H2S04 = 2Na2S04 + 2C102 + HC1 + HC103 + H20 (1)
and
10NaC102 + 5H2S04 = 8C102 + 5Na2S04 + 2HC1 + 4H20 (2)
The first reaction is catalyzed by the chloride ion, which is also a
product of the reaction. Not only has it been reported that the reaction is
accelerated by the product, chloride ion, but also that the Stoichiometry is
altered to that of the second reaction (12). Although chlorine has been
reported among the reaction products of chlorine dioxide generation, most
evidence appears to be against the formation of significant amounts of chlo-
rine (12).
22
-------
TABLE 9. EQUIVALENT WEIGHTS FOR CALCULATING CONCENTRATIONS ON A MASS BASIS
pH
7
2, 0.1
7, 2, 0,
2, 0.1
0.1
Species
Species
C102
C102
.1 C12
C102~
cio3~
Calculations :
Molecular Weight
(mg/mole)
67,450
67,450
70,900
67,450
83,450
Electrons
Transferred
1
5
2
4
6
Equivalent
Weight (mg/eq)
67,450
13,490
35,450
16,863
13,908
Chlorine dioxide: mg/1 CK>2
Chlorine: mg/1 C12
6 *
Chlorite: mg/1 C102"
35'45°
eq
58.26 (absorbance) - 0.255 = A
L (-B(ml) x normality(eq/l) _ mg/1 C1°2 >
^ sample(ml) 67,450 mg/eq'
_ 16,863 mg j-C(ml) x normality(eq/l)
eq *• sample (ml)
mg/1 C12 mg/1 C102
~ 35,450 mg/eq ~ 13,490 mg/eq-'
Chlorate: mg/1 ClOo"
13,908 mg /-D(ml) x normality(eq/l)
eq *• sample(ml)
mg/1 Cl,
mg/1 CIO,
mg/1 C102
16,863 mg/eq ~ 35,450 mg/eq ~ 13,490 mg/eq-
Chloride: mg/1 Cl" - 35.450 mg/eq x E(ml) x normality(eq/l)
° sample(ml)
where A, B, C, D, E denote the respective Readings A, B, C, D, E, as defined
in the text.
Reaction Stoichiometry in Laboratory Generation —
To determine reaction stoichiometries, a mass balance on chlorine around
a reactor must be made using the analytical technique previously described.
Prior to examining a commercial reactor, a laboratory reaction mixture was
prepared, a mass balance performed, and a reaction Stoichiometry determined.
There are significant differences between laboratory reaction conditions and
those typically found in full-scale commercial reactors. Therefore, the
stoichiometries observed would not necessarily be the same. In the labora-
tory, the concentration of sodium chlorite in the reaction vessel was 0.083M,
while in commercial reactors the concentration is typically 2.0M. Concentra-
tion not only affects the time until the reaction is complete but may affect
23
-------
the stoichiometry as well. In commercial reactors the concentrated reaction
mixture is allowed to react for approximately 5 to 25 min and is then diluted
(within the reactor) before being fed to the process stream.
The initial and final compositions of the laboratory reaction mixture are
shown in Table 10. Under these conditions the laboratory reaction was com-
plete within 3 hrs; no change in composition was found after this time. The
recovered mass of chlorine represents 104% of the initial mass of chlorine;
considering experimental error, this value is not significantly different from
the expected value of 100%.
If the initial composition of the reaction mixture is equated to the
final composition, the following equation results:
0.083 C102~ + 0.0004 C103~ + 0.0177 Cl~
- 0.0037 C102~ + 0.0206 C103~ + 0.0424 Cl~
+ 0.0380 C102 + 0.0007 C12 (3)
TABLE 10. COMPOSITION OF REACTANTS AND PRODUCTS IN LABORATORY GENERATION
OF CHLORINE DIOXIDE BY SULFURIC ACID ACTIVATION OF SODIUM CHLORITE
Species
Reactants
cio2
C102~
C103~
C12
Cl"
Total
Final Composition (pH 1.75)
C102
C102~
C103~
C12*
HOC1*
Cl"
Total
Chlorine Atoms
Moles per liter
0.0
0.083
0.0004
0.0
0.0177
0.1011
0.038
0.0037
0.0206
0.0004
0.0005
0.0424
0.1056
Calculated assuming equilibrium is attained in chlorine hydrolysis,
KH = 3 x 10~4.
24
-------
Collecting terms and relating all species to a reaction involving four moles
of chlorite, analogous to Reaction 1, yields:
4 C102~ = 1.92 C102 + 1.02 C103~ + 1.25 Cl~ + 0.04 C12 (4)
This stoichiometry is essentially that of Reaction 1.
Reaction Stoichiometry in Full-Scale, Continuous Generation—
A reaction mixture from a continuous, full-size generator (commercially
available from Rio-Linda Chemical Company) was analyzed. This reactor uti-
lized the sulfuric acid activation of sodium chlorite. The reactor consists
of a section of PVC pipe, 100 mm in diameter and 660 mm long, packed with
turbulence-inducing packing. Its nominal capacity is approximately 2 kg/h
C102 (Fig. 4). The composition of reactants (calculated from the flow rate
and composition of reactant streams) and products (obtained from determination
of reactor product composition) are shown in Table 11. The sample was taken
four hrs before the analyses were performed, and it is not known to what
extent the reaction may have proceeded during transportation. Therefore, the
values reported in Table 11 should be taken as equilibrated values for the
reaction mixture at the time of analysis.
The recovered mass of chlorine atoms represents 104% of the initial mass;
this is not significantly different from 100%. The stoichiometry related to a
reaction involving four moles of C102~ is:
4 C102~ = 2.63 C102 + 0.53 C103~ + 1.01 Cl~ + 0.05 C12 (5)
which can be compared to the result from the laboratory reaction (Eq. 4), or,
on the basis of ten moles of C102~ reacting:
10 C102~ = 6.58 C102 + 1.32 C103~ + 2.52 Cl~ + 0.12 HOC1 (6)
Comparing Eqs. 5 and 6 with Eqs. 1 and 2, it appears that the reactor
stoichiometry falls between stoichiometries 1 and 2. This may indicate that
the overall reaction in an acid-chlorite generator is a combination of Reac-
tions 1 and 2. In addition, from Table 11, it can be seen that only 78% of
the added chlorite is participating in the reaction at all, and of that 78%,
66% is converted to chlorine dioxide. This gives an overall yield of 51%,
i.e., 51% of added chlorite is converted to chorine dioxide, 10% is converted
to chlorate, 22% is unreacted chlorite, and the remainder is accounted for as
chloride and a trace of chlorine. Under the most favorable conditions, the
maximum yield of chlorine dioxide found from the sulfuric acid activation of
sodium chlorite was in the 50-55% range.
25
-------
SODIUM
CHLORI
N
No CIO,
REACTION
COLUMN
PRESSURE
REGULATOR
WATER
FLOW
n
ii
n
ii
ii
VENTURI
FLOW
RATE
METERS
HCI or
No CIO VACUUM LINE
2
-ACID VACUUM LINE
SIGHT TUBE
tt&C
DRAIN
REACTION COLUMN
DIAMETER
LENGTH
CAPACITY
10 cm I.D.
66 cm
5 liters
Figure 4. Chlorine dioxide generator, acid-chlorite process.
Since it had been reported (12) that the chloride ion catalyzes the reac-
tion and alters the stoichiometry to the more favorable Reaction 2, the com-
mercial reactor under study was converted to hydrochloric acid activation from
sulfuric acid activation. To evaluate this generation process, a mass balance
on chlorine atoms was again performed around the commercial reactor. The re-
sults of this mass balance are reported in Table 12.
The recovered mass of chlorine represents 99.7% of the initial mass. If
the stoichiometry of this mixture is determined and related to an equation in-
volving 10 moles of chlorite, the following stoichiometry is found:
10 C102~ =8.04 C102 +0.28 C103~ +1.58 Cl"
26
(7)
-------
TABLE 11. COMPOSITION OF REACTANTS AND PRODUCTS FROM CHLORINE DIOXIDE
GENERATION IN A CONTINUOUS, FULL-SIZED ACID-CHLORITE REACTOR USING H2S04
Chlorine Atoms
Species Moles per liter
Reactants
C102 0.0
C102~ 0.0214
C103~ 0.0003
C12 0.0
Cl~ 0.0021
Total 0.0238
Final Composition (pH 2.1)
C102 0.0109
C102~ 0.0048
C103~ 0.0025
C12* < 0.0001
HOC1* 0.0002
Cl~ 0.0063
Total 0.0247
Calculated assuming equilibrium is attained in chlorine hydrolysis,
= 3 x 10~4.
Reaction 7 very closely approximates Eq. 2, the stoichiometry reported by
Gordon et al. (12) as optimum for acid activation of sodium chlorite.
From Table 12 it can be seen that 97% of the chlorite has reacted, com-
pared to only 77% previously reported for activation of sodium chlorite by
sulfuric acid. Of the feed chlorite 78.2% is converted to chlorine dioxide.
The average conversion of feed sodium chlorite observed in three field experi-
ments (Table 13) was 78.5%. In view of the agreement between laboratory and
field data, it is reasonable to expect consistent reactor yields of 78% con-
version of sodium chlorite to chlorine dioxide by acid activation with hydro-
chloric acid. The complete stoichiometry expected from hydrochloric acid
activation of sodium chlorite is as follows:
5NaC102 + 5HC1 = 4C102 + 5NaCl + HC1 + 2H20 (8)
27
-------
TABLE 12. COMPOSITION OF REACTANTS AND PRODUCTS FROM CHLORINE DIOXIDE
GENERATION IN A CONTINUOUS, FULL-SIZED ACID-CHLORITE REACTOR USING HC1
Chlorine Atoms
Species Moles per liter
Reactants
C102 0.0
C102~ 0.0110
C103~ 0.0000
C12 0.0000
Cl~ 0.0260
Total 0.0370
Final Composition (pH 2.2)
C102 0.0086
C102~ 0.0003
C103~ 0.0003
C12 and HOC1 < 0.0001
Cl~ 0.0277
Total 0.0369
TABLE 13. YIELD OF CHLORINE DIOXIDE FROM ACID ACTIVATION IN
FIELD EXPERIMENTS
Date
1-16-79
2-5-79
2-13-79
Average
C102~ - Feed (mg/1)
2111.4
2072.3
1454.3
conversion
C102~0utput (mg/1)
1573.4
1581.0
1231.0
% yield
74.5
76.3
84.6
78.5
Chlorine-Chlorite Generation of Chlorine Dioxide—
An alternative generation method for chlorine dioxide utilizes the reac-
tion of sodium chlorite with chlorine. A mass balance on chlorine around a
reactor using this reaction was attempted but time limitations of the analyt-
ical procedure precluded a meaningful mass balance. The simplified stoichiom-
etry for this reaction is:
28
-------
2NaC102 + C12 - 2C102 + 2NaCl (9)
From Reaction 9, theoretically 100% of the sodium chlorite can be reacted
to chlorine dioxide. An excess of chlorine above the stoichiometric require-
ment as predicted by Reaction 9 has been found to increase yields of chlorine
dioxide substantially and ensures complete conversion of sodium chlorite. The
excess chlorine (predominately as HOC1) presented the analytical problem men-
tioned above. Hypochlorite ion reacts with chlorine dioxide over time to
produce chlorate ion according to Reaction 10:
2C102 + OC1~ + H20 = 2C103~ + Cl~ + 2E+ (10)
While this reaction is not instantaneous, it is sufficiently rapid to
cause significant decreases in the chlorine dioxide concentration between
sampling time and analysis for mass balance. Therefore, samples were analyzed
immediately for chlorine dioxide to determine yields of chlorine dioxide, and
chlorite measurements were made within 6 hours to estimate extent of reaction.
Chlorine dioxide yields were found to vary from 93 to 98% based on feed
chlorite. Chlorite conversion amounted to 97% of that fed to the reactor.
Chlorine feed was approximately 4% above the stoichiometric requirement based
on chlorite.
Chlorine Dioxide from Sodium Chlorate—
Chlorine dioxide may also be generated directly from chlorate (13),
rather than passing through the intermediate step of reducing the chlorate to
chlorite. Generation of chlorine dioxide directly from chlorate is practiced
widely in the pulp and paper industry, but not in water treatment. Problems
of reactor instability reportedly are encountered. There is an economic
incentive for developing a generation process based on chlorate for use in
water and wastewater treatment. The chemical costs are considerably lower
than in processes based on chlorite.
29
-------
SECTION 6
MODELS OF DISINFECTION KINETICS
FIRST-ORDER KINETICS
To compare the relative disinfection capabilities of chlorine and chlo-
rine dioxide, an empirical model relating degree of bacteria inactivation to
the contact time and residual concentration was used in this study. The model
was first applied to disinfection to approximate kinetics that deviated
greatly from the often assumed first-order disinfection kinetics referred to
as "Chick's law."
In 1908, Harriette Chick published her "Investigation of the Laws of Dis-
infection" (14). She examined the destruction of several bacterial organisms
in buffered distilled water by phenol and other compounds. It is interesting
to note that in only one experiment, the destruction of unicellular anthrax
spores by phenol, did she observe the rate expression which carries her name.
That expression is the familiar first-order kinetic rate expression:
|| = - kN (11)
or in integrated form:
N e~kt (12)
N(0)
where N = number of surviving organisms at time = t;
N(0) = number of organisms at time = 0;
k = rate constant.
DECREASING RATE AND LAG TIME
In the majority of experiments, however, Chick observed that the rate
constant k did not remain constant and in fact decreased over the course of an
experiment. This type of relationship is generally what is observed in waste-
water disinfection: namely, a decreasing rate of bacterial inactivation as
the disinfection process proceeds (15). In wastewater disinfection, however,
the disinfection process begins after some finite time interval, the lag
30
-------
TYPES OF BACTERIAL SURVIVAL CURVES
cr
ID
en
CHICK'S
OBSERVATIONS
WASTEWATER
DISINFECTION
CHICKS LAW
LAG TIME
TIME
Figure 5. Types of bacterial survival curves,
31
-------
time. (Chick also observed a lag time with some pure cultures in clean sys-
tems.) These relationships are shown in Fig. 5.
Several hypotheses have been formulated to explain both the lag time and
the decreasing rate of disinfection:
1. The lag time is indicative of the diffusion to and transport
across the cell membrane and the reaction time for inactiva-
tion within the cell (16).
2. The organisms are not isolated but are associated in groups,
so that the internal organisms are protected by those on the
periphery of the agglomeration. The lag time is then re-
lated to the size of the aggregation and is indicative of
the time to reach the inner-most organism (17).
3. The decreasing rate is due to a decrease in the germicidal
properties of the disinfecting agent, i.e. residual (18).
4. The bacterial population has an ab origine distribution of
susceptibilities to the disinfecting agent so that the ob-
served decreasing rate is the sum of several first-order
rate constants (19).
5. The decreasing rate is the result of the bio-chemical disin-
fection reaction inducing an increased resistance in the
population of organisms, i.e. induced heterogeneity (20).
Chick postulated several mechanisms of bacterial destruction to account
for the decreasing rate of inactivation (14), and attempted to test them. She
found that the rate of inactivation she observed was related to the age of the
pure culture used and the seed material from which the culture was grown.
RETARDANT REACTION APPROACH
In the 1930s, Fair and Thomas (21) observed the same type of kinetic
behavior in studies of biochemical oxygen demand, (BOD). Fair applied an
empirical model, called the "retardant" formula, to his observations of de-
creasing rate of reaction. As applied to the BOD reaction, this rate expres-
sion is:
dL k T
L
dt ~ 1 + at
and in integrated form is :
LQ(1 + at)"k/a (14)
where L = BOD;
LQ = ultimate BOD;
Lt = BOD at time t;
32
-------
k = velocity constant;
a = coefficient of retardation; and
t = time.
An intuitive rationalization of this expression for BOD is based on the rela-
tive rates of oxidation by bacteria of the organics present in a test sample.
The least stable compounds are oxidized at a high rate early in the test. As
these compounds are depleted, the rate of reaction becomes a function of the
more stable organics and the overall rate of reaction slows. This same rela-
tionship was also applied by Fair and Geyer (22) to the removal of BOD through
a trickling filter.
Phelps (23) reported on the self-purification of streams and rivers as
regards bacterial numbers. It was found that the bacteria disappear according
to a decreasing rate of inactivation and that Fair's retardant formula (21) is
applicable to this situation also.
In 1957, at a Ciba Foundation Conference (24), a discussion between Jonas
Salk, the discoverer of polio vaccine, and Sven Card, a Swedish researcher,
centered on the kinetics of chemical virus inactivation. There was more than
just an academic interest in this area since children in the U.S.A. were being
vaccinated with formaldehyde-inactivated polymyelitis virus. Salk contended
that the inactivation proceeded according to first-order kinetics (Chick's
law), while Card presented evidence that the rate of inactivation deviated
from first order in that the rate was decreasing over time. Card then applied
a model identical to Fair's retardant formula (although Card did not call it
that, nor did he reference it) to his data and showed excellent agreement
between his experimental data and this model. In Card's formaldehyde-virus
system, he observed that both an increase in formaldehyde and an increase in
reaction time had the same effect on inactivation. He reasoned that concen-
tration (R) and time (t) should be interchangeable and should appear in the
reaction equation together as a factor "Rt," This is equivalent to the state-
ment that for a given level of inactivation Rt = constant . Card merely
stated this conclusion with no explanation or evidence to support it.
This relationship between concentration and time had been discussed by
both Chick (14) and Watson (25). Watson formulated as a general relationship
that Rnt = constant . He interpreted this relationship as a chemical
reaction between one molecule of one substance with n molecules of a second
substance, or analogous to a chemical process:
[R]°[t] = K (15)
with [t] as an artificial concentration. In kinetic terms, n depicts the
empirical reaction order with respect to disinfection concentration. Fair et
al. (26) reported some experiments on the effect of concentration on bacterial
inactivation and found n to vary from 0.8 to 1.4 for free chlorine and n
to equal 1.3 for chloramines. The value of n is probably temperature-
33
-------
dependent and organism-specific. Incorporating this concept into the retar-
dant formula, Card's model (taking n = 1) as a rate equation for inactivation
is:
dN _ k
dt 1 + a(Rt)
or in integrated form
N(t) = N(0) (1 + a(Rt))~k/a (17)
where N = number concentration of target organism;
N(t) = concentration at time t;
N(0) = initial concentration;
k/a = velocity coefficient;
R = concentration of chemical species held constant over time;
a = coefficient of retardation; and
t = time.
It is important to note that in the integration of the above rate reaction,
the concentration (R) is assumed to be constant during the entire time inter-
val of integration.
Selleck et al. (27) have applied Card's model to the disinfection process
in primary effluent. They have, however, made some modifications and relaxed
the assumptions required for integration of Card's rate equation.
Typically in wastewater disinfection, there is a lag time, t', during
which no disinfection or bacterial inactivation takes place. This is true for
constant concentration of chemical disinfectant. If, however, the disinfec-
tant concentration R is allowed to vary, then a minimum product (Rt)' must
be exceeded before inactivation takes place. Selleck et al. (27) have taken
the basic rate expression as:
|| = - kN (18)
where
k - 0 for (Rt) < (Rt)' (19)
k = k' for (Rt) = (Rt)' (20)
34
-------
and
Equations 20 and 21 above are equivalent to the statement that (Rt)' = a .
Integrating the rate expression for the general case and applying the boundary
condition that at Rt = (Rt)' ; N(t) = N(0) yields:
,_.. -k'/a
N(t) = N(0)[|gjr] (22)
N(t) = N(0) for (Rt) < (Rt)1 (23)
where all variables are defined as before.
An implicit assumption made in the integration of the above rate expres-
sion is that the concentration (R) remains constant. In the case of waste-
water disinfection with either chlorine or chlorine dioxide, the concentration
(R) decreases with time. Selleck et al. (27) use a mathematical extrapolation
to estimate the active residual concentration of chlorine after the immediate
demand of the wastewater has been satisfied and assume the bactericidal prop-
erties of the residual to remain constant.
In this study, the modified retardant formula has been used to describe
the disinfection process in several wastewaters of varying quality. In order
to incorporate into the model some means of adjusting for the difference in
quality among wastewaters and hence the oxidant demand, the residual oxidant
is measured at the actual exposure time of the bacteria. This allows the more
general application of the model and renders more meaningful the comparisons
between wastewaters of different levels of treatment. In addition, measure-
ment of residual at the end of the organism exposure time yields a lower limit
on (Rt) for the measured level of inactivation. The model used in this study
is given by Equations 24 and 25:
N(t) = N(0)P^-] for Rt > b (24)
N(t) = N(0) for Rt < b (25)
where N(0) = initial total coliform bacteria;
N(t) = total coliform bacteria at time = t;
R = residual oxidant in mg/1 as measured at time t;
b = lag time coefficient (Rt)' in mg-min/1;
35
-------
k =-k.'/a = velocity coefficient; and
t = time in minutes.
In all of the above rate expressions, the assumption has been made that
the disinfectant concentration is not a function of time. This is not the
case, but without this simplification, an expression for the residual as a
function of time is required and to date no satisfactory, generally valid
formulation exists.
It must be emphasized that the model presented for use in this study has
no rational, mechanistic basis in describing disinfection by chemical agents.
Nonetheless, it does approximate empirically the behavior of the real system
and as such provides a useful design tool.
36
-------
SECTION 7
DISINFECTION EXPERIMENTS WITH CONVENTIONAL COLIFORM INDICATORS
EXPERIMENTAL DESIGN
For evaluation of chlorine dioxide as a wastewater disinfectant, experi-
ments were performed in wastewaters from three treatment plants in the San
Francisco Bay Area. Identical experiments were run in parallel with chlorine
as the disinfecting agent in order that the relative efficiency of chlorine
dioxide could be determined. The wastewaters used were from the Palo Alto
Water Pollution Control Plant (Palo Alto), the Dublin-San Ramon Services
District Wastewater Treatment Plant (Dublin), and the San Jose-Santa Clara
Wastewater Treatment Plant (San Jose).
Treatment Plant Descriptions
Palo Alto—
The Palo Alto Water Pollution Control Plant is a non-nitrifying
activated-sludge plant with a capacity of 1.5 nr/s (35 mgd). Samples for
disinfection experiments and water-quality characterizations were of secondary
effluent before chlorination. Two sets of disinfection experiments were run
on Palo Alto wastewater. The first set (Palo Alto - 1978) were experiments in
which either chlorine or chlorine dioxide were dosed to the secondary effluent
as received from the treatment plant. For both chlorine and chlorine dioxide,
eight separate experiments were performed on five different days over a two-
week period in July of 1978. On three days, two experiments were run on the
same wastewater. The second set (Palo Alto - 1979) were experiments in which
one disinfectant, either chlorine or chlorine dioxide, was dosed to two sam-
ples of Palo Alto wastewater, one of which had been filtered. The filter
utilized was a laboratory-scale multimedia filter. The filter is shown in
Fig. 6. The filtration rate was 20 liters/hr. The filter was backwashed
immediately before each use and ripened by allowing 20 liters of wastewater
sample to pass to waste before collection of 20 liters for experimentation.
The filter was then rinsed with tap water, drained to the top of the anthra-
cite layer, and allowed to stand until it was needed for the next experiment.
The empty-bed volume is ~ 1.5 liters. These experiments were run on six days
over a three-week period in August 1979. Three experiments were run for chlo-
rine and four for chlorine dioxide, comparing their performance between fil-
tered and non-filtered secondary effluent.
37
-------
overflow
33cm(l3in)
20.3cm(8in )
I6.5cm(6.5in)
diameter
5.1 cm ( 2 in )
I anthracite
2 fine sand
3 garnet
Reservoir
mixer
53.3cm ( 21 in )
69.9cm (27.5 in)
valve to control
filtration rate
filtered
effluent
Figure 6. Multimedia filtration column.
38
-------
Dublin—
The Dublin-San Ramon Wastewater Treatment Plant is a combined-process
activated-sludge/nitrification plant. The effluent from the secondary clari-
fier is filtered through a dual-media anthracite-sand filter. The capacity of
the plant is 0.22 m^/s (5 mgd). Samples for laboratory disinfection experi-
ments and water-quality characterizations were taken after filtration and
before chlorination. These laboratory experiments were run on three days in
January and February 1979.
Experiments were also conducted at full-scale at the Dublin Plant. Plant
flows were controlled manually, as were either the chlorine feed rate or the
chlorine dioxide feed rate. For experiments with chlorine dioxide as the dis-
infectant, the chlorine feed was turned off and chlorine dioxide feed solution
was introduced into the rapid mixing chamber through the same header system
used for normal aqueous chlorine feed. This assured that both disinfectants
would be subjected to the identical conditions of mixing with the wastewater.
Experiments using chlorine dioxide were performed on the same days as experi-
ments with chlorine, allowing sufficient equilibration time between experi-
ments. Chlorine dioxide was generated in a commercially available continuous
packed reactor, utilizing the HCl-acid activation technique (2 kg/hr maximum
capacity). For doses higher than 2.5 mg/1 chlorine dioxide, 2 generators were
used in parallel and plant flow was diverted to a holding pond as required.
Samples for bacterial analysis and residual analysis were taken at the en-
trance to the contact tank, the mid-point of the contact tank and at the over-
flow weir. These points had been chosen to provide reasonable contact time
for the doses used. These experiments were run on five days from January
through April 1979.
San Jose—
The San Jose-Santa Clara Wastewater Treatment Plant is a 7.0 m3/s (160
mgd) plant with separate activated sludge and nitrification processes followed
by multimedia filtration. Samples were taken of secondary effluent, nitrified
effluent, and filtered effluent for laboratory disinfection experiments and
water-quality characterizations. Each sample point was sampled once each week
(not on the same day nor at the same time) and the experiment was repeated
three times for each sample point in June and July of 1979.
Process Sequence and Wastewater Characteristics
The flow diagrams for these three treatment plants and the points from
which samples were taken for laboratory disinfection experiments are shown in
Fig. 7. Average wastewater characteristics for each plant at each sample
point are shown in Table 14. The values given are the means of all experi-
ments at the particular sample point ± the standard deviation of the parame-
ters over the period in which experiments were performed.
Design of Experiments
Laboratory experiments were performed in a randomized complete block
design. Each experiment consisted of dosing wastewater samples with chlorine
or chlorine dioxide at three concentrations, usually 2, 5, and 10 mg/1.
Samples were taken at 5, 15, and 30-min intervals for both total coliform
39
-------
PALO ALTO WWTP
RAW
WASTE
PRIMARY
TREATMENT
CONVENTIONAL
SECONDARY
TREATMENT
DISINFECTION
sample point
DUBLIN -SAN RAMON WWTP
RAW
WASTE
EXTENDED AERATION
SECONDARY TREATMENT
WITH NITRIFICATION
FILTRATION
f
DISINFECTION
sample
point
SAN JOSE-SANTA CLARA WWTP
RAW
WASTE
PRIMARY
TREATMENT
*•
CONVENTIONAL
SECONDARY
TREATMENT
1
samp
poin
NITRIFICATION
PROCESS
s
le
t
I
am
po
FILTRATION |«j-
DISINFECTION
pie sample
nt point
Figure 7. Wastewater treatment plant flow schemes.
analysis and residual analysis. Samples for bacterial analysis were also
taken at time 0. All of the six runs within an experiment (two disinfectants
each at three doses) were done in a randomized order. From three to eight ex-
periments were performed at each sample point (Palo Alto, Dublin, and three
points at San Jose). Wastewater for experiments was collected immediately
before an experiment was conducted. The basic dose-time matrix for an experi-
ment is shown in Fig. 8. In some cases, more than one replicate experiment
could be conducted on one wastewater sample on one day. However, most repli-
cate experiments were conducted on separate days, each with fresh wastewater
samples. Therefore, the wastewater characteristics would not be identical for
all replicates of an experiment. The data from these experiments were used to
evaluate the constants of the disinfection model discussed above. In some
instances, the number of dose levels was augmented to include an additional
experiment at a dose of 1 mg/1, and the contact times were augmented with
samples at 1 and 60 min. This was done to provide an adequate range of data
for use in the model.
The Palo Alto - 1979 experiments were designed to compare one disinfec-
tant between filtered and unfiltered wastewater. The question of concern in
these experiments was the effect of the filtration process on disinfection
kinetics.
Laboratory-Scale Reactor Design
The reactor for all laboratory experiments was a specially modified four-
liter aspirator bottle. The modifications include: four indentations similar
to those of a trypsinizing flask; oxidant injector and sample ports located in
40
-------
TABLE 14. CHARACTERIZATION OF WASTEWATER EFFLUENTS USED IN DISINFECTION EXPERIMENTS
Wastewater
Palo Alto Secondary - 1978
Palo Alto Secondary - 1979
Chlorine, unfiltered
Chlorine, filtered
Chlorine Dioxide,
unfiltered
Chlorine Dioxide, filtered
Dublin Filtered - nitrified
Laboratory experiments
Field Experiments
San Jose
Secondary
Nitrified
Filtered
SS*
mg/1
17.4 ±
13.8 ±
4.4 ±
13.8 ±
3.3 ±
3.0 ±
3.6 ±
42.5 ±
18.3
3.9 ±
8.6
6.5
0.5
3.7
1.2
3.7
2.6
35.3
5.4
COD* Alkalinity*
mg/1 mg/1 as CaCO,
32.7 ± 6.2
47.6 ± 2.7
37.0 ± 4.2
52.6 ± 2.9
45.9 ± 6.0
33.4 ± 1.2
23.3 ± 6.3
68.2 ± 37.5
41.5 ± 4.6
38.2 ± 6.9
234.5
89.8
90.3
112.8
114.3
115.6
161.2
216.3
146.3
184.7
± 10.1
± 6.3
± 7.0
± 15.8
± 13.6
± 38.8
± 57.6
± 117.5
± 79.7
±84.1
PH
(Range)
6.7
6.8
7.0
6.7
7.0
6.9
6.8
7.0
6.9
7.0
to 7.4
to 7.3
to 7.4
to 7.2
to 7.1
to 7.1
to 7.3
to 7.4
to 7.3
to 7.2
NH
mg/1
30.4
24.1
23.9
29.3
29.0
0.4
0.2
16.3
0.8
0.6
oN Total Coliforms
as N log of #/100 ml
± 2.7
± 1.9
± 1.7
± 7.0
± 7.3
± 0.6
± 0.4
± 7.4
± 0.8
± 0.9
6.21
4.97
4.86
5.38
5.27
4.41
4.36
6.51
4.76
4.81
± 0.48
± 0.21
± 0.35
± 0.27
± 0.62
± 0.47
± 0.39
± 0.23
± 0.39
± 0.98
Mean ± standard deviation.
One value only.
-------
CHLORINE
DOSE
mg/l
10
CHLORINE
DIOXIDE
DOSE
mg/l
2 5 10
5 MIN
o 15 MIN
<
Z
8 30 MIN
Figure 8. Experimental dose-time matrix for chlorine-chlorine dioxide
comparison.
the most active mixing zone of the reactor, and all glass or teflon connec-
tions. The reactor was maintained at 184 kPa (12 psig) and both disinfectant
injection and sample delivery were pressure driven. Stirring was provided by
a 60-mm (2-1/2") magnetic stir bar (cylindrical) with rotational speed of ap-
proximately 600 rpm. The reactor was maintained at 24°C for all experiments.
To evaluate the dispersion efficiency of added chemicals within this
batch reactor, a dye-tracer study was performed. The dye-tracer study in-
volved dosing three liters of deionized water with a methylene blue stock
solution of 500 mg/l. Sufficient dye was added to make a final concentration
of 5 mg/l in the reactor. These conditions mimicked those in which actual
disinfection experiments were performed. Samples were taken at 2, 5, 10, 20,
40, 60, and 120 sec and analyzed spectrophotometrically for methylene blue.
The results of this experiment indicate that the half-time of response is less
than 5 seconds. This reactor is shown in Fig. 9.
Characterization of the Dublin (Full-Scale) Disinfectant Contact Tank
Since the residence time varies with the flow rate in a fixed-volume
reactor, the contact time in the disinfectant contact tank can be determined
for any flow if the contact time is known for a known flow rate. The contact
times are related by
t2Q2
(26)
where t. = contact time at flow
42
-------
DISINFECTANT
DELIVERY
PRESSURE
BULB
SUPPLY
12 psig
INDENTATIONS
FOR HIGH
SHEAR MIXING
SAMPLE
PORT
CONSTANT-
TEMP.
H20 BATH
STIR BAR
Figure 9. A rapid-mix, rapid-sampling 4-liter batch reactor.
12 psig = 184 kPa absolute pressure.
A tracer study was performed in the Dublin chlorine contact tank at a
constant flow rate of 0.14 nr/s (3.25 mgd). Sufficient rhodamine-WT dye was
injected through the chlorine injection system to yield a concentration of 100
yg/1 based on dilution in the entire volume of the contact tank. The chlorine
feed was turned off for 15 min before injection and 10 min after injection to
avoid destruction of the rhodamine-WT tracer by high concentrations of chlo-
rine. Figure 10 shows the physical layout of the chlorine contact system.
Figure 11 shows the concentration of dye as a function of time at the three
43
-------
54" PIPE
PROCESS
WATER
INJECTION
POINT
SAMPLE
POINT A
kJUNCTION
(STRUCTURE
INFLUENT
120
C=H
MPLE POINTS
SAMPLE POINT C
MEASURING WEIR
SAMPLE POINT D
Figure 10. Disinfection contact tank at the Dublin-San Ramon wastewater treatment plant.
-------
30 i—
K
W
O
O
K
O
20
10
0
0
ENTRANCE TO
CONTACT TANK
MID-POINT OF
CONTACT TANK
OVERFLOW WEIR
20
40 60
TIME—MINUTES
80
Figure 11. Results of tracer study of disinfectant contact tank.
-------
sample points, inlet to contact tank, midpoint of contact tank, and overflow
weir. The areas under the three curves (i.e., the mass of dye) are not equal.
In fact, the mass of dye as calculated under the first peak is about 80% of
the mass injected. This initial loss of 20% of dye is due to the removal of
water between the injection point and the first sample point for use as pro-
cess water throughout the plant. The continued decrease of dye mass through
the remainder of the contact tank is believed to be caused by photodecomposi-
tion and destruction of the dye by chlorine remaining in the contact tank.
Mean residence times were determined for the influent to the tank, the mid-
point, and the overflow weir. They are 12.3 min, 34.6 min, and 70.1 min,
respectively.
In addition to determining the average contact time of the contact tank,
the dye study also provided a measure of the hydraulic efficiency of the
contact tank. Figure 12 shows the concentration of dye versus time for the
overflow weir sample point. The initial appearance of tracer (t^n) indicates
the minimum contact time of the tank. In this case, t. is approximately 52
min. An exact determination cannot be made due to loss of dye as previously
mentioned. The theoretical contact time for this tank at the constant flow at
which the dye study was performed is 78.9 min, while the average contact time
calculated from Fig. 12 is 70.1 min. The ratio of average contact time to
theoretical contact time is a measure of dead space or inactive space within
the tank. For the Dublin-San Ramon contactor this ratio equals 0.89. (In the
absence of any inactive space the ratio would equal 1.0.)
Another measure of hydraulic efficiency that is widely used in chemical
engineering is the reactor dispersion number "d". The value of "d" approaches
0 for ideal-plug flow and infinity for complete mix. (For a complete discus-
sion of the reactor dispersion number, see Levenspiel (28).) The reactor
dispersion number for this contact tank is 0.0045. A study performed by the
California Department of Health in which dye studies were performed on 15
chlorine contact tanks reports the range of values for dispersion number as
0.005-0.36 and the range of values for the ratio average to theoretical deten-
tion times as 0.63-1.05 (29).
In summary, the Dublin-San Ramon disinfectant contact tank closely ap-
proximates a plug-flow reactor with very little dead space, and as such is an
example of a well-designed disinfectant contactor.
STATISTICAL ANALYSIS OF EXPERIMENTAL RESULTS
The raw data from the experiments described above are given in Appendix
A. The data from the Palo Alto - 1978 experiments were chosen for a statist-
ical analysis intended to investigate the sources of experimental error, the
relationships of dose and contact time with bacterial survival, the variation
between replicate experiments on the same wastewater (i.e., two experiments on
one day performed on different aliquots of the same wastewater sample), the
variation between experiments on wastewater from one source but sampled on
different days, and finally the difference in bacterial response as a function
of disinfecting agent, i.e., chlorine or chlorine dioxide.
46
-------
TRACER STUDY OF DISINFECTANT CONTACT TANK
3.0
2.5
K
W
O
O
K
o
2.0
1.5
1.0
0.0
— OVERFLOW WEIR
50
60 70
TIME—MINUTES
80
90
Figure 12. Dye tracer concentration measured at overflow weir.
-------
TABLE 15. MEAN LOG1Q SURVIVING TOTAL COLIFORMS IN 1978 BATCH EXPERIMENTS
WITH PALO ALTO EFFLUENT
r*/"\Ti t'ftf* i"
V^VJLl Ldl. I
Time,
min.
Disinfectant: C102
2 mg/1 Dose 0*
5
15
30
5 mg/1 Dose 0#
5
15
30
10 mg/1 Dose 0#
5
15
30
Disinfectant: C12
2 mg/1 Dose 0#
5
15
30
5 mg/1 Dose 0*
5
15
30
10 mg/1 Dose 0#
5
15
30
Mean
Value" of log^Q
(N/100 ml)
Experiment Number
5.
5.
5.
4.
5.
3.
3.
3.
5.
2.
1.
0.
5.
4.
4.
<4.
5.
3.
3.
2.
5.
2.
1.
0.
1
90
52
13
39
48
73
14
27
74
02
74
44
88
89
48
34
58
92
16
60
96
86
85
71
2
6.12
4.52
4.41
4.33
6.15
<2.0
2.00
1.86
6.25
0.87
0.54
<0.34
6.18
5.52
4.48
3.30
6.23
3.94
2.19
1.69
6.30
2.25
1.16
0.58
6
4
4
3
6
3
1
1
6
1
0
0
6
5
4
3
6
3
2
1
5
2
1
0
3
.27
.42
.31
.81
.26
.04
.70
.27
.18
.71
.32
.34
.09
.82
.86
.53
.23
.88
.80
.80
.96
.31
.62
.74
4
6.73
5.01
4.96
4.87
6.83
3.64
2.49
3.02
6.83
2.52
2.11
0.32
6.58
6.32
5.46
4.47
6.79
4.19
3.48
3.05
6.70
3.77
1.32
1.42
5
6.79
5.02
5.21
5.01
6.76
3.85
2.98
3.27
6.70
2.00
1.58
0.32
6.81
6.14
4.61
3.85
6.83
4.26
3.50
3.21
6.84
2.64
2.04
0.30
6
6.50
5.14
5.20
5.02
6.52
3.24
3.57
3.37
6.73
2.28
1.45
1.13
6.63
6.29
5.58
4.83
6.67
4.53
4.52
2.86
6.67
3.16
1.97
0.92
6
5
5
5
6
3
2
3
6
2
1
0
6
6
5
4
6
4
3
3
6
3
1
1
7
.48
.11
.23
.25
.67
.66
.52
.08
.93
.28
.19
.90
.62
.34
.48
.77
.63
.08
.80
.05
.60
.24
.97
.24
8
5.84
5.12
4.73
5.03
5.55
3.40
2.76
3.45
5.84
1.72
0.95
0.39
5.76
5.46
4.59
3.98
5.91
3.99
3.34
2.63
5.59
2.84
1.70
0.43
The raw data from which the mean values were calculated are given in Appen-
dix A, Tables A-l and A-2.
'Mean value of 8 replicates.
#.
N(0), prior to disinfection
Statistical Analysis
Preliminary statistical analyses were done on the data from Palo Alto -
1978 shown in Tables A-l and A-2 (Appendix A). To investigate any bias in the
data due to the order of experiments or sampling technique, the data from time
47
-------
Experimental Results — Palo Alto 1978
All samples for bacterial analysis were analyzed by the membrane-filter
technique for total coliform bacteria (1). Generally, four dilutions of each
sample were analyzed in order to obtain an incubated sample plate with a sig-
nificant number of colonies to count, but a sufficiently small number so that
the colonies do not overlap (usually 20-80 colonies are adequate). It is pos-
sible to have from one to four measurements of coliform bacteria for each sam-
ple. The mean values of the logarithms of the numbers of bacteria for each
sample (Palo Alto 1978) are given in Table 15.
Statistical Analysis
Preliminary statistical analyses were done on the data from Palo Alto -
1978 shown in Tables A-l and A-2 (Appendix A). To investigate any bias in the
data due to the order of experiments or sampling technique, the data from time
0 were subjected to an analysis of variance (ANOVA). The grouping variables
used were wastewater, disinfectant, and disinfectant dose. In setting up the
ANOVA in this way, the difference between coliform density at time 0 with re-
spect to wastewater could be tested for all experiments. The results of this
ANOVA are shown in Table 16.
The only effect that is significant is that of wastewater, indicating
that the variance in bacterial numbers at time 0 for each experiment is due
only to the difference inherent in the wastewater used for that particular
experiment. No significant correlation was found between measured wastewater
parameters (total filterable residue, COD, alkalinity, pH, ammonia-nitrogen)
and bacterial counts.
The next step in the statistical analysis was to investigate any differ-
ences between experiments performed on the same wastewater sample, that is the
reproducibility of experiments. Again, analysis of variance was used with
grouping variables of wastewater, experiment, disinfectant, disinfectant dose,
and contact time. The ANOVA is reported in Table 17. The dependent variable
was logiQ[N(t)]. As seen in this ANOVA, all variables except experiment had a
significant effect on the number of surviving bacteria. (These effects will
be examined in subsequent ANOVAs.) This implies that the duplicate experi-
ments run on the same wastewater are not significantly different from each
other. For this reason, the experiments were no longer grouped according to
wastewater, so that variations seen between experiments in subsequent statist-
ical analyses are due to variations in the wastewater. In other words, the
wastewater variable is redundant in that the experiments differ only due to
the effects of wastewater composition.
The statistical analyses shown in Tables 16 and 17 have been calculated
using the logarithms of bacterial numbers from each sample (Appendix A) as the
dependent variable. In order to compare the bactericidal efficiency of chlo-
rine and chlorine dioxide and to examine the relationships of disinfectant
doses and contact times to bactericidal efficiency, a survival ratio for each
experiment was calculated and the logarithms of these survival ratios were
substituted as the dependent variable.
48
-------
TABLE 15.
MEAN LOG1Q SURVIVING TOTAL COLIFORMS IN 1978 BATCH EXPERIMENTS
WITH PALO ALTO EFFLUENT
Contact-
Time,
min.
Mean Value1" of Iog10 (N/100 ml)
Experiment Number
3456
Disinfectant:
2 mg/1 Dose Of
5
15
30
5 mg/1 Dose 0*
5
15
30
10 mg/1 Dose 0#
5
15
30
5
5
5
4
5
3
3
3
5
2
1
0
.90
.52
.13
.39
.48
.73
.14
.27
.74
.02
.74
.44
6.
4.
4.
4.
6.
<2.
2.
1.
6.
0.
0.
<0.
12
52
41
33
15
0
00
86
25
87
54
34
6
4
4
3
6
3
1
1
6
1
0
0
.27
.42
.31
.81
.26
.04
.70
.27
.18
.71
.32
.34
6.73
5.01
4.96
4.87
6.83
3.64
2.49
3.02
6.83
2.52
2.11
0.32
6.79
5.02
5.21
5.01
6.76
3.85
2.98
3.27
6.70
2.00
1.58
0.32
6.50
5.14
5.20
5.02
6.52
3.24
3.57
3.37
6.73
2.28
1.45
1.13
6.48
5.11
5.23
5.25
6.67
3.66
2.52
3.08
6.93
2.28
1.19
0.90
5.84
5.12
4.73
5.03
5.55
3.40
2.76
3.45
5.84
1.72
0.95
0.39
Disinfectant:
2 mg/1 Dose Off
5
15
30
5 mg/1 Dose 0#
5
15
30
10 mg/1 Dose 0#
5
15
30
5
4
4
<4
5
3
3
2
5
2
1
0
.88
.89
.48
.34
.58
.92
.16
.60
.96
.86
.85
.71
6.
5.
4.
3.
6.
3.
2.
1.
6.
2.
1.
0.
18
52
48
30
23
94
19
69
30
25
16
58
6.09
5.82
4.86
3.53
6.23
3.88
2.80
1.80
5.96
2.31
1.62
0.74
6.58
6.32
5.46
4.47
6.79
4.19
3.48
3.05
6.70
3.77
1.32
1.42
6
6
4
3
6
4
3
3
6
2
2
0
.81
.14
.61
.85
.83
.26
.50
.21
.84
.64
.04
.30
6.63
6.29
5.58
4.83
6.67
4.53
4.52
2.86
6.67
3.16
1.97
0.92
6.62
6.34
5.48
4.77
6.63
4.08
3.80
3.05
6.60
3.24
1.97
1.24
5.76
5.46
4.59
3.98
5.91
3.99
3.34
2.63
5.59
2.84
1.70
0.43
The raw data from which the mean values were calculated are given in Appen-
dix A, Tables A-l and A-2.
'Mean value of results from one to four dilutions.
*N(0), prior to disinfection
49
-------
TABLE 16. ANOVA FOR BACTERIAL ANALYSIS BEFORE DISINFECTION
USING LOG1Q[N(0)] AS DEPENDENT VALUE
Source of
Variance
Wastewater (Xi)
Disinfectant (X2)
Dose (X3)
Interactions
\X-t ) x ^Xn )
/v \ v /v "\
\^]_) x \A3/
(X2) x (X3)
(Xt) x (X2) x (X3)
Error
Degrees of
Freedom
4
1
2
4
8
2
8
51
Sum of
Squares
13.2259
0.0002
0.1625
0.4077
0.4897
0.1686
0.2568
2.0911
F-Value
80.64
0.00
1.98
0.25
. 1.49
2.06
0.78
Probability F
Exceeded
0.0001
0.9516
0.1483
0.9092
0.1831
0.1384
0.6197
N(0) = number of bacteria before disinfection.
TABLE 17. ANOVA FOR EXAMINING EXPERIMENTAL REPRODUCIBILITY
Source of
Variance
Time (XL)
Level (X2)
Disinfectant (X.,)
Waste (X4)
Experiment (Xe)
it
Interactions
(V x (X2)
(X,) x (X3)
/v \ x fy \
{A.-^} x ^A4J
(Xj_) x (Xc)
(X2) x (X3)
(X2) x (X4)
(X2) x (X5)
(X3) x (X4)
(X3) x (X5)
Error
Degrees of
Freedom
2
2
1
4
3
4
2
8
6
2
8
6
4
3
355
Sum of
Squares
48.247
781.544
7.569
40.366
0.553
9.229
17.296
2.607
2.038
3.102
2.896
1.774
4.430
1.5064
39.428
F-Value
217.20
3518.42
68.15
90.86
1.66
20.77
77.86
2.93
3.06
13.97
3.26
2.66
9.97
4.52
Probability F
Exceeded
0.0001
0.0001
0.0001
0.0001
0.1735
0.0001
0.0001
0.0035
0.0063
0.0001
0.0013
0.0154
0.0001
0.0042
Interactions of order higher than 2 are included in the error term.
The survival ratio was defined as:
Survival ratio = N(t)/N(0)
(27)
50
-------
where N(t) = number of bacteria measured at time t,
N(0) = number of bacteria measured at time 0.
The logarithm of survival ratio is then
log[N(t)/N(0)] = log[N(t)] - log[N(0)] . (28)
Log[N(t)] was calculated as the arithmetric mean of the logarithms of bacter-
ial numbers from each sample (see Table 15) for each time t. The log[N(0)]
was similarly calculated at time 0. This calculation yields the geometric
mean of the replicate analyses for bacterial numbers.
Experiment, disinfectant, disinfectant dose, and time were used as inde-
pendent variables. The resulting ANOVA is shown in Table 18. This ANOVA
reveals that all of the main effects—disinfectant, disinfectant dose, contact
time, and experiment—were highly significant, as evidenced by the fact that
P(F > Fcalc) = 0.0001 in all cases.
With respect to the comparison between chlorine dioxide and chlorine, the
ANOVA (Table 18) implied that there was a significant difference between the
two disinfectants. The implications of this significant difference are dis-
cussed below. The significant effects of contact time and dose were expected:
The survival ratio declined with increasing disinfectant dose and contact time
for both disinfectants. The significant difference between experiments is be-
lieved to be attributable to variations in properties of the wastewater.
Several of the second-order interactions were found to be significant
(Table 18), notably those between the disinfectant variable on the one hand
and the disinfectant dose, the contact time, and experiment on the other hand.
Clearly, the difference between disinfectants with respect to coliform inacti-
vation was not the same at all doses, at all contact times, and in all waste-
water samples. Also, the interaction between disinfectant dose and contact
time was found to be significant, indicating that the effect of dose on coli-
form inactivation by a given disinfectant depended significantly on the con-
tact time and vice versa. None of the third-order interactions was found to
be significant.
The meaning of the significant effects can be interpreted by comparing
the means of log[survival ratio] for chlorine dioxide with those for chlorine
for various combinations of dose and contact time (Table 19). The values in
Table 19 represent means of the eight experiments. The mean values in Table
19 show the observed differences between chlorine and chlorine dioxide and
between dose levels and contact times for a given disinfectant, corresponding
to the ANOVA main effects in Table 18. From Table 19, the difference between
disinfectant doses is obvious, as is the difference between contact times.
The significant difference between disinfectants seen in the ANOVA of Table 18
is modified by the ANOVA interactions. The relationship of the difference
between disinfectants to the values of dose and contact time can be better
understood with the help of Figs. 13, 14, and 15.
51
-------
TABLE 18. ANOVA FOR COMPARING CHLORINE AND CHLORINE DIOXIDE AS
BACTERICIDES IN PALO ALTO EFFLUENT
Source of
Variance
Disinfectant (X^)
Dose (X2)
Experiment (X3)
Contact Time (X4)
it
Interactions
(X^ x (X2)
(Xj^) x (X3)
(X1) x (X4)
(X2) x (X3)
(X2) x (X4)
(X3) x (X4)
(X,) x (X2) x (X3)
(X,) x (X3) x (X.)
(X2) x (X3) x (X4)
Error
Degrees of
Freedom
1
2
7
2
' 2
7
2
14
4
14
14
14
28
32
Sum of
Squares
4.858
279.297
21.869
37.341
1.339
2.109
6.423
2.463
3.403
1.253
0.904
1.788
3.134
4.052
F-Value
38.36
1102.76
24.67
147.43
5.29
2.38
25.36
1.39
6.72
0.71
0.51
1.01
0.88
Probability F
Exceeded
0.0001
0.0001
0.0001
0.0001
0.0104
0.0445
0.0001
0.2146
0.0005
0.7515
0.9096
0.4691
0.6277
Fourth-order interaction neglected.
TABLE 19. MEAN LOG[SURVIVAL RATIO] OF TOTAL COLIFORMS: COMPARISON OF
DISINFECTANTS AT DIFFERENT COMBINATIONS OF DOSE AND CONTACT TIME
Mean* of log[Survival Ratio] = log[N(t)/N(0)]
Chlorine Dose
mg/1
Chlorine Dioxide Dose
mg/1
^UUCctUL
Time
5 minutes
15 minutes
30 minutes
2
-0.49
-1.38
-2.18
5
-2.23
-2.98
-3.72
10
-3.44
-4.62
-5.53
2
-1.34
-1.43
-1.61
5
-3.01
-3.69
-3.50
10
-4.40
-5.09
-5.81
Calculated means of eight experiment replicates.
Chlorine and chlorine dioxide are compared by dose at 30-min contact
times in Fig. 13. These values are means over eight experiments. The values
in parentheses are the means of the residual over the eight experiments.
Chlorine gives a lower log[survival ratio] at the 2- and 5-mg/l doses, while
chlorine dioxide gives a lower log[survival ratio] at 10-mg/l dose. This
illustrates the disinfectant dose interaction mentioned above. The reason for
52
-------
(.49)
o
DC
CO
o -
• CHLORINE
DIOXIDE
|^ CHLORINE
( ) = RESIDUALS
1.18)
.78)
Figure 13.
0
-1
-2
-3
-4
5
i
-6
7 2 5 10
DISINFECTANT DOSE (mg/l)
Comparison of chlorine and chlorine dioxide bactericidal
effectiveness at 30-min contact time.
(8.08)
this dose-dependent disinfectant difference can be found in the die-away
curves of chlorine and chlorine dioxide. Figure 14 shows typical chlorine and
chlorine dioxide die-away curves in secondary effluent. The initial oxidant
demand reduces the disinfectant concentration significantly. Hence, the rate
of disinfection will be slowed.
When disinfectants are compared with respect to contact times, the disin-
fectant-contact time interaction can be evaluated. In Fig. 15 chlorine and
chlorine dioxide are compared by contact time at 10-mg/l dose. This figure
indicates a different time-dependent mode of action for chlorine versus chlo-
rine dioxide. Chlorine dioxide effects a major portion of bacteria inactiva-
tion at shorter times, while chlorine requires longer contact times to achieve
the same degree of disinfection.
Summary
The conclusions concerning the comparative bactericidal action of chlo-
rine and chlorine dioxide derived from the statistical analysis of the data on
Palo Alto effluent are as follows:
53
-------
14
•=• 8
<
D
== 6
V)
LLJ
cc
I
I
CHLORINE
DIOXIDE
I
10 15 20
TIME IN MINUTES
25
30
INITIAL DOSE OF 12.5 mg/l FOR BOTH
CHLORINE AND CHLORINE DIOXIDE
Figure 14. Chlorine and chlorine dioxide die-away in Palo Alto secondary
effluent.
1. Both chlorine and chlorine dioxide give decreased survival
ratios when dose or contact time is increased.
2. Although some variations exist, chlorine and chlorine diox-
ide give essentially the same survival ratios when compared
on a mass dose basis at 30-minute contact time.
3. Chlorine dioxide is a more rapid disinfecting agent, effect-
ing greater bacterial inactivation than chlorine at the
shorter contact times.
4. Comparing chlorine and chlorine dioxide on a residual basis,
chlorine dioxide effects the same survival ratio as chlorine
but with a much lower residual concentration.
THE EFFECT OF PROCESS SEQUENCE ON BACTERIAL NUMBERS
The experiments with San Jose effluents provided data for evaluation of
process sequence effects on disinfection kinetics. Additionally, analysis of
the bacterial concentrations at each sample point permitted evaluation of pro-
cess sequence effects in removal of bacteria before disinfection. Intuitively,
54
-------
-3
o
I-
cc
cc
13
-4
O -5
-6
(8.64)
CHLORINE
DIOXIDE
CHLORINE
= RESIDUALS
(6.
5 15 30
CONTACT TIME (MINUTES)
Figure 15. Comparison of chlorine and chlorine dioxide bactericidal effec-
tiveness at 10-mg/l dose. Values of residual concentration are
given in mg/1.
decreases in bacterial numbers would be expected to result from additional
treatment, due to physical processes that remove or select for particulates.
For example, the clarification step of activated-sludge treatment selects for
those microorganisms that aggregate into particulates with a density greater
than that of water. Therefore, any bacteria (including total coliform) that
exhibit aggregation or "clumping" would be expected to be removed in the
process.
The logarithms of the initial bacteria counts from each of the sample
points at San Jose were compared. Summary statistics are shown in Table 20
and an ANOVA is shown in Table 21.
55
-------
TABLE 20. INITIAL TOTAL COLIFORMS IN UNDISINFECTED EFFLUENTS FROM THE
SAN JOSE WASTEWATER TREATMENT PLANT
Sample Point
Statistic
Mean, log10[N(0)]
Variance
Standard Deviation
Number of Samples
Secondary
Effluent
6.51
0.0548
0.2342
18
Nitrified
Effluent
4.76
0.1501
0.3875
18
Filtered
Effluent
4.81
0.9516
0.9755
18
TABLE 21. ANOVA OF SAN JOSE INITIAL BACTERIA NUMBERS
Source
Treatment Level
Error
Degrees of
Freedom
2
51
Sum of
Squares
35.5562
20.8183
Mean
Square
17.7781
0.4082
F-Ratio
43.35
Total
53
56.3745
As shown by the F-value [P(F > 43.35) < 0.001] of the ANOVA in Table 21,
there is a highly significant difference in initial bacteria numbers among the
treatment levels. To identify statistically which means are different from
others, critical sums of squares were computed and comparisons were made (a_
posteriori test) on the means of Table 20. These tests showed a statistically
significant difference between bacterial numbers in the secondary effluent and
nitrified or filtered effluent. No difference was found between nitrified
effluent and filtered effluent. This result was surprising. While removal of
bacteria through the nitrification process might be due to the final clarifi-
cation step or predation, it was expected that greater reductions would occur
in filtration than were observed. If the physical processes of particulate
removal are analyzed and filter efficiency is computed as a function of par-
ticulate size (30), a minimum efficiency of removal is found for particulates
on the order of 1 ym. An individual coliform bacteria cell is in the size
range of 1-5 urn (31). It can be inferred from this analysis that the bacteria
in the nitrified effluent exist as singlets or small aggregates.
A similar analysis for the Palo Alto - 1979 experiments is shown in
Tables 22 and 23. As shown, there is no significant difference [P(F > 0.6588)
= 0.429] between the initial bacteria numbers of secondary effluent and lab-
filtered secondary effluent. It can be inferred from these data that the bac-
teria in secondary effluent are also mono-dispersed or nearly so.
At most wastewater treatment plants (all in this study), it is common
practice to use chlorinated finished water as process water throughout the
treatment process. For example, both the Dublin and San Jose plants use water
from the chlorine contact tank (8-10 mg/1 total chlorine residual) to backwash
56
-------
TABLE 22. LOGS OF INITIAL BACTERIAL NUMBERS AT THE
PALO ALTO WASTEWATER TREATMENT PLANT
Statistic
Secondary Effluent,
Unfiltered
Secondary Effluent,
Lab-Filtered
Mean
Variance
Standard Deviation
Number of Samples
4.97
0.1199
0.3462
9
4.86
0.0423
0.2056
9
TABLE 23. ANOVA OF PALO ALTO INITIAL BACTERIAL NUMBERS
Source
Treatment Level
Error
Total
Degrees of
Freedom
1
16
17
Sum of
Squares
0.0601
1.4593
1.5194
Mean
Square
0.0601
0.0912
F-Ratio
0.6588
the filters. This backwash water is settled and blended with the influent to
the primary sedimentation basin. Dublin also uses a small fraction of the
chlorinated water (12-14 mg/1 total chlorine residual) from the chlorine
rapid-mix tank for foam control in the activated-sludge process. The dose and
form of chlorine that is applied to the wastewater in processes other than
disinfection is difficult, if not impossible, to determine. Additionally, the
possible effects of low doses of chlorine species on the morphology of the
coliform bacteria group and the potential consequences for disinfection kine-
tics are largely matters of conjecture. One area of concern in comparing
chlorine and chlorine dioxide as disinfectants in real wastewaters is the
question of induced resistance. If a bacterial population has an induced
resistance to chlorination due to low-level exposure (such as cited above),
the bacteria may appear relatively more sensitive to chlorine dioxide than to
chlorine, a bias which may disappear after long-term use of chlorine dioxide.
COMPARISONS AMONG WASTEWATERS
The data (Appendix A) from the experiments described above were used to
determine the coefficients of the model used in this study. The coefficients
were determined by linear regression of log(residual-time product) on log sur-
vival ratio. This regression yields:
log[N(t)/N(0)] = log bf + k log(RT)
= k log[RT'(b')1/k]
(29)
(29')
or
57
-------
N(t)/N(0) = [RT'(b')1/k]k (30)
where
V = b~k
so that
N(t)/N(0) = [^|]k (31)
An analysis of the standardized residuals from the regressions of each
experiment showed in all cases except chlorine dioxide disinfection of Palo
Alto wastewater - 1979, the simple regression model
y = 3Q + exx + e (32)
was adequate in describing the relationship between survival ratio and
residual-time product (i.e., the standardized regression residuals were nor-
mally distributed). For the chlorine dioxide disinfection of Palo Alto waste-
water - 1979, the doses and sample times provided low enough residual-time
products, so that bacterial samples were taken before any inactivation had oc-
curred. To remove these points from the linear regression, a 95% confidence
interval around N(t)/N(0) = 1.0 was estimated using the variance of the
membrane-filter technique, as calculated from replicates of zero-time bac-
terial numbers. Any survival ratios which fell into this interval were ex-
cluded from the regression, and analysis of the resulting standardized
regression residuals showed conformity to the regression model.
The coefficients of the disinfection model determined by linear regres-
sion are shown in Table 24. Log-log plots of the model predictions and data
points for each experiment are shown in Appendix B.
The San Jose treatment plant had brought their separate nitrification
process and the filtration process on line only six months prior to our labor-
atory disinfection studies. Complete nitrification of the wastestream was not
always achieved during the sample period, so that low levels of ammonia were
measured in both the nitrified effluent and the filtered effluent. The model
coefficients reported in Table 24 for chlorine disinfection of San Jose nitri-
fied and filtered effluents are from linear regressions with all data from
these sample points. If the experiments are segregated according to the chlo-
rine: ammonia ratios, then the experiments in which free chlorine is the pre-
dominant disinfecting agent can be separated from those experiments in which
disinfection is due to combined chlorine. Extensive research has been repor-
ted in the literature on the reactions of chlorine with ammonia, the break-
point phenomenon (11). For complete oxidation of ammonia to nitrogen gas, the
following stoichiometry follows:
58
-------
TABLE 24. VALUES OF FITTING CONSTANTS FOR CHLORINE AND CHLORINE DIOXIDE IN THE DISINFECTION MODEL
VO
Chlorine Dioxide
Experimental
Description
Palo Alto - 1978
Palo Alto - 1979
Unfiltered
Palo Alto - 1979
Lab Filtered
95% CI*
b for b k
1.56 0
0.67 0
0.004 0.
Palo Alto Unfil- 1.29 0
tered 78 & 79 Comb.
Dublin-San Ramon
Lab
Dublin-San Ramon
Field
San Jose
Secondary
San Jose
Nitrified
San Jose
Filtered
CI = confidence
^For definitions
r = correlation
**
0.14 0
0.57 0
0.89 0
0.78 0
0.003 0.
interval.
of b and
coefficient
.99 to
2.47
.32 to
1.41
.002 to
0.010
.88 to
1.91
.07 to
0.28
.37 to
0.87
.33 to
2.43
.18 to
3.30
,001 to
0.01
k , see
-2.90
-2.30
-1.21
-2.75
-1.84
-2.20
-2.18
-2.06
-1.13
Eqs .
95% CI t1i
for k r' n b
-2.49 to 0.86 72 5.17
-3.30
-1.78 to 0.86 29 2.21
-2.83
-1.03 to 0.89 44 0.84
-1.40
-2.42 to 0.86 101 3.95
-3.07
-1.15 to 0.78 22 0.65
-2.51
-1.86 to 0.84 69 1.67
-2.54
-1.41 to 0.78 21 4.06
-2.95
-1.23 to 0.75 19 2.70
-2.88
-0.74 to 0.81 18 0.59
-1.51
24 and 25.
95% CI
for b
2.72 to
7.62
0.61 to
5.82
0.38 to
1.83
2.62 to
5.95
0.37 to
0.86
0.56 to
4.93
1.50 to
10.98
0.83 to
8.81
0.13 to
2.75
Chlorine
95% CI
k for k r
-3.15 -2.82 to 0.92
-3.47
-2.22 -1.67 to 0.83
-2.76
-2.10 -1.65 to 0.86
-2.54
-2.79 -2.50 to 0.88
-3.07
-1.62 -1.31 to 0.91
-1.92
-1.79 -1.32 to 0.93
-2.27
-2.82 -2.14 to 0.86
-3.51
-2.58 -1.67 to 0.79
-3.48
-2.12 -1.16 to 0.69
-3.07
**
n
72
31
32
103
26
12
26
24
21
for log-log transformation of raw data.
n = number of observations.
-------
2NH3 + 3HOC1 * N2 + 3H30+ + 3C1 (33)
The theoretical weight ratio of C1:N is 7.5 for completion of this reaction.
However, the stoichiometry is a simplification of a complex series of reac-
tions, all of which show some pH dependency (11,32). Weight ratios of C1:N up
to 10:1 are required in some situations to achieve breakpoint, depending on
the pH of the reaction and the aqueous system involved. For our purposes, a
C1:N ratio of less than 6.25 was chosen as indicative of combined chlorine re-
sidual, while a C1:N ratio of 12.0 was interpreted as indicating predominately
free chlorine.
The chlorine to ammonia nitrogen ratios for the San Jose nitrified and
filtered experiments are shown in Table 25. Based on these data, the follow-
ing decisions were made to classify the San Jose effluent samples into the
"free residual" or combined residual categories. In nitrified effluent,
Experiment 1 at all chlorine doses and Experiment 2 at 2 mg/1 dose were
considered to have a combined chlorine residual; the rest with the exception
of Experiment 2 at 5 mg/1 dose were considered to have free chlorine resid-
uals. Experiment 2 at 5 mg/1 dose was excluded from the analysis altogether
because of the uncertainty of residual composition near breakpoint. In fil-
tered effluent, all Experiment 1 doses were considered as below breakpoint,
the remaining experiments were used as beyond breakpoint. Linear regressions
of these subgroups yielded model coefficients which are compared in Table 26
with the model coefficients of all experiments within the wastewater of inter-
est. These relationships are shown graphically in Figs. 16 and 17 for nitri-
fied and filtered San Jose wastewater, respectively.
The nitrified and filtered experiments in which the C1:N ratios indicate
disinfection with combined chlorine are compared with disinfection by chlorine
in secondary effluent (C1:N ratios « 6.0, i.e. combined chlorine) in Fig. 18.
The three curves are essentially identical. To simplify further discussion,
"secondary effluent" will be used in referring to disinfection by combined
chlorine whether in partially nitrified or in conventional secondary effluent,
and will be represented by the curve from the secondary effluent experiments.
TABLE 25. CHLORINE TO AMMONIA NITROGEN RATIOS FOR SAN JOSE EXPERIMENTS
Chlorine : Ammonia-N
Experiment
Nitrified Effluent
1
2
3
Filtered Effluent
1
2
3
NH3-N
mg/1
1.7
0.7
< 0.04
1.6
0.12
< 0.04
2 mg/1
1.18
2.86
50
1.25
16.67
50
Dose
5 mg/1
2.94
7.14
125
3.13
41.67
125
10 mg/1
5.88
14.29
250
6.25
83.33
250
60
-------
TABLE 26. COMPARISON OF MODEL COEFFICIENTS, SEGREGATING DISINFECTION
WITH FREE VS COMBINED CHLORINE
Experiment
San Jose Nitrified
All data
Combined residual data
Free residual data
San Jose Filtered
All data
Combined residual data
Free residual data
b*
2.70
6.74
1.33
0.59
5.24
0.44
k*
-2.58
-3.17
-2.40
-2.12
-3.25
-2.16
rt
0.79
0.91
0.87
0.69
0.80
0.77
n
22
11
11
21
6
15
For definitions of b and k , see Eq. 24.
'r = correlation coefficient.
"Nitrified effluent" and "filtered effluent" will be understood to imply dis-
infection by free chlorine and will be represented by the free chlorine curves
in €Figs. 16 and 17.
10"
10
-2
o
55
O
10
-3
10
-4
10
-5
g
co
10
-6
ALL DATA
FREE
CHLORINE
RESIDUAL
COMBINED
CHLORINE
RESIDUAL
10J
10
2
RESIDUAL—TIME IN MG-MIN/L
Figure 16. Coliform interaction by free and combined chlorine in San Jose
nitrified effluent.
61
-------
10"1 b-
10'
10'
s io"
10"
g 10"
ALL DATA
\
\
\
FREE
CHLORINE
\
COMBINED
\ CHLORINE
RESIDUAL \\\ RESIDUAL
i i i i 11
i i i i 111
10° IO1 IO2
RESIDUAL—TIME IN MG-MIN/L
Figure 17. Coliform inactivation by free and combined chlorine in San Jose
filtered effluent.
10-1
10~2
10~3
10'
10
-6
SECONDARY \\\
EFFLUENT
NITRIFIED EFFLUENT
(COMBINED CHLORINE)
FILTERED
EFFLUENT \
(COMBINED CHLORINE)
i i i i 1111
10
0
101
10*
Figure 18.
RESIDUAL — TIA1E IN MG-MIN/L
Coliform inactivation by combined chlorine in San Jose wastewater
at three sampling points.
62
-------
EFFECTS OF TREATMENT LEVEL ON THE DISINFECTION PROCESS
Disinfection with Chlorine
The San Jose experiments provide comparisons of disinfection efficiency
in wastewater effluents of progressively increasing quality. As discussed
above, a significant difference was found in the initial bacterial concentra-
tion between secondary and nitrified wastewater at San Jose. This finding,
coupled with the observation that combined chlorine disinfection in partially
nitrified wastewater and in partially nitrified, filtered wastewater was
nearly identical to disinfection in secondary effluent, suggests that disin-
fection kinetics are not affected by variations in initial bacterial concen-
tration of the magnitude encountered in this study.
Figure 19 compares the disinfection process by chlorine at three treat-
ment levels in San Jose wastewater, after adjustment of the data base by
deleting combined chlorine residual data points in the "nitrified" and "fil-
tered" categories. Hence the "nitrified" and "filtered" categories correspond
to a free chlorine residual, whereas the "secondary" category corresponds to a
combined chlorine residual. Superior disinfection efficiency is observed when
the wastewater has been nitrified compared to conventional secondary effluent.
This increased efficiency is believed to be due to the superior germicidal
properties of free chlorine in nitrified effluent (11). A further increase in
disinfection efficiency is found in comparing the nitrified and filtered ef-
fluents. A decrease in the "non-target organism" chlorine demand (that demand
not associated with bacterial inactivation) might account for the increased
efficiency in filtered effluent vs nitrified effluent. The one-hour chlorine
demands, however, were found to be equivalent or slightly higher in the fil-
tered effluent. The one-hour demands for chlorine in filtered and nitrified
effluents are shown in Table 27.
The Palo Alto - 1979 experiments also indicate an increase in disinfec-
tion efficiency due to filtration, shown in Fig. 20. Hence, the increase in
disinfection efficiency of chlorine achieved by prior filtration is not depen-
dent on the form of chlorine responsible for the bacterial inactivation and
cannot be explained by the removal of bacteria associated with solids, since
no difference was found in initial bacterial concentration between nitrified
and filtered effluents in San Jose wastewater, or between Palo Alto secondary
and Palo Alto filtered secondary. Moreover, the increased efficiency is not a
result of decreased chlorine demand, since comparison of one-hour demands in
both San Jose and Palo Alto wastewaters (see Table 27) indicate that the fil-
tered wastewater demand is equal to—if not greater than—the non-filtered
demand.
Disinfection efficiency is defined in relation to the residual-time
product required to achieve a given level of bacterial inactivation as
measured by the survival ratio N(t)/N(0) . An increase in disinfection
efficiency corresponds to a decrease in residual-time product required to
achieve the same survival ratio.
63
-------
10
10
-1
-2
10"
o 10
-4
10
-5
10
-6
CO
NITRIFIED \
SECONDARY
j I
0
10
RESIDUAL—TIME IN MG-MIN/L
Figure 19. Coliform inactivation by chlorine in San Jose wastewater at
various levels of treatment after segregating data base
according to the form of chlorine residual.
TABLE 27. ONE-HOUR CHLORINE DEMAND
Demand, mg/1
Wastewater and Treatment Level
San Jose, Nitrified
San Jose, Filtered
Palo Alto - 79, Unfiltered Secondary
Palo Alto - 79, Filtered Secondary
2 mg/1
1.71
1.73
1.01
1.38
Doses
5 mg/1
4.36
4.28
1.72
2.32
10 mg/1
6.52
6.92
3.12
4.05
The one-hour demands are calculated in the manner of Feben and Tarus
(33) from residual measurements made over the course of the experi-
ments. This method involves calculation of the demand at time t as
dose minus residual. A regression analysis of log(chlorine demand) on
log(time) yields an equation for demand of the form:
~ one-hour *-
where D = chlorine (or chlorine dioxide) demand; Done-hour = one~
a = a coefficient.
hour demand, mg/1; t = time, hours; and
64
-------
10-1
10"2
10~3
10-4
10
-6
\
\UNFILTERED
\
FILTERED \ x
\
\
\
\
\
i i
i 111 in
I I
10"1 10°
RESIDUAL — TIME
101 102
IN MG-MIN/L
Figure 20. Coliform inactivation by chlorine in filtered and unfiltered
secondary effluent, Palo Alto.
Disinfection with Chlorine Dioxide
The results of the San Jose experiments with chlorine dioxide as disin-
fectant are shown in Fig. 21. In contrast to the corresponding experiments
with chlorine as disinfectant, no increase in disinfection efficiency is seen
when wastewater quality changes from that of secondary effluent to that of
nitrified effluent. This is expected, because the form of the disinfecting
agent, chlorine dioxide, is not affected by NH3~N concentration (34). This
observation further supports the conclusion that variations in initial bac-
terial concentration of the order of magnitude observed in this study (Table
20) do not affect the disinfection process, in view of the fact that the coli-
form concentrations in the undisinfected nitrified effluent were nearly one
hundred times lower than in non-nitrified effluent.
The similarity of disinfection in nitrified and secondary effluent by
chlorine dioxide (Fig. 21) is in good agreement with the results obtained with
chlorine when the experiments with chlorine in nitrified effluent are segre-
gated by Cl:NHo-N ratios. When the species of chlorine is unchanged, no dif-
ference in disinfection efficiency is seen between secondary and nitrified
effluents (Fig. 18).
65
-------
10
1
10
-2
10
-3
10
-4
10
-5
10
-6
SECONDARY
FILTERED
NITRIFIED
10
0
10-
10
2
Figure 21.
RESIDUAL—TIME IN MG-MIN/L
Coliform inactivation by chlorine dioxide in San Jose wastewater
effluents.
The San Jose experiments with chlorine dioxide as disinfectant show an
increase in disinfection efficiency in filtered wastewater as compared with
either secondary or nitrified effluent, similar to that seen for chlorine
disinfection but of greater magnitude. This increase in efficiency is again
seen in the Palo Alto - 79 experiments, shown in Fig. 22, between unfiltered
and filtered secondary effluent, similar to that previously shown (Fig. 20)
for chlorine disinfection in Palo Alto wastewater. However, the increase in
efficiency for chlorine dioxide is again of greater magnitude than that for
chlorine.
The one-hour chlorine dioxide demands are shown in Table 28. Judging
from the data in Table 28, the increase in disinfection efficiency using chlo-
rine dioxide seen between unfiltered and filtered wastewater cannot be attrib-
uted to a decrease in demand.
66
-------
O
i—i
E-i
10
-3
o 10
-4
10
—5
10
-6
\
i i i 11 nl i i i i r i nl i i i i i i ii
10
-1
10
0
Figure 22.
RESIDUAL — TIME IN MG-MIN/L
Coliform inactivation by chlorine dioxide in filtered and
unfiltered secondary effluent, Palo Alto.
TABLE 28. ONE-HOUR CHLORINE DIOXIDE DEMAND
Demand, mg/1
Wastewater and Treatment Level
San Jose, Nitrified
San Jose, Filtered
Palo Alto - 79, Unfiltered Secondary
Palo Alto - 79, Filtered Secondary
2 mg/1
1.45
1.80
1.99
1.94
Doses
5 mg/1
3.56
3.85
4.46
4.57
10 mg/1
4.86
4.19
7.46
6.79
Demand in mg/1 calculated as described in Table 27,
67
-------
COMPARISON OF DISINFECTION WITH CHLORINE AND CHLORINE DIOXIDE
Secondary Effluent
Chlorine and chlorine dioxide disinfection are compared in Palo Alto
secondary effluent (1978 experiments) in Fig. 23. Chlorine dioxide is a more
efficient bactericide than chlorine when compared on the basis of residual-
time product required for a given survival ratio. This is true for the full
range of Rt products (or equivalently survival ratios) of the experiments.
Because of the nature of the log-log representation, however, it is dif-
ficult to visualize how the chlorine dioxide advantage varies with changes in
survival ratio. For illustrative purposes, a ratio of chlorine dioxide resid-
ual to chlorine residual (calculated assuming one-hour contact time) will be
presented at two levels of inactivation, survival ratio of 10~^ and survival
ratio of 10~->. A value of required residual ratio (RR) less than unity indi-
cates superiority for chlorine dioxide compared to chlorine. For example, in
the case of Palo Alto secondary effluent (Fig. 23) at a survival ratio of
10 , the residuals required are 0.06 and 0.18 mg/1 for chlorine dioxide and
chlorine, respectively, yielding a C102: Clo residual ratio of 0.33. At a
survival ratio of 10" , the residuals are 1.38 and 3.33, respectively, yield-
ing a residual ratio of 0.41. The implication is that at lower residual-time
10
-1
1CT
O
10
O
10
-4
10
-5
10
-6
CO
\
\
\
\
CL02 AS \
DISINFECTANT \
CL2 AS
DISINFECTANT
\
\
\
\
\
\
10
0
10'
RESIDUAL—TIME IN MG-MIN/L
Figure 23. Coliform inactivation by chlorine dioxide in 1978 laboratory
experiments with unfiltered secondary effluent from Palo Alto,
68
-------
products, chlorine dioxide has a greater advantage in disinfection efficiency
over chlorine than at higher residual-time products. This convention will be
used for the comparison of chlorine dioxide and chlorine disinfection pre-
sented below. The residual ratios at survival ratios of lO"-*- and 10"^ will be
referred to as RR1 and RR2, respectively.
Disinfection in Palo
ments) is compared in Fig.
Alto unfiltered
24, with RR =
secondary
0.29 and
effluent (1979 experi-
RR
0.25
Chlorine
dioxide is again shown to be a more efficient bactericlde over the entire
range of Rt products studied and the advantage remains relatively constant.
Figure 25 shows comparison of chlorine and chlorine dioxide disinfection
of filtered Palo Alto secondary effluent (1979 experiments). The RR, and RR2
in this case are 0.01 and 0.26, respectively. The large advantage realized
for chlorine dioxide due to filtration is apparent in this comparison. This
advantage, however, decreases with increasing Rt product to the advantage seen
in unfiltered Palo Alto secondary effluent.
The results of disinfection experiments with chlorine and chlorine diox-
ide in San Jose secondary effluent are .compared in Fig. 26. In this case, RR,
and RR2 are 0.28 and 0.73, respectively, which indicates a decrease in chlo-
rine dioxide advantage with increasing Rt product.
O
•—i
E-!
CO
10
-1 L-
10
-3
o 10
-4
10
-5
10
-6
CHLORINE
\
\
\
\
CHLORINE \
DIOXIDE \
\
\
\
\
I I I I IHI
Figure 24.
10"1 10° 101 102
RESIDUAL—TIME IN MG-MIN/L
Coliform inactivation by chlorine and chlorine dioxide in 1979
experiments with unfiltered secondary effluent from Palo Alto.
69
-------
o
I—I
E-«
CO
CHLORINE
10'
RESIDUAL—TIME IN MG-MIN/L
Figure 25.
Coliform inactivation by chlorine and chlorine dioxide in 1979
experiments with filtered secondary effluent from Palo Alto.
10~2
10~3
10-4
10
-5
10
-6
CHLORINE
DIOXIDE
\
\
\
\
\
CHLORINE
\
\
\
\
-i—\ i i 11 nl 1—i i i 11 in i i i i 11 nl i i i i 11 ii
10J
10-=
RESIDUAL—TIME IN MG-MIN/L
Figure 26. Coliform inactivation by chlorine and chlorine dioxide in San Jose
secondary effluent.
70
-------
In general, chlorine dioxide is a more efficient bactericide in secondary
effluent than chlorine when comparison is made on a residual or residual-time
product basis at a given level of inactivation. The advantage for chlorine
dioxide is greatest in filtered secondary effluent and at lower Rt values or
correspondingly higher survival ratios.
Nitrified Effluent
A comparison of bacterial inactivation in San Jose nitrified effluent is
shown in Fig. 27. The advantage of chlorine dioxide over chlorine seen in
secondary effluent has disappeared due to the increase in efficiency of the
free chlorine residual in the nitrified effluent. RR^ and RR2 are 0.69 and
1.29, respectively for these experiments.
Nitrified-Filtered Effluent
The results of experiments with San Jose filtered effluent are compared
in Fig. 28, with corresponding RR.^ and RR2 of 0.02 and 0.88, respectively.
The tremendous advantage of chlorine dioxide disinfection seen due to filtra-
tion at low values of Rt decreases with increasing Rt product. This result is
consistent with the results of experiments in Palo Alto filtered (secondary)
wastewater as discussed above.
The correlations based on Dublin lab experiments are shown in Fig. 29.
In this case, however, the RR, and RR2 of 0.18 and 0.09 indicate that the
chlorine dioxide advantage remains relatively constant over the full range of
Rt products considered. The Dublin field experiments confirm this observation
as seen in Fig. 30. The RRj^ and RR2 for field experiments are 0.27 and 0.10,
respectively.
In general, chlorine dioxide is a more efficient bactericide than chlor-
ine in filtered, nitrified wastewater. The advantage found for chlorine diox-
ide is due to filtration and not nitrification. The advantage for chlorine
dioxide decreases at lower survival ratios in San Jose nitrified-filtered
effluent, while remaining relatively constant in Dublin nitrified-filtered
effluent.
Summary
The results obtained in this work indicate that chlorine dioxide is a
generally superior disinfectant when compared to chlorine on the basis of
residual * time product required to achieve a given total coliform survival
ratio. This conclusion is in agreement with previously reported laboratory
investigations (34-39) and field studies (40-43).
PREDICTING FULL-SCALE PERFORMANCE FROM LABORATORY EXPERIMENTS
A comparison of the laboratory and field (full-scale) experiments in
Dublin wastewater are shown in Figs. 31 and 32 for chlorine dioxide and
chlorine, respectively. In both cases, the regression equations from the
laboratory experiments slightly underestimate the survival observed at the
71
-------
o
1
10
10
10
10
10
2
-3
-4
-5
CHLORINE \\
DIOXIDE
CHLORINE
1 1 ii i _ i i i 1 1 1 il _ i _ i — i i 1 1 ill _ i — i — i i 1 1 ii
_ _ —
10 101 102
RESIDUAL — TIME IN MG-MIN/L
Figure 27. Coliform inactivation by chlorine and chlorine dioxide in San Jose
nitrified effluent.
10-1
ID"2
10~3
o 10
§
_l 10
-4
-5
10
-6
\
\
\
\
\
\
CHLORINE
DIOXIDE
\
CHLORINE
\
10° 101 102
RESIDUAL—TIME IN MG-MIN/L
Figure 28. Coliform inactivation by chlorine and chlorine dioxide in San Jose
filtered effluent.
72
-------
o
I—I
E-i
10
-1
10
-2
10~3
10-4
10"5
10
-6
CL02 AS
DISINFECTANT
CL2 AS
DISINFECTANT
\
10^ 10^ 10^
RESIDUAL—TIME IN MG-MIN/L
Figure 29. Coliform inactivation by chlorine and chlorine dioxide in
laboratory experiments with Dublin effluent.
Figure 30.
10
-1
10
-2
10
-3
o 1Q-
10
-5
10
-6
\
\
\
\
\
CL02 AS
DISINFECTANT
\
\
CL2 AS
)ISINFECTANT
\
\
j—i i i i..i nl 1 i i i 11 ti
10° 101 10
RESIDUAL—TIME IN MG-MIN/L
Coliform inactivation by chlorine and chlorine dioxide in field
experiments with Dublin effluent.
73
-------
10-1
10~2
10~3
10-4
10~5
io~6
\
x
X
X
LAB X
EXPERIMENTS v
X
FIELD
^EXPERIMENTS
x
X
X
X
-I I I I I I III
I I 1 I I 111
Figure 31
10° 101 102
RESIDUAL—TIME IN MG-MIN/L
Coliform inactivation by chlorine in laboratory and field
experiments with Dublin effluent.
o
P
Figure 32.
10
,-1 1-
10
-2
10
-3 __
10
,-4
10
-5
10
-6
\
x
\
X
x
X
X
LAB X \
EXPERIMENTS x
X
FIELD
EXPERIMENTS
x
X
_l I I IJ I 111
I I I I I I II I 1 I I I III
±3° 101 102
RESIDUAL—TIME IN MG-MIN/L
Coliform inactivation by chlorine dioxide in laboratory and field
experiments with Dublin effluent.
74
-------
full-scale treatment plant. The regression lines, however, are not statisti-
cally different from each other for either chlorine dioxide or chlorine. The
agreement is actually quite good when factors such as temperature, contact
time, and chemical mixing are considered. In field experiments, it is ex-
tremely difficult to reproduce the ideal mixing, constant temperature, and
controlled oxidant generation chemistry found in the laboratory experiments.
From Figs. 31 and 32 it can be concluded that laboratory bench-scale experi-
ments of the type described in this report would be a valuable design aid in
predicting the residual and/or contact time required to achieve an effluent
bacterial standard. Bacterial response to disinfectants is sufficiently
varied in different wastewaters and within the same treatment plant at dif-
ferent levels of treatment that no universal disinfection equation could be
expected to yield accurate predictive results. The coefficients of the model
used to describe the experimental results in this study are highly dependent
on wastewater quality and treatment history of the wastewater under considera-
tion, as well as on the choice of disinfectant.
75
-------
SECTION 8
VIRUS INACTIVATION
In view of the controversy over the validity of the coliform group as
adequate indicators of disinfection using chlorine (44-49) or chlorine dioxide
(50,51), additional indicators should be compared in evaluating a disinfec-
tant. The use of animal viruses is suggested, since several authors have
shown that viruses, such as the widely studied poliovirus, are typically more
resistant to disinfectants than are enteric bacteria (52,53). Poliovirus,
while not of epidemiological significance in the United States, may serve as
an indicator for viruses, such as the Hepatitis virus, that can be dissemi-
nated via the water route. Katzenelson and Kedmi (54), however, noted that
poliovirus may not be a suitable indicator since the virus could not be de-
tected consistently in urban sewage. For that reason, and also for ease of
recovery and analysis, other groups such as Kott et al. (55) and Atlantic
Research Corp. (4) have proposed observing coliphages in wastewater disinfec-
tion studies. These bacterial viruses are generally present in number concen-
trations approaching those of the coliforas, and their structure more closely
approximates that of pathogenic viruses than do coliform bacteria. The use of
viruses to monitor water quality in receiving waters has also been suggested,
since human enteric viruses have been observed in water that were determined
to be safe by bacteriological standards (56-58).
Analogous protective mechanisms offered by particulates to bacteria apply
toward the protection of viruses against inactivation. Virions may be protec-
ted against chlorine by aggregation (59-61), adsorption onto particulates
(62), or occlusion within organic solids (63). Also, the physical and chemi-
cal conditions that permit viral aggregation may not only protect against
inactivation, but may also allow partially inactivated virions to reassociate
and to cooperatively infect host cells (64). The same protective mechanisms
apply for disinfection with chlorine dioxide (65) and may be even more impor-
tant, because chlorine dioxide can react very quickly and be consumed in com-
peting reactions with other sources of demand in the wastewater.
RESULTS
Palo Alto (Non-Nitrified) Effluent
Selected experiments were conducted to assess the viricidal effectiveness
of chlorine dioxide as compared with chlorine. In situ coliphage and an
inoculum of Poliovirus I (10° PFU/100 ml) in Palo Alto secondary effluent were
dosed with 5 mg/1 chlorine dioxide and chlorine. Samples were taken at 2-,
76
-------
5-, and 10-minute contact times and analyzed for coliphage (Kott Method),
Poliovirus (standard plaque assay), and for fecal coliforms (MF method). The
results are shown in Fig. 33. In both trials, chlorine dioxide was a much
more effective viricide than was chlorine, although the coliform survivals
would indicate that both disinfectants were performing equally well.
Also, the log reductions of the coliphage and poliovirus by each of the
disinfectants were similar, a response that had been observed in preliminary
experiments not reported. Chlorine dioxide inactivation of both phage and
virus was more rapid and of a much greater magnitude than was inactivation of
these indicators by an equivalent dose of chlorine. Since the test procedure
for poliovirus is much more expensive and time-consuming than the recovery
procedure for coliphage, the coliphage was chosen as an indicator of virus
response to both disinfectants. Therefore, in situ phage, which may be a more
realistic indicator of native populations of other pathogens, was used as
opposed to using inocula of laboratory-grown strains of virus•
The results of the phage analyses that were run parallel with the total
coliform analysis are shown in Figs. 34 and 35. Once again, the total coli-
form bacteria respond similarly to the stress of a given disinfectant dose
(Fig. 34), whereas all three doses of chlorine dioxide (2, 5, and 10 mg/1) are
much more effective at inactivating the coliphage than even the highest dose
(10 mg/1) of chlorine (Fig. 35). The curve for the 10 mg/1 dose of chlorine
dioxide represents the detection limits of the Kott assay for coliphage.
Therefore this curve is an upper limit for the survival ratio and should be
interpreted not as equivalent to, but rather as less than or equal to the
survival at 5 mg/1 chlorine dioxide dose. This suggests that accepted indi-
cator organisms (total coliform bacteria) yield conservative performance data
for the viricidal effectiveness of chlorine dioxide in secondary effluent.
Another set of experiments was conducted to assess the biocidal effec-
tiveness of each disinfectant at very short contact times, since the results
of Fig. 35 indicate that CK^ reacts rapidly, within less than 5 min, to
inactivate phage. These experiments were run at 5-mg/l doses of chlorine
dioxide or chlorine and sampled at 15, 45, 90, and 120 seconds. The results
(Fig. 36) again indicate that chlorine dioxide is a more effective viricide
than a bactericide, and that chlorine dioxide performs better in secondary
effluent than chlorine within a short contact time (< 120 seconds).
Comparison of Virus Inactivation in Palo Alto and Dublin Effluents
Additional experiments were conducted to compare several indicator orga-
nisms with the conventional total coliform group in different wastewaters.
Two effluents, a non-nitrified (Palo Alto) and a nitrified, sand-filtered
(Dublin) secondary effluent were compared in laboratory studies. The nitri-
fied, sand-filtered effluent was used in bench-scale and full-scale studies,
whereas the non-nitrified effluent was used only in bench-scale studies.
Wastewater characteristics for the Palo Alto and Dublin effluents are shown in
Table 14 and Fig. 37. Microbiological analyses consisted of: total coliforms
by MF method, coliphage ($g) by the Kott method, coliphage (*c) by the RPTRR
method, and inoculated Poliovirus 1 (NQ = 10° PFU/100 ml) by the standard
plaque assay.
77
-------
k
-2
-4
2° -6
O
g
n
_i
o:
-8
5 mg/l
C102
-2
-4
-6
-8
LEGEND
FC = FECAL COLIFORM
<£B= COLIPHAGE,E_.CoM B HOST
VIRUS = POLIO VIRUS
\ ^d-
\
\
D
FC
D
D
VIRUS
_1
10
5 mg/l
CL2
10
CONTACT TIME (min)
Figure 33. Comparison of in situ coliphage and an inoculum of Poliovirus I in
non-nitrified secondary effluent. Above, treatment with 5 mg/l
chlorine dioxide; below, treatment with 5 mg/l chlorine.
78
-------
O -
<
o:
Q:
o:
£
d
o
i
Figure 34.
-5
-6
-7
2 mg/l
5 mg/l
}10 mg/l
LEGEND
O--0 C102
15
CONTACT TIME (min)
30
Total coliform survival in non-nitrified secondary effluent,
comparing chlorine dioxide and chlorine at three doses and three
contact times.
The nitrified sand-filtered effluent (Dublin) was markedly different from
the non-nitrified secondary effluent (Palo Alto). Typical characteristics for
each wastewater are summarized in Fig. 37. The results from bench-scale
experiments (Fig. 37) indicate that chlorine dioxide and chlorine perform
equally well at a 2-mg/l dose in the non-nitrified effluent, yet chlorine
dioxide is more than twice as effective as chlorine in the nitrified, sand-
filtered effluent as measured by total coliform survival. Disinfection as
modeled by inactivation of inoculated Poliovirus 1 is much greater for chlo-
rine dioxide than chlorine at the 2-mg/l dose. Similar results were shown
earlier (Fig. 33) for a 5-mg/l dose in the non-nitrified effluent. Therefore,
the total coliform group serves as an adequate, and possibly conservative,
indicator of virus inactivation for chlorine dioxide in both effluents but may
not be an adequate indicator when chlorine is used as a disinfectant.
79
-------
CD
O
I
ct
en
CD
o
LLJ
CD
I
CL
_i
O
o
-3
-4
-5
-7
\
D
2 mg/l
5 mg/l
10 mg/l
2 mg/l
"I0~mg/
.-•^B
I
LEGEND
D---0 C102
15
CONTACT TIME (min)
30
Figure 35. In situ coliphage (_E. coli B host) survival in non-nitrified
secondary effluent, comparing chlorine dioxide and chlorine at
three doses and three contact times.
80
-------
-1
•M
o
Q
C£.
<
O
O
\
\
\
h \
-4
-5
-6
-7
TOTAL COLIFORM
_n
FECAL COLIFORM
TOTAL COLIFORM
D
FECAL COLIFORM
D
— _ COLIPHAGE E.COLI B HOST
LEGEND
O—D C102 5 mg/l
•—• C12 5 mg/l
J_
15 45 90
CONTACT TIME (sec)
120
Figure 36. Responses of several in situ organisms to chlorine dioxide
and chlorine at very short contact times in non-nitrified
secondary effluent.
81
-------
2mg/l CI02 or Cl DOSE
TOTAL COLIFORMS
. , PALO ALTO CI02
PALO A"LTO~~C"£~~
DUBLIN CIO,
D £
-5
POLIOVIRUS IN
DUBLIN EFFLUENT
5 10
CONTACT TIME(min)
TYPICAL CHARACTERISTICS
Palo Alto
Dublin
NH3-N mg/l
35.5
-------
SECTION 9
RECOVERY FOLLOWING DISINFECTION
The need to disinfect wastewater has been questioned (66,67) for a number
of reasons. Firstly, the environmental hazards posed by disinfection (e.g.
carcinogenic byproducts) have been suggested to outweigh the benefits (e.g.,
public health protection). Secondly, there is little epidemiological evidence
that non-consumptive contact with receiving waters that have been contaminated
with fecal material can cause human disease. Although subclinical infections
from recreational contact are possible (68), they are difficult to associate
with contaminated water sources. Thirdly, there is evidence that more re-
growth of bacteria can occur following chlorination than if the effluent were
discharged without disinfection (69-81).
The conditions can exist in receiving waters for coliforms to persist for
long periods of time (82-84). It is also possible for pathogenic strains of
coliforms, such as Klebsiella species, to proliferate (85) and maintain their
virulence (86-88). Several factors affect the survival of bacteria and
viruses in receiving waters, including temperature, light, and predators. The
disinfection process may reduce the controlling biological factors such as
predation and competition, and allow wastewater flora to recover and grow in
the receiving environment.
EXPERIMENTAL
A set of experiments was designed to test for aftergrowth of bacteria
following disinfection. Three replicate experiments were run, each lasting
five days. In each experiment, evidence of bacterial growth was sought in
three reactor vessels: a secondary effluent control (no treatment), chlorine-
dioxide-treated secondary effluent, and chlorine-treated secondary effluent.
Each of the latter two were dosed at 5 mg/1. All reactors were diluted 1:4
with lake water after 30 min to simulate receiving water dilution. All reac-
tors were exposed to the same conditions of temperature and light. Parallel
samples were taken for total coliforms, fecal streptococcus, seeded polio-
virus, and coliphage.
RESULTS
The results (Figure 38) indicate a slight decline in coliforms in the
control. A slight increase (approximately one log unit) was observed in both
chlorine-dioxide- and chlorine-treated samples, although for chlorine dioxide
83
-------
8
UJ 2
O
2
O i
O 6
2 -
TOTAL COLIFORMS
CONTROL
CHLORINE
CHLORINE
DIOXIDE
FECAL STREPTOCOCCUS
CONTROL
CHLORINE
CHLORINE
DIOXIDE
I I
VIRUS
POLIO
CONTROL
COLIPHAGE
CONTROL
30min
TIME (DAYS)
Figure 38. Five-day recovery following disinfection.
84
-------
the initial kill (30 min contact) was greater, and the number of coliforms re-
maining was more suppressed than for chlorine. However, the rate of regrowth
of coliforms after chlorine dioxide treatment was equal to or greater than
that following chlorine treatment. The untreated fecal streptococcus samples
showed a marked decline after 24 hrs that remained relatively stable, while
both disinfected samples showed a slight but perhaps insignificant recovery.
The Poliovirus I and coliphage concentrations declined steadily in the
untreated control samples. The log values for the 30-min treated samples of
polio and coliphage were 1.99 and 1.3, respectively, and remained at or below
these detection limits for the duration of the experiment. Therefore, only
the control values for the phage and virus are shown, since meaningful data
could not be obtained in disinfected samples. The inactivation rate (approx-
imately 0.5 log/24 hr) of Polio 1 at these temperatures (22-24°C) is slightly
lower than has been reported for thermal inactivation and lysis by bacterial
enzymes in a surface water (89). The inactivation rate for coliphage, the
animal virus model, parallels the polio inactivation rate (approximately 0.5
log/24 hr). The mechanisms of inactivation may or may not differ, however,
since the phage is capable of reproduction under these conditions whereas the
poliovirus is not. The fact that the phage did not increase may mean that the
host coliforms were not growing actively enough to allow reproduction, and
hence a measurable increase, in the phage population.
The general trends observed in the recovery experiments agree with pre-
viously reported results (66-71). Namely, a reduction in fecal indicator
bacterial levels, and presumably the levels of enteric pathogens, occurs
within about 24 hrs after discharge of an untreated wastewater into a receiv-
ing water; conversely, an aftergrowth of indicator bacteria occurs following
disinfection by chlorine. This same recovery pattern is observed for an ef-
fluent disinfected with chlorine dioxide. The specific rates of inactivation
or recovery may depend on such experimental variables as temperature, dilution
ratio (wastewater:receiving water), concentration of predators in receiving
water, and water chemistry. These observations confirm the concerns of previ-
ous workers (65-67) that the need to disinfect effluents discharged into a re-
ceiving water with which humans will not experience immediate, direct contact,
should be re-examined with epidemiologically designed experiments similar to
that of Cabelli et al. (104). This conclusion is subject to uncertainty as to
whether enteric pathogens generally exhibit behavior similar to the indicators
studied here with respect to: inactivation during disinfection; in situ
inactivation in receiving waters; and recovery following disinfection.
85
-------
SECTION 10
HALOGENATED ORGANICS PRODUCED DURING DISINFECTION
CHLORINE DIOXIDE VERSUS CHLORINE
Introduction
The concern over halogenated byproducts of chlorine disinfection has
stimulated a search for alternative disinfectants, among which is chlorine
dioxide. In this section chlorine dioxide is compared with chlorine alone on
an equal mass-dose basis with regard to the amount of halogenated organics
formed in wastewaters. Halogenated organics are quantified as total trihalo-
methanes (THMs) and as total organic halogen (TOX). Comparisons between the
two disinfectants were carried out in bench-scale studies and in full-scale
field studies.
Within normal pH ranges encountered in water and wastewater treatment,
chlorine dioxide reduces to chlorite when it reacts with organics (12,13,90).
Chlorine dioxide does not oxidize bromide ion as does chlorine (91).
Chlorine dioxide apparently does not form trihalomethanes (91,92), but it
can produce halogenated organics in some cases. Compounds with carbon-carbon
double bonds are one group of compounds which can be halogenated. Work with
aqueous solutions of cyclohexane and methyl oleate has produced chlorinated
products (13). Chlorine dioxide can also react with phenol to produce chloro-
quinones, chlorohydroquinones, and chlorophenols (13,91). When chlorine
dioxide is used in excess with respect to the organic reactants, such as
cyclohexane and phenol, non-chlorinated organic acids are formed rather than
chlorinated products (13,91).
Studies dealing with the nature of humlc compounds suggests that phenol
derivatives and benzene carboxylic acids are major subunits (93). Therefore,
considering the chlorinated products resulting from chlorine dioxide reaction
with phenol, one can expect to find some halogenated organics resulting from
reactions in water containing humic substances. Manka et al. (94) have shown
that humics are present in treated wastewater effluents. Phenols themselves
have been found in domestic secondary effluent at 0.1 to 1.0 mg/1 concentra-
tions (95).
Experimental
The experiments presented in this section examine the production of halo-
genated organics in wastewater by chlorine or by chlorine dioxide. Two types
86
-------
of wastewaters were used in these experiments: secondary effluent from an
activated-sludge process at the Palo Alto Wastewater Treatment Plant and
effluent from the Dublin-San Ramon Wastewater Treatment Plant which had re-
ceived extended-aeration, activated-sludge treatment (achieving full nitrifi-
cation) and multimedia filtration. The characteristics of the wastewater
effluents are summarized in Table 29. Each experiment represents a wastewater
sample taken on a different day.
The experiments were primarily bench-scale experiments; however, some
full-scale field experiments were conducted to supplement the findings from
the bench-scale experiments. In the bench-scale experiments wastewater -efflu-
ents from both Palo Alto and Dublin-San Ramon were used. As shown in Fig. 39,
three or four liters of wastewater was dosed with either chlorine or chlorine
dioxide. The doses for both disinfectants were 40 mg/1 and 20 mg/1 measured
as ClO^ for chlorine dioxide and as Cl for chlorine. The dosed effluents were
mixed and then transferred into individual 0.5- and 1-liter glass-stoppered
reactor bottles within five minutes after dosing. The reactor bottles were
filled completely to exclude headspace. These reactor bottles were also cov-
ered with aluminum foil to prevent light from influencing the reactions. The
secondary effluent in these bottles was allowed to react for 1 to 24 hrs or
longer at 25°C. At the completion of the respective reaction periods, each
reactor bottle was opened, and samples for THM and TOX analyses were trans-
ferred immediately to 60-milliliter hypovials. Excess sodium thiosulfate
previously introduced into these hypovials quenched any disinfectant residual.
The remaining water from the reactor bottle was analyzed for disinfectant
residuals. TOX analysis was by the Dohrmann method (Section 4).
The field experiments were run at the disinfection facilities of the
Dublin-San Ramon Wastewater Treatment Plant. The influent to the contact tank
was nitrified, filtered activated-sludge effluent. A diagram of the contactor
is shown in Fig. 10. Wastewater characteristics in the chlorine contactor are
summarized in Table 30. At steady state, THM and TOX samples were taken up-
stream from the disinfectant injection point (Point A in Fig. 10) and also at
the end of the contactor just upstream of the outfall (Point D in Fig. 10).
Sodium thiosulfate was used to quench disinfectant residuals. Disinfectant
TABLE 29. TYPICAL CHARACTERISTICS OF WASTEWATER EFFLUENTS USED TO
MEASURE FORMATION OF HALOGENATED ORGANICS
Ranges
Palo Alto Dublin-San Ramon
Parameter (Non-Nitrified) (Nitrified)
COD, mg/1 20-55 20-35
NH3-N, mg/l-N 20-35 < 0.04-0.2
Total Kjeldahl N, mg/l-N 26-42 < 0.8-1.1
Alkalinity, mg/1 CaC03 190-250 100-300
PH 6.9-7.3 6.9-7.7
Suspended Solids, mg/1 10-50 0-13
87
-------
4-liter batch mixer
0.5-or I-liter
reactors
eOmilliliter
sample vials
Figure 39. Procedure for the bench-scale determinations of halogenated
organics formation.
TABLE 30. WASTEWATER CHARACTERISTICS IN THM AND TOX FIELD EXPERIMENTS
AT THE DUBLIN-SAN RAMON CHLORINE CONTACTOR
Parameter
Range in Contactor Influent
COD, mg/1
NH3-N, mg/l-N
Total Kjeldahl-N, mg/l-N
Alkalinity, mg/1 CaCOo
PH
Suspended Solids, mg/1
23-33
< 0.06
0.9-2.0
105-296
7.0-7.4
1-7
doses were chosen to achieve a desired bacterial kill; hence, the two disin-
fectants could not be compared on an equal mass basis in the field experi-
ments. For the two later field experiments, only chlorine dioxide was used.
TOX analysis was by the Jekel method (Section 4).
88
-------
Results
The bench-scale experiments showed that chlorine produced THMs and TOX in
both wastewaters tested. Experimental conditions and the amounts of THM and
TOX formed during 24-hr contact are summarized in Table 31. Typical results
for the THM and TOX formation by chlorine and chlorine dioxide in Palo Alto
TABLE 31. THM AND TOX FORMATION BY CHLORINE AND CHLORINE DIOXIDE
IN BENCH-SCALE EXPERIMENTS
Exp.
Wastewater No.
Chlorine
Chlorine Dioxide
Dose, Dose,
mg/1 Cl mg/1 C107
Average 24-hr
Production
TOX THM
ymol/l-X ymol/l-X
Average 24-hr
Residual
Cl,
CIO,
mg/l-Cl mg/l-C102
1
3
4
Non-
Nitrified
/Tain
Alto)
12
Nitrified 6
(Dublin-
(-1 -n \
ban itamonj1
7
20
none
20
none
40
none
20
none
40
none
20
none
40
none
20
none
40
none
20
none
40
none
20
none
none
20
none
20
none
40
none
20
none
40
none
20
none
40
none
20
none
40
none
20
none
40
none
20
6.0
0.0
—
__
—
— —
10.0
3.0
1.6
.
7.4T
1.7
8.1
-0.8
7.1
0.9
7.1
-0.7
22.4
0.5
26.2
1.4
19.3
-0.2
0.53
-0.03
0.22
-0.06
0.91
-0.02
0.18
0.02
0.85
0.00
0.43T
0.00
1.31
0.10
0.45
-0.01
0.98
0.01
3.97
0.01
4.87
0.03
3.75
0.02
6.1
NA
6.0
NA
9.1
NA
6.5
NA
16.2
NA
6.8T
NA
15.1
NA
9.3
NA
18.5
NA
6.6
NA
19.0
NA
7.5
NA
NA
0.5
NA
8.0
NA
8.0
NA
3.4
NA
3.4
NA
NA
10.4
NA
5.2
NA
3.8
NA
8.4
NA
24.5
NA
11.2
NA = not applicable
All pertinent experiments are shown; missing experiment numbers refer to
experiments conducted with other objectives.
t,
#
'23 hrs.
25 hrs.
89
-------
wastewater are shown in Fig. 40 as a function of contact time up to 24 hrs.
The data in Figure 40 illustrate the lack of production of THMs by chlorine
dioxide and a small apparent TOX production by chlorine dioxide. However,
other bench-scale experiments in the Palo Alto wastewater show no TOX produc-
tion by chlorine dioxide at 24-hr contact time. THM and TOX production in one
of the two bench-scale experiments which used the Dublin-San Ramon wastewater
is shown in Fig. 41. As with the Palo Alto wastewater, chlorine dioxide
caused no production of THMs, but caused a slight TOX production as shown in
Fig. 41 for the 40-mg/l chlorine dioxide dose.
Results of the field experiments, in which THM and TOX formation were
measured during approximately one-hr retention in the Dublin-San Ramon contact
tank, are summarized in Table 32. These experiments were conducted at disin-
fectant doses having approximately equal disinfection effectiveness (survival
ratio); the dose of chlorine dioxide is nearly ten times less than that of
chlorine. Under these conditions, chlorine dioxide clearly formed no signifi-
cant amounts of either THM or TOX, whereas chlorine increased the values of
both parameters substantially.
Chlorine dioxide in some cases produced small amounts of TOX in batch
experiments at doses of 20 and 40 mg/1. Of the 10 values given in Table 31,
one value (Experiment 4, 20 mg/1 dose) showed TOX production by chlorine
dioxide greater than two standard deviations (page 15) for the Dohrmann TOX
procedure, i.e., TOX > 1.96 pmol/1. .The mean values and standard deviations
of TOX and THM production in the bench-scale experiments are shown in Table
33. Apparent negative TOX production values are less than two standard devia-
tions and are therefore statistically indistinguishable from zero production.
TOTAL THM
PRODUCTION
mg/1
20 mg/1 Clz
CIO,
40 mg/1 CI02
0 5 10 15 20 25
TIME IN HOURS
30
Z30
O25-
I*H
-5
TOX
PRODUCTION
4Omg/ICI2
20 mg/1
CIO
40*1*1/1
0 5 10 IS 20 25
TIME IN HOURS
30
Figure 40. TOX and THM production in non-nitrified effluent
(Experiment 4—Palo Alto wastewater).
90
-------
TOTAL THM PRODUCTION 35T
TOX PRODUCTION
40mg/l
^^=^-====?==&
^* on.*t«*/i r*iA
20mg/l CI02
0 5 10 15 20 25
TIME IN HOURS
40mg/1CI02
^— ^^" ^^B«BB*^^_
^rngFobj"
5 10 15 20 25
TIME IN HOURS
Figure 41. TOX and THM production in filtered, nitrified effluent
(Experiment 6—Dublin-San Ramon wastewater).
TABLE 32. SUMMARY OF THM AND TOX FIELD EXPERIMENTS
Average
Production
Field
Exper .
Date
2-6-79
2-15-79
Chlorine
Dose
mg/l-Cl
18.7
none
19.6
none
Chlorine
Dioxide
Dose
mg/l-C102
none
3.1
none
2.3
Contact -
Time
Minutes
55
55
49
46
at Point D
TOX
ymol
10.2
-0.5
8.1
0.4
THM
pmol/l-X
1.02
-0.04
0.94
0.04
Average
Residual
at Point D
mg/l-Cl
12.8
NA
1.0
NA
mg/l-C102
NA*
0.6
NA
0.1
NA = not applicable.
Except for the 20 mg/1 chlorine dioxide dose in Experiment 4, the other in-
stances of significant TOX production mentioned are near to the two standard
deviation limit. Nonetheless, it cannot be excluded from possibility that
chlorine dioxide can form halogenated organics in wastewater. More experi-
ments with greater precision in TOX analysis will be required to make a firm
conclusion as to whether chlorine dioxide does produce halogenated organics in
wastewater. More important, however, is the fact that chlorine dioxide pro-
duced markedly less halogenated organics, if indeed any, compared to chlorine
in the two types of wastewaters tested.
91
-------
The means and standard deviations of THM production shown in Table 33
support previous findings (91,92) that chlorine dioxide does not form THMs.
Of all of the bench-scale experiments (Table 31), only the 40 mg/1 chlorine
dioxide dose in Experiment 8 showed a THM production exceeding two standard
deviations for THM production, i.e., THM > 0.08 pmol/l-X (Table 1). This THM
production value is the only instance exceeding two standard deviations among
many other results to the contrary. Also, this value exceeds the two standard
deviation limit by only a slight margin. Therefore, it is safe to conclude
that chlorine dioxide did not produce THMs in the wastewaters tested. The
apparent negative THM production values in a few experiments are within two
standard deviations of zero production and are therefore statistically indis-
tinguishable from zero.
TABLE 33. MEAN VALUES AND STANDARD DEVIATIONS OF 24-HOUR HALOGENATED
ORGANIC BYPRODUCT FORMATION WITH CHLORINE DIOXIDE AND CHLORINE
Increase During 24-Hr Contact
Mass ^
Dose
Effluent mg/1 Disinfectant
20
20
40
20
20
Dublin 4Q
40
Chlorine
Chlorine
Chlorine
Chlorine
Chlorine
Chlorine
Chlorine
Chlorine
Dioxide
Mean
0
.005
0.39
Dioxide
Dioxide
Dioxide
0
1
0
3
0
4
.05
.15
.015
.9
.03
.9
TTHM
Std.Dev. n~
0
0
0
0
0
0
.02
.15
.06
.23
.007
.14
-
—
4
4
2
2
2
2
1
1
Mean
1.
7.
-1.
7.
0.
20.
1.
26.
3
6
2
6
15
8
4
2
TOX
Std.Dev
1.3
1.7
0.6
0.7
0.5
2.2
-
—
. nf
4
4
2
2
2
2
1
1
Increase calculated as concentration measured after 24 hours less
concentration measured before.
n = number of replicates.
Halogenated Organics Yield Versus Disinfectant Dose
The TOX produced in both the Palo Alto and the Dublin-San Ramon waste-
waters by chlorine is on the order of one percent of the chlorine dose. This
result is the same as that found by Jolley (96) in his work with secondary
effluent. However, the total chlorine residuals had decreased to half of the
initial chlorine dose or less after 24 hrs in nearly all of the bench-scale
experiments (Figs. 42 and 43). This result suggests that 99% of the chlorine
demand is in the form of redox-type reactions, whereas only 1% of the chlorine
demand is in the form of halogenation reactions.
92
-------
DISINFECTANT RESIDUAL
45i
, -a 40 mg/l CI02
20 mg/l CI02
20 mg/l
0 5 10 15 20 25 30
TIME IN HOURS
5 10 15 20 23 30
TIME IN HOURS
Figure 42. Disinfectant residuals in Experiment 8 (Palo Alto wastewater)
DISINFECTANT RESIDUAL
5 10
TIME
15 20 25 30
IN HOURS
20 mg/l CI02
IJI
20 mg/l Cl,
0 5 10 15 20 25 30
TIME IN HOURS
Figure 43. Disinfectant residuals in Experiment 6 (Dublin-San Ramon
wastewater).
93
-------
Since TOX production by chlorine dioxide is not well documented by the
experiments conducted, it is not appropriate at this time to estimate how much
of the initial chlorine dioxide demand is in the form of halogenation pro-
ducts. However, it is worthwhile to point out that chlorine dioxide residuals
also decreased to half (or less in some cases) of the value of the initial
chlorine dioxide doses (Figs. 42 and 43). This result indicates that chlorine
dioxide does enter into redox reactions with the substances in wastewater,
consistent with the findings of Noack and Doerr (97).
Figures 40 and 41 show that higher chlorine doses result in higher THM
productions in both wastewaters. In contrast, from Table 31 (Experiments 8
and 12), it can be seen that in the Palo Alto wastewater, there is no signifi-
cant difference in TOX production between the two chlorine doses used (20 and
40 mg/1). The single experiment (Experiment 6) using Dublin-San Ramon waste-
water and two chlorine doses does show a significantly different TOX produc-
tion between the two chlorine doses. The difference in THM production in the
Palo Alto wastewater experiments is too small to be detected by TOX analysis.
Therefore, the dose effect on halogenated organics production in the Palo Alto
wastewater is very small, indicating that the 20-mg/l chlorine dose is suffi-
cient to nearly exhaust the amount of precursors of halogenated organics
available to it. In order to observe a less ambiguous dose effect, lower
chlorine doses should be used for the Palo Alto wastewater.
The Relationship Between TOX and THM Production
The results summarized in Tables 31 and 32 show the 24-hr TOX production
to be between five and fifty times greater than THM production in the Palo
Alto wastewater and approximately five times greater than THM production in
the Dublin-San Ramon wastewater. These results are consistent with the fact
that TOX represents a broader spectrum of halogenated organics than does THM.
The Behavior of Disinfectant Residuals
At a given chlorine dioxide dose, the chlorine dioxide residuals were
lower than the chlorine residuals in the Palo Alto wastewater (Fig. 42). In
contrast, the chlorine dioxide residuals were higher than the chlorine resid-
uals in the Dublin-San Ramon wastewater (Fig. 43). Of course, the results of
comparing the residuals of different disinfectants depends on the units chosen
to express the residual concentration. In this work, a mass basis generally
was used, because of the direct relation to disinfectant dose and hence to
cost. A better comparison of disinfectant behavior is on an electron-
equivalent basis for disinfectant demand as shown in Table 34. From Table 34
it is clear that the disinfectant demand for chlorine is greater than for
chlorine dioxide in both wastewaters; that is, more electron-equivalents of
chlorine than of chlorine dioxide are used up in both wastewaters. This is
relevant to questions of reactions between oxidants and organics, because the
number of electron equivalents of oxidant that is reduced is a measure of the
extent to which redox reactions have occurred.
94
-------
TABLE 34. DISINFECTANT DEMAND
24-hr Demand in meq/1
20 mg/1 Dose"1" 40 mg/1
Wastewater
Palo Alto
Dublin-
San Ramon
Experiment
1
3
4
8
12
6
7
ci2
0.39
0.39
0.38
0.37
0.30
0.38
0.35
cio2
0.29
0.18
0.25
-
0.22
0.17
0.13
ci2
_
0.59
0.67
0.70
0.61
0.59
~
cio2
—
0.52
0.54
0.44
0.54
0.23
—
Assumes chlorine dioxide undergoes a single electron transfer, i.e., it
is reduced to chlorite ion.
t20 mg/l-Cl =0.56 meq/1; 20 mg/l-C102 =0.30 meq/1.
#40 mg/l-Cl -1.13 meq/1; 40 mg/l-C102 = 0.59 meq/1.
MIXTURES OF CHLORINE DIOXIDE AND CHLORINE
Introduction
One of the methods used to produce chlorine dioxide is the chlorine/chlo-
rite process. Using excess chlorine could lead to a product that is a mixture
of chlorine and chlorine dioxide. Miltner (98) has shown that a mixture of
chlorine dioxide and chlorine would produce less THMs than would chlorine
alone in the same concentration as for the chlorine used in the mixture.
Experimental Design
The experiments using mixtures of chlorine and chlorine dioxide were
bench-scale experiments. The mixtures of chlorine and chlorine dioxide were
based on the mass ratios of chlorine dioxide (as C102) to chlorine (as Cl).
These mass ratios were 1:5, 1:1, and 5:1; experiments with 100% C12 and 100%
C102 were run for comparison. The chlorine dose was kept constant at 5 mg/l-
Cl. Dublin-San Ramon wastewater was added directly to the 0.5- and 1-liter
glass-stoppered reactor bottles identical to those used for the experiments
presented under "Chlorine Dioxide Versus Chlorine." The chlorine and chlorine
dioxide doses were added with mixing, and the reactor bottles were closed and
allowed to react at 25°C for one hour. This short reaction time was chosen
because it approximated the detention time observed in the chlorine contact
basin at the Dublin-San Ramon facility. After the one-hour reaction period,
samples were taken in the same manner as for the bench-scale experiments
described earlier. TOX analysis was by the Dohrmann method.
95
-------
Results
Results of two experiments using mixtures of chlorine and chlorine diox-
ide are shown in Figs. 44 and 45 and are summarized in Table 35. In general,
both experiments show a decrease in THM production as the proportion of chlo-
rine dioxide increases (Fig. 44). The same phenomenon is seen for TOX in Fig.
45. Because of analytical difficulties, the TOX data for Experiment 2 were
unreliable; therefore, TOX production values are not shown for that experi-
ment. The decrease in THM and TOX production appears to require equal mass
doses of chlorine dioxide and chlorine before these decreases become signifi-
cant.
Figure 44 shows an initial increase in THM production from the point with
no chlorine dioxide to the point where the chlorine dioxide to chlorine ratio
is 1:5. Although the increase is greater than two standard deviations for THM
analysis, the general behavior of the rest of the data in Figs. 44 and 45 sug-
gest that this increase is probably due to an operator error in sampling or in
performing the THM analysis.
If the mechanism of THM decrease is by chlorine dioxide reacting with
precursors of halogenated organics to prevent THM production, as Miltner pro-
posed (98), the decrease in TOX production would indicate that this interac-
tion with precursors may prevent halogenation altogether and not just prevent
THM production. However, there still remains a possible reaction between
chlorine and chlorite as proposed by Noack and Doerr (97) and also a possible
direct reaction between chlorine and chlorine dioxide noted by Miltner (98).
Better understanding of the properties and reactions of mixtures of chlo-
rine and chlorine dioxide is necessary. Experiments need to be conducted
using longer reaction times. Non-nitrified as well as nitrified wastewaters
need to be studied. Finally, more investigations into the precise nature of
possible interactions between chlorine and chlorine dioxide in terms of the
species formed and their disinfection capabilities are required.
96
-------
THM PRODUCTION
% CI02 0 10 20 30 40 50 60 70 80 90 100
BY WT.
RATIO
ClOg/C^
BY WT.
1=5
Figure 44. THM production resulting from various mixtures of chlorine and
chlorine dioxide in Experiment 2 (Dublin-San Ramon wastewater),
97
-------
THM PRODUCTION
%CI02 0 10 20 30 40 50 60 70 80 90 100
RATIO CKCl 1^5 hi 5^1
TOX PRODUCTION
RATIO
cio2/q
10 20 30 40 50 60 70 80 90 100
i $5M 5M
Figure 45. TOX and THM production resulting from various mixtures of
chlorine and chlorine dioxide in Experiment 5
(Dublin-San Ramon wastewater).
98
-------
SECTION 11
COST COMPARISON BETWEEN CHLORINE DIOXIDE AND CHLORINE FOR
WASTEWATER DISINFECTION
BASIS OF COMPARISON
An economic comparison is presented between disinfection of wastewater
with chlorine dioxide and with chlorine. In the analysis five treatment plant
sizes, 0.044, 0.22, 0.44, 2.2, and 4.4 m3/s (1, 5, 10, 50, and 100 mgd), are
considered. A nominal contact time of 1 hour is assumed for the disinfection
process. Chemical costs are compared between chlorine and chlorine dioxide
generated from the chlorine-chlorite process.
Three levels of disinfection are compared, corresponding to 2.2/100 ml,
200/100 ml, and 1000/100 ml total coliforms. The latter case (1000/100 ml
total coliforms was included in the analysis at the request of the EPA Project
Officer [M. C. Meckes], who felt that it more nearly corresponds to the efflu-
ent standard imposed for secondary effluent in states such as Ohio. The
1000/100 ml total coliform standard is approximately equivalent to a 200/100
ml fecal coliform standard if the ratio of total to fecal coliforms is approx-
imately 5:1 as found by Hunt and Springer (99). The 1000/100 ml total coli-
form standard also is that reportedly recommended by ORSANCO, and widely
emulated in the U.S. (100).
Two levels of prior treatment are considered: conventional activated-
sludge treatment as typified by the Palo Alto 1978 laboratory experiments, and
nitrified-filtered activated-sludge treatment as practiced at the Dublin-San
Ramon treatment plant. It is believed that the conventional activated-sludge
effluent represents the majority of biologically treated wastewater effluents
in the U.S.A., whereas the nitrified, filtered effluent corresponds to a
higher degree of treatment, such as might be practiced for water reclamation.
All costs, except chemical costs, have been updated to third-quarter 1979
costs by use of the Environmental Protection Agency Sewage Treatment Cost
Index (101,102). The EPA-STP index for third quarter (September) 1979 was
337.8. Chemical costs are based on competitive supplier bids, f.o.b. docks,
January 1980 (103,104). For simplicity, the comparisons will be referred to
as cases A through F, summarized in Table 36.
100
-------
TABLE 36. CASES FOR EVALUATION OF THE RELATIVE COSTS OF DISINFECTION WITH
CHLORINE AND CHLORINE DIOXIDE
Type of EffluentDisinfection Standard
to Be as Total Coliform Count
Disinfected Case (N/100 ml)
Conventional
Activated Sludge
Filtered, Nitrified
Activated Sludge
A
B
C
D
E
F
2.2
200
1000
2.2
200
1000
CAPITAL COSTS
Capital costs were estimated from generalized correlations (101,102).
These estimates were adjusted for inflation as described above. Capital costs
are amortized over 20 years at 11 percent.
Contact Basins
The common current practice in designing disinfectant contact basins is
to provide a serpentine routing of flow following mixing of the disinfectant
and wastewater. For the purposes of this report, costs are estimated for
basins which generally provide a flow routing length to channel width ratio
between 20:1 and 25:1. The Dublin disinfectant mixing and contact facility is
typical of the design considered in this analysis (see Fig. 10).
Chlorine Feed and Storage Facilities
The feed used most widely in wastewater treatment plants is chlorine gas.
The required equipment and storage facilities are well known; commercial
equipment is readily available.
Feed facility costs include distribution panels, cylinder chocks, instal-
lation, manufacturers' preparation of shop drawings, installation check and
startup, and contractor's overhead and profit. Chlorinator costs include one
standby chlorinator and chlorine evapotators on systems having a capacity
greater than 4,000 Ibs per day. Also included are miscellaneous piping and
valves, which typically amount to 5-10% of the installed chlorination equip-
ment costs.
Storage facility costs include housing, hoist equipment, ventilation
equipment and safety monitors.
101
-------
Chlorine Dioxide Generation and Feed Systems
For the purpose of this analysis, it is assumed that chlorine dioxide is
generated by the chlorine-chlorite process. The stoichiometry measured for
this process in a full-scale plant has been reported in Section 5. In theory,
1.34 Ibs of pure sodium chlorite and 0.5 Ibs of chlorine react to give one
pound of chlorine dioxide. However, sodium chlorite is normally purchased
with a purity of 80%. Therefore, assuming 100% conversion of NaClOo. and 100%
yield of CK^ based on NaC102 reacte<* with 4% excess chlorine feed (Section
5), 1.68 Ibs of 80% sodium chlorite and 0.70 Ibs of chlorine are required per
pound of chlorine dioxide generated. This basis is somewhat more optimistic
than generally assumed, but is justified according to the results in Section 5
of this report.
Generation and feed systems costs are based on the costs for a sodium
chlorite mixing and metering system, a chlorine dioxide generator, and the
appropriately sized chlorine feed system. The sodium chlorite system consists
of a polyethylene day tank, a mixer for the day tank, and a dual head metering
pump. The chlorine dioxide generator is a PVC tube filled with porcelain
Raschig Rings or other transport-enhancing media, and is sized for a detention
time of 20 min.
Disinfectant Feed Capacity—
The sizing of generating, feed, and storage facilities is based on the
required disinfectant feed capacity in kg or pounds per day. (A detailed
example of this procedure appears in Appendix C.)
In determining the feed capacity required, the procedure is as follows:
1. Determine the initial bacterial concentration in the waste
stream to be disinfected (Table 14).
2. Calculate the required survival ratio, N(t)/N(0), from the
disinfection standard N(t) and initial bacterial concentra-
tion, N(0).
3. Calculate the required residual-time product using Eqs. 24
and 25 and the model fitting constants from Table 24.
4. Calculate the required residual, based on the provided con-
tact time, namely one hour.
5. Calculate the disinfectant dosage needed to provide the
required residual disinfectant concentration
A summary of the results from steps 1-5 from above for Cases A-F is given
in Table 37. The quantities of disinfectant required per day shown in Table
38 were calculated from the required dosages given in column 5b of Table 37.
OPERATION AND MAINTENANCE COSTS—EXCLUDING CHEMICAL COSTS
Operation and maintenance (O&M) costs include electrical power, mainte-
nance material, and labor costs (101,102).
102
-------
TABLE 37. SUMMARY OF ESTIMATES OF REQUIRED DISINFECTANT DOSAGES USED IN COST EVALUATION CASES
o
u>
(1)
Initial Bacterial
Concentration
„ * ., (-total coliform
Case log^ .
A 6
B 6
C 6
D 4
E 4
F 4
00 ml
.21
.21
.21
.41
.41
.41
(2) (3) (4) (5a)
Residual-Time Residual Required 1-hr Disinfec-
Product at 1-hr Contact tant Demand
Survival Ratio (mg-min/1) Time (mg/1) (mg/1)
) log(N(t)/N(0)) C12 C102 C12 C102 C12 C102
-5
-3
-3
-4
-2
—.1
.87
.91
.21
.07
.11
.41
376.58
90.11
54.01
210.73
13.02
4.82
164.45
34.79
19.95
22.74
1.96
0.82
6.28
1.50
0.90
3.51
0.22
0.08
2.74
0.58
0.33
0.38
0.03
0.01
1.61
0.95
0.80
7.64
2.39
2.52
5.18
2.32
1.84
5.14
0.57
0.13
(5b)
Required
Disinfectant
Dose (mg/1)
ci2 cio2
7.89
2.45
1.70
11.15
2.61
2.60
7.92
2.90
2.17
5.52
0.60
0.14
See text for explanation of cases, procedure, and calculations.
-------
TABLE 38. AMOUNTS OF DISINFECTANTS REQUIRED TO ACHIEVE
TOTAL COLIFORM STANDARDS
Disinfectant Usage, pounds per day
Treatment Plant Size - mgd'
1 5 10 50
100
Case A
ci2
cio2
Case B
ci2
cio2
Case C
ci2
cio2
Case D
ci2
cio2
Case E
ci2
cio2
Case F
C17
cio2
66
66
20
24
18
14
93
46
22
5
22
1.2
329
330
102
121
90
70
465
230
110
25
110
6
658
661
204
242
180
140
930
460
220
50
220
12
3290
3303
1022
1209
900
700
4650
2302
1100
250
1100
60
6580
6605
2043
2419
1800
1400
9300
4604
2200
500
2200
120
1 pound = 0.4536 kg.
^l mgd = 0.0438 m3/s.
Electrical requirements include power for the gaseous chlorination sys-
tem, the sodium chlorite metering and mixing systems, and building heating,
lighting, and ventilation.
Maintenance material requirements are based on experience with gaseous
chlorine systems and liquid metering systems. Costs for chemicals required
for residual monitoring are included.
Labor requirements consist of labor for the gaseous chlorination systems,
plus the labor required to mix the sodium chlorite solution, to adjust its
feed rate, and to maintain the mixing and metering equipment.
CHEMICAL COSTS
All chemical costs are f.o.b. works. It is assumed that the 50- and
100-mgd plants will be able to take advantage of the lower chemical costs
104
-------
associated with larger quantity purchases. Chemical unit costs are shown in
Table 39.
TABLE 39. UNIT COSTS OF CHEMICALS REQUIRED IN WASTEWATER DISINFECTION
Chemical Unit Cost, $ per pound
Chlorine - one ton cylinders 0.155
Chlorine - rail cars 0.075
Sodium chlorite (80%), less than
250,000 pounds annually 2.67
Sodium chlorite (80%), more than
250,000 pounds annually 1.98
Prices are f .o.b. works; 1 pound - 0.4536 kg.
Source: References (103,104).
The price given for sodium chlorite (Table 39) and used in this evalua-
tion is based on manufacturers' quotations for a solution of 25% sodium chlo-
rite content. The equivalent price is $4.36/kg or $5.88/kg ($1.98 per pound
or $2.67 per pound) of sodium chlorite content, depending on the quantity
purchased. The price of dry sodium chlorite is considerably lower: $2.67/kg
($1.21 per pound) of sodium chlorite if purchased in drum quantities at
January 1980 prices (105). The higher price quoted by manufacturers who
supply sodium chlorite to the water industry has been chosen to be conserva-
tive. However, it should be recognized that sodium chlorite possibly could be
purchased for approximately 45 to 60% of the values given in Table 39.
COST SUMMARIES
The cost of wastewater disinfection are summarized in Tables 40 through
45, for Cases A-F, respectively. Disinfection with chlorine dioxide is at
least as expensive as, and generally more expensive than, disinfection with
chlorine under the assumptions embodied in the cases compared here. The rela-
tive cost factor for disinfection with chlorine dioxide compared to chlorine
ranges from a high of 15 to one (Case A, 4.4 m Is [1QO mgd] plant) to a low of
1.0 to one (Case F, 0.044 m3/s [1-mgd] and 0.22 m3/s [5-mgd] plants). The
predominant factor influencing the higher cost of chlorine dioxide disinfec-
tion is the cost of the sodium chlorite.
Case C (conventional activated-sludge effluent, 1000/100 ml total coli-
form standard) is believed to represent most nearly among the cases studied
the typical situation of disinfecting secondary effluent in the U.S. In Case
C, the cost of disinfection with chlorine dioxide ranges from 2 times as ex-
pensive (at 0.044 m /s [1 mgd]) to 5 times as expensive (at 4.4 m /s [100
mgd]) compared to the cost of disinfection with chlorine.
105
-------
TABLE 40. DISINFECTION COST SUMMARY FOR CASE A
Annual
Cost
, Thousand $
Treatment Plant Size
Cost Item
Contact Basin
Generation and
Feed Equipment
0 & M Cost
Chemical Cost
ci2
NaC102
Total Cost
$1000/yr
(t/1000 gal)
ci2
14.7
1.7
7.4
3.7
-
27.5
(7.5)
1
cio2
14.7
6.1
9.9
2.6
108.1
141.4
(38.7)
ci2
36.0
4.0
10.5
18.6
-
69.1
(3.8)
5
cio2
36.0
10.0
18.9
13.0
540.3
618.2
(33.9)
ci2
54
6
13
37
-
Ill
(3
.8
.3
.2
.2
.5
•1)
10
cio2
54.8
12.7
22.5
26.0
1082.2
1198.2
(32.8)
per Year
, mgd
ci2
172.3
18.0
17.1
90.1
-
297.5
(1.6)
50
cio2
172.3
25.9
55.0
63.3
4010.3
4326.8
(23.7)
ci2
313.0
29.7
18.9
180.1
-
541.7
(1.5)
100
cio2
313.0
35.5
81.0
126.6
8019.4
8575.5
(23.5)
-------
TABLE 41. DISINFECTION COST SUMMARY FOR CASE B
Annual
Cost, Thousand $
Treatment Plant Size
Cost Item
Contact Basin
Generation and
Feed Equipment
0 & M Cost
Chemical Cost
ci2
NaC102
Total Cost
$1000/yr
(t/1000 gal)
ci2
14.7
1.6
6.1
1.1
23.5
(6.4)
1
cio2
14.7
4.5
7.8
0.9
39.3
67.2
(18.4)
ci2
36.0
2.0
7.9
5.8
51.7
(2.8)
5
cio2
36.0
7.1
12.2
4.7
198.1
258.1
(14.1)
ci2
54.8
3.0
9.0
11.5
81.3
(2.2)
10
cio2
54.8
8.7
17.1
9.4
396.2
486.2
(13.3)
per Year
> mgd
ci2
172.3
8.1
14.8
28.0
223.2
(1.2)
50
cio2
172.3
16.6
30.0
23.1
1467.9
1709.9
(9.4)
ci2
313
14
16
55
399
(1
.0
.1
.1
.9
.1
•1)
100
cio2
313.0
21.1
43.0
46.2
2937.0
3360.3
(9.2)
-------
TABLE 42. DISINFECTION COST SUMMARY FOR CASE C
o
oo
Annual
Cost, Thousand $
Treatment Plant Size
Cost Item
Contact Basin
Generation and
Feed Equipment
0 & M Cost
Chemical Cost
C12
NaC102
Total Cost
$1000/yr
(t/1000 gal)
ci2
14.7
1.8
6.0
1.0
23.5
(6.4)
1
cio2
14.7
3.7
7.1
0.6
22.9
49.0
(13.4)
ci2
36.0
3.5
7.9
5.1
52.5
(2.9)
5
cio2
36.0
6.4
10.0
2.8
114.6
169.8
(9.3)
ci2
54.8
4.6
8.6
10.2
78.2
(2.1)
10
cio2
54.8
9.4
12.2
5.5
229.2
311.1
(8.5)
per Year
> mgd
ci2
172.3
10.8
15.5
24.8
223.4
(1.2)
50
cio2
172.3
20.5
25.0
13.5
849.9
1081.2
(5.9)
ci2
313.0
17.7
23.0
49.6
403.3
(1.1)
100
cio2
313
27
33
27
1699
2100
(5
.0
.3
.0
.0
.8
.1
.8)
-------
TABLE 43. DISINFECTION COST SUMMARY FOR CASE D
o
VO
Annual
Cost, Thousand $
Treatment Plant Size
Cost Item
Contact Basin
Generation and
Feed Equipment
0 & M Cost
Chemical Cost
ci2
NaC102
Total Cost
$1000/yr
(t/1000 gal)
ci2
14.7
1.80
7.8
5.3
29.6
(8.1)
1
cio2
14.7
5.5
9.0
1.8
75.3
106.3
(29.1)
ci2
36.0
5.0
11.2
26.3
78.5
(4.3)
5
cio2
36.0
8.7
16.5
9.1
376.6
446.9
(24.5)
C12
54.8
7.8
15.8
52.6
131.0
(3.6)
10
cio2
54.8
10.9
20.8
19.3
753.1
857.9
(23.5)
per Year
, mgd
ci2
172.3
21.9
17.9
127.3
339.4
(1.9)
50
cio2
172.3
20.5
45.0
44.1
2794.9
3076.8
(16.9)
ci2
313.0
36.0
19.8
254.6
623.4
(1.7)
100
cio2
313
28
68
88
5589
6087
(16
.0
.6
.0
.2
.9
.7
•7)
-------
TABLE 44. DISINFECTION COST SUMMARY FOR CASE E
Annual
Cost, Thousand $
Treatment Plant Size
Cost Item
Contact Basin
Generation and
Feed Equipment
0 & M Cost
Chemical Cost
ci2
NaCl02
Total Cost
$1000/yr
Ct/1000 gal)
ci2
14
1
7
1
25
(6
.7
.7
.5
.2
-
.1
.9)
1
cio2
14.7
2.6
7.4
0.2
8.2
33.1
(9.1)
ci2
36.0
4.1
10.8
6.2
—
57.1
(3.1)
5
cio2
36.0
4.5
7.8
1.0
40.9
90.2
(4.9)
ci2
54.8
6.6
13.6
12.4
-
86.4
(2.4)
10
cio2
54.8
5.5
9.0
2.0
81.9
153.2
(4.2)
per Year
, mgd
ci2
172.3
20.0
17.3
62.2
—
271.4
(1.5)
50
cio2
172.3
9.3
16.6
4.8
303.5
506.5
(2.8)
100
ci2
313.0
30.5
19.1
124.5
-
487.1
(1.3)
cio2
313.0
11.1
21.8
9.6
607.1
962.6
(2.6)
-------
TABLE 45. DISINFECTION COST SUMMARY FOR CASE F
Annual Cost, Thousand $
Treatment Plant Size
Cost Item
Contact Basin
Generation and
Feed Equipment
0 & M Cost
Chemical Cost
ci2
NaC102
Total Cost
$1000/yr
(471000 gal)
C12
14.
2.
6.
1.
-
24.
(6.
7
2
2
2
3
7)
1
cio2
14.7
1.4
6.0
0.1
2.0
24.2
(6.6)
5
ci2
36.0
3.82
8.0
6.2
-
54.0
(3.0)
cio2
36.0
2.6
6.8
0.2
9.8
55.4
(3.0)
per Year
, mgd
10
ci2
54.8
5.1
9.1
12.4
-
81.4
(2.2)
cio2
54.8
3.5
7.1
0.5
19.6
85.5
(2.3)
ci2
172.3
11.9
16.3
30.3
-
230.8
(1.3)
50
cio2
172.3
6.1
9.5
1.2
72.8
261.9
(1.4)
100
ci2
313.0
21.8
26.2
60.6
-
421.6
(1.2)
cio2
313.0
7.8
12.0
2.3
145.7
480.8
(1.3)
-------
The costs summarized in Tables 40 through 45 include the cost of contact
facilities. The economies of scale in contact basin construction are an im-
portant reason why the unit cost of disinfection decreases with increasing
plant size. In our cost analysis, we have not differentiated between contact
basin designs for chlorine dioxide and chlorine. Rather, we have expressed
the differences between disinfectants as differences in required dose. Hence,
the capital costs of contact basins are identical for the two disinfectants
for all cases and plant sizes.
For large plants, chemical usage will be sufficiently large to permit
purchasing at lower unit costs• This cost advantage is greater for chlorine
than for chlorine dioxide on a percentage basis. Hence, the results of this
analysis show greater cost advantages for chlorine at large plants (50 and 100
mgd) than at smaller ones.
Freight costs have not been included in these cost comparisons. Were
freight costs to be considered, the cost advantage of chlorine might be less-
ened in some cases (e.g., Cases E and F) in view of the lesser quantities of
chemicals that must be transportd when chlorine dioxide is used (Table 38).
DISCUSSION
When the costs of disinfection of municipal wastewater effluents using
chlorine dioxide and chlorine are compared as above (using a coliform standard
as the disinfection criterion and assuming presently available technology and
current market prices), chlorine dioxide generally is not cost-effective in
the usual sense. However, several factors should be considered to put this
conclusion into proper perspective.' These are: (1) the potential for reduc-
ing the costs of chlorine dioxide generation, and (2) the intangible advan-
tages of chlorine dioxide that are not taken into account in the "cost-to-
kill-coliforms" comparison.
Potential for Cost Reduction
The possibility of reducing the unit cost of chlorine dioxide hinges
largely on the potential for reducing the unit cost of the sodium chlorite
reactant, or for developing a process suitable for generating chlorine dioxide
from chlorate at the wastewater treatment plant site. Investigation of these
possibilities was beyond the scope of this work, and hence we can only specu-
late as to the probability of success in these directions.
Sodium chlorite is produced commercially from the reduction of chlorine
dioxide (produced from chlorate) by hydrogen peroxide in sodium hydroxide
solutions (105). With the increase of interest in chlorine dioxide for water
and wastewater disinfection, an economic incentive exists for increasing the
efficiency of the current commercial production techniques for sodium chlorate
or for developing a new technology for sodium chlorite production from chlo-
rate. It is conceivable that an expansion of the market for sodium chlorite
in water and wastewater treatment could result in a relative decrease in the
unit cost. This could occur as a result of economics of scale in production
112
-------
and marketing, as well as by virtue of increased competition. Presently, one
supplier dominates the market.
There is a strong motivation for developing a feasible process to gene-
rate chlorine dioxide from chlorate at the wastewater treatment site. Based
on raw material costs, the unit cost of chlorine dioxide could be reduced by
approximately half if chlorate rather than chlorite were used. Processes
based on chlorate are widely used in the pulp and paper industry where large
quantities of chlorine dioxide are required. However, there are reputed to be
substantial problems of smaller-scale reactor instability, resulting occasion-
ally in minor explosions. From the perspective of the required operator
training, such a generation facility probably would not be feasible in the
context of wastewater treatment. Hence, the wide-spread use of chlorate-based
generation technology would require further process development, the evalua-
tion of which is beyond the scope of our work.
There is no appreciable potential for improving the yield of chlorine
dioxide from chlorite. In the simple, commercially available reactor studied
in this work, the chlorine dioxide yield closely approximated that correspond-
ing to accepted reaction stoichiometry.
Intangible Advantages of Chlorine Dioxide over Chlorine
The additional advantages of chlorine dioxide compared to chlorine that
were not considered quantitatively in the cost analysis in this section are
negligible amounts of halogenated byproducts and superior virus inactivation.
These advantages of chlorine dioxide compared to chlorine have been demonstra-
ted experimentally in this work. The superiority of chlorine dioxide with
respect to virus inactivation can be concluded only tentatively from the re-
sults of this work; a program of experimental verification is necessary. The
advantage of chlorine dioxide over chlorine with respect to formation of halo-
genated byproducts is unequivocal. It should weigh heavily in the ultimate
decision as to whether to employ chlorine dioxide as an alternative to chlo-
rine, in view of the concern over halogenated and other hazardous pollutants
in the nation's water supplies.
113
-------
REFERENCES
1. APHA-AWWA-WPCF, Standard Methods for the Examination of Water and Waste-
water, 14th ed., American Public Health Association, Washington, D.C.,
1976.
2. Kott, Y. Estimation of Low Numbers of _E_. coli Bacteriophage by Use of
the MPN Method. Appl. Microbiol., 14(2):141, 1966.
3. Berg, J., E. M. Aieta, and P. V. Roberts. Effectiveness of Chlorine
Dioxide as a Wastewater Disinfectant. In: Proceedings of National
Symposium on Wastewater Disinfection, A. Venosa, ed. U.S. Environmental
Protection Agency, Cincinnati, Ohio, 1978.
4. Atlantic Research Corp. Phage Testing and Evalutiori Program, Final
Report. Alexandria, Virginia, Oct., 1978.
5. Henderson, J. E., G. R. Peyton, and W. H. Glaze. A Convenient Liquid-
Liquid Extraction Method for the Determination of Halomethanes in Water
at the Parts-per-Billion Level. In: Identification and Analysis of
Organic Pollutants in Water, L.H. Keith, ed. Ann Arbor Science Publish-
ers, Inc., Ann Arbor, Michigan, pp. 105-112.
6. Palin, A. T. Methods for the Determination, in Water, of Free and
Combined Available Chlorine, Chlorine Dioxide and Chlorite, Bromine,
Iodine and Ozone, Using Diethyl-p-phenylene Diamine (DPD). J. Inst.
Water Engr., 21:537, 1967.
7. Palin, A. T. Analytical Control of Water Disinfection with Special
Reference to Differential DPD Methods for Chlorine, Chlorine Dioxide,
Bromine, Iodine and Ozone. J. Inst. Water Engr., 28:139-154, 1974.
8. Miltner, R. J. Measurement of Chlorine Dioxide and Related Products.
In: Proceedings AWWA Water Quality Technology Conf., Dec. 6-7, 1976,
San Diego, California. Paper No. 2A-5; AWWA, Denver, Colorado, 1977.
9. Granstrom, M., and G. Lee. Generation and Use of Chlorine Dioxide in
Water Treatment. J. AWWA, 50:1453, 1958.
10. Ingols, R., and G. Ridenour. Chemical Properties of Chlorine Dioxide in
Water Treatment. J. AWWA, 40:1207, 1948.
11. White, G. Handbook of Chlorination. Van Nostrand Reinhold, New York,
1972.
114
-------
12. Gordon, G., R. G. Kieffer, and D. H. Rosenblatt. The Chemistry of
Chlorine Dioxide. In: Progress in Inorganic Chemistry, Vol. 15., S. J.
Lippard, ed. Wiley-Interscience, New York, 1972.
13. Masschelein, W. J. Chlorine Dioxide: Chemistry and Environmental
Impact of Oxychlorine Compounds. Ann Arbor Science, Ann Arbor,
Michigan, 1979.
14. Chick, H. An Investigation of the Laws of Disinfection. J. Hygiene,
8:92, 1908.
15. McKee, J. E., et al. Report on the Disinfection of Settled Sewage.
California Institute of Technology, Sanitary Engineering Research, 1957.
16. Collins, H. F., and R. E. Selleck. Process Kinetics of Wastewater
Chlorination. SERL Report No. 72-5, University of California, College
of Engineering and School of Public Health, Berkeley, California, 1972.
17. Wei, J. H., and S. L. Cheng. A Multi-Poisson Distribution Model for
Treating Disinfection Data. In: Disinfection—Water and Wastewater, J.
D. Johnson, ed. Ann Arbor Science, Ann Arbor, Michigan, 1975.
18. Collins, H. F., et al. Problems in Obtaining Adequate Sewage Disinfec-
tion. In: Proceedings of the National Specialty Conference on Disin-
fection. American Society of Civil Engineers, 1970.
19. Hall, E. S. Quantitative Estimation of Disinfection Interferences. J.
of Soc. for Water Treatment and Examination, 22(3):153, 1973.
20. Wyss, 0. Theoretical Aspects of Disinfection by Chlorine—Public Health
Hazards of Microbial Pollution of Water. Proc. Rudolphs Res. Conf.,
Rutgers University, New Brunswick, New Jersey, 1961.
21. Fair, G. M., and H. Thomas. Natural Purification of River Muds and Pol-
lutional Sediments. Sewage Works J., 13(270):756, 1941.
22. Fair, G. M., and J. C. Geyer. Water Supply and Wastewater Disposal.
Wiley, New York, 1954.
23. Phelps, E. B. Stream Sanitation. Wiley, New York, 1944.
24. Card, S. Chemical .Inactivation of Viruses. In: Nature of Viruses.
CIBA Foundation Symposium, 1957. pp. 123-146.
25. Watson, H. E. A Note on the Variation of the Rate of Disinfection with
Change in the Concentration of the Disinfectant. J. of Hygiene, 8:536,
1908.
26. Fair, G. M., et al. The Behavior of Chlorine as a Water Disinfectant.
J. AWWA, 40:1051, 1948.
115
-------
27. Selleck, R. E., B. M. Saunier, and H. F. Collins. Kinetics of Bacterial
Deactivation with Chlorine. J. Env. Eng. Div., ASCE, 104(EE6):1197-
1210, 1978.
28. Levenspiel, 0. Chemical Reaction Engineering, 2nd ed. John Wiley &
Sons, New York, 1972.
29. Sepp, E. Tracer Evaluation of Chlorine Contact Tanks. State of Cali-
fornia, Health and Welfare Agency, Department of Health, Sanitary Engi-
neering Section, Berkeley, California, 1977.
30. Yao, K. M., M. T. Habibian, and C. R. O'Melia. Water and Wastewater
Filtration: Concepts and Applications. Environ. Sci. Tech., 5:11,
1971.
31. Buchan, R. E., and N. E. Gibbons, eds. Bergey's Manual of Determinative
Bacteriology, Williams and Williams Co., Baltimore, Maryland, 1974.
32. Saunier, B., and R. E. Selleck. Kinetics of Breakpoint Chlorination and
of Disinfection. Report No. 76-2, University of California, College of
Engineering and School of Public Health, Berkeley, California, 1976.
33. Feben, D., and M. J. Taras. Chlorine Demand Constants. J. AWWA,
43:922, 1951.
34. Berndt, H., and H. J. Linneweh. Comparative Quantitative Studies on the
Bactericide Activity of Chlorine Dioxide and Chlorine. Arch. Hyg.
Bakteriol., 153:41, 1969. In: Chem. Abs., 71:19528e, 1969.
35. Ridenour, G. M., R. S. Ingols, and E. H. Armbruster. Sporicidal Proper-
ties of Chlorine Dioxide. Water and Sewage Works, 96(8):279, 1949.
36. Ridenour, G. M., and R. S. Ingols. Bactericidal Properties of Chlorine
Dioxide. J. AWWA, 39:561, 1947.
37. Walters, G. E. Chlorine Dioxide and Chlorine: Comparative Disinfec-
tion. M.S. Thesis, Johns-Hopkins University, Baltimore, Maryland, 1976.
38. Benarde, M. A., B. M. Israel, V. P. Olivieri, and M. L. Granstrom.
Efficiency of Chlorine Dioxide as a Bactericide. Appl. Microbiol.,
13(5):776, 1965.
39. Benarde, M. A., W. B. Snow, V. P. Olivieri, and B. Davidson. Kinetics
and Mechanism of Bacterial Disinfection by Chlorine Dioxide. Appl.
Microbiol., 15(2):257, 1967.
40. Longley, K.E., B. E. Moore, and C. A. Sorber. Comprison of Chlorine and
Chlorine Dioxide as Wastewater Disinfectants. Presented at Water Poll.
Control Fed. Annual Meeting, Anaheim, California, Oct. 1-6, 1978.
116
-------
41. Aieta, E. A., J. D. Berg, P. V. Roberts, and R.C. Cooper. Comparison of
Chlorine Dioxide and Chlorine in the Disinfection of Wastewater. Pre-
sented at Water Poll. Control Fed. Annual Meeting, Anaheim, California,
Oct. 1-6, 1978.
42. Adams, A. P., et al. Bacterial Aerosols from Cooling Towers. J. WPCF,
50:2362, 1978.
43. Moffa, P. E., and J. E. Smith. Bench-Scale High-Rate Disinfection of
Combined Sewer Overflows with Chlorine and Chlorine Dioxide. EPA-670/2-
75-021, U.S. Environmental Protection Agency, Cincinnati, Ohio, April
1975.
44. Englebrecht, R. S., and E. 0. Greening. Chlorine-Resistant Indicators.
In: Indicators of Viruses in Water and Food, G. Berg. ed. Ann Arbor
Science Publishers, Inc., Ann Arbor, Michigan, 1978.
45. Geldreich, E. Bacterial Populations and Indicator Concepts in Feces,
Sewage, Stormwater and Solid Wastes. In: Indicators of Viruses in
Water and Food, G. Berg. ed. Ann Arbor Science Publishers, Inc., Ann
Arbor, Michigan, 1978.
46. Hackney, C. R., and G. K. Bissonnette. Recovery of Indicator Bacteria
in Acid Mine Streams. J. WPCF, 50:775, 1978.
47. Matson, E. A., S. G. Horner, and J.D. Buck. Pollution Indicators and
Other Microorganisms in River Sediment. J. WPCF, 50:13, 1978.
48. Berg, G., et al. Validity of Fecal Coliforms, Total Coliforms, and
Fecal Streptococci as Indicators of Viruses. Appl. and Environ. Micro-
biol., 36(6):880, 1978.
49. Berg, G., and T. Metcalf. Indicators of Viruses in Waters. In: Indi-
cators of Viruses in Water and Food, G. Berg. ed. Ann Arbor Science
Publishers, Inc., Ann Arbor, Michigan, 1978.
50. Cronier, S., P. V. Scarpino, and M. L. Zink. Chlorine Dioxide Destruc-
tion of Viruses and Bacteria in Water. In: Water Chlorination: Envi-
ronmental Impact and Health Effects, Vol. 2, R. Jolley et al., eds. Ann
Arbor Science Publishers, Inc., Ann Arbor, Michigan, 1978.
51. Berg, J., E. M. Aieta, and P. V. Roberts. . Effectiveness of Chlorine
Dioxide as a Wastewater Disinfectant. In: Progress in Wastewater
Technology. Proceedings of the National Symposium, Cincinnati, Ohio,
September 1978, A. D. Venosa, ed. EPA-600/9-79-018, U.S. Environmental
Protection Agency, Cincinnati, Ohio, June 1979.
52. Snead, M. C., et al. Biological Evaluation of Benefits of Maintaining a
Chlorine Residual in Water Supply Systems. Presented at American Chemi-
cal Society Meeting, Anaheim, California, March 1978.
117
-------
53. Kenyon, K. F., and S. A. Schaub. Free Available Chlorine Disinfection
of Monodisperse Enteric Organisms. Presented at American Chemical
Society Meeting, Anaheim, California, March 1978.
54. Katzenelson, E., and S. Kedmu. Unsuitability of Poliovirus as Indica-
tors of Virological Water Quality. Appl. and Environ. Microbiol.,
37(2):343, Feb. 1979.
55. Kott, Y., et al. Bacteriophages as Viral Pollution Indicators. Water
Res., 8:165, 1974.
56. Goyal, S. M., C. P. Gerba, and J. L. Melnick. Prevalence of Human
Enteric Viruses in Coastal Canal Communities. J. WPCF, 50:2247, 1978.
57. Goyal, S. M., et al. Human Enteroviruses in Oysters and Their Overlying
Waters. Appl. and Environ. Microbiol., 37(3):572, March 1979.
58. Vaughn, J. M., et al. Survey of Human Enterovirus Occurrence in Fresh
and Marine Surface Waters on Long Island. Appl. and Environ. Micro-
biol., 38(2):290, 1979.
59. Floyd, R., D. G. Sharp, and J. D. Johnson. Inactivation by Chlorine of
Single Poliovirus Particles in Water. Environ. Sci. Tech., 13(4):438,
April 1979.
60. Floyd, R., and D. G. Sharp. Viral Aggregation: Buffer Effects in the
Aggregation of Poliovirus and Reovirus at Low and High pH. Appl. and
Environ. Microbiol., 38(3):395, 1979.
61. Sharp, D. G., et al. Virian Aggregation and the Kinetics of Water
Disinfection by Chlorine and Bromine. Presented at American Chemical
Society Meeting, Anaheim, California, March 1978.
62. Stagg, C.H., C. Wallis, and C. H. Ward. Inactivation of Clay-Associated
Bacteriophage MS-2 by Chlorine. Appl. and Environ. Microbiol.,
33(2):385, Feb. 1977.
63. Hejkal, T. W., F. M. Wellings, P. A. LaRock, and A. L. Lewis. Survival
of Poliovirus Within Organic Solids During Chlorination. Appl. and
Environ. Microbiol., 38(1):114, 1979.
64. Young, D. C., and D. G. Sharp. Partial Reactivation of Chlorine-Treated
Echovirus. Appl. and Environ. Microbiol., 37(4):766, April 1979.
65. Scarpino, P. V., et al. Effect of Particulates on Disinfection of
Enteroviruses in Water by Chlorine Dioxide. EPA-600/2-79-054, U.S.
Environmental Protection Agency, Cincinnati, Ohio, July 1979.
66. Buxton, G. V., and S. A. Ross. Wastewater Disinfection—Toward a
National Policy. J. WPCF, 51(8):2033, 1979.
118
-------
67. Kinney, E. C., D. W. Drummond, and N.B. Hanes. Effects of Chlorination
on Differentiated Coliform Groups. J. WPCF, 50:2307, 1978.
68. Cabelli, V. J. Relationship of Microbial Indicators to Health Effects
of Marine Bathing Beaches. Amer. Jour. Public Health, 69(7):690-696,
1979.
69. Davis, E. M., et al. Bacterial Relationships in Storm Waters. Water
Res.Bull., 13(5):895, Oct. 1977.
70. Eliassen, R. Coliform Aftergrowth in Chlorinated Storm Overflows. J.
San. Eng. Div., ASCE, 94(2):371, 1968.
71. Kittrell, F. W., and S. A. Furfuri. Observations of Coliform Bacteria
in Streams. J. WPCF, 35(11):1361, 1963.
72. Mitchell, R., and C. Chamberlin. Survival of Indicator Organisms.
In: Indicators of Viruses in Water and Food, G. Berg. ed. Ann Arbor
Science Publishers, Inc., Ann Arbor, Michigan, 1978.
73. Grabow, W. 0. K., 0. W. Prozesky, and J. S. Burger. Behavior in a River
and Dam of Coliform Bacteria with Transferable or Non-Transferable Drug
Resistance. Water Res., 9:777, 1975.
74. Gordon, R. W., and C. B. Fliermans. Survival and Viability of _E_. coli
in Thermally Altered Reservoir. Water Res., 12:434, 1978.
75. Deaner, D. G., and K. D. Kerri. Regrowth of Fecal Coliforms. J. AWWA,
61:465, 1969.
76. Camper, A. K., and G. H. McFeters. Chlorine Injury and the Enumeration
of Waterborne Coliform Bacteria. Appl. and Environ. Microbiol.,
37(3):633, 1979.
77. Hendricks, C. W.,and S. M. Morrison. Multiplication and Growth of Se-
lected Enteric Bacteria in Clear Mountain Stream Water. Water Res.,
1:567, 1967.
78. Smith, P. R., E. Farrell, and K. Dunican. Survival of R j^. coli in Sea
Water. Appl. Microbiol., 27(5):983, 1974.
79. Verstraete, W., and J. P. Voels. Comparative Study of _E_. coli Survival
in Two Aquatic Ecosystems. Water Res., 10:129, 1976.
80. Cooke, E. M. Escherichia coli and Man. Churchill Livingstone, London,
1974.
81. Welch, W. J., and J. S. Lee. Modeling Techniques for Estimating Fecal
Coliforms in Estuaries. J. WPCF, 50:862, 1978.
119
-------
82. Standridge, J. H., et al. Effects of Waterfowl (Anas platiphynehas) on
Indicator Bacteria Populations in a Recreational Lake in Madison,
Wisconsin. Appl. and Environ. Microbiol., 38(3):547, 1979.
83. Seidler, R. J., J. E. Morrow, and S. T. Bagley. Klebsiella in Drinking
Water Emanating from Redwood Tanks. Appl. and Environ. Microbiol.,
33(4):893, April 1977.
84. Knittel, M. D., et al. Colonization of the Botanical Environment by
Klebsiella Isolates of Pathogenic Origin. Appl. and Environ. Micro-
biol., 34(5):557, Nov. 1977.
85. Seidler, R.J. Coliforms in Drinking Water Emanating from Redwood Reser-
voirs. EPA-600/2-79-049, U.S. Environmental Protection Agency,
Cincinnati, Ohio, July 1979.
86. Brown, C., and R. J. Seidler. Potential Pathogens in the Environment:
Klebsiella pneumoniae, a Taxonomlc and Ecological Enigma. Appl. Micro-
biol., 25(3):900, June 1973.
87. Katzenelson, E. Survival of Viruses. In: Indicators of Viruses in
Water and Food, G. Berg. ed. Ann Arbor Science Publishers, Inc., Ann
Arbor, Michigan, 1978.
88. Bitten, G. Survival of Enteric Viruses. In: Water Pollution Micro-
biology, Vol. 2, R. Mitchell, ed. John Wiley, New York, 1978.
89. Obrien, R. T., and J. S. Newman. Inactivation of Polioviruses and
Coxsachieviruses in Surface Water. Appl. and Environ. Microbiol.,
33(2):334, 1977.
90. Myhrstad, J. A., and Samdal, J. E. Behavior and Determination of Chlo-
rine Dioxide. J. AWWA, 61(4):205-208, 1969.
91. Miller, W. G., R. G. Rice, C. M. Robson, R. L. Scullen, W. Kuhn, and H.
Wolf. An Assessment of Ozone an Chlorine Dioxide Technologies for
Treatment of Municipal Water Supplies, EPA-60012-78-147, Municipal
Environmental Research Laboratory, Office of Research and Development,
U.S. Environmental Protection Agency, Cincinnati, Ohio, August 1978.
pp. 225-290.
92. Symons, J. M. Ozone, Chlorine Dioxide, and Chloramine as Alternatives
to Chlorine for Disinfection of Drinking Water. In: Water
Chlorination: Environmental Impact and Health Effects, Vol. 2, R.
Jolley, H. Gorchev, and D. H. Hamilton, eds. Ann Arbor Science
Publishers, Inc., Ann Arbor, Michigan, 1978. pp. 555-560.
93. Schnitzer, M. Humic Substances: Chemistry and Reactions. In: Soil
Organic Matter, M. Schnitzer and S. U. Khan, eds. Elsevier Scientific
Publishing Company, Amsterdam, 1978. pp. 1-64.
120
-------
94. Manka, J., R. Menachem, Mandelbaum, and A. Bortinger. Characterization
of Organics in Secondary Effluents. Environ. Sci. Tech., 12(8):1017-
1020, November 1974.
95. Rook, J. J., A. A. Gras, B. G. Van der Heyden, and J. de Wee. Bromide
Oxidation and Organic Substitution in Water Treatment. J. Environ. Sci.
and Health, A, 13(2):91-116, February 1978.
96. Jolley, R. L. Chlorine-Containing Organic Constituents in Chlorinated
Effluents. J. WPCF, 47(3):601, 1975.
97. Noack, M. G., and R. L. Doerr. Reactions of Chlorine, Chlorine Dioxide
and Mixtures Thereof with Humic Acid: An Interim Report. In: Water
Chlorination: Environmental Impact and Health Effects, Vol. 2. Ann
Arbor Science Publishers, Inc., Ann Arbor, Michigan, 1978. pp. 49-58.
98. Miltner, R. J. The Effect of Chlorine Dioxide on Trihalomethanes in
Drinking Water. M.S. Thesis, Department of Civil and Environmental
Engineering, University of Cincinnati, Cincinnati, Ohio, 1976.
99. Hunt, D. A., and J. Springer. Preliminary Report on a Comparison of
Total Coliform and Fecal Coliform Values in Shellfish Growing Area
Waters. In: Proceedings 8th National Shellfish Sanitation Workshop.
U.S. Dept. of Health, Education, and Welfare; Food and Drug Administra-
tion, January 1974. pp. 97-104.
100. McKee, J. E., and H. W. Wolf. Water Quality Criteria. California State
Water Quality Control Board, Sacramento, California, 1963. p. 119.
101. Process Design, Performance and Economic Analysis Handbook, Biological
Wastewater Treatment Process. Gulp, Wesner, Gulp, El Dorado Hills,
California, 1977.
102. Estimating Costs for Water Treatment as a Function of Size and Treatment
Plant Efficiency. EPA-600/2-78-182, U.S. Environmental Protection
Agency, Cincinnati, Ohio, 1978.
103. Estimate from Olin Water Services, Overland Park, Kansas. W. J. Ward,
Director, Technology, April 1979.
104. Estimate from Rio Linda Chemical Company, Rio Linda, California.
J. Hicks, President, January 1980.
105. Chemical Marketing Reporter, 7 January 1980.
106. Gall, R. Chlorine Dioxide. An Overview of Its Preparation, Properties,
and Uses. In: Ozone/Chlorine Dioxide Oxidation Products of Organic
Materials, Rice and Contruvo, eds. Ozone Press Internation, Cleveland,
Ohio, 1978.
121
-------
APPENDIX A
COLIFORM INACTIVATION DATA
Table A-l. Log[Surviving Bacteria] in Eight Replicate Experiments Using
Chlorine Dioxide as Disinfectant
Table A-2. Log[Surviving Bacteria] in Eight Replicate Experiments Using
Chlorine as Disinfectant
Table A-3. Full-Scale Experiments Conducted at Dublin-San Ramon Treatment
Facility on Nitrified, Filtered Secondary Effluent Using Chlorine
Dioxide and Chlorine
Table A-4. Results of Laboratory Experiments with Dublin Filtered Nitrified
Activated-Sludge Effluent
Table A-5. Results of Laboratory Experiments with Unfiltered and Filtered
Palo Alto Secondary Effluent Using Chlorine Dioxide
Table A-6. Results of Laboratory Experiments with Unfiltered and Filtered
Palo Alto Secondary Effluent Using Chlorine
Table A-7. Results of Experiments with San Jose Secondary Effluent
Table A-8. Results of Experiments with San Jose Nitrified Effluent
Table A-9. Results of Experiments with San Jose Filtered, Nitrified Effluent
122
-------
TABLE A-l. LOG[SURVIVING BACTERIA] IN EIGHT REPLICATE EXPERIMENTS
USING CHLORINE DIOXIDE AS DISINFECTANT
DOSE :
EXPERIMENT:
N>
CO
0 MIN
5 MIN
I-
o
O 15 MIN
O
30 MIN
2mg/l
5 mg/l
10 mg/l
1|2|3|4|5|6|7|8
5.91 6.05 6.09 6.73 6.79 6.41 6.48 5.85
5.90 6.18 6.45 6.67 5.91
5.90
5.70
508 4.41 434 5.00 499 521 5.08 503
5 95 4 36 4 50 4 97 4 99 5 02 4 94 5 04
4.30 5.08 5.11 5.34 5.41 5.30
5.00 5.00 5.00 5.00 5.00
5.26 4.38 4.38 5.06 5.25 5.17 5.20 493
5.00 4.38 4.26 4.95 5.44 5.00 5.09 4.61
4.48 4.30 4.78 5.15 5.48 5.38 4.60
5.04 5.00 5.15 5.26 4.78
4.00 4.20 3.60 4.78 4.81 5.09 5.15 4.91
4.78 4.32 3.04 4.81 4.90 4.92 5.05 5.08
4.48 4.30 4.90 5.26 5.08 5.45 4.11
4.30 5.00 5.08 5.00 5.36
1
5.83
5.60
5.00
399
3 91
3.23
3.20
3.00
3.33
3.37
3.11
2
60?
6.28
2.00
?no
1.78
1.90
1.90
3
6.26
278
3 30
1.70
1.20
1.30
1.30
4 I
6.83
3.76
390
3.60
3.30
2.38
2.27
2.30
3.00
2.89
3.06
3.11
5 I
6.76
345
375
4.15
4.04
2.89
3.15
2.90
3.15
3.15
3.50
6
6S?
341
330
3.00
in?
3.82
3.72
3.70
3.12
3.15
3.83
7 i 8
6.67 5.77
5.88
5.00
3.98 3 20
4 08 3 30
3.30 3.70
3.30
2.28 2.68
2.28 3.04
3.00 2.30
3.00
2.99 3.46
3.24 3.56
3.00 3.32
1 | 2 I 3 |
5.87 6.12 6.15
5.60 6.38 6.20
1 90 1.30 1 78
1 95 1 00 1 78
2.18 0.30 1.56
2.04
1.98 0.60 0.30
1.75 0.48 0.34
1.49
0.31 0.34 0.34
0.60
0.40
4 | 5 |
6.83 6.70
2.62 2 15
246 2 15
1.70
2.11 1.40
1.68
1.66
0.34 0.34
0.30 0.30
6 | 7
6.73 6.93
2.26 2.45
2 56 2 43
2.32 1.95
1.98
1.60 1.18
1.56 1.30
1.20 1.08
1.11 0.90
1.15
8
5.93
5.75
1.60
208
1.65
1.53
0.70
1.00
1.15
0.48
-------
TABLE A-2. LOG[SURVIVING BACTERIA] IN EIGHT REPLICATE EXPERIMENTS
USING CHLORINE AS DISINFECTANT
DOSE:
2mg/l
5 mg/l
10 mg/l
EXPERIMENT:
0 MIN
5 MIN
H
O
o
o
15 MIN
30 MIN
1
5 .85
5.90
4.78
5.00
4.48
4.34
1213,4,5,6,7,8
6.09 6.00 6.43 6.81 6.63 6.62 5.89
6.28 6.18 6.74 5.90
5.78
5.48
5.75 5.95 6.16 6.13 6.30 6.34 5.37
5.30 5.70 6.48 6.15 6.28 5.40
5.60
4.70 4.92 5.34 4.72 5.45 5.41 4.60
4 62 4 79 5 25 4 46 5.62 5.55 4.58
4.30 4.90 5.72 4.78 5.68
4.30 4.84 5.53 4.48
3.30 3.30 4.70 4.00 4.78 4.41 4.00
3.30 4.58 3.70 4.70 4.80 3.95
4.00 4.60 5.00 4.78 4.00
4.00 4.85 5.08
1 I 2
5.78 6.14
5.66 6.32
5.30
4.09 3.83
4.08 3.90
3.60 4.08
3.95
2.84 2.28
3.15 2.00
3.48 3.30
2.71 1.88
2.48 1.90
1.30
1 3
6.10
6.36
3.96
4.08
3.78
3.70
2.36
3.23
1.76
1.30
2.15
2.00
,4,5
6.79 6.83
4.15 4.15
4.06 4.15
4.41 4.30
4.14 4.49
2.78 3.49
3.34 3.70
3.48 3.30
4.30
2.91 3.23
2.87 3.21
3.38 3.20
,6,7
6.67 6.63
4.53 3.90
4.26
4.26 3.85
4 79 3 75
2.86 3.30
2.89 3.11
2.84 3.00
1 8
5.83
6.30
5.60
3.86
4.06
4.08
3.95
3.01
340
3.34
3.60
2.34
2.73
2.86
2.60
1
5.92
6.00
791
2.81
??7
1 81
1.76
1.57
1.04
0.60
0.48
,2,3
6.30 6.08
5.84
2.51 2.34
2.26 2.40
2.16 2.42
2.07 2.09
1.00 1.74
1.41 1 76
1.20 1.50
1.04 1.48
0.60 1.15
0.60 0.34
0.54
I 4
6.70
3.77
1.34
1.30
1.53
1.30
I 5 | 6
6.84 6.67
281 3 24
2.69 3.08
2.42
2.04 2.24
1 91
1.75
0.30 0.95
0.70
1 7 | 8
6.60 5.73
5.75
5.30
3 24 3 00
2.67
2.22 2.04
1 79 1 66
1.91 1.41
1.23 0.48
1.26 0.34
0.48
-------
TABLE A-3. FULL-SCALE EXPERIMENTS CONDUCTED AT DUBLIN-SAN RAMON
TREATMENT FACILITY ON NITRIFIED, FILTERED SECONDARY EFFLUENT USING
CHLORINE DIOXIDE AND CHLORINE
Date
1/16/79
2/6/79
2/13/79
2/13/79
Repli-
Disin- Dose cate
fectant mg/1 No.
ClOo 4.76 1
2
3
4
5
6
C102 3.12 1
2
3
4
5
6
C102 2.25 1
2
3
Cl, 4.60 1
2
3
Average
0
4.28
4.38
4.30
4.30
4.45
4.48
3.82
3.90
3.79
3.87
4.04
4.93
4.70
4.75
4.99
4.84
4.69
4.59
Log Surviving Total
log10[N/100ml]
Contact
8
<0.30
0.78
<0.30
<0.30
<0.30
<0.30
0.30
<0.30
0.30
0.30
<0.30
0.30
2.41
2.72
2.33
2.70
2.72
2.51
Time , min
24
<0.30
<0.30
0.30
0.30
<0.30
0.30
<0.30
0.30
0.30
0.30
0.30
<0.30
2.41
2.27
2.36
2.45
2.40
2.61
Coliforms
48
<0.30
0.30
<0.30
<0.30
0.30
0.30
0.30
<0.30
0.30
0.30
0.30
<0.30
2.28
2.22
2.39
2.24
2.36
2.34
2/13/79
Cl.
19.60
4.83
2.08
1.04
0.30
125
-------
TABLE A-4. RESULTS OF LABORATORY EXPERIMENTS WITH DUBLIN FILTERED
NITRIFIED, ACTIVATED-SLUDGE EFFLUENT
Disin- Dose
fectant mg/1
C102 2
5
10
C12 2
5
10
Repli-
cate
No.*
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
Average
0
3.74
4.08
5.18
3.78
4.20
4.90
3.93
4.08
4.99
6.07
4.24
4.98
3.93
4.54
4.95
3.88
4.43
4.86
Log Surviving Total
log10[N/100ml]
Contact
5
<0.30
4.01
4.38
<0.30
2.37
2.06
<0.30
0.30
0.88
3.30
3.84
3.85
_t
2.69
2.08
1.65
2.02
1.37
Time, min
15
<0.30
4.20
4.43
<0.30
2.39
2.09
0.95
<0.30
0.98
3.65
3.87
3.64
2.23
2.76
1.90
1.38
0.30
0.26
Coliforms
30
<0.30
4.18
4.13
1.30
2.35
2.26
<0.30
<0.30
0.34
3.30
3.68
3.74
1.60
2.26
2.62
1.10
<0.30
0.79
Dates of replicate experiments: 1, 1/9/79; 2, 2/16/79; 3, 2/22/79.
'- signifies sample lost.
126
-------
TABLE A-5. RESULTS OF LABORATORY EXPERIMENTS WITH UNFILTERED AND
FILTERED PALO ALTO SECONDARY EFFLUENT USING CHLORINE DIOXIDE
Average Log Surviving Total Coliforms
log10[N/100ml]
Type
of Dose
Effluent mg/1
Unfiltered 2
5
10
Filtered 2
5
10
Repli-
cate
No.*
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
0
5.80
5.06
5.02
5.55
5.84
5.24
5.29
5.72
5.24
5.26
5.26
5.31
5.82
5.00
4.82
5.07
6.22
5.01
4.72
5.31
6.68
4.58
4.86
5.15
Contact
1 5
4.35 4.32
4.40 4.30
4.50 4.36
4.80 4.88
<2.0
4.62
3.14
3.34
1.00
1.84
1.59
2.54
3.04 3.35
3.48 3.28
3.09 2.74
3.98 3.19
<1.0
0.78
0.30
1.64
<1.0
1.00
<0.4
<0.4
Time,
15
4.82
4.13
4.43
4.82
<2.0
2.56
2.46
2.52
<0.4
1.30
0.59
<0.4
3.09
3.46
2.86
3.50
<1.0
0.60
<0.4
0.98
<1.0
<0.4
<0.4
<0.4
min
30
4.04
3.86
4.22
4.60
<2.0
2.52
1.84
2.35
<0.4
<0.4
<0.4
<0.4
3.00
3.22
2.74
2.65
1.0
<0.3
0.3
0.79
<1.0
<0.4
<0.4
<0.4
60
_t
-
-
—
<2.0
2.5
1.08
2.84
—
-
-
—
—
-
-
—
<1.0
<0.3
0.6
0.98
_
-
-
—
Dates of experiments: 1, 8/8/79; 2, 8/14/79; 3, 8/17/79; 4, 8/21/79,
'- signifies that experiment was not conducted.
127
-------
TABLE A-6. RESULTS OF LABORATORY EXPERIMENTS WITH UNFILTERED AND
FILTERED PALO ALTO SECONDARY EFFLUENT USING CHLORINE
Average Log Surviving Total Colifonns
log10[N/100ml]
Type
of
Effluent
Unfiltered
Filtered
Repli-
Dose cate
mg/1 No.
2 1
2
3
5 1
2
3
10 1
2
3
2 1
2
3
5 1
2
3
10 1
2
3
0
4.69
4.92
4.95
4.82
4.78
5.04
4.96
5.23
5.38
5.78
4.50
4.65
4.81
4.74
4.77
4.66
4.89
4.89
Contact
1 5
4.80 4.39
4.48 4.67
5.10 4.48
3.19
2.95
3.34
0.79
1.03
2.72
4.64 4.23
5.09 4.40
4.57 4.42
1.53
1.86
1.64
0.39
0.83
1.47
Time,
15
3.04
3.52
2.04
1.17
1.00
1.40
<0.3
<0.3
1.46
2.69
3.23
3.50
0.67
1.06
1.38
0.70
<0.3
0.66
min
30
3.12
3.30
2.95
0.39
1.00
1.44
<0.3
0.80
2.20
1.30
1.47
2.70
<0.3
0.24
1.22
<0.3
0.15
0.26
60
_t
—
—
0.30
1.00
1.26
—
-
—
_
-
—
<0.3
0.15
0.30
—
-
—
*Dates of duplicate experiments: 1, 8/23/79; 2, 8/24/79; 3, 8/22/79.
- signifies that experiment was not conducted.
128
-------
TABLE A-7. RESULTS OF EXPERIMENTS WITH SAN JOSE SECONDARY EFFLUENT
Disin- Dose
fectant mg/1
C102 2
5
10
C12 2
5
10
Repli-
cate
No.*
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
Average
0
6.08
6.81
6.67
6.40
6.63
6.58
6.16
6.75
6.49
6.30
6.76
6.45
6.27
6.94
6.30
6.29
6.69
6.54
Log Surviving Total
log10[N/100ml]
Contact
5
5.11
3.10
5.71
3.64
4.57
4.56
1.48
2.60
2.59
5.80
6.51
6.42
5.70
4.62
5.06
3.18
2.49
2.15
Time , min
15
4.70
6.60
5.62
3.40
4.21
3.76
_t
1.32
2.76
5.71
5.43
3.23
2.52
2.28
1.70
2.10
1.54
Coliforms
30
4.30
6.30
5.40
3.52
3.77
3.43
—
1.48
—
4.00
5.55
4.45
3.84
2.41
1.78
_
2.30
1.40
Dates of experiments: 1, 6/28/79; 2, 7/10/79; 3, 7/17/79.
- signifies that sample was lost.
129
-------
TABLE A-8. RESULTS OF EXPERIMENTS WITH SAN JOSE NITRIFIED EFFLUENT
Disin- Dose
fectant mg/1
C102 2
5
10
C12 2
5
10
Repli-
cate
No.*
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
Average
0
4.98
4.98
4.25
4.96
4.66
4.53
5.34
4.73
4.21
4.78
4.72
4.35
5.67
4.72
4.40
4.87
5.26
4.30
Log Surviving Total
log10[N/100ml]
Contact
5
3.88
3.55
3.04
2.08
1.04
1.11
1.30
0.85
0.84
4.76
4.43
3.60
3.38
1.78
2.04
1.78
0.84
0.90
Time, min
15
3.85
3.41
3.78
1.50
1.38
0.30
0.30
0.30
0.30
3.77
4.19
3.15
2.00
-
—
0.30
<0.3
0.64
Colifonns
30
3.84
3.47
2.78
1.48
1.95
0.30
_t
0.30
<0.3
3.77
3.35
3.04
_
0.30
2.11
0.30
<0.3
0.75
*Dates of experiments: 1, 7/7/79; 2, 7/12/79; 3, 7/19/79.
'- signifies that sample was lost.
130
-------
TABLE A-9. RESULTS OF EXPERIMENTS WITH SAN JOSE FILTERED, NITRIFIED EFFLUENT
Disin- Dose
fectant mg/1
C102 2
5
10
C12 2
5
10
Repli-
cate
No.*
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
Average
0
2.79
5.46
4.73
4.90
5.38
4.72
4.81
5.71
4.59
4.75
4.91
4.76
1.90
5.84
4.86
4.89
5.84
4.78
Log Surviving Total Coliforms
log10[N/100ml]
Contact
5
_t
4.12
—
0.48
1.11
<0.3
<0.3
0.45
<0.3
4.4
4.26
2.30
0.30
2.53
1.11
0.20
1.94
0.69
Time,
15
1.30
2.95
—
<0.3
0.85
<0.3
<0.3
0.30
<0.3
3.59
3.18
2.00
0.30
0.84
0.95
0.30
1.51
0.65
min
30
3.18
2.00
<0.30
0.75
<0.3
0.30
<0.3
<0.3
2.41
1.74
—
0.3
0.45
0.84
<0.3
1.21
0.52
*Dates of experiments: 1, 7/2/79; 2, 7/13/79; 3, 7/20/79.
- signifies that sample was lost.
131
-------
APPENDIX B
LOG-LOG PLOTS OF MODEL PREDICTIONS AND DATA POINTS FOR ALL EXPERIMENTS
Figure B-l. Inactivation of coliform bacteria by chlorine dioxide in Palo
Alto wastewater - 1978.
Figure B-2. Inactivation of coliform bacteria by chlorine in Palo Alto
wastewater - 1978.
Figure B-3. Inactivation of coliform bacteria by chlorine dioxide in Palo
Alto wastewater - 1979.
Figure B-4. Inactivation of coliform bacteria by chlorine in Palo Alto
wastewater - 1979.
Figure B-5. Inactivation of coliform bacteria by chlorine dioxide in
filtered Palo Alto wastewater.
Figure B-6. Inactivation of coliform bacteria by chlorine in filtered Palo
Alto wastewater.
Figure B-7. Inactivation of coliform bacteria by chlorine dioxide in Palo
Alto wastewater - 1978 and 1979.
Figure B-8. Inactivation of coliform bacteria by chlorine in Palo Alto
wastewater - 1978 and 1979.
Figure B-9. Inactivation of coliform bacteria by chlorine dioxide in Dublin
lab experiments.
Figure B-10. Inactivation of coliform bacteria by chlorine in Dublin lab
experiments.
Figure B-ll. Inactivation of coliform bacteria by chlorine dioxide in Dublin
field experiments.
Figure B-12. Inactivation of coliform bacteria by chlorine in Dublin field
experiments.
Figure B-13. Inactivation of coliform bacteria by chlorine in San Jose
secondary effluent.
Figure B-14. Inactivation of coliform bacteria by chlorine dioxide in San
Jose filtered effluent.
Figure B-15. Inactivation.of coliform bacteria by chlorine dioxide in San
Jose nitrified effluent.
Figure B-16. Inactivation of coliform bacteria by chlorine dioxide in San
Jose secondary effluent.
132
-------
10
-1
10
-2
10
-3
10
-4
10
-5
10
-6
N(T)/N(0)=[(RT)/1.56]
"r=0.86
-2.90
i i i I i I
10
-1
10J
10*
RESIDUAL — TIME IN MG-MIN/L
Figure B-l. Inactivation of coliform bacteria by chlorine dioxide in Palo
Alto wastewater - 1978.
io-i ,
o
K
t)
CO
10'
10
10J
RESIDUAL — TIME IN MG-MIN/L
Figure B-2. Inactivation of coliform bacteria by chlorine in Palo Alto
wastewater - 1978.
133
-------
10-1
10~2
10~3
S iO-4
10
-5
10
-6
4-
N(T)/N(0)=[(RT)/0.67]-2-30
"r=0.86
1 I I 1 M 111 I I I 1 1 I 1 11 1 I I 1 I I I 11 I I I 1 I I I 11 I I I I I I I I
10
-1
10-=
RESIDUAL — TIME IN MG-MIN/L
Figure B-3. Inactivation of coliform bacteria by chlorine dioxide in Palo
Alto wastewater - 1979.
O
I—I
HI
10
1
10
-2
10
-3
10
-4
10
-5
10
-6
N(T)/N(0)=[(RT)/2.21]-2'22
"r=0.83
i i i 1 1 1 i
10
— 1
10
10
10
Figure B-4.
RESIDUAL — TIME IN MG-MIN/L
Inactivation of coliform bacteria by chlorine in Palo Alto
wastewater - 1979.
134
-------
o
I
10
~2
10"3
10-4
10~5
10"6
N(T)/N(0)=[(RT)/.004]'1-81
"r-0.89
10"1 10° 101 10£
RESIDUAL — TIME IN MG-MIN/L
Figure B-5. Inactivation of coliform bacteria by chlorine dioxide in
filtered Palo Alto wastewater.
10
-1
10
-2
10
-3
10
-4
$ 10
-5
10
-6
: N(T)/N(0)=[(RT)/0.84]
Fr-0.86
-3.10
IQ-I 10° IQ! io2
RESIDUAL — TIME IN MG-MIN/L
Figure B-6. Inactivation of coliform bacteria by chlorine in filtered Palo
Alto wastewater.
135
-------
io-i ._
io
I—I
i
3
10-4 1_
10'
10
-6
+ \, + + +
: N(T)/N(0)=[(RT)/1.29]
Fr=0.86
j—i i 11
10-
10
10
10*
RESIDUAL — TIME IN MG-MIN/L
Figure B-7. Inactivation of coliform bacteria by chlorine dioxide in Palo
Alto wastewater - 1978 and 1979.
10-1
10~2
10~3
10-4
10~5
_6 r N(T)/N(0)=[(RT)/3.95]~2-78
10 ^~r=0.88
_j i i 111
ic
1
10
°
io
RESIDUAL — TIME IN MG-MIN/L
Figure B-8. Inactivation of coliform bacteria by chlorine in Palo Alto
wastewater - 1978 and 1979.
136
-------
10"2
ID"3
10-4
10~5
10
-6
N(T)/N(0)=[(RT)/0.
"r=0.78
i i i i mil i i i i mil 1—i i i ii
10
-1
10
0
10
2
RESIDUAL — TIME IN MG-MIN/L
Figure B-9. Inactivation of coliform bacteria by chlorine dioxide in Dublin
lab experiments.
10-1
10"2
10
10
-3 _
-4
10~5 fe-
10"6 fer-
N(T)/N(0)=[(RT)/0.65]-1-62
"r=0.91
10
-1
10
0
10
1
10*
Figure B-10.
RESIDUAL — TIME IN MG-MIN/L
Inactivation of coliform bacteria by chlorine in Dublin lab
experiments.
137
-------
g
cfl
Figure B-ll.
10
-1
10
-2
10
-3
10
-4
10
-5
10
-6
N(T)/N(0)=[(RT)/0.57]-2'20
'r=0.84
I I I
10
-1
i ii i i i i 111
10
0
10J
10
2
RESIDUAL — TIME IN MG-MIN/L
Inactivation of coliform bacteria by chlorine dioxide in Dublin
field experiments.
10
10
-2
10
-3 _
10
-4
10
-5
10
-6
N(T)/N(0)=[(RT)/1.67]-1'79
"r=0.93
10
-1
10°
10J
Figure B-12.
RESIDUAL — TIME IN MG-MIN/L
Inactivation of coliform bacteria by chlorine in Dublin field
experiments.
138
-------
g
io-i
10~2
10~3
10-4
10~5
10~6
N(T)/N(0)=[(RT)/4.06]-2'82
"r-0.86
10
-1
10
101
10<
RESIDUAL — TIME IN MG-MIN/L
Figure B-13. Inactivation of coliform bacteria by chlorine in San Jose
secondary effluent.
10
10
~2
10'
10
-4
10
-5
10
-6
: N(T)/N(0)=[(RT)/.003]
"r=0.81
-1.13
10
-1
10
°
10*
RESIDUAL — TIME IN MG-MIN/L
Figure B-14. Inactivation of coliform bacteria by chlorine dioxide in San
Jose filtered effluent.
139
-------
10
-1
ID
-2
10
3
g
10
-5
10
_6
N(T)/N(OH(RT)/0.78]-2-06
"r=0.75
i i i 1111 ii i i i 111 ni i i i 111 n i i i 111 ni i i i 11111
10
-1
10*
Figure B-15.
RESIDUAL — TIME IN MG-MIN/L
Inactivation of coliform bacteria by chlorine dioxide in San
Jose nitrified effluent.
g
10
10
-2 _
10
10
10-
-4 _
10
-6
- N(T)/N(0)=[(RT)/0.89]
Fr=0.78
-3.18
10"1 10° 101 102
RESIDUAL — TIME IN MG-MIN/L
Figure B-16. Inactivation of coliform bacteria by chlorine dioxide in San
Jose secondary effluent.
140
-------
APPENDIX C
A DETAILED EXAMPLE OF THE CALCULATIONS TO DETERMINE THE REQUIRED
DISINFECTANT FEED CAPACITY—POUNDS PER DAY
These example calculations are for the disinfection by chlorine dioxide
of a conventional activated-sludge treatment effluent with a 1000/100 ml total
coliform standard at the treatment plant outfall. This corresponds to Case C
discribed in the main body of this report, Table 36.
1. Determine the initial bacterial concentration in the waste
stream to be disinfected.
For conventional activated-sludge treatment, the value
read from Table 14 is 1.62 x 106 total coliforms/100 ml.
2. Calculate the required survival ratio, N(t)/N(0), from the
initial bacterial concentration, N(0). For Case C, the
disinfection standard is 1000 total coliforms/100 ml. The
survival ratio is then:
1000 total coliforms/100 ml
1.62 x 106 total coliforms/100 ml = 6'17 X 10
3. Calculate the required residual time product using Eqs . 24
and 25 and the model-fitting constants from Table 24.
N(t)
N(t) = N(0) for Rt < b
where N(0) = initial total coliform bacteria;
N(t) = total coliform bacteria at time t;
R = residual oxidant in mg/1 as measured at time t;
b = lag time coefficient (Rt)' in mg'min/1;
k =-k'/a = velocity coefficient; and
t = time in minutes.
141
-------
The model fitting constants for Case C, disinfection with
chlorine dioxide are:
b - 1.56
k = -2.90
The required residual-time product is then:
N(t)/N(0) = [^]
Solving for Rt:
Rt = [N(t)/N(0)]1/k • b
Rt = [6.17 x 10~4]~1/2'90 • 1.56
Rt = 19.95 mg C102 * min/1
4. Calculate the required residual, based on the provided
contact time, namely one hour. The required residual is
then:
Rt 19.95 mg C102 * min/1
60 min 60 min
R = 0.33 mg C102/l
5. Calculate the disinfectant dosage needed to provide the
required residual disinfectant concentration.
In order to be able to calculate the disinfectant dosage, an empirical
relationship relating disinfectant dose to residual has been developed using
the observed values of oxidant demand (see Table 27 and Reference 33). The
dose required is then given by the sum of the required residual to achieve the
disinfection standard plus the one-hour oxidant demand. The one-hour demand,
however, is also a function of dose, as it must be since demand equals dose
minus residual. A linear regression of log[1-hour oxidant demand] on
log[dose] yields the equation:
Dl-hour ' 3(Dose)a
where Di_ho r = one-hour oxidant demand-mg/1,
3,a = constants, dependent on the oxidant and the wastewater,
Dose = oxidant dose-mg/1.
142
-------
The equation to be solved is:
Dose = Residual required + 3(Dose)
a
The values of 3 and a and the regression correlation coefficients for Palo
Alto and Dublin Wastewaters are shown in Table C-l.
TABLE C-l. CONSTANTS FOR ONE-HOUR DISINFECTANT DEMAND EQUATION
AS A FUNCTION OF DOSE
Case
Type of Effluent)
A, B, and C
(conventional
activated sludge)
D, E, and F
(filtered, nitrified
activated sludge)
. cio2
gar
0.99 0.80 0.97
0.94 0.995 0.99
ci2
6 a
0.63 0.46
1.11 0.80
*
r
0.81
0.98
correlation coefficient
For Case C, the required dose is:
0.80
Dose = 0.33 + 0.99(DOSE)
Dose =2.17 mg/1 C102
The required feed capacity in pounds per day is given by:
2.17[g/m3] x l/454[lb/g] x l/264.2[m3/gal] x 106[million gal/gal]
= 18.1 [pounds C102 per day]/[mgd wastewater treated]
The corresponding values are entered in the row designated "Case C, Cin" i-n
Table 38.
143
------- |