oEPA
United States
Environmental Protection
Agency
Office of Drinking
Water
Washington DC 20460
Center for Environmental
Research Information
Cincinnati OH 45268
Technology Transfer
CERI-88-23
Workshop on
Emerging
Technologies for
Drinking Water
Treatment
-------
CERI-88-23
WORKSHOP ON EMERGING TECHNOLOGIES FOR
DRINKING WATER TREATMENT
April 1988
Sponsored by:
U.S. EPA Office of Drinking Water
U.S. EPA Office of Research and Development
and
Association of State Drinking Water Administrators
-------
TABLE OF CONTENTS
Presentation
SDWA Amendments and Regulations I~l
Stephen W. Clark, Chief, Technical Section, Office of
Drinking Water, US Environmental Protection Agency,
Washington, DC
State Implementation of the SDWA II-l
Frederick A. Marrocco, Vice President, Association of
State Drinking Water Administrators and Chief,
Division of Water Supplies, Pennsylvania Department
of Environmental Resources, Harrisburg
Current and Emerging Treatment Technologies III-l
Robert M. Clark, Director, Drinking Water Research
Division, US Environmental Protection Agency,
Cincinnati, OH
Barriers to Implementing New Technologies IV-1
G. Wade Miller, Executive Director, Association of
State Drinking Water Administrators and President,
Wade Miller Associates, Inc., Arlington, VA
Technical Session: Filtration V-l
Sigurd P. Hansen, Senior Engineer, CWC-HDR, Inc.,
Cameron Park, CA
Technical Session: Disinfection/By-Products VI-1
Rip G. Rice, President, Rice International
Consulting Enterprises, Ashton, MD
Technical Session: Organics VII-1
John E. Dyksen, Senior Project Manager, Malcolm
Pirnie, Inc., Paramus, NJ
Technical Session: Inorganics VIII-1
J. Edward Singley, Vice President, James M. Montgomery
Consulting Engineers, Inc., Gainesville, FL
-------
SDWA Amendments and Regulations
Stephen W. Clark, Chief, Technical Section, Office of Drinking
Water, US Environmental Protection Agency, Washington, DC
1-1
-------
EMERGING TECHNOLOGIES -
REGULA TOR Y IMP A C TS
-------
SOW A -- MAJOR REQUIREMENTS
• 83 CONTAMINANTS BY JUNE 19, 1989
• TWO TREATMENT TECHNIQUES
• FILTRATION (SURFACE SOURCES)
• DISINFECTION (ALL)
» 25 CONTAMINANTS IN 1991, 1994, . . . etc,
-------
EPA's CURRENT SCHEDULE
• JUNE 1987 — FLUORIDE & 8 VOCs
• DECEMBER 1987 ~ FILTRATION
• JUNE 1988 — 40 COMPOUNDS
• JUNE 1989 ~ 34 COMPOUNDS
• JUNE 1991 ~ DISINFECTION & BY-PRODUCTS
-------
BASIS OF MAXIMUM CONTAMINANT
LEVELS (MCLs)
9 MCL GOAL (RISKS)
• MONITORING FEASIBILITY
• TREATMENT FEASIBILITY
• COSTS TO METROPOLITAN SYSTEMS
e NATIONAL IMPACTS
-------
MAJOR CONTAMINANT CATEGORIES
• MiCROBIALS
• DISINFECTION BY-PRODUCTS
• SYNTHETIC ORGANIC CHEMICALS (SOCs)
* INORGANIC CHEMICALS (!OCs)
-------
MICROB1ALS
TREATMENT TECHNIQUES REQUIRED:
• MANDATED BY CONGRESS
• CANNOT MEASURE
• OUTBREAKS (POOR OR NO TREATMENT)
• INCREASED OUTBREAKS
-------
SURFACE WATER TREATMENT RULE --
PERFORMANCE CRITERIA
• 99.9% REMOVAL / INACTSVATION GIARDIA
• 99.99% REMOVAL / KNACTIVATION VIRUSES
-------
TECHNOL OGY IMP A C TS -
SURFACE WATER TREATMENT RULE
INCREASED USE OF :
• MULTIMEDIA FILTERS
o OZONE
• POLYMERS
-------
DISINFECTANTS / BY-PRODUCTS
SET MCLs FOR :
• CHLORINE * HALOACIDS
« CHLORAMINES * HALOKETONES
e CHLORINE DIOXIDE * HALOALDEHYDES
• HALOMETHANES • HALOPHENOLS
-------
DISINFECTANTS / BY-PRODUCTS »
HEALTH CONCERNS
BLOOD
• CENTRAL NERVOUS SYSTEM DEPRESSION
CANCER
-------
CONTROL APPROACHES
REMOVE PRECURSORS
• ALTERNATE DISINFECTANTS
• REMOVE BY-PRODUCTS
-------
IMPACTS
• AFFECTS >200M PEOPLE
• HIGH COST
• BALANCE MICROBIAL SAFETY
• INCREASED USE OF :
• COAGULANTS / POLYMERS
• OZONE
-------
ORGANICS --
HEALTH CONCERNS
• VOCs : LIVER, KIDNEY, CNS, CANCER
e PESTICIDES : CNS, CANCER
• OTHER SOCs : CNS, LIVER, CANCER
-------
BEST AVAILABLE TECHNOLOGIES (BAT)
• GRANULAR ACTIVATED CARBON (GAC)
PACKED TOWER AERATION
-------
IMPACTS
• MORE GROUND WATER TREATMENT
* AERATION & GAG
• ALSO; PAG, OZONE, MEMBRANES,
AND RESINS
-------
KEY INORGANICS (lOCs)
• LEAD
• NITRITE / NITRATE
• RADIONUCLIDES (Rn, U, & Ra)
• OTHERS (As, Se, Ba, F, etc.)i
-------
LEAD --
HEALTH CONCERNS
CNS
BLOOD FORMATION
» BLOOD PRESSURE (?)
-------
NITRITE / NITRA TE --
HEALTH CONCERNS
• ACUTE TOXICITY IN NEWBORNS
• BLUE BABY SYNDROME
• FATAL IF UNTREATED
-------
RADIONUCLIDES -
HEALTH CONCERNS
KNOWN HUMAN CARCINOGENS
• RADON - LUNG
• RADIUM - BONE
-------
BEST AVAILABLE TECHNOLOGIES
• LEAD : CORROSION CONTROL
NITRATE : ION EXCHANGE, OSMOSIS
RADON : AERATION
RADIUM : ION EXCHANGE, OSMOSIS
-------
IMPACTS
LEAD :
• CORROSION CONTROL - UNIVERSAL
INCREASED MONITORING
OTHER IOCS :
• INCREASED TREATMENT
-------
POINT OF ENTRY CRITERIA
• CENTRAL CONTROL
• EFFECTSVE MONITORING
• EFFECTIVE APPLICATION
• MECROBIAL SAFETY
« PROTECT ALL CONSUMERS
-------
SMALL SYSTEM TECHNOLOGIES -- e.g.,
LIMESTONE BED (CORROSION)
TANK AERATION (Rn & VOCs)
SLOW SAND (FILTRATION)
-------
SUMMARY"
REGULATORY IMPACTS
• 83 CONTAMINANTS BY 1989
• DISINFECTION & FILTRATION TREATMENT
• DISINFECTION BY-PRODUCTS BY 1991
• FUTURE CONTAMINANTS : SOCs
-------
SUMMARY --
TECHNOLOGY IMPACTS
® MORE TREATMENT
* MORE USE OF OZONE, RESINS,
MEMBRANES, etc.
• GROUND WATER TREATMENT
• SIMPLE TECHNOLOGIES (SMALL
-------
FACT SHEET 5 {957
Drinking Water Regulations
under
1986 Amendments to SDWA
Significant directives to EPA's standard-setting program for
drinking water contaminants included in the 1986 Amendments
to the SDWA are provided below:
o EPA is to set MCLGs and National Primary Drinking Water
Regulations for 83 specific contaminants and for any other
contaminant in drinking water which may have any adverse
effect upon the health of persons and which is known or
anticipated to occur in public water systems.
o Recommended Maximum Contaminant Levels (RMCLs) are now termed
Maximum Contaminant Level Goals (MCLGs). No changes were made
in the basis of an MCLG; i.e.:
MCLGs are non-enforceable health goals which are to be
set at the level at which no known or anticipated adverse
effects on the health persons occur and which allows an
adequate margin of safety.
o Maximum Contaminant Levels (MCLs) are to be set as close to
MCLGs as is feasible. The definition of "feasible" was
changed to the following:
Feasible means with the use of the best technology,
treatment techniques and other means, which the
Administrator finds, after examination for efficacy
under field conditions and not solely under laboratory
conditions, are generally available (taking costs into
consideration).
Granular Activated Carbon (GAC) is stated in the SDWA as feasible
for the control of synthetic organic chemicals (SOCs), and any
technology or other means found to be best available for control
of SOCs must be at least as effective in controlling SOCs as GAC.
o MCLGs and MCLs are to be proposed at the same time and also
promulgated simultaneously.
o MCLGs and MCLs/Monitoring requirements are to be set for 83
contaminants listed in the SDWA. The best available technology
(BAT) is also to be specified for each.
The 83 contaminants are shown in Table 1. Seven substitutes
are allowed if regulation of any seven other contaminants would
be more protective of public health. The list of substitutes
must be proposed by June 19, 1987.
-------
-2-
o The timetable to produce the MCLGs and MCLs/Monitoring is as
follows:
- 9 by June 19, 1987
- 40 by June 19, 1988
- 34 by June 19, 1989
o MCLGs and MCLs/Monitoring are also to be set for other contaminants
in drinking water that may pose a health risk.
- The 1986 Amendments require the EPA to publish a list
(Drinking Water Priority List) of drinking water contaminants
that may require regulation under the SDWA.
- The list must be published by January 1, 1988, and every 3
years following.
- MCLGs and MCLs/Monitoring are to be set for at least 25
contaminants on the list by January 1, 1991.
- MCLGs and MCLs/Monitoring are to be set for at least 25
contaminants every 3 years following January 1, 1991 (i.e.
1994, 1997,...)-
o Criteria must be established from which States can determine
which surface water systems must install filtration. The criteria
are to be set by December 19, 1987.
o A treatment technique regulation is to be set that will require
all public water systems to use disinfection.
- Variances are available. EPA will specify variance criteria.
- The disinfection treatment rule must be promulgated by
June 19, 1989.
o Requirements are to be set for water systems to monitor for
unregulated contaminants.
- Minimum monitoring frequency would be five years.
- State can add/delete contaminants from list.
- Monitoring regulations are to be promulgated by
December 19, 1987.
o MCLGS and MCLs/Monitoring requirements are to be reviewed by
EPA every three years.
-------
-3-
o Other requirements/provisions of the 1986 Amendments:
- Public notification regulations are to be changed to
provide for different types and frequencies of notice
depending upon the potential health risk. Final regulations
are due September 19, 1987.
- BAT for issuance of variances is to be set when MCLs are
set. BAT may vary depending upon the size of systems and
other factors, including costs.
- Exemptions can be extended for systems with 500 connections
or less. No limit is placed on the number of extensions but
certain criteria will have to be met.
o A summary of deadlines pertinent to standard-setting is presented
in Table 2.
For additional information, contact:
Joseph A. Cotruvo, Director
Craig Vogt, Deputy Director
U.S. Environmental Protection Agency
Criteria and Standards Division
Office of Drinking Water (WH-550D)
401 M Street, S.W.
Washington, D.C. 20460
202/382-7575
-------
-4-
TABLE 1
Contaminants Required to be Regulated
under the SDWA of 1986
Vp1atile Organic Chemicals
Trichloroethylene
Tetrachloroethylene
Carbon tetrachloride
1,1,1-Trichloroethane
1,2,-Di chloroethane
Vinyl chloride
Methylene chloride
Benzene
Chlorobenzene
Di chlorobenzene
Trichlorobenzene
1,1-Dichloroethylene
trans-1,2,Dichloroethylene
cis-1,2,-Dichloroethylene
Microbiology and Turbidity
Total coliforms
Turbidity
G i a r d i a lamb1ia
Arsenic
Barium
Cadmium
Chromi um
Lead
Mercury
Nitrate
Selenium
Silver
Fluoride
Aluminum
Antimony
Endrin
Lindane
Methoxychlor
Toxaphene
2,4,-D
2,4,5-TP
Aldicarb
Chlordane
Dalapon
Diquat
Endothall
Glyphosate
Carbofuran
Alachlor
Epichlorohydrin
Toluene
Adipates
2,3,7,8-TCDD (Dioxin)
Inorganics
Organics
Radionuclides
Viruses
Standard plate count
Legionella
Molybdenum
Asbestos
Sulfate
Copper
Vanadium
Sodium
Nickel
Zinc
Thallium
Beryllium
Cyanide
1,1,2-Trichloroethane
Vydate
Simazine
PAH's
PCB's
Atrazine
Phthalates
Acrylamide
Dibromochloropropane (DBCP)
1,2-dichloropropane
Pentachlorophenol
Pichloraro
Dinoseb
Ethylene dibromide (EDB)
Dibromomethane
Xylene
Hexachlorocyclopentadiene
Radium 226 and 228 G
*e_a,particle and Ph°ton radioactivity088
Radon
alpha particle activity
Uranium
-------
-5-
TABLE 2
Summary of Deadlines
for Standards under SDWA of 1986
When
9 MCLGs and MCLs/Monitoring June 19, 1987
Propose Seven Substitutes June 19, 1987
Public Notice Revisions Sept. 19, 1987
Filtration Criteria Dec. 19, 1987
Monitoring for Unregulated Contaminants Dec. 19, 1987
List of Contaminants Jan. 01, 1988
40 MCLGs and MCLs/Monitoring June 19, 1988
34 MCLGs and MCLs/Monitoring June 19, 1989
Disinfection Treatment June 19, 1989
25 MCLGs and MCLs/Monitoring Jan. 01, 1991
Status; National Primary Drinking Water Regulations
Volatile Organic Chemicals (VOCs)
0 ANPRM March 4, 1982 (47 FR 9350)
0 Proposed MCLGs June 12, 1984 (49 FR 24330)
Final MCLGs, proposed MCLs, Monitoring Nov. 13, 1985
0 November 13, 1985 Federal Register
- Extension of public comment period for 45 days on
tetrachloroethylene MCLG.
- NTP Report recently released
0 Public Briefing: December 19, 1985 Washington, D.C.
0 Public Hearing: January 13-14, 1986 Washington, D.C,
-------
-6-
Final MCL and Monitoring June 1987; Summary
VOCs; Final MCLGs and MCLs (in mg/1)
Trichloroethylene
Carbon Tetrachloride
Vinyl Chloride
1,2-Dichloroethane
Benzene
para-Dichlorobenzene
1,1-Dichloroethylene
1,1,1-Trichloroethane
Final
MCLG*
zero
zero
zero
zero
zero
0.075
0.007
0.2
Final
MCL
0.005
0.005
0.002
0.005
0.005
0.075
0-007
0.2
*Final MCLGs were published Nov. 13, 1985. The MCLG and MCL
for p-dichlorobenzene were reproposed at zero and 0.005 jng/1
on April 17, 1987; comment was requested on levels of 0.075
mg/1 and 0.075 mg/1, respectively.
BAT for 1412 (MCLs);
e Packed tower aeration (PTA) and granular activated carbon
(GAC) for the eight VOCs, except vinyl chloride.
0 PTA for vinyl chloride. ,
BAT for 1415 (Variances);
0 Same as BAT for 1412.
Compliance Monitoring:
0 Initial Monitoring: All systems must monitor each sour
at least once in four years.
- Surface waters: 4 quarterly samples
- Ground waters: 4 quarterly samples; state can exempt
systems from subsequent monitoring if no VOCs detected
in first sample,
- Composite samples of up to five sources allowed
-------
-7-
Phase in by system size (start monitoring:
January 1, 1988).
.Size Complete by - • Date
> 10,000 i year December 31, 1988
3300-10,000 2 years December 31, 1989
< 3300 4 years December 31, 1991
Repeat monitoring: varies from quarterly to once per
five years. The frequency is based on whether VOCs are
detected in the first round of monitoring and whether
system is vulnerable to contamination.
Monitoring for Unregulated VOCs;
0 Initial monitoring: all systems required to sample each
drinking water source once for unregulated VOCs during
a four year period.
* Same phase-in schedules as compliance monitoring.
0 50 VOCs specified:
- List 1: required for all systems (33 VOCs)
- List 2: required for vulnerable systems (2 VOCs)
- List 3: required at Stat° Discretion (15 VOCs)
0 Repeat monitoring: Every five years; EPA will specify
a new list.
Analytical Methods; GC or GC/MS
e Methods 504, 502.1, 503.1, 524.1, 524.2, 502.2
Laboratory Certification Criteria;
0 Seven VOCs: + 20% _>_ 0.004 mg/1
+ 40% < 0.004 mg/1
0 Vinyl Chloride: + 40% < 0.004 mg/1
Non-transient Non-community Water Systems (NTWS);
0 Non-community water systems which regularly serve at
least 25 of the same persons over 6 months per year
(i.e., NTWS) are required to meet all requirements in
this rule.
-------
-8-
Point-of-Entry (POE), Point-of-Use (POU), and Bottled Water;
0 POE may be used to achieve compliance with MCLs but is
not BAT.
0 POU and bottled water cannot be used to meet MCLs.
Variances and Exceptions
e As a condition of issuing a variance or exemption, states
have the authority to require the water system to implement
additional interim control measures. If an unreasonable risk
to health exists, the state must require either installation
of point-of-use devices or distribution of bottled water to
each customer.
IQCs, SOCs, Microbials
0 ANPRM October 5, 1983
0 Proposed MCLGs, November 13, 1985, Federal Register
(50 FR 46936)
0 Repoprosed MCLGs, proposed MCLs/Monitoring scheduled for
September 1987. Final June 1988.
Proposed MCLGs for SOCs
Existing Proposed
SOC NIPDWR (mg/1) MCLG (mg/1)
Acrylamide — zero
Alachlor — zero
Aldicarb, aldicarb — 0.009
sulfoxide and aldicarb
sulfone
Carbofuran — 0.036
Chlordane -- zero
cis-1,2-Dichloroethylene — 0.07
DBCP — 2ero
1,2-Dichloropropane — 0.006
-------
-9-
Proposed MCLGs for SOCs continued
SOC
Existing
NIPDWR (mg/1)
Proposed
MCLG (mg/r)
1, 2-Dichloropropane
o-Dichlorobenzene —
2,4-D 0.1
EDB
Epichlorohydrin —
Ethylbenzene —
Heptachlor —
Heptachlor epoxide —
Lindane 0.004
Methoxychlor 0.1
Monochlorobenzene
PCBs
Pentachlorophenol —
Styrene
Toluene
2,4,5-TP 0.01
Toxaphene 0.005
trans-1 , 2-Dichloroethylene —
Xylene
0.006
0.62
0.07
zero
zero
0-68
zero
zero
0.0002
0.34
0.06
zero
0.22
0.14
2.0
0.052
zero
0-07
0.44
Proposed MCLGs for
Existing
IOC NIPDWR (mg/1)
Arsenic 0.05
Asbestos —
Barium 1«0
Cadmium 0.010
Chromium 0.05
Copper
Lead 0-05
Mercury 0.002
Nitrate 10
Nitrite
Selenium 0.01
MFL = million fibers per liter
IOCS
Proposed
MCLG mg/1
0.050
•7.1 MFL
1.5
0.005
0.12
1.3
0.020
0.003
10
1.0
0.045
-------
-10-
Proposed MCLGs for Microbiological Parameters
Existing
Parameter NIPDWR Proposed MCLG
Total Coliforms 1-4/100 ml Zero
Turbidity 1-5 NTU 0-1 NTU**
Giardia* — Zero
Viruses* — Zero
*Analytical methods do exist but they are not considered
to be technically and economically available for
Giardia or viruses for use in compliance monitoring.
Are included in filtration rule.
**Nephelometric Turbidity Unit. _._
Fluoride
0 ANPRM October 5, 1983
0 Proposed MCLG May 14, 1985
" Final MCLG, Proposed MCL, SMCL, Monitoring Nov. 14, 1985
c Final MCL, SMCL, Monitoring April 2, 1986
Final MCLG 4 mg/1
Final MCL 4 mg/1
Final SMCL 2 mg/1
Final Monitoring 1 per year surface waters
1 per 3 years ground waters
Minimum repeat: 1 per 10
years
Radionuclides
ANPRM September 30, 1986
Draft MCLGs for Radionuclides
Radionuclide Draft MCLG
Radium 226 zero
Radium 228 zero
Uranium zero
-------
-11-
Draft MCLGs for Radionuclides (continued)
Radionuclide Draft MCLG
Uranium
Radon
Gross Alpha particle article
Beta particle and photon radioactivity
zero
zero
zero
zero
Proposed MCLGs/MCLs/Monitoring scheduled for January 1988
Final December 1988.
Other IQCs and SOCs
0 Proposed MCLGs/MCLs/Monitoring scheduled for June 1988. Final
June 1989.
IOCS and SOCs:
Methylene Chloride
Antimony
Endrin
Dalapon
Diquat
Endothall
Glyphosate
Adipates
2,3,7,8-TCDD (Dioxin)
Trichlorobenzene
Standard plate count
Legionella
Sulfate
Nickel
Thallium
Beryllium
Cyanide
1,1,2-Trichloroethane
Vydate
Simazine
PAH's
Atrazine
Phthalates
Pichloram
Dinoseb
Hexachlorocyclopentadiene
Substitutes and Drinking Water Priority List
0 Proposal June 19, 1987. Final January 1, 1988.
° Candidates for removal from SDWA List of 83:
Zinc
- Silver
- Sodium
- Molybdenum
- Vanadium
- Dibromomethane
-------
-12-
In addition, other candidates listed for public comment are
- Sulfate - Phthalates
Candidates for Substitution into List of 83:
- Aldicarb sulfoxide
Aldicarb sulfone
- Ethylbenzene
Heptachlor
- Heptachlor epoxide
- Styrene
- Nitrite
Drinking Water Priority List
Zinc
Silver
Sodium
Aluminum
Molybdenum
Vanadium
Dibromomethane
Chlorine
Hypochlorite ion
Chlorine dioxide
Chlorite
Chi ora mine
Ammonia
Trihalomethanes (chloroform,
dibromochloromethane, bromo-
dichloromethane, bromoform)
Chlorcphenols
Haloni tri1es
Selected disinfection related
chlorinated acids, alcohols,
aldehydes, and ketones
Chloropicrin
2,4-Dinitrotoluene
1,3-Dichloropropane
Bromobenzene
Chioromethane
Bromomethane
1,2,3-Trichloropropane
1,1,1,2-Tetrachloroethane
Chloroethane
2,2-Dichloropropane
o-chlorotoluene
p-chlorotoluene
hexachlorobenzene
hexachloroethane
hexachlorobutadiene
1,1-dichloropropene
2,4,5-T
Isophorone
Ethylene thiourea
Roron
Strontium
Cryptosporidium
Filtration/Col iforms
0 Surface Water Treatment Rule {filtration and Disinfection) to
be proposed August 1987. Draft includes following:
General Requirements and Definitions
Coverage: all public water systems using any surface
water.
-------
-13-
Treatment technique requirements in lieu of MCLs for
Giardia, viruses, heterotrophic plate count bacteria,
Legionella, Turbidity
Treatment definition is at least 99.9 percent removal
or inactivation of Giardia lamblia cysts and at least
99.99 percent removal or inactivation of enteric viruses.
Technical definitions are provided for "surface" water
and appropriate treatment processes.
Criteria under which filtration is required
Source water quality conditions under which disinfection
only is considered safe (fecal coliform concentrations
less than 20 per 100 ml or total coliforms less than
100 per 100 ml in 90 percent of the samples and a
maximum turbidity level of less than 5 NTU).
Site-specific conditions to be met including disinfectant
levels, monitoring, watershed management programs,
sanitary surveys, waterborne disease outbreak history,
total coliform and total trihalomethane MCL compliance
history.
Primacy Agency must determine that all conditions are
met or filtration is required.
0 Disinfection requirements for unfiltered water supplies
All systems must disinfect.
Performance criteria: achieve 99.9 percent inactivation
of Giardia lamblia cysts and 99.99 percent inactivation
of enteric viruses (as appropriate).
Operating criteria: disinfectant concentration and
disinfectant contact times required to obtain the
inactivation rates specified in the performance criteria
are published in the rule.
Design criteria: redundancy of components, and
auxiliary power supply.
Monitoring requirements: continuously monitor
disinfectant concentration at the plant and demonstrate a
disinfectant residual concentration in 95 percent of the
distribution system samples.
* Disinfection requirements for filtered water supplies
Disinfection with filtration must achieve at least
99.9 percent and 99.99 percent removal/inactivation of
-------
-14-
Giardia and viruses, respectively. Primacy agency defines
level of disinfection required, depending on technology and
source water quality.
Monitoring requirements: continuously monitor
disinfectant concentration at the plant and demonstrate
a disinfectant residual concentration in 95 percent of
the distribution system samples.
Filtration requirements
- Filtration must be installed when the system fails to meet
source water quality or site-specific conditions.
- Performance requirements in terms of turbidity levels are
set for commonly used filtration technologies (e.g./
conventional, direct, slow sand and diatomaceous earth
filtration technologies).
Other technologies (including emerging and innovative
technologies) can be approved by the Primacy Agency if they
achieve the performance criteria for Giardia lamblia and
virus removal/inactivation.
Monitoring and Analytical Requirements
EPA approved methods for total coliforms, fecal coliform,
turbidity, disinfectant residual, temperature, and pH
are proposed.
Specified frequency of monitoring for each parameter and
numbers of analyses required, varies according to population
C O T- \T Ct rl .
Specified
numbers
served.
Reporting, Public Notice and Recordkeeping
Requirements include monthly reports to the Primacy Agency
on coliform concentrations, turbidity levels, disinfectant
concentration, pH, water temperature, disinfectant contact
time and disinfectant residuals.
Unfiltered water systems must also report annually on
their watershed control program and sanitary surveys.
Water-borne outbreaks must be reported within 48 hours.
c Violations
- All systems with unfiltered surface water sources must
meet source water quality and site-specific conditions
within 48 months of promulgation. If they fail to
meet these criteria within 30 months, filtration would
be required, but they would not be in violation until
failing to meet such criteria after 48 months.
-------
-15-
Monitoring requirements for unfiltered, raw water
quality must be met within 30 months of promulgation.
Performance criteria and monitoring/reporting requirements
for the filtered and disinfection treatment techniques
must be met within 48 months of promulgation.
Total Coliform MCL
* Maximum contaminant level goal—zero
e Maximum contaminant levels
Based on presence or absence of total coliforms in sample,
rather than density.
- Monthly MCL
No more than 1 coliform-positive sample/month for
systems which analyze fewer than 40 samples/month.
No more than 5% of samples can be coliform-
positive if system analyzes at least 40 samples/
month.
Long-term MCL
No more than 5% of most recent 60 samples coliform-
positive if system analyzes fewer than 60 samples/
year.
No more than 5% of all samples in most recent
12-month period can be coliform-positive if system
analyzes at least 60 samples/year.
0 Monitoring frequency
For systems serving 3,300 persons or fewer:
5 samples/month, with less monitoring for systems
which:
(a) filter and disinfect surface water and
disinfect ground water; and
(b) have a sanitary survey at the frequency
specified in the proposed regulation.
For systems serving more than 3,300 persons:
Based on population served
Similar to current minimum monitoring requirements
for coliforms, but with smaller number of population
categories.
-------
-16-
Repeat samples
If system has coll form-positive sample, system must
collect five repeat samples all on the same day from
same location as original sample, except some may be
from the next service connection.
If any repeat sample is coliform-positive, system
must:
— Analyze positive culture medium to determine if it
contains fecal coliforms; and
— Collect and analyze another set of five repeat
samples all on the same day, unless the MCL has
been violated and the system has notified the
State-
Heterotrophic bacteria (HPC)
No MCLG/MCL is proposed.
Proposed regulation based on HPC interference with
total coliform analysis and considers effectiveness
of disinfection and filtration controlling HPC levels.
If coliform sample produces a turbid culture in the
absence of gas production, using the Multiple-Tube
Fermentation Technique or produces confluent growth or
a colony number that is "too numerous to count" using
the membrane filter technique, the system may either
accept the sample as coliform-positive or declare the
sample invalid and collect and analyze another water
sample. Second sample is analyzed for both total
coliforms and HPC. If HPC is greater than 500
colonies/ml, then sample is considered coliform-
positive, even if total coliform analysis is negative.
Variances and exemptions—none allowed
Waterborne disease outbreaks—States must search for and
investigate possible occurrences of such outbreaks.
-------
State Implementation of the SDWA
Frederick A. Marrocco, Vice President, Association of State
Drinking Water Administrators and Chief, Division of Water
Supplies, Pennsylvania Department of Environmental Resources,
Harrisburg
Handout materials not available.
II-l
-------
Current and Emerging Treatment Technologies
Robert M. Clark, Director, Drinking Water Research Division, US
Environmental Protection Agency, Cincinnati, OH
III-l
-------
Treatment Techniques (Robert M. Clark, Director, Drinking Water
Research Division, US Environmental Protection Agency, Cincin-
nati, OH) .
-------
The Safe Drinking Water Act (SDWA) of 1974 requires that EPA establish
maximum contaminant level goals (MCLG) for each contaminant in drinking water
which may have an adverse effect on the health of persons. Each Goal is
required to be set at a level at which no known or anticipated adverse
effects on health occur, allowing an adequate margin of safety. The SDWA
also requires that National Primary Drinking Water Regulations (NPBWRs)
establishing maximum contaminant levels (MCLs) or treatment techniques, and
secondary drinking water regulations be established. The purpose of this
paper is to describe the activities of the Drinking Water Research Division
(DWRD) of EPA in evaluating the various technologies that may be available
for achieving these MCLs.
In addition to conducting research into technological alternatives, the
DWRD provides technical advice and assistance to State agencies, EPA regional
offices, water utilities and professional organizations. Amendments to the
SDWA have increased regulatory pressure and accelerated the timetable for
establishing MCLs- DWRD's activities and its role in satisfying the amend-
ments will be discussed in terms of the following SDWA admendment categories:
contaminants to be regulated; surface water filtration; mandatory disinfec-
tion; and prohibition of use of lead materials. In addition DWRD's work
which is intended to study and characterize the factors that lead to deterio-
ration of water quality in distribution systems will be discussed.
aDirector, Drinking Water Research Division, WERL, U.S.EPA, 26 W. St. Ciair St
Cincinnati, Ohio 45268 *'
-------
REGULATED CONTAMINANTS
DWRD is conducting research in three major areas associated with regu-
lated contaminants. These areas are: disinfection by-products; organic
contaminants; and, inorganic contaminants.
DISINFECTION BY-PRODUCTS
Disinfection by-products research is perhaps the most significant and
controversial area of research ever undertaken by EPA in the drinking water
field. The major and most important disinfection by-products are total
trihalomethanes (TTHM). In addition over 500 additional disinfection by-
products have been identified. An objective of this research is to establish
the feasibility of lowering the TTHM standard which of course will have many
ramifications for the water supply industry. In addition, a limited number
of additional disinfection by-products are being quantified.
Two major pathways of research are currently being pursued. One is the
identification and characterization of the by-products of chlorination disin-
fection and the second is the evaluation of various treatment techniques to
control byproducts using results from in-house pilot plant studies. By-
products of all other types of disinfection are being studied as well.
Because of the complexity of the research effort a disinfection by-product
workgroup has been formed to bring various groups within ORD together with
the Office of Drinking Water to maintain communication in this research area.
-------
Identification of By-Products
DWRD has been actively developing methods for identifying the various
by-products associated with chlorine disinfection. Most of this work has
been conducted at the bench scale. The Division has also been working closely
with the Technical Services Division of ODW to isolate the various by-products
identified in field studies.
The major disinfection by-products found to-date include the following:
trihaolmethanes (4 compounds), dihaloacetonitrile (3 compounds), chloroacetic
acid (3 compounds), chlorohydrate, chloropicrin, 1,1,1-trichloropropenone,
Table 1 lists these compounds and the general ranges within which they have
been found. From 30-60% of TOX in drinking water is associated with these
disinfection by-products.
TABLE 1- COMMON DISINFECTION BY-PRODUCTS
General Compound
Identification
Number of
Individual Compounds
Nominal Concentration
< 10
in ug/L
10-100
> 100
Trihalomethanes
Dihaloacetonitriles
Chloracetic Acids
Chloral Hydrate
Chloropicrin
1,1,1-Trichloropropanone
4
3
3
X
X
X
X
-------
Control of Disinfection By-Products
Figure 1 is a schematic of DWRD's in-house pilot plant. Two parallel
channels are being evaluated. The channel at the top of the figure is a
control in which the pilot-plant is being operated as a conventional filtra-
tion plant with chlorine added at the head works. In the bottom channel the
point-of-chlorination has been moved to minimize disinfection by-products,
primarily total trihalomethanes, and to simultaneously maintain control of
microbiological quality throughout the plant. The influent to both channels
is spiked with coliform, phage and raw sewage. A side stream from the lower
channel is passed through granular activated carbon (GAC).
Based on work conducted to-date, microbiological quality has been main-
tained and TTHMs minimized when chlorination was moved to a point just before
the filter. Assimable Organic Carbon (AOC) is also being measured throughout
the plant in order to characterize the biological quality of the water that
might enter the distribution system. Based on preliminary work AOC is also
reduced when chlorination is moved to a point just before filtration.
Recent results indicate that with a properly optimized treatment system
chloramines can minimize TTHM formation and also maintain microbiological
control. This area of investigation is continuing
ORGANIC CONTAMINANTS
Despite the complexity of setting MCLs for health threatening agents,
Congress has been placing great pressure on EPA to regulate organic contam-
inants more extensively. A number of drinking water contaminants have yet to
be regulated. At this time only a few revisions of existing regulations have
been made although the SDWA requires a review of regulations at least every
three years.
-------
(PHAGE)
(SEWAGE)
i
STORAGE
FUOC
SETHE
t
CHLOWE
ALUM
ALUM
I
M!X
FUDC
SETTLE
FUER
CLEAR
WELL
(CHLORIC
FUER
CLEAR
WELL
GAC
FOER
f
CHLORIC
CLEAR
WELL
FIGURE 1. PiLOT PLANT SCHEMATIC FOR DISINFECTION BY-PRODUCT CONTROL
-------
Traditionally EPA's technology program has utilized the approach of
carefully evaluating process kinetics on the bench, scaling up for engineer-
ing feasibility at the pilot-scale, and making field-scale evaluations to
evaluate process economics. This approach may take several years for comple-
tion and involve several million dollars for each unit process examined.
Regulatory pressure will no longer allow for a contaminant-by-contaminant and
process-by-process evaluation.
As mentioned previously, recent amendments to the SDWA require a large
number of Synthetic Organic Chemicals (SOCs) to be considered for regulation
under stringent time deadlines.^ A key provision for potential regulation
under the SDWA is that a feasible technology be available to remove each
contaminant regulated. For SOCs, granular activated carbon (GAG) by law is
feasible technology. Table 2 summarizes the synthetic organic compounds for
which DWRD is currently developing carbon usage rate data. In order to
calculate these data several approaches are being taken. One approach is to
develop usage rate data based on microcolumn data and the other is to calcu-
late usage rates based on field scale studies. These approaches will be
described in the following sections.
TABLE 2. CARBON USAGE RATES FOR SELECTED ORGANICS
2,4-D
Silvex
Llndane
Methoxychlor
Toxaphene
Chlordane
Heptachlor
Heptachlor Epoxide
PCS(s)
Pentachlororophenol
Alachlor
Carbofuran
Aldicarb
Aldicarb Sulfone
Aldicarb Sulfoxide
Endrin
Phthalate(s)
Adipate(s)
PAH(s)
Hexachlorocyclo pent ad i e ne
Dioxin
Simazine
Atrazine
Diquat
Endothall
Picloram
Dalapon
Dinoseb
Glyphosate
Oxamyl
EDB
DBCP
1,2-Dichloropropane
Cis-1,2-Dichloroethylene
Trans-1,2-dichloroethylene
Chlorobenzene
o-Dichlorobenzene
Toluene
Styrene
Ethyl Benzene
o-Xylene
m—Xylene
p-Xylene
Dibromomethane
1,1,2-Trichloroethane
-------
Microcolumn Data
Microcolumns are being used to study _GAC adsorption. The adsorption
protocol involves developing a set of single solute isotherms in both dis-
tilled and natural waters. Using the Freundlich isotherm parameters, the
homogeneous surface diffusion adsorption model (HSDM) is employed to generate
breakthrough curves and predict carbon usage rates for fixed-bed GAC column
operation both at microcolum and at full-scale. Other inputs to the HSDM are
kinetic parameters estimated from correlations given in the literature and
parameters defining the carbon type and the system hydraulics.
For some SOCs, predicted full-scale usage rates are found to be cost
effective, i.e., typically beyond a two-year service life, and these full-scale
usage rates are used as inputs to cost models. A typical HSDM-generated full-
scale breakthrough curve is given in Figure 2.
For poorer adsorbed SOCs, full-scale predications show shorter service
lives, and microcolumn operations are conducted. Because microcolumn break-
through curves represent empirical kinetic phenomenon rather than correlated
estimates, the microcolumn breakthrough curves are scaled-up to full-scale
conditions. These scaled-up usage rates are used as inputs to cost models.
The microcolumns were designed to give constant-pattern breakthrough
curves based on the HSDM. The apparatus consists of a 4 mm Inside Diameter
(ID) glass column packed with US Standard Seive 100 x 200 mesh Filtraaorb 400
GAC at a depth sufficient to contain the mass transfer zone for the compound.
Other requirements are a prefilter to prevent excessive head loss across the
column, dampers to smooth out pump delivery and maintain constant pressure on
the column, and a flexible head space bag to restrict loss of volatile contam-
inants. The apparatus is described in Figure 3. For SOCs having full-scale
service lives of less than two years, microcolumn breakthrough curves may be
-------
Predicted Effluent Concentration Profile for cis-DCE
EBCT=15 mln 4 gpm/sqft Loading
200 n
180^
5> 160-
C 140-1
O
120 H
£ 100-
80-^
60-^
40-
—i
S 20-j
O
O
O
0)
' , , , . < | I
100 105
• ' •
nr i i "1 i i i
T—T
110 115 120 125 130 135
Throughput Time (days)
140 145
150
FIGURE 2. TYPICAL BREAKTHROUGH CURVE FOR FULL SCALE SYSTEM
-------
PRESSURE GAGE
PULSE
DAMPERS
»
\ NFUJENT
GLASS WOOL
PREFUER
ADSORPTON
MICROOOLJJMN
FIGURE 3. ADSORPTION MICROCOLUMN SCHEMATIC
-------
generated in less than 5 days. This is compatible with the short regulatory
time frame. Figure A shows a typical microcolumn breakthrough curve.
Once the full-scale breakthrough curve's are developed for each compound,
field-verified cost equations are used to generate full-scale costs.2 A com-
puter program has been developed that uses breakthrough curve data to calculate
carbon usage rates, and ultimately calculate costs for a wide number of carbon
systems. A set of costs generated from this procedure at various sizes of
application has been developed. For the data prepared for this analysis, it
is assumed that the carbon systems were consistent with the following condi-
tions: EBCT - 15 minutes; loading rate - A gpm/ft^; carbon type - Filtersorb
AOO. These were full-scale conditions that were used in the HSDM.
Several treatment plant sizes were assumed with emphasis on small systems.
Steel pressure contactors were assumed for GAG adsorption in systems with a
design capacity less than or equal to 1 mgd. Concrete gravity contactors
were assumed in larger systems. All systems were assumed to operate at 50%
plant capacity. Spent carbon replacement versus reactivation alternatives
were selected based on least cost. Cost parameters used in the analysis are
presented in Table 3.
*
TABLE 3. COST PARAMETERS USED IN ANALYSIS
Item Value
GAG Price $0.70/lb @ AO.OOO Ibs
Capital Recovery Interest Rate 10%
Amortization Period 20 years
ENR Construction Cost Index A229
Producers Price Index 29A
Labor Rate $ll/oanhour
Electric Rate $0.07/kwh
Fuel Oil Rate $0.90/gallon
Process Water Rate $0.50/1000 gallons
GAG Bulk Density 30 lbs/ftj
GAG Loss Rate from Reactivation HZ
-------
100X200
bed
-------
An example of the SOCs examined, their influent and effluent concentra-
tions, and type of water studied in each case are given in Table 4. Most of
the scenarios are worst-case situations which provide conservative cost esti-
mates for removal of the single-solute SOCs. Table 4 presents a matrix of
GAG system costs for various usage rates and plant capacities. All estimates
are determined using 1986-year prices and cost indices. Using the data in
Table 4, two sets of cost-curves were generated: (1) a set of graphs display-
ing cost as a function of system capacity at a fixed usage rate (designated
by specific SOC removal), (2) a set of graphs displaying cost as a function
of usage rate or GAG bed life at a fixed system capacity. These cost-curves
can be very useful in analyzing cost sensitivity with variations in system
size, use rate and bed life.
In general, the most significant cost impacts appear to be with systems
that have a capacity of less than 5 mgd and/or a GAG bed life shorter than 4
months (>0.4 lbs/1000 gal). At system capacities greater than 40 mgd the
cost curve has a fairly flat response showing small changes in cost with
variations in system size. Likewise, when the GAG bed life is longer than 1
year, the change in cost with variation in usage-rate is small. For systems
larger than 10 mgd, the fairly flat response of the cost curve extends down
to a GAG bed life of 6 months (<0.23 lbs/1000 gal). It appears that for
strongly adsorbed compounds the GAG bed life could be reduced by as much as
100% and the cost would only increase by approximately 10 to 20%. An interest-
ing example is as follows: if the GAG usage-rate increases by a factor of 10
(1000Z increase) from 0.0572 lbs/1000 gal (2-year bed life) to 0.6 lbs/1000
gal (69-day bed life), the cost increases by only a factor of 2 (100% cost
increase)•
-------
Table 4. GAG COSTS FOR VARIOUS USE RATES AND SYSTEM SIZES
Use Rate, lbs/1000 gal .0572 .0626 .0734 .0832 .1176 .178 .2191 .
Service Life, days 730 667 569 502 355 234 191
A B CD E F G
2336
179
H
Design Capacity (MGD) Total System Cost (^/lOOO
A
B
C
D
E
P
G
H
I
J
K
L
M
.25 83.8 84.6 86.2 87.6 92.6 101.1 106.9
.75 79.1 79.8 81.3 82.7 87.4 95.6 101.1
1 67.9 68.6 70.1 71.5 76.1 84.2 89.7
2 40.2 40.9 42.3 43.7 48.2 56.1 61.4
3 35.2 35.9 37.3 38.6 43.1 50.8 56.0
5 30.3 31.0 32.4 33.6 38.0 45.6 48.5
7 27.7 28.4 29.8 31.0 35.4 39.4 40.9
10 25.4 26.1 27.4 28.7 31.1 33.4 34.9
20 21.1 21.3 21.7 22.0 23.3 25.4 26.9
30 18.0 18.2 18.6 19.0 20.2 22.3 23.7
50 15.2 15.4 15.7 16.1 17.3 19.4 20.8
70 13.7 13.9 14.3 14.6 15.8 17.8 18.9
100 12.4 12.6 13.0 13.3 14.5 16.1 18.9
« two-year service life
- chlorobenzene, Cl - 600 ug/L, Ce = 6 ug/L, Mllli-Q distilled water
- chlorobenzene, Ci - 600 ug/L, Ce - 6 ug/L, filtered Ohio River water
- aldicarb, Ci - 500 ug/L, Ce - 1.3 ug/L, filtered Ohio River water
- EDB, Ci - 100 ug/L, Ce - 1, Milli-Q distilled water
- 1,2-dichloropropane, Ci - 100 ug/L, Ce - 2 ug/L Milli-Q distilled water
- trans-DCE, Ci - 100 ug/L, Ce - 70 ug/L filtered Ohio River water
108.9
103.0
91.6
63.2
57.9
49.1
41.5
35.4
27.4
24.2
21.2
19.2
19.4
.2713
154
I
gal)
114.
108.
96.
68.
62.
50.
42.
36.
28.
25.
22.
20.
20.
1
0
5
0
6
5
9
7
7
5
2
2
2
.378
110
J
128.
122.
110.
81.
71.
54.
46.
40.
32.
29.
25.
25.
23.
7
1
4
5
1
4
7
5
4
2
0
0
1
.45
93
K
138.
132.
119.
90.
73.
57.
49.
43.
34.
31.
30.
27.
24.
5
9
6
5
7
0
3
1
9
1
5
5
9
.6046
69
L
159.
151.
139.
109.
79.
62.
54.
48.
39.
35.
34.
3V
30.
3
3
3
7
4
5
7
5
9
2
6
5
8
1.2
35
M
237.8
228.0
213.7
121.0
100.8
83.4
75.4
68.8
55.4
56.8
53.6
51.9
49.5
- 1,2-dichloropropane, Ci - 100 ug/L, Ce - 2 ug/L filtered Ohio River water
« toluene, Ci « 5000 ug/L, Ce *> 100 ug/L, distilled Wausau water
- cis-DCE, Ci - 200 ug/L, Ce - 70 ug/L, filtered Ohio River water
« 93 day service life
» trans-DCE, Ci - 500 ug/L, Ce - 70 ug/L, filtered Ohio River water
- 35 day service life
-------
This cost analysis suggests that highly accurate predictions of GAG
adsorption performance for many SOCs may not be necessary because of the
apparent insensitivity of cost with mild changes in usage rate. From the
cost curves this appears to be valid for situations where the GAG bed life
is longer than 6 months. It was found that capacities and thus costs are
most sensitive to the parameters in the Freundlich isotherm equation. The
costs presented herein were generated using a computer program written for
the IBM-PC/AT.
Other Technologies Studies
Other technologies being examined at the bench and pilot scale for control
of organics include ozone oxidation, reverse osmosis, ultraviolet treatment, and
ultrafiltration.
Ozone Oxidation
Ozone oxidation is being studied extensively in DWRD's in-house pilot
plant facilities.^ Controlled pilot plant ozone treatment tests have been
conducted on 29 VOCs in distilled water and groundwater. Results showed that
aromatic compounds, alkenes and certain pesticides are well removed by ozone
treatment but that alkanes are poorly removed. Also, removal efficiency
improved for the alkenes and aromatic compounds with increasing ozone dosage
and for some alkanes with increasing pH. For most compounds, the efficacy of
ozone was not affected by the background water matrix. Information from the
literature concerning the ozone treatment of pure materials in the gaseous or
liquid phase generally predicted the effectiveness of ozone in treating
aqueous solutions.
-------
Reverse Osmosis
Revesrse osmosis (RO) has shown some promise in removing both VOCs and
SOCs from groundwater.^ Most of the efforts by DWRD to date have been on a
pilot- or bench-scale basis, with some limited application of reverse osmosis
for the removal of organics at one field site. Primary indications are that
certain RO membranes are very effective in removing a wide range of organic
chemicals.
Ultraviolet Treatment
Ultraviolet light also shows some promise for removing organic contam-
inants, particularly when combined with ozone. DWRD is in the process of
funding a project with the Los Angeles Water and Power Company that deals
with the removal of VOCs from groundwater using these two technologies. If
successful, these chemicals will be oxidized to C0£ and water, and the need
to deal with off-gas control problems will be eliminated.
Ultrafiltration
Studies have been conducted to evaluate the costs and performance of low
pressure membrane processes (ultrafiltration) for TTHM precursor reduction in
small systems. A 150 day pilot-plant of two highly organic contaminanted
ground waters (both of which produced more than 400 ug/L THMs when conven-
tionally treated) produced a finished water that easily met the Trihalomethane
Maximum Contaminant Level (100 ug/L). Costs appears to be reasonable.^
More extensive testing of this technology is underway.
Field-Tested Technologies
Carbon Adsorption
Extensive field scale studies have been conducted on the use of granular
activated carbon (GAG) treatment with onsite regeneration. Past research has
-------
devoted to demonstrating the effectiveness of GAG for surface water treatment.5
Extensive field testing has been conducted -at Jefferson Parish and Cincinnati,
Ohio. Other field sites include: Manchester, New Hampshire; Evansville,
Indiana; Miami, Florida; Huntington, West Virginia; Beaver Falls, Pennsylvanis;
and Passaic, New Jersey.
More recently, DWRD has been conducting studies that incorporate the use
of carbon treatment for removal of VOCs and SOCs from groundwater at Suffolk
County, New York, California's San Joaquin Valley, and Wausau, Wisconsin.
Each of these projects is designed to examine a different aspect of GAG
application and except for Wausau, are intended to deal with the little-under-
stood pesticide contamination problem.
A typical project is in Suffolk County, New York, where the removal of
organics, pesticides, and nitrates is being studied under various flow situa-
tions. Two parallel treatment systems (one consisting of GAG and ion exchange
and the other consisting of reverse osmosis) are being operated at low flows
similar to home usage. Costs for these two systems will be established along
with unit operating efficiency so that a large public water supply system can
be designed and tested. The results from this study will be applicable to
other areas, especially in farming communities, where multiple contamination
of groundwater is identified.
Two principle objectives of a project being conducted in the San Joaquin
Valley study are to (1) develop cost-effective design criteria for the removal
of DBCP and other pesticides from water supplies by GAG, (2) to compare treat-
ment methods other than GAG for the removal of these compounds, and (3) to
improve and strengthen existing administrative guidelines and jurisdictional
responsibilities pertaining to both community water systems and private wells
-------
containing the compounds of interest. Pilot studies will be conducted with
GAC, and results will be compared with those existing operating systems.
These comparisons should lead to recommendations for more effective design
and operational criteria for GAC units used by water systems. The focus in
this study will be on point-of-use GAC units that are now in use by small
systems, individual homeowners, and farmers in the area. Disposal of treat-
ment residues is also being studied.
The third site, in Wausau, Wisconsin, has multiple contaminants in its
groundwater source from a nearby superfund site. Figure 5 shows a typical
breakthrough curve from the Wausau carbon field study.
GAC is the primary technology being studied, but air-stripping is being
examined as a companion technology. The Wausau project is unique in that
modeling techniques are being used to predict full-scale design criteria for
a GAC plant. These predictions will be evaluated against the actual cost and
performance associated with building of GAC contactors onsite. the results
from this study are expected to provide a useful methodology for extension to
other GAC applications. Modeling techniques will be particularly useful in
studying the long list of SOCs and VOCs that may be proposed for regulation
under the current and future provisions of the Safe Drinking Water Act. In
addition to modeling, rainicolumn technology is also being studied. This
methodology allows investigators to acquire water from a given site and to
study the performance of a small, high-pressure column in the presence of a
natural water background. Results from these minicolumn experiments are
expected to reproduce the breakthrough curves normally seen in pilot- and
full-scale facilities.
-------
3,0
CM-
T
c
o
• *» Influent
D» Effluent; EBCT - .97 mln
0«= Effluent; EBCT •= 2.93 mln
A » Effluent; EBCT « 4.96 mln
o» Effluent; EBCT = 9.98 mln
x m Effluent; EBCT » 20.3 mln
• « Effluent; EBCT - 30.9 mln
CIa~1,2-Dlchloroethene
WVQ (12x40)
Loading Rate « 4.05 m/hr
Bulk Density « 417 kg/m3
Ave. Influent Temp. » 13 °C
Detection Limit » .3 ug/L
30
90 120 150 180 210 240 270 300 330 360
Elapsed Time of Operation (days)
FIGURE 5. PILOT SCALE DATA FROM WAUSAU, WISCONSIN
-------
Packed Tower Aeration
Aeration technology has proved to be especially effective for the removal
of VOCs. However, the research at Wausau, Wisconsin (which also incorporates
air stripping and off-gas control technology) indicates that aeration may in
some cases be effective for removing compounds that have somewhat lower
Henry's Law Constants than would normally be expected to be removed by this
process. In this project, as with carbon facilities, modeling is being used
to predict the performance of packed-tower aeration and results are promising.
The off-gas control portion of the air stripping project at Wausau is being
conducted in cooperation with the American Water Works Association Research
Foundation.
*
DWRD also has another air-stripping research project currently underway
at Baldwin Park, Calfifornia which is examinaing the removal of VOCs from a
groundwater supply. As in Wausau, the Baldwin Park project is exmaining the
problems of off-gas control technology.^
Another field-scale activity dealing with air-stripping has been con-
cluded at Brewster, New York. In this project, modeling techniques and
pilot-scale facilities were used to determine the scale-up relationships to
be used for full-scale, air-stripping facilities. The purpose of the project
was to develop a technique that could be used by consulting engineers to
adequately predict the cost and performance of full-scale facilities using
pilot aeration columns.
Conventional Treatment
Conventional treatment is unlikely to be used for removing organic
contaminants from groundwater. However, field studies are being conducted
at Tiffin, Ohio, where the river source contains high spikes of pesticides
-------
because of local seasonal agricultrual applications. Powdered activated
carbon, added to the water normally treated only by conventional treatment,
appears to be quite effective for removing synthetic organic chemicals.
Secondary Discharge Problems
When evaluating treatment technology one has to be aware of the possi-
bilities of creating secondary discharge problems. For example, while
conducting GAC research on surface water supplies, it was discovered that
dioxins were formed in the reactivation process. An extensive evaluation
led to the installation of an afterburner, which was found to eliminate
those dioxin byproducts when operating at a temperature of 2400CF.'
At Baldwin Park, California, the use of high-stack dispersion was inves-
tigated as a means of minimizing the impact of VOC removal from groundater.
At Baldwin Park and at Wausau, Wisconsin, gas-phase carbon adsorption is
being investigated for removal of both VOCs and SOC from air stripping waste
gases.
The project with the Los Angeles Water and Power Company is intended to
investigate the water-phase oxidation of VOCs and SOCs to C(>2 and ^0 using
ozone in combination with ultraviolet light. Gas-phase oxidation of VOCs and
SOCs by ozone and ultraviolet light also appears promising.
Future work will concentrate on residuals control from various unit
processes such as reverse osmosis and ultrafiltration.
Inorganic Contaminants
Several field scale studies are being conducted for inorganics control.
These technologies include; ion exchange and reverse osmosis. Specific
-------
contaminants being removed Include nitrates, radium and radon. Each of these
technologies is discussed in the following.paragraph.
Ion Exchange
Ion exchange technology has been field-tested for nitrate removal under
a DWRD-supported cooperative agreement and has been pilot-tested for uranium
removal at sites in the western part of the United States.®»^
The nitrate removal plant has been operating automatically for about 3
years. The 1-mgd (0.38 ML/d) demonstration plant is located in McFarland, CA
and consists of three anion exchange vessels that are designed to reduce
nitrate levels to below 10 mg N03-N/L, the EPA MCL and also the California
requirement. Currently, about 500 gpm (32 L/S) of water is being treated and
i
about 200 gpm (13 L/S) is bypassed and later blended with the treated water,
resulting in a total product flow of 700 gpm (44 L/S). The blended water
adequately meets the nitrate MCL and EPA's Secondary Regulation (Fed. Reg.
1979) for chloride and sulfate levels.10
Bench-scale studies were initally conducted at EPA-Cincinnati for uranium
removal and later pilot-tested. The laboratory work showed that when drinking
water containing 300 ug/L uranium was passed through anion exchange resin,
more than 9,000 bed volumes were treated before breakthrough was observed.
DWRD then evaluated the performance of twelve 1/4-cubic-foot (0.007 m^) anion-
exchange systems installed in New Mexico, Colorado, and Arizona at sites
where uranium levels in the raw water exceeded 20 ug/L. Results confirmed
the findings shown in the laboratory. Because of the high loading capacity
of the anion resins for uranium, these units are well suited for point-of-use
applications where on-site regeneration is not feasible. For centralized
treatment, the resin may be regnerated and recycled by backwashing it with
sodium chloride solution.
-------
Reverse Osmosis
Groundwater sources in Illinois, lowa^ Florida, Texas, Wisconsin, and
some Rocky Mountain states contain radium in excess of the 5 pCi/L MCL. RO
treatment of radium-laden ground water was demonstrated in 1977 when DWRD and
Sarasota County, FL water supply staff undertook a cooperative effort to
study the operation of eight RO systems.^ The systems were located in small
communities serving a population from 39 to 15,000 and the design capacities
varied from 800 gpd (3 kL/d) to 1 mgd (3.8 ML/d). Six different manufacturers
of RO systems were represented and both the hollow fiber and spiral wound
cellulose acetate membranes were used. The study showed from 82-961 Ra-226
removal for all systems, resulting in treated water that contained below the
EPA MCL of 5 pCi/L.
RO technology has also been pilot tested for nitrate removal as part
of DWRD cooperative agreements at three locations.^, 13,14 ^8 mentioned
previously, high nitrate levels exist in the well waters of Suffolk County,
LI, NY, along with several SOCs.12 This combination of organic and inorganic
contaminants is the reason that RO was selected as one of the treatment
technologies to be studied there. Seven commercially available membranes
were evaluated for their rejection capabilities. Nitrate removals ranged
from 75-95%, and research is continuing with one of the polyaraide membranes
that proved relatively efficient for both nitrate and SOC rejection. At
Charlotte Harbor, FL both high pressure (265-359 psig) and low pressure
(163-187 psig) RO systems were studied for the removal of several spiked
inorganic contaminants, including nitrate, from a natural groundwater.^ The
investigation showed that the high pressure system was significantly more
effective for removing all substances measured. The comparison for nitrate
-------
removals, for example, was 80% vs. 6-24%. Before ion exchange was selected
for full-scale evaluation at McFarland, CA^ a 20,000 gpd (77 m^/d) RO system
was examined for nitrate removal.^ Even though the system experienced
frequent electrical and mechanical failures, nitrate rejection of about 65%
was achieved.
Radium Removal Technologies
DWRD has sponsored research to study promising methods for radium
removal. 15,16 -j^e jon exchange process, with both weak acid resin and strong
acid resin were investigated. Both resins effectively removed radium from
water to well below the 5 pCi/L MCL and, in most cases, to <0.5 pCi/L, respre-
senting over 96% removal. The weak acid resin in the hydrogen form also
removed hardness, which was not the case for the strong acid resin in the
calcium form. The maximum capacity of the weak-acid resin was about 2.3
times that of strong acid resin and much less spent regenerate per unit
volume of water treated was produced from the weak-acid column than from the
strong-acid column. Another part of this project was to determine the feasi-
bility of Dow Chemical Company's Radium-Selective Complexer (RSC) for remov-
ing radium from brines compared with typical groundwaters. The RSC is a
synthetic resin that has a high affinity for radium. The capacity of RSC was
observed to have been about 200 times greater in 450 mg/L TDS water than in
40,000 mg/L TDS brine (51,000 pCi/dry g vs. 300 pCi/dry g). However, the
effect of other parameters, including calcium, sodium, and other ions, in
addition to EBCT, needs to be determined. The application of the RSC for treat-
ment of radium from brines is presently being studied at a small community in
Colorado under a research cooperative agreement with DWRD.
-------
Radon Removal Technologies
In response to the Agency's recent emphasis on radon in the environment,
DWRD has funded a project to evaluate several treatment techniques for the
removal of radon from community water supplies. Three treatment methods,
packed tower aeration, diffused bubble aeration and granular activated carbon
(GAC) will be evaluated for the removal of radon from two community water
supplies in New Hampshire. The study will compare the methods for effective-
ness, costs, operation, maintenance and other related factors. To accomplish
the objectives, the three treatment systems will be constructed and operated
at each of the two sites. The two sites selected are trailer parks, one
whose water supply serves 40 homes with an average daily flow of 4.6 gpm and
an average radon concentration of 155,000 pCi/L. The other location has 56
homes, the average daily water usage is 6.3 gpm, and the average radon con-
centration is 40,000 pCi/L. After construction of the treatment systems,
they will be operated for 6-12 months. At the end of this project, a report
will be developed as a guideline to help State Agencies and small communities
select an appropriate technology for radon gas treatment in public water
supplies.
Summary of Technologies Studied
Table 5 summarizes the treatment technoliges that DWRD is evaluating
for removal of VOCs, synthetic organic chemicls (SOCs), nitrates, and radio-
nuclides from water supplies. The table indicates carbon adsorption is
effective for removing both VOCs and SOCs. Packed tower and diffused aera-
tion are best suited for removing VOCs. Ion exchange has been field-tested
to show effective removal of nitrates and pilot-tested for uranium removal
Reverse osmosis (RO) has proven to be effective in the field for radium
-------
removal and pilot-tested for nitrate removals. Of the technologies that show
promise and are being tested at the bench-.and pilot-scale, conventional treat-
ment with powdered activated carbon (PAC) is effective for removing a few of
the SOCs, ozone oxidation is effective for removing certain classes of VOCs
and SOCs, and certain reverse osmosis membranes and ultraviolet treatment are
also potentially effective against VOCs and SOCs. Aeration and carbon adsorp-
tion are being examined for their radon removal capabilities.
Table 6 summarizes the various technologies examined for contaminant
removal, and based on data gathered to date, it attempts to characterize
their relative performances for both Phase I and Phase II organics.
TABLE 5. TREATMENT TECHNOLOGIES EVALUATED BY DWRD FOR REMOVING
VOCs, SOCs, NITRATES, AND RADIONUCLIDES FROM DRINKING WATER
Technology
Status
Field-tested 1.
2.
3.
4.
Pilot-tested 1.
2.
Promising Technologies 1.
2.
3.
4.
5.
6.
7.
8.
Technology Contaminant Class or
Specific Contaminant
Removed
Carbon Adsorption
Packed Tower and
Diff used-Air Aeration
Ion Exchange
Reverse Osmosis
Reverse Osmosis
Ion Exchange
Conventional Treatment
with Powdered Activated
Carbon
Ozone Oxidation
Reverse Osmosis
Ultraviolet Treatment
Ion Exchange
Selective Comp lexer
Aeration
Carbon Adsorption
1.
2.
3.
4.
1.
2.
1.
2.
3.
4.
5.
6.
7.
8.
VOCs, SOCs
VOCs
Nitrates
Radium
Nitrates, Uranium
Uranium
SOCs
VOCs, SOCs
VOCs, SOCs
VOCs, SOCs
Radium
Radium
Radon
Radon
-------
TABLE 6. PERFORMANCE SUMMARY FOR ORGANIC TECHNOLOGIES EXAMINED
REMOVAL EFFICIENCyt
Granular
Activated
Carbon
Regulatory Adsorption
Phase Organic Compounds Filtrasorb 400a
VOLATILE ORGANIC CONTAMINANTS
Alkanes
I Carbon Tetrachloride 44
I 1,2-Dichloroethane 44
I 1,1,1-Trichloroethane 44
II 1,2-Dichloropropane 44
II Ethylene Dibromide 44
II Dibromochloropropane 44
Alkenes
I Vinyl Chloride 44
II Styrene NA
I 1,1-Dichloroethylene 44
II cis-l,2-Dichloroethylene 44
II trans-l,2-Dichloroethylene 44
I Trichloroethylene 44
Aromatics
I Benzene 44
II Toluene ++
II Xylenes -H-
II Ethylbenzene ++
II Chlorobenzene 44
II o-Dichlorobenzene 44
I p-Dichlorobenzene 44
PESTICIDES
II Pentachlorophenol -H-
II 2,4-D ++
II Alachlor ++
II Aldicarb NA
II Carbofuran ++
II Lindane ++
II Toxaphene ++
II Heptachlor ++
II Chlordane ++
II 2,4,5-TP -H-
II Methoxychlor ++
Packed
Tower
Aeration
44
44
44
44
44
4
44
NA
44
44
44
44
44
44
44
44
44
44
44
0
0
44
0
0
0
44
44
0
NA
NA
Reverse
Osmosis
Thin
Film
Composite
44
4
44
44
44
NA
NA
NA
NA
0
NA
44
0
NA
NA
0
44
4
NA
NA
NA
•H- .
NA
44
NA
NA
NA
NA
NA
NA
Ozone
Oxidation
(2-6 ppm)
0
0
0
0
0
0
44
44
44
44
44
+
44
44
4+
44
4
4
4
44
4
44
NA
44
0
NA
44
NA
4
NA
Conventional
Treatment
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
NA
0
0
NA
0
0
0
NA
NA
NA
NA
(See footnotes at end of table)
(Continued)
-------
TABLE 6. PERFORMANCE SUMMARY FOR ORGANIC TECHNOLOGIES EXAMINED (Contd.)
REMOVAL EFFICIENCY*
Granular Reverse
Activated Osmosis
Carbon Packed Thin Ozone
Regulatory Adsorption Tower Film Oxidation Conventional
Phase Organic Compounds Filtrasorb 400a Aeration Composite (2-6 ppm) Treatment
OTHER
II Acrylamide NA 0 NA NA NA
II Epichlorohydrin NA 0 NA 0 NA
II PCB's -H- -H- NA NA NA
Excellent 70% - 100%
+ = Average Removal 30% - 69%
0 - Poor 0% - 29%
NA ™ Data not available or compound has not been tested by EPA Drinking Water
Research Division
a - Excellent removal category for carbon indicates compound has been demonstrated to
be adsorbable onto GAG, in full- or pilot-scale applications, or in the
laboratory with characteristics suggesting GAC can be a cost-effective technology.
-------
Although Table 6 only provides a general guideline for removal of com-
pounds, several interesting trends are note'd. Carbon adsorption appears to
provide removal for a wide range of organics whereas conventional treatment
is revealed as a poor treatment for those compounds listed in the table.
Packed tower aeration manifests itself as an excellent technology for volatile
organic compounds and may have application for a limited number of pesticides.
Ozone oxidation appears to be a good treatment technology for certain classes
of organics such as simple alkenes and aromatlcs, as well as certain similar,
but more complex organic structures. Although only a few organics have been
subjected to long-term testing via reverse osmosis, promising removals for
several low molecular weight organics can be seen.
Table 7 shows that ion exchange and reverse osmosis each result in
excellent contaminant removals.
TABLE 7. PERFORMANCE SUMMARY FOR INORGANIC TECHNOLOGIES EXAMINED
Regulatory
Phase
II
III
III
III
Inorganic
Compound
Nitrate
Radium
Uranium
Radon
Removal Efficiency
Reverse Ion Carbon
Osmosis Exchange Aeration Adsorption
++ -H-
++ 4+
++ -H-
* *
-H- - Excellent 70%-lOOZ
* - Research being conducted by DWRD
Tables 6 and 7 were generated using a variety of sources, including
EPA-DWRD pilot- and field-scale studies, Henry's Law Constants for predic-
tion of removal of some of the pesticides, as well as the use of oxidative
-------
trends for predicting the removal of complex pesticides. These tables,
therefore, carry the caveat that the cited-removals should not be used for
design purposes, but that each technology must be tested on compounds, under
field conditions, before the EPA-DWRD will advocate a technology's use.
SURFACE WATER FILTRATION/MANDATORY DISINFECTION
DWRD's primary activities in this area are the studies related to Giardia
and viral inactivation. Current studies on Giardia inactivation are being
conducted at Ohio State. Hepatitus A and phage inactivation studies are
being conducted at the University of North Carolina. Field studies related
to inactivation of Giardia are being planned. A major result of these studies
are the calculation of CT values for various organisms at different levels of
temperature, pH and disinfectant dose. The following projects are ongoing in
this area:
0 Reports on concentration and time values for Hepatitis A
and coliphage inactivation by:
Chlorine and Chloramines
Chlorine dioxide and ultraviolet light
° Laboratory - scale studies on the inactivation of Giardia Muris
by chlorine, chlorine dioxide, and chloramines.
0 Pilot scale study at Harrisburg, PA on inactivation 'of virus
and Giardia by chlorine and chloramines.
0 Laboratory study on inactivation of Cryptosporidium by chlorine
and chloramines.
e Evaluating treatment plants and distribution systems for
Legionella.
-------
PROHIBITION OF LEAD MATERIALS
The amendments to the SDWA prohibit the use of lead in any material in
contact with potable water. DWRD's major current project dealing with lead
solder is being conducted with the South Huntington Water District on Long
Island, New York.
Ninety sites were identified and selected for this study. Ten homes
were selected in each of nine age groups from zero to over 20 years old.
These sites have been sampled three times at different pH levels: less than
6.2, 7.2 and 8.2. The homes were randomly sleeted to obtain a geographic
distribution In the 19.4-square-mile service area of the District. The type
of existing solder was verified in each home through scraping of an exposed
solder joint and testing by an atomic absorption spectrophotometer. Of 95
homes tested, only one had less than 0.5 percent lead In the solder; 67.3
percent of the homes had lead content ranging from 55 to 65 percent In the
solder.
First-draw samples were tested for copper and cadmium in addition to
lead. An additional sample was drawn for determination of various water
quality parameters including pH. Copper values above the 1 milligram per
liter (mg/L) secondary Maximum Contaminant Level (up to 7.77 mg/L at 6.9 pH)
were found. These sites were checked for stray electric currents in the
water service pipe to determine whether or not the high copper values were
caused through grounding of electric systems to the water plumbing system.
During low pH sampling of 64 first-draw tests for cadmium, only one
sample at 42 ug/L was above the drinking water standard. Since the second
sample at 10 seconds contained only 3 ug/L cadmium, it Is assumed the presence
of cadmium was caused by the faucet.
-------
On the first-draw lead samples taken, 61.8 percent were above the drink-
ing water standard. The results ranged from a low of 2.0 ug/L in a 1968 home
at pH - 5.9, to values ranging from 500 to 1,200 ug/L (pH ranges from 5.6 to
6.9) at homes constructed between 1981 and 1983. However, the highest lead
value (1,300 ug/L) was obtained at pH - 5.9 in a 1968 home.
There appears to be a correlation between the age of the lead solder and
its ability to leach lead into the drinking water. Ignoring the first zero-
second flush and allowing the first 300 milliliters (mL) of flush for the
faucet's leaching of lead, the second sample at 10 seconds were compared for
75 homes. For homes constructed in 1980, 1981, 1982, and 1983, the second
125 mL sample (taken in March and April, 1984) indicated that 23 homes, or 70
percent, exceeded the 50 mg/L drinking water standard, and 10 homes were less
than this maximum contaminant level. For homes constructed between 1955 and
1979, only seven, or 17 percent, exceeded the drinking water standard at a
low pH with 35, or 85 percent, less than the maximum contaminant level.
Table 8 summarizes some of the first draw results.
TABLE 8. RESULTS FROM HOUSEHOLD LEAD STUDY FIRST DRAW SAMPLES
pH % of Homes Exceeding 50 ug/L
< 6.2 58.7
6.2 - 8.2 26.6
78.2 24.4
-------
Leaching of metallic solders, particular/ lead solder, appears to be
affected by the following factors:
pH of water
0 Plumbing workmanship,
0 Hardness of water
0 Time since last use of water
Percentage of lead in solder.
Where these conditions are conducive to lead leaching, large numbers of
consumers may be exposed to high-lead, first-draw water.
Several interesting results have come from this ongoing study that may
provide information on leaching from various kinds of solder from special
pipe loops. Tables 9 and 10 provide some useful summary data from the project.
As can be seen there is lead in the various alternative solders being tested.
In addition all four pipe loops have experienced copper leaching at very high
levels.
TABLE 9. LEAD SOLDER PROJECT
Solder Type
Tin/ Silver
Tin/ Silver
Tin/ Silver
Highest Lead Value
0.015 mg/L
0.042 mg/L
0.057 mg/L
PH
5.3
5.4
5.3
Standing Time
4 hrs
4 hrs
4 hrs
-------
TABLE 10. COPPER LEACHING IN FOUR PIPE LOOPS
pH
5.5
5.5
5.5
5.5
Solder
Tin/ Silver
Tin/ Antimony
Tin/Copper
Tin/Lead
Copper
(fflg/L)
3.80
4.28
4.50
4.28
Tables 11, 12 and 13 illustrate some of the problems to be encountered
when trying to meet the presently discussed 20 ug/L lead level in households.
These data are from other ongoing research projects. As can be seen at lower
pHs for short sampling periods in new homes it will be very difficult to meet
the proposed standard. At higher pH's for longer sampling periods and for
Increasing age of test site the proposed standard can be met although it is
only met consistently in older homes.
TABLE 11. PERCENTAGE OF TEST SITES WITH LEAD IN DRINKING WATER
GREATER THAN 20 ug/L AT LOW pH (6.4 & LESS)
AGE OF
TEST SITE
(Years)
0-1
1-2
2-3
3-4
4-5
6-7
9-10
15 - 16
20 & Older
FIRST
DRAW
100%
100%
86%
100%
86%
78%
71%
57%
86%
10
SEC
100%
71%
86%
86%
57%
44%
29%
14%
27%
20
SEC
100%
86%
57%
100%
29%
33%
14%
14%
29%
30
SEC
100%
57%
57%
71%
43%
33%
14%
14%
0%
45
SEC
100%
57%
43%
71%
43%
11%
14%
14%
14%
60
SEC
86%
29%
43%
71%
43%
11%
14%
14%
0%
90
SEC
86%
43%
43%
29%
14%
11%
0%
14%
14%
120
SEC
88%
14%
29%
29%
0%
0%
0%
14%
0%
-------
TABLE 12. PERCENTAGE OF TEST SITES WITH LEAD IN DRINKING WATER
GREATER THAN 20 ug/L "AT MEDIUM pH (7.0 - 7.4)
AGE OF
TEST SITE
(Years)
0-1
1-2
2-3
3-4
4-5
6-7
9-10
15 - 16
20 & Older
FIRST
DRAW
100%
80*
40%
50%
30%
10%
20%
40%
20%
10
SEC
90%
60%
20%
20%
10%
0%
0%
20%
0%
20
SEC
90%
30%
10%
20%
10%
0%
0%
20%
0%
30
SEC
60%
10%
10%
30%
0%
0%
0%
10%
0%
45
SEC
30%
20%
10%
20%
10%
0%
0%
0%
10%
60
SEC
20%
0%
0%
30%
0%
0%
0%
0%
0%
90
SEC
10%
10%
0%
30%
0%
0%
0%
0%
0%
120
SEC
10%
0%
0%
20%
0%
0%
0%
0%
0%
TABLE 13. PERCENTAGE OF TEST SITES WITH LEAD IN DRINKING WATER
GREATER THAN 20 ug/L AT HIGH pH (8.0 & GREATER)
AGE OF
TEST SITE
(Years)
0-1
1-2
2-3
3-4
4-5
6-7
9-10
15 - 16
20 & Older
FIRST
DRAW
100%
67%
30%
25%
30%
20%
10%
33%
20%
10
SEC
100%
22%
10%
0%
10%
0%
0%
22%
0%
20
SEC
60%
11%
10%
0%
0%
0%
10%
11%
0%
30
SEC
10%
11%
0%
0%
0%
0%
0%
11%
0%
45
SEC
20%
11%
0%
0%
0%
0%
0%
0%
0%
60
SEC
10%
0%
0%
0%
0%
0%
10%
0%
0%
90
SEC
20%
11%
0%
0%
0%
0%
0%
0%
0%
120
SEC
0%
0%
0%
13%
0%
0%
10%
0%
0%
DETERIORATION OF WATER QUALITY
DWRD is and has been conducting extensive studies into the deterioration
of water quality in distribution systems. Methodology is being developed for
-------
calculating the AOC associated with treated and distributed water. Work is
being conducted into the identification of legionella and other organisms
that may exist in distribution systems. The following general areas are
being studied:
1. Microbial Contamination of Distribtulon Systems
Research has focused on corrosion inhibitors and the possibility of
their acting to promote microbial growth In water mains. Other research
(2 projects) are examining protection of bacteria by higher organisms as a
mechanism for bacteria to be transported into the distribution system In a
viable state. Still another project has been funded to learn about protec-
tion and transport of bacteria by GAG fines. Present work includes testing
of sampling concepts to detect coliform bacteria and research on the assimil-
able organic carbon (AOC) in water and its relationship to bacterial growth
and proliferation in mains. Inhouse and extramural work on AOC started in
FY'86 and will continue into FY'88. Additional work related to biological
stability of treated water will be needed, especailly if ozone is to be used
more extensively In water treatment.
2. General Corrison Control Studies
In addition to the lead study mentioned previously, field studies on
corrosion control methods for lead, galvanized steel and copper pipes continue
to take place at various communities in the State of Washington. The project
emphasizes the practicality of providing realistic assistance to utilities to
implement corrosion control measures. Finally, bench-scale studies on the
chemistry of silicates and phosphates for corrosion control has been completed
at the University of Missouri.
-------
3. Modeling and Cost Research
Extramural and inhouse research efforts have focused on the following
four aspects of water main distribution system deterioration: (1) Computer
models for mainframe and microcomputers to predict water quality and cost
have been developed. This model, for example, estimates TTHM levels through-
out the distrlbtuion system based on time-of-travel and mixing of water from
wells of varying quality. Cooperative work dealing with the efforts of
hydraulic behavior on the propagation of contamination in distribution systems
has been initiated with members of a French-U.S. bilateral team; (2) Case
studies have collected main break data and applied statistical techniques to
identify system characteristics, environmental factors and installation
practices that contribute to abnormally high break rates that can impair
water quality; (3) Case studies have also analyzed repair, replacement and
renovation costs and have to incorparate data base collection and analysis
into a decision making strategy for utilities to minimize costs and insure
hydraulic integrity and reliability; and (4) A study is underway to develop
standardized distribution system costs and includes new construction, repair
and rehabilitation.
SUMMARY AND CONCLUSIONS
DWRD is active in conducting treatability and technology studies to
assist In development of maximum contaminant levels under the Safe Drinking
Water Act. The Division is also involved in giving technical advice to
states, EPA Regional Offices, water utilities and professional organizations.
An organization of major interest to EPA is the Association of State Drinking
-------
Water Administrators. In many ways the State Drinking Water Administrators
are the first line of implementation of the SDWA. It is the purpose of this
paper to help provide information to the State Adminstrators and to provide
a point of contact for future information concerning the evaluation of various
kinds of technologies for meeting SDWA maximum contaminant levels.
-------
REFERENCES
1. Safe Drinking Water Act 1986 Amendments, United States Environmental
Protection Agency, Office of Drinking Water, EPA 570/9-86-002, August
1986, Washington, DC 20460.
2. Clark, Robert M. and Dorsey, Paul, "A Model of Costs for Treating
Drinking Water", Journal of the American Water Works Association,
Vol. 74, No. 12, pp. 618-d627, December 1982.
3. Fronk, Carol Ann, "Removal of Volatile Organic Chemicals in Drinking
Water by Ozone Treatment". Submitted for publication to the Journal
of the American Water Works Association.
4. Sorg, Thomas J., and Love, 0. Thomas Jr., "Reverse Osmosis to Control
Inorganic and Volatile Organic Contamination" Submitted for publication
to the Journal of the American Water Works Association.
5. Lykins, Benjamin W., Geldreich, Edwin E., Adams, Jeffrey Q., Ireland,
John C., and Clark, Robert M., "Granular Activated Carbon for Removing
Nitrohalomethane Organics from Drinking Water (Project Summary)".
EPA-600/S2-84-165, WERL, Cincinnati, OH, 1984.
6. U.S. EPA Research Cooperative Agreement CR809974, "Evaluation of Packed
Tower Aeration for Removal of Volatile Organics from Drinking Water",
Project Officer: R. J. Miltner, Drinking Water Research Division,
Cincinnati, OH, 1982-1985.
7. Miller, S. E., DeRoose, F. L., Howes, J. E., Tabor, J. E., Hatchel, J. A.,
Sueper, C. V., Kohler, D. F. and Degner, K. B., "Determining the Effective-
ness of an Afterburner to Reduce Dioxins and Furana". EPA/600-2-86-039,
WERL, Cincinnati, OH, March 1986.
8. Lauch, Richard P. and Guter, Gerald A-, 1986. "Ion Exchange for the
Removal of Nitrate from Well Water", Journal American Water Works Associa-
tion, Vol. 78, No. 5, pp. 83-88.
9. Reid, George W., Lassovszky, Peter, and Hathaway, Steven, 1985. "Treat-
ment, Waste Management and Cost for Removal of Radioactivity from Drink-
ing Water", Health Physics, Vol. 48, No. 5, pp. 671-694.
10. Federal Register, 1979. National Drinking Water Regulations, 40CFR Part
143, Vol. 44, No. 140, Thursday.
11. Sorg, Thomas J., Forbes, Robert W. and Chambers, David S., 1980. "Removal
of Ra-226 from Sarasota County, Florida Drinking Water by Reverse Osmosis",
Journal American Water Works Association, Vol. 72, No. 4, pp. 230-237.
12. Lykins, Benjamin W., Jr. and Baler, Joseph A., 1985. "Removal.of
Agricultural Contaminants from Groundwater". In Proceedings: American
Water Works Association Annual Conference, June 23-27, 1985, pp. 1151-1164.
-------
13. Huxstep, Martin R., 1981. "Inorganic Contaminant Removal from Drinking
Water by Reverse Osmosis", (Project Summary, EPA-600/S2-81-115, WERL,
Cincinnati, OH.
14. Guter, Gerald A., 1982. "Removal of Nitrate for Contaminated Water
Supplies for Public Use: Final Report". (Project Summary), EPA-600/S2-
82-042, Cincinnati, OH.
15. Myers, Anthony G., Snoeyink, Vernon L. and Snyder, David W., 1985.
"Removing Barium and Radium Through Calcium Cation Exchange", Journal
American Water Works Association, Vol. 77, No. 5, pp. 60-66.
16. Syner, David W., Snoeyink, Vernon L. and Pfeffer, Julie L., 1986.
"Weak-Acid Ion Exchange for Removing Barium, Radium and Hardness,
Journal American Water Works Association, Vol. 78, No. 9, pp. 98-104.
-------
Barriers to Implementing New Technologies
G. Wade Miller, Executive Director, Association of State Drinking
Water Administrators and President, Wade Miller Associates, Inc.,
Arlington, VA
IV-1
-------
AN INFORMATION SERVICE OF THE AWWA RESEARCH FOUNDATION
Research
Quarterly
PROJECT REPORT
Overcoming Barriers to the Introduction of
New Technologies
G. Wade Miller
President
Wade Miller Associates, Inc.
Arlington, Virginia
In June 1984, Wade Miller Associates, Inc., and the
Utah Bureau of Public Water Supplies initiated a
project to identify barriers to the introduction of new
technologies into the drinking water field and to
determine how some or all of these barriers can be
eliminated. Sponsored by the AWWA Research
Foundation, this project focused on the roles and
interactions of equipment suppliers, state plan review
engineers, consulting engineers, and utilities in the
process of designing and constructing water treatment
facilities and appurtenances. The project was
designed to identify solutions since it was known that
barriers to new technologies exist.
If the results of the study could be distilled into one
statement, it would be that there are no heroes or
villains responsible for introducing or blocking new
technology in the potable water supply industry. Each
of the involved parties operates from a different
perspective and also has a different set of concerns.
These concerns often are either not appreciated or
fully understood by the other parties. These
perceptions and concerns impact specific actions
taken or not taken. All parties are sincerely interested
in seeing that water systems have well-designed, cost-
effective technologies. The failure to achieve this goal
is rooted in the basic decision structure in which all
the major participants operate.
The rate of introducing new technologies into water
supply, and in fact into all facets of local government,
has historically been slower than it has been in many
other sectors of the economy. There are several
reasons why this is true.
« Several types of organizational entities are
engaged in the process, and their interactions
often lead to miscommunications and
breakdowns in the system.
• The water industry itself is a mature industry; the
emphasis is on reliability and use of proven
treatment techniques and products. Although
there is innovation and change in some aspects
of water supply, the rate of change is
considerably slower than in the field of
microcomputers or telecommunications, for
example.
• Because of public health concerns and the
emphasis on proven techniques, the number of
new products being introduced is low relative to
other industries. Some products that are
considered new, such as ozone and reverse
osmosis, have actually been used in water
treatment plants since 1906 and 1967,
respectively.
Organizational Entities
The five major organizational entities that interact
continuously in the course of water plant design,
construction, and operation, are: state drinking water
officials who review and approve plans and
specifications, consulting engineers who design the
facility, equipment suppliers, the general contractor,
and the utility.
State regulatory agencies. State drinking water
administrators are public health officials charged with
the primary responsibility for protecting the public
health by ensuring that utilities deliver a potable, safe
supply of water to the consumer. They view their role
as one of primary oversight of designers, equipment
suppliers, and water utility owners and operators. In
this role, they often must evaluate complex
technologies and rule on the use of substances whose
chemistry is not always well understood. Long after
the engineer and contractor have completed their
work and the warranty on the equipment has expired,
state drinking water officials must monitor drinking
water quality and work with problem systems.
As a group, state regulators perceive themselves as
understaffed and less than well equipped in terms of
resources and access to knowledge needed for the
review process. Consequently, there is a strong
predisposition to favor "tried and true" technologies.
The principal concern is system operation; cost is
-------
Water
Research
Quarterly
AN INFORMATION SERVICE OF THE AWWA RESEARCH FOUNDATION
relegated to a secondary consideration. Thus, state
regulators tend to favor old-line technologies with
conservative design parameters. By requiring unit
processes to be lightly loaded and/or to have very
long detention times, it is believed that the operator
will have more time to respond to problems. Because
of their lack of resources and their commitment to
protection of the public health, state regulators often
appear to be arbitrary and conservative in their
judgments.
Consulting engineers. Design engineers are concerned
with designing a water treatment plant that will
function efficiently. This generally means that
equipment selected is tried and proven. The engineer
has little incentive to specify new, often unproven
technology. In fact, the engineer has a disincentive; if
the plant does not function properly, liability claims
and/or damage to the firm's reputation can be
significant.
The consulting engineer is the lead professional in
most water supply system projects. He is responsible
for determining the client's needs and designing and
overseeing the installation of a water treatment system
that will meet those needs within the regulatory and
public health framework. In effect, he is the primary
procurer of products and technologies.
The survey of 20 consulting engineers conducted in
this study served to confirm that designers are highly
conservative in their philosophy and not inclined to
take a leading role in advocating new technologies
and products. The key factor that drives consulting
engineers to rely on well-established older
technologies is liability. The engineering firm generally
holds virtually all of the liability associated with the
design of a treatment plant. State regulatory personnel
have no liability. Equipment supplier liability is limited
to product warranties. So long as the plant is
constructed properly, the contractor is not liable for its
performance. As a result of this situation, engineering
firms must carry expensive liability insurance;
insurance carriers are quick to cancel or raise fees to
exorbitant levels if a claim is filed. Hence, the engineer
has a substantial amount to lose if a new technology
fails to perform. Since many of the smaller water
systems are designed by small engineering firms, a
major problem can force the firm to close down.
Equipment suppliers and general contractors. The most
elegant plans and specifications will not treat a single
drop of water until they are converted into the
physical structure of the treatment plant. The
contractor and the equipment supplier, working with
the engineer, create the actual system from the paper
abstraction. General contractors are the favorite target
of criticism by those who conveniently ignore the
highly competitive, low-bid environment in which
general contractors must work.
The equipment supplier is the primary source of
innovation and forward movement in water treatment
technology. The supplier's environment stimulates
change and innovation since the company's survival
and prosperity depend to a large part on its ability to
improve on the state of the art and thereby create a
salable advantage over competitors in a highly
competitive and price-driven environment. The
supplier performs many valuable services, including
research and development (R&D), pilot testing,
training and supporting design engineers, and
providing supplies and services to operators. Much of
the funding for professional activities, publications,
and exhibitions comes directly or indirectly from
suppliers seeking to advertise their products.
Equipment suppliers are concerned primarily with
selling a product that satisfies the demands of the
marketplace. The supplier is profit motivated.
However, his profits are often invested in research or
product development that ultimately benefits the
industry.
Utilities. Utilities are often the forgotten organizational
entity. Water utilities are usually as conservative as
their state counterparts. Like the consulting engineers,
they have little incentive to try new products or
process equipment. When they do opt for installation
of an innovative process or treatment technique, the
incentive is the promise of a significant reduction in
costs—capital costs, operation and maintenance costs,
or both.
Others. The role of the U.S. Environmenta! Protection
Agency (USEPA) in product introduction should be
mentioned briefly. Although the USEPA is the primary
developer of drinking water regulations, it specifically
does not get involved in product testing or
endorsement and is not directly involved in the
"decision loop" of states, consulting engineers,
utilities, and equipment suppliers. It does get involved
in the approval-acceptance of direct and indirect
additives to drinking water, but only on an advisory
basis. The USEPA's advisory opinions on treatment
chemicals and paints and coatings have never been
mandatory. The state regulatory agencies have
primary authority and responsibility in this area. Many
choose to rely primarily on USEPA's advisory rulings,
making them mandatory for approval. Since the
USEPA limits its work to issuing advisory opinions on
new products that are essentially identical in chemical
formulation to previous products, states that limit their
own approvals to products that have been "approved"
by the USEPA basically bar the introduction of any
new products of this type in their states.
Specific Barriers
Certainly one of the principal barriers to introduction
of new technologies is the methodology by which
(continued on page 8)
-------
8
AN INFORMATION SERVICE OF THE AWWA RESEARCH FOUNDATION
(continued from page 7)
water treatment systems are designed, purchased, and
operated. It has been called the "great American low-
bid process," the "eternal triangle" of the consulting
engineer who designs the plant, the general
contractor who builds it (using equipment from the
supplier), and the owner who operates it. While having
no direct fiduciary interest, the state drinking water
administrator, as the principal guardian of the public
health, is also very much involved.
When something goes wrong, an all too frequent
occurrence, the situation quickly dissolves into a
triangle of "finger pointing," each party trying to shift
the responsibility to the other. The entire system is full
of loopholes: the contractor may substitute cheaper,
inferior equipment that local or federal procurement
regulations force the engineer or owner to accept; the
engineer fails to "do his homework," resulting in less-
than-efficient processes and hardware. The owner
may take the best designed and constructed plant in
the world and produce poor quality water by using
unqualified, untrained operators.
This process does not establish a clear, single source
of responsibility that can efficiently design, construct,
and operate a plant. The net effect of the system is to
drive the technology and equipment quality to the
lowest common denominator, and drive the costs up,
to the benefit of no one.
Another barrier is the lack of a uniform set of
guidelines for reviewing new or old technologies.
Consequently, the supplier finds that he must meet
widely divergent standards for treating waters of
similar composition that simply happen to be located
in different geographical and/or political areas. This
serves to drive up the cost of equipment and to
complicate system design.
State regulatory agencies that do not periodically
review and update their standards may appear to have
little technical basis for existing standards. In some
extreme cases, an entire policy having major
economic impact on the state's water utilities may be
rooted in the personal prejudices of a long-term
senior employee who has not kept abreast of technical
evolution. Policies in states operating on this basis
may be characterized by abrupt changes that
correspond to the retirement or reassignment of key
personnel.
About one-half of the states rely on the "Ten States
Standards" (TSS) for all or a major part of their basic
criteria. Those states that are not a part of the TSS
committee that generates these guidelines tend to
interpret the standards in a much more rigid manner
than was intended by the authors. This has led to the
phenomenon of new technologies being approved by
TSS committee states, while the same are rejected by
an outside state citing the TSS as the reason. Use of
the TSS by so many states reflects a strong desire
among state officials for greater commonality of basic
standards. The problem with the TSS is that they have
institutionalized some basic criteria that are not based
on best available technology. For example, even
though modern filters provide effective treatment at
rates far in excess of the two-gallons-per-minute-per-
square-foot guideline listed in the TSS, this
requirement remains in the standards long after it has
been rendered obsolete.
The combination of the low-bid environment and the
lack of serious prequalification of suppliers has placed
severe profit constraints on the equipment supply
industry. Suppliers have responded by cutting back on
services, reducing warranties, substituting cheaper
designs and materials in products, and reducing
product R&D efforts. The first cost of the product may
be lower; however, the service life cost and
maintenance headaches associated with cheap
equipment may dwarf any up-front savings.
The great lack of communication between equipment
suppliers and state regulatory officials contributes to
an atmosphere of distrust and misunderstanding
between them. Survey responses and conversations
with members of both groups during this study clearly
established that each party suffers from considerable
misconceptions about the other and these misconcep-
tions are allowed to interfere in the working
relationship to the detriment of both parties.
Conclusions and Recommendations
Even though there are clearly several barriers to the
introduction of new products, the system does work.
As noted earlier, most water treatment professionals-
state administrators, consulting engineers, equipment
suppliers, and utility managers—have a common goal
of providing a high quality product to the consumer.
The procurement system, as currently structured, is
inefficient and unwieldy, however. Virtually no
incentives exist for risk taking, either by the consulting
engineers or the state plan review engineers. Utilities
sometimes can benefit from risk taking in the form of
lower costs of an innovative technology. For the most
part, however, the water treatment equipment supplier
is selling to a conservative constituency that places a
high premium on reliability and, under most state
laws, low-bid prices.
Of all the key actors surveyed'during this study, only
the consulting engineers appear to accept the current
system used for procurement, design, and
construction of water treatment facilities. The state
drinking water administrators recognize that they are
perceived as "barriers to innovation" and would like to
take actions to ensure that they are not barriers. The
more progressive utilities may be frustrated by
conservative state guidelines and would like to see
-------
water
Research
Quarterly
AN INFORMATION SERVICE OF THE AWWA RESEARCH FOUNDATION
changes. Most equipment suppliers are frustrated by
the present way of doing business, but are relatively
powerless to change the status quo.
In addition to the inefficiency of the procurement
process, the industry is plagued by a clear lack of
communication. This lack of communication cuts
across all groups. The states, until the recent
formation of the Association of State Drinking Water
Administrators (ASDWAi, have not had a forum for
sharing information with each other. Equipment
suppliers often appear to develop new products
without consideration for the concern of state
administrators. They market their products to the
consulting engineer and typically consider the
engineer, and not the utility, to be the client.
Consulting engineers are criticized by states and
equipment suppliers alike for not being willing to "take
more chances." Engineers are especially criticized for
not sticking by a specification once written.
Principal recommendations for actions that can break
down some of the existing barriers are as follows:
1. Consideration should be given to the
development of a uniform set of national guidelines
that can be used by all states in the plan review
process. The feasibility of such a set of standards is a
controversial topic; however, the states appear to want
and need the consistency that would result from this
action. It is exceedingly important that any such
guidelines or standards be kept as current as possible.
Continued use of outdated design parameters based
on outmoded technology can be a major barrier to the
introduction of new technologies if such parameters
are incorporated into national guidelines or standards.
2. There is a clear consensus among the majority of
state agencies and most of the other participants for a
national "clearinghouse" that would support direct
communication between state regulators, equipment
suppliers, consulting engineers, and water system
operators. The clearinghouse could serve as a vehicle
for disseminating new product information and test
data, and generally foster the spread of design,
product, and operational information within the
industry.
3. The current practice of low-bid purchasing
inhibits the introduction of new technologies by
causing credible suppliers to expend unnecessary
resources while watching low-bid suppliers win
procurements on the basis of the first cost, not service
life costs. The process makes it more difficult for state
officials to approve new technologies, particularly
packaged treatment systems, for fear of inferior
substitutions to approved products at the time of
bidding. Numerous examples of successful equipment
and service procurement on a negotiated or
prequalified bid basis exist in the municipal sector
overseas (i.e., as in France where two large, vertically
integrated companies dominate the water industry),
and in the municipal privatization and industrial
sectors in the United States.
Feasible alternatives to the current practice should be
defined, and steps should be taken to educate
municipal buyers and consulting engineers in a more
effective procurement approach. The lead role would
have to be taken by USEPA, the American Water
Works Association (AWWA), and ASDWA since
consulting engineers seem to accept the current
procurement process.
4. In the current environment, exhorting the
consulting engineer to "take more risks" on new
technology will have no significant impact unless
some mechanism for reducing his liability in the event
of failure is also developed. A definitive analysis of the
current liability situation and development of
alternative models that would share liability among the
principal participants (i.e., designers, equipment
suppliers, contractors, system owners, and states)
should be undertaken by AWWA, ASDWA, or other
interested parties.
5. State regulatory agencies need to look closely at
their decision making to assure that it is consensus-
based and not dominated by one, or several, strong-
willed staff members who have not kept pace with the
development of water treatment technology. States
should emphasize, even require, participation in
professional educational activities by those staff
members who are reviewing plans and specifications
and who are responsible for design criteria. Resources
should be made available for site visits to treatment
plants that use new technologies, for participation in
pilot studies, for sharing information and exchanging
ideas with plan review engineers in other states, and
for travel to regional or national conferences or
seminars dealing with new technologies.
6. Equipment suppliers need to recognize the
importance of state drinking water officials and make
a concerted effort to work with and understand their
constraints and perspectives. In particular,
manufacturers need to understand the public health
implications that drive the regulatory process and
address those issues in their product submittals and
research efforts. Also, new product development
should include input from key state officials as early
as possible, so that health and operational concerns
can be recognized and incorporated.
7. One of the most important and immediate actions
that should be taken is sponsorship of a workshop (or
a series of workshops) that would bring equipment
suppliers, states, and consulting engineers together in
a forum conducive to constructive dialogue. USEPA's
Office of Drinking Water in conjunction with the
Drinking Water Research Division should provide
funding for such workshops, which should be
conducted by ASDWA alone or with the AWWA
Research Foundation.
-------
Technical Session: Filtration
Sigurd P. Hansen, Senior Engineer, CWC-HDR, Inc., Cameron Park,
CA
Material not enclosed — see separate handout.
V-l
-------
Technical Session: Disinfection/By-Products
Rip G. Rice, President, Rice International Consulting Enterpri-
ses, Ashton, MD
VI-1
-------
CONTENTS
I. INTRODUCTION 1
II. DEFINITIONS 2
III. EPA DISINFECTION REGULATIONS 4
A. STATUTORY REQUIREMENTS 4
1. Variances and Exemptions 5
2. Anticipated Timing of Disinfection Regulations 6
B. THE PROPOSED SURFACE WATER TREATMENT REGULATION 7
1. Applicability 7
2. Microorganisms to be Regulated 7
3. Appropriate Treatment Technologies 9
4. Turbidity Limitations 10
5. Minimum Disinfection Requirements 10
a. For Systems Which Do Not Filter 10
b. For Systems Which Filter 13
IV. EPA DISINFECTION BY-PRODUCTS REGULATIONS '15
A. STATUTORY REQUIREMENTS 15
B. CANDIDATE DISINFECTANT/OXIDANT BY-PRODUCTS TO BE
REGULATED 16
C. ANTICIPATED TIMING OF REGULATIONS 18
e
V. BEST AVAILABLE TECHNOLOGIES AND TREATMENT TECHNIQUES 20
A. STATUTORY REQUIREMENTS 20
B. DEVELOPMENT OF CRITERIA FOR BEST AVAILABLE
TECHNOLOGIES 21
C. BAT FOR DISINFECTION 21
D. BAT TECHNOLOGIES FOR DISINFECTION BY-PRODUCTS . 22
E. STRATEGIES FOR CONTROLLING DISINFECTION BY-PRODUCTS
23
VI. DISCUSSION OF DISINFECTANTS AND OXIDANTS 24
A. BASIC ISSUES 24
1. First Approach 25
2. Second Approach 25
3. Third Approach 27
-------
B. TECHNICAL ISSUES » DISINFECTION AND DISINFECTANTS 28
1. For Systems Treating Surface Waters 28
a. The "CT" Value . 28
b. Suggested No Adverse Response Levels (SNARLs) ... 29
c. Possible Regulatory Consequences 30
1. EPA'S PERSPECTIVE ON RECOMMENDED NAS
SNARLS 31
2. For Systems Treating Groundwater 33
a. "CT" Values and SNARLs 33
b. Possible Regulatory Consequences 34
C. TECHNICAL ISSUES -- DISINFECTION/OXIDATION
BY-PRODUCTS 34
1. Non-Halogenated DOBs 35
a. From Ozonation 35
b. From Chlorination 37
D. COMPARISON OF DISINFECTANTS - OXIDANTS 37
1. General Considerations 37
2. Chlorine 39
3. Chlorine Dioxide 40
41
4. Monochloramine 42
a. Chloramine Summation 43
5. Ozone 44
a. General Considerations 44
b. Disinfection With Ozone 45
c. Microflocculation 46
d. Promotion of Biodegradability 46
e. By-Products of Ozonation 47
f. Catalytic Ozonation 47
g. Ozone Oxidation of Bromide Ion 48
h. Summation for Ozone 48
6. Ultraviolet Radiation for Groundwaters 49
VII. DISCUSSION OF TREATMENT TECHNOLOGIES FOR
DISINFECTION AND TO MINIMIZE PRODUCTION OF DOBs 51
A. THE PROBLEM 51
B. THE STRATEGIES 51
ii
-------
C. DISINFECTION TECHNOLOGIES 51
1. Primary Disinfectants 52
a. Chlorine 52
i. Chemistry of Chlorination 53
ii. Establishing A Chlorine Residual 55
iii. Factors Affecting Disinfection Efficiency of
Chlorine 57
iv. Disinfection With Chlorine Gas 57
v. Disinfection With Sodium Hypochlorite Solution . 58
vi. Disinfection With Solid Calcium Hypochlorite . . 60
vii. Chlorination System Design 60
viii. Chlorination Systems Costs 67
b. Ozone 86
i. Characteristics and Properties of Ozone 86
ii. Generation of Ozone 90
iii. Contacting of Ozone With Water 91
iv. Destruction of Contactor Exhaust Gas 93
v. Chemistry of Ozone in Water 93
vi. Establishing An Ozone Residual 94
vii. Factors Affecting the Disinfecting Efficiency of
Ozone .97
viii. Ozonation System Design 97
ix. Costs of Ozonation Systems 107
c. Ultraviolet Radiation 120
i. General Description of the UV Process 120
ii. UV Disinfection System Design 122
iii. Estimating the Average Intensity in a UV Reactor
124
iv. Water Quality Considerations in the Design of a
UV Disinfection System 124
v. System Design and O&M Considerations for the
UV Process 125
vi. System Design Considerations for Effective
Maintenance 127
vii. Costs for Disinfection With UV Radiation .... 128
2. Secondary Disinfectants 130
a. Chlorine Dioxide 133
i. Generation of Chlorine Dioxide 136
ii. Oxidation-Reduction Reactions of Chlorine Dioxide
• "-. 139
iii. Establishing a Chlorine Dioxide Residual .... 141
111
-------
iv. Factors Affecting the Efficiency of Secondary
Disinfection With Chlorine Dioxide 141
v. Chlorine Dioxide Systems Design 142
vii. Costs of Chlorine Dioxide Generating Systems . . 146
b. Chloramination 152
i. Chemistry of Chloramination 153
ii. Establishing a Chloramine Residual 153
iii. Chloramination System Design 154
iv. Costs for Chloramination 155
3. Oxidants 157
a. Potassium Permanganate 157
VIII. CASE EXAMPLES OF EMERGING TECHNOLOGIES . 162
A. OZONE CASE HISTORIES 162
1. North Andover, Massachusetts
Ozone Disinfection for Giardia lamblia 162
a. The Problem 162
b. The Interim Solution 163
c. The Results .....' 164
d. For The Future 164
e. Note - Sturgeon Bay, WI 165
2. Kennewick, Washington (Cryer, 1986) -
Preozonation For THM Control 165
a. The Problem 165
b. Pilot Plant Study Results 166
c. Plant Design 166
d. Operational Experiences 166
i. General 166
ii. Water Quality 168
iii. Applied Ozonation Dosages - Dissolved Ozone
Residuals 168
iv. Ozone Equipment Operational Experience .... 168
v. Costs Of Ozonation 169
e. Implications of Proposed Surface Water Treatment Rule
Disinfection Conditions 169
iv
-------
B. UV RADIATION CASE HISTORY 17°
1. Ft. Benton, Montana
UV Radiation for Primary Disinfection 170
a. The Problem 170
b. The Solution 170
c. UV Radiation Conditions 171
d. Costs 171
e. Operating Experience 171
C. CHLORINE DIOXIDE CASE HISTORIES 172
1. Evansville, Indiana
Predisinfection for THM Control 172
a. The Problem 172
b. Pilot Plant Study 172
i. Optimization Phase 172
ii. Long-Term Evaluation 172
c. The Full-Scale Plant 173
d. Operating Experience 173
e. Implications of the Proposed Surface Water Treatment
Rule 174
2. Hamilton, Ohio
Primary/Secondary Disinfection With Chlorine
Dioxide 175
a. The Problem 175
b. The Treatment Process 175
c. Generation of Chlorine Dioxide 175
d. Effects of Installation of Chlorine Dioxide 176
e. Costs for Chlorine Dioxide 176
f. Implications of the Proposed SWTR CT Values ... 176
3. Galveston, Texas
Preoxidation With C1O2
Post-Disinfection With C1O2 + Chloramine 177
a. The Problem 177
b. The Original Treatment Process 177
c. Study of Alternatives 178
i. Study Results 178
d. The Results 180
-------
C. CHLORAMINE CASE HISTORIES 181
1. Bloomington, Indiana
Prechlorination, Post-Chloramination 181
a. The Problem 181
b. The Solution 181
c. Performance 181
2. Philadelphia, Pennsylvania
Pre-Chlorine Dioxide + 1-Hour Prechlorination
Post-Chloramination 183
a. The Problems 183
b. Process Modifications 183
i. Chloramination of Finished Water (Reduction of
Free Chlorine Contact Time from 96-hours to 24-
hours) 184
ii. Reduction in Chlorine Treatment at Raw Water
Basin 184
iii. Utilization of Chlorine Dioxide at Raw Water
Basin Inlet (5 Hours Free Chlorine Contact Time)
184
iv. Installation of a New Chlorine Application Point
(One Hour Free Chlorine Contact Time) .... 184
v. Ten Minutes Free Chlorine Contact Time . . . . • 185
c. Economics 185
d. Operational Improvements 185
e. For the Future 186
IX. SUMMARY RECOMMENDATIONS FOR DISINFECTION STRATEGIES
AND FOR THE CONTROL OF DISINFECTION/OXIDATION BY-
PRODUCTS 186
A. For Disinfection 186
B. For Controlling Disinfectant/Oxidant By-Products 188
X. REFERENCES 189
VI
-------
EMERGING TECHNOLOGIES FOR COMPLIANCE WITH
DISINFECTION AND DISINFECTION BY-PRODUCTS REGULATIONS
Rip G. Rice. Ph.D.
Rice International Consulting Enterprises
1331 Patuxent Drive
Ashton, MD 20861
I. INTRODUCTION
The Safe Drinking Water Act Amendments of 1986 as related to Disinfection
and Disinfection By-Products will require all community and non-community
water supplies to disinfect their drinking water supplies. However, this must
be accomplished in such a manner as to minimize or eliminate the formation
of by-products of the disinfection process. Since some of the disinfectants
employed in drinking water treatment (e.g., chlorine, chlorine dioxide, and
ozone) also are used for oxidative purposes other than disinfection, the same
by-products can be produced during such non-disinfection applications.
Historically, chlorine has been the primary disinfectant utilized by American
water utilities. Its use has produced (and continues to produce) excellent
results in terms of killing and/or inactivating pathogenic microorganisms, and
also has provided many additional water treatment benefits (e.g., iron, and
manganese removal, color removal, sometimes taste and odor control, etc.).
On the other hand, chlorine also is an excellent chlorinating agent,
producing trihalomethanes (THMs) and other halogenated organic materials
during water treatment. Consequently, other disinfectants are being studied
and utilized (chlorine dioxide, monochloramine, ozone) as well as combina-
tions of oxidants/disinfectants (i.e., ozone followed by chlorine, chlorine
dioxide, or monochloramine) to "replace" chlorination, or at least to lower
the amounts of chlorine applied rather cavalierly in the past without regard
to the chemical by-products produced by such practice.
Analogously, each of these alternate disinfectants or combinations thereof
can be expected to produce its own set of oxidation/disinfection by-products.
Therefore, EPA's major problem with respect to this subject area is to study
the efficacy of each potentially promising alternative water treatment
scheme, identify the by-products which can be or are produced, and
determine their health effects.
Disinfection By-Products currently are discussed by some as a topic related
solely to the "disinfection1',, step, without considering that the identical by-
products can be produced- during non-disinfection applications of the same
oxidant/disinfectant. When breakpoint chlorination is employed for ammonia
removal early in the treatment process, the same types of halogenated by-
1
-------
products can be produced as when chlorine is added for disinfection after
filtration, given the presence of the same types and concentrations of
organic materials, and the same water parameters.
Ozone and potassium permanganate (KMnO4) are prime candidates for coping
with many Synthetic Organic Chemicals (SOCs) to be regulated under the
SDWA by oxidation early in the water treatment process. As a result of
adding ozone to raw water supplies (for turbidity, color, and/or taste and
odor control, microflocculation, oxidation of THM precursors, or iron and
manganese oxidation, for example), oxidation by-products will be produced
which are of the same general types as when ozone is employed late in the
water treatment process as the primary disinfectant.
Similarly, when potassium permanganate, rarely used, if ever, as a
disinfectant, is applied for oxidative purposes (iron, manganese, color
removal, taste and odor control), many of the same oxidation products are
produced as when ozone, chlorine dioxide, or even chlorine are employed as
disinfectants or oxidants (see Rice & Gomez-Taylor, 1986, and references
cited therein).
II. DEFINITIONS
A. DISINFECTION ~ The killing or inactivation of (pathogenic)
microorganisms.
Attainment of the desired degree of disinfection is a function of the
concentration of the disinfecting agent (in mg/L) times the contact time (in
minutes) (the "CT1 value, in mg/Lrmin). For example, with chlorine at a
concentration of 2 mg/L and a contact time of 30 minutes, the product of
[concentration x contact time] is 60. On the other hand, with ozone applied
under viral inactivation conditions (0.4 mg/L maintained over 4 minutes;, the
"CT" product is 1.6.
B. PRIMARY DISINFECTION - The major disinfection step normally
practiced after filtration in a surface water treatment plant
employing conventional or direct filtration. In groundwater
treatment plants, primary disinfection also can be applied post-
filtration (if filtration is used). If filtration is not incorporated in
groundwater treatment, primary disinfection is practiced in the
treatment plant, prior to the water entering the distribution system.
Oxidants/disinfectants which are used currently for primary disinfection are:
chlorine, ozone, ultraviolet radiation, chlorine dioxide, and monochloramine.
C. SECONDARY DIS|NFECTANT (RESIDUAL DISINFECTANT) - A
disinfectant addecp to finished water to maintain a disinfecting
residual concentration throughout the water distribution system.
-------
Current secondary (residual) disinfectants are: chlorine, chlorine dioxide,
and monochloramine.
D. POINT OF DISINFECTION - The point in the water treatment
process at which disinfectant is added for the purpose of killing or
mactivation of microorganisms.
E. DISINFECTION/OXIDATION BY-PRODUCTS -- Materials produced or
formed whenever a disinfectant or oxidant is added to water.
These by-products can be materials formed by:
(a) Reduction or disproportionation of the disinfectant (or oxidant)
itself, e.g., chlorite and chlorate ions from chlorine dioxide,
dissolved oxygen from ozone, chloride ion from chlorine,
(b) Oxidation of organic materials present in the water, i.e., aldehydes,
ketones, alcohols, and carboxylic acids from the reactions of humic
acids with ozone, chlorine, chlorine dioxide, KMnC>4, or
(c) Halogenation of organic materials upon the addition of halogen-
containing disinfectants, e.g., trihalomethanes and other halogenated
organics during chlorination, organic chloramines produced by
reaction of monochloramine with organonitrogen compounds, etc.
If bromide ion is present in the raw water, it can be oxidized by ozone or
chlorine (not by chlorine dioxide or chloramine) to form hypobromous acid,
which, in turn, can brominate many organic materials. Bromine-containing
trihalomethanes, for example, are known to be formed by this reaction
mechanism.
Until recent years, it had been considered that chlorine is added to drinking
water supplies primarily to disinfect or to remove ammonia. When chlorine
is added to raw water to remove ammonia by breakpoint chlorination, the
oxidative function of chlorine is utilized. Although disinfection also is
obtained when the chlorine is added, nevertheless the primary purpose of
breakpoint chlorination early in the water treatment process is oxidative
removal of nitrogen (ammonia). On the other hand, since the chlorine is
added at the point of greatest concentration of organics capable of being
chlorinated, the maximum concentration of halogenated oxidation products
will be produced simultaneously under these circumstances.
Therefore, and in light of current, more complete understanding of the roles
of various disinfectants as also being strong chemical oxidants, it should be
recognized that anywhere in the water treatment process a strong oxidant is
added for the purpose of oxidation, disinfection may occur simultaneously.
The degree of disinfection gbtained under these circumstances will depend
upon a number of factors, "Including the types of organisms present, the
specific oxidant/disinfectant added, the water temperature, and the time the
-------
disinfectant is in contact with the microorganisms present (i.e., the "CT"
value).
Conseouently, the point(s) of disinfection today should be redefined as any
point(s) in the water treatment process at which a specific oxidant/-
disinfectant is added for any purpose. The same "disinfectant by-products"
can and will be produced, given the presence of the same types and levels of
"disinfectant by-product precursors", and the same water conditions (i.e.,
temperature, pH, ionic strength, alkalinity, hardness, etc.).
From a nomenclature point of view, therefore, it is more appropriate to
refer to "Disinfectant/Oxidant By-Products" (DOBs), rather than merely as
Disinfection By-Products (DBFs).
III. EPA DISINFECTION REGULATIONS
A. STATUTORY REQUIREMENTS
Section 1412(b)(8) of the Safe Drinking Water Act Amendments of 1986
(enacted June 19, 1986) requires the EPA Administrator to propose and
promulgate national primary drinking water regulations requiring disinfection
as a treatment technique for all public water systems within 36 months (by
June 19, 1989). Simultaneously, the Administrator is required to promulgate
a rule specifying criteria that will be used by EPA or delegated State
authorities to grant variances from this requirement, according to . the
provisions of Section 1451(a)(l)(B) and 14l5(a)(3). This disinfection
requirement applies to all waters used for drinking purposes, including
groundwaters.
Section 1445(a) requires the EPA Administrator to promulgate regulations,
not later than 18 months after enactment of the Safe Drinking Water Act
Amendments of 1986 (by December 19, 1987), requiring every public water
system to conduct a monitoring program for unregulated contaminants.
These regulations shall require monitoring of drinking water supplied by the
system, and shall vary the frequency and schedule of monitoring
requirements for systems based on the number of persons served by the
system, the source of supply, and the contaminants likely to be found. Each
system shall be required to monitor once every five years after the effective
date of the Administrator's regulations, unless the Administrator requires
more frequent monitoring.
This monitoring requirement will apply to all disinfection and disinfection
by-product regulations to be discussed in this presentation, as well as to
compounds listed as Volatile and Synthetic Organic Chemicals (VOCs and
SOCs).
-------
1. Variances and Exemptions
A water utility will be required to disinfect unless it successfully applies to
the state primacy agent for a variance. At this time, criteria for obtaining
variances from disinfection requirements are not yet available. However,
proof of microbiological safety during the variance period can be assumed to
be a determining factor, and the State must find that in granting the
variance, an unreasonable risk to the public health will not result.
The respective roles of variances and exemptions under the SDWA
Amendments differ, but the logic is clear. Variances can be granted only on
the basis that because of poor raw water quality, the MCLs cannot be met,
even after application of Best Available Technology. There does not appear
to be any endpoint for compliance under the variance provision, although the
states must impose a compliance schedule as a condition of the variance
[Section 1415].
In EPA's proposed Surface Water Treatment Rule (U.S. EPA, 1987a), EPA
states:
"... due to the acute and high risk associated with poor disinfection of
surface waters, EPA is proposing that no variances be allowed."
EPA also states that no variances will be allowed in the requirements for
filtration of surface water supplies, in the same proposed Rule (U.S. EPA,
1987a). This position is taken under Section 1412(b)(7)(C)(ii) of the SDWA,
which states that in lieu of the variance provisions, the EPA is to specify
criteria by which States will determine which public water supplies will be
required to filter.
Exemptions from any requirement respecting a maximum contaminant level or
any treatment technique requirement, or from both, are allowed based on the
inability to comply due to "compelling factors", which may include economic
factors. As with variances, the state must find that the granting of an
exemption shall not result in an unreasonable risk to health [Section 1416(a)-
(1)]. However, the maximum length of time for an exemption is three years,
and at the time the exemption is granted, the State must prescribe a
schedule for compliance [Section 1416(b)(2)(B)]. For smaller connection
systems (under 500), the exemption can be extended for several two-year
periods [Section 1416(b)(2)(C)], if the system establishes that it is taking all
practicable steps to meet the requirements of Section 1416(b)(2)(B).
In the proposed Surface Water Treatment Rule (U.S. EPA, 1987a), EPA is
proposing that no exemptions be allowed from the requirement of providing
disinfection, but that exemptions be allowed for the degree of disinfection
required and for meeting the filtration requirements. For example, under
certain conditions, an unfiltered system might obtain an exemption if it
achieved a 99% inactivation of Giardia. but not the required 99.9%
inactivation. Guidance for determining conditions under which an exemption
might be granted is discussed in the Surface Water Treatment Rule Guidance
Manual (US. EPA, 1987e).
-------
Additional consequences of requiring disinfection also can be anticipated.
For example, many groundwaters are not disinfected today. Those which
contain iron and manganese will form insoluble oxides when a
disinfectant/oxidant such as chlorine, chlorine dioxide, or ozone is added. In
turn, this will require the addition of a filter for their removal. Thus the
affected groundwater systems may also be forced to install filtration as a
result of being required to disinfect.
Having installed filtration to cope with insoluble materials produced upon
oxidation, the question now becomes, should primary disinfection gjso be
practiced after filtration? If so, this will open the door to two-stage
oxidation/disinfection. This is a prevalent treatment practice in many
modern groundwater treatment plants.
EPA also may require redundancy of the necessary disinfectant feed
equipment, along with an auxiliary power source, automatic start-up (and
alarm) to ensure that continuous disinfection is provided. Many utilities fall
short of having such facilities today.
2. Anticipated Timing of Disinfection Regulations
EPA anticipates promulgating the Surface Water Treatment Rule (SWTR) in
mid-late 1988, which will include requirements for filtration and disinfection.
Treatment processes (filtration and disinfection) will be discussed, individual-
ly and in combination, for removal of specific contaminants. Such processes
will include the use of powerful oxidants (e.g., ozone, chlorine, and chlorine
dioxide), which also can provide primary disinfection at their point(s) of
addition.
In addition to the Surface Water Treatment Rule, EPA intends to promulgate
additional regulations specifying disinfection requirements for systems using
groundwater sources, and possibly more specific disinfection requirements for
surface water systems. This Groundwater Treatment Rule currently is
planned to be proposed early in 1991. It will include regulations for
Disinfection By-Products (when health risks associated with disinfection by-
products are adequately evaluated), revision of THM regulations, and
proposed MCLs for the disinfectants themselves (e.g., chlorine, chlorine
dioxide, chlorite/chlorate ions, and monochloramine).
In early 1988, the Environmental Protection Agency (U.S. EPA, 1988a)
promulgated the First Drinking Water Priority List, which contains 53
substances or classes of substances that are candidates for regulation. These
include the above disinfectants, ammonia, the four THMs currently regulated,
ozone by-products, cryptosporidium. and many additional halogenated organic
materials, some of which are by-products of chlorine disinfection.
At the same time (U.S. EPA, 1988a), EPA finalized the list of 83 substances
mandated by the Safe Drjijking Water Act Amendments to be regulated by
June 1989. Many of these substances are at least partially reactive with
chlorine dioxide, ozone and potassium permanganate. Consequently, they may
be destroyed, or at least converted in chemical form by pretreatment with
-------
strong oxidants, giving rise to byproducts. Such preoxidation will change
the way these oxidized derivatives are removed in subsequent processing
steps.
B. THE PROPOSED SURFACE WATER TREATMENT REGULATION (U.S.
EPA, 1987a)
1. Applicability
This regulation will be applicable to all water treatment systems which use
surface water, because all of these are at risk, at least to some degree, from
contamination by Giardia lamblia. enteric viruses, and pathogenic bacteria.
Surface water is defined as all water open to the atmosphere (e.g., rivers,
lakes, streams, reservoirs, impoundments), and all springs, infiltration
galleries, wells, or other collectors that are directly influenced by surface
water.
"Directly influenced by surface water" means that the source is subject to
pathogen contamination from surface waters. This determination is to be
made on a case-by-case basis. The State will be responsible for determining
which systems use surface water and, therefore, will be subject to the
requirements of this rule.
EPA estimates that more than 90% of the systems which do not filter
currently serve less than 10,000 people. Therefore, the impact of these
proposed regulations will be maximum on small treatment systems.
2. Microorganisms to be Regulated
EPA plans to regulate the microorganisms listed in Table I, plus turbidity,
which is included because it is an indicator of disinfectant performance.
MCLGs for Giardia cysts and enteric viruses are given as "zero", not with
respect to a specific water volume, but rather conceptually. EPA's intent is
to approach the MCLG by requiring treatment technologies (e.g., filtration
and disinfection).
Under the Surface Water Treatment Rule, all community and noncommunity
public water systems using any public water source will be required to treat
their surface water source(s) so as to achieve at least 99.9 percent removal
and/or inactivation of Giardia cysts, and at least 99.99 percent removal
and/or inactivation of enteric viruses. This can be accomplished by
disinfection alone, or by a combination of filtration and disinfection. No
variances or exemptions will be allowed for disinfection requirements.
At a minimum, treatment for any surface water will include disinfection. In
addition, unless the system meets certain site-specific criteria, treatment also
will include filtration. No variances from filtration will be allowed; but
exemptions will be permitted^
-------
TABLE I. MICROORGANISMS AND DISINFECTION GOALS
Parameter
Total Coliforms
Turbiditv
Giardiad
Proposed11
MCLG
0/100 mL
zero
Proposed11
MCL
b
0 S NTTJC
1,000 fold removal
Pathogenic
Viruses" zero
Heterotrophic
Plate Count (HPC)e
Legionella zero
and/or inactivation
10,000 fold removal
and/or inactivation
compliance with filtration and
disinfection criteria will ensure
HPC control.
compliance with filtration and
fection requirements will
remove and/or inactivate.
a) U.S. EPA, 1987a.
b) Proposed MCL is based on presence or absence of coliforms in a sample,
rather than on density. No more than 5 percent of the most recent 60
consecutive samples can be coliform-positive for systems analyzing fewer
than 60 samples/year (3,300 persons or fewer), and no more than 5
percent of the total number of samples analyzed in the most recent 12-
month period can be coliform-positive for systems analyzing at least 60
samples/year (3,300 persons or more). In addition, no more than one
sample/month if less than 40 samples are collected per month, or 5% of
the samples/month if 40 or more samples/month are collected, can be
coliform-positive.
c) NTU = Nephelometric Turbidity Unit; filtered water turbidity. States may
allow less stringent levels (<, I NTU) for systems using ozonation at CT
values achieving 99.9 percent inactivation of Giardia cysts. All systems
are expected to optimize treatment so as to achieve the lowest turbidities
feasible at all times of the year.
d) Existing analytical methods are not considered to be technically and
economically available for use in compliance monitoring. Consequently, a
treatment technique regulation is proposed.
e) Former terminology - Total Plate Count
Systems with very clean and i protected source waters (i.e., low total coliform
or fecal coliform levels and low turbidity levels as specified in the Rule)
would only be required to use disinfection to achieve the required degrees of
inactivation of Giardia cysts and viruses. If such systems are continually
8
-------
able to meet certain disinfectant "CT" requirements specified in the Rule, the
water system can be assumed to be in compliance with the above inactivation
requirements for Giardia and viruses without having to monitor for these
organisms.
Systems with no potential sources of human enteric viruses within the
watershed (at the discretion of the State) would not be subject to the
requirement for 99.99 percent removal and/or inactivation of enteric viruses.
It should be noted, however, that this provision will only benefit systems
using chloramination for primary disinfection, since other disinfectants (e.g.,
ozone, chlorine, chlorine dioxide), which achieve 99.9 percent inactivation of
Giardia cysts also achieve much greater than 99.99 percent inactivation of
enteric viruses at the same time.
Systems required to filter could use a variety of technologies to meet the
minimum 99.9 and 99.99 percent performance levels. To satisfy this
requirement, systems would be required to meet certain turbidity removal and
disinfection performance criteria, and comply with design and operating
criteria specified by the State.
3. Appropriate Treatment Technologies
Conventional treatment (which includes coagulation, flocculation,
sedimentation, rapid granular filtration, and disinfection) has been
demonstrated to achieve at least 99.9 percent removal and/or inactivation, of
Giardia cysts and 99.99 percent removal and/or inactivation of enteric
viruses, under appropriate design and operating conditions (U.S. EPA, 1987a).
Filtration systems without disinfection, but with proper pretreatment - (to
produce water having turbidity levels < 0.5 NTU), can be assumed to provide
99 to 99.9 percent inactivation of Giardia lamblia and 90 to 99.9 percent
removal of viruses. Disinfection is needed to supplement filtration so that
the overall treatment achieves greater than 99.9 percent removal and/or
inactivation of Giardia cysts and 99.99 percent removal and/or inactivation of
viruses.
EPA considers conventional treatment with disinfection to be the Best
Available Technology (BAT) for most source waters in the United States
because of the multiple barriers of protection that it provides.
Direct filtration (which includes coagulation), slow sand filtration, and
diatomaceous earth filtration, each with disinfection, have been demonstrated
to achieve at least 99.9 percent removal and/or inactivation of Giardia cysts
and 99.99 percent removal and/or inactivation of viruses under appropriate
design conditions (U.S. EPA, 1987b). The Surface Water Treatment Rule will
allow their use under certain source water quality conditions as determined
by the State.
The same comments about the effectiveness of the filtration step to remove
Giardia cysts and viruses, arid the need for supplemental disinfection, apply
equally to direct filtration, slow sand filtration, and diatomaceous earth
filtration.
-------
Filtration technologies other than those specified above can be used if it is
first demonstrated through pilot plant challenge studies, using Giardia cysts
and viruses, or equivalent indicators, that the filtration technology, in
combination with disinfection, can achieve at least 99.9 percent and 99.99
percent removal and/or inactivation of Giardia cysts and viruses,
respectively.
4. Turbidity Limitations
For systems using conventional treatment or direct filtration, the proposed
Surface Water Treatment Rule requires that filtered water turbidity be £ 0.5
NTU in 95% of the measurements taken every month. Other levels could be
specified by the State after on-site demonstration that effective
removal/inactivation of Giardia lamblia cysts, or effective removal of Giardia
lamblia cyst-like particles can be obtained at other filtered water turbidity
levels. For example, the State could allow less stringent turbidity
performance criteria (up to 1.0 NTU) for systems demonstrating effective
Giardia removal by pilot plant studies. However, in such cases, the proposed
rule will require that the maximum filtered water turbidity level be <, 1 NTU
in 95% of the measurements taken each month, and at no time exceed 5
NTU.
All systems will be expected to optimize their treatment so as to achieve the
lowest turbidities feasible at all times of the year. This would promote
optimal removal of Giardia cysts and other pathogens, and provide optimum
conditions for disinfection.
For systems using slow sand filtration, the filtered water turbidity must be <
1 NTU in 95% of the measurements taken each month, and at no time exceed
5 NTU. However, the State could allow a turbidity value greater than 1
NTU in 95% of the measurements if the filter effluent at the plant prior to
disinfection consistently meets the long-term coliform MCL as required for
the disinfection system (see footnote b, Table I).
For systems using diatomaceous earth filtration, the filtered water turbidity
must be < 1 NTU in 95% of the measurements taken each month, and at no
time exceed 5 NTU.
For systems using other filtration technologies, the performance criteria
would be the same as for conventional treatment and direct filtration. The
State could allow a filtered water turbidity value greater than 0.5 NTU in
95% of the measurements taken (at no time exceeding 5 NTU) if the system
demonstrates effective performance at such levels.
5. Minimum Disinfection Requirements
a. For System,? Which Do Not Filter
At a minimum, a surface water treatment system, or a groundwater treatment
system "directly influenced by surface water" (to be determined by the
10
-------
State), and which does not filter will be required to provide disinfection
operational conditions which inactivate 99.9% inactivation of Giardia cysts
and 99.99% inactivation of enteric viruses. Both of these levels of
disinfection can be attained by application of chlorine, chlorine dioxide, and
ozone, according to the "CT values listed in Table IIA.
Table IIA. CT VALUES FOR ACHIEVING 99.9% INACTIVATION OF
Giardia lambliaa (U.S. EPA, Nov. 3, 1987)
Temperaturr
Disinfectant pH 0.5°C 5°C 10°C 15°C 20°C 25°C
Free Chlorine'3
(2mg/L)
6
7
8
9
171
261
377
521
122
186
269
371
91
140
201
279
61
93
134
186
46
70
101
139
30
47
67
93
Ozone 6-9 4.5 3 2.5 2 1.5 1.0
Chlorine Dioxide 6-9 81 54 40 27 21 14
Chloraminesc 6-9 3,800 2,200 1,850 1,500 1,100 750
(preformed)
a These CT values for free chlorine, chlorine dioxide, and
ozone simultaneously will guarantee greater than 99.99%
inactivation of enteric viruses.
b CT values will vary depending on concentration of free
chlorine. Values indicated are for 2.0 mg/L of free chlorine.
CT values for different free chlorine concentrations are
specified in tables in the Guidance Document (U.S. EPA,
1987d).
c To obtain 99.99% inactivation of enteric viruses with
preformed chloramines requires CT values > 5,000 at tempera-
tures of 0.5, 5, 10, and 154C.
To avoid the filtration requirement, EPA's proposed Surface Water Treatment
Rule will require that a system practice disinfection and have backup
disinfection capability provided by backup components, including an auxiliary
power supply, with automatic startup (and alarm), to ensure that continuous
disinfection is provided.
The unfiltered system also will be required to demonstrate by monitoring
certain disinfection parameters that it is achieving disinfection operational
conditions which inactivate 99.9 percent of Giardia cysts and 99.99 percent
of enteric viruses on a daily basis. To make this demonstration without
resorting to isolating and counting cysts and viruses, the system must
monitor and report the disinfectants) used, disinfectant residual concentra-
tion^), disinfection contact time(s), pH, and water temperature. The system
11
-------
must apply these data to determine if their "CP value exceeds that required
by the Rule for the specific disinfectant or combination of disinfectants to
determine if the minimum percent inactivations required are being achieved.
A subset of these "CT" values for the various disinfectants to provide 99.9
percent inactivation of Giardia lamblia cysts is shown in Table IIA. These
values are based on laboratory studies, and include appropriate safety
factors. Since Giardia cysts are much more resistant to free chlorine,
ozone, and chlorine dioxide than are enteric viruses, it can be assumed that
if a system achieves 99.9 percent inactivation of Giardia cysts using these
disinfectants, it will achieve much greater than 99.99 percent inactivation of
enteric viruses simultaneously.
On the other hand, enteric viruses are significantly more resistant to
chloramine residuals than are Giardia cysts. Systems using chloramination
for primary disinfection, therefore, will have to demonstrate that they are
achieving 99.99 percent of inactivation of enteric viruses as well as 99.9
percent inactivation of Giardia cysts. On-site pilot plant challenge studies,
using seeded organisms, will be needed to determine whether lower CT values
for chloramines achieve the required levels of inactivation. Such studies
require a high level of expertise to carry out, and specialized independent
(commercial) laboratories or university research personnel to make such
determinations.
For the purpose of calculating CT values, disinfection contact time, in
minutes (for chlorine, chlorine dioxide, or monochloramine) is the time
required for the water to move between the point of disinfection application
and the first customer. For ozone, because of its short half-life in water,
disinfection contact time must be defined as the time the water is exposed
to a continuous ozone residual concentration during the water treatment
process.
Residual disinfectant concentration (in mg/L) for chlorine, chlorine dioxide,
or chloramine is the concentration of the disinfectant at a point in the
distribution system before or at the first customer. Contact time in
pipelines must be calculated based on "plug flow" (i.e., all water moves
homogeneously in time between two points) by dividing the internal volume
of the pipeline by the peak hourly flow through that pipeline. Contact time
within mixing basins and storage reservoirs must be determined by tracer
studies or an equivalent demonstration.
For ozone, the residual of disinfectant concentration probably will be defined
as that ozone residual measured between two specific points in the treatment
process. These two points of ozone residual measurement will also define
the contact time for purposes of calculation of CT-values.
If disinfectants are applied at more than one point, only the percent
inactivation of all disinfectant sequences prior to the first customer may be
considered in the determination of total percent inactivation. In making this
determination, the disinfectant residual of each disinfection sequence and
corresponding contact time must be measured before subsequent disinfection
application point(s).
12
-------
If multiple disinfectants are used (e.g., primary disinfection with ozone,
followed by later secondary disinfection with chloramines or chlorine), the
percent inactivation achieved by each of the disinfectants is additive and
would apply in determining the overall disinfection performance provided.
b. For Systems Which Filter
For systems which filter, and which produce a filtered water turbidity < 0.5
NTU, it may be assumed that 99 to 99.9% removal of Giardia cysts and 90 to
99% removal of viruses is attained by the filtration system. EPA
recommends that systems credit filtration systems with 2-logs of Giardia
inactivation, but only one-log of virus inactivation. Therefore, disinfection
will be required to provide an additional 90% inactivation (1-log inactivation)
of Giardia cysts and 99.9% inactivation (3-logs inactivation) of viruses.
Table IIB lists CT values necessary to achieve 90 percent inactivation of
Giardia lamblia cysts. With the exception of chloramines, where higher CT
values than those indicated in Table IIB might be needed, these conditions
will achieve greater than 99.99% inactivation of viruses simultaneously.
If multiple disinfectants are used, the percent inactivation achieved by each
of the disinfectants is additive and will apply in determining the overall
disinfection provided. The clarification/filtration processes used, the degree
of fecal contamination in the water at the point(s) of primary disinfection,
and the formation of disinfection by-products (currently trihalomethanes), will
influence the type(s) of disinfectant used and the level of CT value(s).
Under the proposed SWTR, the entire treatment train must achieve a
minimum overall removal and/or inactivation of 99.9% for Giardia cysts and
99.99% removal of enteric viruses, respectively.
If a water treatment system has demonstrated that filtration is achieving
greater than a 2-log removal of Giardia cysts, the Primacy Agency may
permit the disinfection requirement to be reduced to the Ct requirements for
0.5 log reduction of Giardia cysts -- but not for systems using direct
filtration!. CT values which provide 0.5-log reduction of levels of Giardia
cysts simultaneously provide greater than 4-logs reduction in levels of
viruses, but only with ozone, chlorine and chlorine dioxide. Table IIC lists
CT values for attaining 0.5-log inactivation of Giardia cysts.
Based upon raw water total coliform levels at the point of disinfection, EPA
recommends that for systems which filter, sufficient disinfection be provided,
in addition to the 2-logs of inactivation of Giardia cysts and one-log of
inactivation of viruses credited to the filtration system, the following levels
of additional disinfection be provided :
13
-------
Table IIB. CT VALUES FOR ACHIEVING 90% INACTIVATION OF
Giardia lambliaa (U.S. EPA, 1987a,d)
Temoerature
Disinfectant
Free Chlorine0
(2 mg/L)
pH
6
7
8
9
0.5°C
60
90
130
170
5°C
40
60
90
120
10°C
30
40
60
90
15°C
20
30
50
60
20°C
15
23
34
46
25°C
10
15
22
31
Ozone
Chlorine Dioxide
Chloraminesc
(preformed)
6-9
6-9
6-9
1.5
27
1,270
1
17
730
0.8
13
620
0.7
9
500
0.5
7
366
0.3
5
260
a
These CT values for free chlorine, chlorine dioxide, and
ozone simultaneously will guarantee greater than 99.99%
inactivation of enteric viruses.
CT values will vary depending on concentration of free
chlorine. Values indicated are for 2.0 mg/L of free chlorine.
CT values for different free chlorine concentrations are
specified in tables in the Guidance Document (U.S. EPA,
1987d).
To obtain 99.99% inactivation of enteric viruses with
preformed chloramines requires CT values > 5,000 at
temperatures of 0.5, 5, 10, and 15°C.
Raw Water Total
Coliforms *
(#/100 mL)
< 100
< 500
< 1,000
< 5,000
< 10,000
Giardia Cyst
Inactivation
(logs)
1.0
1.5
2.0
2.5
3.0
Enteric Virus
Inactivation
(logs)
3.0
3.5
4.0
4.5
5.0
at the point of disinfection
14
-------
Table IIC. CT VALUES FOR ACHIEVING 0.5-LOG INACTIVATION
OF Giardia lambliaa (U.S. EPA, 1987d)
Temoerature
Disinfectant pH 0.5°C
Free Chlorineb
(2 mg/L)
Ozone
Chlorine Dioxide
Chloraminesc
(preformed)
6
7
8
9
6-9
6-9
6-9
28
43
63
87
0.8
13
690
5°C 10°C 15°C 20°C 25°C
20
31
45
62
0.5
9
363
15
23
34
46
0.4
7
337
10
15
22
31
0.4
4.5
250
8
12
17
23
0.3
3
181
5
8
11
15
0.2
2
130
a These CT values for free chlorine, chlorine dioxide, and
ozone simultaneously will guarantee greater than 99.99%
inactivation of enteric viruses.
b CT values will vary depending on concentration of free
chlorine. Values indicated are for 2.0 mg/L of free chlorine.
CT values for different free chlorine concentrations are
specified in tables in the Guidance Document (U.S. EPA,
1987e).
c To obtain 99.99% inactivation of enteric viruses with
preformed chloramines requires CT values > 5,000 at tempera-
tures of 0.5, 5, 10, and 15*C.
IV. EPA DISINFECTION BY-PRODUCTS REGULATIONS
A. STATUTORY REQUIREMENTS
Section 1412(b)(3)(A) requires the EPA Administrator to publish Maximum
Contaminant I^evel Goals (MCLGs) and promulgate national primary drinking
water regulations for each contaminant which, in the opinion of the
Administrator, may have any adverse effect on the health of persons, and
which is known or anticipated to occur in public water systems. Not later
than January 1, 1988, and at 3-year intervals thereafter, the Administrator
must publish a list of contaminants which are known or anticipated to occur
in public water systems and which may require regulation under this Act. It
is this Section under which disinfection/oxidation by-products will be
identified and regulated by the EPA, although the specific wording also
includes synthetic organic chemicals (SOCs) which are not produced during
the water disinfection process.
15
-------
These regulations will be subject to the monitoring requirements of Section
1445(a) of the Safe Drinking Water Act Amendments, which have been
discussed earlier.
B. CANDIDATE DISINFECTANT/OXTOANT BY-PRODUCTS TO BE
REGULATED
At present, EPA is regulating only the four trihalomethanes as DOBs. The
total concentration of these four compounds currently is regulated at 100
jug/L. Other compounds will be listed for regulation by EPA as their health
effects become known.
In early 1988, the Environmental Protection Agency (U.S. EPA, 1988a)
promulgated the First Drinking Water Priority List, which contains 53
substances or classes of substances that are candidates for regulation. These
include the above disinfectants, ammonia, the four THMs currently regulated,
ozone by-products, cryptosporidium. and many additional halogenated organic
materials, some of which are by-products of chlorine disinfection. EPA
currently is conducting research on health effects, analytical chemistry, and
treatability for most of these compounds. Within three years of publication
of this first Drinking Water Priority List (by January 22, 1991), EPA is
required to promulgate regulations for 25 of the substances listed.
Ten operating water utilities were surveyed recently by EPA scientists
(Stevens et al., 1987a,b,c). Some 200 discrete DOBs present in drinking
water and produced by chlorination were isolated, and many have been
identified. Although many of these DOBs are chlorinated, many are
oxygenated with no chlorine atoms in their molecular structures.
Table IV lists the frequency of occurrence in the 10-utility survey of 22
chlorinated by-products of greatest current interest to EPA's Office of
Drinking Water (Stevens et al., 1987a,b,c). Other disinfectants/oxidants can
be expected to produce corresponding lists of compounds which may be
proposed for regulation as DOBs in the future. Many of the compounds
listed in Table IV have been proposed for possible regulation (see Table III).
Table V summarizes the currently known health effects associated with
several types of chlorination by-products (Akin et al., 1987). Much of the
information learned about disinfection by-products in drinking water has
been inferred from studies of chlorination of naturally occurring organic
materials present in source waters.
Studies on the toxicity of the readily identifiable compounds are continuing
with the recognition that there is no reliable way to determine whether all
substances found to possess some biological activity are being identified.
Because the concentrations of disinfection by-products in drinking water are
low, it may not be possible to detect any biological activity directly. Thus
it is not possible to determine if efficient recovery of biologically active
substances is being achieved by the standard or advanced isolation and
identification methods that have been investigated for this purpose.
16
-------
Table III. DISINFECTANTS AND DISINFECTANT BYPRODUCTS
LISTED IN THE FIRST DWPL (U.S. EPA, 1988a)
Disinfectants:
Chlorine, Hypochlorite Ion, Chlorine Dioxide, Chlorite Ion,
Chlorate Ion, Chloramine, and Ammonia
Four Trihalomethanes
Chloroform, Bromoform, Bromodichloromethane, Dichlorobromo-
methane,
Haloacetonitriles:
Bromochloroacetonitrile, Dichloroacetonitrile, Dibromoacetonitrile,
Trichloroacetonitrile
Halogenated Acids. Alcohols. Aldehydes, Ketones. and Other Nitriles
Others:
Chloropicrin (trichloronitromethane), Cyanogen Chloride, Ozone
Byproducts
However, after chlorination, mutagenic activity can be determined directly in
solutions containing about 2 g/L of humic acid. Thus recovery of mutagenic
activity during subsequent extraction and fractionation can be followed
unambiguously. From these types of studies, it has been found that the
total mutagenic activity of the compounds in those samples that could be
identified by GC/MS and for which sufficient quantities were available for
Ames testing, accounted for only a few percent of the total activity
observed in the solution. Obviously some biologically active substances were
produced that were not being extracted and characterized.
Extensive studies of the by-products of ozonation, chloramination, and
treatment with chlorine dioxide have not yet been conducted (Akin, 1987).
However, it is well-known that the by-products of ozonation, for example,
are more highly oxygenated and polar in nature than the starting materials.
However, even fewer of these by-products of ozonation have been identified
than in the case of chlorination. Currently, studies are underway to
determine if ozonized, concentrated humic acid solutions are mutagenic, or
produce any effects in subchronic bioassays in rodents. These same solutions
will have to be tested for biological activity after subsequent treatments
with either chlorine, monpchloramine, or chlorine dioxide, because in actual
practice, ozonation applied as the primary disinfectant after filtration
normally is followed by one of these secondary disinfectants to provide a
bacteriostatic residual for the distribution system.
17
-------
Table IV. OCCURRENCE OF CHLORINATED DISINFECTION BY-
PRODUCTS AT 10 WATER UTILITIES (Stevens et al,
1987a,b,c)
~ Number of Range of Values ~
Compound Locations (ug/L)
High Confidence
Chloroform 10 of 10 2.6 to 594
Bromodichloromethane 10 of 10 2.6 to 77
Chlorodibromomethane 10 of 10 0.1 to 31
Bromoform 6 of 10 0.1 to 2.7
Dichloroacetonitrile 10 of 10 0.2 to 9.5
Dibromoacetonitrile 3 of 7 0.4 to 1.2
Bromochloroacetonitrile 7 of 7 0.2 to 4.0
Chloropicrin 8 of 10 0.2 to 5.6
Low Confidence
Chloroacetic Acid 6 of 10 < 10
Dichloroacetic Acid 10 of 10 < 10 to > 100
Trichloroacetic Acid 6 of 10 10 to 100
Trichloroacetaldehyde
(as Chloral Hydrate) 10 of 10 10 to 100
1,1,1-Trichloropropanone 10 of 10 10 to 100
2-Chlorophenol 0 of 10
2,4-Dichlorophenol 0 of 10
2,4,6-Trichlorophenol 0 of 10
Qualitative Only
1,1-Dichloropropanone 0 of 8
l,l-Dichloro-2-butanone 0 of 8
3,3-Dichloro-2-butanone 1 of 8
l,l,l-Trichloro-2-butanone 0 of 8
Cyanogen Chloride 1 of 7
Dichloroacetaldehyde 0 of 10
C. ANTICIPATED TIMING OF REGULATIONS
Before DOBs can be proposed for regulation, they must be identified as being
present in drinking water supplies and produced during the disinfection (or
oxidation) process, isolated, and their health effects determined. At present,
only the four THMs are being regulated as disinfection by-products. The
current MCL for TTHMs is 100 /ng/L. However, the National Academy of
Sciences (1987) has recommended that this level be lowered. EPA regulators
concur in this thinking, and plan to propose a new THM standard when the
Groundwater Treatment Rule is proposed in late 1990 or early 1991 (U.S.
EPA, 1987a; 1987d).
18
-------
Table V. SUMMARY OF HEALTH EFFECTS ASSOCIATED WITH
CHLORINATION BY-PRODUCTS (Akin et al, 1987)
Chemical
Class
Example
Toxicological
Effects*
Trihalomethanes
Haloacetonitriles
Haloacid
Derivatives
Chlorophenols
Chloroform
Dichlorobromomethane
Dibromochloromethane
Bromoform
Chloroacetonitrile
Dichloroacetonitrile
Trichloroacetonitrile
Bromochloroacetpnitrile
Dibromoacetonitrile
Dichloroacetic Acid
Trichloroacetic Acid
2-Chlorophenol
2,4-Dichlorophenol
2,4,6-Trichlorophenol
C, H, RT
H, RT
H, RT
H, RT
G, D
M, G, D
G, D
M, G, D
G, D
MD, C, N, OL, A
HPP
F, TP
F, TP
C
Chlorinated
Ketones
Chlorinated
Furanones
Chlorinated
Aldehydes
1,1-Dichloropropanone
1 , 1 , 1-Trichloropropanone
1 , 1,3,3-Tetrachloropropanone
MX
2-Chloroacetaldehyde
M
M
M
M, Cl
G
C = Carcinogenic; H = Hepatotoxic; RT = Renal Toxic; G =
Genotoxic; D = Developmental; M = Mutagenic; MD =
Metabolic Disturbance; N = Neurotoxic; OL = Ocular
Lesions; A = Aspermatogenesis; HPP = Hepatic Peroxisome
Proliferation; F = Fetotoxic; TP = Tumor Promoter; Cl =
Clastogenic
From the number of halogenated compounds listed for possible regulation in
the First Drinking Water Priority List (Table III), from the 22 chlorinated
compounds listed in Table IV, and from the nearly 200 chlorination by-
products identified in the 10-utility survey (Stevens et al., 1987a,b,c), it is
apparent that most of EPA's current concerns are for by-products of
chlorination.
19
-------
V. BEST AVAILABLE TECHNOLOGIES AND TREATMENT TECHNIQUES
A. STATUTORY REQUIREMENTS
The SDWA Amendments require the EPA Administrator to specify MCLs for
each contaminant to be regulated at a level which is as close to the MCLG
as is feasible with the use of best available technology, treatment
techniques, and other means which the Administrator finds, after examination
for efficacy under field conditions, are available (taking cost into
consideration) [Section 1412(b)(4)]. The Administrator also is required to list
the Best Available Technologies (BAT) which are capable of meeting MCL
regulations.
Granular Activated Carbon (GAC) technology is defined by the Safe Drinking
Water Act Amendments of 1986 as being feasible for the control of synthetic
organic chemicals, and any technology, treatment technique, or other means
found to be best available for the control of synthetic organic chemicals
must be at least as effective in controlling synthetic organic chemicals as
GAC [Section 1412(b)(5)].
At first glance, this last paragraph may not appear to be applicable to this
discussion of disinfectant/oxidant by-products. However, primary disinfection
usually is practiced after all filtration steps, including after GAC adsorption.
Many of the organic precursors of DOBs will be adsorbed by the GAC; thus
when disinfection now is practiced, lower quantities of DOBs will be
produced.
Conversely, the practice of preoxidatipn, using ozone, KMnO4, C1O2, or even
chlorine, for the purposes of oxidation of color, iron and manganese, , taste
and odor materials, microflocculation, or of synthetic organic chemicals, will
provide at least partial disinfection, according to the CT values cited in
Tables HA and HB, and depending on the total coliform levels at the point
of preoxidation/predisinfection. In some cases, particularly those involving
ozone, chlorine dioxide, and chlorine at the lower pH value, it is possible
that the primary disinfection requirement may be satisfied at this point in
the water treatment process (prior to filtration). This is especially true in
the case of ozone, for which the CT values necessary to provide 99.9%
inactivation of Giardia cysts and greater than 99.99% enteric viruses range
from 1 to 4.5 mg/L-min (see Table IIA).
If the water treatment process is modified to include the addition of strong
oxidants/disinfectants ahead of filtration (including GAC filtration/adsorp-
tion), the chemical nature of many of the organic contaminants (VOCs and
SOCs) may be altered, thus promoting their biodegradation in the filters and
changing their GAC adsorption characteristics. At the very least, even
partial oxidation of those easily oxidized VOCs and SOCs will decrease the
adsorptive loading on the GAC filters advantageously.
The Administrator also is .authorized [Section 1412(b)(7)(A)] to promulgate a
national primary drinking--* water regulation that requires the use of a
treatment technique in lieu of establishing an MCL, if the Administrator
makes a finding that it is not economically or technologically feasible to
20
-------
ascertain the level of the contaminant. In such case, the Administrator shall
identify those treatment techniques which, in the Administrator's judgement,
would prevent known or anticipated adverse effects on the health of persons
to the extent feasible. Such regulations shall specify each treatment
technique known to the Administrator which meets the requirements of this
paragraph, but the Administrator may grant a variance from any specified
treatment technique.
B. DEVELOPMENT OF CRITERIA FOR BEST AVAILABLE
TECHNOLOGIES
EPA's overall approach to the development of BAT criteria currently includes
the following five major steps (U.S. EPA, 1987c):
1. Identification of all candidate treatment technologies, or combinations of
treatment technologies;
2. Initial screening of technologies;
3. Identification of available treatment technologies (or combinations of
treatment technologies) for major contaminant categories and
subcategories;
4. Evaluation of available technologies (or combinations of treatment
technologies) in each category;
5. Determination of BAT for each major contaminant category or
subcategory.
C. BAT FOR DISINFECTION
The following processes for removing or inactivating microbiological
contaminants already have been identified by EPA during the initial
screening of technologies:
1. Coagulation Sedimentation/Filtration
2. Direct Filtration
3. Pressure Diatomaceous Earth Filtration
4. Slow Sand Filtration
5. Cartridge Filtration
6. Disinfection/Oxidation
7. Ultraviolet Radiation
Evaluation of these technologies by EPA is proceeding with the objective of
developing BAT Criteria Documents for each. UV radiation has been found
to be ineffective for providing 99.9% inactivation of Giardia lamblia cysts
(Rice & Hoff, 1981). Therefore its use as a primary disinfectant probably
21
-------
will be limited to groundwaters which are not "directly influenced by surface
waters".
Of the seven technologies listed, #1 (Conventional Treatment), #2, and #6
can be considered as well-known and in widespread practice. The other four
can be considered as "Emerging Technologies", even though slow sand filters,
for example, are utilized widely in other parts of the world. For example,
the United Kingdom relies extensively on this technology in both small and
large municipalities (Rachwal, et al., 1987).
D. BAT TECHNOLOGIES FOR DISINFECTION BY-PRODUCTS
At the present time, a BAT document exists only for THM removal (U.S.
EPA, 1981). Available or Potentially Available Treatment Methods described
in this 1981 document are grouped into the following three categories:
1. Best Generally Available Treatment Methods for Reducing (concentrations
of) TTHMs;
2. Additional Treatment Methods for Reducing (Concentrations of) TTHMs;
3. Granular Activated Carbon (GAC) and Biological Activated Carbon (BAG).
Under category 1 are the following Best Generally Available Treatment
Methods:
a. Use of chloramines as an alternate or supplemental disinfectant or
oxidant,
b. Use of chlorine dioxide as an alternate or supplemental disinfectant
or oxidant,
c. Improving existing clarification for THM precursor removal,
d. Moving the point of chlorination to reduce the amount of TTHM
formation and, where necessary, substituting for the use of chlorine
as a preoxidant, chloramines, chlorine dioxide, hydrogen peroxide, or
potassium permanganate,
e. Use of powdered activated carbon for reduction of concentrations of
THM precursors or TTHMs seasonally or intermittently at dosages
not to exceed 10 mg/L on an annual average basis.
Under Category 2 are the following Additional Treatment Methods for
Reducing (concentrations of) Trihalomethanes:
a. Off-Line Water Storage,
b. Aeration (after production of THMs),
c. Introduction of clarification where not currently practiced,
22
-------
d. Alternate source of raw water,
e. Ozone as an alternate or supplemental disinfectant or oxidant.
At least items a, c, and d are technologies which are well-known. Only
aeration and the use of ozone can be considered as "Emerging Technologies ,
even though the use of ozone in Europe has been extensive, particularly
since World War II.
Finally, GAC and BAG technologies, because of their higher capital costs and
lack of widespread adoption in the USA, both must be considered as
"Emerging Technologies".
E. STRATEGIES FOR CONTROLLING DISINFECTION BY-PRODUCTS
Currently, EPA has identified three distinct strategies for controlling
disinfection by-products:
1. Application of alternative disinfectants,
2. Removal of precursors prior to disinfection,
3. Removal of disinfection by-products once they are formed.
Initial screening of technologies has identified the following available
treatment processes or unit process steps for controlling disinfection by-
products, specifically chlorination by-products (THMs):
1. Oxidation
2. Coagulation, Sedimentation, Filtration
3. Direct Filtration
4. Lime Softening
5. GAC Adsorption (Precursor Removal)
6. PAC Adsorption
7. GAC Adsorption (By-Product Removal)
8. Packed Column Aeration
9. Diffused Aeration
Items #5, #7, and #8 are "Emerging Technologies", whereas the others
generally are well-known.
However, it is logical to expect that EPA may expand this list to include
combinations of treatment technologies for removing synthetic organic
contaminants, including disinfectant/oxidant by-products, such as:
10. Oxidation, Coagulation, Sedimentation, Filtration (Conventional
Treatment with Preoxidation)
23
-------
11. Oxidation, Direct Filtration (Direct Filtration with Preoxidation)
12. Oxidation, Slow Sand Filtration
13. Oxidation, GAC Adsorption
In at least half of the more than 2,000 water treatment plants throughout
the world using ozonation, for example (including more than some 35 of the
40 ozone plants currently operating in the U.S.A.), ozone oxidation is
practiced prior to conventional treatment or direct filtration. Preozonation
is practiced early in the treatment process for microflocculation, turbidity
removal, taste and odor control, color removal, iron and manganese
oxidation, oxidation of THM precursors, etc. In this manner, combinations
of oxidation followed by conventional treatment or direct filtration reduces
the concentrations of chlorine-demanding organic materials at the terminal
chlorination (currently the primary disinfection) step.
Although these process combinations are well-known throughout Europe, they
must be considered as "emerging" in the U.S. today.
In process combinations #12 and #13, the oxidation step is designed to
convert normally biorefractory organic materials into more readily
biodegradable materials. During the slow sand filtration or GAC adsorption
steps which follow oxidation, considerable biochemical conversion of dissolved
organic materials to CC>2 and water occurs.
In subsequent studies, EPA plans to evaluate the sensitivity of process design
criteria as functions of treatment performance and cost.
VL DISCUSSION OF DISINFECTANTS AND OXIDANTS
A. BASIC ISSUES
The current interest in regulating disinfectants and disinfectant/oxidant
by-products has been instigated by the recognition that when chlorine is
added for disinfection purposes, it also produces halogenated organic
materials. Their formation is a function of a number of factors, in
particular the concentration and types of organic materials present in the
water when chlorine is added, and the chlorinating conditions (amount of
chlorine dosed, pH, temperature, reaction time, etc).
Therefore, the salient objective of water treatment now is to reduce the
amounts of halogenated materials formed anywhere during the treatment
process. In turn, this has prompted the three strategies identified by EPA:
1. Remove the undesired compounds after they have been produced;
2. Change disinfectants tp those which do not produce the undesired
by-products;
24
-------
3. Reduce the concentrations of organics in the water before chlorine is
added.
Modifications of these same three strategies also can be considered as
applicable to coping with SOCs and their disinfection/oxidation by-products.
Of the three approaches, the first two can be criticized as disregarding the
problem and treating the symptom ... keep the water treatment process the
same, but either incorporate additional processing to remove the undesired
materials produced by water treatment, or switch to another disinfecting
material which will not produce the undesired materials.
Only the third approach recognizes and treats the problem itself, that is, the
concentration of organics in the water, and how to remove these
"precursors" of undesired by-products or lower their concentrations before
adding chlorine. If one assumes that technological logic will win the day
over partial solutions which only treat the symptoms, it can be expected that
most water utilities will opt to apply the third strategy, provided that it is
affordable.
1. First Approach
Once the halogenated organics have been produced, they are very difficult
and/or costly to remove from water. Oxidation is ineffective, even with
ozone, for removal of THMs. Adsorption by GAC is effective only for srjort
periods of time, and GAC reactivation costs are quite high. Adsorption by
powdered activated carbon (PAC) is costly and increases the quantities of
sludge for disposal. Air stripping of volatile organic chemicals (VOCs) is
efficient, but GAC may be required to adsorb the compounds (to prevent
transfer of the pollutant from water to air), and the same problems of
reactivation are present. Besides, after removal from the water, the VOCs
now must be disposed of, and most of them can be expected to be subject to
RCRA compliance under the appropriate hazardous waste disposal regulations.
2. Second Approach
Switching to an "alternative disinfectant" to chlorine (used as the primary
and secondary disinfectant) is a logical approach, provided that the
alternative disinfectant will:
(a) not produce halogenated organics,
(b) not produce other undesired disinfection by-products,
(c) provide guarantees of microbial disinfection or inactivation at least
equal to those provided by chlorine, and
(d) provide a stable residual disinfectant in the distribution system.
25
-------
For economic reasons, most utilities also would like the alternative
disinfectant to be no more expensive than chlorine. All of this, particularly
the "same cost" requirement, is asking a great deal of any single candidate
alternative disinfectant. Given the fact that chlorine has been applied as
the disinfectant of choice throughout the world since the early 1900s, if a
cheaper and better disinfectant were available, it should have been in
commercial use by this time.
Candidate "alternative disinfectants" currently being considered by many
water treatment specialists are: chlorine dioxide, monochloramine, ultraviolet
radiation, and ozone. However, both ozone and UV radiation can be rejected
as candidates because neither provides a stable residual level for distribution
systems. Miller et al. (1978), after studying European and Canadian
applications of ozone for disinfection, concluded that ozone can be used as
the secondary, residual disinfectant, but only if five conditions are met
simultaneously:
1. Water temperatures must be cool (to slow biological regrowths),
2. Water must be free of iron (to eliminate growth of iron-consuming
bacteria),
3. Water must be free of ammonia (to eliminate growth of nitrobacter
microorganisms),
4. TOC values must be less than 1 mg/L (to minimize its accumulation as
food for microorganisms),
5. Residence time in the distribution system must be less than 12 hours (so
that microbial regrowths will be minimal).
Even in France, where post-filtration ozonation for primary and some
secondary disinfection has been practiced since the early 1900s, all five
conditions were found by Miller et al. (1978) to coexist only rarely.
Chlorine dioxide can be used as the secondary disinfectant in distribution
systems, and in the water treatment plant for the primary disinfection step.
However, it must be realized that C1O2 is reduced when it disinfects and/or
oxidizes, producing some level of chlorite ion, C1O2", from which chlorine
dioxide is generated.
Chlorine dioxide produces hematological effects in both humans and
laboratory animals (NAS, 1987, pp. 89-90). The mechanism of these effects
is not known; however it is believed to be related to the oxidant properties
of chlorine dioxide and its aqueous reaction products, chlorite and chlorate
ions. In addition, thyroid and developmental neurological effects have been
observed in laboratory animals. These thyroid effects are thought to be
caused by its oxidation of dietary iodide ion in the gastrointestinal tract.
The oxidized iodine then binds to either food or tissue, and is unavailable
for absorption. The mechanisms of the neurological anomalies are unknown.
26
-------
Because of this behavior, EPA currently recommends that the total oxidant
levels C1O2 usage (chlorine dioxide + chlorite ion + chlorate ion) should not
exceed 1 mg/L. This means that usage of chlorine dioxide as the primary
disinfectant must be limited to rather clean waters, requiring very low
applied dosages, i.e., 1.2-1.4 mg/L.
In Germany, for example, CIO? is employed for distribution system residual
at dosage levels not to exceed t).3 mg/L (Miller et al, 1978). In Switzerland,
restrictions on the use of CIO? are even more rigid, 0.15 mg/L being the
maximum allowable dosage (Schalekamp, 1986).
In laboratory studies, Werdehoff and Singer (1986) have shown that the 1
mg/L total oxidants from chlorine dioxide will not be exceeded if the dosage
of CIO? does not exceed 1.2 to 1.4 mg/L. This level has been confirmed in
pilot plant and full-scale plant studies at Evansville, Indiana by Lykins and
Griese (1986) (see Case History of Evansville, IN ~ Section VIII.C.1).
By elimination, this appears to leave monochloramine as the only apparent
candidate "ideal alternative disinfectant" to chlorine. On the other hand,
monochloramine is a very much weaker disinfectant than chlorine (see Tables
IIA, IIB, IIC, and Section VLB on Technical Issues - Disinfection and
Disinfectants). Because of the very high CT values of monochloramine for
inactivating 99.9% of Giardia lamblia cysts and 99.99% inactivation of enteric
viruses (see Table IIA), it is impractical to consider its use as a primary
disinfectant for surface waters or groundwaters "directly influenced by
surface water". Therefore, monochloramine should be considered only as the
secondary disinfectant for these types of systems (U.S. EPA, 1987a,d).
3. Third Approach
Without much argument, except perhaps cost, the concept of reducing the
concentration(s) of organics in water before chlorine is added will produce
fewer halogenated organic DOBs. This approach addresses the basic problem
of removal of organics. A more restrictive subset of this approach is to
remove organics or lower their concentrations before any strong oxidant is
added. This is because the organic oxidation products of disinfectants/-
oxidants other than chlorine will not be well identified for some time, since
more sensitive analytical procedures and health effects data continue to be
developed.
A compromise approach is to employ non-chlorinous oxidants in the early
stages of water treatment (e.g., ozone, potassium permanganate, hydrogen
peroxide, before or during rapid mix, and/or prior to filtration) to assist in
the removal of organic materials by partial oxidation of the organics, more
efficient flocculation and filtration. In such cases, the non-chlorinous
oxidants do not produce halogenated oxidation products, except for those
waters which may contain substantial amounts of bromide ion (which can be
oxidized by ozone ~ as well as chlorine -- to hypobromous acid, which then
may produce brominated organics, such as bromoform — see Section VI.C,
Technical Issues - Disinfection By-Products).
27
-------
This third approach, although installed in many water treatment plants for
the purpose of lowering the concentrations of disinfection/oxidation by-
products, also is applicable to the simultaneous removal or reduction in
concentration of many SOCs and VOCs.
B. TECHNICAL ISSUES « DISINFECTION AND DISINFECTANTS
It is paramount that the disinfectant, applied for the purpose of disinfection
(as opposed to being applied for oxidation) be capable of providing the
required levels of microorganism kills or inactivations. In addition to
reducing levels of total cohforms, heterotrophic plate count organisms, and
legionella bacteria, disinfection currently is defined by EPA (U.S. EPA,
1987a) to mean a 99.9% reduction in Giardia lamblia cyst levels, and a 99.99%
reduction in enteric virus concentrations.
1. For Systems Treating Surface Waters
a. The "CT Value
Disinfection should be recognized as a [concentration x contact time]
phenomenon (C x T product). That is, the weaker the disinfectant, the
longer it must contact the water being disinfected. Therefore, one of the
consequences of changing disinfectants will be the appropriate lengthening or
shortening of the contact times, thus changing the sizes of disinfectant
contact chambers.
In EPA's proposed Surface Water Treatment Rule (U.S. EPA, 1987a), several
tables of CT values are provided as guidance for water treatment officials in
coping with the microorganisms to be regulated (see Table I). Table IIA
compares the CT values for the four major disinfectants over the pH range
of 6 to 9 at 0.5°C to 25°C with respect to their abilities to achieve 99.9%
inactivation of Giardia lamblia cysts: free chlorine (at 2 mg/L), ozone,
chlorine dioxide, and preformed chloramines.
Several important conclusions can be drawn from Table IIA:
1. Ozone is by far the most efficient disinfectant of the four (CT values of
4.5 at 0.5°C to 1 at 25°C). This means that a 0.5 mg/L residual ozone
concentration would have to be maintained over 9 minutes at 0.5°C
ranging down to 2 minutes at 25°C in order to guarantee 99.9%
inactivation of Giardia lamblia cysts and more than 99.99% inactivation of
virus inactivation.
2. Chlorine dioxide over the pH range of 6-9 is at least twice as effective
as free chlorine at pH 6 only. At pH 7, 8, and 9, free chlorine becomes
progressively less effective than chlorine dioxide.
3. A residual chlorine dioxide level of 1 mg/L would have to be maintained
for 81 minutes at 0.5°C ranging down to 14 minutes at 25°C in order to
guarantee 99.9% inactivation of Giardia lamblia cysts.
28
-------
4. A residual of 2 mg/L of free chlorine at pH 7 would have to be
maintained for 130 minutes at 0.5°C ranging down to 23.5 minutes at 25°C
in order to guarantee 99.9% inactivation of Giardia lamblia cysts. If the
water being disinfected also contains high trihalomethane formation
potentials, using chlorine as the primary disinfectant under these
conditions will produce high THM and other halogenated organic (see
Table IV) concentrations.
5. When CT values are attained with free chlorine, chlorine dioxide, and
ozone to inactivate 99.9% Giardia lamblia cysts, simultaneous attainment
of 99.99% inactivation of enteric viruses is assured.
6. Preformed chloramines are by far the weakest disinfectant of all. To
achieve 99.9% inactivation of Giardia lambiia cysts with a 2 mg/L residual
would require contact times of 1,650 minutes at 0.5°C ranging down to
375 minutes at 25°C.
7. Much longer contact times are required with preformed chloramines to
attain 99.99% inactivation of enteric viruses (CT values > 5,000 at 0.5-
15°C) than to attain 99.9% inactivation of Giardia lamblia cysts. This
means that attainment of the required 99.9% inactivation of Giardia
lamblia cysts with preformed monochloramines does not assure 99.99%
inactivation of enteric viruses.
This has led EPA to the recommendation that surface water systems wishing
to use monochloramine as their primary disinfectant will have to demonstrate
efficacy by conducting pilot scale tests with challenge organisms (Giardia
cysts and viruses) ~ a costly and time-consuming process. Thus, from a
practical and cost standpoint, monochloramine should not be considered as a
primary disinfectant (U.S. EPA, 1987a;d).
b. Suggested No Adverse Response Levels (SNARLs)
For some substances not regulated as known or suspected carcinogens and
for which there are adequate toxicity data available from prolonged ingestion
studies in humans or animals, the Subcommittee on Disinfectants and
Disinfectant By-Products of the Safe Drinking Water Committee of the
National Academy of Sciences has calculated Suggested No-Adverse-Response-
Levels (SNARLs). In a recent publication (National Academy of Sciences,
1987), SNARLs for adults and children are presented for chlorine dioxide,
chlorite ion, chlorate ion, and monochloramine. These specific values are
based on the assumption that 20% of the daily intake of these substances is
by ingestion of drinking water, and are summarized in Table VI.
Also included in Table VI are recommended SNARLs for some of the
halogenated disinfection by-products listed by the EPA in the first Drinking
Water Priotiry List for possible regulation (U.S. EPA 1988a).
29
-------
Table VI. NAS RECOMMENDED SNARLs FOR DISINFECTANTS
AND DISINFECTANT BY-PRODUCTS (NAS, 1987)
Disinfectant/By-Product for 70-kg adult for 10-kg child
Chlorine Dioxide 0.21 mg/L 0.06 mg/L
Chlorite & Chlorate Ions 0.024 mg/L 0.007 mg/L
Monochloramine 0.581 mg/L 0.166 mg/L
Dichloroacetonitrile 0.056 mg/L
Dibromoacetonitrile 0.161 mg/L 0.046 mg/L
The Subcommittee does not recommend this SNARL because
of concern that DCAN may be carcinogenic.
c. Possible Regulatory Consequences
Values for SNARLs recommended in Table VI are quite significant, by
themselves, but also when considered in light of the CT values for the
disinfectants listed in Table IIA. Considered by themselves, they are
indications of possible MCLs (or at least MCLGs) which EPA may propose
and promulgate in the future (i.e., when the Groundwater Treatment Rule is
proposed in 1991). If these SNARL values do become the basis for the
corresponding MCLs, it is also probable that EPA will set the MCLG and/or
MCL to protect the 70-kg adult, because exposure times would be calculated
over a human lifetime.
In this event, a projected MCL of 0.21 mg/L for chlorine dioxide may
eliminate this material from consideration as a primary disinfectant, except
in very clean waters, and will reduce its utility as a secondary disinfectant
(except in very clean waters).
At the same time, a projected MCL of 0.581 mg/L for monochloramine will
virtually eliminate this material from consideration as a primary disinfectant
because of the excessive contact times which would be required to attain
99.9% and 99.99% inactivations of Giardia lamblia cysts and enteric viruses,
respectively, at this concentration. For example, at 0.5°C, a contact time of
5,690 minutes (94 hours = 3.95 days) would be required to assure 99.9%
inactivation of Giardia lamblia cysts. At 25°C, a contact time of 1,293
minutes (21.6 hours) would be required. To attain 99.99% inactivation of
enteric viruses at 0.5-15°C would require a contact time with monochlor-
amine of > 8,620 minutes (> 143 hours = > 2.4 days).
The hypothesized 0.581 mg/L MCL for monochloramine would not eliminate
this material from consideration as a secondary disinfectant, however, since
EPA is likely to require only a "detectable residual" in the distribution
system when the Surface Water Treatment Regulation is promulgated during
1988. Under these projected regulatory restrictions, the concept of a single
30
-------
"alternative disinfectant" disappears, to be superseded by the concept of
"alternative disinfecting systems". Such alternative disinfecting systems
would involve the use of combinations of a primary disinfectant (e.g., ozone,
ultraviolet radiation) to be followed by a secondary (residual) disinfectant
(e.g., chlorine, chlorine dioxide, or chloramine) for the majority of cases.
For systems whose raw waters have very low THM formation potentials, or
in which THMFP can be lowered appropriately prior to disinfection, then
chlorine can continue to be both the primary and secondary disinfectant.
However, the luxury of continued application of chlorine as the
predisinfectant and as the primary and secondary disinfectant will be
dictated by the future THM regulations, as well as future regulations for
other halogenated organic by-products of chlorination (see next section).
1. EPA'S PERSPECTIVE ON RECOMMENDED NAS
SNARLS
In the proposed SWTR (U.S. EPA, 1987a), EPA rejects the NAS assumption
that only 20% of the daily exposure to chlorine dioxide and monochloramine
comes from drinking water. EPA expects that exposure to these
disinfectants from sources other than drinking water will be minimal.
Therefore, assuming that nearly 100% of exposure comes from drinking water,
the relative SNARLs might be estimated as 1 mg/L for chlorine dioxide
(including its oxidation/reduction products, chlorite and chlorate ions) and
2.5 mg/L for chloramine. This is the current recommendation that
concentrations of these two disinfectants do not exceed 1.0 and 2.5 mg/L,
respectively. These levels were proposed by EPA earlier (U.S. EPA, 1979).
It is also possible that EPA might apply a different uncertainty factor
(1,000) than used by the NAS (100) in calculating SNARLs (actually MCLs)
for chlorine dioxide, chlorite and chlorate ions, and monochloramine. The
significance of this possibility is illustrated in the following example SNARL
calculation for monochloramine:
8.3 mg/kg bw/day x 70 kg x 0.2 = 0.581 mg/L, or
581 ug/L
100 x 2 liters
where:
8.3 mg/kg bw/day = dosage in mg/kg of body weight per day
70 kg = average adult weight
0.2 = 20% of total daily intake
100 = uncertainty factor
2 liters = daily adult water consumption
31
-------
It can be seen that changing the estimate of percent of total intake from
0.2 to 1.0 multiplies the calculated SNARL by a factor of five. On the other
hand, changing the uncertainty factor from 100 to 1,000, decreases the
calculated SNARL by an order of magnitude (a factor of 10).
For example, using an uncertainty factor of 1,000, the NAS SNARL for C1O2
for adults becomes 0.021. When this figure is multiplied by 5 (to correct for
100% daily exposure rather than 20% exposure), the SNARL becomes 0.105
mg/L. Similar calculations for monochloramine gives a revised projected
SNARL of 0.29 mg/L.
To put these potential regulatory limits into perspective, the three
possibilities discussed above are summarized in Table VII for chlorine
dioxide, chlorite and chlorate ions, and monochloramine.
From this table, the range of highest and lowest potential MCL levels can be
projected, Variations A and B. In the proposed SWTR, EPA has rejected the
NAS-proposed values and is proposing values close to the higher (Variation
A) levels. However, further health effects studies are in progress, and these
issues are to be revisited by EPA in about a year. At that time, it is
possible that the current recommended maximum levels for chlorine dioxide
and monochloramine may be lowered.
TABLE VII. PROJECTED MCLs FOR CHLORINE DIOXIDE AND
CHLORAMINE
Disinfectant Projected MCLs Based on Variations in Adult
SNARLs
or By-Product June 87 NASa Variation Ab Variation Bc
chlorine dioxide
chlorite/chlorate
ions
0.21 mg/L
0.024 mg/L
1.05 mg/L
0.12 mg/L
0.105 mg/L
0.012 mg/L
monochloramine 0.581 mg/L 2.9 mg/L 0.29 mg/L
a 20% daily exposure; uncertainty factor = 100. The 20% daily
exposure level number has been rejected by the EPA in favor
of 100% (U.S. EPA, 1987a).
b 100% daily exposure; uncertainty factor = 100
c 100% daily exposure; uncertainty factor = 1,000
Therefore, the following possibilities appear with respect to chlorine dioxide
and monochloramine: •••"••
32
-------
1. If an MCL for chlorine dioxide is set at 0.10 mg/L (Variation B), chlorine
dioxide would be eliminated as a primary and probably as a secondary
disinfectant, except in very clean waters.
2. An MCL for chlorite/chlorate ions at either the highest or lowest value
given may eliminate chlorine dioxide from consideration as a primary
disinfectant.
3. For monochloramine, an MCL of 2.9 mg/L (Variation A) will allow this
material to be applied both as a primary and secondary disinfectant. In
fact, EPA's current proposal (U.S. EPA, 1987a;d) is to allow a maximum of
2.5 mg/L.
On the other hand, an MCL of 0.29 mg/L (Variation B) would eliminate
monochloramine from consideration as a primary disinfectant, but would
not eliminate monochloramine as a secondary disinfectant.
Under these assumptions of total human exposure from drinking water,
chlorine dioxide currently can be considered for primary and secondary
disinfection, within the constraints of the appropriate "CT values", and
monochloramine can be considered as a secondary disinfectant. However,
there is a strong probability that EPA will propose MCLs by 1991 (Ground
Water Treatment Rule) which will be lower than currently recommended
levels, and which will reduce the practicality of both disinfectants as
primary, and probably as secondary disinfectants as well.
2. For Systems Treating Groundwater
As pointed out earlier, many groundwater systems which do not now
disinfect will be required so to do, but the Groundwater Treatment Rule will
not be proposed by EPA until 1991. Since cysts normally are not present in
groundwaters (except in those groundwaters "directly influenced by surface
waters), requirements to inactivate these types of microorganisms are not
expected. However the other microorganisms listed in Table I (total
coliforms, pathogenic viruses, heterotrophic plate counts, and legionella
organisms can be expected to be present, and therefore to be regulated. If
viruses are regulated, the same 99.99% inactivation requirement can be
expected.
a. "CT" Values and SNARLs
It can be anticipated that CT values similar to those given in the Surface
Water Treatment Rule will apply to the various disinfectants, as well as the
SNARLs and/or the resulting MCLs/MCLGs. Therefore, the same reasoning
discussed above with respect to chlorine dioxide, chlorite and chlorate ions,
and monochloramine should apply to eliminate or reduce the viability of
these materials from consideration as primary and secondary disinfectants.
33
-------
b. Possible Regulatory Consequences
This reasoning leads to the same conclusions regarding disinfection of
groundwater systems as for surface water systems: in the majority of
systems, the application of a primary disinfectant (e.g., ozone, UV radiation)
will be necessary, followed by a secondary disinfectant (e.g., chlorine,
chlorine dioxide, or chloramine) for the "detectable residual" to be carried
throughout the distribution system.
As before, the ability to use chlorine as both the primary and secondary
disinfectant will rest on the future MCLs for THMs and for other by-
products of chlorination to be proposed by EPA in 1990 or 1991.
Ground water systems may face an additional complication, however, when
they are required to disinfect. Those ground waters which contain
significant levels of iron and manganese will produce insoluble iron and
manganese oxides when treated with ozone or chlorine. This will necessitate
the addition of filters to remove oxidized insoluble materials. In addition,
both ozonation and chlorination also may cause flocculation of dissolved
organics, and precipitate turbidity-causing particulates.
Formation of these types of insoluble materials will require secondary
disinfection after filtration.
One major advantage of UV radiation as the primary disinfectant for
groundwaters is that at appropriate applied energjy levels, bacterial disinfec-
tion occurs readily, without extensive oxidation. Therefore, for groundwater
systems with only traces of iron and manganese in their raw waters, use of
UV radiation followed by secondary chlorination might be the least-cost
alternative disinfecting system.
C. TECHNICAL ISSUES - DISINFECTION/OXIDATION BY-PRODUCTS
Although to date, EPA is regulating only the four THMs as disinfection by-
products, several additional halogenated organics have been listed in the first
Drinking Water Priority List (U.S. EPA, 1988a) for possible regulation. Also
included on the first DWPL are the three disinfectants themselves (chlorine,
chlorine dioxide, chloramine), chlorite and chlorate ions, ammonia, and
"ozone byproducts". This latter expression is intended as a catchall
category. In the event that specific byproducts of ozonation are identified
and shown to be detrimental to public health, they will be proposed for
regulation individually.
Akin et al. (1987) describe the current EPA Health Effects Research Program
dealing with compounds which have been identified as disinfection by-
products and for which health effects data currently are being developed.
All of those under current investigation are halogenated compounds which
are listed in Tables IV and V.
The major technological issue in controlling current DOBs is how to reduce
the concentrations of these halogenated organic materials without
34
-------
compromising microbiological safety. Some will add, "...and without
producing other by-products which are as potentially toxic (or of public
health concern) as those produced from chlorine". Still others will add, "...
and at no increase in costs .
1. Non-Halogenated DOBs
As has been concluded in an earlier discussion, only chlorine and ozone
(possibly UV radiation for groundwater) appear to be the major primary
disinfectants of the future for public water supplies. Ozone is known to
produce oxygenated materials as oxidation products. High energy UV
radiation also can produce oxidation products, provided that appropriate
dissolved oxygen and peroxide levels are present. What types of non-
halogenated oxidation products are produced by ozone and/or UV radiation,
and what is known of their toxicities? Are similar types of non-halogenated
oxidized organic materials produced from the use of chlorine?
a. From Ozonation (Rice & Gomez-Taylor, 1986)
Ozone generally is not reactive with saturated aliphatic hydrocarbon
compounds. However, other types of aliphatic compounds, particularly
olefins, are readily oxidized to unstable ozonides, peroxides, and
hydroperoxide intermediates to form primarily aldehydes, ketones, and acids.
A by-product of nearly all (if not all) organic oxidations with ozone, is
hydrogen peroxide. In the presence of additional ozone, however, the H2O2
is decomposed into hydroxyl free radicals, which are more powerful oxidizing
agents than the ozone molecule itself. Hydroxyl radicals are rapidly
destroyed by bicarbonate and carbonate ions, which comprise the natural
alkalinity of raw waters.
Although ozone is capable of oxidizing many organic compounds completely
to CO2 and water, this conversion usually requires large doses of ozone (> 3
moles O3/mole of organic compound) and long reaction times (sometimes
hours). Under drinking water treatment plant ozonation conditions (1 to 5
mg/L applied ozone dosage; 5 to 20 minutes contact time), organic
compounds usually are only partially oxidized. The oxidized organic
materials are more polar, of lower molecular weight, more biodegradable, and
usually more readily removed from solution by flocculation and filtration.
Oxalic acid (HOOC-COOH) is found commonly as a "final" oxidation product
of many aliphatic, aromatic, and heterocyclic organic materials, because of
its very slow rate of oxidation with ozone. Acetic acid is another "final"
oxidation product which also is resistant to further oxidation, even with
ozone, under drinking water treatment plant conditions.
Formic acid and formaldehyde also are formed during the later stages of
ozone oxidation of many organic materials. However, these two by-products
are readily oxidized to CO2 and water, provided sufficient ozone is present
in solution. Ozone oxidation rates of intermediate organic oxidation products
usually are slower than the oxidation rates of the original compounds.
35
-------
In a few instances, ozone oxidation products have been isolated which are
more toxic than the original compounds. For example, ozonation of the
pesticide heptachlor produces heptachlor epoxide, almost quantitatively.
Heptachlor epoxide is quite stable to continued ozonation (Hoffman &
Eichelsdorfer, 1971).
Parathion and malathion produce paraoxon and malaoxon, respectively, as
initial oxidation products upon ozonation. Both of the oxons are at least as
toxic than are the original thion pesticides. However, upon continued
ozonation, the intermediate oxons continue to oxidize, producing innocuous
final oxidation products (Laplanche et al., 1983).
Ozonation of phenol can produce resorcinol, which is a known THM
precursor. This may explain the observations made by several groups that in
following ozonation of some surface waters directly with chlormation, higher
levels of THMs have been produced than by chlorination alone (Rice, 1980).
In most cases, however, ozonation followed immediately by chlorination
produces lower concentrations of THMs.
Infrequent finding of higher THM levels after ozonation and immediate
chlorination has led to moving the initial ozonation point(s) to the rapid mix
or before, and just sometimes just prior to filtration, to allow flocculation
and even biodegradation of the more polar initial ozone oxidation products
during filtration. Chlorination then follows at a later treatment stage.
Under these conditions, lower THM levels always are found.
The National Academy of Sciences (1987) report (p. 67) notes that oxidation
products of ozone (as well as non-chlorinated organic oxidation products
from chlorine and chlorine dioxide) are similar to the organic compounds
formed by natural oxidation processes. In other words, a surface water
source such as a lake will be experiencing prebiological and chemical
oxidative processes for months, perhaps longer, by processes which are
similar in their chemistries to oxidation processes used in water treatment.
It is concluded by the National Academy of Sciences (1987, p. 195) that little
is known about the types of by-products produced by ozonation of natural
organics, and that well-conceived studies need to be conducted which will
focus on the stable compounds expected from ozonation reactions with humic
materials. Since saturated aldehydes (non-toxic) are well-known oxidation
products of organic materials, it is suggested (NAS, 1987) that particular
attention should be paid to the search for unsaturated aldehydes (some of
which are known toxicants) and for hydroxyhydroperoxides.
However, unsaturated aldehydes, if formed during ozone oxidation, should
oxidize rapidly at the double bond with ozone, thus destroying them.
Finally, the National Academy of Sciences (1987, p. 196) states:
"Notwithstanding the fact that these studies need to be carried out,
drinking water suppliers should not dismiss the possibility of using
ozone as an alternative to chlorine and chloramines in water
treatment. Ozone is an excellent disinfectant (although it must be
36
-------
used in combination with a secondary disinfectant to maintain a
residual in the distribution system); ozone is an excellent oxidant for
the various needs of water treatment; it does not form chlorinated by-
products; and the admittedly inadequate studies now available point to
lower toxicities of ozonated water than of chlorinated water."
b. From Chlorination
From the aqueous chlorination of various humic materials, Stevens et al.
(1985) identified 98 specific organic compounds, including 47 discrete
compounds containing chlorine and 51 discrete compounds that did not
contain chlorine. Christman et al. (1980) showed that the products of
chlorination of aquatic humic materials generally fall into three broad
structural categories: non-chlorinated substituted aromatics, chlorinated
straight chain acids, and non-chlorinated straight chain aliphatic acids.
Seeger et al. (1984) showed that at low chlorine doses, which are typical of
those conditions currently found in drinking water treatment plants and their
distribution systems, many ring-chlorinated aromatic acids are obtained, in
addition to the numerous nonchlorinated aromatic and aliphatic compounds
identified by earlier investigators.
In reviewing the available literature on products isolated and identified from
the treatment of humic and fulvic acids in aqueous solution with ozone,
chlorine, chlorine dioxide, and potassium permanganate, Rice & Gomez-Taylor
(1986) concluded that the non-chlorinated oxidation products formed by all
four oxidizing agents are similar.
More recently, Stevens et al. (1987a,b,c) isolated nearly 200 discrete
compounds from 10 operating water utilities practicing chlorine disinfection.
Of those compounds which were positively identified, 31 are non-chlorinated,
and are similar to those identified in other studies of ozonation of
humic/fulvic acids.
Therefore, consideration of ozonation as a water treatment agent should not
be postponed simply on the basis of the myriad of oxidation products which
are formed. Many of the same compounds are formed during chlorination.
D. COMPARISON OF DISINFECTANTS - OXIDANTS
1. General Considerations
The capability of one substance to oxidize another is measured by its
Oxidation Potential, normally expressed in volts of electrical energy
(referenced to the hydrogen electrode). The oxidation potential is a measure
of the relative ease by which a substance is able to lose electrons, thereby
being converted to a higher state of oxidation. If the oxidation potential of
substance A is higher than that of substance B, then substance B can be
oxidized by substance A. Oxidation potentials of representative
37
-------
oxidants/disinfectants commonly used or encountered in water treatment are
listed in Table VIII.
Table VIII. OXIDATION POTENTIALS OF WATER TREAT]
OXIDANTSa
Species
hydroxyl free radical
* ozone
hydrogen peroxide
permanganate ion
* hypochlorous acid
* chlorine
* hypobromous acid
* bromine
hypoiodous acid
* chlorine dioxide
* iodine
oxygen
Oxidation
(OH)-
(03)
(H202)
(MnO4')
(HOC1)
(C12)
(HOBr)
(Br2)
(HOI)
(C102)(aq)
(12)
(02)
Potential. Volts
2.80
2.07
1.76
1.68
1.49
1.36
1.33
1.07
0.99
0.95
0.54
0.40
a Source: Handbook of Chemistry & Physics, CRC Press, Inc.
* excellent disinfecting agents
Although the relative position of an oxidant in Table VIII is indicative of its
ability to oxidize other materials, its oxidation potential does not indicate
how fast one material will be oxidized by another, nor how far toward
completion the oxidation reaction will proceed. One cannot tell from
oxidation potentials alone whether a specific organic compound will be
oxidized completely (to CO2 and water) or only to the first of several
intermediate stages.
One significant fact can be learned from Table VIII, however, at this point.
As has been discussed in earlier sections of this report, it is rare that
organic compounds treated with an oxidant even as powerful as ozone will be
converted totally to CO2 and water, under conditions normally encountered
in water treatment plants. Therefore, no other commonly employed and less
38
-------
powerful water treatment oxidant (such as chlorine, bromine, chlorine
dioxide, etc.), all of which have lower oxidation potentials than ozone, will
oxidize an organic material completely to CO2 and water if ozone will not.
All oxidants weaker than ozone will be less effective than ozone in
converting organic compounds to carbon dioxide and water, and thus may
produce higher quantities of partially oxidized organic materials under water
treatment plant conditions. It is important, therefore, for the water
treatment professional to understand the chemistry of the organic com-
ponents of the particular water supply when considering the use of any
oxidant in the processing.
When oxidizing agents are added to water supplies containing specific
organic compounds as impurities, it is not sufficient simply to follow the
reduction in concentration of that compound. Other parameters, specifically
the total organic carbon (TOC), provide needed information. With chlorine,
the TOX (total organic halogen) also is a necessary parameter. In terms of
the newer halogenated compounds to be regulated by EPA, TOX Formation
Potential (TOXFP) will become important, particularly the non-purgeable
TOXFP (the NPTOXFP). This is a measure of the non-volatile TOXFP in the
water.
It is also desirable to know the products of oxidation of the organic
impurities which are being treated by the oxidizing agent. When a discrete
organic compound is oxidized, it may be totally destroyed by oxidation, but
without a decrease in TOC content, In this instance, the concentrations of
oxidation products may be as significant as the concentration of the original
impurity.
It is also important not to attempt to relate the disinfection capability of a
specific oxidant to its oxidation potential. Ozone is the most powerful
oxidant listed in Table VIII, and it is also the best disinfectant (i.e., its CT
values are less than for any other disinfectant for a given species of
microorganism). However hydrogen peroxide and potassium permanganate,
which follow ozone in Table VIII, are known to be poor disinfectants, while
iodine and chlorine dioxide, which have oxidation potentials less than half
that of ozone, are very good disinfectants.
Table VIII is useful to the water treatment specialist in understanding the
degree of chemical transformation which can be expected when various
oxidizing agents are used at various stages of the water treatment process.
Table IX lists data for the various disinfectants, comparing the dosages and
contact times required for 99% inactivation of Escherichia Coli. Table X
lists similar data for 99% inactivation of Poliovirus Type I.
2. Chlorine
This material is an excellent disinfectant, an excellent chemical oxidant, but
unfortunately also is an excellent chlorinating agent. Its production of
THMs is well-known, but it also produces a wide variety of additional
39
-------
halogenated compounds (TOX). In addition, chlorine is known to produce
significant quantities of non-halogenated oxidation products. Some of these,
from humic and fulvic acids for example, are identical to those produced by
potassium permanganate, ozone, and chlorine dioxide (Christman, et al.,
1980).
TABLE IX. COMPARISON OF DISINFECTION PERFORMANCE TO £,
COLI a
Disinfecting
Agent
HOC1
(OC1)-
NH2C1
ozone
ozone
C102
C102
iodine
iodine
H2O2
KMnO4
KMnO4
a From
Concen- Contact
tration Time,
mg/L min
0.1
1.0
1.0
0.04
0.0125
0.30
0.80
1.3
0.30
90
1
16
0.4
0.92
175.0
0.50
0.33
1.8
0.35
1
2
-360
45
25
CxTb
0.04
0.92
175.0
0.02
0.004
0.54
0.28
1.3
0.60
32,400
45
400
Drinkine Water and Health. Vol.
National Academy
Press, 1981
0), Chapter
Temperature
pH *C
6.0 5
10.0 5
9.0 5
7.2 1
7.0 12
7.0 5
7.0 25
6.5 2-5
9.1 20-25
6.5 ambient
5.9 0
9.2 20
2 (Washington, DC:
2.
b Concentration of disinfectant times contact time.
On the positive side, chlorine provides a stable residual for the water
distribution system, provided that the water is free of chlorine-demanding
ammonia and organic materials.
3. Chlorine Dioxide
This chemical is a powerful oxidant and disinfectant. Because of its
instability, it must be generated on-site, which can be considered a
disadvantage. In its pure state, C1O2 does not produce trihalomethanes in
the presence of organic materials which do produce THMs with chlorine.
Some procedures for synthesizing chlorine dioxide from sodium chlorite and
elemental chlorine involve the use of excess chlorine. Thus, in these
40
-------
instances, some free chlorine is present in the water, resulting in the
production of some THMs. The more that C1O2 can be synthesized without
the need for excess chlorine, the lower the concentrations of THMs which
will be produced by the excess chlorine. One procedure for generating C1O2
free of excess chlorine is by addition of mineral acid to solutions of sodium
chlorite.
TABLE X. COMPARISON OF DISINFECTION PERFORMANCE TO
POLIOVIRUS TYPE I a
Disinfecting
Agent
HOC1
HOC1
(OC1)-
(OC1)-
NH2C1
NH2C1
ozone
ozone
C1O2
C1O2
iodine
iodine
Concen-
tration
rng/L
0.5
1.0
0.5
1.0
10
10
0.042
<0.3
0.3
0.8
1.25
20
Contact
Time,
min
2.1
2.1
21
3.5
90
32
10
0.13
16.6
1.5
39
1.5
CxTb
1.05
2.1
10.5
3.5
900
320
0.42
<0.04
5.0
1.2
49
30
Tempera-
pH ture, °C
6.0
6.0
10.0
10.0
9.0
9.0
7.0
7.0
7.0
7.0
6.0
7.0
5
5
5
15
15
25
25
5
5
25
25-27
25-27
a From Drinking Water and Health. Vol. 2 (Washington, DC:
National Academy Press, 1980), Chapter 2.
b Concentration of disinfectant times contact time.
Although chlorine dioxide does not produce the variety of chlorinated
organic materials as does chlorine, nevertheless, some organic compounds do
form chlorinated oxidation products, although in much smaller quantities than
from chlorine (Rice & Gomez-Taylor, 1986). Mostly, however, C1O2 oxidation
products of organic materials are non-chlorinated.
A major disadvantage of chlorine dioxide is that as it performs its oxidation
or disinfection work, about half of it reverts back to the chlorite ion.
Although not all of the lexicological parameters of chlorite ion are as yet
known, it does produce hematological effects (Condie, 1986). Consequently,
EPA currently recommends that the total residual concentration of chlorine
dioxide, chlorite ion and chlorate ion not exceed 1 mg/L in finished water
41
-------
because of concerns about hematopoietic and possibly thyroid effects
(Cotruvo and Vogt, 1985).
On the other hand, chlorine dioxide is not known to oxidize bromide ion in
water to produce hypobromous acid (as does chlorine and ozone), which can
form bromine-containing organic materials.
4. Monochloramine
This material is synthesized at the water treatment plant by the reaction of
elemental chlorine with ammonia in equimolar quantities by one of three
procedures:
1. Ammonia can be added to water already containing chlorine,
2. Chlorine can be added to water already containing ammonia,
3. A preformed solution of monochloramine can be added to the process
water.
Procedure #1:
If chlorine is first added to the water to be treated to attain disinfection
over a specific period of time, a CT contribution toward disinfection from
free chlorine will be attained. However, the longer the free chlorine
residual is sustained, the more halogenated organic materials which are >not
desired will be produced during the disinfection contact time. Thus there
will not be major benefit to the use of chloramine, in that instantaneous
halogenated organic levels will be the same. However, addition of ammonia
will insure that THMs will not continue to be generated after monochlor-
amine has been formed. In this manner, TTHM levels can be held to their
initial levels.
Procedure #2:
If ammonia is first added to the water, then chlorine, provided that very
good mixing is available, none of the water should be exposed to significant
concentrations of free chlorine. This will lower concentrations of
halogenated materials, but not necessarily guarantee disinfection since
monochloramine now is the disinfectant.
In certain cases, however, this approach may be even less effective. When
chlorine is added to waters which contain organic nitrogen materials, such as
proteins and aminoacids, the chlorine reacts much more rapidly with the
organo-nitrogen compounds to form organic N-chloramines than with free
ammonia to form the inorganic monochloramine (Weil & Morris, 1949; Morris,
1967). In turn, the organic N-chloramines are even weaker disinfectants
than is monochloramine.
42
-------
Even worse, standard methods of analysis for monochloramine based on
iodometry (i.e., DPD) do not distinguish between inorganic mpnpchloramine
and organic chloramines, because both oxidize iodide ion to iodine in the
determination of "combined residual" chlorine (NAS, 1987, pp. 62-63).
Thus, a water utility with raw water containing organic nitrogenous materials
which adds ammonia, then chlorine to produce "monochloramine" for primary
disinfection may be seriously overestimating the ability of its treatment
system to provide the desired degree of disinfection.
Procedure #3:
To be assured of producing the minimum quantities of chlorinated organics,
the chlorine and ammonia can be mixed in organic-free water, and the
monochloramine solution can be added to the water being treated. This
approach "preforms" the chloramine. However, this approach produces
minimal disinfection and also is subject to the same competition with organic
nitrogen compounds as discussed above. The organo-nitrogen compounds still
present can steal chlorine from the inorganic monochloramine, producing the
much less effective organic chloramines, which are not distinguishable from
inorganic chloramines by field iodometric analytical procedures.
a. Chloramine Summation
Because chloramine is a much weaker disinfectant than is free chlorine, or
any of the other disinfectants, the contact time required to assure
disinfection of any particular organism will be longer than with chlorine,
chlorine dioxide, or with ozone.
In some cases, such as with patients who are on kidney dialysis machines,
the ingestion of chloramine-containing water can be fatal. In municipalities
using C1NH2 as the terminal disinfectant, the utility generally advises local
hospitals and health maintenance centers of the presence of monochloramine
in their tap waters, and advises these institutions to employ distilled water
for their dialysis patients.
As with chlorine dioxide, the use of monochloramine as a bacteriostat can
follow post-filtration disinfection by means of ozone, for example, to provide
minimum production of halogenated organics. Monochloramine is a very poor
chemical oxidizing agent. However, it is known to dissociate slowly in water
to produce small quantities of hypochlorous acid, which in turn produce
traces of halogenated organic materials (Rice & Gomez-Taylor, 1986).
In assessing all currently known facts about monochloramine, the
Subcommittee on Disinfectants and Disinfectant By-Products of the National
Academy of Sciences' Safe Drinking Water Committee has stated recently
(NAS, 1987, pp. 194-195): '
"Because it is a much weaker disinfectant than chlorine, chloramine
must be used at higher concentrations and for longer periods of
43
-------
contact to achieve sufficient disinfection. Even with extended contact
time and higher concentrations, however, chloramination is not
recommended as a primary disinfectant, especially where virus or
parasitic cyst contamination is potentially present."
"Preformed monochloramine is undesirable as a primary disinfectant."
"The use of marginal chlorination as a method of introducing
chloramines into a water supply system is specifically not recommended
because, along with depletion of chlorine to produce inorganic
monochloramine, organic chloramines that have even lower efficacy as
disinfectants are formed. Organic chloramines have also been
implicated as major contributors to the mutagenicity of chlorinated
drinking and natural waters."
There are currently no suitable methods for fully quantifying the
organic chloramine fraction in the presence of inorganic
monochloramine. Until such methods are developed, utilities that
handle water supplies containing high concentrations of organic
nitrogen run the risk of overestimating the ability of their systems to
maintain adequate disinfection."
"When free chlorine is used as the primary disinfectant, an amount
should be used that is sufficient to produce a slight residual of free
chlorine above that required to oxidize nitrogen, followed by addition
of ammonia to form monochloramine and limit THM formation.
5. Ozone
a. General Considerations
This gas also must be generated on site, at the water treatment plant,
because it is too unstable to be stored for significant periods of time in
cylinders. Additionally, it is only partially soluble in water (about 13 times
the solubility of oxygen). Therefore, one key to its successful use is proper
gas/liquid contacting. If all of the ozone generated is not solubilized and
reacted with water constituents, the excess ozone present after contacting
must be reused or destroyed, in order to prevent unnecessary exposure of
plant operating personnel to this material.
Because the generation, application, and handling of ozone is so foreign to
the classical water treatment processor in the USA, and because of the
relatively high capital cost of ozone generation and application equipment,
the acceptance of ozone in North America has been slow to develop, even
though it is used abundantly in other countries of the world (Miller et al.,
1978).
On the positive side, ozone is the most powerful disinfectant and oxidizing
material which is available to the water processor. It kills or inactivates all
organisms tested in shorter periods of time than does any other
disinfectant/oxidant available (see Tables IIA, IIB, IIC, VIII - X).
44
-------
Ozone cannot produce any halogenated organic materials, other than
indirectly, e.g., by oxidation of bromide ion (if that material is present in
raw waters being treated) followed by bromination of organic precursors.
Being a powerful oxidizing agent, ozone used in very low applied dosages is
able to neutralize charges on colloidal particles, thus causing them to
precipitate. In higher applied dosages, ozone is quick to oxidize organic
materials, but usually only partially. In rare cases, organic matter can be
converted quickly to CC>2 and water by ozone. However, under normal water
treatment conditions (i.e., dosages of 1-5 mg/L of ozone and contact/reaction
times of 5-15 minutes), most of the organic materials contained in water are
only partially oxidized.
b. Disinfection With Ozone
For purposes of disinfection, it is necessary to generate ozone, on-site, and
apply the gas (normally air + 1-5% ozone) to the water to be disinfected
(using an appropriate gas/liquid contacting device) for a length of time
appropriate to provide the CT values given in Table IIA for 99.9%
inactivation of Giardia lamblia cysts. In practice, dissolved ozone levels of
approximately 0.4 to 0.5 mg/L are attained and maintained during the
disinfection process. Ozone contact times in currently operating water
treatment plants using ozone for bacterial disinfection and viral inactivation
are a minimum of 10 minutes, and range up to 15-20 minutes.
EPA's proposed Surface Water Treatment Rule requires attainment of 99.9%
inactivation of Giardia lamblia cysts. From Table IIA, the highest CT value
for ozone is 4.5 (for water temperatures of 0.5°C). This means that for a
dissolved ozone concentration of 0.5 mg/L, a maximum contact time of 9
minutes would be required at 0.5°C to attain the CT value of 4.5. This
length of contact time is easily within the design capabilities of the
technology, and would not produce unique consequential problems of design
and installation of ozonation systems.
As will be discussed in the following subsections, ozone has many
applications in drinking water treatment which utilize its oxidation
capability, and which are entirely unrelated to its disinfection capabilities.
On the other hand, whenever ozone is used for oxidative purposes, it is
possible to attain primary disinfection simultaneously, for most oxidative
applications of ozone. The simplest method of attaining simultaneous
primary disinfection is to extend the contact time appropriate to attain the
necessary CT value at the particular water temperature.
However, knowledge of the dissolved ozone concentration during oxidation
also is critical to attainment of the appropriate CT value. In certain
oxidative applications, specifically ozone oxidation of waters containing high
levels of iron and manganese (groundwaters), and preozonation for
microflocculation and turbidity control (surface waters), measurement of a
dissolved ozone residual is inappropriate.
45
-------
In the case or iron and manganese, the dark, precipitating oxides produced
upon ozonation will interfere with measurement of dissolved ozone. In the
case of microflocculation or turbidity control, the objective is to apply a
very low dosage of ozone, which means that a measurable concentration of
dissolved ozone may never be attained. In these cases, ozone disinfection
would have to be provided as a separate treatment step, normally practiced
just before or after filtration.
However, in all other oxidative applications, it is not only possible, but
actually feasible to design the ozone dosage and contact time so as to
produce the CT values required in Tables IIA, IIB, and IIC. In taking credit
for ozone disinfection prior to filtration, attention must be paid to the total
coliform levels at the point of primary disinfection. Successful attainment of
both oxidation and primary disinfection before filtration means that only
secondary disinfection will be required after filtration.
c. Microflocculation
Ozone oxidation of humic/fulvic materials can proceed to a variety of
endpoints. With low applied ozone dosages, oxidation occurs primarily on
the pendant groupings without cleaving the high molecular weight
humic/fulvic polymer chains.
Partially oxidized organic materials contain a plurality of polar, oxygen-
containing groupings, which now allow ready combination with cationic
flocculating agents. Consequently, ozone can be effective in aiding in the
removal of organics when used in or before the rapid mix step, provided that
ozone oxidation is followed by coagulation then conventional or direct
filtration.
On the other hand, when larger amounts of ozone are applied to solutions of
humic/fulvic acids, not only does the oxidation of pendant groupings occur,
but also oxidative scission of the high molecular weight humic/fulvic polymer
chains. This produces low molecular weight, polar compounds, which are not
as easily flocculated with cationic flocculating agents.
Thus, for optimum application as a flocculation aid, low applied ozone
dosages should be employed. German experience in the use of ozone for
microflocculation shows that the ratio of ozone applied to DOC (dissolved
organic carbon) should be in the range of 0.1 (mg/L of applied ozone per
mg/L of DOC) (Sontheimer, 1985).
d. Promotion of Biodegradability
These same partially oxidized organic materials also are more biodegradable,
and advantage is taken of this behavior in many European water treatment
plants. For example, slow sand filters remove organics from drinking water
by operating in a biological mode. Preozonation of the water fed to the
slow sand filters increases the biodegradability of organic materials, and
46
-------
makes the combined oxidation/filtration step more effective (Rachwal, et al.,
1987).
In a similar manner, if GAC adsorption is required for removal of refractory
organic materials, and if some or all of the organic materials can be
partially oxidized by ozonation, then preozonation prior to GAC filtration
can promote biological removal of the partially oxidized organic materials in
the GAC medium. The adsorptive capacity of such Biological Activated
Carbon filters thus can be restricted to the ozone-refractory organics, while
the ozone-sensitive organics (partially oxidized) are converted to carbon
dioxide and water biologically.
The consequences of not providing a biological treatment step after
ozonation in the water treatment plant are biological regrowths in the
distribution system, unless a sufficient level of secondary disinfectant is
present. If the secondary disinfectant is chlorine, then it is important that
sufficient removal of organics has occurred before the chlorine is added.
Otherwise, chlorinated organics still can be produced, sometimes in higher
quantities than without ozonation (Rice, 1980).
e. By-Products of Ozonation
However, the partial oxidation of organic compounds has raised some
concerns as to the possible toxicities of the intermediate oxidation products
formed upon ozonation. In the majority of cases, these oxidation products
are oxygen-containing derivatives of the original organic materials, mostly
aldehydes, ketones, alcohols, and carboxylic acids. Ozone is quick to rupture
many unsaturated linkages in organic molecules, producing aldehydes - and
ketones.
There are, however, a few organic compounds from which ozone has been
shown to produce toxic oxidation products. For example, the pesticide
heptachlor, although rapidly oxidized to "destruction" by ozone, produces
heptachlorepoxide nearly quantitatively. This example illustrates the absolute
necessity of knowing what specific compounds are present in the water being
oxidized and/or disinfected, in order to determine appropriate pre- and/or
post-treatment procedures to cope with these undesired oxidation products.
f. Catalytic Ozonation
For many organic compounds refractory even to so strong an oxidizing agent
as ozone, the simultaneous application of ultraviolet radiation or hydrogen
peroxide along with the ozone can accelerate otherwise sluggish reaction
rates significantly. Acceleration is brought about by catalytic formation of
hydroxyl free radicals, which are stronger oxidizing agents than ozone itself
(see Table VIII).
47
-------
g. Ozone Oxidation of Bromide Ion
Ozone does not form halogenated organic compounds, unless bromide ion is
present in the raw water. If bromide ion is present, hypobromous acid will
be produced, which can form brominated organic materials:
Br" + 03 > (OBr)-
(OBr)' + H+ > HOBr
If bromide ion is present in sufficient quantity to result in the formation of
bromine-containing organics in quantities which exceed current or projected
MCLs, either the bromide ion should be removed, or chlorine dioxide should
be considered as the primary disinfectant, followed by chloramination for
secondary disinfection, or an alternate source of raw water should be
utilized.
h. Summation for Ozone
Because it is such a powerful oxidizing agent and disinfectant (the strongest
available for water treatment) and does not form halogenated by-products
(except when bromide ion is present), ozone is the most versatile
oxidant/disinfectant for water treatment.
As an oxidizing agent, ozone is useful added in very small quantities .for
turbidity control and microflocculation. In higher applied dosages, ozone
oxidizes many troublesome inorganic materials (iron, manganese, sulfide,
nitrite, arsenic), and destroys or alters the chemical structures of many
organics responsible for tastes and odors, colors, THM formation, and other
precursors of halogenated organics. Such alteration of chemical structures
can make the oxidized organic materials readily removable from the. treated
water by flocculation/filtration, and/or by biological means. Many of the
compounds listed as Synthetic Organic Chemicals by the EPA in the list of
83 compounds to be regulated by 1989 (U.S. EPA, 1988a) are at least
partially oxidizable by ozone, although at varying reaction rates.
The more refractory, halogenated, organics which may be present in the raw
water, which may not be oxidized with ozone at a practical rate, can be
oxidized at greatly increased rates by combining ozone with UV radiation or
with hydrogen peroxide.
The recent National Academy of Sciences Report (NAS, 1987, pp. 195-196)
summarizes the latest thinking of the Subcommittee on Disinfectants and
Disinfectant By-Products of the NAS Safe Drinking Water Committee with
respect to water treatment and ozonation in particular as follows:
"When possible, organic precursors should be removed prior to the
disinfection process. Thi$ can be achieved by changing the order of
the procedures of conventional treatment. A better approach,
however, is to improve specific conventional treatment processes to
remove organic compounds and to add processes such as carbon
48
-------
adsorption and preoxidation. Initial removal of organic by-product
precursors precludes the need for reducing (disinfectant) contact tune,
thus improving the efficacy of the disinfection processes and
minimizing formation of organic chlorine by-products."
The use of alternative oxidants, especially ozone and chlorine dioxide,
will increase in the United States in the coming decades. Little is
known about the types of by-products produced by ozonation of
natural organics. Well-conceived studies need to be conducted that
will focus on the stable compounds expected from ozone reactions with
humic material. ... Particular attention should be given to the search
for unsaturated aldehydes and the hydroxy-hydroperondes."
"Following these studies, further health effects studies are needed to
determine whether ozone by-products are mutagenic or carcinogenic or
produce other adverse effects. These studies should take into account
variations that are likely to occur when the oxidation process is
carried out in different matrices (pH, O^/TOC ratio, alkalinity)/
"Notwithstanding the fact that these studies need to be carried out,
drinking water suppliers should not dismiss the possibility of using
ozone as an alternative to chlorine and chloramines in water
treatment. Ozone is an excellent (primary) disinfectant (although it
must be used in combination with a secondary disinfectant to maintain
a residual in the distribution system); ozone is an excellent oxidant
for the various needs of water treatment; it does not form
chlorinated by-products; and the admittedly inadequate studies now
available point to lower toxicities of ozonated water than of
chlorinated water."
6. Ultraviolet Radiation for Groundwaters
The effectiveness of ultraviolet radiation as a bactericide and virucide has
been well established (U.S. EPA, 1986), but is not appropriate for the
inactivation of Giardia lamblia cysts (Rice & Hoff, 1981). It is a physical
disinfecting agent compared to the other disinfecting agents being used,
whose actions are chemical. Radiation at a wavelength of 254 nm penetrates
the microorganism cell wall and is absorbed by the cellular nucleic acids.
This can prevent replication (reproduction) and cause death of the cell.
Since UV radiation is not a chemical agent, no potentially toxic residuals are
produced. Although the structures of certain chemical compounds may be
altered by the UV radiation, the energy levels used for disinfection are too
low for this possibility to be a significant cause for concern.
UV radiation at a wavelength of 254 nm is readily available commercially in
standard equipment from a number of suppliers. It was reported (Angehrn,
1984) that in 1984, some 2,000 communities in Western Europe now disinfect
their drinking water supplies .by means of UV radiation.
Major advantages of UV disinfection are its simplicity, lack of impact on the
environment, and minimal space requirements. There is a negligible
49
-------
likelihood of producing harmful chemicals in the water. Required contact
times are very short, on the order of seconds rather than minutes. The
equipment is simple to operate and maintain, but fouling of the quartz
sleeves or Teflon tubes (housing the UV bulbs) must be dealt with on a
regular basis. Fouling of these items normally is controlled by mechanical,
sonic, or chemical cleaning.
On the other hand, waters containing high suspended solids concentrations,
color, turbidity, and soluble organic matter can react with or absorb the UV
radiation, thus reducing the disinfection performance. In addition, if the
amount of radiation received by the organism is not a lethal dose, photo-
enzymatic repair can occur, and the organism can become viable again. The
phenomenon is termed "photoreactivation", and the enzymatic mechanism
generally requires subsequent exposure to light at wavelengths between 300
and 500 nm (available in sunlight, incandescent, and fluorescent light
sources).
The major operating costs are power consumption and annual replacement of
the UV lamps. Increased popularity and lowered costs have occurred due to
improvements in modern lamp and system designs, increased competition, and
improved reliability and simplicity of operation.
The biocidal properties of UV radiation make it a candidate for use as a
primary disinfectant, which must be followed by a secondary disinfectant
since UV radiation provides no disinfecting residual. In addition, at biocidal
wavelengths (254 nm) and energies, UV radiation does not produce much
oxidation of organic materials. However, at higher energy intensities and
low UV wavelengths (i.e., 184.9 nm), oxidation of organic materials can occur
readily to produce CC>2 and water.
The UV light spectrum is broad enough to include wavelengths of unequal
effects upon different organisms. Even for efficacious wavelengths, the
influence of the UV radiation may differ. Also, the devices designed to
discharge UV light are not all alike in the radiant energy they generate.
Moreover, the UV emanations are susceptible to absorption by the molecules
of the organismsuspending medium. Their germicidal influence may be
reduced by the time they reach the microbe.
That UV radiation is of practical utility in controlling most organisms is
beyond challenge. The devising of effective UV systems requires being
engineered, however. More than an ultraviolet radiation source is needed.
Finally, where the organisms are killed by the UV radiation, subsequent
removal of the dead microbes may be required, as by filtration, when the
presence of particles is undesired, or when their catabolic products may
manifest total organic carbon (TOC) levels, or pyrogenic lipopolysaccharides.
50
-------
VII. DISCUSSION OF TREATMENT TECHNOLOGIES FOR
DISINFECTION AND TO MINIMIZE PRODUCTION OF DOBs
A, THE PROBLEM
In providing disinfection under the projected EPA drinking water regulations,
water treatment systems will be faced with the challenge of how to disinfect
surface and groundwaters to:
1. assure 99.9% and 99.99% inactivation of Gjardia lamblia cysts and enteric
viruses, respectively,
2. assure control of other microorganisms,
3. not impart toxicity to the disinfected water (mutagenicity,
carcinogenicity, etc.),
4. minimize formation of chlorination and other undesired disinfection by-
products, and
5. do all this without exceeding the projected MCL levels for the candidate
disinfectants.
B. THE STRATEGIES
Before primary disinfection is practiced, the concentrations of organic
materials should be as low as practicable, so as to minimize the production
of oxidation and/or halogenation by-products. For secondary disinfection,
the turbidity should be < 0.5 NTU.
If the raw groundwater has the appropriate levels of turbidity and
concentrations of dissolved organics, pretreatment can be eliminated.
Otherwise, at least filtration must be applied.
Concentrations of dissolved organics can be reduced by applying conventional
treatments of flocculation, perhaps sedimentation, then filtration. Newer
techniques of coping with organics concentrations include oxidation (before
filtration, with ozone, potassium permanganate) and adsorption (powdered or
granular activated carbon).
An emerging oxidation technology is the use of ozone coupled simultaneously
with UV radiation or with hydrogen peroxide. This technique generates
hydroxyl free radicals, which are more powerful oxidizing agents than ozone
alone.
C DISINFECTION TECHNOLOGIES
According to the reasoning given earlier in this document, the use of
chlorine for pretreatment and primary disinfection should be minimized. The
current 100 ^g/L MCL for THMs is expected to be lowered. In addition,
51
-------
regulation of some of the halogenated by-products of chlorination listed in
Table III may place even greater restrictions on the use of chlorine for some
utilities.
Projected MCLs for chlorine dioxide and its decomposition products (chlorite
and chlorate ions) effectively eliminate this material as a primary
disinfectant, except for very clean waters and/or short distribution systems.
The known limited capability of monochloramine to achieve the mandated
percent inactivations of Giardia lamblia and enteric viruses eliminate this
material from the role of primary disinfectant.
Therefore, for surface water treatment, only two primary disinfectants
remain: chlorine and ozone. For groundwater (unaffected by surface water),
the same two primary disinfectants remain, plus UV radiation. For
secondary (residual) disinfection, chlorine remains as the most viable, with
occasional use of chlorine dioxide and monochloramine. The conditions of
chloramination should be those recommended by the National Academy of
Sciences (NAS, 1987): chlorination to the nitrogen breakpoint, followed by
addition of ammonia.
Each disinfection approach will be discussed in this section. Much of this
discussion is adapted from an earlier EPA publication (U.S. EPA, 1983) and
supplemented by a more recent publication (U.S. EPA, 1986) which is a
design manual for disinfection of wastewaters by means of chlorination,
ozonation and UV radiation. Nearly all of the factors appropriate to
disinfection of wastewaters is directly applicable to drinking water disinfec-
tion as well. Since most of the utilities which will be affected by the
upcoming Surface Water and Groundwater Treatment Rules serve less . than
10,000 persons, the emphasis of this discussion will be on the smaller
utilities.
1. Primary Disinfectants
a. Chlorine
Chlorine, symbolized chemically as Cl2, is the disinfectant most commonly
used by U.S. water utilities. It is available in three forms:
Form Formula Name
gas Cl2 chlorine gas
solid Ca(OCl)2 calcium hypochlorite
aqueous NaOCl sodium hypochlorite
solution
The gaseous form is used most frequently, especially by larger water
utilities, because it is the lowest cost form of chlorine.
52
-------
i. Chemistry of Chlorination
In the gaseous form, chlorine will react with water to form hydrochloric acid
and hypochlorous acid:
C\2 + H2O <-—> HC1 + HOC1
chlorine water hydrochloric hypochlorous
acid acid
The hypochlorous acid then will react with the water by dissociation to an
extent determined by the pH of the solution:
HOC1 <--—> (OC1)- + H+
hypochlorous hypochlorite hydrogen
acid ion ion
pH is a measure of the concentration of hydrogen ion in the water. The
more hydrogen ion present, the lower is the pH. Conversely, the lower the
hydrogen ion concentration, the higher will be the pH value.
At neutral pH (pH = 7.0), almost 80% of the chlorine is present in its most
effective disinfecting form, hypochlorous acid; the remainder exists in the
less effective hypochlorite ion form. As the pH increases, however, an
increasing amount of HOC1 will react with water to form more hypochlprite
ion. At pH 8.0, for example, nearly 80% of the chlorine present exists as
the hypochlorite ion, almost a complete reversal of the situation which exists
at pH 7.
From the CT data given in Tables IIA, IIB, and IIC for chlorine inactivation
of Giardia lamblia cysts, it can be seen that the CT values for a 2 mg/L
concentration of chlorine increase rapidly for each temperature as the pH
rises above 7.0. Consequently, effective pH control is essential in order to
guarantee the amount of disinfection designed in chlorination systems.
Figure 1 shows the relationship between pH and the concentration of
hypochlorous acid, the desired chlorine species in aqueous solution at 0°C
and 20°C.
When the chlorination step is conducted by adding either sodium hypochlorite
or calcium hypochlorite, the chemical reactions which occur result in an
alkaline (basic) product as compared to the acidic product obtained when
using the gas:
NaOCl + H2O >HOC1 + NaOH
sodium water hypochlorous sodium
hypochlorite acid hydroxide
53
-------
Ca(OCl)2
calcium
hypochlonte
H20 --> 2HOC1
water hypochlorous
acid
Ca(OH)2
calcium
hydroxide
IOC
10 11
Figure 1. Distribution of hypochlorous acid and hypochlonte ions in
water at different pH values and temperatures of 0°C and 20°C
(AWWA, 1973).
The resulting hydroxides increase the pH values of the aqueous solutions
Since an increase in pH results in lower concentrations of HOC1 and
therefore, poorer disinfection (requiring much longer contact times to
54
-------
provide the appropriate CT value to guarantee 99.9% inactivation of Giardia
lamblia cysts and 99.99% inactivation of enteric viruses), the ability to adjust
and control pH becomes critically important when using the hypochlonte
forms of chlorine.
ii. Establishing A Chlorine Residual
Hypochlorous acid is one of the most powerful oxidizing agents known. This
means that it will react with many substances present in the water being
disinfected, in addition to the target organisms. This fact is clear from the
known production of trihalomethanes, and many of the halogenated
compounds listed in Tables III and IV.
In order to achieve a concentration of chlorine sufficient to guarantee the
required level of disinfection, it is necessary to add enough chlorine to react
with all of the reactive substances which are likely to be present, and then
provide the excess "residual" chlorine. These reactions consume chlorine and
are collectively called the "chlorine demand" of the water. Thus the
chlorine demand of a water must be satisfied before a residual of free
chlorine can be provided and an adequate job of disinfection can be attained.
The concentration of chlorine determined by an analytical procedure is called
the "available chlorine residual", and it means only that amount of chlorine
originally added which remains available for disinfection. This residual may
be either a free available residual, a combined available residual, or a
combination of the two. Free available chlorine is essentially the sum of
concentrations of hypochlorous acid and hypochlorite ions. Combined
available chlorine is the sum of the concentrations of mono- and. di-
chloramines, plus nitrogen trichloride and organic nitrogen chlorine-
containing compounds (see later discussion of Chloramines — Section
VILClb).
Intuitively, one would expect that each mg/L of chlorine added to water
would be measurable as hypochlorous acid or hypochlorite ion. This is not
the case, because chlorine reacts with many substances present in the water
in complex manners. To understand some of these complex reactions better,
Figure 2 shows what is called a "breakpoint curve". The amount of chlorine
added is shown on the horizontal scale and the amount of available chlorine
determined by an analytical procedure is shown on the vertical scale.
Assume that the chlorine is added slowly and that the water contains small
amount of reduced substances such as sulfides, ferrous iron, organic
materials, organic nitrogen materials (aminoacids and proteins), and some
ammonia, all of which exert a chlorine demand. The initial amount of
chlorine added will be taken up by reactions with the reduced substances,
and the analysis for free available chlorine [HOC1 + (OC1)'] will show that
none is present.
After the chlorine demands of the reduced substances have been satisfied,
then the HOC1 will react with ammonia, organic nitrogen materials, and some
55
-------
of the organics present to yield chloramines and chlorinated organic
compounds.
-I 6
V
i 4
in
\tt
K
U
i 2
3
z
u
ImrrndioU
•-Dtmond —
H, S,Ft",ttc
Fret Chlorine
4 6 8 10
CHLORINE APPLIED, rog/L
12
14
Chlorine and
Ammonia
or similar compounds
Free Rtsiduals
Figure 2. Graphical representation of the breakpoint chlorination reaction
(U.S. EPA, 1983, p. 111-10). The straight line at the left shows
that chlorine residual is proportional to dosage in pure water.
When impurities are present, they exert a chlorine demand.
When all of the ammonia and other chlorine-demanding organics have reacted
with chlorine, the addition of more chlorine results in the hypochlorous acid
oxidizing some of the same materials it just helped to create. The strange
phenomenon observed is that the addition of more chlorine results in a
decrease in the amount of residual (at this point a combined residual)
indicated by the analytical procedure. When this oxidation is complete
(called the breakpoint), then the addition of still more chlorine results in an
increase in the amount of available chlorine measured. Note that the
breakpoint must be surpassed before a free residual of chlorine can
accumulate and persist.
It is important to be aware that the above illustration is considerably more
complex than described, because the reactions taking place are time-
dependent. For this reason, a breakpoint curve is difficult to recreate and
predict. Individual tests must be run, seasonally, and the data plotted to
define the breakpoint for each water.
56
-------
iii. Factors Affecting Disinfection Efficiency of Chlorine
As indicated by the data of Tables IIA, IIB, and IIC, chlorine in the free
state [HOC1 + (OC1)'] is quite an effective disinfectant. Rapid inactivation
of most microorganisms can be obtained in a matter of minutes. However,
effective disinfection with chlorine requires careful attention to:
1. Concentration of free available chlorine - high enough in the plant so
that it is always detectable at the farthest points in the distribution
system (time basis),
2. Maintaining the pH as close to 7.0 as is practical or consistent with
other water quality aspects, so as to maintain as much of the chlorine
residual in the HOC1 (hypochlorous acid) form,
3. Time of contact long enough to achieve the desired degree of microbial
inactivation (i.e., to attain the CT value commensurate with the
concentration of chlorine at the appropriate temperature).
4. Mixing -- baffle the chlorine contactor well to eliminate the possibility of
short-circuiting of flow or provide an external mixing device.
Other factors also influence the chlorine disinfection process. Temperature
has some effect on the ratio of HOC1 to hypochlorite ion (see Figure 1) and
has more effect on the disinfection rate (see Tables IIA, IIB, IIC, and Tables
XI through XVI - taken from the October 10, 1987 Guidance Manual for the
Proposed Surface Water Treatment Rule, U.S. EPA, 1987d), disinfection being
faster at the higher temperatures. However, usually there is no means
available to the operator to control temperature. Other tables are given in
the Guidance Document (U.S. EPA, 1987d) relating residual chlorine levels to
pH, concentration, and logs of Giardia inactivation.
iv. Disinfection With Chlorine Gas
Chlorine is a toxic, yellow-green gas at ordinary temperatures and pressures.
It is supplied in high strength steel cylinders, under sufficient pressure to
liquefy the chlorine. When chlorine is required, simply opening the gas
valve allows rapid vaporization of the liquid. As the liquid evaporates, its
temperature falls. This normally results in a slower rate of evaporation,
thus requiring manifolding of containers or use of a vaporizer.
There are two basic types of gas chlorinators: (1} pressure operated, direct
gas feed and (2) vacuum operated, solution feed. The former allows chlorine
gas, under pressure, to be fed directly into the water to be disinfected.
Solution feed units mix the gas with a side stream of water to form a
solution of hypochlorous acid and hypochlorite ion, which then is mixed with
the main stream.
57
-------
TABLE XI. CT VALUES FOR 99.9% INACTIVATION OF Giardia
CYSTS BY FREE CHLORINE AT 0.5°C *
Free
Residual
(mg/L)
< 0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
6.0
129
138
145
151
156
160
164
168
171
174
176
179
181
183
6.5
160
172
181
188
194
200
205
209
213
216
220
223
226
228
7.0
196
211
222
231
238
245
251
256
261
265
269
273
277
280
pH
7.5
238
255
268
279
288
296
303
310
315
321
326
330
335
339
8.0
284
305
321
333
344
354
362
370
377
383
389
395
400
405
8.5 9.0
335
360
379
394
407
418
428
437
445
453
460
466
472
478
392
421
443
461
476
489
501
511
521
530
538
546
553
559
These CT values achieve greater than a 99.99% inactivation
of enteric viruses.
v. Disinfection With Sodium Hypochlorite Solution
Liquid chlorinators meter a previously prepared hypochlorite solution directly
into the water to be disinfected. If the water supply system cannot afford
the capital costs (see later sub-section) and requirements associated with
storing and handling chlorine gas, solutions of sodium hypochlorite can be
purchased. It must be remembered that sodium hypochlorite solutions are
more costly per pound of available chlorine and do not contain the
concentrations of chlorine available in cylinders of chlorine gas. Also,
hypochlorite solutions decompose if stored for prolonged periods. Thus,
small systems using sodium hypochlorite should plan to store no more than a
one-month supply.
In recent years, methods for on-site electrolytic generation of aqueous
solutions of hypochlorite ion have been developed. In a two-cell unit, a
brine solution (salt in water) is electrolyzed, producing a solution of
hypochlorous acid in one cell and a solution of caustic (sodium hydroxide) in
the other:
58
-------
Na+ + Cl- + 2H2O + e' —>
sodium chloride water electron
HOC1 + NaOH+ H2
hypochlorous sodium hydrogen
acid hydroxide
TABLE XII.CT VALUES FOR 99.9% INACTIVATION OF Giardia
CYSTS BY FREE CHLORINE AT 5°C *
Free
Residual
(rng/L)
< 0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
nH
6.0
92
98
104
108
111
114
117
119
122
124
126
127
129
131
6.5
114
123
129
134
139
142
146
149
152
154
157
159
161
163
7.0 7.5
140
150
158
165
170
175
179
183
186
189
192
195
197
200
* These CT values achieve greater
of enteric viruses.
169
182
191
199
206
211
216
221
225
229
232
235
239
242
than
8.0 8.5 9.0
202
217
229
238
245
252
258
264
269
273
277
281
285
288
a 99.99%
239
257
270
281
290
298
305
311
317
323
328
332
337
341
280
300
316
329
339
349
357
365
371
378
383
389'
394
399
inactivation
The advantages of this procedure are that purchasing and storing of gaseous
chlorine and hypochlorite solutions are avoided. The primary technical
disadvantages are the generation of hydrogen (which poses fire and explosion
hazards), and the need to dispose of the caustic generated. In addition,
the cost per pound, on a chlorine basis, typically is more than double for
on-site electrolytic generation of hypochlorite ($0.30 to $0.35/lb) versus the
cost of gaseous chlorine ($0.08 to $0.15/lb). However, site-specific
considerations may make on-site hypochlorite generation the process of
choice.
59
-------
TABLE XIII. CT VALUES FOR 99.9% INACTIVATION OF Giardia
CYSTS BY FREE CHLORINE AT 10°C *
Free
Residual
(mg/L)
< 0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
* These
enteric
oH
6.0
69
74
78
81
83
86
88
90
91
93
94
96
97
98
6.5
86
92
97
101
104
107
109
112
114
116
117
119
121
122
7.0 7
105
113
119
123
128
131
134
137
140
142
144
146
148
150
CT values achieve greater
viruses.
.5
127
136
144
149
154
158
162
166
169
172
174
177
179
181
than
8.0
152
163
171
178
184
189
194
198
201
205
208
211
214
216
a 99.99%
8.5
179
193
203
211
218
224
229
234
238
242
246
249
253
256
9.0
210
225
237
247
255
262
268
273
279
283
288
292
296
299
inactivation of
vi. Disinfection With Solid Calcium Hypochlorite
Solid calcium hypochlorite is stable when properly packaged and sealed. A
water supply system can purchase its annual requirements in a single
procurement. Simply mixing the proper amounts of solid and water to allow
metering without clogging of pumps or metering valves is all that is required
for use. Normally, an entire drum of calcium hypo-chlorite is made into
solution. This avoids the partial use of a container, with attendant
uncertainties of proper resealing and loss of strength.
vii. Chlorination System Design
Choice of the form of chlorination system to be used, whether gaseous
chlorine, dry calcium hypochlorite, sodium hypochlorite solution or on-site
generation, depends upon a number of factors which include the following:
o availability of chlorine source chemical,
o capital cost of the facility,
o operation and maintenance costs for the equipment,
60
-------
o chemical costs,
o location of the facility,
o operator skills available,
o safety.
TABLE XIVCT VALUES FOR 99.9% INACTIVATION OF Giardia
CYSTS BY
Tree
Residual
(mg/L)
<, 0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
FREE CHLORINE
AT 15°C
*
DH
6.0 6.5
46
49
52
54
56
57
59
60
61
62
63
64
65
65
* These CT values
of enteric viruses.
57
61
65
67
69
71
73
74
76
77
78
79
80
81
7.0 7.5
70
75
79
82
85
87
89
91
93
95
96
97
99
100
achieve greater
85
91
96
100
103
106
108
110
112
114
116
118
119
121
than
8.0
101
109
114
119
123
126
129
132
134
137
139
141
142
144
a 99.99%
8.5
120
128
135
140
' 145
149
153
156
159
161
164
166
168
170
9.0
140
150
158
164
170
174
179
182
186
189
192
194
197
199 .
inactivation
Each of the methods of chlorination will provide the same disinfecting power
on a pound for pound basis of available chlorine when utilized at the same
pH. However, each of the systems must be approached differently in terms
of basic design and safety.
Sufficient chlorine must be provided to satisfy the chlorine demand of the
water at the point of chlorine addition, plus an additional amount to
maintain the required residual after a specified contact time (see CT values
in Tables XI to XVI). The relative dosages of the various chemical sources
of hypochlorite ion [(OC1)-] in solution can be determined; these frequently
will depend upon the point of chlorine application in the process.
The chlorine demand of raw water usually is far higher than that of finished
water. In any case, a minimum contact time, commensurate with the CT
61
-------
value, the pH of the water, and the temperature, must be maintained to
assure the appropriate inactivation of Giardia lamblia and viruses.
TABLE XV.CT VALUES FOR 99.9% INACTIVATION OF Giardia
CYSTS BY FREE CHLORINE AT 20°C *
Free
Residual
(rng/L)
< 0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
6.0
34
37
39
40
42
43
44
45
46
46
47
48
48
49
6.5
43
46
48
50
52
53
55
56
57
58
59
60
60
61
7.0
53
56
59
62
64
66
67
68
70
71
72
73
74
75
pH
7.5
64
68
72
75
77
79
81
83
84
86
87
88
90
91
8.0
76
82
86
89
92
95
97
99
101
102
104
106
107
108
8,5
90
96
101
105
109
112
114
117
119
121
123
125
126
128
9.0
105
113
119
123
127
131
134
137
139
142
144
146
148
150
These CT values achieve greater than a 99.99% inactivation
of enteric viruses.
Chlorination With Gaseous Chlorine:
Chlorine is supplied in high strength steel cylinders with minimum chlorine
capacities of 100 and 150 pounds, up to one ton, and in tank cars under
sufficient pressure to liquefy the chlorine. Major manufacturers of gaseous
chlorine are listed in Table XVII. However, the quantity consumed by small
water systems normally would be purchased from local suppliers which are
listed in the local telephone directory yellow pages under "Chemical
Suppliers" or "Swimming Pool Suppliers".
Direct feed chlorinators add gas under pressure directly into the water to be
disinfected. This type of unit normally is used only when electrical power is
unavailable or insufficient water pressure differential to operate a solution
feed unit. This is a site-specific application which will not be discussed
further.
62
-------
TABLE XVICT VALUES FOR 99.9% INACTIVATION OF Giardia
CYSTS BY FREE CHLORINE AT 25°C *
Free
nH
Residua]
(mg/L) 6.0 6.5 7.0 7.5
< 0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
*
23
25
26
27
28
29
29
30
30
31
31
32
32
33
These CT values
of enteric viruses.
29
31
32
34
35
36
36
37
38
39
39
40
40
41
achieve
35
38
40
41
43
44
45
46
47
47
48
49
49
50
greater
42
46
48
50
51
53
54
55
56
57
58
59
60
60
than
8.0
51
54
57
59
61
63
65
66
67
68
69
70
71
72
a 99.99%
8.5 9.0
60
64
68
70
73
75
76
78
79
81
82
83
84
85
70
75
79
82
85
87
89
91
93
94
96
97
99
100
inactivation
Solution feed units mix chlorine gas with a side stream of water to form a
hypochlorous acid solution, which then is injected into the main stream.
Solution feed chlorinators operate on a vacuum controlled basis, auto-
matically shutting off if the side stream flow is interrupted. This type of
unit, shown in Figure 3, is preferable for safety reasons over direct feed
units.
The basic solution feed gas chlorinator includes the following components:
o gas shut-off valve to interrupt gas flow;
o vacuum regulator, or
o gas flow indicator;
o adjustable gas flow controller;
o check valve;
o venturi type gas injector.
The market for supply of gas chlorinators is quite competitive, as illustrated
by Table XVIII, which provides a partial listing of suppliers of these types
of units.
63
-------
TABLE XVII.
MANUFACTURERS OR PACKAGERS OF GASEOUS
CHLORINE
Name
Ashland Chemical Company
Petrochemical Division
Dow Chemical U.S.A.
FMC Corp.,
Industrial Chemical Group
Georgia-Pacific
Kaiser Aluminum & Chemical Corp.
Industrial Chemical Division
Kaiser Chemicals
Cleveland, OH 44124
Kuehne Chemical Co., Inc.
Kearny,
Occidental Chemical Corp.
Industrial & Specialty Chemicals
Olin Corporation
Pennwalt Corporation
Inorganic Chemicals Division
Stauffer Chemical Co.
Vulcan Materials Co.,
Chemicals Division
Address
P.O. Box 2219, Columbus, OH
43216
2020 Dow Center, Midland, MI
48640
2000 Market Street, Philadel-
phia, PA 19103
P.O. Box 105605, Atlanta, GA
30348
300 Lakeside Drive, Oakland,
CA 94643
30100 Chagrin Boulevard,
86 Hackensack Avenue,
NJ 07032
360 Rainbow Blvd. South, P.O.
Box 728, Niagara Falls, NY
14302
120 Long Ridge Road,
Stamford, CT 06904
3 Parkway, Philadelphia, PA
19102
Westport, CT 06881
P.O. Box 7689, Birmingham, AL
35253
Well established standards for design of gas chlorination systems exist in
standard waterworks industry literature (Great Lakes - Upper Mississippi
River Board of State Sanitary Engineers, 1985; Am. Water Works Assoc.,
1985). The following points are provided to highlight possible design
questions:
64
-------
O'
c
•-I
ft
TO VENT
O\
s.°
Q.
i—i
a
OOP
c
3.
i
^-»
§
i
o>
LEAD
GASKET
CHLORINE
CYLINDER
VALVE
YOKE
CLAMP
RATE
VALVE
VACUUM SEAL
"0" RING
OUTLET CONNECTION
' VENT VALVE
RATE INDICATOR
REGULATING
DIAPHRAGM
ASSEMBLY
VACUUM LINE
EJECTOR AND
CHECK VALVE
ASSEMBLY
WATER SUPPLY —
CHLORINE
SOLUTION
CHLORINE CYLINDER
a
8
-------
TABLE XVIII. SOME GAS CHLORINATOR MANUFACTURERS
Name Address
Capital Controls Co., Inc. Box 211, Colmar, PA 18915
Chlorinators Inc. 733 NE Dixie Highway, Jensen
Beach, FL 33457
Fischer & Porter Co. County Line Rd., Warminster,
PA 18974
Hydro Instruments Inc. 492 Richlandtown Pike, Box
615, Quakertown, PA 18951
Modern Process Systems Inc. 14834 Highland Road, Baton
Rouge, LA 70810
Wallace & Tiernan, 25 Main Street, Belleville, NJ
Division of Pennwalt Corp. 07109
Chlorination With Sodium Hvpochlorite
Sodium hypochlorite, usually supplied in concentrations of 5 and 15%
available chlorine, is available commercially only in solution. In this form, it
is easier to handle than gaseous chlorine or calcium hypochlorite. However,
sodium hypochlorite solutions will lose their disinfecting (oxidizing) power
during storage, and should be stored in a cool, dark, dry area. The material
is supplied in glass or plastic bottles, carboys, or lined drums ranging in size
from 0.5 to 55 gal. Bulk shipment by tank truck also is a common form of
transport. No more than a one-month supply of the chemical should be
purchased, to prevent loss of available chlorine.
Chlorination With Calcium Hvpochlorite
This material is supplied as a white solid which is quite corrosive, and gives
off a strong chlonnous odor. It contains approximately 65% available
chlorine, is readily soluble in water, and is available in granular, powdered,
or tablet form. It is provided in 2, 5, 8, and 35 Ib cans and in 100, 300,
and 800 Ib drums. The containers generally are resealable.
Calcium hypochlorite is hygroscopic (readily absorbs moisture), and reacts
slowly with atmospheric moisture to form chlorine gas. Therefore, shipping
containers must be emptied completely or carefully resealed. It is not
feasible to handle this material in bulk handling systems.
The contents of a calcium hypochlorite container are emptied into a mixing
tank where it is readily and completely dissolved in water. The resulting
corrosive solution is stored in and fed from a stock solution vessel
constructed of corrosion-resistant materials such as plastic, ceramic, glass,
66
-------
or rubber-lined steel. Dosage of the solution at 1% or 2% available chlorine
content is by a diaphragm type, chemical feed/metering pump.
viii. Chlorination Systems Costs (U.S. EPA, 1983)
In the following discussion, cost data are presented for the three types of
chlorination systems discussed above, and as related to their use at small
water treatment systems, sized to treat water volumes up to at least 1 mgd.
The larger systems will have to change chlorine cylinders more frequently
than will the smaller systems. For example, a 1 mgd water treatment plant
using an average chlorine dosage of 5 mg/L will use nearly 42 Ibs of
chlorine per day. Thus, a 150-lb cylinder of chlorine will last between three
and four days at this size plant.
Operating and maintenance costs presented do not include chemical costs.
These will vary depending upon the volume of water treated and the dosage
required.
Cost data were obtained primarily from two sources (Hansen et al., 1979;
U.S. EPA, 1983), and updated by calling various vendors in mid-1987.
However, a more recent publication by Gumerman et al. (1986) contains
additional cost data developed specifically for small water systems. Cost
curve figures from this later publication are included, as appropriate.
Solution-Feed Chlorination With Gaseous Chlorine
Equipment Costs
Table XIX shows a detailed cost breakdown obtained during May 1980 from
three vendors of chlorination equipment. Data are presented in terms of a
basic gas chlorination system, as well as costs for five increasingly complex
systems. The basic system includes equipment to handle two 150-lb chlorine
cylinders, two cylinder-mounted chlorine gas regulators, automatic
changeover valve, and chlorine gas flow and rate valve ejector (with system
backup). Alternate #1 adds two scales, a gas mask, and a diffuser
corporation cock (to allow connection under water line pressure). Alternate
#2 adds a flow-pacing chlorine addition system. Alternate #3 adds a flow
meter. Alternate #4 adds a booster pump and piping. Alternate #5 adds a
chlorine leak detector.
The cost comparisons in Table XIX present a basic-to-most-sophisticated
comparison between the various system configurations in which gaseous
chlorination systems can be purchased. Costs are comprised of equipment,
installation, safety enclosure, contractor's overhead and profit, plus 10%
engineering fee for the basic system estimates.
The basic (lowest cost) gaseous chlorination system costs about $9,350; with
all options added, the most sophisticated gaseous chlorination system costs
$16,050, in May 1980 dollars.
67
-------
TABLE XIX.
CAPITAL COSTS FOR GAS CHLORINATION (1980
Dollars*)
EQUIPMENT COSTS FOR A SYSTEM OF 100 Ibs/day (2 kg/h) OR
LESS
Basic System**
Alternate #1 - add scales, mask
diffuser, corporation cock
Alternate #2 - add flow pacing -
existing signal
Alternate #3 - add flow meter
& signal, 8 in. or less
Alternate #4 - add booster
pump & piping
Alternate #5 - add Cl2 gas
detector
INSTALLATION
SAFETY ENCLOSURE
Average $1.873
High 2,300
Low 1,320
Average
Average
Average
Average
Average
Average
High
Low
Average
High
Low
770
1,694
2,068
792
1,382
1167
1,500
1,000
3.500
6,000
2,000
CONTRACTOR'S OVERHEAD AND PROFIT (20%) 1,869
ENGINEERING FEES (10%)
TOTAL CAPITAL COST
Basic System
Most Sophisticated
(with Alternate #5)
* May 1980 quotes (three vendors)
934
$9,343
$16,049
**
Basic system includes two 150-lb chlorine cylinders, two
cylinder-mounted regulators, automatic changeover valve,
chlorine gas flow rate valve, and ejector (system with
backup)
68
-------
Construction Costs
Figure 4 shows construction cost curves for gas feed chlorination systems up
to a maximum chlorine feed rate of 80 Ibs/day (Gumerman et al., 1986).
Operation and Maintenance Costs
Hansen et al. (1979) state that in general, operation and maintenance costs
for chlorination systems treating 2,500 gpd to 1 mgd are independent of
flow. Process energy requirements are for the booster pump only, and are
about 1,630 kWh/yr. Building energy requirements for a 25 ftz building to
house the system would be 2,560 kWh/yr. Maintenance material requirements
would be only for miscellaneous repair of valving, electrical switches, and
other equipment, and would total about $40/yr. Labor requirements are for
periodic checking of equipment, with an average requirement of 0.5 h/day, or
183 h/yr.
O&M costs of $2,45 7/yr are summarized in Table XX. Note that power costs
were taken at $0.07/kWh and labor at $10.00/h. These were prevalent rates
in 1982 and were used (U.S. EPA, 1983) to update the corresponding energy
and labor costs made earlier by Hansen et al. (1979).
TABLE XX.O&M SUMMARY FOR SOLUTION-FEED GAS
CHLORINATION
Item Requirements* Costs
ELECTRICAL ENERGY:
Process 1,630 kWh/yr x $0.07 = $ 114.10
Building 2,560 kWh/yr x $0.07 = 179.20
Total 4,190 kWh/yr x $0.07 = $ 293.30
MAINTENANCE MATERIAL: = $ 40/yr
LABOR: 183 h/yr x $10/h = $ 1,830
TOTAL ANNUAL O&M COST $2.457
* Estimates of energy, maintenance, and labor made by Hansen et
al. (1979).
O&M cost curves are depicted in Figure 5 (building energy, process energy,
and maintenance material) and Figure 6 (labor and total O&M costs), both
taken from Gumerman et al., 1986).
69
-------
to,OOP.OOP
\
6
5
4
3
2
1.
ft
* «
* 3
H
W „
0 2
0
Z
0 1<
CONSTRUCT
> w * atat^toMi
-------
1C
3
a
6
S
4
3
2
1
9
I
*:
•»
i 3
j
< 2
oc
Ul
H-
< 1(
-1
Ii
2 4
5 »
i *
11
]
5
4
3
2
1
,000 1,
: &
8
a
s
4
3
2
300 1
5
4
3
2
>.
»0 | 1(
i~ I i i i i i i
ENERGY - 1
•H •* «* * o»«^o»«
E f
a
9
4
3
2
1<
1
000.00
00.000
ag-M«ae^i*if •;.?-.
^
).000
100^
fiiyiiiiii.uiniiini 'iii
HM«TH*itfflnaiiaa^a
)0
0
mmnm
marrm
asm
mm
SOS
BHBtm
BEBM
X
/"
H
aoooi
^
ma)
V
(
r
ta
•
^
•
h
• •
?
r
B
• •
(AINTEN/
ROCESS
>
UILDING
iNCE
ENE
ENE
: M
RQ
RG
A"
••M
Y
Y
re
••
R
••
IA
•
(,
• •
2 3 4 S 678i 2 3439 789 2 3 4 S 3 789
.0 10 100 1000
AVERAGE PEED RATE • It/day
10 100
AVERAGE FEED RATE - kg/day
Figure 5. Operation and maintenance requirements for gas feed
chlorination systems -- building energy, process energy, and
maintenance material (Gumerman et al., 1986).
71
-------
100.000 100.000
I
3
4
3
2
1C
S
CO
O
° 1C
rr
O 7
O 4
"3
2
1<
8
S
4
3
2
11
lows*- » « •» « w S IQ
8
1 1 1 i i i i i 3i
E »
6
S
4
3
i
£
»00 » 1C
•i OC 9
0 1
07
< 8
4
3
2
4
K) 1C
f 8
5
4
3
2
) K
•
[•HI Illl Illl i
1.000
p_»
00
Mmmaum
»0
3
«^ *
*»aaj
•si^ai
™-^--
l—
Effil?
—f-
j
i
J
B»Oi
jass
^-i
tat
SiESi.
=«
'
J
—
n-gan-J
,-
-°
s»
•5S-—
~~.
JHE
ra
.
a
8
Jeaswassss*
H^H
BOTSSBffil
"—
B^l
fcr
—
ntt«
1st
F
"
s
m
B8
f
(
2 349 8789 2 3 4 S 6 78S
18 1
AVERAGE FEED RATE - It)/
TOTAL O&M COST
'(EXCLUDING
CHEMICALS)
LABOF
B B
t 2 3 4 S@7SS
00 1060
day
10 100
AVERAGE FEED RATE - kg/day
Figure 6. Operation and maintenance requirements for gas feed
chlorination systems -- labor and total O&M cost (Gumerman et
ah, 1986j.
72
-------
Chemical Costs
In 150-lb cylinders, chlorine cost $0.47/Ib in the Washington, DC/Baltimore
area in January 1983. In 1987, the cost for gaseous chlorine in 150-lb
cylinders is $0.37/lb, when purchased in ton quantities. For purposes of this
discussion applicable to small water systems, the $0.47/lb will be considered
as current.
Assuming a chlorine dosage of 5 mg/L, gaseous chlorine chemical costs would
be about $18/yr to treat 2,500 gpd, and about $7,150/yr to treat 1 mgd.
To calculate costs for gaseous chlorine at lower or higher dosages, the
following formula can be used:
dosage (mg/L) x no. of L dosed/day x C\2 cost/lb
———— = C\2 cost/day
1000 (mg/g) x 454 g/lb)
Sodium Hvpochlorite Solution Feed
Equipment Costs
Table XXI displays estimates obtained during May 1980 from two vendors of
sodium hypochlorite chlorination equipment. Data are presented for the
basic liquid hypochlorination system which includes two metering pumps (one
serves as standby), solution tank, diffuser, and appropriate quantities of
tubing. However, two types of basic system are costed, one activated
electrically, the other activated hydraulically. The basic systems can be
supplemented, and costs of four increasingly sophisticated alternatives also
are presented in this table. Alternate #1 adds a diffuser corporation cock
and anti-siphon backflow preventer. Alternate #2 adds a safety- housing
enclosure. Alternate #3 adds a flow-pacing system, and Alternate #4 adds a
flow meter and signal.
The total capital costs for the basic and most sophisticated systems are:
most sophisticated
basic system system
Electrically Activated $3,643 $10,738
Hydraulically Activated $5,023 $15,121
Construction Costs
Figure 7 is a cost curve (Gumerman et al., 1986) which shosw that the
construction cost for hypochlorite solution chlorination systems is constant
for systems feeding up to 100 Ibs/day.
73
-------
TABLE XXI. CAPITAL COSTS -- LIQUID CHLORINATORS*
EQUIPMENT COST
(basic system**)
INSTALLATION
SITE WORK
Electrically
Activated
Avg. $ 1,800
High 2,300
Low 1,300
500
250
Hydraulically
Activated
$ 2.266
2,782
1,750
1,000
250
CONTRACTOR'S OVERHEAD & PROFIT (20%) 729 1,004
ENGINEERING FEES (10%) 364
Alternate #1: add diffuser 165
corporation cock & anti-siphon
backflow preventer
Alternate #2: add safety 6,930
enclosure (housing)
Alternate #3: add flow pacing
existing signal
Alternate #4: add flow meter
signal, 8 in. or less
TOTAL CAPITAL COST:
Basic System (Equipment + $3,643
Installation + Site Work +
Overhead & Profit +
Engineering Fees
Most Sophisticated $10,738
(with Alternate #2)
(with Alternate #4)
* May 1980 quotes (two vendors)
**
503
231
6,930
1,485
1,452
$5,023
$15,121
Basic System includes two metering pumps (one standby),
tubing, solution tank, and diffuser
74
-------
1,
8
S
4
3
2
D
8
S
» 4
1 3
CO -
0 *
0
i t(
i-TT
o i
JE «
co 8
i ^
0 3
2
1C
9
3
4
3
2
1C
1
000.00
^
M^MHHHM*
30,000
1,000
—
100
>0
0
aE=E£T'TE^
, L
T
L
pz
i
U
;'
I
m
1!
i!
—- -1
i
lUt
1
JL
i
^^r^fsoii
F
|
z
I
FT
2 3 4 S 87i$ 2 3 4 S i 7i
10
MAXIMUM PEED RATE - 1
-— — — -
UUKl
L
5Bfl^
H
r
•
sea
•a
t
JHI
^
^a«
^
M
V
aw
s
|
EK6
B
MOB
(B 09
91 B
K D
HI •
• •
m v
E as
m as
•
9 2 3 4 36799
100 1900
b/day
1 10 100
MAXIMUM FEED RATE ' kg/day
Figure 7. Construction cost for hypochlorite solution chlorination systems
(Gumerman et al., 1986).
75
-------
Operating and Maintenance Costs
As with solution-feed gas chlorinators, O&M requirements are independent of
flow for plants treating 2,500 gpd to 1 mgd (Hansen et al., 1979). Process
energy requirements are for the diaphragm metering pump and amount to 570
kWh/yr. Building energy requirements for a 25 ft^ building would be 2,560
kWhyyr. Maintenance materials would be required only for minor component
repair -- costs which are estimated at $20/yr.
Labor is required for periodic mixing of the sodium hypochlorite solution, as
well as for checking of the equipment. Based on a requirement of 1 h/day,
the annual labor requirement would be 365 h/yr.
Annual O&M costs of $4,108 are summarized in Table XXII. Note again that
power costs are based on $0.07/kWh and labor costs of $10.00/h.
TABLE XXII. O&M COST SUMMARY FOR SODIUM HYPOCHLORITE
SOLUTION FEED (U.S. EPA, 1983)
Item Requirements* Cost
ELECTRICAL ENERGY:
Process 570 kWh/yr x $0.07 = $ 39.90
Building 2,560 kWh/yr x $0.07 = $ 179.20
Subtotal 3,130 kWh/yr x $0.07 = $ 219.10
MAINTENANCE MATERIAL $ 20/yr
LABOR 365 h/yr x $10/h = $3,650
Total Annual O&M Cost $4.108
* amounts estimated by Hansen et al. (1979)
Figure 8 shows constant O&M requirements for hypochlorite solution
chlorination systems (building energy, process energy, and maintenance
material) for systems feeding up to 100 Ibs/day of chlorine. Figure 9 shows
similar constant O&M maintenance requirements for hypochlorite solution
chlorination systems (labor and total O&M costs). Both curves are from
Gumerman et al. (1986).
Chemical Costs
Sodium hypochlorite customarily is sold as a 15% (by weight) solution. In
July 1987, the cost of a 15% solution in 1,500 gal tanks in the Washington,
DC/Baltimore area was $0.50/gal. A small water utility treating 2,500
gal/day with 5 mg/L of chlorine as sodium hypochlorite solution would
require about 64 gal/yr. At $0.50/gal, this utility would spend about $32/yr.
76
-------
1(
6
9
4
3
2
1(
* 9
*» 4
: *
E 2
UJ
t-
<
a K
ui "T
0 8
Z 7
< 0
z a
Ul a
5 4
< 3
a
2
K
1
3
2
1
3,000 1
- 9
: *
a
9
4
3
2
)00 1
- 9
: i
a
9
4
i
3
2
>»
io i K
(£. 9
UJ
^ Z 4
3
2
<
i
1 K
8
» S
4
- 3
- 2
1G
1.
000. OC
M§^^BC"EI»TK-
30.000
.— «x
).000
B»Bs«3Ki
fclUHimillKH
)00
0
0
sasrssa
t-
F
i
|
EKSffia'.I^
jg^SSSGESL
WLS^T
f™1'^
C3!:
i' '
L
I
!=•
r
K:
r~
_1T~ j
i! i
5H3S&&S3S
^,i:.-,'i:.-
^^^^
rn
-44
r:t
fl
rcT~~-"
t i^i !
S j
!
"^IR3 "i*Y*as&~li—
"TT
rr"
iEjraesiinaJi^JDsWsKSJii-
t )
"""^IH^"1
assaa
»»|n>o
|
[ T
BKfl
«
6%
IKpS!
U J^**
r
•J MAJNl
" ' RJATI
- i
I
jLzE'JZJSTi.
_
PEN.
fl9A!
ESS
INQ
Jst.i3W
tfywm.
JU
rT1
4_
%N(
D
^
$£&£'-
^
t- — B
KB H E
ENERGY
i i i i
ENERGY
2 3 4 S 6 7 Si 2 3 4 9 8 789 2 3 4 i€7i
0 10 100 tO
AVERAGE FEED RATE - Ib/tiay
^"I
T^
n wi
n n
H •
!
IS
00
1.0
10
AVERAGE FEED RATE
100
Figure 8. Operation and maintenance requirements for hypochlorite
solution chlorination systems - building energy, process
energy, and maintenance material (Gumerman et al., 1986).
77
-------
100.000 100,000
i
0
5
4
3
2
J<
9
4
£ 3
*»
1 2
t-
CO
O
0 1(
II
<:
2 »
2
1C
I
4
3
2
1C
= I
"- 7
a
5
4
3
2
I.OOO 1C
S
4
3
2
h»
>
>.
IOO f 1(
id i I I l I I 1 1
LABOR
!;• MO*, »»-*tant
: 1
0
5
4
3
2
> 1C
1.
1,000
100
1 "f<^--~
H
-"
-
;
_IIIII^ ' •' !
H
10
I
JL
^^LTTTT
L_._L.IJ_
LJ
SSB
0^
2 3 4 S 6789
0 11
AVEI
«^-—
. -(EXC
CHgl
kL 0
LUC
MGt
)R
&M
n»^
^L£
c
3
1}
;c
>s
T
2 3430 789 2 3 4 58788
3 100 1000
*AGi PEED RATE • Ib/day
i 1
1.0 10 100
AVERAGE FEED RATE - kg/day
Figure 9. Operation and maintenance requirements for hypochlorite
solution chlorination systems - labor and total O&M cost
(Gumerman et al., 1986).
78
-------
A 1 mgd facility using 5 mg/L of chlorine would require 400 times more
hypochlorite solution, and would spend $12,800/yr.
However, it should be realized that the cost for sodium hypochlorite solution
is about three times the cost of gaseous chlorine in equivalent quantities,
and that 0.9 Ib of 15% NaOCl solution is equivalent in oxidation potential to
one pound of gaseous chlorine. These ratios between the two forms of
chlorine will assist in obtaining more accurate cost calculations.
Calcium Hypochlorite Solution Feed
Equipment costs and operating and maintenance costs for this method of
disinfection should be very close to those for sodium hypochlorite feed
systems. Solutions of calcium hypochlorite are prepared in a mixing tank,
then transferred to a day tank (a tank which holds enough solution to last
for one day), then injected into the water stream using a diaphragm
metering pump.
Chemical Costs
Solid calcium hypochlorite, Ca(OCl)2, contains 65% available chlorine.
Therefore, one pound contains 65 Ib of available chlorine. Since a 2,500
gal/day treatment plant requires 0.104 Ib of chlorine per day (at an average
dosage of 5 mg/L), 0.104/0.65 = 0.16 Ib of Ca(OCl)2 per day is required with
which to prepare a solution for metering into the water to be treated.
During July 1987, calcium hypochlorite was selling for $0.95/lb in, the
Washington, DC/Baltimore area. Over a one year period, the 2,500 gal/day
facility will require 0.16 Ib x 365 days = 58.4 Ibs of Ca(OCl)2, x $0.95/lb =
$55.48/yr. A 1 mgd facility, which uses 400 times the amount of chlorine,
therefore will spend $22,192/yr for Ca(OCl)2.
Pellet Feed Chlorinators
Figure 10 shows that construction costs for pellet feed chlorinators are
constant for systems feeding up to 15 Ibs/day. Figure 11 shows operation
and maintenance requirements for pellet feed chlorinators (process energy
and maintenance material); Figure 12 shows operation and maintenance
requirements for labor and total O&M cost. These curves are taken from
Gumerman et al. (1986).
Erosion Feed Chlorinators
Figure 13 shows constant construction costs for erosion feed chlorinators
feeding up to six Ibs/day of chlorine. Figure 14 shows operation and
maintenance requirements for erosion feed chlorinators (process energy and
maintenance material); Figure 15 shows O&M requirements for labor and
total O&M cost. All three figures are taken from Gumerman et al., 1986.
79
-------
100,000
6
9
4
3
2
1(
6
9
4
3
2
1
i\
e
9
4
3
2
1(
^
6
9
4
3
2
1'
0.
LSpJL.
JOO
Vt?
—
•MB9
iimmtMfflg
«
=
)
3
_™
™
«
EBC3
-
^SSE
_
""•
"
?se
23SU
ESS
_
E
"
SJL.
r~
i
:(
,:
"•"""-f
i rf
-
^
-
-vTB&l
4
)
^±
=«j
I
!
"—
ISKLlT^W
Eaasaai
i s
T I ' _ ~^1
r ~*i
-
_
I
:
i
rift
i_LX
h
1
4
1
J^
1 "
jj
1
^SB
2 3 4 9 8789 2 3498 789 2 3498 71®
1 1.0 10 101
MAXIMUM FEED RATE • Ib/day
*»
I
t-
co
O
o
z
o
H
u
oc
t-
w
z
o
o
0.1
1.0 10
MAXIMUM FEED RATE • kg/day
Figure 10. Construction cost for pellet feed chlorinators (Gumerman et al.,
1986). v
80
-------
10(
1
a
9
4
3
2
10
9
MAINTENANCE MATERIAL - $/yr
p M ** *• W»-*aHt»|-. M (•» * (HOt^OMCJO M U *> UK_-4fe
)0 10
•» ^
a
9
4
3
2
9
4
3
r jl
* in
IO 1 1 1 1 1 1 1
ENERGY *
JO M U *. MtOt-^OH
a
4
3
1 lo
0.
000
12_
— J
1)
^
y
6:
®m
>R<
•2 »
1
^^^.-.^=5— .,--•.- .•
fc:"™^^
" „, i
iNTENA
E^S
3CESS i
p-
ICE
•Nil
_„„„.„_,[_.,„__!
--^-^^IL™,
"~~™"™[™"
M
IQ1
.; .^HS
IBM
AT
'
UOf
BBB
•«•
m
E
E
•a
3
BS
•B
IB
H
^fl
iTl,
i
tunJ—
j
j
i
T* f
HOT
• M
• •
L
i
-hn
i »
H
2 34387§a 2 3498 789 2 3 4 3S7§§
T 1.0 10 100
AVERAGE CHLORINE FEED RATE • ib/day
0.1 1.0
AVERAGE CHLORINE FEED RATE - kg/timy
Figure 11. Operation and maintenance requirements for pellet feed
chlorinators -- process energy and maintenance material
(Gumerman et al., 1986).
81
-------
100.000 100.000
1
8
S
4
3
2
12
9
4
5 3
4»
* 2
CO
O
o io<
< s
i;
2
1
S
4
3
2
IP.
8
9
4
3
2
.goo 10
8
8
4
3
x 2
JO ^ 10
tD 7
: 5 :
4
3
2
I m
i i
4
3
2
10
0,
noM^noan^.
(MMBVH^^DW*
•^••••••iii
GPP
DO
MWH
^^•MKE£^X5
B^1BMK&££3?3
5
•••in imt'n imimm
BSggg
I-— •—•
I — -
L ™
msst
_
••«••••
r""L±C"~*c'n'*'
teHSEuBafeti-sa^Kt. h_i.
™~L!_
ffi«^«
ncrfa*;*
=_
fM»J!
~"
L
GK.TJ
-.
-
~
bd
h
_
_
__,L,i._L«
i
^i
»sa
!*•
a
1
I
36
^
^^K—'
.„ ...IIM-iJ
B^e£!^
X
.^^
riid
™
•MHMMI
•i^H^HH
X*
^
pi
>
h-
^
TO
MEffi
^
f
T"
i
1
1
T
(
f
*
01
EX
t*
**
\
fAL 0«D
CLUOIN
®
4 C<
3 C
LABOR
X
3S'
HE
I
r
M
C
At
BKBSBSffi;
.S)
2 3498789 2 3498 789 2 3 4 8 8 7SS
1 1.0 10 100
AVERAGE CHLORINE FEED RATE • Ib/day
1 ' ' ' ' 1 ' ''''''I ' . . i i i i • , 11
0.1 1.0 10
AVERAGE CHLORINE PEED RATE - kg/day
Figure 12. Operation and maintenance requirements for pellet feed
chlorinators - labor and total O&M cost (Gumerman et al,
1986).
82
-------
1.
4
•f
6
9
4
3
2
1
I
6
S
4
r '
§ «
o
i H
1 S
O 4
° 3
2
000.00
mmrmnBX^
00,000
"""""" (at
~ c
c
3.000
1000
8
9
4
3
2
1C
0.
(0
0
fe=^
taHEWia^-
I
URVi
URVI
URVi
^E^EEJSS
pHBSBm
l"™"
: A
• B
: C
t
i^mK
a
a
•
h
P
•
•
j
<
f
i , i
h-
rf
PRE
ro
PRE
^FT
iftA1
40UI
SSURE W,
WELL B
SSURE M
ER AN Ai
IflTY OPI
NTED AB(
^Ef
EFOf
^TEF
30VI
•RAT
3VE
t S
IE
I S
: G
ID
A
YS
A
YS
RO
F
SI
toast
Ti
S!
11
UE
E!
Qi
i mini™
EM
JB^i
EM '
€; i
•DEI
lASi
- CHLORINE
ERSIBLE W
. CHLORINE
'UHP
1-CHLORINA7
1 TANK
i
AD
IP
AD
'OR
DI
DI
T
T;
[ON
[ON
2 349 6739 2 3 14 9 8 r88| 2 3 4 5i7SS
1 1.0 10
MAXIMUM FiiO RATg • Ib/day
.
i
100
0.1
1.0 10
MAXIMUM FiEO RATE • kg/day
Figure 13. Construction cost for erosion feed chlorinators (Gumerman et
83
-------
1000
100.000
AVERAGE FEED RATE • kg/day
Figure 14. Operation and maintenance requirements for erosion feed
chlorinators -- process energy and maintenance material
(Gumerman et al., 1986).
84
-------
1C
6
5
4
3
2
JC
5
4
» 3
<0
ft
O «
2
1C
T
6
4
3
2
10
10,000 1
6
$
4
3
2
>,000 11
a
6
S
4
3
2
5*
JOO f 1(
tnof~9 W « » M ^j
HO8V1
II 1 1 1 1 1 I 5i
^ 3
S
4
3
- 2
1C
C
30,000
-»S_B«=4=— .
~ '^""^^
5 000
)00
5i^*1
0
a£LU,
^
+
a*4&
T.1"™:1 T-fr1"1 ""
.... _.j«^
«*8B*fjp
OB*35a!5a^»^sai3»
1
J»—
f-
^
f
M
'
0
*'
>*
/
{
_jt
X
J*
4
r
mA
(/
.
/-
.39eLC£
«,
__J
/"
•BH
A
*
UC5BC
•»Pui-
W
••
rjj
r
""
F-
1 TOTAL
M«=
IBB
f"?
. O&M COST
UQING
CAL5)
• LABOR
Kupn
• m
• •
• •
2 34SS789 2 3489 789 2 3 4 i i 789
.1 1.0 10 100
AVERAQg FEED RATE - Ib/day
0.1
1.0
10
AViBAQE FEED RATE • kg/day
Figure 15. Operation and maintenance requirements for erosion feed
chlorinators - labor and total O&M cost (Gumerman et al.,
85
-------
On-Site Electrolytic Hypochlorite Ion Generation
Figure 16 shows construction costs for on-site hydrochlorite generation
systems producing up to 100 Ibs/day of chlorine. Figure 17 shows operation
and maintenance requirements for on-site hypochlorite generation systems-
building energy, process energy, and maintenance material; Figure 18 shows
operation and maintenance requirements for labor and O&M cost. These
curves are taken from Gumerman et al. (1986).
b. Ozone
i. Characteristics and Properties of Ozone
Ozone (03) is a very powerful oxidizing agent, second only to elemental
fluorine among readily available chemicals. Because of its strong oxidizing
ability, ozone also is a powerful disinfectant. It is an unstable gas at
ambient temperatures and pressures, and decomposes rapidly to oxygen (from
which it is made) at temperatures above 35°C. For this reason, it cannot be
manufactured and packaged at a central manufacturing plant, as can chlorine.
Therefore, ozone must be generated on-site and used at once.
Ozone has a characteristic odor which can be detected by most humans at
low concentrations (0.01 to 0.05 ppm by volume), far below the levels of
acute toxicity. However, olfactory fatigue has been noted in some instances.
This means that as the length of exposure to an ozone-containing
atmosphere increases, the odor of ozone may become less noticeable to the
individual being exposed.
>
When added to water, ozone dissolves only partially. That is to say that
ozone is only slightly soluble in water, about 2-10 times the solubility of
oxygen, depending on the water temperature. The solubility of ozone in
water is governed by Henry's Law, which states that the mass of ozone that
will dissolve in a given volume of water, at constant temperature, is directly
proportional to the partial pressure of the ozone gas above the water.
Consequently, proper design of the ozone contacting system is very
important to the proper application of ozone to water (see later discussion-
iii).
The stability of ozone is greater in air than in water, but is not excessively
long in either case. The half-life of ozone in water has been reported to
range from 8 minutes to 14 hours, depending on the phosphate and carbonate
concentrations of the water (Grunwell et al., 1983). With no phosphates or
carbonates present and the water adjusted to pH 7.0 with sodium hydroxide,
the half-life was 8 minutes.
86
-------
0)
O
o
z
o
H
o
(C
h-
co
2
O
O
1(
7
6
5
4
3
2
t
8
9
4
3
2
1C
i
a
9
4
3
i
2
1(
6
5
3
2
1C
1
), 000,0
.~-™-~-
-?nrg^ ^aiggtaa^ijCE
)0,000
^•B^SSS!ES!S
J.OOO
00
i
2ESQSH
I
)00
•S3SB2L
ft^H
_
^c_
"™"
„«
««S2ffl
1 1
=>-»
--
#*>
P
«=™*l=r~'
.a.,
*
T
onJ
^-
1
p-
57^1
^5^a^:JGa^^.^^raa»J5»?iJMigJ^a^
|
£!
-.
L ~ I
->
1
-n^sea^
t^uatiffitf
jagwran-gr
^ai-JM"
"
EBDBCTt^
m^ra^njr
jnv
L
=
sar-
I
i
_
:
6
n
r"
m»t— 1
[i
•i
*m
M
U
1
1
i
B M
P M
• Pi
B
2 3436789 2 3436 789 2 34517 89
.0 10 100 1000
MAXIMUM FEED RATE - Ib/day of chiorin* equivalent
1.0 10 100
MAXIMUM FEED RATE - kg/day of chlorine equivalent
Figure 16. Construction cost for on-site hypochlorite generation systems
(Gumerman et al., 1986).
87
-------
10.000
1,000.000
MAINTENANCE MATERIAL
PROCESS ENERGY
BUILOINQ ENERGY
2 3 4 SS788
1000
AVERAGE FEED RATE • Ib/day o* ehlorlne- equivalent
i I i I I f i i i '[ i I < I i i i rr *
1.0 10 100
AVERAGE FEED RATE « kg/day oJ chlorine equivalent
Figure 17. Operation and maintenance requirements for on-site
hypochlorite generation systems -- building energy, process
energy, and maintenance material (Gumerman et al., 1986).
-------
CO
O
2
•a
O
_J
<
t-
o
1C
6
S
4
3
2
JJ
6
3
4
3
2
1(
T
6
9
4
3
2
1C
T
s
4
3
2
1C
10,000 K
7
8
5
4
3
2
XOOO K
e
s
4
3
„ 2
w
)00 . 1C
IQ 1 1 1 1 1 1 I 1
LABOR
>|-^ M U * OK»-^OHC
e
s
4
•» 3i
2
» K
1
30,000
3,900
__m™»_
"cr
—
(
)00
^
!0
mim^maammgam
***•£
>
-*-**1
Kn_
z&>
I
^
r
c:
h
-
i
|
^
'
/
*«s
^*
grf
I*
T-S
!»•
K1
j?
rc^=ZL~ ' »— — i-^—
i
, t
--— H--=— •
™(
y
"«^™"J
""^^
1
^
rf
1
/
/
i
1
Cj— :f—
II
,/
n
,
. TOTAI
• (EXCL
. Ot
.UDI
ICA
M
|
r
•»
(
?BB^E5
K*
•1
1
>
LABOR
i
I
kM
NG
LS]
C
35
r
iT
™
EV M H
•MBMM
2 3 4 5 6789 2 3458 739 2 3 4 58789
.0 10 100 1000
AVERAGE FEED RATE - tb/day of chlorine equivalent
1.0
I
10
I
100
AVERAGE FEED RATE • kg/day of chlorine equivalent
Figure 18.
Operation and maintenance requirements for on-site
nypochlorite generation systems -- labor and O&M cost
(Gumerman et al., 1986).
89
-------
ii. Generation of Ozone
For the on-site generation of ozone by means of corona discharge (the
procedure used most for water treatment), an oxygen-containing gas (air,
oxygen-enriched air, or pure oxygen) is dried and cooled, then passed
between two electrodes separated by a discharge gap and a dielectric
material across which high voltage potentials are passed. In recent years,
the more modern ozone generating equipment has been designed to operate
at high frequencies rather than high voltages.
Figure 19 is a schematic diagram of the essential components of a corona
discharge ozone generator. For small water plants, ambient air dried to a
dew point of minus 65°C (-65°C) will be the source gas fed to the ozone
generator. It is imperative that the ambient air be dried rigorously in order
to maximize the yield of ozone produced, as well as to minimize the
formation of corrosive nitrogen oxides.
i HEAT 1
T1 ?
%
— ELECTRODE
— DIELECTRIC
02
DISCHARGE
GAP
Figure 19. Typical corona discharge cell ozone generating configuration.
With properly dried air, the output from a corona discharge ozone generator
will be dried, cooled air, containing \% to 3.5% of ozone, which is partially
soluble in water (about 2-10 times the solubility of oxygen). This produced
mixture of ozone in air then must be mixed with the water to be treated, by
a process known as contacting (see later discussion - iii).
Ozone also can be generated by passing UV radiation through ambient air.
However, in this procedure, only small quantities of ozone can be generated,
and at much lower concentrations (0.1 to 0.001%) than can be generated by
corona discharge. This concentration is considered insufficient to provide
enough ozone in water to cause a significant amount of oxidative or
disinfective treatment.
90
-------
When oxygen is used as the feed gas to the ozone generator, the
concentration of ozone produced is effectively doubled over that produced
when properly dried air is employed. This means that oxygen containing 2%
to 7% (by weight) of ozone is generated in the same size corona discharge
ozone generator and for the same expenditure of electrical energy as when
dried air is used.
In the United States today, there are at least 35 operational water treatment
plants using ozone. Only the new 540 mgd plant in Los Angeles is using
oxygen as the feed gas. In the rest of the world, where ozone is used more
extensively than currently in the U.S., fewer than a dozen water treatment
plants are known to be using oxygen as the feed gas, although more interest
is being shown in its use.
When the properly dried air or oxygen passes through the ozone generator,
part of the oxygen dissociates as a result of being exposed to the high
energy electrical field of the corona discharge:
O2 + e' > 2[O] "fragments"
oxygen
These oxygen "fragments" are highly reactive, and they combine rapidly .with
molecular oxygen, forming the triatomic molecule, ozone:
2[O] + 2 O2 > 2 03
oxygen oxygen ozone
fragments
The overall reaction to produce ozone is the sum of the above reactions:
3 O2 + e' <— -> 2 03
oxygen energy ozone
The reaction to produce ozone is reversible, meaning that once formed,
ozone decomposes back to oxygen. This reverse reaction occurs quite
rapidly above 35°C. Therefore, because reactions involving high energy
electrical discharges also are accompanied by generation of considerable heat,
ozone generators are designed to include a high degree of cooling, in order
to minimize ozone losses by decomposition.
iii. Contacting of Ozone With Water
Because ozone is only partially soluble in water (2-10 times that of oxygen),
the manner with which it is contacted with the water to be treated is of
primary importance. In turn, the manner of contacting ozone with water
depends upon the particular ozonation job or jobs to be accomplished.
91
-------
For purposes of illustration, there are basically two types of reactions
involving ozone: fast and slow. Fast ozonatipn reactions include the
destruction or inactivation of microorganisms (disinfection), oxidations of
iron, manganese, sulfide ions, nitrite ions, some organics, and lowering
turbidity levels. On the other hand, other ozone oxidations, such as
oxidation of many organic materials (specifically many pesticides and VOCs),
are kinetically rather slow.
As a result, for the fast ozone reactions, the contacting objective is to add
the requisite amount of ozone to the water as rapidly as possible, so that
the solute being oxidized or inactivated will react as rapidly as possible.
The rate of this type of ozonation reaction is dependent on the rate of
transfer of ozone into solution. This type of reaction is termed "mass
transfer controlled", and contact times, therefore, are relatively short.
For example, during disinfection with ozone, ozone is added first to satisfy
the ozone demand of the water and to create a residual dissolved ozone
level, then to maintain the requisite residual of ozone for a given period of
time.
Slower ozone oxidations are dependent upon maintaining a low residual
concentration of ozone in the water, but over a much longer length of time,
because the chemical reaction rates are slow. These types of reactions are
termed "reaction rate controlled".
Ozone can be generated under positive or negative air pressure. . If
generated under positive pressure, the ozone-containing air normally exits
the ozone generator at approximately 15 psig. This is sufficient pressure for
the gas to pass through porous diffusers installed at the base of a 16 foot
column or tank of water. Fine bubbles containing ozone and air (or oxygen)
rise slowly through this water column. During the time of bubble rise, the
ozone contained in the bubbles diffuses from the gas phase into the water
phase, where oxidation and/or disinfection takes place. The 16 ft height of
water column maximizes the amount of ozone transferred from the tiny
bubbles as they rise.
Other types of positive pressure ozone contactors include packed columns,
static mixers, and high speed agitators. The water to be treated also can be
sprayed through small orifices into an ozone-containing atmosphere (the
atomizer, or spray-drying principle).
When ozone is generated under negative pressure, the air or oxygen is drawn
through the feed gas preparation system, then through the ozone generator
and into the contactor. The sub-atmospheric pressure can be created by
employing a submerged turbine as the contactor. Thus, when the turbine is
turned on, its whirling action creates a slight vacuum, drawing air through
the entire ozonation system.
Other methods of creating sub-atmospheric pressure which are in common
practice are by use of injectors or Venturi-type nozzles. Water to be
treated is pumped past a small orifice (injector) or through a Venturi nozzle.
In either case, a slight vacuum is created, which can be used to draw the
92
-------
air/oxygen/ozone output of the ozone generating system. Contacting takes
place during the mixing of gas with flowing water.
It can be appreciated that the diffuser and packed tower contactors require
no energy to operate, whereas high speed agitators, static mixers, and all of
the negative pressure contactors require energy for their operation.
As excellent discussion of the various types of ozone contactors has been
assembled by Masschelein (1982).
iv. Destruction of Contactor Exhaust Gas
No matter how well designed the ozone contactor, it is rare, if ever, that
100% transfer of ozone to the water occurs, unless insufficient ozone is
provided. Therefore, there will always be some excess ozone in the gases
which exit the ozone contacting system. Particularly for disinfection with
ozone, during which a dissolved residual of ozone is present for a
predetermined length of time, there will always be ozone present in the
contactor exhaust gases (off-gases). This excess ozone either must be
reused or destroyed before discharge of the gases to the plant atmosphere.
Reuse of ozone can be effected if two stages of ozonation (one for
oxidation, one for disinfection) are designed into the water treatment
process. In this event, contactor exhaust gases from the later ozonation
step are drawn into the earlier ozonation stage, for example, by means of a
submerged turbine. However, even well-designed reuse of ozone in a second
stage still will leave some excess ozone in the final exhaust gases, and this
must be destroyed before discharge. Destruction of ozone (to produce
oxygen from which it is generated) is accomplished by several methods:
o thermal destruction -- at 300-350°C, ozone decomposition .requires
only a few seconds,
o catalytic destruction - at ambient temperatures with the appropriate
catalyst (metal or metal oxide),
o adsorption and decomposition on moist granular activated carbon,
o exposure to UV radiation (254 nm wavelength).
This latter method is in use only in very large treatment plants, because
even at 254 nm wavelength, there is a small amount of ozone generated.
v. Chemistry of Ozone in Water
Unlike chlorine, ozone does not react with water to produce disinfecting
species. However, above pH about 6, ozone does decompose in water to
produce the more reactive hydroxyl free radical:
03 + H20 —> O2 + 2(OH)-
93
-------
This reaction is accelerated at the higher pH values (above 8). In addition,
hydroxyl free radicals are produced whenever ozone is exposed to UV
radiation, or when ozone is applied in the presence of hydrogen peroxide.
Ozone oxidation rates of some refractory organic materials are greatly
accelerated by the simultaneous addition of UV radiation or H2®2 because of
the formation of hydroxyl free radicals.
Thus, when ozone is added to water it can react as the 03 molecuJe, as the
hydroxyl free radical, or as a mixture of both.
Bicarbonate and carbonate ions are excellent scavengers of free radicals;
therefore waters containing high alkalinity levels will quickly destroy the
hydroxyl free radicals produced by the decomposition of ozone, and eliminate
the benefits of the free radical oxidative mechanisms.
vi. Establishing An Ozone Residual
In order to attain the appropriate CT value to achieve 99.9% inactivation of
Giardia lamblia cysts (Table IIA) a measurable residual level of ozone must
be developed in water and maintained for the appropriate period of time. In
operating drinking water treatment plants using ozone for disinfection
(primarily in France, but also in many other countries of the world), it is
customary to attain a dissolved ozone residual concentration of approximately
0.4 mg/L, then to maintain this concentration for a period of time of at
least four minutes (laboratory) and 6-8 minutes in operating water treatment
plants.
It can be appreciated that the water entering the ozone-for-disinfection
contacting chamber has the highest ozone demand, whereas water exiting the
contacting chamber has the lowest ozone demand. Therefore, part of the
function of the ozone contacting system is to satisfy the short term, or
immediate, ozone demand of the water. Since disinfection is a mass transfer
controlled reaction (fast), a certain percentage of ozone is utilized initially
to satisfy the short term ozone demand. After this ozone demand is
satisfied, the contacting task becomes one of maintaining a specific ozone
concentration for a specific time period.
These two tasks normally are conducted in two different contacting
chambers, with the larger amount of ozone being applied to the first
chamber. This procedure, which is employed to minimize the amount of
ozone added, is shown in Figure 20. Approximately two-thirds of the total
ozone required is added to the first chamber, in which the ozone demand of
the water is satisfied, and a residual ozone level of about 0.4 mg/L is
attained. The balance of the ozone (one-third) is added to the second
chamber, in which the 0.4 mg/L ozone residual is maintained.
To attain the necessary contact time for the ozone with a 0.4 mg/L ozone
residual, it is simply a matter of dividing 0.4 into the CT value from Table
IIA taken at the appropriate temperature. The product is the minimum
contact time (in minutes) required. Table XXIII shows the CT values and
94
-------
contact times required for an ozone residual of 0.4 mg/L at various
temperatures.
UNOZONATED
WATER
CONTACT
CHAMBER
OFF-GAS
U'AvB^
«.* it >'".!«• ..vv
OZONE-RICH
AIR
FLOW METER (TYPICAL)
VALVE (TYPICAL)
Figure 20. Two-compartment ozone contactor with porous diffusers.
TABLE XXIII. CT VALUES AND CONTACT TIMES FOR OZONE
TO ACHIEVE 99.9% INACTIVATION OF Giardia
lamblia AND > 99.99% INACTIVATION OF ENTERIC
VIRUSES FOR 0.4 mg/L OZONE
Temperature (°Q
Disinfectant pH
Ozone (0.4 mg/L) 6-9
CT value
Contact Time (min)
0.5
4.5
11.25
5
3
7.5
10
2.5
6.25
15
2.0
5
20
1.5
3.75
25
1.0
2.5
95
-------
If the dissolved ozone residual can be raised to 0.5 mg/L, the contact times
will be somewhat lower. If the ozone residual is lower (i.e., 0.3 mg/L), the
contact time will be somewhat longer. If only a 1-log or 0.5-log inactivation
of Giardia is required, as when preozonation disinfection is practiced prior
to filtration), the CT value (and therefore, the ozone contact time) required
will be even lower (see Table IIB and IIC, respectively).
Many current ozonation systems in the U.S. today are designed to optimize
the oxidative capabilities of ozone, not necessarily to optimize ozone
disinfection. In these instances it is customary to provide a single ozone
contact chamber. If single contact chambers are used for ozone disinfection,
and the residual ozone concentration is measured at the outlet of the
contact chamber, how should the ozone concentration and contact time be
determined?
The initial response is to assume that the ozone residual concentration at
the outlet is present throughout the contact chamber. Were this to be true,
one would simply multiply the detention time in the contactor by the
concentration of ozone in the water exiting the contactor to determine the
CT value.
However, this assumption is incorrect. Considering only the first
compartment of the 2-compartment ozone contactor in Figure 20, it must be
realized that the water entering the upper part of the compartment contains
no ozone and constantly exerts an ozone demand. Therefore it can be
concluded that the upper portion of this compartment contains much .less
ozone than the lower portion. However, the ozonized water exits from the
lower section with a measurable dissolved ozone residual.
It is clear from this reasoning that the true CT value of a single contact
chamber is lower than that obtained by multiplying the detention time by the
ozone concentration in the water exiting the chamber. As a first
approximation only, one should assume that the average concentration of
ozone throughout the contact chamber is, say, 50% of that measured at the
outlet. But even this approximation may not be true.
If single ozone contact chambers are to be supplied, it is advisable to
measure the dissolved ozone concentrations at various points in the chamber
at the various water flow rates and ozone dosages with the actual water to
be treated. In this manner, an ozone concentration profile can be developed
for the particular contactor, and CT values can be calculated which are
closer to the true value.
It must be recognized that if ozone is employed early in the water treatment
process to oxidize iron and manganese, and if the levels of iron and
manganese are sufficiently high, copious amounts of brown/black precipitate
will be produced. These insoluble materials will interfere with the
measurement of a dissolved ozone residual, even though a significant residual
may be present.
96
-------
vii. Factors Affecting the Disinfecting Efficiency of Ozone
It should be appreciated that because ozone is such a powerful oxidizing
agent, it is not particularly selective. In other words, if ozone is used early
in the water treatment process, for example, to oxidize iron and manganese,
or color, or tastes and odors, a high degree of disinfection also will be
obtained. Conversely, if ozone is used after filtration for disinfection, it
will also oxidize any easily oxidizable materials still present.
Ozone's ability to disinfect is affected little by changes in water
temperatures or pH, and it does not react with water, except as noted
earlier. Also, ozone does not react with ammonia at any significant rate,
except above pH 9. However, ozone does have a short half-life in water,
decomposing to oxygen at a rate dependent upon the water temperature.
This means that ozone will not have a lasting residual in the distribution,
and therefore normally cannot be employed as the secondary disinfectant.
Although there are a number of European, Canadian, and even two U.S.
water treatment plants which do use ozone as the secondary disinfectant,
these are the exception rather than the rule. In those cases in which ozone
is the last treatment step, a combination of five factors must occur
simultaneously (Miller et al., 1978):
o cool water temperatures,
o clean and short distribution system,
o short residence time (< 12 hours)
o low levels of organics,
o no ammonia present.
In all other cases, a secondary disinfectant which provides a stable residual
is added after ozone has been utilized as the primary disinfectant. One
advantage of employing ozone as the primary disinfectant is that the oxidant
demand of the water is satisfied simultaneously, which allows much lower
dosages of chlorine, chlorine dioxide, or monochloramine to be utilized,
normally less than 0.5 mg/L, to provide a stable, detectable, residual for the
distribution system.
viii. Ozonation System Design
Figure 21 illustrates the five basic components of an ozonation system which
employs ambient air as the generator feed gas. The essential components
include air preparation, electrical power supply, ozone generation, ozone
contacting, and ozone contactor exhaust gas destruction. The amount of
ozone produced by an ozone generator is affected by the physical
characteristics of the equipment, the power supply to the ozone generator,
the moisture content and dust content of the feed-gas, the temperature of
the ozonized gas, and the oxygen content of the feed gas. Each of these
basic system components will be discussed below.
97
-------
THE FIVE BASIC
COMPONENTS
OF THE
OZONATION
PROCESS
Electrical
Power
Supply
2
Gas
Preparation
1
dry
gas
Ozone
Generator
3
A
HoO
o,
Ozonated Water
Exhaust Gas
Destruction
5
Contactor
4
Figure 21. The five basic components of an air-fed ozonation system.
Air Preparation
Ambient air must be dried to a maximum dew point (a unit of moisture
content of air) of minus 60°C. Even drier gas is preferable (lower than-
60°C). This unit process must be designed conservatively, especially for
warm, humid climates. Use of air having a dew point higher than -60°C will
result in lower ozone production, slowly foul the ozone production
(dielectric) tubes or plates, and increase corrosion problems in the ozone
generator unit and downstream equipment as well.
Air feed systems typically are classified by their operating pressure. The
most common is a low pressure system, which operates at pressures ranging
from 69 to 103 kilopascals (kPa = 10 to 15 psig), although pressures up to
275 kPa (40 psig) have been reported when the gas pressure is reduced prior
to the ozone generator. High pressure systems operate at pressures ranging
from 480 to 690 kPa (70 to 100 psig), reduce the pressure prior to the ozone
generator, and typically are used in small to medium sized applications.
Either system may be used in conjunction with most ozone generators and
with all of the contacting systems described.
98
-------
The decision to use a high or low pressure air preparation system often is
based on a qualitative evaluation of potential maintenance requirements, in
addition to the quantitative capital cost evaluation. Some of the issues to
consider are listed below:
1. High pressure air pretreatment equipment generally has higher
maintenance requirements for the air compressors;
2. High pressure air pretreatment equipment generally has lower maintenance
requirements for the desiccant dryers.
3. High pressure air pretreatment equipment generally has a lower capital
cost. At small to medium sized installations this lower capital cost may
offset the additional maintenance required for the air compressors and
associated equipment, such as filters for the oil-type compressors. The
design engineer should investigate the potential maintenance associated
with the high and low pressure systems rather than evaluating the design
on the basis of capital cost alone.
Another type of air feed-gas treatment system is the "nominal pressure"
system, which typically operates at a negative or in some cases a slightly
positive pressure. This type of air pretreatment system is designed to
operate in conjunction with aspirating turbine, injector, and/or Venturi type
contactors, all of which create the partial vacuum necessary to draw ambient
air through the air pretreatment and ozonation systems. Since ozone
disinfection of drinking water involves maintenance of a specific ozone
contacting time to attain the required CT value, this type of contacting' will
require higher energy expenditures than the positive pressure contactors (and
positive pressure air pretreatment systems). Therefore, nominal pressure air
pretreatment systems are not considered appropriate for drinking water
disinfection.
Figures 22 and 23 are schematic diagrams showing low pressure and high
pressure air feed gas pretreatment systems, respectively. Figure 22
illustrates a dual component process, showing desired flexibility for the
equipment provided. The precompressor filters are provided to protect the
air compressors from damage due to large particles. The air compressors
typically are positive displacement, oil-less units. Positive displacement
compressors are used in order to obtain constant air flow at variable
operating pressures. Variable pressures often are encountered due to
variable pressure losses in downstream equipment and processes such as
filters and ozone contact basins. Oil-less compressors are used to eliminate
oil contamination of the downstream desiccant dryer medium and ozone
generator dielectrics. Liquid-seal and rotary lobe compressors have been
used most frequently.
The compressors may be followed by an after-cooler or a refrigerant dryer.
These components are depicted by dotted lines in Figure 22, which indicate
that they are optional. Typically, either one or the other option is provided.
These cooling mechanisms are used to remove moisture in the air at minimal
operating expense.
99
-------
Pre-Compressor
Filter
Compressor
— — —>— — —
After-Cooler
__ _ ___.
._„]- Optional
Pre-Compressor
Filter
Compressor
,___L__.
• »
1 After-Cooler !
I __ .„.. j
Refrigerant
. - Dryer -—.
I
___J__...
• Refrigerant
L-- Dryer
I
Pre-Desiccant
Filter
HH
Towe
rA
To
/ver B
Post-Desiccant
Filter
Heat-Reactivated
Desiccant Dryers
Pre-Desiccant
Filter
Tower A
Tower 8
Post-Desiccant
Filter
To Ozone Generator
To Ozone Generator
Figure 22. Example low pressure air feed-gas treatment schematic for
ozone generation (U.S. EPA, 1986, p. 125).
100
-------
Pre-Compressor
Filter
Compressor
After-Cooler
Pre-Oesiccant
Filter
Tower A
Tower B
Post-Desiccant
Fitter
Heat-Less
Desiccant Dryers
Pra-Compressor
Filter
Compressor
After-Cooler
Pre- Desiccant
Filter
1
Tower A
Tower B
I
t_ _r
Post-Desiccant
Filter
PRV - Pressure Relief Valve
To Ozone Generator
To Ozone Generator
Figure 23. Example high pressure air feed-gas treatment schematic for
ozone generation (U.S. EPA, 1986, p. 126).
101
-------
The compressed, cooled air is directed to a predesiccant filter, which is used
to remove dust and dirt particles greater than 3 to 5 microns in diameter.
Particulate removal prior to the desiccant dryers reduces plugging in the
desiccant medium.
Probably the most important component of the air treatment process is the
desiccant dryer, which consists of two towers containing moisture-adsorbing
media. One tower operates in the adsorption mode while the other tower is
being regenerated. The low pressure system desiccant dryer uses heat for
reactivation of the desiccant.
Post-desiccant filters are installed to remove particulates smaller than 0.3 to
0.5 micron in diameter. Two-stage filtration is preferred. The first stage
filter removes particulates greater than 1 micron and the second stage
removes particulates less than 0.3 to 0.4 micron in diameter.
In the high pressure air pretreatment system (Figure 8), the precompressor
filters are used to remove larger particulates and protect the air
compressors. These compressors typically are oil-less units; however, oil-seal
compressors can be used if followed by extensive oil removal equipment.
Following the high pressure compressors, it is essential to provide after-
coolers, to remove the heat of compression. The filter(s) before, the
desiccant dryer are used to remove particulates less than 3 to 5 microns in
diameter when oil-less compressors are used. When oil-seal compressors are
used, filtration to remove oil droplets less than 0.03 micron is provided.
The high pressure system desiccant dryer consists of two towers with
moisture adsorbing media. One tower operates in the adsorption mode while
the other is regenerating. Regeneration is accomplished without additional
heat. Thus, the high pressure desiccant dryers are called heat-less units.
The post-desiccant filters remove particulate matter less than 0.3. to 0.4
micron in diameter. The high pressure system also has a pressure reducing
valve to regulate operating pressures in the ozone generator.
Air compressor, refrigerant dryer, and desiccant dryer design considerations
are discussed in detail in U.S. EPA, 1986.
Electrical Power Supply
Supply line voltage (220/440 V), or frequency in some cases, is varied to
control the amount of ozone being generated and its rate of generation.
Because these two parameters are varied in many ozone generation systems,
the electrical power subunit can represent a proprietary product of the
ozonation system supplier. As a result, the electrical power system must be
specified as an integral power supply system specifically designed for the
ozone generator to be supplied. In other words, the power system should be
designed for and purchased from the ozone generating system supplier.
Because the ozone generator uses high voltage (> 10,000 V) and in some
cases also high frequency electrical current (up to a maximum of 2,000 Hz),
102
-------
special electrical design considerations must be implemented. For example,
special insulation must be provided for the electrical wire. A cool
environment for the high voltage transformers should be provided, and the
electrical transformers should be protected from ozone contamination due to
minute ozone leaks which might occur on a periodic basis.
The electrical considerations for an ozone system should receive special
attention. For example, a number of problems have been reported with dry-
type potential transformers. Oil-cooled transformers apparently have
performed more reliably. In view of the dependence of ozone generation on
high frequency or high voltage electrical energy, the ozone generator
supplier should be responsible for designing and supplying the electrical
subsystems. However, the specifications should require that the frequency
and voltage transformers be high quality units designed for ozone service.
The ozone generator supplier should be requested to provide a record of
successful electrical equipment performance.
Another item to consider in the design of an ozone generating system is the
power factor. This is the ratio between the apparent power (kw) measured
by a watt-hour meter and actual power (kVA) measured in terms of voltage
and amperage. The relationship is shown by the equation:
Power Factor = Apparent Power/Actual Power
pf = kW/kVA
The power factor is UNITY when the voltage and current of an alternating
current power supply are "in-phase" with each other, for example in a purely
resistive circuit like a heating element. In a purely capacitive circuit, such
as an ozone generator which has not reached the ionization potential of the
dielectric, the voltage and current are 90 degrees out of phase. In this
case, the power factor is ZERO. For a generator producing ozone, the
voltage and amperage will be somewhere between 0 and 90 degrees out of
phase; thus, the power factor will be less than 1.0. The actual power factor
will vary depending upon the power supply to the ozone generator and the
amount of electrical resistance developed within the electrical circuit.
The power factor may be corrected by installing inductors in the electrical
circuit or by using the inductance created by the operation of motors in
other areas of the treatment plant. However, caution must be exercised
when using other plant equipment for power factor correction, because of
the variable operating conditions of the equipment from hour to hour and
from day to day. The inductance of other equipment should be used only to
control the low power factor of the ozone generator when consistent
equipment operation can be assured.
An operating ozone generator can decrease the power factor to 0.3 to 0.5,
depending on the generator setting. Corrections normally will be cost-
effective, since utilities that supply electrical power typically impose
penalties for a low power factor.
103
-------
Ozone Generators
The most common commercially available ozone generators can be classified
as follows:
o horizontal tube; one electrode water-cooled,
o vertical tube; one electrode water-cooled,
o vertical tube; both electrodes cooled,
o plate; water- or air-cooled.
The operating conditions of these ozone generators can be subdivided as
follows:
o low frequency (50/60 Hz), high voltage (> 20,000 V),
o medium frequency (600 Hz), medium voltage (< 20,000V),
o high frequency (> 1,000 Hz), low voltage (< 10,000 V).
Currently, low frequency - high voltage units are most common, but recent
developments in electronic circuitry are resulting in higher frequency units
(and lower voltages) being used.
To determine the amount of ozone required to provide disinfection, the
following steps are recommended:
1. Estimate the initial ozone demand of the water at the point of entry into
the first chamber of the ozone contactor. The initial ozone demand is
that quantity of ozone necessary to develop a stable residual ozone
concentration at the outlet of the first contact chamber. In most ozone
disinfection processes, the ozone residual attained is 0.4 to 0.5 mg/L, and
the initial ozone demand is between 1 and 6 mg/L, depending upon the
water quality at this point.
2. If the initial ozone demand is estimated, assume a 90% ozone transfer
efficiency for the 2-chamber bubble diffuser contactor.
3. An additional amount of ozone will be required in the second chamber of
the ozone contactor to maintain the residual of ozone for the appropriate
period of time to attain the necessary CT value. Assume this amount of
additional ozone will be 50% of that applied in the first chamber.
4. Determine the peak water flow through the plant, and from this, calculate
the peak ozone requirement.
In operating ozone treatment systems, a single generator designed for peak
use is never installed, in order to minimize energy costs during operation,
provide backup capability, and to facilitate maintenance operations. The
minimum production rate from an ozone generator is 10 to 25 percent of its
104
-------
maximum production capability. On the other hand, the most cost-effective
production rate of an ozone generator occurs at 60-70% of its maximum
production rate.
Therefore, if the treatment plant requires, say, 100 Ibs/day of ozone during
normal production, and 150 Ibs/day for peak operation, it is cost-wise to
purchase three 60 Ibs/day generators, and operate all three at about 65% of
their capacity for normal production. This approach allows satisfying the
occasional peak ozone demands, provides one standby generator during
normal flows, and allows the down generator to be maintained on a periodic
schedule.
Qgone Contactors
The different types of contactors have been discussed earlier. Selection of
the specific type of contactor should be based upon the specific type of
ozonation task contemplated. For disinfection, the most commonly used
contactor is the two-chamber, porous plate or porous tube diffuser contactor
shown in Figure 5. Housings for such contactor systems normally are
constructed of concrete.
Destruction of Excess Ozone
Ozone destruction removes excess ozone in a contact basin off-gas prior to
venting or prior to recycle or reuse of the exhaust gas. Safety is the major
consideration. The maximum ambient air ozone concentration currently
allowed by the OSHA for exposure of human beings over an 8-h working day
is 0.0002 g/m3 (0.1 ppm by volume). This concentration is significantly less
that the ozone concentrations in contactor exhaust gases, normally greater
than 1 g/rn3 (500 ppm by volume).
The primary methods for treating excess ozone in the exhaust gases are:
thermal destruction (300-350°C for 3 sec), thermal/catalyst destruction, and
catalyst destruction (metal catalysts or metal oxides). Moist granular
activated carbon is used quite extensively in small scale European ozonation
systems generating g/h quantities of ozone (< 2 mgd water supplies; large
public swimming pools, etc.), but is not recommended (U.S. EPA, 1986, p.
137) for this purpose.
The most favored procedure currently involves passing the contactor exhaust
gases through a catalytic ozone destruction unit, which contains a
proprietary catalyst system based on manganese dioxide. Excess ozone is
converted to oxygen, which may be discharged safely to the ambient
atmosphere.
Ozone contactor exhaust gases are treated in this manner, rather than
drying or recirculating through the ozone generator for economic reasons.
It is more cost-effective to dry ambient air than to clean and redry the wet
air exiting the ozone contactor.
105
-------
Materials of Construction
Care must be taken in selecting materials of construction for those portions
of the ozonation system in direct contact with either "dry" (before the
contactor) or "wet" (after the contactor) ozone-containing gas. While
reinforced concrete is an appropriate material for ozone contactors, the
ozone-containing gas piping system should be 304-L or 316-L stainless steel
for dry and wet services, respectively.
Monitoring the Ozonation System Operation
Equipment should be provided to monitor the operation of the components of
the system. The minimum degree of instrumentation, all of which can be
provided by the ozone equipment supplier as part of the package units, is as
follows:
o Gas pressure and temperature devices at key points in the air preparation
system. Simple pressure gauges and mercury thermometers will suffice.
o Continuous monitoring of the dew point measuring device to determine
the moisture content of the dried air feed gas to the ozone generator.
High dew point indications should be designed to sound an alarm and shut
down the generator. Equipment for calibration of the dew point monitor
should be provided as well.
o Means of measuring inlet/discharge temperatures of the ozone generator
coolant medium (water and/or oil, or air) is required, as is a means of
determining whether coolant is actually flowing through the generator.
An automatic system shutdown should be provided if coolant flow is
interrupted or if its discharge pressure exceeds specified limits.
o A means of measuring flow rate, temperature, pressure, and ozone
concentration of the ozone-containing gas discharged from the ozone
generator is required to determine the ozone production rate.
o A means of measuring the power input to the ozone generator is required.
Ozone Equipment Suppliers
Table XXIV is a listing of the major suppliers of corona discharge ozone
generation equipment in the United States (Source: International Ozone
Association, Pan American Committee, Norwalk, CT). Each of these suppliers
provides the total ozonation system, including electrical power supply, air
preparation equipment, contacting apparatus, and contactor off-gas destruc-
tion devices, as well as the appropriate monitoring and automation
equipment.
106
-------
TABLE XXIV. MAJOR U.S. OZONATION SYSTEMS
SUPPLIERS
Name
BBC Brown Boveri Inc.
Capital Controls Co., Inc.
Emery Chemicals Inc.
Gould & Eberhardt Industries
Griffin Technics Corp.
Hankin Environmental Systems
Infilco Degremont Inc.
Mitsubishi, Electric Sales,
MEDAMA, Inc.
Ozone Research and Development
Corporation
PCI Ozone Corporation
Trailigaz Ozone of America
Welsbach Ozone Systems Inc.,
Division of Polymetrics, Inc.
Address
1460 Livingston Ave., North
Brunswick, NJ 08902
Box 211, 3000 Advance Lane,
Colmar, PA 18915
4900 Este Avenue, Cincinnati,
OH 45232
P.O. Box 190, Sutton Road,
Webster, MA 01570
178 Route 46, P.O. Box 330,
Lodi, NJ 07644
71 Route 206 South, Somer-
ville, NJ 08876
P.O. Box 29599, 2828
Emerywood Parkway,
Richmond, VA 23229
5757 Plaza Drive, Cypress, CA
90630-0007
3840 North 40th Avenue,
Phoenix, AZ 85019
One Fail-field Crescent, West
Caldwell, NJ 07006
1 Jenkintown Plaza, 211 West
Avenue, Suite 210, Jenkintown,
PA 19046
101 Nicholson Lane, San Jose,
CA 95134
ix. Costs of Ozonation Systems
Equipment Costs
These will include estimates for ozone generation equipment and ozone
contacting systems, both of which are supplied by the ozone systems
manufacturers. Water supply systems treating 0.5 mgd and less will require a
107
-------
daily ozone generation capacity from three to 21 pounds, and will be able to
dose ozone at average levels of up to 3-5 rag/L. At these production levels,
ozone normally will be generated from dried air, not oxygen, in order to
avoid the costs of oxygen generation, recovery and recycle equipment.
As mentioned earlier, the new 540-mgd Los Angeles drinking water plant
uses oxygen as the feed gas to the ozone generators. A new German ozone
generator, currently marketed in the U.S. by Capital Controls Co. (the
CAPOZON system) is designed to generate ozone from pure oxygen, in small
diameter dielectric tubes. Because of this new design, this system is claimed
to be capable of generating unusually high concentrations of ozone (14%) in
oxygen. This system may find application for drinking water treatment
plants in the future.
Ozonation equipment to be purchased includes the following:
o air preparation equipment (drying and chilling)
o ozone generator
o ozone contactor
o ozone destruction unit
o instrumentation and controls
For generation of large quantities of ozone (100 Ibs/day and higher),
approximately $1,200 per pound of ozone generation capacity per day wul be
required to procure the air preparation, ozone generation, and contacting
equipment. Ozone destruction and instrumentations and controls will be in
addition to this figure.
For smaller quantities of ozone, costs will be higher, but will vary
significantly from site to site.
For the small production quantities of ozone required by small water
treatment plants serving less than 10,000 persons per day (three to 21
pounds per day), all items except the contactor can be assembled into a
single skid-mounted unit. If the contactor selected is the turbine type, it
also can be included in the skid-mounted assembly unit.
Diffuser contactors for small ozonation systems generally are constructed of
polyvinyl chloride (PVC) pipe standing on end, or of fiberglass reinforced
plastic (FRP) tanks. A contact chamber containing diffusers should be
approximately 18-ft high, providing a water depth of 16-ft and a detention
time of 10-20 minutes, to attain the appropriate CT value required for 99.9%
inactivation of Giardia lamblia cysts and/or 99.99% inactivation of enteric
viruses. These conditions will maximize the transfer of ozone from gas
phase to aqueous solution when employing diffuser contacting systems.
Tables XXV and XXVI list equipment cost estimates obtained from two
ozonation systems suppliers in 1982, for various daily ozone generation rates.
Ozone supplier A (Table XXV) provided estimates for ozone dosages provided
estimates for ozone dosages of 3 and 5 mg/L at water flow rates of 500,000
gal/day, 350,000 gal/day, and 180,000 gal/day. This breakdown shows that
equip-ment costs for air pretreatment and ozone generation capacity
available from this supplier depend upon the dosage required at a particular
108
-------
TABLE XXV. COSTS OF OZOHATION EQUIPMENT FOR SMALL MATES SUPPLY SISTER (Company A - May 1982)
Size of water supply 500.000 gpd 350.000 gpd 180,000 gpd
Maximum dosage of ozone (mg/L)
at peak flow 53 53 53
Daily ozone requirement (Ibs) 21 14 14 7 75
Contact chamber diameter (14 ft
high, 4 compartments, 4 dif-
fusers, Derakane fiberglass
reinforced plastic) 6 ft 6 ft 5 ft 5 ft 4 ft 4 ft
EQUIPMENT COSTS
Air preparation + ozone
generation unit
Contact chamber with diff users
Monitoring Instrumentation
$31
$11
,500
,500
$25
11
,000
,500
$25
10
,000
,200
$22
10
,000
,200
$22
9
,000
,900
$19,500
9,900
1) Ozone in generator product
2) Ozone in ambient plant air
3) Ozone dissolved in water
4) Dew point monitor in air
preparation unit
$15,000 15,000 15,000 15,000 15,000 15,000
Ozone Destruction Unit
TOTAL EQUIPMENT COSTS
Power requirement kWh
$6,700
(10 cfm)
$64,700
13.3
5,000
(7 cfm)
$56,500
10.1
5,000
(7 cfm)
$55,200
10. 1
4,200
(3 cfm)
$51,400
5.0
4,200
(3 cfra)
$51,100
5.0
4,200
(3 cfm)
$48,600
3.65
-------
TABLE XXVI. COSTS OF OZONATION EQUIPMENT FOR SMALL WATER SUPPLY SYSTEMS (Company B - May 1982)
Flow Rate 0.1 mgd 0.2 mgd 0.3 mgd 0.4 mgd
Maximum ozone dosage ,
(mg/L), at peak flow 3333
Daily ozone requirement
(Ibs/day) 3 6 7 12
Equipment Costs LP** HP*** LP HP LP HP LP HP
Air preparation +
ozone generator* - $17,500 §33,200 $30,200 $38,500 $35,500 $43,000 $40,000
Power requirements
(kWh/lb of 03
generated) 10.5 20 10.5 13.5 10., 5 13.5 10.5 13.5
Ozone contactor
with diffusers $8,500 $12,000 $16,000 $21,000
Ozone Monitors $4,000 $4,000 $4,000 $4,000
-f 0$ generation**** or 2,000 or 2,000 or 2,000 or 2,000
- O3 room or contact
chamber exhaust 2,200 2,200 2,200 2,200
- dew point 3,500 3,500 3,500 3,500
0. 5 mgd
3
14
LP HP
$49,800 $46,800
10.5 13.5
$29,000
$4,000
or 2,000
2,200
3,500
TOTAL EQUIPMENT
COSTS - $33,700- $54,900- $51,900- $64,200- $61,200- $73,700- $70,700- $88,500- $85,500-
35,700 52,900 49,900 62,200 59,200 71,700 68,700 86,500 83,500
* includes air preparation, ozone generation, ozone destruction and system controls.
** air preparation unit includes air filters or separators, compressor delivering air at 8-12 psig to a
refrigerative cooler and a dual tower desiccant dryer.
*** same as LP air preparation system, except compressor delivers air at 80-120 psig. HP system takes less
space requires less maintenance, but requires more energy.
**** $4,000 instrument IB an automatic, continuous reading in-line aonitor.
$2,000 instrument is not automatic and utilizes wet chemistry.
-------
water flow rate. In addition, the size (and cost) of the ozone destruction
units required also varies, as does the power requirement to operate the
total ozonation system.
Ozone Supplier A can provide four monitors with his system. All are
optional, but a!3 are recommended for optimal performance and minimal labor
and downtime. These will monitor:
o the dew point in the air preparation unit,
o ozone output of the generator,
o ozone in the plant ambient air (in case of leaks),
o dissolved ozone residual in the water.
The cost of these four monitors is constant at $15,000, regardless of system
size in the range shown in Table XXVI (5 to 21 pounds per day).
Table XXVI shows similar data for Ozone Supplier B. In this case,
equipment costs are presented for water flows of 100,000, 200,000, 300,000,
400,000, and 500,000 gal/day. Average ozone dosages are taken as 3 mg/L,
and the daily ozone output required varies from 3 Ibs/day for treating
100S000 gal/day to 15 ibs/day for treating 500,000 gal/day.
Ozone Supplier B offers two types of air preparation equipment, however,
and estimates are presented for each. One type operates at high pressures
(80-120 psig), and the other at low pressures (8-12 psig). The high pressure
air treatment units are lower in capital cost, but require more energy for
their operation.
Ozone Supplier B does not normally provide a residual dissolved ozone
monitor, but offers two types of monitor for ozone output from the
generator. The automatic, m-line continuous reading monitor costs $4,000;
the non-automatic monitor requires wet chemistry determinations to .develop
data at some period of time after the sample has been taken, and costs
Therefore, cost data presented in Table XXVI vary by the differences
between costs for high and low pressure air preparation equipment, and by
the costs of the two ozone generator monitors.
In 1987, Ozone Supplier A estimated that the cost for a 5-lb/day ozone
generator, sufficient to treat 180,000 gal/day at an applied ozone dosage of 3
mg/L, to be $20,500. This confirms that costs for ozone generation
equipment in 1983 are still current, for the smaller water treatment plants.
Installation Costs
Costs for Installation of the ozonation equipment include labor and material
costs for piping water to and from the ozone generators (if they are water
cooled), for piping ozone-containing air to the contactor chamber, for piping
water to and from the contactor, and for piping contactor off-gases to and
from the ozone destruction unit. Electrical wiring costs also must be
considered in these costs. Ozonation equipment suppliers contacted advise
111
-------
that for production of up to about 30 Ibs/day of ozone, installation costs
will be roughly the same, and will average 15% to 25% of the equipment
costs of the largest units estimated in Tables XXV and XXVI. The actual
figures for the two equipment suppliers then become:
Supplier Cost of Equipment Installation Cost
(500.000 gpd plant) (15% to 25%)
A $64,700 $ 9,705 - $16,175
B $85,000 $12,750 - $21,250
Housing Costs
The power supply, air preparation equipment, ozone generation equipment,
and turbine contacting units can be installed rela-tively easily, in areas on
the order of 10 x 17 feet. However, diffuser contacting units are tall (18
feet) and bulky, and normally are installed outside existing buildings (above
ground) or underground inside buildings being constructed. Alternatively, a
170 ft2 Butler building can house the ozonation system, except for above-
ground diffuser unit. Such a building costs about $6,000.
Construction Costs for Ozone Contactors
Construction costs for ozone contactors are based upon concrete contactors
which use serpentine flow in an upflow/downflow configuration. Ozone
dissolution equipment is located at the bottom of the downflow pass, thus
providing countercurrent flow. The conctruction costs include the ozone
dissolution equipment and manways which can be used for access to the
interior of the contactor. Figure 24 shows the construction costs for ozone
contactors having volumes between 2,000 and 18,000 gal (Gumerman et al.,
1986).
Construction Costs for Ozone Generation and Feed Systems
Construction costs for ozone generation and feed systems using air as the
feed gas include air preparation equipment, the ozone generator, ozone
destruction equipment for contactor off-gases, and a building to house the
equipment. For oxygen feed installations, the construction costs include the
ozone generator, ozone off-gas destruction equipment, and the building.
Ozone generators and the contactors are housed in a sealed room, which
contains an ozone leak detector. In the event of an ozone leak, this
detector will activate an alarm and the building ventilation system, and shut
down the ozone generator.
Piping is included for the ozone generator cooling water, ozone, and water
to and from the ozone contactor. Off-gases from the ozone contactor are
destroyed using a combination thermal/catalytic destruct device.
112
-------
CO
O
O
O
u
c
(0
O
O
1(
i
6
5
4
3
2
1
6
8
4
3
2
1(
\
9
4
3
2
1C
9
4
3
2
1<
I
1
),000,(
000, 0<
30.000
).000
)00
)00
00
i
**>
MMM
*
*
\
r-
_
!
J,
^
-i
^1
!
I
i
I
i
J_J—
J-^^
*> ^
1
2 349 9789 2 3 4 9 9 789 2 3 4 9 9 789
00 1000 10.000 100.000
CONTACTOH VOLUME • gal
1 » ' '1 • '
1 10 100
CONTACTOR VOLUME
Figure 24. Construction cost for ozone contactors (Gumerman et al, 1986).
113
-------
The building ventilation system consists of an intake louver and motorized
damper on one side of the building, and a fan, ductwork, and motorized
damper on the opposite side of the room. This system is designed to change
the room air once every two minutes, when activated. The ventilation
system can be activated by the ozone detector, a switch activated by
opening the building door, or a switch located on the building exterior. A
shutdown switch for the ozone generator is also located on the building
exterior.
Construction costs are presented in Figure 25 for ozone generation and feed
systems (Gumerman et al., 1986).
Operation and Maintenance Costs
Operating costs for ozonation systems vary, and depend upon a number of
factors:
o method of air preparation,
o method of cooling the generator (water or air),
o if water-cooled, the amount of refrigeration required for cooling water,
o method of contacting,
o dosage of ozone required,
o pumping of generator coolant,
o method of contactor off-gas destruction.
Air Preparation: High pressure versus low pressure, versus sub-atmospheric
pressure desiccant system with or without addition of air chiller.
Ozone Generator Cooling: Air versus water. If water, the amount of
cooling required. In northern climates, water produced at the plant
generally is cold enough to serve as the generator coolant the year round.
In southern climates, generator cooling water must be refrigerated most, if
not all, of the year.
Method of Contacting: Diffuser contactors require no added energy.
Ozone/air mixtures normally are generated under a sufficient pressure to
overcome the head of 16 feet of water. On the other hand, turbine
diffusers require energy for their operation, but take up much less space
than diffuser contactors. For disinfection with ozone, however, the
appropriate reaction time must be provided after initial contacting has been
achieved to attain the appropriate CT value to assure inactivation of Giardia
lamblia cysts and/or enteric viruses.
Contactor Off-Gas Destruction: Thermal versus catalytic destruction.
Operating costs of these techniques vary.
114
-------
JOO.OQO.OOO
MAXIMUM OZONE FEED RATE - Ib/day
1.0
10
100
MAXIMUM OZONE FEED RATE • kg/day
Figure 25. Construction cost for ozone generation and feed systems
(Gumerman et al., 1986).
115
-------
Maintenance material requirements are for periodic equipment repair and
replacement of parts. Air preparation systems contain air prefilters which
must be replaced frequently. Tube-type ozone generators normally are shut
down once per year for cleaning of the tubes and other general maintenance.
This can require several man-days of time, depending upon the number of
ozone generators in the system. Spare parts normally consist of replacement
tubes, which can be broken during cleaning, or which can deteriorate after
years of operation at high voltages, or more rapidly if the air is improperly
treated.
Labor requirements are for periodic cleaning of the ozone generation
apparatus, annual maintenance of the contacting basins, and day-to-day
operation of the generating equipment (average 0.5 h/day).
Operating and maintenance costs for equipment of Ozone Suppliers A and B
are summarized in Table XXVII. Also included are building heating costs
(which are taken to be the same up to 0.5 mgd) and costs for maintenance
materials and O&M labor. There are no chemicals costs related to ozone
generation, except for periodic changing of desiccant in air preparation
systems (normally after 10 years of use).
Electrical energy is a major component of operating costs, representing 26%
to 43% of total O&M costs at small plants (0.1 mgd) increasing to 59% to
65% at the larger plants (0.5 mgd).
Building energy costs (which are approximately the same for all small size
plants up to 0.5 mgd) include energy costs for heating, lighting, and
ventilation. Labor costs (which are independent of the plant sizes listed-
0.1 to 0.5 mgd) account for 54% to 70% of total O&M costs at the small
plants, but only 30% to 36% at the 0.5 mgd plant.
Figure 26 shows operation and maintenance requirements for ozone
generation, feed, and contacting systems ~ building energy, process energy,
and maintenance material for systems feeding 5 to 250 Ibs/day of ozone.
Figure 27 shows O&M requirements for labor and total O&M cost. Both
figures are taken from Gumerman et al. (1986).
Summary Statement Regarding Costs for Ozonation Systems
Because of the many differences in methods of air pretreatment, ozone
contacting, contactor off-gas destruction, monitoring, and other operational
parameters, equipment costs given above should not be considered as more
than general guidelines. Vendor quotes should be obtained at the time ozone
is being considered by the small water supply system.
It should be noted that vendor quotes obtained for estimating purposes are
likely to be somewhat higher than firm bids made to specifications. This is
because the market for ozonation systems currently is quite competitive, and
suppliers usually bid their best prices when responding to clear specifica-
tions.
116
-------
TABLE XXVII. OPERATING AND MAINTENANCE COSTS FOR SMALL OZONE SYSTEMS APPLYING 3 mg/L DOSAGE
Water Flow
Rate (ngd)
Supplier A
0.18
0.35
0.50
Supplier B
0.10
0.20
0.30
0.40
0.50
Electrical
Building*
6,
6,
6,
(High
6,
6,
6,
6,
6,
570
570
570
Pressure
570
570
570
570
570
Energy
Process
6,661
12,775
51,611
(kWh/yr)
Total x
13,231
19,345
58,181
Air Preparation)
21,900 28,470
29,565
34,493
59,130
68,985
36,135
41,063
65,700
75,555
$0.
$
$1
$4
$1
$2
$2
$4
$5
07/kWh -
926
,354
,073
,993
,529
,874
,599
,289
Maintenance
Material
$120
200
300
$120
120
200
250
300
Labor x
(hrs/yr)
250
250
250
250
250
250
250
250
$10/hr -
$2
2
2
2
2
2
2
2
,500
,500
,500
,500
,500
,500
,500
,500
Total Cost
($/yr)
$ 3,546
4,054
6,873
$ 4,613
5,149
5,574
7,349
8,089
Estimated from data of Hansen et al. (1979)
-------
100.000 10.000.000
10 100
AVERAGE OZONE FEED RATE • kg/day
Figure 26. Operation and maintenance requirements for ozone generation,
feed, and contacting systems -- building energy, process
energy, and maintenance material (Gumerman et al., 1986).
118
-------
1,000,000 1.000.000
"|F^ ai i :£=FF3=
a
9
4
3
2
1(
9
6
9
4
i 3
* *
ta
§ 1C
< 4
H
O 3
K
2
11
T
9
4
3
2
1C
I 8
6
9
4
3
2
10.000 1(
6
9
4
3
2
»t
>.
1,000 £ 1(
^ CC *
"* O
W
—
«
mm
LJ
sS
/
^
rfi
^
^
uan
T01
t
jg
*
1"
*
V
EX
^'
?'
\
*T
PAL O&M COSTS
CLUOING CHEMICALS]
I 1
OXY(
3.S kw
M^^tf—
ff
* 8
3
r=
IR F
kwl
3£f
h/l
EE
i/IE
4 F
I
D
i C
:E
3;
)Z
H
'C
-
O
D
>r
h
N
4E
f
E"
IR FEED_ ' |-
kwh/ib OZONE
f
I
ABOR
1
1
FIELD DATA LEGEND
A « LABOR - STRASBURG
2 3493 7S9 2 3490 789 2 3496789
10 100 1000
AVERAGE OZONE FEED RATE • Ib/day
I . • • , . , o j r i ,.,,,, . . .
1.0 10 100
AVERAGE OZONE FEED RATE • kg/day
Figure 27.
Operation and maintenance requirements for ozone generation
feed, and contacting system -- labor and total O&M cost
(Gumerman et al, 1986).
119
-------
c. Ultraviolet Radiation
Much of this discussion is taken from U.S. EPA, 1983, and 1986.
The effectiveness of UV radiation as a bactericide and virucide has been
well established. It is a physical disinfecting agent compared to the
chemical disinfectants, chlorine, chlorine dioxide, ozone, and monochloramine.
Radiation at a wavelength of 254 nm penetrates the cell wall and is absorbed
by the cellular nucleic acids. This can prevent replication and cause death
of the cell. Since UV radiation is not a chemical agent, no toxic residual
can be produced. Although certain chemical compounds can be altered by
the radiation, the energy levels used for disinfection are far too low for this
to be a significant cause for concern.
Major advantages of UV radiation are its simplicity, lack of impact on the
environment and aquatic life, and minimal space requirements. Required
contact times are very short, on the order pi seconds rather than minutes.
The equipment is simple to operate and maintain, but fouling of the quartz
sleeves or Teflon tubes must be dealt with on a regular basis. Fouling
normally is handled by mechanical, sonic, or chemical cleaning. High
concentrations of suspended solids, color, turbidity, and soluble organic
matter in the water can react with or absorb the UV radiation, reducing the
disinfection performance. <
i. General Description of the UV Process
Disinfection by UV radiation relies on the transference of electromagnetic
energy from a source (lamp) to organism cellular material (specifically the
cell's genetic material). The lethal effects of this energy result primarily
from the cell's inability to replicate. The effectiveness of this radiation is a
direct function of the quantity of energy, or dose, absorbed by the organism.
This dose is described by the product of the rate at which energy is
delivered, or intensity, and the time to which the organism is exposed to
this intensity.
The primary artificial source of UV energy, at present, is the low pressure
mercury arc lamp. It is almost universally accepted as the most efficient
and effective source for disinfection systems application. The primary
reason for its acceptance is that approximately 85% of its energy output is
nearly monochromatic at the wavelength of 253.7 nm, which is within the
optimum wavelength range of 250 to 270 nm for germicidal effects. The
lamps are long [standard lengths are typically 0.75 and 1.5 m (2,5 and 4.9 ft)
arc lengths] thin tubes [typically 1.5 to 2 cm (0.6 to 0.8 in) in diameter].
The radiation is generated by striking an electric arc through mercury vapor;
discharge of the energy generated by excitation of the mercury results in
the emission of the UV radiation.
The UV demand of a water is quantified by a spectrophotometric
measurement at the key wavelength of 253.7 nm; this expresses the
absorption (or transmittance) of energy per unit depth. The output is
120
-------
absorbance units/cm, or a.u./cm. The percent transmittance can be
determined from this unit by the expression:
% Transmittance = 100 x 10-(a-u-/cm)
The term most often used for design purposes is the UV absorbance
coefficient, a, expressed in base e:
UV absorbance coefficient, a = 2.3 (a.u./cm)
The unit for a is cm'l.
A second major concern is the provision of adequate exposure time to the
microorganisms in order to meet the dose requirement at a given intensity.
The key is to have plug flow through the system, such that each flow
element resides in the reactor for the same amount of time. Perfect plug
flow is not possible to achieve, of course. Some dispersion will exist, such
that there will be a distribution of exposure times about the ideal,
theoretical exposure time. A design objective is to minimize this
distribution.
The basic premise to understand is that the UV radiation must be absorbed
before it can have a disinfecting effect. Photochemical damage occurs to
the deoxyribonucleic acid (DNA) macromolecules which interferes with the
ability of the cells to replicate. Cell death following UV radiation is almost
entirely attributable to the photochemical damage of these compounds.
On the other hand, if the amount of radiation received by the organism is
not a lethal dose, but only damaging, photoenzymatic repair can occur, and
the effects of the UV radiation can be reversed. The phenomenon has been
termed "photoreactivation". This repair mechanism is unique to UV
radiation, but is not universal, and there is no clearly defined delineation of
characteristics which suggest which species would have the ability to repair
and which would not.
The enzymatic mechanism generally involved in photoreactivation requires
subsequent (or concurrent) exposure to light at wavelengths between 300 and
500 nm. Such light is available in sunlight and in most incandescent and
fluorescent light sources.
Therefore, to assure that disinfection occurs with no chance for photoreac-
tivation, attention must be paid to designing the UV disinfecting system to
provide sufficient UV dose for an appropriate period of time. As a general
rule of thumb (U.S. EPA, 1986, p. 215), if a 3-log reduction of micro-
organisms is required to meet the disinfection criteria, the system should be
designed to provide a 4-log reduction, to account for the effect of
photoreactivation.
Table XXVII lists some of the major suppliers of UV disinfection equipment
for water treatment systems.
121
-------
TABLE XXVII. MAJOR SUPPLIERS OF UV DISINFECTION
^ EQUIPMENT
Name Address
Aquafine Corp. 25230 W, Ave Stanford, Valencia,
CA 91355
Aquionics Inc. 21 Kenton Lands Road, Box 18395,
Erlanger, KY 41018
Atlantic Ultraviolet 250 Fehr Way, Bay Shore, NY
11706
Ultra Dynamics Corp. 1631 Tenth St., Santa Monica, CA
90404
Ultraviolet Purifications 299 Adams St., Bedford Hills, NY
10507
Ultraviolet Technology Inc. 8930 Osage Ave., Sacramento, CA
95828
ii. UV Disinfection System Design
In all, the design of a UV system must accommodate a few simple
considerations:
o satisfy the UV demand of the water,
o maximize the use of the UV energy being delivered by the lamps, and
o provide the conditions which encourage plug flow.
UV lamps can be suspended outside the liquid to be treated or submerged in
the liquid. In either design, the intent is to get the energy into the liquid
as efficiently as possible. Typically, if the lamp is to be submerged in the
liquid, it is inserted into a quartz sleeve to minimize the cooling effects of
the water. Lamps can be placed in the liquid with the lamp perpendicular to
the direction o! water flow. Other configurations may have the lamp
parallel to the flow, or the lamp may be suspended above the flowing liquid.
As the lamp emits radiation, the intensity will attenuate as the distance from
the lamp increases; this is simply due to the dissipation or dilution of the
energy as the volume that it occupies increases. A second attenuation
mechanism involves the actual absorption of the energy by chemical con-
stituents contained in the water. This, analogous to the chlorine or ozone
demands, is the "UV demand" of the water.
122
-------
The key points to be addressed when evaluating or specifying the design of
a UV reactor are as follows:
1. Residence Time Distribution fRTD). This should be constructed at a
number of flow conditions for an existing system. It should also be required
when specifying commercial systems. The RTD provides key information on
the actual or anticipated hydraulic behavior of a reactor.
2. Plug Flow. As a guideline, an aspect ratio greater than 15 should be
incorporated into the reactor design. Maintenance of plug flow within a
reactor will be influenced by the approach and exit conditions. The design
should have minimal disturbances at the inlet and exit planes of the lamp
battery. Directional changes in the flowpath would best be made outside of
the lamp battery.
3. Dispersion Number: A key goal is to minimize the dispersion number, d.
A design goal should be to have d between 0.02 and 0.05. This would be
representative of a plug flow reactor with low to moderate dispersion.' This
can be accomplished by increasing the product of ux (u = velocity of water
as it travels through the reactor, in cm/sec; x = the average distance
traveled by an element of water while under direct exposure to UV radia-
tion, in cm), even in a system with a relatively high dispersion coefficient.
The designer should be aware, however, that extended lengths and higher
velocities will cause higher head losses. In certain situations, some
adjustment of the dispersion number may be necessary in order to meet
specific head loss requirements.
4. Turbulence: Radial turbulence is necessary due to the non-uniform
intensity field. The reactor design should induce an estimated Reynold's
Number greater than 6,000 at minimum flow. If possible, it would be
beneficial to confirm the laminar/turbulent flow transition velocity by direct
head loss measurements on the lamp battery.
5. Head Loss: Direct measurements should be required for full-scale modules
or scaleable pilot units as part of commercial equipment specifications.
These should be determined over a wide velocity range and should exclude
entrance and exit losses.
6. Effective Volume: Maximal use of the reactor lamp battery is essential
to keep the process cost-effective. This will be related directly to the
reactor's inlet and outlet design. The goal must be to have equivalent
velocities at all points upon entering and upon exiting the lamp batteries.
Stilling walls (perforated baffles), and weirs should be incorporated into
reactor designs to assure this. Guidelines for specifying commercial
123
-------
equipment should require the ratio 0/T to be greater than 0.9 and/or the
ratio of tp/T to be greater than 0.9.
iii. Estimating the Average Intensity in a UV Reactor
The second element of UV dose, after time, is the intensity of energy during
the exposure time. Intensity is the rate (or flux) of delivery of photons to
the target. In UV process design models, the rate of bacterial mactivation
is described as a function of the intensity. By this fact, it becomes
important to be able to quantify the intensity in a given system. The
intensity in a reactor is a function of the UV source (output), the physical
arrangement of the source relative to the water (the arrangement of the
lamps and their placement in or out of the liquid), and the energy sinks
present which will attenuate the source output before it can be utilized for
disinfection processes.
Lamps used in UV disinfection systems generally have lengths of
approximately 0.9 and 1.6 m. The arc length defines the active, light-
emitting portion of the lamp [0.75 m and 1.5 m (2.5 and 4.9 ft., resp.),
respectively]. The diameter of the lamp is small, typically 1.5 and 1.9.. cm.
The lamp envelope is made of fused quartz or other highly transparent (to
the 253.7 nm wavelength) glass material, such as Vycor. ;
In the quartz systems, the individual lamps are sheathed in quartz sleeves
only slightly larger in diameter (2.3 cm) than the lamp, and the entire
lamp/quartz bundle is submerged in the flowing liquid. In systems where the
water does not contact the quartz or lamp surface, separate conduits carry
the waters. The conduits are translucent to the UV light, with the lamps
placed near the outside conduit wall.
iv. Water Quality Considerations in the Design of a UV
Disinfection System
Without question, a major element in the effective design of a UV
disinfection system is a clear and concise understanding of the
characteristics of the water to be treated. These are directly related to the
degree of pretreatment the material will receive before the disinfection step
and will affect the sizing and performance of the disinfection system.
Necessary pretreatment ranges from very minimal, as in the case of a
groundwater which requires no treatment other than disinfection, to full
conventional treatment of a polluted surface water.
The four water quality parameters which most affect the design or
performance of a UV disinfection system are the flow rate, initial
microorganism density, suspended solids (or turbidity), and the UV
absorbance of the water at the point of application of UV radiation.
i
1. Flow Rate: The flow rate is set by design of the main plant and
projections of the hydraulic load to the plant. In evaluating the design
124
-------
requirements for the disinfection process, some consideration should be given
to the equalization effects of the treatment processes before disinfection.
This can have an effect on the sizing of the UV system.
Flow estimates should be for the design year of the plant. There should be
some knowledge of the progressive increase in the flows through the design
life of the plant in order to determine if the system can be phased in by
the addition of modules as the demand increases. Some consideration also
should be given to the hydraulic load to the unit.
For disinfection, average flows are not critical to the design sizing; rather
they are important to estimating average utilization of the system for
operation and maintenance needs. Peak flows should be used for sizing,
particularly reflecting diurnal variations.
2. Initial Microorganism Density: The performance of a UV disinfection unit
is related directly to the initial density of the indicator organisms. This is
not a parameter which is generally monitored at a treatment plant,
particularly one already employing chlorine for disinfection. In the case of
disinfection by UV, however, it is critical. Performance is given by the log
of the survival ratio, N/Ng, or by the number of "logs the density is
reduced.
3. Suspended Solids (Turbidity); From the development of the ,UV
disinfection model, it is clear that the occlusion of bacteria in the
particulates will have a significant effect on the design of a UV system. It
is recommended (U.S. EPA, 1986) that the turbidity measurement be used as
the primary indicator to quantify these particulates.
4. UV Absorbance: The one parameter which is solely in the venue of UV
disinfection is the "UV demand of the water. Some organic and inorganic
compounds in the water may absorb energy at the 253.7 nm wavelength.
This absorbance will affect the intensity of the radiation within the reactor;
in specific design situations, the level of absorbance will affect the sizing of
the system and possibly the configuration (spacing) of the lamps. The final
product of these calculations is the average nominal intensity as a function
of the UV absorbance coefficient.
v. System Design and O&M Considerations for the UV
Process
The design basis and the process elements which are key to the design of a
UV system are:
o the hydraulic behavior of the unit,
o calculation of the intensity in the reactor, and
125
-------
o generation of the appropriate water characterization data.
In this section are presented other peripheral topics which the designer (and
operator) must consider. These factors will affect the operation and
maintenance of the system, and the overall economics. Specifically, these
factors include the elements which affect the reactor intensity (lamp output,
lamp aging, and attenuation of intensity due to fouling of reactor surfaces),
and methods for monitoring for lamp aging and unit fouling.
Factors Affecting UV Intensity in a Reactor
UV Lamp Output: Output at any given time will be influenced by lamp
temperature and by the voltage potential applied across the lamp.
Additionally, output at the resonant frequency will always degrade with time
of operation due to any number of "aging" factors. In the submerged
systems, it is not practical under most design conditions to control the lamp
temperature. In the non-contact systems, such as the tubular arrays, it is
possible to maintain the lamps at their optimum wall temperature by
controlling the temperature of the ambient air surrounding the lamps. This
is being practiced currently in commercial applications. Heat given off. by
the lamp ballasts is circulated into the lamp reactor in cases where heat is
required; otherwise fans vent the reactor with cooler outside air. These
operations are controlled thermostatically.
Voltage: Radiance is a function of the arc current. This can be exploited
by adjusting the voltage, in order to vary the output of the lamp.
Decreasing the voltage will result in a decrease in the current. Such a
control mechanism has been installed at full-scale facilities as an energy
conservation measure. During periods of low UV demand, the lamps are
"dimmed" by slowly turning the lamp supply voltage down. This results in a
reduction in the power draw of the lamp. Generally, the lamp intensity can
be reduced to levels no less than 50% before the lamp current becomes too
low and the lamp will begin to flicker and eventually turn off.
Lamp Aginp: A number of factors combine to effectively age a lamp and
limit its useful life. These include failure of the electrodes, plating of the
mercury to the interior lamp wall (blackening), and solarization of the lamp
enclosure material (reducing its transmissibility). These all cause a steady
deterioration in the lamp's output at the 253.7 nm wavelength, such that its
output at the end of the lamp s life can be 40 to 60 percent of its nominal
output. ,
The output of the lamp throughout its life is affected primarily by the
extent of blackening and solarization of the glass tube; the actual life of the
lamp is governed by the condition of the electrodes. Germicidal lamps are
typically of the hot cathode, type; these will deteriorate progressively with
increasing number of starts. Thus, the lamp life expectancy generally is
rated according to the number of times the lamp is started, or the burning
cycle. The lamp life normally cited by most manufacturers is 7,500 hours,
126
-------
based on a burning cycle of eight hours. The average UV output at this
point is estimated to be 70% of the lamp output at 100 hours (note that the
nominal output of the low pressure mercury arc lamps represents its output
after a 100 hour "burn-in" period).
Monitoring the Lamp Intensity
The procedure for monitoring the lamp intensity is by comparison with that
of a new lamp, but after the 100 hour burn-in time. The first step is to
measure the intensity, at a fixed distance, of three to five new lamps which
have been burned for about 100 hours. The average of the five then
becomes the benchmark to determine the relative output of the lamps in use
(percent of new lamp average). Each lamp should be tagged and given an
I.D. number; this allows direct monitoring of individual lamps and allows the
operator to keep an appropriate mix of lamps in a system and to know when
to discard a lamp.
The same procedure is used to monitor the transmittance of a quartz sleeve.
In this case, a single lamp is used. First the intensity is measured with and
without a new, clean quartz sleeve in place over the lamp. Similar
measurements then are taken with the unit's quartz sleeves and compared to
the transmittance of the new quartz. This can be done before and after \ the
quartz is cleaned.
vi. System Design Considerations for Effective
Maintenance
An overriding concern in the proper maintenance of the UV reactor for
effective performance is to keep all surfaces through which the radiation
must pass as clean as possible. The effects of surface fouling on energy
utilization efficiency are critical, and very often can be pointed to as the
reason for non-performance of a particular system. Other concerns relate
primarily to the accessibility to UV reactors and to keeping adequate records
to control replacement cycles and maintenance schedules.
Reactor Maintenance
The most reliable method to determine if a reactor is becoming dirty and
requires cleaning is by visual inspection. The unit should be drained and the
surfaces observed for fouling. In open systems, this can be done rather
conveniently and quickly. Reactors which are sealed vessels can be difficult.
These designs should accommodate such visual inspections by incorporating
large portholes or manways in the reactor shell.
Generally the surfaces of submerged quartz systems contacted by the water
will become coated by inorganic scale, very much like boiler scale. This will
be especially the case in areas where there is hard water. Additionally, the
inside surface of the quartz and the outer surfaces of the Teflon tubes
127
-------
eventually will develop a grimy dust layer, primarily from airborne dirt and
water vapor.
Fouling of the reactor's internal surfaces will be signaled by reduced
performance efficiency, or by reductions in the intensity measured by in-
line probes. While these may provide some signal of fouling, it is still
necessary to be able to inspect the surfaces physically.
It is appropriate to completely overhaul the reactor, cleaning all interior
surfaces, and determining the lamp outputs and quartz (or Teflon)
transmittances. Each lamp is removed from the reactor and washed with a
mild soap solution, rinsed, and swabbed with an alcohol (isopropyl) soaked
rag (cheesecloth). Then the interior surfaces of the quartz sleeves are
cleaned by the same procedure by using a gun-barrel type cleaning rod to
swab the interior surfaces. At the same time, each lamp, which is tagged
with an I.D. number, is measured for relative output. Those which are below
a specified level are discarded and replaced with new lamps. These new
lamps also are tagged with a number. In this manner, each lamp can be
traced on the basis of operating time and output. A reactor lamp inventory
then can be mixed and controlled to maintain a minimum average output
level.
In similar fashion, the quartz should be monitored for transmittance. It ;may
be cumbersome, however, to remove all the quartz from a system. It is
recommended instead (U.S. EPA, 1986, p. 233) that a representative fraction
of the quartz sleeves be monitored; 10-15% of the quartz inventory would be
sufficient. The same quartz should always be monitored; these would be
considered as representative of all quartz in the system. If the tagged
quartz begins to show marked deterioration due to aging and wear, it then
may be appropriate to broaden the monitoring and to begin replacing the
quartz sleeves. The replacement can be accomplished gradually. As with
new lamps, eventually there will be a mix of old and new quartz in the
system. There is little experience in determining the effective life of the
quartz sleeves; certainly it will vary by site, but generally should be between
four and seven years.
In Teflon systems, the lamps are removable on racks; they should be cleaned
and monitored in the same manner as the quartz systems. The Teflon tubes
should be cleaned on occasion; this can be done by swabbing the tubes with
soapy water/alcohol. A non-abrasive material should be used. Each tube
also should be monitored for transmittance, just as with the quartz sleeves.
This may not be as straightforward, however, because of their limited
accessibility and problems in getting direct measurements with a UV
radiometer/detector.
vii. Costs for Disinfection With UV Radiation
Construction Costs
Table XXVIII summarizes costs developed in 1978 by Hansen et al. (1979).
By 1982, these costs had increased about 15% (U.S. EPA, 1983) for the size
128
-------
TABLE XXVIII. CONSTRDCTION COSTS FOR ULTRAVIOLET LIGHT DISINFECTION
Plant Capacity (gpd)
Cost Category
Excavation and
Sitework*
Manufactured
Equipment
Concrete*
Labor*
t~) Pipe and Valves*
SO
Electrical and
Instrumentation*
Hous ing*
SUBTOTAL
Miscellaneous and
Contingency*
14,400
$ 60
800
250
110
60
430
1.500
3,210
470
28,800
$ 60
1,125
250
170
150
430
1,500
3,885
560
187,200
$ 60
4,485
250
250
350
430
1.500
7,225
1,010
374,400
$ 60
8,685
250
300
450
430
1.500
11,675
1,580
748,800
z$ 80
17,365
280
400
750
480
1.800
21,155
2,830
1,123,200
$ 110
26,050
300
500
1,000
480
2.000
30,440
4,060
TOTAL $3,680 $4,445 $8,335 $13,255 $24,085 $34,500
Data from Hanaen et al. (1979).
-------
ranges listed. Data presented are for single and multiple UV lamp
disinfecting units ranging in water throughput capacity from 14,400 gal/day
to 1,123,200 gaVday.
All UV generating units are quite compact; for example, the 1,123,400
gal/day unit occupies an area of less than 24 ft2. Costs listed in Table
XXVIII include equipment costs of the UV units, and the related costs of
piping, electrical equipment, equipment installation, and a building to house
the equipment.
Figure 28 shows construction cost curves for UV disinfection systems
(Gumerman et ah, 1986).
Operating and Maintenance Costs
These are shown in Table XXIX (for 1979) for the same size plants as in
Table XXVIII. Process energy is required for the mercury lamps operating
inside of the UV generating units. Continuous 24-hr/day operation is
assumed, with only occasional shutdown to clean cells and replace UV lamps
which have become weakened by lengthy use. Building energy requirements
are for heating, lighting, and ventilation.
Maintenance materials are related to the replacement cost of the UV lamps,
which usually are replaced after operating continuously for about 2,000 hours
(about eight months). Labor requirements are related to occasional cleaning
of the quartz sleeves which surround the mercury vapor lamps, and periodic
replacement of the weak UV bulbs.
It is noteworthy that replacement bulb costs at the smallest plant (14,400
gal/day) are only about 9% of the total O&M costs, whereas at the largest
plant size (1,123,200 gal/day), replacement bulb costs are about 48% of the
total O&M costs. This reflects the fact that the larger UV generating units
contain a greater number of UV bulbs per unit volume of water treated.
Figure 29 plots estimated labor requirements for the operation and
maintenance of UV systems (Scheible et al., 1985). Figure 30 shows
operating and maintenance requirements for UV disinfection systems-
building energy, process energy, and maintenance materials -- treating 5,500
gal/day to 1 million gpd. Figure 31 shows O&M requirements for labor and
total O&M cost. Figures 30 and 31 are taken from Gumerman et al. (1986).
2. Secondary Disinfectants
There are disinfectants which are added in small dosages to waters which
have been treated with primary disinfectants. The purpose of secondary
disinfectants is to maintain a small residual disinfectant in the distribution
system.
130
-------
1
7
5
4
3
2
9
I
8
5
4
*» 3
CO
O
0 1C
M>r- w *-J
rfOliOOHiSNOO
*
4
10
1
3.000.C
000.00
10.000
1.000
00
100
0
•^
.Ml
I-*""
^
^
+
5
"
ml
•••i^H
^
X
^j
4
i*T
di
i •
2 349 8788 2 3 4 9 9 789 2 3 4 8 8 788
000 10.000 100,000 1,000.001
PLANT CAPACITY - gal/day
10
1 I
100 1000
PLANT CAPACITY • m^/day
Figure 28. Construction cost for ultraviolet light disinfection (Gumerman
et al, 1986).
131
-------
TABLE XXIX. OPERATION AND MAINTENANCE SUMMARY FOR ULTRAVIOLET LIGHT DISINFECTION
to
Plant Flow
Rate (gpm)
14,400
28,800
187,200
374,400
748.800
1,123,200
Energy (kWh/yr*
Building
10,260
10,260
10,260
10,260
12,310
13,340
Process
440
800
5,260
10,510
21,020
31,540
Total x
10,700
11,140
15,520
20,770
33,330
44,880
$0.07/kWh -
$ 749
780
1,086
1,454
2,333
3,142
Maintenance
Matl.*($/yr)
$ 100
140
600
1,120
2,250
3,300
Labor
(hr/yr)
24
24
24
30
36
42
Total Cost
x $10/hr = ($/yr)
$240
240
240
300
360
420
$ 1,089
1,160
1,926
2,874
4,943
6,862
* Data from Hansen et al. (1979).
-------
1000
BOO
600
400
200
"5"
I 100
I BO
£ 60
40
10
30
60
Approximate Numtwr ol Limpi (1.5m Arc)
120 300 600 1200
3000 6000
I
I
I
NoM: Labor Bl»d on
Ton! Sytum KW.
I I I
I
10 20 40 60
Svtltrn Si», Total KW
100
200
400 600 1000
Figure 29. Estimate of labor requirements for the O&M of UV systems
(Scheible et ah, 1985).
In point of fact, there are three secondary disinfectants in prevalent use:
chlorine, chlorine dioxide, and monochloramine. Considerations pertaining to
chlorine as a secondary disinfectant are essentially the same as were
discussed earlier for chlorine as a primary disinfectant.
a. Chlorine Dioxide
Chlorine dioxide (C1C<2) is an unstable, greenish-yellow gas, explosive in air
in concentrations above 4%. Because of this instability, it is always
generated in solution, on-site, and is used immediately without storage. As
long as care is taken to keep chlorine dioxide in solution and storage of
solutions is avoided, there will be no explosion hazards. Chlorine dioxide is
readily soluble in water and is decomposed by sunlight.
133
-------
10.000
1,000,000
MAINTENANCE MATERIAL
PROCESS ENERGY
S&^SSftJHH
T 2
1000
345 67SS
10.000
2 3 4 S 0 ?8®[ 1
100.000
3 4 86788
1.000.000
PLANT CAPACITY • gel/day
10
I
100
1000
PLANT CAPACITY • m3/day
Figure 30. Operation and maintenance requirements for ultraviolet light
disinfection -- building energy, process energy, and
maintenance material (Gumerman et al, 1986).
134
-------
JOQ.OOQ
| 2343 eras 2 3 4 s e 719 2 3 4 s«7Si
1000 10.000 100,000 1.000.000
PLANT CAPACITY - gal/day
T
TT
10
100 1QQO
PLANT CAPACITY • n»3/day
Figure 31. Operation ar\d maintenance requirements for ultraviolet light
disinfection -- labor and total O&M cost (Gumerman et al.
1986). *'
135
-------
The material is a more powerful oxidizing agent and a better biocide than is
chlorine. In addition, when chlorine dioxide is prepared in the absence of
excess free chlorine, its use will not produce trihalomethanes, or other
chlorinated organic by-products of current public health concern.
Additionally, chlorine dioxide has been used in pretreatment to oxidize
phenolic compounds and to separate iron and manganese from organic
complexes which are stable to chlorination. No oxidation of bromide ion to
produce hypobromous acid has been observed using chlorine dioxide free of
excess free chlorine.
Distribution system residuals of dissolved chlorine dioxide are longer-lasting
than those of chlorine, because there is no reaction with ammonia or
formation of chlorinated organic materials. Additionally, chlorine dioxide is
not known to impart tastes and odors to water, as does chlorine.
Recent health effects studies have shown (NAS, 1987) that chlorine dioxide
produces hematological effects in both humans and laboratory animals. For
these reasons, the U.S. EPA (1981) currently advises that the total
concentration of chlorine dioxide and its decomposition products (chlorite
and chlorate ions) be maintained below 1 mg/L. This allows a maximum
applied dosage of 1.2-1.4 mg/L (Werdehoff & Singer, 1986).
Gaseous chlorine dioxide has a strong, disagreeable odor, similar to that of
chlorine gas, and is toxic to humans when inhaled. It is detected by the
human nose at concentrations between 1.4 and 1.7%. When present at 4.5%
concentrations, it irritates the respiratory mucous membranes and may cause
severe headaches. At concentrations below 6% in air, it may be compared
with chlorine with respect to its toxicity (Masschelein, 19/9). Eventual
intoxications appear by local irritations of the nervous system, ocular and
respiratory mucous membranes, without substantial resorption or systemic
poisoning (Ehrlicher, 1964). There are no cumulative effects in cases of
repeated exposure (Haller & Northgraves, 1955).
i. Generation of Chlorine Dioxide
For drinking water treatment, chlorine dioxide is generated from solutions of
sodium chlorite, NaClO2- This material, purchased as a solid (80% NaClO2)
or most normally, as a 25% aqueous solution, is treated with aqueous
solutions of chlorine or hypochlorous acid), and sometimes with a strong
mineral acid such as sulfuric or hydrochloric acids.
Three processes used in water treatment plants for the generation of C1C>2
employ (a) gaseous chlorine, (b) sodium hypochlorite solution and mineral
acid, and (c) mineral acid. Each process is summarized below.
1. Gaseous Chlorine
This is a two-step procedure, beginning with the formation of hypochlorous
acid upon dissolution of gaseous chlorine into water:
136
-------
C12 + H2O > HOC1 + HC1
chlorine water hypochlorous hydrochloric
acid acid
These intermediate products then react with sodium chlorite to produce
chlorine dioxide:
HOC1 + HC1 + 2NaClO2 ~-> 2C1O2 + 2NaCl + H2O
hypochlorous hydro- sodium chlorine sodium water
acid chloric chlorite dioxide chloride
acid
The overall result of these reactions is summarized by the equation:
C12 + 2NaClO2 —-> 2C1O2 + 2NaCl
chlorine sodium chlorine sodium
chlorite dioxide chloride
According to the stoichiometry of this equation, one mole of chlorine rfeacts
with two moles of sodium chlorite to produce two moles of chlorine dioxide.
In water supply practice, some molar excess of chlorine actually is employed
so as to insure conversion of the maximum amount of chlorite ion to
chlorine dioxide. Therefore, the recommended ratio of reactants is two
moles of chlorine per mole of sodium chlorite. On a weight basis, 1.57 parts
of chlorine gas are added per part of NaQO2 (calculated on a 100% solids
basis when solutions of sodium chlorite or 80% solids materials- are
employed).
Under these conditions of excess chlorine gas being added, the product C1O2
solution also will contain an amount of hypochlorous acid/hypochlorite ions.
These can react with THM and TOX precursor materials to produce THMs
and TOX materials.
2. Sodium Hypochlorite and Mineral Acid
This also is a two-step process, in which sodium hypochlorite solution reacts
with hydrochloric acid to form hypochlorous acid, which then reacts with
sodium chlorite to form chlorine dioxide:
NaOCl + HC1 --> NaCl + HOC1
sodium hydrochloric sodium hypochlorous
hypo- acid chloride acid
chlorite
137
-------
HOC1
hypo-
chlorous
acid
HC1 + 2NaC102 --> 2C1O2
2NaCl
hydro-
chloric
acid
sodium
chlorite
chlorine sodium
dioxide chloride
H2O
water
In this procedure, as in the gaseous chlorine and sodium chlorite procedure,
excess chlorine is utilized to insure conversion of the maximum amount of
chlorite ion to chlorine dioxide. The C1O2 solution so produced also will be
able to form some THMs and TOX because of the presence of excess
hypochlorous acid.
3. Mineral Acid
This process involves mixing a solution of acid with a solution of sodium
chlorite, and the reactions are as follows:
5NaClO2
sodium
chlorite
4HC1 --> 4C1O2
SNaCl
hydrochloric
acid
chlorine sodium
dioxide chloride
2H2O
water
When sulfuric acid is the strong mineral acid, the reactions can be depicted
as follows:
10NaClO2
sodium
chlorite
8C1O2 + 5Na2SC>4 + H2O + HC1
sulfuric
acid
chlorine
dioxide
sodium
sulfate
water hydro-
chloric
acid
The exact ratios of reactants will depend upon which mineral acid is
employed for the production of chlorine dioxide.
When generated using mineral acid, excess chlorine is not required.
Consequently, solutions of chlorine dioxide prepared in this manner do not
contain free residual chlorine, thus no more than trace amounts of
chlorinated organic by-products can be synthesized. In addition, pure
chlorine dioxide does not oxidize bromide ion to produce hypobromous acid,
which then can produce brominated organic derivatives. With excess free
chlorine available, however, bromide ion can be oxidized in water to form
hypobromous acid.
In all three cases, the appropriate aqueous solutions of reactants are metered
into a chlorine dioxide reactor (a cylinder containing Raschig rings, glass
beads, or hollow glass cylinders) where intimate mixing of the reacting
solutions occurs (see Figures 32 and 33). The size of the reactor and the
residence time of the reacting solutions are such that after a few seconds,
the solution exiting the reactor displays a strongly yellow color of chlorine
dioxide. This solution then is pumped directly into the water to be treated.
138
-------
Water flow mew
4 - 20 nA
signal
o
CIO,
I
Chlorine dioxi<$»
gentrating lower
Hatch ig
ring*
c t
I >
}
c >
c >
"" V"
f,^^
--*
1 stroke
Mixer Sodium
chlorite
tolution
Point of
application
Injector water supply -
Figure 32. Schematic diagram of an automatic feed, automatic flow-
proportional chlorine dioxide system: generation from chlorine
and sodium chlorite (courtesy Capital Controls Co., Inc.).
In this manner, solutions of chlorine dioxide are generated as the material is
required and used immediately, without storage. Appropriate metering and
control instrumentation can be installed with the chlorine dioxide reactor so
that the generation and addition of chlorine dioxide is paced by the flow
rate of the water to be treated. As a result, the unit operates without the
need for constant manual attention.
ii. Oxidation-Reduction Reactions of Chlorine Dioxide
At neutral pH (= 1) chlorine dioxide dissolves totally in water without
hydrolysis reactions. On the other hand, when performing its function as a
chemical oxidant or disinfectant in water, chlorine dioxide is chemically
139
-------
reduced. One of the reaction products is the chlorite ion, from which CIO?
was formed initially:
C1O2 + e-
chlorine reducing
dioxide agent
(C102)'
chlorite
ion
Chlorin* dioxidt
gerurtting toww
Figure 33. Manual feed equipment arrangement for generating chlorine
dioxide from sodium hypochlorite solution and mineral acid
(U.S. EPA, 1983).
In strongly acid ( pH < 2) or strongly alkaline (pH ^ 11) solutions, chlorine
dioxide undergoes disproportionation (self oxidation and reduction):
2C1O2 + H2O ---> (C1O2)-
chlorine
dioxide
water
chlorite
ion
(cio3)-
chlorate
ion
2H+
hydrogen
ion
140
-------
These disproportionation reactions are accelerated by the presence of
hypochlorous acid and hypochlorite ion. Therefore, when chlorine dioxide is
generated in the presence of excess chlorine and at pH below 2 or above 11,
chlorite and chlorate ions will be formed as well
iii. Establishing a Chlorine Dioxide Residual
Because the cost of chlorine dioxide is higher than that of chlorine, and
because of the hematological effects upon both humans and laboratory
animals, only small dosages of chlorine dioxide (maximum residuals of 1 mg/L
total of C1C«2, chlorite, and chlorate ions) currently are recommended by EPA
for drinking water treatment (U.S. EPA, 1981; 1987a). However, a primary
advantage of pure chlorine dioxide is that it does not react with ammonia or
with THM precursors to produce THMs, or with TOX precursors to produce
TOX, as does chlorine. This means that if the water has been pretreated to
remove most of the oxidant-demanding constituents (or does not contain
them initially), the total C1O2 dosage can be used to provide primary or
secondary disinfection at the recommended 1 mg/L maximum oxidant residual
level.
t
In laboratory studies, Werdehoff & Singer (1986) have shown that about 70%
of the C1O2 added is converted to chlorite ion. Therefore a chlorine dioxide
dosage of 1.2 to 1.4 mg/L is the maximum practical dosage in order not to
exceed the recommended 1 mg/L residual of total oxidant. This value has
been confirmed by Lykins & Griese (1986) in pilot plant and full-scale plant
studies at Evansville, IN (see Case History -- Evansville, IN, Section
VIII.C.1).
Another major advantage of pure chlorine dioxide is that it does not oxidize
bromide ion to produce hypobromous acid. This means that bromine-
containing THMs and TOX derivatives will not be formed when this material
is used for oxidation/disinfection.
iv. Factors Affecting the Efficiency of Secondary
Disinfection With Chlorine Dioxide
Chlorine dioxide is a more effective disinfectant than chlorine or
hypochlorous acid. Because it does not react with water, ammonia, bromide
ions, or most organic nitrogen compounds, it is not "wasted" in extraneous
reactions of this type. It is less sensitive to changes in pH (except at very
low and very high values), maintaining its disinfection capabilities over the
pH range of 6 to 10.
On the other hand, because it is a powerful oxidizing agent, chlorine dioxide
can and will react with oxidizable impurities contained in a raw or treated
water. Thus, it is important to ensure that oxidant-demanding components
of the water have been removed to as low a level as is feasible, consistent
with the costs involved, before chlorine dioxide is added.
141
-------
v. Chlorine Dioxide Systems Design
Table XXX is a partial listing of current suppliers of chlorine dioxide
equipment. Several different types of generation equipment are available,
which vary depending upon the supplier, but also upon the generation
process chosen (gaseous chlorine versus sodium hypochlonte plus acid, versus
acid plus sodium hypochlorite, for example). Pertinent aspects of each type
of chlorine dioxide generation equipment will be discussed in this subsection.
TABLE XXX. PARTIAL LISTING OF CHLORINE DIOXIDE
EQUIPMENT SUPPLIERS
Name Address
Capital Controls Co., Inc. P.O. Box 211, Colmar, PA
18915
CIFEC 10 Avenue de la Porte
Molitor, F-7500, Paris, France
Clow Corporation 408 Auburn Ave., Pontiac, MI
48058
Fischer & Porter Co. County Line Road, Warminster,
PA 18974
Olin Water Services Co. 9393 W. 110th St., Overland
Park, KS 66210
ProMinent Fluid Controls 1005 Parkway View Drive,
Pittsburgh, PA 15205
Rio Linda Chemical Co., Inc. 410 North Tenth Street,
Sacramento, CA 95813
Wallace & Tiernan Division 25 Main St., Belleville, NJ
Pennwalt Corp. 07109
1. Gaseous Chlorine Plus Hvpochlorite Solution
All three of the chlorine dioxide generation procedures described earlier are
in general use in the United States at this time for application in small
water treatment facilities. The most commonly used procedure involves
addition of gaseous chlorine to sodium chlorite solutions. A schematic
diagram of chlorine/sodium chlorite equipment is shown in Figure 32.
The gaseous chlorine procedure is particularly applicable when a gaseous
chlorination system already exists at the treatment plant. The reaction by
which chlorine dioxide is generated is as follows:
142
-------
+ 2NaClO2 —-> 2C1O2 + 2NaCl
chlorine sodium chlorine sodium
chlorite dioxide chloride
This equation indicates that 71 Ibs (26.48 kg) of chlorine mixed with a
solution containing 181 Ibs (67.5 kg) of 100% NaQC>2 will produce 135 Ibs
(50.36 kg) of chlorine dioxide. However, the ratio of reagents recommended
by most suppliers of chlorine dioxide generating equipment is 1:1 by weight.
This means that more than double the stoichiometric amount of chlorine
required by the above equation is utilized. This excess of chlorine over that
required results in a faster reaction rate and insures a more complete
conversion of chlorite ion to C1C>2.
However, because a large excess of chlorine is employed, solutions of
chlorine dioxide prepared by this technique also will contain some free
available chlorine, mostly as hypochlorous acid. This amount of free
chlorine will produce some quantity of trihalomethanes and other halogenated
(TOX) materials.
2. Hypochlorite. Chlorite, and Mineral Acid
The production of chlorine dioxide using sodium hypochlorite solution with
sodium chlorite and strong mineral acid is illustrated in Figure 33, and is
well suited to most small water systems. Dosages of each chemical can be
derived from the equations given earlier.
In this system, all three reactants are in solution. Utilization of strongs acid
increases the conversion of chlorite ion to chlorine dioxide. Solution feed
pumps of equal capacities can be used by adjusting the solution strength of
each of the reactants. Thus, the chlorine dioxide production and addition
rates can be paced by the flow rate of the water being treated and/or by its
secondary disinfectant demand.
Sodium chlorite is available in 55-gal (208.33 L) drums, either as a solid (80%
active NaClO2) or as a solution containing 30% NaClO2 (33% solids). If not
used directly from the drum, sodium chlorite solution is stored in polyfvinyl
chloride) (PVC) or fiberglass tanks and transferred by means of PVC, rubber,
or Tygon tubing systems. Diaphragm pumps incorporating PVC as the
material in contact with the solutions are used for pumping sodium chlorite
solutions. Provision must be made for immediate washdown of any spills of
the chemical. This precaution is generic to all chlorine dioxide generating
systems.
3. The CIFEC System for Generating Chlorine Dioxide
A schematic diagram of this system, developed in France, but in use at
several U.S. water treatment plants, is illustrated schematically in Figure 34.
The system produces C1O2 from gaseous chlorine, in high yield, and
containing little excess free chlorine.
143
-------
00 2 EXIT
VMUUH UMC Qf CHLORINE
CHLORINATOR
EJECTOR WITH CHECK WIVE ASSEMBLY
SOOIUM CHLORITE METERING PUMP
ELECTRIC VALVE
Figure 34. Schematic of the CIFEC chlorine dioxide generating system.
Gaseous chlorine is passed into water which is circulated continuously in
what is referred to as an "enrichment loop". Under these conditions,
dissolved chlorine (hypochlorous acid) concentrations become higher than can
be achieved in a single pass. As a result, the pH of the hypochlorous acid
solution is lowered to below 4. This solution then is pumped into the C1O2
reactor along with a solution of sodium chlorite. As long as the pH of the
hypochlorous acid solution is below 4.0, conversion of chlorite ion to
chlorine dioxide is significantly higher than the single pass method employing
elemental chlorine gas. Therefore, chlorine dioxide is produced which is free
of significant quantities of free chlorine.
144
-------
4. Rio Linda Chlorine Dioxide Generator
Figure 35 shows a schematic diagram of a newer acid/chlorite chlorine
dioxide generator marketed by Rio Linda Chemical Co., Inc. Chlorine dioxide
is generated by addition of dilute hydrochloric acid to sodium chlorite
solution. The novel principle of this generator is the mixing of acid with
concentrated sodium chlorite solution just before the two solutions reach the
reactor. The two solutions are brought together in an eductor by means of
a vacuum created by water flow through the eductor. Such a system
eliminates a pump and allows the system to occupy a smaller space.
FLOW DIAGRAM
REACTION
COLUMN
FLOW RATE
METERS
SODIUM
CHLORITE
SUPPLY
Figure 34. Schematic of the Rio Linda Chemical Co. acid/NaC!O2 chlorine
dioxide generator.
vi. Miscellaneous Comments
Because several types of chlorine dioxide equipment are available, as well as
three processes for its production, it is considered inappropriate to attempt
to provide detailed instructions in this document for the preparation of
chemical solutions and feed rates. However, the small water utility choosing
to install chlorine dioxide generating equipment can have confidence that
145
-------
each equipment vendor will provide detailed recipes for preparing and
metering the appropriate solutions to his chlorine dioxide reactor so as to
produce an aqueous solution of chlorine dioxide of known and constant
concentration for addition to the plant process water.
A final point to be noted is that the currently recommended maximum
concentration of total oxidants (chlorine dioxide, chlorite and chlorate ions)
of 1 mg/L means that a water utility processing 0.5 mgd (1,893 m3/day) and
dosing a maximum of 1 mg/L of chlorine dioxide will require a maximum
ClO^ production rate of only 4 Ibs (1.82 kg)/day. Smaller systems will
require even less C1C«2. At such low dosage levels, two of the three vendors
of chlorine dioxide contacted recommend that their units be operated inter-
mittently, collecting C1C»2 solution in an enclosed holding tank for metering
into the water being processed. This is because at the low flow rates of
reactant solutions, mixing is less efficient in the chlorine dioxide reactor.
Consequently, conversion of chlorite ion to chlorine dioxide is less efficient.
vii. Costs of Chlorine Dioxide Generating Systems
Hansen et al. (1979) summarized costs for the generation of chlorine dioxide
from equal parts of 2.4% sodium chlorite solution, 25% sulfuric acid solution
and 1% sodium hypochlorite solution. Suppliers contacted in 1982 had
changed the design of their generation systems for small water supply
systems to use 33% hydrochloric acid rather than 25% sulfuric acid.
Equipment costs estimated by Hansen et al. (1979) assumed the use of a dual
head diaphragm pump for simultaneous addition of hypochlorite and acid
solutions, and a single head pump for the addition of sodium chlorite
solution. Detention time in the chlorine dioxide reactor is estimated at 12
seconds, and the generating equipment costs are assumed to be constant up
to 50 Ibs/day of chlorine dioxide.
At a maximum chlorine dioxide dosage rate of 1 mg/L, a 1 mgd water
treatment plant would dose 8 Ibs/day. At the same dosage level, a 2,500
gal/day water facility would require only 0.2 Ib/day of chlorine dioxide.
Equipment Costs
Quotes were obtained (U.S. EPA, 1983) from three suppliers of chlorine
dioxide generation equipment sized so as to prepare C1C>2 at the rate of 8
Ibs/day (for a 1 mgd water treatment plant). These are shown in Table
XXXI. Supplier A's recirculating loop system (CIFEC) is the highest in
equipment price; their lowest cost unit was priced at $34,000 in 1982. This
unit operates with a special recirculating pump designed to handle
hypochlorous acid below pH 4, plus a sodium chlorite solution pump and all
necessary instrumentation to allow automatic operation, with shutdown
provisions in the event of cessation of water flow.
146
-------
TABLE XXXI. 1982 VENDOR QUOTES -- CHLORINE DIOXIDE
GENERATORS
Vendor
Recircu-
lating loop
Supplier A
(French)
C102
production
capacity
flbs/davl
1-10
space
reauired*
2x3x6 ft
Reactants
Cb gas +
NaClO2 solu-
tion
Unit Cost
$34,000 (1
rate, adjust
manually)
dt^rft f\f\/\ /*\
automatically)
(prices delivered to New York)
Supplier B 4 3.5x4x1.5
ft (wall-
mounted
automatically)
HC1, NaOCl
solutions
Supplier C 14-140
Supplier C 14-140
4x3x1.5 ft Cb gas +
Nad
37.5x27x
6.5 in.
1O2 soln.
same
4x3x1.5 ft HC1 +
NaClO2 soln.
37.5x27x same
6.5 in.
rates (adjust
$41,700 (3
rates (adjust
$25,000
(installed)
$ 4,320**
(floor mount)
$ 3,600**
(wall mount)
$ 4,320
(floor mount)
$ 3,600
(wall mount)
**
all units require additional space for solution tank(s).
this unit requires a chlorinator for operation, which is not
included in price estimates.
The next lowest in price is the system from Supplier B, which generates
chlorine dioxide from 33% hydrochloric acid, 12% sodium hypochlorite, and
25% sodium chlorite solution. This unit cost $25,000 (installed) in 1982, and
includes three solution pumps, water flow rate detector, and switches to shut
down the unit if the water flow stops. This unit is wall mounted and
requires 3.5 x 4 ft of wall space, plus floor space for drums of the three
chemical solutions used to feed the generator. For volumes of C1O2
sufficient to treat flows in communities of 5,000 and 2,500 population, this
unit is said to be capable of continuous operation, with no loss in efficiency
of conversion of chlorite ion to chlorine dioxide. However, to supply the
needs of systems serving as few as 25 persons, the unit would have to be
operated intermittently, with C1O2 solution being stored in a holding tank
for later metering into the water.
147
-------
Supplier C provides two types of chlorine dioxide generators for small water
supply systems. One uses acid/sodium chlorite, the other uses chlorine gas
and sodium chlorite. These units cost $3,600 in 1982, if wall-mounted, and
$4,320 for a floor-mounted cabinet. The single size unit offered by this
supplier is designed to generate up to 140 Ibs/day. In order to produce 8
Ibs/day or less, a small water utility would have to install a holding tank
and operate the generator intermittently.
The chlorine gas/sodium chlorite generator of Supplier C requires a gas
chlorinator to feed chlorine gas. Therefore, in new plants considering use
of this type of equipment, the cost of a chlorinator must be added to the
cost of the chlorine dioxide generator. In existing plants currently using
gas chlorination, the chlorinator already is in place and would not represent
additional equipment cost.
Because equipment quotes for generating chlorine dioxide vary so widely,
water treatment personnel are advised not to try applying past equipment
cost estimates. Technology for generating and applying chlorine dioxide is
changing rapidly (as opposed to technologies for addition of gaseous or
aqueous chlorine), and new suppliers enter the market from time to time. It
is more advantageous to seek quotations from the various suppliers as to the
various methods for generating C1O2- Select the methods most appropriate
to the specific water treatment plant, then determine what piping and wiring
will be needed to install the equipment selected.
Figure 36 (Gumerman et al., 1986) shows that the construction costs for
chlorine dioxide generating and feed equipment are constant for feed rates
of up to about 45 Ibs/day.
Operation and Maintenance Costs
Hansen et al. (1979) concluded that, in general, O&M costs for generating
QO^ are independent of the quantities generated. Process energy
requirements, which are for metering pumps and mixer for preparing chlorite
solution from solid sodium chlorite, are estimated at 1,240 kWh/yr. Energy
requirements for 40 ft^ of building space to house the equipment would be
4,100 kWh/yr, resulting in total energy requirements of 5,340 kWh/yr. Main-
tenance material requirements would be for minor equipment repair only,
amounting to about $100/yr. Labor is required for preparation of solutions
and periodic maintenance of the equipment. Annual labor requirements are
estimated to be 1 h/day, or 365 h/yr.
Annual O&M costs of $4,124/yr (based on $0.07/kWh power cost and $10.00/h
labor cost) are summarized in Table XXXII. Figures 37 and 38 (Gumerman et
al., 1986) show the estimated O&M requirements for chlorine dioxide
generation and feed equipment - building energy, process energy, and
maintenance material (Figure 37) and labor and total O&M cost (Figure 38).
148
-------
(0
O
o
i
o
«E
H
(O
O
1.
7
6
S
4
3
2
1<
8
9
4
i
3
2
1(
i
a
9
4
3
2
t(
fl
9
4
3
2
10
0.
300.000
DO.OOO
»n™
WOO
)00
_
_
U-
T~~ i rrn
i
l
!!
I »
i
^ ? ll
_ ...I,,..., .
777"^
s
0
!
i
1
J,i -K,. ^
'
f~
'r
L,
hh
•MB
•••
___ i
•••
^B •!
2 3 496789 2 3 456789 2 3 49878
1 1.0 10 1
MAXIMUM FEED RATE • Ib/day
9
00
0.1
1.0
MAXIMUM FEED RATE • kg/day
10
Figure 36. Construction cost for chlorine dioxide generation and feed
equipment (Gumerman et al., 1986).
149
-------
10,000
10.000.000
FIELD DATA LEGEND
A « MAINT. MAIL. - VILLA RICA
• TOTAL ENERGY ~ VILLA RICA
C * TOTAL ENERGY - FAYETTEVILLE
MAINTENANCE MATERIAL
PROCESS ENERGY
BUILDING ENERGY
2 34$ 8788 2 3 4 8 i 7S9J 2 3 4 5 §788
1.0 10 1QO
AVERAGE FEED RATE - kg/day
I
0.1
I I
1.0 10
AVERAGE FEED RATE • Ib/day
Figure 37. Operation and maintenance requirements for chlorine dioxide
generation and feed equipment -- building energy, process
energy, and maintenance material (Gumerman et al., 1986).
150
-------
(0
O
O
o
H
1,000,000 1,000.000
1
0
5
4
3
2
1
5
4
3
2
1(
1
4
3
2
11
T
0
S
4
3
2
1C
'- f
0
5
4
3
2
00.000 1
9
: •
0
s
4
3
- 2
.s
),000 . i(
• K T
5 8
j 3
4
3
2
)00 1C
0
S
4
3
2
10 1C
I
.=-~__
00,000
3,000
"»••••
>00
)Q
sam
Lj-
BH
Bl
,
z±h
MB •
aaMBBM^
ri&n
HB
ma
••
- F
A
Ji
TO
Ci>
~T
•*"
I ELD DATA LEGEND
• LABOR - VILLA RICA
. LABOR - FAYETTEVILLE
TAL O&fc
(CLUD)N
^£0^
^P«-
*/
1 CC
3 C
onJUs
/&
»w
/
/
>ST
HEM
^E
LmJ— -
Lt
»,e
J
T\
iff
/
f
c
0
Al
R
.S]
2 343 9789 2 3 4 S 6 789 2 3 4 9 S 7S9
1.1 1.0 10 100
AVERAGE FEED RATE • its/day
0.1
1.0 10
AVERAGE PEED RATE • kg/day
Figure 38. Operation and maintenance requirements for chlorine dioxide
generation feed equipment - labor and total O&M cost
(Gumerman et al., 1986).
151
-------
TABLE XXXII. OPERATION AND MAINTENANCE SUMMARY FOR
C1O2 GENERATING AND FEED SYSTEMS
Item Requirements* Cost
ELECTRICAL ENERGY
Process 1,240 kWh/yr x $0.07 = $ 86.8
Building 4,100 kWh/yr x $0.07 = $ 287.0
TOTAL 5,340 kWh/yr x $0.07 = $ 373.8
MAINTENANCE MATERIAL $ 100/yr
LABOR 365 h/yr x $10.00 = $3,650
Total Annual O&M Cost $4.124
* based on estimates of Hansen et al., 1979
Chemical Costs
At a production rate of only 8 Ibs/day (maximum for a 1 mgd water
treatment plant at a 1 mg/L applied chlorine dioxide dose), chemical costs
are not as significant as pumping costs. Nevertheless, 1982 chemical costs
were as follows:
gaseous chlorine $0.47/lb
sodium chlorite 1.55-1.65/lb
HC1 0.10/lb
NaOCl 0.93/gal (15% solution)
Prices for chlorine and sodium chlorite in 1987 were about 30% lower.
A chlorine dioxide production rate of 8 Ibs/day equates to 2,920 Ibs/yr. If
the chemicals cost for C1O2 is arbitrarily assumed to be $l/lb, a 1 mgd
water treatment plant can expect to pay about $2,920 in addition to the
annual O&M costs. On the same basis, chemical costs at a 2,500 gal/day
plant would be $2,920/400 = $7.30/yr for producing chlorine dioxide at
$1.00/lb.
b. Chloramination
Chloramines are formed when water containing ammonia is chlorinated, or
when ammonia is added to water containing chlorine (hypochlorite or
hypochlorous acid). This is accomplished currently by adding gaseous
ammonia (purchased as the anhydrous liquid, NH3, in 150 Ib cylinders for
small water treatment systems) directly to the water, or by adding a solution
of ammonium sulfate, (NH4)2SO4 (purchased in 100 Ib bags, 98% pure; 25%
available NH3).
152
-------
i. Chemistry of Chloramination
Three chloramine compounds can be produced, depending on the ratios of
chlorine and ammonia which are utilized:
NH3 + HOC1 —-> H2O + NH2C1
ammonia hypochlorous water monochloramine
acid
NH2C1 + HOC1 —> H2O + NHC12
dichloramine
NHC12 + HOC1 —-> H2O + NC13
nitrogen
trichloride
The distribution of the chemical species of chloramines is a function of pH
and of the amount of chlorine added. For example, in the pH range of 7 to
8 and a chlorine to ammonia weight ratio of 3:1, monochloramine is the
principal product. At higher chlorine to ammonia ratios or at lower pH
values (5 to 7), some dichloramine will be formed. If the pH drops below 5,
some nitrogen trichloride (often erroneously called "trichloramine) may be
formed. Formation of this compound should be avoided because it imparts
undesirable taste and odor to the water.
Figure 39 (National Academy of Sciences, 1980) shows the relative
percentages of monochloramine and dichloramine produced as the pH changes,
for different weight ratios of chlorine to ammonia. At a pH value of about
5.7, approximately equal amounts of mono- and dichloramines are present in
solution.
Care also should be taken not to exceed chlorine to ammonia ratios of 5:1.
This is the ratio existing at the peak of the breakpoint curve, above which
all of the ammonia will have been removed, chloramines will be absent, and
free residual chlorine will be present.
ii. Establishing a Chloramine Residual
Generation of chloramines is conducted on-site, in solution, as required,
simply by adding the appropriate amount of chlorine to waters already
containing ammonia, or by adding ammonia to waters already containing
chlorine, then allowing a short holding time to be certain that the chemicals
have had time to react with each other to form chloramines. Usually,
chloramine-forming reactions are at least 99% complete within a few minutes.
153
-------
pH
Figure 39. Proportions of mono- and dichloramines in water with
equimolar concentrations of chlorine and ammonia (NAS, 1980).
However, the National Academy of Sciences (NAS, 1987) has recommended
rejection of the first approach (adding chlorine to ammonia-containing
water). Furthermore, when using the second approach, the NAS recommends
addition of sufficient chlorine to produce a slight residual of free chlorine
above that required to oxidize nitrogen (particularly the organic nitrogen
compounds), followed by addition of ammonia to form monochloramine.
iii. Chloramination System Design
Ammonia is available as the anhydrous gas (NH3), as a 29% water solution
(aqua ammonia), or in powdered form as ammonium sulfate (NH^SO^
Gaseous ammonia is supplied in 150-lb (68.1 kg) cylinders (for small water
systems), similar to gaseous chlorine. Aqua ammonia is supplied in 55 gallon
154
-------
f208.33 L) drums. Ammonium sulfate is available in 100-lb (45.4 kg) bags
(98% pure, 25% available ammonia).
Gaseous ammonia normally is added to the treated water using systems and
equipment similar to those used for gaseous chlorine. Aqua ammonia and
ammonium sulfate solutions are handled using systems and equipment similar
to those for sodium hypochlorite and calcium hypochlonte solutions,
respectively. Aqua ammonia is basic and odorous, but is non-corrosive.
Sizing of the treatment facility must take into consideration the intended 3:1
chlorine/ammonia ratio.
A 25% to 30% solution of ammonium sulfate in water is prepared in a plastic
or fiberglass container and added to the water by means of a chemical
metering pump. Solutions of ammonium sulfate are stable, but are acidic,
and therefore can be corrosive to some metals. Materials which will
withstand dilute sulfuric acid also will easily resist any possible corrosion
effects of dilute ammonium sulfate solutions.
iv. Costs for Chloramination
Generation of chloramines requires the same equipment for chlorination
(gaseous or aqueous hypochlormation) plus equipment for the addition of
ammonia (gaseous or aqueous). Costs for chlorination equipment and for its
operation and maintenance have been presented earlier. In this section,
costs for addition of ammonia are presented.
During January 1983, costs for liquid ammonia were $0.40/lb (in 150-lb
cylinders), $0.70/lb of contained ammonia in 28% solution (purchased in 55
gal drums), and $0.51/lb for solid ammonium sulfate (purchased in 100 Ib
bags), in the Washington, DC - Baltimore area.
Cost calculations given below are based on the following reaction of chlorine
gas and ammonia to produce monochloramine:
NH3 + C12 -—> C1NH2 + HC1
In addition, the calculations assume a dosage of 2.5 mg/L of chloramine as
the secondary disinfectant. This is the maximum level currently
recommended by the EPA to be dosed into water supplies (U.S. EPA, 1987a).
It is further assumed that the chloramines will be produced by adding
ammonia to water already containing free available chlorine. Finally, cost
calculations are based upon chlorine added as the gas. By using previously
described calculations involving solutions of sodium hypochlorite or of
calcium hypochlorite, the amounts of these chlorinating agents required to
produce monochloramine can be calculated readily.
155
-------
1. Costs for Monochioramine From Anhydrous Ammonia + Chlorine Gas
A 2,500 gal/day water treatment plant will require 0.05 Ib/day (23.66 g) of
monochloramine at a dosage of 2.5 mg/L. According to the stoichiometry of
the above equation for the generation of monochloramine from ammonia and
chlorine, each gram-molecular weight (weight of one mole of compound
expressed in grams) of monochloramine will require 1 gram-molecular weight
each of ammonia and chlorine. Thus:
17 g NH3 + 71 g C12 —-> 51.5 g C1NH2 + 36.5 g HC1
Since 23.66 g of monochloramine are required each day, the corresponding
amounts of ammonia and chlorine required are:
Ammonia: (23.66/51.5) x 17 = 7.81 g/day
Chlorine: (23.66/51.5) x 71 = 32.61 g/day
Dividing the grams of each reactant by 454 (the number of grams per
pound), gives a daily requirement of 0.017 Ib of ammonia and 0.0715 Ib of
chlorine. Multiplying each of these figures by 365 days yields the annual
number of pounds of ammonia and chlorine required. Finally, annual costs
for each are calculated by multiplying the annual requirements by the
current costs:
Ammonia: 0.017 Ib/day x 365 = 6.205 Ib/yr x $0.40/lb =
$2.48/yr
Chlorine: 0.0715 Ib/day x 365 = 26.10 Ib/yr x $0.47/lb =
$12.27/vr
The sum of these two numbers ($2.48 + $12.27) = $14.75, total annual costs
for the 2,500 gal/day facility.
A 1 mgd facility will require 400 times the amounts of chemicals at the same
2.5 mg/L dosage, therefore:
$14.75 x 400 = $5.898. annual chemical costs.
2. Costs for Monochloramine From Aqua Ammonia + Chlorine Gas
From the preceding calculations, the 2,500 gal/day facility will require 6.205
Ibs/yr of anhydrous (gaseous) ammonia. If the source of ammonia is 28%
aqueous ammonia, the calculation of costs is as follows:
1 gal of 28% ammonia weighs 8.34 Ibs and contains 8.34 x 0.28 = 2.34 Ibs
of ammonia.
6.205 Ibs/yr ammonia requires 6.205/2.34 = 2.55 gal/yr of aqua ammonia.
156
-------
At $0.70/lb of ammonia contained in aqua ammonia, the annual cost of aqua
ammonia is:
6.205 Ibs x $0.70/lb = $4.35/yr.
The annual cost of gaseous chlorine is $12.27/yr, therefore the total annual
chemical costs are:
$4.35 + $12.27 = $16.62/vr.
The 1 mgd water treatment plant will require 400 times the amounts of
chemicals:
$16.62/yr at the 2,500 gal/day plant x 400 = S6.648/vr at the 1 mgd plant.
3. Costs for Monochloramine From Ammonium Sulfate + Chlorine Gas
One pound of ammonium sulfate contains 0.2576 Ib (28.76%) of available
ammonia. The 2,500 gal/day plant using solid ammonium sulfate will require
6.205 Ibs of anhydrous ammonia annually. To obtain this amount of available
ammonia requires 6.205/0.2576 = 24.09 Ibs/yr of ammonium sulfate. At
$0.51/lb, the 2,500 gal/day plant will require:
24.09 Ibs/yr x $0.51/lb = $12.29/yr + $12.27 for chlorine = $24.55/yr.
The 1 mgd facility will require 400 times as much chemicals, or:
$24.55 x 400 = $9,820/yr.
3. Oxidants
a. Potassium Permanganate
Although potassium permanganate is not recommended as a primary
disinfectant for coping with Giardia cysts or enteric viruses, it can be quite
effective as a non-halogenated, preoxidizing agent for removal of many
tastes and odors, colors, iron, manganese, sulfide, nitrite, and many dissolved
organic materials. As such, the use of potassium permanganate in place of
prechlorination can be viewed as a procedure to control disinfection by-
products. The major supplier in the United States is Cams Chemical
Company, P.O. Box 1500, LaSalle, IL 61301 (815-223-1500). Literature
describing this versatile is readily available from this supplier.
This chemical is shipped as a solid, dark purple in color. It is readily
soluble in water, and solutions of 1-4% in water are recommended for the
treatment of potable water supplies. For example, to make a 4% solution
0.33 pound of solid potassium permangante are added per gallon of water]
and the mixture agitated 15-30 minutes using a 750-1000 rpm mixer.
157
-------
Addition of permanganate solutions to the raw water or in the rapid mix is
recommended for water treatment applications. This is because as the
permanganate ion performs its oxidative work, it is converted into lower
valent manganese compounds, which are insolble. Thus potassium
permanganate is always added ahead of filtration.
To feed 1 mg/L of permanganate to the raw water using a 4% solution, the
metering pump should be set to dispense 9.5 mL/min for every 100 gal/min
of flow.
Figure 40 shows construction costs for potassium permanganate systems
feeding 1 to 250 Ibs/day. Figure 41 shows operation and maintenance
requirements for potassium permanganate feed systems — process energy and
maintenance material. Figure 42 shows O&M requirements for labor and
total O&M costs. These three figures are taken from Gumerman et al.
(1986).
158
-------
(O
O
u
z
2
H
o
D
cc
z
O
o
Tf*]
i
e
5
4
3
2
1(
^^•v
6
S
4
2
7
6
S
4
3
2
I
1(
5
3
2
1C
•
000.00
JCLOOfi
p^
I 000
Hiaiiinrre
)00
)0
0
HBBB
wo
0*
pi
1C
0a •>
.--^
<*"
P0P
•**
UBMEK
I?
ra
i
i
i
0^
TT^P8^^^
^!^3L_ •
™»»
P»
r^jn_L_
.^^T^
~
i
I
"T"
2 34S67S9 2 3 4 5 S 7Si 2 3 4 S 6 7SS
.0 10 100 1000
MAXIMUM FEED RATE • ib/day
1.0 10 100
MAXIMUM FEED RATE • kg/day
Figure 40. Construction cost for potassium permanganate feed systems
(Gumerman et al., 1986).
159
-------
UJ
111
u
z
<
z
Ul
I-
z
"<
2
10
jj
e
s
4
3
2
1(
I
6
S
4
3
2
1(
7
8
S
4
3
2
T
6
S
4
3
2
0.
00 U
: 1
N- 6
S
4
3
2
>0 _M
i- 1 i 1 1 1 1 II
• kwh/yr
|^ r» u * « o»~4!C
E
*'
E MATERIAL
\^f\
NERGY
i rfi
^^
^^
-^~
_^Jra
U™i— .
j
— -L
1
nn,
2 345 6789 2 3 4 S 6 789 2 3 4 i S 789
.0 10 100 1000
AVERAGE FEED RATE • Ib/tiay
"T~' ' ' \ ' • • ' J " " ' | • « P r i i v . , , .
1.0 10 100
AVERAGE FE£D RATE - kg/day
Figure 41. Operation and maintenance requirements for potassium
permanganate feed systems -- process energy and maintenance
material (Gumerman et al., 1986).
160
-------
100,000 100,000
V)
o
o
2
•0
O
TOTAL O&M COST
(EXCLUDING CHEMICALS)
3 456789 2 3 496739 2 3 45370®
10 100 1000
AVERAGE FEED RATE • Ib/day
1.0
I
10
I
100
AVERAGE FEED RATE • kg/day
Figure 42.
Operation and maintenance requirements for potassium
permanganate feed ststems -- labor and total O&M cost
(Gumerman et al., 1986).
161
-------
VIII. CASE EXAMPLES OF EMERGING TECHNOLOGIES
Several examples of the use of disinfectants other than chlorine are
presented in this section. The objectives of the different approaches
described are (1) to attain the desired degree of disinfection and (2) to
minimize the formation of disinfection by-products. Some of the examples
illustrate changes in disinfection/oxidation techniques to cope with problems
in addition to disinfection by-products, e.g., taste and odor, color, algae, etc.
Case histories presented include examples of the use of ozone as a primary
disinfectant for controlling Giardia lamblia cysts at one plant and as a
preoxidant for lowering THM levels at another.
Another case history involves a surface water treatment plant which has
installed UV for primary disinfection, followed by chlorination for secondary
disinfection.
Other examples involve uses of chlorine dioxide for preoxidation and for
post-disinfection, of chloramines to lower THM concentration, and of
combinations of chlorine dioxide and chloramine to lower THM
concentrations.
A. OZONE CASE HISTORIES
1. North Andover, Massachusetts^
Ozone Disinfection for Giardia lamblia
a. The Problem
During early 1986, 18 cases of Giardiasis were reported in the North Andover
area. On May 1, residents there were instructed to boil their tap water for
at least five minutes while public health officials sought to locate and
correct the problem.
North Andover's water supply, Lake Cochichewick, was found to contain
Giardia cysts in samples taken 4n April 1986. Local officials confirmed that
the lake's watershed has a large muskrat population, and there has been
concern that residential septic systems were not working well enough to
prevent harmful effects in the lake.
author is indebted to Messrs. Ross Hymen and Paul Anderson of
the Massachusetts Department of Environmental Quality Engineering, Mr.
Tom Boshar of Lally Associates, the consultants on this project, and to Bill
Nezgod and Dr. Carl Nebel, of PCI Ozone Corporation, suppliers of the
ozonation equipment, for providing information for this case history.
162
-------
At the time of the Giardiasis outbreak, treatment of North Andover water
supply involved pumping of Lake Cochichewick water through two pumping
stations into the distribution system without filtration, but with chlorination.
Over the years, high humic concentrations have resulted in the formation of
a significant scale m the pipes, containing a significant amount of biofilm as
well.
Heavier chlorination overcame the immediate Giardia problem, but caused a
second water quality problem. Descaling in the old distribution system
piping released coliform organisms into the water supply. The presence of
coliforms caused the State DEQE to continue the rtboil water" notice until
the coliform organisms could be shown to be absent.
In addition, THM levels of the heavily chlorinated water rose to above 200
Aig/L. Also, residents began complaining about high chlorine concentrations.
Ozonation was investigated and was found to have promise.
Total coliforms in Lake Cochichewick raw water normally are between 50 and
500 per 100 mL; raw water turbidities normally are between 1 and 2 NTU.
b. The Interim Solution
In early October, 1986, two ozonation systems (ozone generators, contactors,
and exhaust gas destruction) began operating at the two Lake Cochichewick
pumping stations in response to the emergency situation. Two discrete
ozonation systems, one capable of generating 50 Ibs/day of ozone, the second
capable of generating 150 Ibs/day, were rented from the ozonation equipment
supplier for a period of time. Later, when the efficacy of ozonation was
proven, the town purchased and installed two 150-lb/day ozonation systems
in the two pumping stations.
Therefore, as an interim solution, the lake water is being treated with ozone
at the two pumping stations. Each pumping station handles 2.5 to 3 mgd.
At four points in the distribution system, chlorine is added to provide a
residual disinfectant.
The state DEQE provided emergency funding of $2.5 million, which provided
for the ozonation systems, for connecting pipelines to the three neighboring
communities, for relining pipes with cement, and replacing water mains.
Rental fees for the two ozonation systems totaled approximately $90,000.
The two 150-lb/day ozonation systems were purchased for $325,000. This fee
included the air preparation system, ozone generation, diffuser contactors,
and ozone destruction equipment, plus the appropriate instrumentation,
including a monitor for measuring residual ozone at the outlet of the contact
chamber.
These two ozonation systems were installed as a stop-gap measure to control
Giardia cysts until a proposed $10.5 million water 12-mgd plant is designed
and constructed, and which will provide complete treatment, including
filtration and ozonation (in mid-1989).
163
-------
The two ozone contacting chambers (each 10 ft wide x 20 ft long, with a
16-ft water depth) are designed with baffles, to provide each contactor with
five complete mix ports, with ozone being applied equally in each section.
Thus, there is plug flow throughout the ozone contactors. Applied ozone
dosages are 5 mg/L.
At the outlet of each contact chamber, the concentration of ozone lies
between 0.9 and 1.0 mg/L. System designers assume that the average ozone
concentration in each contactor is 0.5 mg/L. Total residence time of water
in the ozone contact chamber is 10 minutes at full pumping rate (summer
time). During winter, with lower water demand, pumping rates are reduced
by 50%, thereby doubling the residence time in the contactors to 20 minutes.
Water temperatures of the lake vary from about 5°C in winter to just under
20°C during summer. Therefore, the appropriate CT values listed in Table
IIA for 99.9% inactivation of Giardia lamblia range from 3 to 1.5 mg/L-min.
Assuming the average dissolved ozone concentration of 0.5 mg/L carried
through the 10 minute contact period, the summertime CT value attained is 5
mg/L-min.
During winter, when the contactor residence time is 20 minutes, the CT
value attained is 10 mg/L-min. Both of these CT values are well in excess
of the 3 to 1.5 mg/L-min required to guarantee 99.9% inactivation of Giardia
lamblia and 99.99% inactivation of viruses.
No filtration is provided during the current interim period. Instead, the
community is placing total reliance on ozone for primary disinfection plus
secondary disinfection with chlorine.
c. The Results
After approximately 90-days of ozone treatment, the State DEQE uncon-
ditionally lifted the boil water order, which had been in effect for nine
months. Not only are Giardia cysts absent from the North Andover water
supply, but also coliform organisms.
In addition, several additional benefits have been obtained as a result of
installing ozonation. Prior to ozonation being installed, THM values were in
the range of 8 to 120 /ig/L. Since ozonation has been installed, THM values
now are in the range of 1.1 to 2 A^g/L. In addition, the color of the treated
water has improved significantly (65% to 95% lower). Finally, taste and odor
levels in the finished waters are greatly improved.
d. For The Future
As indicated earlier, a new 12-mgd conventional treatment plant, including
ozonation and granular activated carbon adsorption, has been designed.
Preozonation will be applied before the rapid mix. After dual media
filtration, GAC adsorption is incorporated, followed by post-chlorination for
residual. In addition to providing an adsorption capability, the GAC step
164
-------
also will allow biological decomposition of the easily biodegraded organic
fractions of the water present at that treatment point.
Bids were let for this plant in the summer of 1987; the plant is scheduled to
be on-line by mid-1989.
e. Note - Sturgeon Bay. WI
A second plant is installing ozone specifically for controlling Giardia lamblia
contamination in Sturgeon Bay, WI. This is a groundwater plant, without
filtration, located on a Lake Erie shore, which has been shown to contain
Giardia cysts. Water flow at this treatment plant is 100,000 gal/day. Water
temperature is about 10°C the year round.
Two 4-minute residence time ozone contactors are being installed through
which the water flows sequentially. After contacting, the ozone-containing
water will be stored in a filling well. Thus, the total ozone contact time is
estimated to be 10 minutes. Dissolved ozone concentrations at the outlet of
both contactors will be adjusted to 1 mg/L.
Thus the CT value in the second ozone contact chamber will be 4 mg/L-min
(4 minutes times 1 mg/L), plus a conservatively estimated 2 mg/L-min in the
first chamber (0.5 mg/L times 4 minutes), plus an additional 1 mg/L-min in
the filling well, during ozone decay from 1 mg/L. The total CT value
designed is 7 mg/L-min. Table IIA requires a CT value of only 2.5 mg/L-min
for 10°C.
The retrofitted ozonation system is expected to be operating in April/May
1988.
2. Kennewick, Washington (Cryer, 1986) -
Preozonation For THM Control
a. The Problem
Prior to 1977, the City of Kennewick had been drawing essentially all of its
municipal water supply from a system of five Ranney collectors located along
the Columbia River, followed by chlorination. When initially installed, these
were capable of producing approximately 20 mgd; however, their output had
deteriorated to about 15 mgd by 1977. By 1978, the maximum daily system
demand essentially had reached the capacity of the Ranney system. It was
determined that direct utilization of the Columbia River would be the only
reliable source of supply for long term development.
A consequence of this decision was the need to provide consider-ably more
treatment to achieve the same or better finished water quality. Thus a pilot
plant study was undertaken to test alternative water treatment processing
steps. This study included the use of preozonation and of post-filtration
GAC adsorption, in addition to conventional and direct filtration procedures.
165
-------
Average values of the water quality parameters of the raw Columbia River
are:
TTHM 1
TTHMFP* 136 /ig/L
TOC 2.4 mg/L
No. Particles 11,650 /mL
Particle Volume 160,700 nanoL/L
Turbidity 1.7 NTU
Chlorine contact time: 1 week
b. Pilot Plant Study Results
Both the preozonation and coagulation/filtration steps provided approximately
30% reduction in levels of TTHM formation potential and 10% reduction in
TOC levels. The combined processes gave approximately 60% reduction in
levels of TTHMFP and 20% reduction in TOC levels. The
preozonation/coagulation/filtration processes was . determined to be
operationally equivalent to activated carbon adsorption for the removal of
THM precursors; it was also determined to be the most cost-effective method
of treating organics in the Columbia River water supply.
c. Plant Design
Approval was given for a 30 mgd water treatment plant to be constructed in
four stages of 7.5 mgd each, based on the following processing sequence:
preozonation/flash mix/coagulation/flocculation/filtra'tion and post-chlorina-
tion. Design criteria for the preozonation facilities are given in Table
XXXIII.
At the point of application, ozone dosage rates were 1.5 mg/L (average) and
4.0 mg/L maximum. The contactors provide 10 minutes of detention time.
Raw water total coliform levels are consistently less than 50 per 100 mL.
Raw water turbidities are in the range of 1.5 to 2.0 NTU.
d. Operational Experiences
i. General
The new 7.5 mgd treatment plant currently is operated from May through
October of each year, when system demand exceeds 10 mgd. For the balance
of the year, water demand is about 8 mgd. This can be satisfied by the
Ranney collector system, which operates at lower cost than the new
treatment plant.
166
-------
TABLE XXXIII. DESIGN CRITERIA FOR KENNEWICK WATER
TREATMENT PLANT PREOZONATION FACILITIES
(Cryer, 1986)
U.S.
Item Units
Plant Capacity
Design Flow mgd
cfs
gpm
Ozone Contact Basins
No. of basins
Detention min
time @ design
flow
Basin dimen- ftxft
sions-inside
Av. water ft
depth
Basin volume ft-3
gal
Total basin ft3
volume
Metric
Initial Ultimate Units Initial Ultimate
7.5
11.6
5,200
4
10
14x8
16
1,792
13,400
7,168
30.0
46.41
20,800
16
10
14x8
16
1,792
13,400
28,672
ra3/day
rn3/sec
28,300
0.33
m3/min 19.70
min
m x m
m
m3
m3
4
10
4.2x2.4
4.8
50.75
203
113,300
1.33
78.62
16
10
4.2x2.4
4.8
50.75
812
Chemical Feed Rate (Max. Dosace (3) Design Flow"!
Ozone plant Ib/day
influent
Chemical Feeders
Ozone Ib/day 2
250
x 125
Chemical Storage Capacity (Max
Ozone No. 03
generators
Total Ibs
2
250
1,000
3 x 250
2 x 125
. Dose (a)
5
1,000
kg/day
kg/day
113
57
454
3 x 113
2 x 57
Design Flow")
No. O3
Total kg
2
113
5
456
167
-------
ii. Water Quality
Kennewick's new Columbia River Water Treatment Facilities came on-line in
1980. Generally, the City and the service area customers have been satisfied
with the quality of water provided. Treated river water has a higher
customer acceptance, judging by the limited number of complaints, than
Ranney collector water alone.
THM analyses of the new treated water and chlorinated Ranney collector
water indicate that the plant has been able to maintain the TTHM
concentration below the current 100 /ig/L standard, while the quality of
water from the Ranney collector system remains very similar to the levels
determined during pilot plants studies, for raw^ .Columbia River^ wate.r. The
City's Ranney collector wa^rs" have had "an average THM'concentration of
approximately 107 Aig/L, while the ozone treatment plant water has averaged
approximately 14 /xg/L.
iii. Applied Ozonation Dosages - Dissolved Ozone
Residuals
Applied ozone dosage rates have ranged from 1.7 to 2.5 mg/L. Until 1983,
ozone residual levels were maintained at approximately 0.5 mg/L exiting the
contactor. In 1983, the City installed a dissolved ozone analyzer to control
the ozone dosage, which has resulted in lowering the dissolved ozone
residual concentrations to 0.1 mg/L, thus saving ozone, and still control
biological growth in the filters and basins prior to chlorination.
iv. Ozone Equipment Operational Experience
Operationally, the ozone generation equipment has performed very well. The
compressors have required only preventive maintenance. The ozone
generators have required the replacement of only three burned out tubes
during the first six years of operation. The major maintenance problem
appears to be tube fouling which was found to be a result of high moisture
in the feed gas. This situation was caused by two extenuating
circumstances. After several years of operation, it was discovered that the
refrigerant air dryer unit had developed a small leak which reduced the
effectiveness of the air preparation system. It was then determined that the
absorptive medium in the desiccant drier must be replaced when its
regeneration capacity is reduced to 40% of its originally specified capability.
Cleaning of dielectric tubes has become an annual maintenance procedure.
The only other significant operational problems concerned the ozone
contactors. Excess foaming and scum production can occur during spring
and late summer (algae destruction, primarily). This may require the
installation of surface skimmers and froth spray equipment. In addition, the
stainless steel tubes holding the ceramic diffusers corrode after about two to
three years of use, and must be checked and occasionally replaced when the
diffusers are cleaned.
168
-------
v. Costs Of Ozonation
The Kennewick facility is able to produce ozone at approximately $0.20/ib.
However, Cryer (1986) notes that the ozonation system has never been
operated at 100% of the high voltage output, and that ozone generation
efficiency has averaged approximately 16-17 kWh per pound of ozone
produced.
Based upon the City's very limited experience in operating the facility during
periods of time when the ozonation system has been off-line, the plant
operator believes that the addition of ozone prior to flocculation/filtration
results in a savings in flocculation chemicals of approximately 10%.
Currently, the treatment plant is producing finished water at a cost of
approximately $14.10 per million gallons ($0.141/1,000 gallons).
e. Implications of Proposed Surface Water Treatment Rule
Disinfection Conditions
During the May/October period of plant operation, the temperature of the
waters treated is approximately 20°C. With a 10 minute ozone contacting
time, and assuming an average residual ozone concentration of 0.5 mg/L (as
was the case until 1983), the CT value of 5 mg/L-min was being obtained,
compared with a value of 1.5 mg/L-min required (Table IIA) by the proposed
SWTR for 99.9% inactivation of Giardia lamblia cysts and 99.99% inactivation
of enteric viruses at 20°C. On the other hand, because the Kennewick plant
filters (thereby providing 2-logs of Giardia inactivation), only a single log of
additional Giardia inactivation is required. This means (Table Iffi) a CT
value of less than 0.7 mg/L-min, which corresponds to an ozone contact time
of only 1.4 minutes at the same residual ozone concentration.
After filtration, gaseous chlorine is added. A residual of 0.6 mg/L free
chlorine is maintained for 60 minutes at pH of approximately 8.0. The
contribution of chlorination to the required disinfection CT value thus is 36
mg/L-min. From Table IIA, the CT value for chlorine at pH 8.0 and 20°C is
101 mg/L-min.
Therefore, maintenance of the 0.5 mg/L ozone residual for 10 minutes
provides over 300% of the required disinfection capability, without reliance
on the post-chlorination. In addition, post-chlorination provides a
supplemental 36% disinfection credit (36/101).
However, since 1983 the ozone residual concentration leaving the ozone
contact chambers has been reduced to 0.1 mg/L. Assuming this ozone
concentration is present throughout the 10 minutes of ozone contacting, a
CT value of only 1 mg/L-min is obtained, versus the 1.5 mg/L-min required
by Table IIA. This means that only 67% of the disinfection capability is
provided by the preozonation step to achieve 99.9% inactivation of Criaidia.
but 130% of the inactivation required to obtain 90% additional inactivation
because of the filtration capability. The 60 minute chlorine contact time at
pH 8.0 (0.6 mg/L) provides an additional 36% of the required disinfection.
169
-------
This means that the plant currently provides at least 100% of the amount of
disinfection stipulated under the proposed Surface Water Treatment Rule,
disregarding the contribution of filtration.
B. UV RADIATION CASE HISTORY
1. Ft. Benton. Montana 2
UV Radiation for Primary Disinfection
a. The Problem
The City of Ft. Benton, MT takes water from the Missouri River and treats
it in a 20-30 year old filtration plant. This plant needed to be upgraded.
Costs for installing a new filtration plant were deemed to be excessive.
b. The Solution
A new 2 mgd peak flow treatment plant was installed in mid-1987. Water is
drawn through Ranney collectors installed 20-25 feet below the river bed.
This allows the river bed to filter the raw water. Turbidities of water
entering the treatment plant average 0.08 NTU. No Giardia cysts have been
found in the filtered river water.
Plant intake water is treated with UV radiation for primary disinfection,
then chlorinated for residual. If chlorination were to be relied on for
primary disinfection, a 2-hour contact time would be required by the State
authorities. Even a 30-minute contact time would require two-stage pumping
with long concrete pipes to provide sufficient contact time.
After primary UV disinfection, the water is chlorinated to a 1 mg/L residual
with no contact time. The water being chlorinated is quite clean, in that
the applied chlorine dosage is just over 1 mg/L. The pH is about 7.4 and
the water temperature is approximately 20°C year around.
The entire water treatment process is housed in a 32 x 32 ft building.
author is indebted to Fred Zinnbauer of Aquionics, Incorporated,
Erlanger, KY (supplier of the UV radiation equipment) and to Gary Swanson,
Project Engineer at Robert Peccia Associates, Helena, MT, consultants for
this project, for information on this plant.
170
-------
c. UV Radiation Conditions
The UV disinfection system consists of six irradiation chambers, two control
cabinets with alarms, chart recorders, relays, hour-run meters, lamp and
power on-lights, six thermostats, electrical door interlocks, mimic diagrams,
and six UV intensity motors measuring total UV output.
Four irradiation chambers are on-line at all times, with two chambers
reserved for backup. The alarm system is tied into the automatic butterfly
valves to provide for a fully automatic backup system. Each irradiation
chamber contains one 2.5 kW mercury vapor, medium pressure arc tube,
generating UV radiation at 253.7 nm.
The initial UV dosage is 41,000 jiW sec/cm2 at maximum water flow (1,650
gal/min) through each irradiation unit. Expected arc tube life is 4,500
operating hours providing a minimum UV dosage of 25,000 /iW sec/cm^.
These conditions are designed to reduce concentrations of E. coli organisms
by a minimum of 5-logs (1(P reduction).
The UV irradiation system is interfaced with the telemetry control system
activated when the low tank pump start set-point is activated on the set-
point controller. Sufficient warmup time is maintained before pump startup
is initiated. The UV system is deactivated once the high tank stop set-point
is activated on the set-point controller and the pump stops running.
Sufficient time delay is built into the system to ensure that water flow is
completely stopped before the UV system is deactivated.
A fully automated backup system is provided. Each bank of three irradiation
chambers has two units on-line at all times, with the third unit serving as
backup. In the event that the UV intensity drops below acceptable limits
(20,000 /tW sec/cm2) in any of the chambers, the automatic butterfly valve
will close, stopping flow through the chamber; the automatic butterfly valve
on the standby unit will open. The alarm system also is activated if UV
intensity drops below acceptable limits in any of the chambers. The UV
alarm system is interfaced with the automatic dialer and alarm system.
Replacement of the UV lamps is quite simple, requiring no more than a few
minutes.
d. Costs
Total equipment costs for the 6-unit UV irradiation system with butterfly
valves was $74,587.
e. Operating Experience
The new Ft. Benton water treatment plant started up in July 1987; no
operational data are yet available.
171
-------
C. CHLORINE DIOXIDE CASE HISTORIES
1. Evansville. Indiana (Lykins & Griese, 1986)
Predisinfection for THM Control
a. The Problem
Because of the November 1979 amendment to the National Interim Primary
Drinking Water Regulations (U.S. EPA, 1979), the Evansville (Indiana) Water
and Sewer Utility was faced with reducing the levels of trihalomethanes in
their finished drinking water. At the time, raw water was being
prechlorinated and treated with alum prior to primary settling, treated with
lime (to pH = 8), passed through secondary settling, fluoridated, rapid sand
filtered, then post-chlorinated before passage into the clearwell. The treat-
ment process was conducted in two separate 30 mgd treatment lines (total
production 60 mgd), and THM levels exceeded the THM standard of 100 /ig/L.
Prechlorination doses averaged 6 mg/L. Distribution system residence time
averages three days.
b. Pilot Plant Study
With the assistance of a cooperative agreement from the Environmental
Protection Agency (CR811108 - Sept. 1983), the Evansville utility initiated
research to evaluate the use of chlorine dioxide in a 100-gal/min pilot plant
adjacent to the full-scale plant. One train of the full-scale treatment plant
served as the control for the pilot plant study.
The pilot plant study was conducted in two phases, optimization, and long-
term. Chlorine dioxide was substituted first for prechlorination, then for
post-chlorination. GAC adsorption also was evaluated prior to the post-
chlorination step.
i. Optimization Phase
When predisinfection was eliminated and the pilot plant effluent was
disinfected and stored three days (to simulate distribution system residence
time), TTHM concentrations averaged 141 ju,g/L with chlorine post-
disinfection (2.5 mg/L residual chlorine concentration), and 1.2 /ig/L with
chlorine dioxide post-disinfection (1.9 mg/L residual).
Predisinfection with chlorine dioxide to maintain a residual through the pilot
plant did not increase the THM concentration and provided adequate
disinfection. The C1C>2 residual decreased from 4 to 0.3 mg/L through the
pilot plant.
ii. Long-Term Evaluation
In this phase the pilot plant procedures were evaluated during each season
of the year to determine the extent of seasonal effects. In this phase of
172
-------
the study, predisinfection with chlorine dioxide (1.1 mg/L average applied
dosage) reduced the amount of THMs formed during post-chlormation by
approximately 60%. The concept of post-disinfection with chlorine dioxide
was abandoned because of the difficulty in maintaining a disinfectant residual
in the distribution system (three days average residence time), keeping in
mind that the current EPA recommendation is that total oxidants from
application of chlorine dioxide should not exceed 1 mg/L in the distribution
system. The average chlorine residual concentration in the clearwell was 2.1
mg/L.
c. The Full-Scale Plant
Based on data obtained from the pilot plant study, Evansville utility officials
concluded that the MCL for TTHMs would be exceeded if the then-current
treatment was not altered. Changing the primary disinfectant from chlorine
to chlorine dioxide was judged to be the most effective procedure for
control at the least cost. A separate building was constructed to house the
chlorine dioxide generation facility, One portion contains the C1O2
generator, a second portion houses two 1-ton cylinders of chlorine.
This installation is capable of generating 14.24 Ibs of C1O2 per hour, which
can be divided in any proportion between the two halves of the treatment
plant. Gaseous chlorine and 25% NaClC>2 solution are delivered to the
chlorine dioxide reactor under partial vacuum generated by an eductor. Both
reagent flows are controlled by flow-rate meters, and the system is designed
to shut down if the eductor water supply fails or if chemical feed lines
break.
This production rate has been achieved with over 95% conversion to C1C>2
over the first 18 month period of operation.
d. Operating Experience
Chlorine dioxide began being added as a predisinfectant prior to any
additional chemical treatment in August 1983. The C1O2 produced is divided
evenly between the two treatment lines. During the first five months of
use, various CIO? dosage levels were used to determine the resultant
reductions in THM concentrations, to gather data on what percentage of the
C1O2 dosage would appear as total oxidant residuals in the finished water,
and to review the general operation of the entire system.
During this period, only one major problem was encountered. Brass
corporation cocks used to connect the PVC C1C>2 feed lines to the raw water
influent piping were oxidized by the concentrated disinfectant. This
oxidation and subsequent leaking of C1O2 solution resulted in temporary
disruption of the new treatment technique. The problem was resolved by
sliding a section of PVC pipe through new corporation cocks into the main
stream of the raw water lines. This modification permitted the PVC piping
to serve as an inductor while preventing direct contact of the brass
corporation cocks with concentrated C1O2 solution.
173
-------
Since the implementation of the treatment change, total oxidant levels from
chlorine dioxide have been maintained consistently below the 1.0 mg/L
recommended by the U.S. EPA. With an average applied C1C>2 pretreatment
dosage of 1.2 mg/L, total oxidant concentrations in the finished water have
averaged 0.5 mg/L. These data show that approximately 42% of the QO£
dosage is present as total oxidant.
Since the installation of predisinfection with chlorine dioxide at Evansville,
annual TTHM levels in the distribution system have been maintained between
50 and 80
e. Implications of the Proposed Surface Water Treatment
Rule
Raw water total coliform levels at Evansville are in the range of 3,400 to
5,400 per 100 mL. The filtration system provides 2-logs of Giardia and virus
inactivation. However, because of the higher raw water total colifonn
levels, EPA recommends (U.S. EPA, 1987d) that primary disinfection be
provided to attain 99.9% inactivation of Giardia and 99.99% inactivation of
viruses, in addition to those provided by filtration.
Chlorine dioxide predisinfection at Evansville employs a dosage of 1.2 mg/L,
followed by chlormation in the clearwell at an average concentration of 2.1
mg/L. Although the contact time after adding C1O2 is not stated in the
reference, attainment of the required CT value of 21 mg/L-min at 20°C
(Table IIA) and an average ClO^ concentration of 1.0 mg/L means that a 21
minute contact time would provide 99.9% inactivation of Giardia cysts and
99.99% inactivation of viruses, independently of the added degree of disinfec-
tion which can be credited to post-chlorination. Since the average total
oxidant (from chlorine dioxide) concentration in the plant finished water is
0.5 mg/L, an actual CT value of 10.5 mg/L-min would be expected in a 21
min contact time. This represents about 50% of the required degree of
primary disinfection.
With respect to post-chlorination, at an average 2.1 mg/L chlorine residual
in the clearwell at pH 8 (CT value = 101 mg/L-min at 20°C) means that a
contact time of 48 minutes would be required to provide 99.9% inactivation
of Giardia cysts and 99.99% inactivation of viruses. If only a 30-minute
clearwell contact time is provided, still 63% of the required disinfection is
provided by chlorine in the clearwell. Therefore, it is likely that the
required amount of disinfection, even with the high level of raw water total
coliforms present, is attained from both the chlorine dioxide and chlorine
disinfection steps.
174
-------
2. Hamilton. Ohio (Augenstein, 1974; Miller et al.,
1978; U.S. EPA, 1983)
Primary/Secondary Disinfecting Wi^ Chlorine
Dioxide
a. The Problem
In 1956 a 15 mgd lime softening plant began operating in Hamilton, OH
treating groundwater (18 wells, 200 ft deep). Chlorine was used as the sole
disinfectant when the plant started operating. However, because of customer
complaints of chlorinous tastes and odors, a change was deemed necessary.
Chlorine dioxide was tested, and in 1972 this material was substituted for
chlorine for primary and secondary disinfection.
b. The Treatment Process
Hamilton's treatment process in 1987 is as follows: aeration, lime addition,
flash mixing, sedimentation, recarbonation (with food grade CO?), filtration,
fluoridation (sodium silicofluoride), disinfection (with C1O2J, clearwell
storage. Raw water turbidities are below 1 NTU, and raw water total
coliforms are less than 1 per 100 mL.
c. Generation of Chlorine Dioxide
Chlorine dioxide is generated by mixing 37% aqueous sodium chlorite and
aqueous chlorine in a ratio of 1:1 by weight (2 Ibs of each reagent per
million gallons of water to be disinfected). This provides an applied C1O2
dosage of 0.25 mg/L. Chlorine gas is delivered to the site in 150-lb
cylinders. Aqueous NaClO2 solution (37%) is delivered in 200-lb drums.
The chlorine dioxide generation system consists of one plant-fabricated
reactor vessel for C1O2 production, one peristaltic pump for NaClO2 solution,
two chlorinators (one serves as standby). Two 150-lb liquid chlorine
cylinders are positioned next to the chlorinators. The weight of the cylinder
contents is measured by a scale. Switchover from one cylinder to the other
is manual. PVC tubing is specified between the chlorinator and C1O2 reactor
vessel; heavy Tygon tubing transports the NaClO2 solution from the drum to
a small plastic day-tank and to the reactor vessel. This Tygon tubing loses
its rigidity and must be replaced after about a month of use. The semi-
transparent day tank allows visual inspection of the level of liquid NaClO2
and thereby enables the operator to maintain an acceptable suction head on
the peristaltic pump.
The chlorine dioxide reactor vessel is constructed of Schedule 80 PVC piping,
18 inches high and approximately 6 inches in diameter. The vessel is filled
with PVC rings, 1-inch in diameter. The chamber is opaque except for the
sight glass mounted in-line on the discharge piping. A white card is
positioned behind the sight glass for better observation of the C1O2 color.
175
-------
Applied dosage of C1O2 in the clearwell is 0.25 mg/L. The residual leaving
the clearwell is approximately 0.15 mg/L, and is 0.10 mg/L at the extremities
of the distribution system.
d. Effects of Installation of Chlorine Dioxide
Prior to installation of C1C>2, customers had complained about brown stains
(iron) during washing of clothes. When C1C«2 treatment was initiated, brown
slimes were loosened from the mains. Plant personnel then flushed the
distribution system. Shortly thereafter, the brown stain problems
disappeared, and complaints ceased. Plant personnel attributed the source of
the problem to crenothrix and leptothrix bacteria (iron bacteria) that had
been present in the extremities of the distribution system before introduction
of chlorine dioxide.
Since 1972, C1O2 has been the sole disinfectant in Hamilton's drinking water.
e. Costs for Chlorine Dioxide
In 1977, the costs for C1O2 in Hamilton were determined (Miller et al., 1978)
to be about 3.60/capita/year. Chlorine and NaQO2 together cost $6,540^rr
(1977). Total chemicals costs in 1977 for the finished water averaged
190/1,000 gal; the fraction due to CIO? was considered negligible. The
operating and maintenance costs for CIO? generation at Hamilton were
estimated to be less than $50/yr in 1977. The plant-fabricated C1C>2 reactor,
piping, hardware, and installation were estimated by the plant supervisor to
cost around $400 in the 1977 market. Installation was done by plant
personnel. The peristaltic pump for NaClC>2 solution cost less than $200
(1977). To this must be added the cost of the two chlorinators ($600),
which were already on line at the plant.
f. Implications of the Proposed SWTR CT Values
Although the Hamilton raw water is groundwater, and therefore probably will
not be subject to the requirements of the proposed Surface Water Treatment
Rule (U.S. EPA, 1987a), it is interesting to consider the effects if such
disinfection requirements as listed in Table IIA were to be levied on this
water supply system.
Chlorine dioxide is added to the Hamilton water in applied doses of 0.25
mg/L as it enters the clearwell. The water temperature is about 20°C year
round, and the pH is 9.4 to 9.5. Hamilton's first customer is located about
0.5 mile from the plant. Thus there is very little contact time in the
distribution system.
The plant filters efficiently, and therefore only 1-log additional inactivation
of Giardia cysts and 2-logs inactivation of viruses need be provided by the
chlorine dioxide. Table IIB shows that at 15°C, a CT value of 9 mg/L-min
will provide the required degree of disinfection.
176
-------
The Hamilton clearwell holds 500,000 gallons. During periods of high water
use, water is produced at the rate of 19 mgd. In periods of low water use,
only 8 mgd are produced. Thus the contact time in the Hamilton clearweiJ
ranges from 30 to 90 minutes, at the high and low production rates,
respectively. Assuming that the average concentration of C1O2 in the
clearwell is 0.2 mg/L, then the CT values provided are 6 and 18 mg/l^min,
respectively.
Thus the current disinfection conditions using chlorine dioxide meet the
proposed CT requirements, except at the most rapid production of water,
when only two-thirds of the disinfection capability is provided by CIO? ui
the clearwell. However, simply increasing the C1C>2 dosage by one-third (i.e.,
to 0.32-0.33 mg/L) during periods of highest flow will provide the additional
amount of disinfection required by the proposed SWTR.
3. Galveston. Texas (Myers et al., 1986)
Preoxidation With CIO?
Post-Disinfection With C1O2 + Chloramine
a. The Problem
The Galveston County Water Authority (GCWA) owns and operates an 18 mgd
capacity water treatment plant in Texas City, TX, currently (1986) producing
12 mgd of finished water from the Brazos River. Raw water characteristics
include high color, variable turbidities (68-111 NTU), high organic contents,
high iron (2.7-3.8 mg/L), and seasonally high algae content, sometimes
reaching levels of 5,000 blue/green algae/mL. Such high algae and organic
contents create the potential for unpalatable tastes and odors to develop
during treatment and distribution.
Total trihalomethane formation potentials for Brazos River water, measured
periodically during September 1983 through April 1984 (20-day chlorination
period), were in the range of 400 to 650 /u,g/L. Finished water THM
concentrations measured during the same period were in the range of 180 to
350 ug/L. During periods of intermittent raw water algae blooms and
associated high organic carbon contents, numerous consumer complaints were
received regarding tastes and odors.
These problems prompted an investigation of alternative disinfectant
strategies.
b. The Original Treatment Process
The treatment process included addition of cationic polymer for primary
coagulation, lime for pH adjustment, prechlorination for taste and odor
control, and ferric sulfate as a flocculant aid prior to upflow reactor/-
clarifiers. These provide the dual functions of flocculation and
sedimentation. Dual media filtration follows, then disinfection (chlorination
177
-------
to a 2.4-5.0 mg/L total available chlorine residual in the finished water), and
fluoridation.
c. Study of Alternatives
Chloramines and chloride dioxide were selected for further study and pilot
plant testing. Ozone was eliminated from further study because of
implementation considerations. Also, ozone does not maintain a residual
during distribution, and therefore would require the use of a secondary
disinfectant anyway. Chlorine dioxide for both prepxidation and post-
disinfection achieved the goals of GCWA better than did chloramines (poor
preoxidant).
Preoxidation/post-disinfection combinations were studied in the following
sequences:
Phase 1) Chlorine/chlorine
Phases 2a & b) Chloramine/chloramine
Phase 3) Chlorine dioxide/chlorine
Phase 4) Chlorine dioxide/chlorine dioxide
Phase 5) Chlorine dioxide/combined chlorine + chlorine dioxide, and
Phase 6) Chlorine dioxide/combined chloramine + chlorine dioxide.
i. Study Results
Phase 1: Chlorine - Chlorine
Finished water exhibited intermittent algae-related taste and odor, and THM
levels were in excess of federal standards (350 /tg/L); however bacterial
quality was excellent. This approach was abandoned.
Phases 2a and 2b: Chloramine - Chloramine
Phases 2a and 2b employed chlorine:ammonia weight ratios of 3:1 and 7:1,
respectively. Although THM levels were lowered to about 60 /ig/L, bacterial
counts for coliforms demonstrated that confluent growth was occurring on
several cultures, indicating that inadequate residual was being maintained in
the distribution system (3:1 chlorine:ammonia ratio). This approach was
abandoned.
The experiments were repeated using the 7:1 chlorine/ammonia ratio.
Acceptable bacterial quality was achieved, but numerous taste and odor
complaints were received during this period. This approach was abandoned.
178
-------
Phase 3: Chlorine Dioxide - Chlorine
Chlorine dioxide was installed in May 1984, using a generator with a
conversion efficiency (chlorite ion to C1O2) of approximately 80%>. Alter
C1C«2 preoxidation, the clarified water showed no traces of THMs. No taste
and odor complaints were received, despite very high raw water algae counts
(up to 5,000 blue/green algae/mL). However, finished water THM levels
sometimes persisted above 100 /ig/L.
In November 1984, a chlorine dioxide generator with a 98+% conversion
efficiency was installed, and this phase was continued for an additional two
months. THM levels averaged 102 /ig/L), and no taste and odor complaints
were received. This approach was abandoned.
Phase 4: Chlorine Dioxide - Chlorine Dioxide
This approach of using chlorine dioxide both as the preoxidant and post-
disinfectant was authorized on an interim basis by the Texas Department of
Health in early March 1985 with the stipulation that a maximum C1C«2
residual of 1.0 mg/L be maintained and that finished water quality be
monitored throughout the distribution system.
Finished water THM concentrations during the test period averaged 60 /ig/L,
and finished water turbidities were the lowest of any of the alternative
disinfectant scenarios. Bacterial counts generally were excellent, but inter-
mittent elevated counts were noted at three different points in the
distribution system: at the clearwell, and at locations 2 and 5 miles from
the plant.
Additionally, the bacterial species distribution also changed upon C1C>2
disinfection, both at the plant and in the distribution system. Bacterial
counts displayed a shift from orange to yellow-staining gram negative (-)
rods to white-staining gram positive (+) rods, similar to slime-forming
Bacillus (sp.).
This approach, although promising, also was abandoned.
Phase 5: Chlorine Dioxide - Combined Chlorine + Chlorine Dioxide
Phase 4 was repeated except that chlorine was applied in conjunction with
ClO^ until the bacteria could be identified. Excellent bacterial quality was
obtained. The bacterial plate counts remained at or below the guideline of
500 colonies per 100 mL for all monitoring locations. The shift in bacterial
species distribution continued as the plate counts decreased, so that over
95% of all colonies examined were either yellow gram (-) rods or white gram
(+) rods.
Finished water THM levels rose to an average of 81 /ug/L. This approach
was rejected, in favor of Phase 6.
179
-------
Phase 6: Chlorine Dioxide - Combined Chloramine + Chlorine Dioxide
This phase was implemented on December 6, 1985. With this post-disin-
fection combination, finished water THM levels averaged 68 /x,g/L, and the
bacteriological quality remained excellent. No coliforms were found in the
clearwell or in the distribution system during this phase, and bacterial
counts have ranged from < 1 to 30 colonies per 100 mL. Of those colonies
identified, over 95% were the white-staining, gram (+) Bacillus (sp.) variety
with less than 1% belonging to the orange-staining gram (-) type.
d. The Results
Preoxidation with chlorine dioxide and post-disinfection with C1O2 in
combination with either chlorine or chlorammes has provided effective taste
and odor control, maintained an active disinfectant residual and mitigated
THM formation (to a current level of about 68
Chlorine dioxide preoxidation has proven to be an excellent algicide and
biocide. The effectiveness of C1O2 in removing algae in flocculation/-
sedimentation has been demonstrated in a decrease in filter fouling and
improved finished water turbidities since implementing its use.
Odors experienced at the flocculating clarifiers and taste and odor
complaints in the distribution system have been reduced. The unexpected
benefit of improved treatment plant performance has been maintained for at
least the first seven months of operation of the treatment modifications.
Finished water turbidities decreased from an average range of 0.3 to 0.65
NTU during 1983 (chlorine pre- and post-treatment) to 0.09 to 0.2 NTU in
1986.
Bacterial counts using the membrane filter method for coliforms have
declined to excellent levels. Since the advent of C1O2 post-disinfection, all
bacterial counts obtained from samples collected at the GCWA distribution
point have been below 30 colonies/100 mL, and all but two samples have
been below 10 colonies/100 mL. This indicates a substantial residual present
at that distribution system point. The bacterial counts obtained from the
GCWA distribution system continually are below the guideline of 500/mL, and
often are below 5 colonies/100 mL, indicating the maintenance of a good,
active residual.
The Galveston County Water Authority has successfully implemented the first
chlorine dioxide disinfection program in the State of Texas.
180
-------
CHLORAMINE CASE HISTORIES
1. Bloomington. Indiana (Singer, 1986)
Prechlorination. Post-Chloramination
a. The Problem
Bloomington takes its water from a lake having TOC levels of 4-6 mg/L.
This water is treated by conventional coagulation with alum, flocculation,
sedimentation, and filtration through pressure filters. Prior to September
1984 the utility applied chlorine to the raw water and again ahead of the
pressure filters. Average chlorine dosages were 1.8 and 1.0
. .
respectively. Typical average TTHM concentrations reported by the utility
as part of its quarterly compliance monitoring program were as follows:
March 1983 77 /ig/L
June 1983 168 jig/L
Sept 1983 142 ^tg/L
Dec 1983 50 /xg/L
March 1984 79 /ig/L
May 1984 113 jtg/L
It is clear that the current 100 jug/L TTHM running annual average standard
was being exceeded.
b. The Solution
In September 1984, the Bloomington water utility changed from post-
chlorination to post-chloramination. An average 0.54 mg/L of ammonia was
applied along with 1.5 mg/L of chlorine ahead of the pressure filters. The
desired residual chlorine concentration leaving the plant of 1.0 mg/L of free
chlorine was changed to 1.5 mg/L of combined chlorine.
c. Performance
Quarterly compliance monitoring data for TTHMs subsequent to the
changeover are as follows:
Sept. 1984 47
Dec. 1984 24 /tg/L
Feb. 1985 43 /*g/L
April 1985 56 jug/L
Aug. 1985 57 jug/L
Clearly, TTHM concentrations have been brought down to about 50 /ig/L
from well over 100 ju,g/L
Table XXXIV summarizes THM and TOX (Total Organic Halide) data for
samples collected as major points in the treatment system when chlorine
181
-------
alone was used for pre- and post-treatment. The data show that TOX levels
increased in parallel with increases in TTHM levels. Table XXXV summarizes
similar data taken using combined chlorine residuals. These data show that
although the TTHM formation ceases after the addition of ammonia, the
production of TOX continues, although at a greatly reduced rate. This
illustrates that as EPA sets MCLs for additional halogenated organic
materials other than THMs, utilities opting for post-chloramination must plan
to determine the makeup of their TOX fraction.
TABLE XXXIV. SUMMARY OF THM DATA AT BLOOMINGTON,
INDIANA WITH FREE CHLORINATION 8/16/84
(Singer, 1986)
Sampling Point
Residual Free TOC TTHM TOX
Chlorine, mg/L mg/L ju,g/L ju,g/L
Raw Water
Settled Water
0.25
4.3
3.6
1
23
48 127
Filtered Water
Dist. System #1
Dist. System #2
1.0
1.8
0.65
2.4 81 205
110 291
151 363
TABLE XXXV. SUMMARY OF THM DATA AT BLOOMINGTON,
INDIANA WITH POST-CHLORAMINATION 8/26/85
(Singer, 1986)
Sampling Point
Residual
Chlorine, mg/L
TOC TTHM
mg/L /Ag/L
TOX
Raw Water
Settled Water
Filtered Water
Dist. System #1
Dist. System #2
—
Trace, Free
1.2 Combined
1.0 Combined
0.9 Combined
4.1
2.8
2.8
—
—
0
53
55
52
57
17
94
91
115
116
Since switching to post-chloramination, the utility has experienced no
adverse effects in operations or in finished water quality. According to
distribution system monitoring records, the microbiological safety of the
water has been maintained.
182
-------
2. Philadelphia. Pennsylvania (McKeon et al., 1986)
Pre-Chlorine Dioxide + 1-Hour Prechlorination
Post-Chloramination
This is a case history which is still evolving, but one which has resulted in
modifications to attain current regulatory and water quality objectives.
However, the types of studies which have been conducted by the Philadelphia
Water Department point to the directions this utility probably will take in
response to continued tightening of federal regulations and changes in water
quality.
a. The Problems
The Baxter Water Treatment Plant (formerly called Torresdale), was built in
1960 on the site of a slow sand filtration plant that had been in service
since 1903. The current plant is rated at 282 mgd, and supplies potable
water to over 800,000 people, from the Delaware River by conventional
treatment.
Chemicals fed routinely include chlorine, ferric chloride or ferrous sulfate,
lime, fluoride, and ammonia. Powdered activated carbon is fed on demand
for control of taste and odor events, and chloride dioxide is fed currently
for control of THMs and tastes and odors. The C1O2 feed system was
incorporated in the original plant design in the 1950s to oxidize phenolic
compounds found in the watershed. Since that time, the sources of phenol
have been eliminated and the CICb feed system has found use as part of the
program to reduce concentrations of trihalomethanes.
Prior to 1976, the Baxter plant practiced breakpoint chlorination at the raw
water basin and maintained free chlorine in the distribution system. From
raw water chlorination to the consumer, a total of 96 hours of free chlorine
contact time normally elapsed,
Initial analyses for THMs in 1978 showed peak concentrations above 300 /ig/L
with a running annual average of 140 /ig/L. In light of these results, the
Philadelphia Water Department began to reevaluate its disinfection strategies.
The selection of an alternative disinfectant strategy at the Baxter plant not
only must take into consideration bacterial kills, but taste and odor control,
suppression of algae, corrosion control, residual duration, and economics.
This case history reviews Philadelphia's experiences over ten years in
modifying chlorination practices, and the problems which surface with
reduction of chlorine contact time.
b. Process Modifications
Initial efforts to reduce formation of THMs concentrated on reduction in
free chlorine contact times. From the original 96-hour free chlorine contact
time, today (1986) chlorine contact time has been reduced to 1-hour,,
183
-------
i. Chloramination of Finished Water (Reduction of Free
Chlorine Contact Time from 96-hours to 24-hours)
Chloramination of the finished water was initiated in 1976. Ammonia was
added to convert free chlorine to monochloramine, to yield a longer lasting
residual in the distribution system, to improve the organoleptic properties of
the water, and to reduce the corrosion rates associated with the use of free
chlorine. Adequate disinfection was assured by maintaining a 2-3 mg/L free
chlorine residual throughout the plant treatment process.
Trihalomethane formation within the treatment process was reduced by 40%
under summertime conditions in comparison to the ultimate (96-hour) THMFP
reached previously (from 231 to 174 ju-g/L).
ii. Reduction in Chlorine Treatment at Raw Water Basin
A gradual phasing out of chlorine addition at the raw water basin inlet took
place from 1976 through 1979. Whereas in 1975, 60 Ibs of chlorine per
million gallons of chlorine was added at the intake, by the end of 1979,
chlorine addition was limited to 20-30 Ibs/million gallons of water. From
1977 to 1979, the marginal chlorination of the intake followed by five hours
of free chlorine contact time in the floe/sedimentation basins and filters
produced summertime average THM concentrations of 200 /Ltg/L and yearly
averages of 140 /ig/L. While this treatment regimen produced positive results
in the control of algae blooms in the raw water basin and minimization of
taste and odor events, its debit was high formation of THMs. The decreased
use of chlorine in the raw water basin also had the effect of reducing the
amount of chlorine loss due to evaporation, and thus the quantity of chlorine
used.
iii. Utilization of Chlorine Dioxide at Raw Water Basin
Inlet (5 Hours Free Chlorine Contact Time)
Beginning in 1980, routine use of chlorine at the raw water intake basin was
abandoned. In its place, chlorine dioxide was added in dosages between 0.5
and 1.0 mg/L. Summertime THM values thus were reduced from 200 /ig/L to
140 /ig/L. This treatment was sufficient to control algae, and thus tastes
and odors, at all times except during the spring algae bloom. For that
period of time, breakpoint chlorination of the intake water and/or 100-200
Ibs of powdered activated carbon per million gallons is required to eliminate
vegetative tastes and odors.
iv. Installation of a New Chlorine Application Point (One
Hour Free Chlorine Contact Time)
In the fall of 1980, a chlorine application point was installed in the "applied
to filters" channel which allowed for increased flexibility in the use of
chlorine. Free chlorine contact time was reduced from five hours to one.
184
-------
Chlorine was added at the rapid mix to barely achieve breakpoint and
provide a residual which dissipated within a few minutes. Chlpramines were
carried across the floe/sedimentation basins. Sufficient chlorine then was
added at the new application point to achieve a free chlorine residual of
1.5-2.0 mg/L. This residual was converted to chloramines one hour later as
the water flow left the filter building.
This treatment regimen gave adequate control of taste and odor, except for
the previously mentioned outbreaks associated with spring algae blooms,
which forced reversion back to free chlorine at the intake. THM levels,
with only one hour free chlorine contact time, resulted in summertime values
averaging 100 /tg/L with a yearly running average of 60
v. Ten Minutes Free Chlorine Contact Time
A plant trial was carried out in November 1982 utilizing a 10-minute chlorine
contact time. Results indicated that satisfactory disinfection could be
achieved with only chloramines carried through the floe/sedimentation basins
when water temperature was below 60°F. This strategy was initiated on a
plant-scale in December 1982. Adequate disinfection was achieved, but
periodic taste and odor results were less than satisfactory. Medicinal and
chemical taste complaints were received and several large taste and odor
events occurred following storms.
THM results under the reduced chlorine treatment reduced running annual
average concentrations from 60 to 50 ^tg/L.
The trial was terminated in December 1983 because the disinfection scheme
did not adequately address the taste and odor qualities of the water.
Treatment regimen returned to 1-hour free chlorine contact time.
c. Economics
Realization of 70% reduction in THM concentrations over the ten year period
was obtained at minimum cost. The 1978 disinfection cost was $5.01/million
gallons; in 1986 the cost was $9.53. Converting the 1986 costs into 1978
dollars, the 1986 adjusted cost is $5.52/million gallons. Cost increases were
minimized because the reduced chlorine contact times resulted in less
evaporative losses of chlorine, which netted a 20% decrease in the amount of
chlorine utilized. This partially offset the increased costs incurred by the
use of chlorine dioxide.
d. Operational Improvements
Free chlorine contact times at the Baxter plant have been reduced from 96-
hours in 1975 to 1-hour in 1986. Trihalomethane concentrations in the
finished water have been reduced from an estimated annual average of 220
jig/L in 1975 to 60 /*g/L in 1986. Treatment strategy is most critical during
the spring and summer months when vulnerability to taste and odor events is
185
-------
high and THM formation within the treatment process is accelerated due to
warm water temperatures.
e. For the Future
The near term goal of the Philadelphia Water Department is to reduce the
annual running THM concentration average below 50 £tg/L, a 15% reduction
from current (1986) levels. This can be achieved at minimum expenditure by
installation of a pH adjustment point at post-treatment. The existing
treatment scheme calls for raising the raw water pH to 8.4 at the rapid mix
and carrying this high pH through the distribution system for corrosion
protection. Addition of a pH adjustment point at post-treatment will allow
for a pH of 7.5 to be used through the floe/sedimentation basins and filters,
with pH adjustment to 8.4 after chloraminatipn. Plant scale trials using this
treatment regimen yielded a 20% reduction in THM formation (to about 40
If EPA lowers the THM MCL to below 50 /itg/L, ozone and/or granular
activated carbon become the likely alternatives at the Baxter plant.
Extensive laboratory and pilot plant evaluations (Neukrug et al, 1984) have
developed conceptual full-scale plant designs incorporating these two
treatment techniques.
Estimated annual amortized capital and operating costs for ozone at the
Baxter plant, spread over the lifetime of the equipment, are estimated to be
about $50/million gallons. The associated costs for GAC post-contacting (15-
minute empty bed contact time) with a 75-day regeneration frequency would
be about $212/million gallons. This design configuration is capable of
producing THM concentrations of less than 10 /ig/L.
IX. SUMMARY RECOMMENDATIONS FOR DISINFECTION STRATEGIES
AND FOR THE CONTROL OF DISINFECTION/OXIDATION BY-
PRODUCTS
Consideration of all of the preceding discussions, including the case studies,
leads to the following general and specific conclusions and recommendations:
A. For Disinfection
1. From review of EPA's proposed Surface Water Rule (U.S. EPA, 1987a) and
the National Academy of Sciences recommendations (NAS, 1987), ozone
and chlorine should be viewed as the primary disinfectants of choice for
surface waters and ground waters directly influenced by surface waters.
However, chlorine can be used only when the THM and TOX formation
potentials are sufficiently low to meet current and projected disinfection
by-product regulations, and to insure that no other halogenated products
are formed in quantities sufficiently high that they will be subject to
regulation. UV radiation can be used as a primary disinfectant, but for
groundwater only, since it does not provide the required amount of
inactivation of Giardia cysts for surface water treatment.
186
-------
2. For groundwater disinfection, ozone, chlorine, and UV radiation are the
primary disinfectants of choice, with the same restrictions on chlorine as
noted above.
3. Monochloramine is not recommended by EPA or by NAS as a primary
disinfectant because of its very high CT values with respect to Qiardia
cysts and because inactivation of Giardia cysts with chloramine does not
guarantee inactivation of viruses.
4. NAS also recommends against the use of preformed monochloramine and
against marginal chlorination as methods of introducing chloramines.
Organic N-chloramines, which are even weaker disinfectants than
monochloramine, and which cannot be distinguished from inorganic
chloramines by current field analytical techniques, are formed by these
chloramination techniques.
5. For surface or groundwater, chlorine is the most effective secondary
disinfectant, provided the water is sufficiently clean that significant
quantities of THMs and other halogenated organic compounds are not
produced.
5. If an MCL is set for chlorine dioxide at 1.0 mg/L, this material can be
considered for primary disinfection, and definitely can be employed as a
secondary disinfectant. A ClO^ dosage of 1.2-1.4 mg/L will produce
residual oxidant (C1C>2 + chlorite ion + chlorate ion) levels of ~ 1 mg/L.
6. On the other hand, if an MCL for chlorine dioxide is set at 0.2 mg/L
(Table VII), use of C1O2 as a primary disinfectant will be severaly
restricted to exceptionally clean waters, because the maximum applied
dosage then would be equivalent to about 0.3 mg/L. Its use as a
secondary disinfectant also will require a high quality finished water in
order to provide a stable, detectable residual throughout the distribution
system.
7. If an MCL is set for monochloramine at 2.5 mg/L, it can be used as a
secondary disinfectant. However, if an MCL is set at 0.29 mg/L (Table
VII), its use as a secondary disinfectant, while not eliminated, will be
limited to high quality finished waters.
8. If EPA sets low MCLs for both C1O2 and chloramine, then chlorine may
become the only, practical secondary disinfectant available. In this event,
the water will have to be pretreated more efficiently in order to lower
TOX precursor levels.
9. The National Academy of Sciences (1987) recommends that when chlor-
amine is used as the secondary disinfectant, chlorine should be added
first to produce a slight residual of free chlorine above that required to
oxidize nitrogen, followed by addition of ammonia to form monochloramine
and limit THM formation.
10. CT values for different disinfectants employed in a treatment process are
additive. For example, preoxidation with ozone or chlorine dioxide for a
187
-------
non-disinfection application nevertheless provides some CT benefit which
can be added to that obtained in the primary and secondary disinfection
steps. When attempting to take disinfection credit for ozone, a
measurable residual of dissolved ozone must be maintained for a specific
contact time in the treatment plant. Preferably, dissolved ozone should
be monitored at two points, and the detention time between then should
be known as accurately as possible. This will allow more accurate
depiction of CT values than those calculated from measurement of
residual ozone at a single point.
11. When using ozone for disinfection, contact time must be defined in terms
of detention time in the ozone contacting chamber or subsequent ozone
reaction chamber. This is because ozone residuals dissipate rapidly after
application of ozone to water ceases. To calculate the disinfection credit
of ozone under such conditions requires making assumptions regarding the
period of time over which the ozone residual concentration is established
and maintained (if only a single ozone contactor is employed). Satisfying
the ozone demand of the water and attaining the desired residual of
ozone in a first contact chamber, followed by a second chamber in which
the desired residual ozone is maintained, is a more quantitative and
reliable disinfection design approach.
12. Changing disinfection procedures usually must be made with attention
being paid to overcoming other water quality problems, such as control of
color, taste and odor, iron and manganese, synthetic or volatile organic
contaminants, etc.
13. Whenever ozone is employed, for oxidation or disinfection purposes, it
should be added in the absence of chlorine, chlorine dioxide, or
chloramine. This is because ozone oxidizes these three chemicals (very
rapid oxidation of C1C>2), and both reagents are mutually destroyed, to no
benefit to the water treatment process. When primary disinfection with
ozone is followed by chlorine, chlorine dioxide, or chloramine as the
secondary disinfectant, it is important that no residual ozone be present,
particularly in the case of chlorine dioxide, for the same reason.
B. For Controlling Disinfectant/Oxidant By-Products
14. Currently, halogenated by-products are the only compounds being
regulated or selected for regulation in the next few years.
15. Attention to minimizing the use of chlorine during disinfection will also
minimize the production of disinfection by-products.
16. Conventional physical water treatment processes (e.g., coagulation,
flocculation, sedimentation, filtration) should be improved to remove or
minimize concentrations of organic precursors of disinfection by-products
prior to the addition of chlorine. Preoxidation and activated carbon
adsorption are treatment steps which can be effective in this regard.
Ozone, potassium permanganate, hydrogen peroxide can be effective
188
-------
preoxidizing agents. Chlorine dioxide also is an effective preoxidant, if
its use at 1.2-1.4 mg/L applied dosages continues to be allowed.
17. When relying on monochloramine as the secondary disinfectant, close
attention should be paid to the TOX levels and the halogenated organic
products which are contained in this group parameter. TOX levels are
known to increase slowly in distribution systems employing chloramine as
the secondary disinfectant.
18. Most current preoxidation applications of ozone are designed to provide
only a trace residual of ozone at the exit of the contact chamber. In
plants currently employing ozone for preoxidation using a single
contactor, simply increasing the dissolved ozone residual to the level
appropriate for the designed ozone contact time and the water
temperature can provide at least the major portion, if not all, of the
primary disinfection required to guarantee inactivation of Giardiq cysts
and enteric viruses.
X. REFERENCES
Akin, E., et al., 1987, 'The (U.S. EPA) Office of Research and Development
Health Research Program on Drinking Water Disinfectants and Their By-
Products: An Issue Paper Prepared for a Science Advisory Board Program
Review".
American Water Works Association, 1973, Water Chlorination Principles and
Practice. AWWA Publication No. M20, p. 12.
American Water Works Association, 1985, Water Treatment Plant Design
(Denver, CO: Am. Water Works Assoc.)
Angehrn, M, 1984, "Ultraviolet Disinfection of Water", Aqua 2:109-115.
Augenstein, H.W., 1974, "Use of Chlorine Dioxide to Disinfect Water
Supplies", J. Am. Water Works Assoc. 66:716-717.
Christman, R.F.; Johnson, J.D.; Pfaender, F.K.; Norwood, D.L.; Webb, M.R.;
Hass, J.R.; Bobenreith, M.J., 1980, "Chemical Identification of Aquatic
Humic Chlorination Products", in Water Chlorination. Environmental
Impact and Health Effects. Vol. 3. R.L. Jolley, W.A. Brungs, and R.B.
Gumming, Editors (Ann Arbor, MI: Ann Arbor Science Publishers, Inc.,
1980), pp. 75-84.
Condie, L.W., 1986, "Toxicological Problems Associated with Chlorine
Dioxide", J. Am. Water Works Assoc. 78(6):73-78.
Cotruvo, J.A.; Vogt, C.D., 1985, "Regulatory Aspects of Disinfection", in
Water Chlorination: Chemistry. Environmental Impact, and Health Effects.
Volume 5. R.L. Jolley, RJ. Bull, W.P. Davis, S. Katz, M.H. Roberts, Jr.,
and V.A. Jacobs, Editors (Chelsea, MI: Lewis Publishers, Inc., 1985), pp.
91-96.
189
-------
Cryer, E.T., 1986, "Preozonation at Kennewick, Washington, Case History",
presented at Intl. Ozone Assoc. Workshop on Drinking Water Treatment
With Ozone, Perrysburg, OH, April, 1986.
Duguet, J.P.; Brodard, E.; Dussert, B.; Mallevialle, J., 1985, "Improvement in
the Effectiveness of Ozonation of Drinking Water Through the Use of
Hydrogen Peroxide", Ozone: Science & Engineering 7(3):241-258.
Ehrlicher, H., 1964, Zentr. Arbeitsmed. Arbeitsschutz 14:260.
Glaze, W.H.; Kang, J.-W.; Aieta, M., 1987, "Ozone-Hydrogen Peroxide Systems
for Control of Organics in Municipal Water Supplies", in Proc. Second
Intl. Conf. on The Role of Ozone in Water and Wastewater Treatment
D.W. Smith & G.R. Finch, Editors (Kitchener, Ontario, Canada: TekTran
International Ltd., 1987), pp. 233-244.
Great Lakes - Upper Mississippi River Board of State Sanitary Engineers,
1985, Recommended Standards for Water Works - 1985 Edition (Albany,
NY: Health Education Service, 1985).
Grunwell, J.; Benga, J.; Cohen, H.; Gordon, G., 1983, "A Detailed Comparison
of Analytical Methods for Residual Ozone Measurement", Ozone Science &
Engineering 5(4):203-223.
Gumerman, R.C.; Bums, B.E.; Hansen, S.P., Small Water System Treatment
Costs (Park Ridge, NJ: Noyes Data Corporation, 1986).
Haller, J.F.; Northgraves, W.W., 1955, Tappi 38:199.
Hango, R.A.; Doane, F.; Bollyky, L.J., 1981, "Wastewater Treatment for ReUse
in Integrated Circuit Manufacturing", in Wasser Berlin '81. 5. Ozon-
Weltkongress (Berlin, Federal Republic of Germany: Colloquium Verlag
Otto H. Hess, 1981), pp. 303-313.
Hansen, S.P.; Gumerman, R.C.; Gulp, R.L., 1979, "Estimating Water Treatment
Costs. Volume 3. Cost Curves Applicable to 2,500 to 1 mgd Treatment
Plants", U.S. EPA Report No. EPA-600/2-79-162c (Cincinnati, OH: U.S.
EPA, Water Engineering Research Laboratory, 1979).
Hoffman, J.; Eichelsdorfer, D., 1971, "On the Action of Ozone on Pesticides
with Chlorinated Hydrocarbon Groups in Water", Wasser 38:197-206.
Laplanche, A.; Martin, G.; Tonnard, F., 1983, "Ozonation Schemes of
Organophosphorus Pesticides. Application in Drinking Water Treatment",
in Sixth Ozone World Congress Proceedings (Norwalk, CT: Intl. Ozone
Assoc., 1983), pp. 94-95.
Lykins, B.W., Jr.; Griese, M.H., 1986, "Using Chlorine Dioxide for
Trihalomethane Control", J. Am. Water Works Assoc. 78:88-93.
190
-------
Masschelein, W.J., 1979, Chlorine Dioxide: Chemistry and EnwoamentaJ
Impact of Oxychiorine Compounds (Ann Arbor, MI: Ann Arbor Science
Publishers, Inc., 1979).
Masschelein, W.J., 1982, "Contacting of Ozone with Water and Contactor Off-
Gas Treatment", in Handbook of Ozone Technology and Applications.
Volume One. R.G. Rice & A. Netzer, Editors (Ann Arbor, MI: Ann Arbor
Science Publishers, Inc., 1982), pp. 143-224.
McKeon, W.R.; Muldowney, JJ.; Aptowicz, B.S., 1986, "The Evolution of a
Modified Strategy to Reduce Trihalomethane Formation", in Proc. Annual
Meeting (Denver, CO: Am, Water Works Assoc., 1986), pp. 967-997.
Miller, G.W.; Rice, R.G.; Robson, C.M.; Scullin, R.L.; Kiihn, W.; Wolf, H., 1978,
"An Assessment of Ozone and Chlorine Dioxide Technologies for
Treatment of Municipal Water Supplies", EPA Report No. EPA-600/2-78-
147.
Morris, J.C., 1967, "Kinetics of Reactions Between Aqueous Chlorine and
Nitrogen Compounds", in Principles and Applications of Water Chemistry.
S.D. Faust & J.V. Hunter, Editors (New York, NY: John Wiley & Sons,
Inc., 1967), pp. 23-53.
Myers, G.L.; Thompson, A.; Owen, D.M.; Baker, J.M., 1986, "Control of
Trihalomethanes and Taste and Odor at Galveston County Water
Authority", in Proc. Annual Meeting (Denver, CO: Am. Water Works
Assoc., 1986), pp. 1667-1675.
National Academy of Sciences, 1980, Drinking Water and Health. Vol! 2
(Washington, DC: National Academy Press, 1980), Chapter 2.
National Academy of Sciences, 1987, Drinking Water and t^ealth.
Disinfectants and Disinfectant By-Products. Volume 7 (Washington, DC:
National Academy Press, 1987).
Neukrug, H.M.; Smith, M.G.; Maloney, S.W.; Suffett, I.H., 1984, "Biological
Activated Carbon - At What Cost?", J. Am. Water Works Assoc.
Rachwal, A.J.; Bauer, M.; Chipps, M., 1987, "Ozone's Role in Biological
Filtration Processes", presented at Second Intl. Conf. on The Role of
Ozone in Water and Wastewater Treatment, Edmonton, Alberta, Canada.
Rice, E.W.; Hoff, J.C., 1981, "Inactivation of Giardia lamhlia Cysts by
Ultraviolet Radiation", Applied & Environmental Microbiology, 42(3):546-
547, Sept. 1981.
Rice R.G., 1980, "The Use of Ozone to Control Trihalometfianes in Drinking
Water Treatment", Ozone Science & Engineering 2:75-99 (1980).
Rice, R.G, 1987a, "Rationales for Multiple Stage Ozonation in Drinking Water
Treatment Plants", Ozone: Science & Engineering 9(l):37-62.
191
-------
Rice, R.G., 1987b, "Applications of Ozone in Soft Drink Bottling Plants",
presented at 34th Annual Meeting, Soc. Soft Drink Technologists, Las
Vegas, NV.
Rice, R.G.; Gomez-Taylor, M, 1986, "Occurrence of By-Products of Strong
Oxidants Reacting with Drinking Water Contaminants - Scope of the
Problem", Environmental Health Perspectives 69:31-44.
Schalekamp, M., 1986, "Pre- and Intermediate Oxidation of Drinking Water
with Ozone, Chlorine, and Chlorine Dioxide", Ozone: Science &
Engineering 8(2):151-186.
Scheible, O.K.; Kreft, P.; Venosa, A.D., 1985, "Demonstration of Process
Design and Evaluation Procedures for Ultraviolet Disinfection", Summary
prepared for U.S. EPA, Water Engineering Research Laboratory,
Cincinnati, OH, Contract No. 68-03-1821.
Scheible, O.K.; Casey, M.C.; Forndran, A., 1986, "Ultraviolet Disinfection of
Wastewaters from Secondary Effluent and Combined Sewer Overflows",
EPA Report No. EPA-600/2-86/005 (Cincinnati, OH, U.S. EPA, Water
Engineering Research Laboratory). NTIS No. PB86-145182.
Seeger, D.R.; Moore, L.A.; Stevens, A.A., 1984, "Formation of Acidic Trace
Organic By-Products From the Chlorination of Humic Acids", U.S. EPA
Report No. 600/D-84-159; NTIS Report No. PB84-201722.
Singer, P.S., 1986, "THM Control Using Alternate Oxidant and Disinfectant
Strategies: An Evaluation", in Proc. 1986 Annual Conference (Denver, CO:
Am. Water Works Assoc.), pp. 999-1017.
Sontheimer, H., 1985, in "Trends in Ozonation: Roundtable Discussion", J. Am.
Water Works Assoc. 77(8):30.
Stevens, A.A.; Moore, L.; Dressman, R.C.; Seeger, D.R., 1985, "Disinfectant
Chemistry in Drinking Water - Overview of Impacts on Drinking Water
Quality", in Safe Drinking Water: The Impact of Chemicals on A Limited
Resource. R.G. Rice, Editor (Chelsea, MI: Lewis Publishers, Inc., 1985),
pp. 87-108.
Stevens, A.A.; Miltner, R.J.; Moore, L.A.; Slocum, C.J.; Nash, H.D.; Reasoner,
D.J.; Berman, D. (1987a), "Detection and Control of Chlorination By-
Products in Drinking Water", in Proc. Conference on Current Research in
Drinking Water Treatment (Cincinnati, OH: U.S. EPA, Water Engineering
Research Lab., 1987).
Stevens, A.A.; Moore, L.A.; Slocum, C.J.; Smith, B.L.; Seeger, D.R.; Ireland,
J.C., 1987b, "Chlorinated Humic Acid Mixtures Establish Criteria for
Detection of Disinfection By-Products in Drinking Water", Drinking Water
Research Division, Water Engineering Research Laboratory, U.S. EPA,
Cincinnati, OH.
192
-------
Stevens, A.A.; Moore, L.A.; Slocum, C.J.; Smith, B.L.; Seeger, D.R.; Ireland,
J.C., 1987c, "By-Products of Chlorination at Ten Operating Utilities
presented at Sixth Conference on Water Chlorination: Environmental
Impacts and Health Effects, May 3-8, 1987, Oak Ridge Associated
Universities, Oak Ridge TN.
U.S. Environmental Protection Agency, 1979, "Control of Trihalomethanes in
Drinking Water; Final Rule", Federal Register 44(231):68623-68642.
U.S. Environmental Protection Agency, 1981, "Technologies and Costs for the
Removal of Trihalomethanes from Drinking Water" (Washington, DC:
Science and Technology Branch, Criteria and Standards Division, Office of
Drinking Water, Dec. 9, 1981).
U.S. Environmental Protection Agency, 1983, "Microorganism Removal for
Small Water Systems", Report No. EPA 570/9-83-012 (Washington, DC:
U.S. EPA, Office of Drinking Water, June 1983).
U.S. Environmental Protection Agency, 1985, "National Primary Drinking
Water Regulations: Synthetic Organic Chemicals, Inorganic Chemicals and
Microorganisms; Proposed Rule", Federal Register 50(219):46935-47022, Nov.
13, 1985.
U.S. Environmental Protection Agency, 1986, "Design Manual: Municipal
Wastewater Disinfection", Report No. EPA/625/1-86/021 (October 1986).
U.S. EPA, Office of Research and Development, Water Engineering
Research Laboratory, Center for Environmental Research Information,
Cincinnati, OH 45268.
U.S. Environmental Protection Agency, 1987a, "Water Pollution Control:
National Primary Drinking Water Regulations; Filtration, Disinfection,
Turbidity, Giardia lamblia. Viruses, Legionella. Heterotrophic Bacteria;
Proposed Rule", Federal Register 52(212):42177-42222.
U.S. Environmental Protection Agency, 1987b, 'Technologies and Costs for
the Removal of Microbiological Contaminants from Potable Water
Supplies", Draft Document, available from Office of Drinking Water,
Criteria and Standards Division.
U.S. Environmental Protection Agency, 1987c, "Concept Outline: Development
of Best Available Technology Criteria", Science and Technology Branch,
Criteria and Standards Division, Office of Drinking Water, March 31,
1987.
U.S. Environmental Protection Agency, 1987d, "Guidance Manual for
Compliance With the Filtration and Disinfection Requirements for Public
Water Systems Using Surface Water Sources", draft dated October 10,
1987. Science and Technology Branch, Criteria and Standards Division,
Office of Drinking Water, U.S. EPA, Washington, DC.
193
-------
U.S. Environmental Protection Agency, 1988a, "Drinking Water; Substitution
of Contaminants and Drinking Water Priority List of Additional Substances
Which May Require Regulation Under the Safe Drinking Water Act",
Federal Register 53(14):1891-1902.
Weil, I.; Morris, J.C., 1949, "Kinetic Studies on the Chloramines. I. The
Rates of Formation of Monochloramine, N-Chloromethylamine and N-
Chlorodimethylamine", J. Am. Chem. Soc. 71:1664-1671.
Werdehoff, K.S.; Singer, P.S., 1986, "Effects of Chlorine Dioxide on
Trihalomethane and Total Organic Halide Formation Potentials and on the
Formation of Inorganic By-Products", in Proc. Annual Conference (Denver,
CO: Am. Water Works Assoc., 1986), pp. 347-364.
194
-------
DISINFECTION / BY-PRODUCTS WORKSHOP
SURFACE WATER PROBLEM
I. RAW WATER PARAMETERS
Source : Unprotected Watershed
Turbidity : 5-10 NTU
Total Coliforms : 500 - 5,000 per 100 mL
THMFP : 350 /tg/L
TOXFP : 500 jig/L
TOC : 5-10 mg/L (mostly humics)
[NH4+] : 0.1 mg/L
[organic-N] : 0.5 mg/L
[Br] : 0.1 mg/L
C\2 Demand : 7 mg/L
C1O2 Demand : 2 mg/L
KMnC«4 Demand : 4 mg/L
Ozone Demand : 3 mg/L
II. TREATMENT OBJECTIVES
Disinfection
TTHMs : 100 /ig/L
III. REGULATORY FACTORS
(C102 + C1O2- + C103-) MCL : 1.0 mg/L
Chloramine MCL : 2.5 mg/L
[Residual Disinfectant] : detectable residual
-------
SURFACE WATER TREATMENT PROBLEM
Modification # 1
I. RAW WATER QUALITY - Same
II. TREATMENT OBJECTIVE - TTHMs = 50 /tg/L
III. REGULATORY FACTORS
(C1C-2 + C1O2' + 003-) MCL : 0.5 mg/L
Chloramine MCL : 1.0 mg/L
[Residual Disinfectant] : detectable residual
SURFACE WATER TREATMENT PROBLEM
Modification # 2
I. RAW WATER QUALITY - Same
II. TREATMENT OBJECTIVE - TTHMs = 25
III. REGULATORY FACTORS
(C1O2 + C1O2- + C1O3-) MCL : 0.2 mg/L
Chloramine MCL : 0.25 mg/L
[Residual Disinfectant] : detectable residual
3/30/88
-------
DISINFECTION / BY-PRODUCTS WORKSHOP
GROUNDWATER PROBLEM
I. RAW WATER PARAMETERS
Source : Aquifer - Not Directly Influenced
by Surface Water
Turbidity : 0.3 NTU
Total Coliforms : 0-100 per 100 mL
THMFP : 200 /tg/L
TOXFP : 300 /ig/L
TOC : 3-5 mg/L
[NH4+] : 0.1 mg/L
[organic-N] : 0.5 mg/L
[Br] : 0.1 mg/L
Iron : 0.6 mg/L
Manganese : 0.30 mg/L
C\2 Demand : 4 mg/L
C1O2 Demand : 1.3 mg/L
KMnO4 Demand : 2 mg/L
Ozone Demand : 3 mg/L
II. TREATMENT OBJECTIVES
Disinfection
TTHMs : 100 /ig/L
III. REGULATORY FACTORS
(C1O2 + C1O2- + C1O3') MCL : 1.0 mg/L
Chloramine MCL : 2.5 mg/L
[Residual Disinfectant] : detectable residual
-------
GROUNDWATER TREATMENT PROBLEM
Modification # 1
I. RAW WATER QUALITY - Same
II. TREATMENT OBJECTIVE - TTHMs = 50 fig/L
III. REGULATORY FACTORS
(C1O2 + C1O2' + C1O3-) MCL : 0.5 mg/L
Chloramine MCL : 1.0 mg/L
[Residual Disinfectant] : detectable residual
GROUNDWATER TREATMENT PROBLEM
Modification # 2
I. RAW WATER QUALITY - Same
II. TREATMENT OBJECTIVE - TTHMs = 25
III. REGULATORY FACTORS
(C1O2 + 002" + C1O3-) MCL : 0.2 mg/L
Chloramine MCL : 0.25 mg/L
[Residual Disinfectant] : detectable residual
3/30/88
-------
Technical Session: Organics
John E. Dyksen, Senior Project Manager, Malcolm Pirnie, Inc.,
Paramus, NJ
VII-1
-------
WORKSHOP ON
EMERGING TECHNOLOGIES FOR
DRINKING WATER TREATMENT
ORGANICS TREATMENT TECHNIQUES
OVERVIEW
Conducted by:
United States Environmental Protection Agency
and
Association of State Drinking Water Administrators
Presented By:
John-E. Dyksen
Senior Manager, Water Treatment
Malcolm Pirnie, Inc.
Paramus, New Jersey
-------
WORKSHOP ON
EMERGING TECHNOLOGIES FOR
DRINKING WATER TREATMENT
ORGANICS TREATMENT TECHNIQUES
OVERVIEW
TABLE OF CONTENTS
Page
I. INTRODUCTION 1
A. Purpose and Scope 1
II. AVAILABLE ORGANICS TREATMENT TECHNOLOGIES 2
A. Summary of Available Technologies 2
B. Other Applicable Technologies 2
III. GRANULAR ACTIVATED CARBON 9
A. Process Description 9
B. Process Design Consideration 9
C. Testing to Evaluate Process Design Parameters 11
D. Optimizing Design Criteria 13
E. Facility Design Considerations 14
F. Applicability to Various*System Sizes 15
G. Treatment Economics , ^
-Lb
IV. PACKED COLUMN AERATION
A. Process Description
B. Process Design Considerations 18
C. Testing to Evaluate Process Design Parameters 19
D. Development of Design Criteria 20
E. Facility Design Considerations 21
-------
TABLE OF CONTENTS (Continued)
IV. PACKED COLUMN AERATION (Continued)
F. Applicability to Various System Sizes
G. Treatment Economics
V. BIBLIOGRAPHY
Page
24
24
26
Table
No.
2
3
4
5
6
7
8
9
Figure
No.
1
2
3
4
LIST OF TABLES
Description
Typical Performance of PAC and Conventional
Treatment Processes
Typical Performance of Diffused Aeration
Typical Performance of Slat Tray Aeration
Typical Performance of Ozonation Process
Typical Performance of Reverse Osmosis Process
Readily Adsorbed Organics
Poorly Adsorbed Organics
Summary of Carbon Usage Rates
Freundlich Isotherm Constants for Carbon Adsorption
LIST OF FIGURES
Description
Schematic of PAC Adsorption Process
Schematic of Diffused Aeration Process
Home Diffused Aeration System
Schematic of a Redwood Slat Tray Aerator
Following
Page
3
4
4
5
9
9
10
12
Following
Page
2
3
3
4
-------
TABLE OF CONTENTS (Continued)
Figure
No.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
LIST OF FIGURES (Continued)
Description
Schematic of Ozonation Process
Schematic of Reverse Osmosis Process
Schematic of Mechanical Aeration Process
Schematic of Catenary Grid System
Schematic of Higee System
Effect of Type of Compound on Carbon Life
Effect of Contaminant Levels on Carbon Life
Effect of Carbon Type on Facility Cost
Empty Bed Contact Time versus Percent Radon Remaining
Steady-state Adsorption/Decay Curve for Radon
Carbon Isotherm Test Results
Diagram of Dynamic Mini-Column Adsorption
Technique System
Diagram of Pilot Column Test System
GAG Facility Cost Components
Breakthrough Curves for c-1,2 DCE
Breakthrough Curves for TOG
EBCT versus Carbon Usage for c-1,2 DCE
EBCT versus Carbon Usage for TOC
Parallel Mode of Operation
Series Mode of Operation
Cost Comparison of Parallel versus Series Flow
•p /^ v I-. T "i r\/~n™i
Following
Page
4
5
6
6
7
10
10
10
11
11
12
12
13
13
13
13
13
13
13
14
14
-------
TABLE OF CONTENTS (Continued)
LIST OF FIGURES (Continued)
Figure Following
No. Description Page
26 Cost Comparison of Parallel versus Series Flow 14
for TOC
27 Schematic of Pressure GAC Contactors 14
28 Schematic of Downflow Gravity GAC Contactors 14
29 Cost Comparison of Carbon Recharge Options 15
30 Typical Home GAC Unit 15
31 Capital Cost Curves for GAC Facilities 16
32 O&M Cost Curves for GAC Facilities 16
33 Diagram of Packed Column Aeration 18
34 Effect of Compound on Packed Column Design 18
35 Effect of Temperature on Packed Column Design 19
36 Schematic of Pilot Aeration Column 19
37 Mass Transfer Relationships for Packed Columns 21
38 Packing Height versus Air to Water Ratio 21
39 Schematic of Typical Packed Column Facility 21
40 Distributor Types 22
41 Schematic of Catalytic Incineration Process 23
42 Schematic of Vapor Phase GAC System 23
43 Capital Cost Curves for Packed Column Aeration Facilities 24
44 O&M Cost Curves for Packed Column Aeration Facilities 24
-------
I. INTRODUCTION
A. Purpose and Scope
The June 1986 Amendments to the Safe Drinking Water Act (SDWA) require
the United States Environmental Protection Agency (EPA) to set maximum contam-
inant levels (MCLs) for several contaminants found in drinking water. The
MCLs are to be established based upon:
1. Health goals
2. Effectiveness of treatment technologies in removing the contaminants
3. Level of treatment that is affordable for the water supply systems
EPA is currently establishing MCLs for radon and a number of organic
chemicals occurring in contaminated water supplies. According to the
Amendments, all public water systems will be required to come as close as
possible to meeting the MCLs by using the Best Available Treatment (BAT)
technology.
The purpose of this document is two-fold:
1. To assist water utilities in selecting appropriate treatment methods
to meet the regulations.
2. To assist the regulatory agencies in assessing the feasibility of
treatment methods proposed by the water utilities.
Provided in this document is a review of all the emerging treatment
technologies for removing organic contaminants and radon from drinking water.
Emphasis is given to treatment methods that have been proven effective for
their removals. The discussion on these treatment methods includes the
process selection criteria, treatability tests, design considerations, opera-
tional issues and facility costs.
-1-
-------
II. AVAILABLE ORGANICS TREATMENT TECHNOLOGIES
A. Summary of Available Technologies
1. Most Applicable Technologies
a. Granular activated carbon (GAC) adsorption
b. Packed column aeration
2. Other Applicable Technologies
a. Powdered activated carbon (PAC) and conventional treatment
b. Diffused aeration
c. Multiple tray aeration
3. Additional Technologies (Emerging)
a. Ozone/UV
b. Reverse osmosis
c. Mechanical aeration
d. Catenary grid (aeration)
e. Higee (aeration)
f. Resins
B. Other Applicable Technologies
1. Powdered Activated Carbon. (PAC) and Conventional Treatment
a. Based on the principles of adsorption, PAC is generally added
at one or more application points in a conventional treatment
train (coagulation, flocculation, sedimentation and filtra-
tion) . Principal design considerations are:
Dosage
Contact time (at least 15 minutes)
- Point(s) of application (1 to 3)
A schematic of the PAC treatment process is shown on Figure 1.
b. Typical performances of PAC and conventional treatment process-
es are presented in Table 1. As can be seen, PAC addition
achieves higher removals than conventional treatment alone.
c. Principal advantages of PAC:
Better organics removals than conventional treatment
- Acts as coagulant aid
- Taste and odor removal
-2-
-------
CHEMICALS-POWDERED ACTIVATED CARBON
DISINFECTANT
FILTER AID
RAPID MIX
FLOCCULATION
SEDIMENTATION FILTRATION
FIGURE 1 - SCHEMATIC OF P.A.C.
ADSORPTION PROCESS
-------
TABLE 1
TYPICAL PERFORMANCE OF PAC AND
CONVENTIONAL TREATMENT PROCESSES
Conventional
Treatment
Percent PAC
Removal Dosage (mg/L) Percent Removal
VOCs
Carbon Tetrachloride - 9.6-30 0-25
1,1,1-Trichloroethane - 7 40-65
SOCs
Acrylamide 58 13
Alachlor <50 4-34 36 - 100
Carbofuran 54-79 9-25 45-75
o-Dichlorobenzene - 8-27 38-95
2,4-D 0-3 11 - 306 69 - 100
Ethylbenzene - 8-27 33 - >99
Heptachlor 64 11 - 97 53 - 97
Lindane 10-20 2-34 82-97
Monochlorobenzene - 8-27 14 - >99
Toluene - 8-27 0-67
2,4,5-TP 63 1.5 - 17 82 - 99
Toxaphene - 1-44 40-99
Xylenes - 8-27 60 - >99
Note;
(-) Information not available.
-------
- Useful for short-term, emergency applications in conjunc-
tion with conventional treatment
d. Limitations:
- Excessive dosages may be required for certain contaminant
removal.
Limited applicability in locations which have certain
constraints (e.g., hydraulic, space, sludge handling) and
also where conventional treatment is not in place or is
not required.
Diffused Aeration
a. Based on bubbling air into a water-filled contact chamber
through a diffuser mechanism usually located at or near the
bottom of the chamber. Principle design considerations are:
- Basin depth (5 to 10 feet)
- Diffuser type
- Air/Water ratio (5:1 to 15:1)
- Detention time (10 to 15 minutes)
Basin geometry
A schematic of diffused aeration process is shown on Figure 2.
A typical home diffusion aeration unit for radon removal is
shown on Figure 3.
b. Typical performance of diffused aeration at an air to water
ratio of 5:1 to 15:1 and a contact time of 10 to 15 minutes is
presented in Table 2.
c. Principle advantage of diffused aeration:
- Existing basins such as a chlorine contact basin can be
retrofitted with diffusers and converted to aeration
basins.
d. Limitations:
- Less effective than packed column aeration (discussed
later).
- Generally used only where existing basins are available.
Multiple Tray Aeration
a. Water falls over a series of trays with slats, perforations, or
wire mesh. The counter-current flow of air (either from
-3-
-------
AIR SUPPLY
INFLUENT
DIFFUSER GRID -^3
."««•"'
<.
RIO
T-
' •
c.
. „
I
•
•
*r-
-,
•
-
•
.
• — . _: •_ : —
'M
•'•f
.
. i
- 1,
. 5f
• V
'•'• ',
' r
EFFLUENT
FIGURE 2- SCHEMATIC OF
DIFFUSED AERATION PROCESS
-------
PL'MP/LEVEL -
CONTROL SOX
RAW WATER FROM
PUMP
VARIABLE TIMER ~P~j
AIR-STRIPPED RACON
TO CUTSiDE VEST
• AERAT'.CN TANK
UCL'.O LEVEL
PROSE
PUMP
TO S1ETERED
HOUSEHOLD
WATER USE
AIR CIFFUSER
HTCRCPS'EUMATiC
TANK
( )
AIR COMPRESSOR
OR PUMP
AIR FLOW
REGULATOR
FIGURE 3- HOME DIFFUSED AERATION SYSTEM
-------
TABLE 2
TYPICAL PERFORMANCE OF DIFFUSED AERATION
% Removal
VOCs
Trichloroethylene 53-95
Tetrachloroethylene . 73-95
1,2-Dichloroethane 42-77
1,1-Dichloroethylene 97
1,1,1-Trichloroethane 58-90
SOCs
Carbofuran 11-20
1,2-Dichloropropane 12-79
cis-l,2-Dichloroethylene 32-85
trans-l,2-Dichloroethylene 37-96
o-Dichlorobenzene 14-72
Ethylbenzene 24-89
Monochlorobenzene 14-85
Toluene 22-89
Xylenes 18-89
-------
blowers or natural ventilation) removes the volatile organics
in the water. Principal design considerations are:
- Tray type (wood, plastic)
- Tray height (12 to 16 feet)
- Natural or forced draft
- Air/Water ratio (30:1)
A schematic of a slat tray aeration system is shown on Fig-
ure 4.
b. Typical performance of slat tray aeration at an air to water
ratio of about 30:1 and tray height of 12 to 16 feet is
presented in Table 3.
c. Principle advantages of multiple tray aeration:
Less susceptible to clogging, from iron and manganese
precipitation, than packed column aeration.
Readily available as package units from manufacturers.
d. Limitations:
Not as effective as packed column aeration (discussed
later)
- Control of biological growth and corrosion may be
required.
B. Additional Technologies (Emerging)
1. Ozone/UV
a. Ozone provides organics removal by acting as an oxidant,
generally breaking down unsaturated bonds within the molecules.
Ultra-violet (UV) radiation catalyzes this breakdown reaction.
Principal design considerations are:
Ozone dosage
Contact time (15 minutes)
Radiation intensity
A schematic of the ozonation process is shown on Figure 5.
b. Typical performance of the ozonation process at a contact time
of about 15 minutes is presented in Table 4.
c. Principle advantages of Ozonation/UV process:
- Can combine organics removal with disinfection, and taste
and odor control
-4-
-------
INLET
CHAMBER
DISTRIBUTOR
NIPPLES
STAGGERED
SLAT TRAYS
AIR INLET
BLOWER
AIR
SEAL
WATER
INLET
-Tt
AIR
— OUTLET
'
-- u,
-T»-
•— tb 1)
'i 1 *^~
— ; . — .
•
MLJ: T
LI"- 'T.T~".r!'' i
c\y '
-»-= — -n-1
- BAFFLES
AiR STACKS
AIR SEAL
WATER
OUTLET
FIGURE 4- SCHEMATIC OF A REDWOOD SLAT TRAY AERATOR
-------
TABLE 3
TYPICAL PERFORMANCE OF SLAT TRAY AERATION
VOCs % Removal
Trichloroethylene 30-90
Tetrachloroethylene 20-85
-------
OXIDANT
(OZONE)
RAW WATER
CONTACT BASIN
•^•PRODUCT WATER
FIGURE 5- SCHEMATIC OF OZONATION PROCESS
-------
TABLE 4
TYPICAL PERFORMANCE OF OZONATION PROCESS
Dosage (mg/L)
SOCs
Carbofuran 9
1,2-Dichloropropane 0.9-6
cis-1,2 Dichloroethylene 2-10
trans-1,2 Dichloroethylene 0.9-6
o-Dichlorobenzene 9
Ethylene Dibromide 0.9-6
Ethylbenzene 1.5-9
Heptachlor 17
Heptachlor epoxide 17
Lindane 0.4 - 150
Monochlorobenzene 0.4-6
Toluene 1.5 - 12
Xylenes 1.5 - 12
% Removal
100
8-22
87 - 93
100
88
8-9
47 - 95
100
26
0 - 100
86 - 98
49 - 98
54 - 98
-------
- Waste disposal not required
- No THM formation
d. Limitations:
- May not achieve complete oxidation, concern about inter-
mediate breakdown products.
- Background matrix and presence of other contaminants will
impact effectiveness.
2. Reverse Osmosis
a. Based on the application of hydrostatic pressure to drive the
feed water through a semipermeable membrane while a major
portion of its impurity content remains behind and is dis-
charged as waste. Principal design considerations are:
Operating pressure (150 to 250 psi)
Percent recovery (10 to 50 percent)
Membrane type
Pretreatment requirements
A schematic of the reverse osmosis process is shown on Fig-
ure 6.
b. Generally, reverse osmosis shows promise for removing low
molecular weight, volatile organic contaminants and certain
pesticides. Typical performance of reverse osmosis process at
an operating pressure of about 200 psi is presented in Table 5.
c. Principal advantage of reverse osmosis process:
- It is a proven technology for inorganics removal and may,
therefore, be very useful in situations where a dual
inorganic/organic problem is present.
d. Limitations:
Considerable pretreatment may be required, depending on
the raw water quality.
Low recovery.
3. Mechanical Aeration
a. Surface or submerged mechanical aerators are used to entrain
atmospheric air into the water by creating a region of intense
turbulence in the water. Volatile organics are removed by the
dispersed air. Principal design considerations are:
-5-
-------
LEGEND
I. SOURCE WATER PUMPING
2. POLYMER FEED SYSTEM '
3. FILTRATION
4. HIGH PRESSURE PUMP
5. pH AND SCALE CONTROL
6. REVERSE OSMOSIS
7. FILTER WASH WATER STORAGE
8. DISINFECTION-
9. FILTER BACKWASH
10. WASTEWATER DISPOSAL
I I. CARTRIDGE FILTRATION
EXISTING FACILITIES
FIGURE 6- SCHEMATIC OF REVERSE OSMOSIS PROCESS
-------
TABLE 5
TYPICAL PERFORMANCE OF REVERSE OSMOSIS PROCESS
% Removal
VOCs
1,2-Dich'loroethane 15-70
1,1,1-Trichloroethane 15-100
Carbon tetrachloride 95
Trichloroethylene 0-75
Tetrachloroethylene 70-90
Benzene 2-18
p-Dichlorobenzene 0-10
SOCs
Acrylamide 0-97
Aldicarb 94-99
Alachlor 100
Carbofuran 86-99
1,2-Dichloropropane 10-90
cis-1,2-Dichloroethylene 0-30
trans 1,2-Dichloroethylene 0-30
2,4-D 1-65
o-Dichlorobenzene 65
Ethylbenzene 30
Ethylene dibromide 37-84
Lindane 50-75
Methoxychlor >90
Monochlorobenzene 50-100
PCBs 95
Xylenes 10-85
-------
- Aerator type
- Detention time (1 to 2 hours)
Energy requirements
Schematics of the two types of mechanical aerators are shown on
Figure 7.
b. Performance of mechanical aerators for the removal of specific
organics is not available.
c. Principal advantages of mechanical aerators appear to be:
- Simplicity of operation
- Applicable where there is a risk of chemical precipitation
or biological growth that would tend to clog a packed
column aerator
d. Limitations:
High energy requirement
- Long detention time and, therefore, large area requirement
Potential freezing of open aeration basins in cold climate
Generally used for wastewater treatment
4 . Catenary Grid Aeration
a. Based on bringing the raw water in contact with air and trans-
ferring the organics from water to air phase. Water flows down
a column which is divided into sections by catenary grids. The
shape of the catenary grid is designed so that a zone of
turbulence is created above each grid section. Principal
design considerations are:
Air/water ratio
- Number of grids
Hydraulic loading rate
A schematic of the catenary grid aerator is shown on Figure 8.
b. Catenary grids can provide VOC removals that are comparable to
packed column aeration. In general:
Removal increases with increasing air/water ratio.
Removal increases with increasing number of grids.
- Mass transfer coefficient decreases with increasing number
of grids.
-6-
-------
Drive
Mechanical Surface Aerator
Drive
Compressor
. . i.arapi:c
I '.Nl
Turbine
Sparger
Suboerged Turbine Aerator
FIGURE 7- SCHEMATIC OF MECHANECAL
AERATION PROCESS
-------
DEMISTER
FLUIDIZED
ZONE
CATENARY
GRID (TYP.)
TREATED WATER
SAMPLE COLLECTOR
MANOMETER FOR
AIR FLOW RATE
MEASUREMENT-r
BLOWER 7
AIR FLOW
DAMPER
WATER INLET
(TYP.)
RAW WATER
ROTAMETER
rRAW WATER
\SAMPLE TAP
*
TREATED WATER
SAMPLE TRAP
WATER FLOW
METERING VAL
RAW WATER
FROM WELL
TREATED WATER
TO DRAIN
FIGURE 8- SCHEMATIC OF CATENARY GRID SYSTEM
-------
c. Principal advantage of catenary grid aerator:
Compact design: a smaller diameter and lower column
height than packed column aeration.
Lower capital cost than packed column because of more
compact design.
d. Limitations:
The scale-up procedure from pilot to full-scale is not
fully understood and requires further development.
The air flow pressure drop is relatively high, resulting
in larger blowers and higher energy costs than packed
column aeration.
Limited treatability data are available for several
organics.
5. Higee
a. The process uses a rotating bed of packing with a high surface
area per unit volume. Air is forced into the rotating packing
element while the contaminated water flows from the center of
the packing outward in a counter current manner. Principal
design considerations are:
- Air/Water ratio
Pump (Higee) size
A schematic of the Higee system is shown on Figure 9.
b. Very little treatability data is currently available for
organics removal.
c. Principal advantages of Higee system appear to be:
- Packing requirements are less than in packed column
aeration for equal removal efficiencies.
High liquid loading rates are possible with relatively
compact equipment
Air volumes required are reduced.
d. Limitations:
- Limited treatability data are available for organics
removal.
-7-
-------
EXHAUST AIR
AIR IN
BLOWER
GROUNDWATER
FILTER
HIGEE
PRODUCT
WATER
o
PUMP
FIGURE 9- SCHEMATIC OF HIGEE SYSTEM
-------
Generally considered only under special requirements such
as space constraint, height limitations or exhaust air
treatments.
Currently, treatment cost is high. Appears to be
cost-effective only for high organic concentrations
(greater than 1,000 ppm) and low flow rates (less than
100 gpm).
6. Resins
a. Process uses special synthetic resins that have high adsorptive
capacity for organic compounds. Principal design considera-
tions are:
Resin type
Empty bed contact time (EBCT)
Regeneration frequency
b. Studies indicate that Ambersorb XE 340 (made by Rohm and Haas)
is effective in removing VOCs.
c. Principal advantages of synthetic resins appear to be:
- Shorter EBCT required than GAC
Longer life than GAC
Resin regeneration on site using steam
d. Limitations:
High cost of resins - $10 per pound as compared to $.80
per pound for GAC.
Disposal of highly concentrated regenerant water.
-------
III. GRANULAR ACTIVATED CARBON
A. Process Description
1. Based on the principle of adsorption - the transfer of a dissolved
contaminant (adsorbate) from a solvent (solution) to the surface of
an adsorbent (carbon).
2. Factors Affecting Adsorption Process
a. Type of contaminant (adsorbate) - Tables 6 and 7 present lists
of readily adsorbed and poorly adsorbed organics, respectively.
branched-chain compounds more adsorbable than straight-
chained compounds
increasing molecular weight increases adsorption
lower solubility increases adsorption
greater concentration, increased adsorbability
b. Adsorbent
high degree of porosity
extensive internal surface area
affinity of adsorbate for absorbent (polar, nonpolar)
c. Aqueous Solution
- temperature
- pH
dissolved solids
other adsorbates
B. Process Design Considerations
1. Key Parameters
a. contaminant
b. levels
c. GAG type
d. empty bed contact time
e. carbon usage rate - pounds of carbon per 1,000 gallons of water
treated
f. surface loading rate (2 to 10 gpm/sf)
-9-
-------
TABLE 6
READILY ADSORBED ORGANICS
Aromatic Solvents
Benzene, toluene, nitrobenzenes
Chlorinated Aroraatics
PCBs, chlorobenzenes, chloronapthalene
Phenol and chlorophenols
Polynuclear Aromatics
Acenapthene, benzopyrenes
Pesticides and herbicides
DDT, aldrin, chlordane, heptachlor
Chlorinated non-aromatics
Carbon tetrachloride, chloroalkyl ethers
High MW Hydrocarbons
Dyes, gasoline, amines, humics
TABLE 7
POORLY ADSORBED ORGANICS
Alcohols
Low MW Ketones, Acids, and Aldehydes
Sugars and Starches
Very High MW Or Colloidal Organics
Low MW Aliphatics
-------
g. carbon depth (10-30 ft)
h. radon removal
2. Contaminant and Levels
a. Effect of different organics on GAG design is shown on Fig-
ure 10.
b. Effect of contaminant levels on GAG design is shown on Fig-
ure 11.
c. Selected pretreatment can lower contaminant loading on carbon
and thereby increase carbon life. Example:
- Packed column for volatile organic chemical (VOC)
reduction.
- Coagulation/filtration for turbidity and total organic
chemical (TOG) reduction.
2. GAG Type
a. Low cost GAG - low capital cost, but more frequent reactivation
and, therefore, higher operating cost.
b. Total cost for different GAG types are shown on Figure 12.
4. Empty Bed Contact Time (EBCT)
a. Affects capital costs
b. Average - 10 minutes for most organics
c. Calculated based on flow rate and GAG volume
EBCT (min) = GAG volume (cu ft) 7.48 _gal_
Flow rate (gal per mm) cu ft
5. Carbon Usage Rate
a. Rate at which carbon capacity will be exhausted
b. Affects facility capital and operating costs of the facility
c. Carbon usage rates for several organics are shown in Table 8.
In general, pesticides are better adsorbed than VOCs.
-10-
-------
400
w
>
u
0.
300
o 200
CQ
cc
o
100
0
EFFLUEMT CONCENTRATION 10ug/l
EBCT-10 MINUTES
TETRACHLOROETHYLENE
TRICHLOROETHYLENE
1,1,1,-TRICHLOROETHANE
0 IOO 200 300 400 500600 700 800 900 IOOO
INFLUENT CONCENTRATION/ug/l
FIGURE 10- EFFECT OF TYPE OF COMPOUND OH CARBON LIFE
-------
200
u
u.
o
CD
cc
<
o
100
0
TRICHLOROETHYLENE
10 MINUTE EBCT
EFFLUENT CONCENTRATION
50jug/i
10.ug/l
1
IOO 200 300 400 500 600 700 800 900 IOOO
INFLUENT CONCENTRATION, oig/I
FIGURE 11- EFFECT OF CONTAMINANT LEVELS ON CARBON LIFE
-------
70-
£
o
03 CARSON COST S 1 . 0 0 / I b
b= CARBON COST J 0 . I 0 / I b
« = C A S 8 ON COST 1 0 .60 / I b
a
b
c
1
4.
~2 4- 6 B X)
REACTIVATION FREQUENCY, MONTHS
T2
FIGURE 12- EFFECT OF CARBON TYPE ON FACILITY COST
-------
TABLE 8
SUMMARY OF CARBON USAGE RATES
Concentration (ug/L)
Influent Effluent
Volatile Organic Chemicals (VOCs):
Tetrachloroethylene (PCE)
Trichloroethylene (TCE)
Trichloroethane (TCA)
Pesticides:
Chlordane
Dibromochloropropane (DBCP)
Aldicarb
Chlorinated Aromatics:
Dichlorobenzene
PCB (Aroclor 1016)
100
100
100
100
100
100
100
100
2
2
2
1
1
1
2
2
Carbon Usage
(lbs/1,000 gal)
0.08
0.16
0.96
0.012
0.016
0.02
0.01
0.015
-------
6. Radon Removal
a. EBCT
- Requires large EBCT (100 to 200 minutes). EBCT required
for various removals are presented on Figure 13.
b. Adsorption/decay
Radon that is adsorbed on the carbon decays continuously
- Comparison of breakthrough curves for radon and a nonde-
caying adsorbate is shown on Figure 14.
- Decay of radon acts to continuously self-regenerate the
carbon bed.
C. Testing to Evaluate Process Design Parameters
' * . .*.*.*., >
1. Isotherm Testing , . . •
a. Useful screening tool for:
- Determining preliminary carbon usage rates.
- Evaluating relative adsorbabilities of different contami-
nants.
Evaluating the effects of temperature and pH on adsorption
Comparing the performances of different carbons
Evaluating the relative effects of -other contaminants
present in raw water
b. Procedure:
- Mix a measured weight of pulverized carbon in water of
known organic concentration and agitate for a certain
contact time.
- Measure resultant effluent organic concentration and
calculate equilibrium capacity from the amount of organic
adsorbed and the known weight of carbon in solution.
c. The relationship between equilibrium capacity and effluent
concentration has been found to generally follow Freundlich
isotherm relationship:
/ v l/n
x/m = Kc
-11-
-------
100
90
80
70
60
SO
40
30
20
CZi
= 10
a
I I
\
20 40 °° a° 100 120 MO 100 100 :oo
EMPTY DEO CONTACT TIME. MINUTES
FIGURE 13- EMPTY BED CONTACT TIME
VS PERCENT RADON REMAINING
-------
Lu
UJ
LJ
-100
75
so:
0
-TYPICAL GAC BREAKTHROUGH CURVE
FOR NGN- DECAYING A3SCR3ATE
;—TYPICAL STEADY-STATE,
25;
-
0 .
/
/ /
^Z— *—
* 1 I IVJ-^L. *J i u_^-^u i v > r^ J b«^
ADSORPTION/DECAY
OF RADON ON GAC
10 [5 20 25 30 35 40
TIME, days
FIGURE 14- STEADV-STATE ADSORPTION/DECAY
CURVE FOR RADON
-------
where:
x/m = equilibrium capacity (mg organic/gm carbon)
c = organic concentration (mg/L)
K, 1/n = Freundlich constants
d. Isotherm constants for several organics have been reported by
Dobbs and Cohen (Carbon Adsorption Isotherms for Toxic
Organics, EPA-600/8-80-023, April 1980). Examples of the test
results are shown on Figure 15 and in Table 9.
e. An approximate carbon usage can be calculated based on the
isotherm equation. For example, assume:
Trichloroethylene (TCE) influent concentration (C.) = 100 ug/L
Trichloroethylene (TCE) effluent concentration (C ) = 5 ug/L
Safety factor (SF) =0.75
Isotherm equation: x/m = Kc
Where:
K = 28 (mg) (L) 1/n (From Table 9)
(gin) (mg)
1/n =0.62 (From Table 9)
Rearranging:
Carbon Usage = Ci - Co x 8.34 (Ibs) (L)
(lbs/1,000 gal) K(Ci)1/nSF (1,000 gal) (mg)
= (100 - 5) ug I mg 8.34 Ibs L
L 1,000 ug X 1,000 gal X mg
28 ( IPO)0'62 0.75
(1000) X
=0.16 lbs/1,000 gal
2. Minicolumn Testing
a. Used to determine:
- Feasibility of carbon treatment for a given water
- Preliminary process design criteria
- Rough estimate of system economics
b. A typical setup for dynamic minicolumn test (DMCT) is shown on
Figure 16. The procedure involves running a sample of raw
water through a short GAC column (about 70 mm deep) and analyz-
-12-
-------
TABLE 9
FREUNDLICH ISOTHERM CONSTANTS
FOR CARBON ADSORPTION
COMPOUND- Tetrachlorcethene (Tetrachloroethvlene)
STRUCTURE:
Cl^ /Cl
c=r
cr
MOL. Wt.
165.83
FIEUNOIICH
PA«AME1E«S
K
l/n
pH
5.3
50.8
0.56
Coir. Co.l. r j g.96
INItlAl CONC. -g/l
1.0
ADSOtPTION CAPACITY, mg/g™
51
0.1 : 14.0
0.01 ' 3.9
0.001
1.1
COMPOUND: Triehloroethene (Trichioroethylene)
STRUCTURE:
Cl.
Cl
/
FORMULA:
CyHCIi
131.39
COMPOUND: 1,1,1-Trichlcroethane
STRUCTURE:
Cl H
I I
Cl—C—C—H
Cl H
FORMUIA: C?H3C13
FBEUNDLICH
PARAMETEtS
K
l/n
Corr. Co.l. F
INITIAl CONC. mg/l
1.0
0.1
0.01
0.001
5.3
28.0
0.6?
0.99
pH
ADSOIMION CAPACITY, mg/jm
28
6.7
-'•6
0.38
~
--—
133.41
fKUNDLICH i
PAIAMETEIS '
K
l/n
Coir. Co«l. r
INIIIAl CONC ~,,l
1.0
O.I
?.OI
O.C01
pH
5.3
2. S3
0.21
0.97 j
AOiOI'IION CAPACllY, mg/gm
2.5 {
!. 1
0.5! I
0.23
REFERENCE: OOBBS AND COHEN (1880)
-------
COMPOUND- — .-Te-trJf^ijorogthene (Tetrachiorcethy 1 e
OUND: Trichloroethene (TricMoroethylene)
ma ADSORBED / gm CARBON
5 i
^ ,
<
o.o
too
to
1.0
.xf
1
W"
ii>
k^-i
! i
i
I
i;
i .
1 I
[Till
i ' '
u^
*r!
i
|
s M'
^1
D! ' ' ' ' '"(J'.Ol ' ' ' ""&'.] ' ' ' '
RESIDUAL CONC (Cf|, mg/l
uf ~
i!
y
!l
ij. .
I
i
i
Hi
Y.'o
Tft
'fd
!
•
z ,
0
S '
<
u
I '
n '
at
O ,
t/t
O
<
a '
O.C
10
1.0
0.1
"<
jf
*
Slboi
Y
>
'6
—JL
If
.01
^
a
1
-t-
-W 1
1
1
'o.'i '
/i
* *
L J.
4-L
Tnf
J j
.0
RESIDUAL CONC (C,), mg/l
COMPOUND- .„..,„? tK|-Trich1oroetriane
m CARBON
mg ADSORBED / t
< '
i
i
o!
. — . — . ; i : .; i
10
! !
1
! i .
l.C
--•'
0.1
J
j^
i
1
"T i i
j
I i
i
T""^^
j
T
It"
L-ftp
D01 ' ' ""d 01 ' ' ' ' "o.l ' ' ' '
RESIDUAL CONC (C(). mg/l
I
*JL
^
H
==:
"l.O ' ' '
~T
-b
10
FIGURE 15- CARBOM ISOTHERM TEST RESULTS
REFERENCE: DOBBS AND COHEN (1980)
-------
11.1
YtSlOM OAI tAMTI MO 1AO
11
I . u Knout
nn ON njBjHO i«~« c • >.»«• oo
=4=1
btf OLAU (Aitftl COLLf CTKM CVMMO4
WITH ADATTU) U^TUH HU*
-QO
\
-4-
\
$
x •
V
\
X
^
s.
1
\
$
\
Ij
t&
•IJ
iv(
$
$
•3
T\
'••i
&
t^r
***•
•\
^
x-
V
v
1
-**^
\;
s
^Y
\'
X
^X
^r*
X
ii trta cotm
^«
D > i *M oo
- rntu woo.
1
I u ITAMLIU 8TUL f
DUCAT COLUMN
FIGURE 16- DIAGRAM OF DYNAMIC MINI-COLUMN
ADSORPTION SYSTEftl
-------
ing the effluent concentration. Analysis of the test results
is discussed in the following section.
3. Dynamic Column
a. Used to simulate full-scale GAC facility operation and develop
following design criteria:
- Effect of various contact times
Optimum carbon bed depth
- Effect of hydraulic loading
- Number of vessels needed
Type of GAC needed
- Carbon life/changeout frequency
Carbon exhaustion rate
Contaminant loading rates
b. A typical set-up for pilot column test is shown on Figure 17.
Raw water is run through GAC columns (about five-foot deep and
4-inch I.D.). Effluent is sampled to obtain the breakthrough
curve (time versus concentration relationship).
D. Optimizing Design Criteria
1. Optimize key design parameters to minimize treatment costs. Cost
components shown on Figure 18 indicate that key design parameters
are:
- EBCT
Carbon usage
2. Optimizing EBCT
a. Plot breakthrough curves - examples of breakthrough curves for
cis-1,2 dichloroethylene (DCE) and TOC are given on Figures 19
and 20, respectively.
b. Calculate carbon usage rates:
CU (lbs/1000 gal) = Mass of Carbon (Ibs) x 1,000 gallons
Volume Treated to Breakthrough (gallons)
c. Plot carbon usage versus EBCT - example for DCE and TOC are
shown on Figures 21 and 22, respectively.
d. Select optimum EBCT - 15 minutes for DCE, 20 minutes for .TOC
3. Parallel versus series flow
a. Parallel flow (shown on Figure 23) necessitates carbon
replacement at breakthrough
-13-
-------
FEED PRE
TANK- TREATMENT1
CARBON
FEED
TANK
CARBON COLUMNS
SAMPLE
TAPS -
Pretreatment tanks optional depending on the
suspended solids concentration in feed.
PRODUCT
WATER
BACK WASH
WATER
FIGURE 17- DIAGRAM OF PILOT COLUMN TEST SYSTEM
-------
CARBON CONTACTOR
(50%)
Captiai Cost
CARBON REGENERATION
(30%)
CARBON TRANSFER
AND STORAGE (5%)
SITEWORK ETC.
(15%)
LABOR (10%)
O&M Cost
CARBON REGENERATION
(75%)
POWER (10%)
MAINTENANCE
MATERIAL (5%)
FIGURE 18- GAC FACILITY COST COMPONENTS
-------
CT
•-*
o
i-
<
LU
o
6
o
60
50
40
30
20
10
EBCT -
7.2 MINUTES -
/
V
/
-
rEBCT =
15.2 MINUTES — v
\
I
1
,
,'
//
/^/" EBCT =
' -y- 16.4 MINUTES
/
J
/
j
P
j
t
too
200
TIME (DAYS)
Note: Ci = Influent Concentration
Ce= Effluent Concentration
1.O
0.8
0.6
Ce
tO.4
300
FIGURE 19- BREAKTHROUGH CURVES FOR c-1,2 DCE
-------
V.
Cl
E
2
LJ
O
2
O
O
EBCT = i
10 MINUTES!
EBCT - •
20 MINUTES
_ E2CT =
;30 MINUTES!
40 MINUTES |
40
80
120
160
200
i.o
0.75
0.5
0.25
Ce
CJ
240
TIME (DAYS)
Note: Ce = Effluent Concentration
Ci = influent Concentration
FIGURE 20- BREAKTHROUGH CURVES FOR TOO
-------
O)
o
O
o
LU
O
<
CO
D
Z
O
CQ
DC
<
O
0.25
0.2
BREAKTHROUGH
10
15
20
EBCT (MINUTES)
FIGURE 21- EBCT VERSUS CARBON USAGE FOR c-1,2 DCE
-------
o
O
o
O
HJ
o
:D
2
O
03
cc
<
o
BREAKTHROUGH
10
20
30
40
50
EBCT (MINUTES)
FIGURE 22- EBCT VERSUS CARBON USAGE FOR TOC
-------
INFLUENT
EFFLUENT
o
z
Q
o
m
cc
<
o
MAXIMUM LOADING
UNUSED CAPACITY
BED DEPTH
FIGURE 23- PARALLEL MODE OF OPERATION
-------
b. Series flow (shown on Figure 24) provides for carbon
replacement at exhaustion
c. Optimize between number of contactors (capital cost) and
regeneration frequency (operating cost)
d. Cost comparison between the two modes are shown on Figures 25
and 26 for DCE and TOC, respectively.
E. Facility Design Considerations
1. Major Process Elements
a. Carbon contactors
b. Regeneration system
2. Carbon Contactor Configuration
a. Upflow
Can be operated in parallel or series mode
Bed allows suspended solids to pass without developing
excessive pressure drop
Release of carbon fines in effluent stream is of concern
More widespread use for wastewater treatment
b. Downflow
Can be operated in parallel or series mode
Small amount of suspended matter can be handled by back-
washing
Can be operated in pressure or gravity mode
Pressure contactor (shown on Figure 27) offers more
flexibility as the system can be operated with higher head
loss
- Gravity contactor (shown on Figure 28) are less expensive
due to common wall construction
3. GAC regeneration:
a. On-Site Regeneration - economical where carbon exhaustion rate
is greater than 1,500 pounds per day.
-14-
-------
INFLUENT
o
EFFLUENT
O
o
_J
2:
o
CQ
CC
<
O
MAXIMUM LOADING
UNUSED CAPACITY
BED DEPTH
FIGURE 24- SERIES MODE OF OPERATION
-------
10,000
1000
£T
<
UJ
C/)
o
u
100
10
/
/,
/.
/
SERIES
/.
/.
PARALLEL
10
100
PLANT CAPACITY (MGD)
FIGURE 25- COST COMPARISON OF PARALLEL
VERSUS SERIES FLOW FOR c-1,2 DCE
-------
cc
<
LU
03
O
O
10,000
1000
100
10
PARALLEL
SERIES
10
100
PLANT CAPACITY (MGD)
FIGURE 26» COST COMPARISON OF PARALLEL
VERSUS SERIES FLOW FOR TOC
-------
INFLUENT
I
20,000 LB EA.
GRANULAR
ACTIVATED CARBON
90p6i
T
T T T
I * ^- - - - 5 o « f &f:
i^ ,-.1*4 •> \'i
.=•:;•'
?,
COLLECTOR
SYSTEM
TO SYSTEM —
• TREATED WATER
FIGURE 27- SCHEMATIC OF PRESSURE GAC CONTACTORS
-------
SURFACE
WASH
INFLUENT
EFFLUENT-
GRAVEL
FILTER BLOCKS
'DRAIN
FIGURE 28- SCHEMATIC OF DOWNFLOW
GRAVITY GAC COMTACTOR
-------
b. Off-Site Regeneration - economical where carbon exhaustion rate
falls between 500 and 1,500 pounds per day-
c. Off-Site Disposal - economical where carbon exhaustion rate is
less than 500 pounds per day.
d. Cost comparison of different carbon recharge options are shown
on Figure 29.
4. Operational Issues
a. Desorption
Due to decrease in influent concentration
- Due to chromatographic effects (displacement of an organic
compound by another that is more favorably adsorbed)
b. Bacterial growth
Could enhance removals by biodegradation
Causes clogging of the carbon bed
- Causes higher plant count in treated water
c. Backwash
- Minimized by providing adequate pretreatment. Backwashing
should be gentle in order not to fluidize the carbon bed.
5. Waste Disposal
a. Backwash
- Carry-over of carbon fines during initial backwashing
b. Spent carbon
Regeneration on-site or off-site
- Disposal in a hazardous waste landfill
F. Applicability to Various System Sizes
1. Organics Removal:
a. <1 mgd, package pressure contactors
b. 1 to 10 mgd, pressure contactors
c. >10 mgd, gravity contactors
2. Radon Removal:
a. Small systems, home GAC units - see Figure 30
b. Large systems, not economically feasible
-15-
-------
120
100
80
Z
o
03
< 60
O
CQ
z
UJ
O
40
20
100
CARBON REPLACEMENT
FLUIDIZED
BED REACTOR
MUTIPLE
HEARTH
1000
10,000
CARBON REQUIREMENT (LBS/DAY)
100,000
FIGURE 29- COST COMPARISON OF
CARBON RECHARGE OPTIONS
-------
(tn
i
. • •
• •
. * •
. • •
• • *
• • "
. •
> * *
• * *
. • •
* * '
— . •
, • •
• • "
• • '
. • •
. • •
1
T
/
S
•"'•'•'*•
".*••' * "."•••
^\
2
f- -r
i
— r-"
7ntt£^h-
- -^--,
3
\
LEGEND: 1 - MAIN WATER LINE
2 • HYDROPNEUMATIC TANK
3 - CAC UNIT
4 - TREATED WATER
-3»»- A 5 • 3ACKWASH TO DRAIN
5
.^:-::i:-:::-::-:-^:-:::::l:i--:::l:l--:::l-i
FIGURE 30- TYPICAL HOME GAC UNIT
-------
G. Treatment Economics
Capital cost components include:
Basic Site Specific
Contactors Special sitework
Activated carbon Raw water holding tank (for ground water
systems)
Regeneration New/restaged well pump (for ground water
facility systems)
Carbon storage GAC contactor building
Carbon transport Chemical facility
facilities
Clearwell
Finished water pump(s)
Backwash storage
Operation and maintenance (O&M) cost components include:
carbon make-up
labor
- fuel
steam
power
maintenance
laboratory analyses
Cost curves for different contaminant groups are shown on Figure 31
(capital cost) and Figure 32 (O&M cost). In general:
Pesticides - least costly
Chlorinated aromatics - I
VOCs - y
Radon - most costly
-16-
-------
1,000,000
100,000
^ 10,000
S£
^^
frt
t-
co
o t.ooo
_J
<
H-
o.
<
o
100
1 ft
0.
01
J
4-k™™_
'
II
[ '
|H X
K>y
, ^* ^
iih
Hi
nil
i
100
/*
r
s
I
I
r i j
^
/•
1 [ 1 '
' : -
11 • /
\]//
\ •
lit
III
i - :
i i •'
ii
h ii
i
M
III
1000
Pt
#
I
1
1
"•
1 . . 1
! i
1!
• ii
! •• i 1 1
^_
* f f
1 1
!-lT[
; i
i
PLANT DESIGN CAPACITY (MGD)
._.».» ALACHLOR (PESTICIDE)
— TCE (VOC)
- — RAOON
FIGURE 31 - CAPITAL COST CURVES
FOR GAG FACILITIES
-------
10,000
•x.
V>
X 1,000
CO
H
W
O
u
5 100
08
O
10
1
0.
-
01
i 1
I :
j
""T~
'
1
I
1 1 "
1,
|
TT"""1
T •
I t 1
I:
i
i '
1 1
1 1
ii
i
1
1 t
!
!
I
0.
/
S
^
1
v
^i
^^
,<
i » ' t
1
'J
Xj
1
.^
^
^
11
-11
II
III
III
/
mj/
'if r
i i
1 1 !
• jr
i! j^
$'
II-
1.0
/
/
jt
+*
f
s
P^3
-/•
r"
<
i*~
/
:
ll
i .S
/
;-•
ii
10 .
s
r
P^TT
jg~
ki
i , -
i
^
IX
i^r • ^
1 ' '^^
rfi
1
I t
100
>^
S\
>|'
^ . j ,
'
!
Ii :
^ •
!••
in
•'HI
t ; ' .
ili'
i
1 I I rl
Illi
1 1 1 M
} 1
III!
1000
I
1
1
"T
i ,
1 1
1
1
I '
i j
i •
-1?
I!
: I
1 \
1
II
i
I
!
1 1
1
i
T
T
PLANT OPERATING CAPACITY (MGD)
ALACHLOR CPESTCOO
TCE (VOC)
—' RADON
FIGURE 32 - O & M COST VURVES FOR
GAC FACILITIES
-------
IV. PACKED COLUMN AERATION
A. Process Description
1. Based on the principle of mass transfer from liquid phase to gas
phase. Rate of mass transfer proceeds according to following
equation:
M = KT a A P
j_j
Where: M = mass of substance transferred per unit time and
volume (Ib/hr/cf)
K = coefficient of mass transfer (Ib/hr/sf)
LJ
a = effective area (sf/cf)
A P = concentration difference or driving force
2. Driving force is the differencte between actual conditions in the air
stripping unit and conditions associated with equilibrium between
the gas and liquid phases.
3. Henry's Law
a. Equilibrium concentration follows Henry's Law, which states
that the amount of gas dissolves in a given quantity of liquid,
at constant temperature and total pressure, is directly propor-
tional to the partial pressure of the gas above the solution.
Henry's constant calculated as follows:
H (dimensionless units) = (16.04) (P) (M)
(T) (S)
P = vapor pressure in mm
M = gram molecular weight of solute
T = temperature in degrees Kelvin
S = solubility in mg/L
b. A compound's Henry's Law constant indicates relative volatility
of the compound; high Henry's Law constant - easily removed by
air stripping.
c. Henry's Constants for several chemicals (at 20 C):
Henry's Constant
(a tin)
VOCs
Vinyl chloride: 2,985
- Trichloroethane (TCA): 286
-17-
-------
- Tetrachloroethylene (PCE): 274
- Trichloroethylene (TCE): 155
Pesticides
- Chlordane: 5.35
Dibromochloropropane (DBCP): 13.8
- Aldicarb: 2.32 x 10
Chlorinated Aromatics
- Dichlorobenzene: 108
- PCB (Arochlor 1242): 78.6
Radon 2,260
4. A schematic of the packed column aeration system is shown on Fig-
ure 33.
B. Process Design Considerations
1. Key Parameters:
a. contaminant
b. levels
c. type of packing material
d. A:W ratio (cubic feet per cubic feet)
e. Liquid loading rate (gpm/sf)
f. Packing height (ft)
g. water temperature
2. Contaminant and Levels
a. Higher the volatility of a compound, the more easily it is
stripped by aeration.
b. Effect of different organics on packed column design is shown
on Figure 34.
3. Type of Packing Material
a. Packing materials are designed to provide:
maximum air-water contact area
low air pressure drop
-18-
-------
INFLUENT
EFFLUENT
.r/^r/V-"^/^
"j^-^'-'^Vx^
u
LIQUID DISTRIBUTOR
PACKING MATERIAL
PACKING SUPPORT
AIR IN
FIGURE 33- DIAGRAM OF PACKED COLUMN AERATOR
-------
100
80
H-
LL
I
Q. 60
LJ
Q
0
Z
u
<
Q.
40
20
CHLOROFORM
TCE
1,2 DICHLOROETHANE
95% REMOVAL
55°F
20/1
40/1
60/1
80/1
100/1 120/1
A/W RATIO
FIGURE 34- EFFECT OF COMPOUND ON PACKED COLUMN DESIGN
-------
b. Generally made of plastic or ceramics, and vary in sizes
(1/2 inch to 4 inches)
c. Commonly used packing materials include:
Super intalox
Tellerettes
Tri-packs
Pall rings
Berl saddles
Raschig rings
4. Air to Water Ratio
a. Function of:
- Contaminant type
- Water temperature
b. Determines the blower size and operating cost of the system
c. Typical range of air/water ratio are:
30:1 - Highly volatile compounds (e.g., TCE)
200:1 - Highly soluble compounds (e.g., DBCP)
5. Liquid Loading Rate
a. Determines the column diameter
b. Typically 25-30 gpm/ft
6. Packing Height
a. Function of:
- Air/Water ratio
Required removal
b. Determines the capital cost of the system
7. Water Temperature
a. Solubility of organic compounds in water generally decreases
with increase in temperature
b. Impact of water temperature on column design is shown on
Figure 35.
C. Testing to Evaluate Process Design Parameters
1. A schematic of the pilot column used for testing packed column
aeration is shewn on Figure 36.
-19-
-------
o
o
O
<
Q_
20
2C
I C
C
c
5
55
BO
LIQUID
AIR : w;
CONTAM
LOADING RATE-3Qgpm/sf
;TER RATIO = so: i
INANT = TCE
^^'^^*'
^^
^'
^^
^'
jf
•*"
--— ^^
^^
_- "
~. — "^
'
^^
0 75 90 95 97.5 99 99.5 99.8
Defl F REMOVAL EFFICIENCY (%)
Deg F
FIGURE 35- EFFECT OF TEMPERATURE
ON PACKED COLUMN DESIGN
-------
-ATEH FLOW
".ETERISG VALVE
HATER METER
rVo_v^V
*" i^-^t—a /r
FROM RAH WATER ^^
SOURCE
RAW WATER
SAMPLE TAP
EFFLUENT ilR
i
EFFLUENT WATER
SAMPLE TAP
EFFLUENT WATER
TO DRAIN
a
\
/
-HATER DISTRIBUTION PLATE
-PACKING MEDIA
INTERMEDIATE SAMPLE TAP (TYP.)
•SUPPORT PLATE (TYP.)
IONNECTING FLANGE (TYP.)
•TOWER SUPPORT TABLE
AIR FLOW
METERING VALVE
PRESSURE
GAGE
120 VOLTS
SINGLE FH«*:
15 AMPS
BLOntfi-
•INFLUEUT AIR
FIGURE 36- SCHEMATIC OF PILOT AERATION COLUMN
-------
2. Test variables are:
a. Hydraulic loading rate
b. Air flow rate
c. Air:Water (A:W) ratio
d. Type of packing
3. Testing Procedure:
a. Water and air flowrates are varied to obtain several A:W
ratios. For each run, the column is operated for about 30 min-
utes to achieve steady state. Eight to twelve runs are recom-
mended for each test.
b. For each run, influent and effluent samples are collected,
along with air and water flowrates, and air and water tempera-
tures.
c. The procedure is repeated for different packing materials.
d. Precautions: In order to ensure reliable data from the pilot
study, the following factors have to be considered:
Packing material should be dumped carefully to avoid
channeling and void spaces in the column.
The pilot column should be on level ground to minimize
channeling or wall effects.
Sampling points should be located so that the samples
truly represent the process performance.
Samples should be collected in duplicates.
- Accuracy of the laboratory analysis method should first be
verified.
D. Development of Design Criteria
1. Design of packed column aeration system involves the following
relationships:
a. Z = (HTU) (NTU), where:
Z = packing height (ft)
HTU = Height of transfer unit (ft)
NTU = Number of transfer units
-20-
-------
b. HTU = L , where:
K aC
L O
L = liquid flow, Ib mole/hr/sq ft
C = molar density of water, Ib mole/cu ft
o
NTU = R
R-l
In
(x./x ) (R-l) + 1
a. o
R
Where: x./x = Ratio of influent to effluent liquid phase
concentration
R = Dimensionless stripping factor
d. (A) (H) MW DensityAir
(W) MW . Density
Air Water
= A (H) (7.512 x Id"4)
Where: A = Air Flowrate (cfm)
W = Water Flowrate (cfm)
H = Henry's Constant (atm)
2. Develop mass transfer relationships as shown on Figure 37.
3. Select the packing material with largest K a, and optimum liquid
loading rate (gpm/sf).
4. Estimate optimum A/W ratio based on A/W ratio - packing height
relationship, as shown on Figure 38.
5. Nomographs for preliminary design of packed columns have been
developed by Speece, Niamalakhandan and Lee (Nomograph for Air
Stripping of VOC from Water, Journal of Environmental Engineering,
Vol. 113, No. 2, April 1987).
E. Facility Design Considerations
1. iMajor Process Elements
a. Column and column intervals
b. Packing medium
c. Blower
d. Booster pump
e. Instrumentation and control
2. A schematic diagram of a packed column facility is shown on Fig-
ure 39.
-------
200
150
V)
V.
13
o
-C
100
90
80
70
60
50
40
30
10
40
TETRACHLOROETHUENE
THEORETICAL
2" TELLERETTES
2" TRI-PACKS
15 20 30 40 50 60 70 80 90 100
LIQUID LOADING RATE (gpm/sf)
10 15 20 30 40 50 60 70 80 90 100
LIQUID LOADING RATE (gpm/sf)
FIGURE 37- MASS TRANSFER RELATIONSHIP
FOR PACKED COLUMN
-------
10
20 30 40 50
AIR TO WATER RATIO (cf : cf)
60 70
LIQUID LOADING RATE = 30 GPM/SF
FIGURE 38- PACKING HEIGHT VS AIR TO WATER RATIO
-------
TREATED WATER
TO RESERVOIR
EXHAU3T
AIR A
INFLUENT
WATER
WELL NO.6
FIGURE 39- SCHEMATIC OF A TYPICAL
PACKED COLUMN FACILITY
-------
3. Location/Site Constraints
a. Zoning requirements
b. Height restrictions
c. Location of air intake louvers
4. System Hydraulics
a. Restaging well pumps
b. Flow and system pressure
c. Repumping to distribution system
5. Housing
a. Freezing potential
b. Noise
c. Security
d. Equipment maintenance
6. Column and Column Internals
a. Column Construction
FRP (fiberglass-reinforced plastic)
Aluminum
Stainless steel
- Concrete
b. Mist eliminator
c. Liquid distributor
orifice plate (shown on Figure 40)
- trough-type distributor (shown on Figure 40)
orifice headers
spray nozzles
d. Support grid
e. Packing Media
7. Air Quality
a. Intake air - air-bourne contaminants
b. Exist air - discharge regulations
8. VOC Emissions
a. Emission rate in the exit air is calculated based upon:
-22-
-------
Orifice - type distributor
Trough - type distributor
FIGURE 40~ DISTRIBUTOR TYPES
-------
Emission rate (Ibs/hr) = (C. - C ) x V x 5
io7
Where: C. = Influent concentration in raw water (ug/L)
i
C = Effluent concentration in treated water (ug/L)
e
V = Water flow rate (gpm)
b. Ambient concentrations
c. Modeling
d. Column modifications
- Height
Air flowrate
Exist velocity
e. Treatment Options:
Thermal incineration of the organics - Disadvantage of
high energy requirements
Catalytic incineration of the organics (shown on
Figure 41) - Lower temperature requirements than thermal
incineration, but currently not effective for removing
chlorinated organics at low levels.
Ozone destruction - catalyzed by UV radiation. At
present, limited application for vapor-phase treatment.
Vapor phase carbon adsorption (shown on Figure 42)
Appears to be the most feasible method at present. Carbon
replacement frequency can be estimated by:
Mass balance based on packed column exit air
concentration and flow rate.
Monitoring using gas chromatography or mass
spectrometry. Samples can be collected in bags and
steel cannisters or using carbon and synthetic
resins.
- Using a combination of mass balance (initially) and
monitoring (towards the anticipated carbon exhaustion
period).
9. Clogging of Packing
a. Iron
-23-
-------
Recirc. damper
or secondary
heat exchanger
Filter/
mixer
~\ Secondary
n n "f—air
\J (/ addition
FIGURE 41 - SCHEMATIC OF CATALYTIC
INCINERATION PROCESS
-------
TREATED
CLEAtt
H£ATfN<3 BLOWER
SLOWEf?
TREATED
FIGURE 4 2- SCH6WAT1C OF VAPOft
PHASE <5AG SYSTEM
-------
b. Solids
c. Biological growth
d. Pretreatment requirements may have to be considered for any one
of these problems
10. Corrosivity of Treated Water
a. Problem: increased DO, reduced CO
b. Solution: increase pH; provide post treatment
F. Applicability to Various System Sizes
1. Organics removal
a. All system sizes
b. Typically used for ground water systems
2. Radon removal
a. Small systems: home diffused aeration units preferred
b. Large systems: use may be limited by radon emission in exit
air
G. Treatment Economics
1. Packed column cost components.
Basic Site Specific
Column Structure Special Sitework
Internals Raw water holding tank
Packing New/restaged well pump
Blower(s) Blower building
Clearwell Booster pump building
Booster pump(s) Chemical facility
Piping Noise control installation
Air emissions control
2. Cost curves for different contaminant groups are shown on Figure 43
(capital cost) and Figure 44 (O&M cost). In general:
Vinyl Chloride - least costly to remove
PCE
TCE
Carbon Tetrachloride
1,2-Dichloroethane
DBCP - most costly to remove
-24-
-------
1,000,000
100,000
-> 10,000
CO
h-
to
o 1,000
D.
u
100
10
0.
[ ' 1 r~.~7";~."r~ r — ; — : — r~r
1 ' r f ' •* ' 1 1 — f-f-
; 1
'.i
M
1
! "
1
! 1 :
I'll
T7T — " ! ; — . 7 i -, , . : ; ; ;;•;"•",
II
i i
i
|:
I
1
i
i
.; :; .. ;. / , .... — : : . . :. ..j
, ,
1 "l 1 1 i : 1
• / ; • •
X,
i
i t
1 1 . 1 1
'!
-T- 1 i !/i* Ss : —
i Six.''
.
Xi
1
i i S
i
11 ' i
iii
i!
i!
i
i 1 ( ; . I ; i i • 1 — i | | |
• 1
• ! , ^^
i i ^ ^
\ Jxi
,'
— , — , , , ., ; — e, — , — ^_|_
i
I
1 i
1 :
~J
^'
• : . 'jT'. JT .
I/ - -^^ ^
/
II X^X
X^^
Ssj^. ,
' 1 1 ^ i ^ '
1 {
i
f , i
X X" : • i
• X.; ,.
I i
ij
:'"
il
i
il
•'
IP
i
!
j ; • [ :
1 ;
!
1 I
1 jl
01 0.1 1.0 10 100
i
H
1!
• :" i •
~T
i
i
M
1000
PLANT DESIGN CAPACITY (MGD)
__ VINYL CHLORIDE. RADON
' TCE
DBCP (PESTICIDE)
FIGURE 43 - CAPITAL COST CURVES FOR
PACKED COLUMN AERATION FACILITIES
-------
10,000
cc
<
ai
* 1,000
I-
W
O
o
08
O
100
10
0.01
V
0.1
1.0
10
100
1000
PLANT OPERATING CAPACITY (MGD)
I
IT
—• VINYL CHLORIDE. RAOON
— TCE
— DBCP (PESTICCE)
FIGURE 44 - O & M COST CURVES FOR
PACKED COLUMN AERATION FACILITIES
-------
3. Vapor Phase GAC Cost
a. May dictate system economics and choice of the process.
b. Typical costs for different system sizes at air to water ratio
of 40:1 are:
System Size (MGD) Capital Cost (K$)
1 40
10 200
100 1,400
1,000 9,000
c. Cost includes carbon contactors, initial carbon charge, gas
heater and installation.
-25-
-------
V. BIBLIOGRAPHY
PAC Adsorption Processes
Aly, 0. M. and S. D. Faust, "Removal of 2,4 Dichlorophenoxyacetic Acid Deriva-
tives from Natural Waters." JAWWA, Vol. 57:221-230, 1965.
Baker, D., "Herbicide Contamination in Municipal Water Supplies in North-
western Ohio." Final Draft Report 1983. Prepared for Great Lakes National
Program Office, U.S. Environmental Protection Agency, 1983.
Cohen, J. M. , L. J. Kamphake, A. E. Lemke, C. Henderson and R. L. Woodward,
"Effects of Fish Poisons on Water Supplies", Part 1, Removal of Toxic Mate-
rials JAWWA 52:12, 1551-1566, 1960.
Croll, B. T., G. M. Arkell, and R. P. J. Hodge, "Residues of Acrylamide in
Water", Water Research. Vol. 8:989-993, 1974.
ESE, Alternative Powdered Activated Carbon Study, Prepared for the City of
North Miami Beach, Gainesville, Florida, 1979.
Lettinga, G., W. A. Beverloo, W. C. van Lief, "The Use of Flocculated Powdered
Activated Carbon in Water Treatment". Prog. Wat. Tech., Vol. 10, Nos. 1-1/2,
pp. 537-554, 1978.
Miltner, ?.. J. and C. A. Fronk, "Treatment of Synthetic Organic Contaminants
for Phase II Regulations", Internal Report DWRD, EPA, December 1985.
Robeck, G. G., K. A. Dostal, J. M. Cohen and J. F. Kreissl, "Effectiveness of
Water Treatment Processes in Pesticide Removal". JAWWA 57:2, 181-199, 1965.
Singley, J. E., B. A. Beaudet, and A. L. Ervin, "Use of Powdered Activated
Carbon for Removal of Specific Organic Compounds," Proceedings of Seminar,
Controlling Organics in Drinking Water, AWWA Annual Conference, 1979, San
Francisco, California.
United States Environmental Protection Agency (EPA), 1982a, Drinking Water
Cost Equations, Municipal Environmental Research Laboratory, Cincinnati, Ohio,
PB83-181826.
Diffused Aeration
Love, 0. T. and R. G. Eilers, Treatment of Drinking Water Containing
Trichloroethylene and Related Industrial Solvents, JAWWA, Vol. 74, No. 8,
pp. 413-425, 1982.
Miltner, R. J. and C. A. Fronk, "Treatment of Synthetic Organic Contaminants
for Phase II Regulations", Internal Report DWRD, EPA, December 1985.
-26-
-------
Ruggiero, D. D., "Removal of Organic Contaminants from the Drinking Water
Supply at Glen Cove, New York, USEPA Document, EPA-600/2-84-029, January 1984.
Ruggiero, D. D. et al., Removal of Organic Contaminants from Drinking Water
Supply at Glen Cove ~New York. Presented at the 100th Annual American Water
Works Association National Conference, June 18, 1980, Atlanta, Georgia.
Multiple Tray Aeration
Hess, A. F., J. E. Dyksen and G. C. Cline, Case Studies Involving Removal of
Organic Chemical Compounds from Ground Water, Presented at the Preconference
Seminar Concerning Organic Chemical Contaminants in Ground Water at the 1981
Annual Conference of the American Water Works Association, June 7-11, 1981,
St. Louis, Missouri.
Hess, A. F. J. E. Dyksen, and H. J. Dunn, Control Strategy — Aeration
Treatment Technique, in AWWA Research Foundation, Occurrence and Removal of
Volatile Organic Chemicals from Drinking Water, Denver, Colorado, pp. 125-127,
1983.
Joyce, M., Smyrna, Delaware Solves a Water Problem, Water and Sewage Works,
March 1980.
Ozone/UV
Buescher, C. A., J. H. Dougherty, and R. T. Skrinde, "Chemical Oxidation of
Selected Organic Pesticides". JWPCF, Vol. 36, No. 8, 1964.
Gilbert, E, "Chemical Changes and Reaction Products in the Ozonization of
Organic Water Constituents" In; Oxidation Techniques in Drinking Water
Treatment, Office of Drinking Water, Washington, D.C., Report No. EPA-
570/9-79-020, 1979.
Hoigne, J. and H. Bader, "Rate Constants of Reactions of Ozone with Organic
and Inorganic Compounds in Water - I". Water Research, Vol. 17, 173-183, 1983.
Legube, B., S. Guyon, H. Sugimitsu, and M. Dore, "Ozonation of Some Aromatic
Compounds in Aqueous Solution: Styrene, Benzaldehyde, Naphthalene, Diethyl-
phthalate, Ethyl and Chloro Benzenes", Ozone: Science and Engineering,
Vol. 5, pp. 151-170, 1983.
Miltner, R. J. and C. A. Fronk, "Treatment of Synthetic Organic Contaminants
for Phase II Regulations", Internal Report DWRD, EPA, December 1985.
Yocum, Floyd H., "Oxidation of Styrene with Ozone in Aqueous Solution", In:
Ozone/Chlorine Dioxide Products of Organic Material, International Ozone
Institute, Cleveland, Ohio, 1978.
-27-
-------
Reverse Osmosis Processes
Berkau, E. E., C. E. Frank, and I A. Jefcoat, "A Scientific Approach to the
Identification and Control of Toxic Chemicals in Industrial Wastewaters",
AlChe Symposium Series No. 197, Volume 76, 1980.
Cabasso, I., E. Klein, C. Eyer, and J. Smith, "Trace Organic Contaminants in
Drinking Water; Evaluation of Semi-permeable Membranes and Osmotic Pumping to
Achieve Concentration", Presentation before the Division of Environmental
Chemistry (American Chemical Society), 204-210, 1974.
Edwards, V. and P. F. Schubert, "Removal of 2,4-D and Other Persistent Organic
Molecules from Water Supplies by Reverse Osmosis", JAWWA, October 1974.
Hinden, E. et al., "Organic Compounds Removed by Reverse Osmosis" Water and
Sewage Works, 1968.
Malaiyandi, M., P. Blais, V. S. Sastri, "Note: Separation of Lindane from Its
Aqueous Solutions by Reverse Osmosis System". Separation Science and Tech-
nology 15:7, 1483-1488, 1980.
Miltner, R. J. and C. A. Fronk, "Treatment of Synthetic Organic Contaminants
for Phase II Regulations", Internal Report DWRD, EPA, December 1985.
Perry, D. L., J. K. Smith, and S. C. Lynch, 1981, Development of Basic Data
and Knowledge Regarding Organic Removal Capabilities of Commercially Available
Home Water Treatment Units Utilizing Activated Carbon, Phase 3/Final Report,
Performed for Criteria and Standards Division, Office of Drinking Water, U.S.
Environmental Protection Agency.
Sorg, T. J. and O. T. Love, Reverse Osmosis Treatment to Control Inorganic and
Volatile Organic Contamination, Proceedings: Experiences with Ground Water
Contamination, Annual AWWA Conference and Exposition, Dallas, Texas, 1984.
Mechanical Aeration
Roberts, P.V. and Levy, J.A., "Air Stripping of Trihalomethanes", Presented at
AWWA Seminar, Controlling Trihalomethanes, Las Vegas, Nevada, May 5, 1983.
Catenary Grid
CHEM-PRO CORPORATION, 27 Daniel Road, P.O. Box 1248, Fairfield, NJ 07007.
Malcolm Pirnie, Inc., Control of Organic Chemicals in Ground Water Supply.
Report submitted to Jamaica Water Supply Company, Lake Success, NY, October
1983.
HI GEE
GLITSCH, INC. P.O. Box 660053, Dallas, TX 75266-0053.
-28-
-------
GAC Adsorption
Dobbs, R.A. and J.M. Cohen, "Carbon Adsorption Isotherms for Toxic Organics,"
EPA Report 600/880-023, Office of Research and Development, MERL, Cincinnati,
Ohio, April 1980.
Dyksen, J.E., Raman, K., Raczko, R.F. and Clark, R.M. "GAC Treatment Costs -
Minimizing Them", presented at the AWWA Annual Conference, Kansas City,
June 14-18, 1987.
Fiessinger, F. and Y. Richard, "International Experience with Granular
Activated Carbon." Paper presented at the AWWA meeting, Atlanta, Georgia,
June 15-20, 1980.
Hess, A.F., "GAC Treatment Designs and Costs for Controlling Volatile Organic
Compounds in Ground Water". Paper presented at the National American Chemical
Society Meeting, Atlanta, Georgia, March 30 - April 3, 1981.
Hiltebrand, D.J.; Dyksen, J.E. and Raman, K. "Treatment Alternatives and
Associated Costs for the Removal of Radon from Ground Water" In: Radon in
Ground Water. Ed: Barbara Graves, Lewis Publishers, 521, 1987.
Love, O.T. and R.G. Filers, "Treatment for the Control of Trichloroethylene
and Related Industrial Solvents in Drinking Water," Drinking Water Research
Division, U.S. EPA, Cincinnati, Ohio, February 1981, JAWWA, 74:413, 1982.
Lowry, J.D. and Brandown, J.E. "Removal of Radon from Ground Water Supplies
using Granular Activated Carbon or Diffused Aeration", University of Maine,
Department of Civil Engineering, Orono, Maine 06469.
Malcolm Pirnie, Inc., "Preliminary Treatment Designs and Costs for Control of
Organic Compounds." Report prepared for the Office of Drinking Water, U.S.
EPA, Washington, D.C., April 1981.
McCarty, P.L., D. Argo and M. Reinhard, "Operational Experiences with
Activated Carbon Adsorbers at Water Factory 21," JAWWA, 71:683-689, November
1979.
Meijers, A.P., "Objectives and Procedures for GAC Treatment," JAWWA, 71:628,
1979.
O'Brien, R.P., D.M. Jordan and W.R. Musser, "Trace Crganics Removal from
Contaminated Ground Waters with Granular Activated Carbon," Calgon
Corporation, Pittsburgh, Pennsylvania. Presented at the National Ameri-
can Chemical Society Meeting, Atlanta, Georgia, March 29 - April 3, 1981.
Peel, R.G. and A. Benedek, "Attainment of Equilibrium in Activated Carbon
Isotherm Studies," Environmental Science and Technology, 14:66-79', 1980.
Randtke, S.J. and C.P. Jepsen, "Effects of Salts on Activated Carbon
Adsorption of Fulvic Acids," JAWWA, 74:84-93, February 1982.
-29-
-------
Roberts, P.V. and R.S. Summers, "Performance of Granular Activated Carbon for
Total Organic Carbon Removal," JAWWA, 74:113-118, February 1982.
Ruggiero, D.D. and R. Ausubel, "Removal of Organic Contaminants from Drinking
Water Supply at Glen Cove, New York, Phase II," Report No. EPA-600/2-82/027,
U.S. EPA, March 1982, and Phase I, Report No. EPA-600/2-80-198, 1980.
Schulhof, P., "An Evolutionary Approach to Activated Carbon Treatment," JAWWA,
71:648, 1979.
Schalekamp, M., "The Use of GAC Filtration to Ensure Quality in Drinking Water
from Surface Sources," JAWWA, 71:638-647, November 1979.
Sontheimer, H. , "Applying Oxidation and Adsorption Techniques: A Summary of
Progress," JAWWA, 71:612-617, November 1979.
Sontheimer, H., "Design Criteria and Process Schemes for GAC Filters," JAWWA,
71:618, 1979.
Symons, J.M., "Removal of Organic Contaminants from Drinking Water Using
Techniques Other than Granular Activated Carbon Alone - A Progress Report,"
Drinking Water Research Division, U.S. EPA, Cincinnati, Ohio, May 1979.
U.S. EPA, "Process Design Manual for Carbon Adsorption," Technology Transfer,
U.S. EPA, October 1973.
Weber, W.J., Jr. and M. Pirbazari, "Adsorption of Toxic and Carcinogenic
Compounds from Water," JAWWA, 74:203, 1982.
Weber, W.J., Jr. and B.M. VanVliet, "Synthetic Adsorbents and Activated
Carbons for Water Treatment: Overview and Experimental Comparisons," JAWWA,
73:420, 1981.
Packed Column Aeration
Camp, T.R., "Gas Transfer To and From Aqueous Solutions," Journal Sanitary
Engineering Division, Proc. ASCE, 84:SA4:1701, 1958.
Cummins, M.D. and J.J. Westrick, "Packed Column Air Stripping for Removal of
Volatile Compounds." Presented at the 1982 Conference on Environmental
Engineering, ASCE, July 14-15, 1982.
Dyksen, J.E. and A.F. Hess, "Aeration Techniques for Removing Trace Organic
Compounds from Drinking Water." Paper presented at the 1981 ASCE National
Conference on Environmental Engineering, Atlanta, Georgia, July, 1981.
Dyksen, J.E., A.F. Hess, M.J. Barnes and G.C. Cline, "The Use of Aeration to
Remove Volatile Organics From Ground Water." Presented at the 1982 Annual
Conference of the American Water Works Association, Miami Beach, Florida, May
1982.
-30-
-------
Haney, Paul D., "Theoretical Principles of Aeration," JAWWA, 46:4:353, 1954.
Hess, A.F., J.E. Dyksen and G.C. Cline, "Case Studies Involving Removal of
Organic Chemical Compounds from Ground Water." Presented at the 1981 Annual
Conference of the American Water Works Association, St. Louis, Missouri, June
1981.
Hiltebrand, D.J.; Dyksen, J.E. and Raman, K. "Treatment Alternatives and
Associated Costs for the Removal of Radon from Ground Water" in Radon in
Ground Water. Ed: Barbara Graves, Lewis Publishers, 521, 1987.
Kavanaugh, M.C. and R.R. Trussel, "Design of Aeration Towers to Strip Volatile
Contaminants from Drinking Water," JAWWA, 72:12:684, 1980.
Kavanaugh, M.C. and R.R. Trussel, "Air Stripping as a Treatment Process."
Paper presented at the AWWA conference, July 1981.
Langelier, W.F., "Theory and Practice of Aeration," JAWWA, 24:1:62, 1932.
Love, O.T. and R.G. Eilers, "Treatment for the Control of Trichloroethylene
and Related Industrial Solvents in Drinking Water," U.S. EPA, Drinking Water
Research Division, Cincinnati, Ohio, October 1980.
Mackay, D. et al., "Determination of Air-Water Henry's Law Constants for
Hydrophobic Pollutants," Environmental Science and Technology, 13(3), 33-337,
1979.
McCarty, P.L., K.H. Sutherland, J. Graydon and M. Reinhard, "Volatile Organic
Contaminants Removal by Stripping," Proc. AWWA Seminar, Controlling Organics
in Drinking Water, San Francisco, June 1979.
McKinnon, R.J. and J.E. Dyksen, "Aeration Plus Carbon Adsorption Remove
Organics from Rockaway Township (NJ) Ground Water Supply." Presented at the
1982 Annual Convention of the American Society of Civil Engineers, New
Orleans, Louisiana, October 25-27, 1982.
Metcalf and Eddy, Inc., "Volatile Organics Removal: Two Ground Water Supply
Case Histories." Presented at the New York Section AWWA, 1980.
Munz, C. and P.V. Roberts, "Transfer of Volatile Organic Pollutants into a Gas
Phase During Bubble Aeration," Technical Report No. 262, Department of Civil
Engineering, Stanford University, 1982.
Nebolsine, Kohlman and Ruggiero Engineers, "Removal of Organic Contaminants
from Drinking Water Supply at Glen Cove, New York," Interim Report on U.S.
EPA Agreement No. CR806355-01, Office of Research and Development, MERL,
Drinking Water Research Division, Cincinnati, Ohio, July 1980.
Raczko, R.F., J.E. Dyksen and M.B. Denove, "Pilot Scale Studies of Air
Stripping for Removal of Volatile Organics from Ground Water." Presented at
the 14th Mid-Atlantic Industrial Waste Conference, University of Maryland,
College Park, Maryland, 1982.
-31-
-------
Roberts, P.V. and P. Dandliker, "Mass Transfer of Volatile Organic
Contaminants During Surface Aeration." Presented at the 1982 Annual
Conference of the American Water Works Association, Miami Beach, Florida,
May 1982.
Singley, J.E., A.L. Ervin, M.A. Mangone, J.M. Allan and H.H. Land, "Trace
Organics Removal by Air Stripping," AWWA Research Foundation, 1980.
Symons, J.M., A.A. Stevens, R.M. Clark, E.E. Geldreich, O.T. Love and
J. DeMarco, "Treatment Techniques for Controlling Trihalomethanes in Drinking
Water," U.S. EPA, Drinking Water Research Division, 1981.
Treybal, R.E., "Mass Transfer Operations," McGraw-Hill Book Co., New York (3rd
edition), 1980.
Warner, H.P., J.M. Cohen and J.C. Ireland, "Determination of Henry's Law
Constants of Selected Priority Pollutants," Wastewater Research Division,
Municipal Environmental Research Laboratory, Cincinnati, Ohio, 1980.
-32-
-------
WORKSHOP ON EMERGING TECHNOLOGIES
FOR DRINKING WATER TREATMENT
TECHNICAL SESSION ON ORGANICS REMOVAL
CASE STUDY
EXAMPLE 1:
A municipal ground water supply has been contaminated with several
organic chemicals. Based on the raw water quality given below, determine;
An appropriate treatment method
Size and process design of the treatment facility
ORGANIC CONTAMINANT LEVELS
IN LITTLETOWN WELLS
Trichloroethylene
Tetrachloroethylene
1,1,1-Trichloroethane
Notes:
1. Total Combined Flowrate = 225 gpm
Consider:
Treatment objective
Treatment options
Treatability testing required
- Cost
1 units in micrograms/liter)
Well No.
10
200
1,000
750
11
200
1,000
600
14
100
200
300
-1-
-------
TREATMENT OBJECTIVES
VOC
Trichloroethylene
Tetrachloroethylene
1,1,1-Trichloroethane
Composite of
All Wells
160
650
500
Design
Effluent
Concentration
5
10
10
Required
Percent
Removal
97
98.5
98
Note:
1. All concentrations given in ug/L.
Define:
- What are the treatment options?
-------
CONTAMINANT CHARACTERISTICS
VOC
Trichloroethylene
Tetrachloroethylene
1,1,1-Trichloroethane
* at 20 C
Henry's Constant
(H>*
(atm)
155
274
288
Adsorption
Isotherm Carbon
Constants Usage Rate
K 1/n (lbs/1,000 gal)
28 0.62
50.8 0.56
2.48 0.34
0.19
0.18
2.8
Define:
Which treatment process will work (aeration or GAG)?
Consider:
Volatility of the contaminants
Carbon usage rates
-3-
-------
PACKED COLUMN - PILOT STUDY RESULTS
:3 23 40 50 60 70 80 90 100
LIQUID LOADING RATE (gpci/sf)
30
Define:
15 ID 33 <0 50 60 70 80 90 100
LIQUID LOADING RATE (gpm/sf)
- Optimum loading rate
Type of packing material
- A:W Ratio
Packing height
:a 40 50 BO ;c so 90 100
LIQUID LOADING RATE (gpm/sf)
-4-
-------
Air:Water Ratio Versus Packing Height
40
35
30
0)
O
-------
PACKED COLUMN PROCESS DESIGN CRITERIA
A sample calculation for developing the process design criteria is shown
below:
Pilot Study Results
VOC = Tetrachloroethylene
Packing height = 10 ft
Column diameter = 1 ft
Packing material = 2-inch Tri-packs
Water Air
Run Flow Rate Flow Rate
'Io. (gpm) (gpm) A/W
PCE (ug/L)
Influent Effluent % Removal
16
43
20:1
100
93
1.
Calculation K a (Refer to workshop notes for formulas)
L
a. R = (A) (H) x 7.5212 x 10
(W)
= 20 x 274 x 7.5212 x 10
= 4.12
b. NTU = R
-4
-4
In
R-l
(x./x ) (R-l) + 1
i o
R
4.12
4.12 - 1
In (100/7) (4.12 - 1) + 1
4.12
= 3.2
c. HTU = 2
NTU
10
3.17
= 3.2 ft
d. K a = L
L
HTU
16 gpm x 1 x 60 min x 1 ft
~
3.2 ft
hr 7.48 ga
-1
= 51 hr
K a for other loading rates are calculated similarly.
/ersus loading rate is made.
A plot of K a
-6-
-------
2. Calculation of Packing Height
2
Optimum loading rate = 25 gpm/ft
2 -1
K a at 25 gpm/ft for 2-inch Tri-paks = 58 hr (from graph)
Li
Assume A/W ratio = 40
% removal required (PCE) = 98.5%
a. R = (A) (H) x 7.5212 x .10~4
(W)
= 40 x 274 x 7.5212 x 10~4
= 8.24
b. x./x = 1 = 1 = 66.7
10 1 - % Removal 1 - 98.5
100 100
c. NTU = 8.24 hr (66.7) (8.24 - 1) + 1
8.24 - 1 8.24
= 4.64
d. HTU = L (gpm/ft ) x 60 min x 1_ ft
K a (hr"1) hr 7.48 gal
L
= £5_ gpm/ft x 60 min x 1 ft
58 -1 hr 7.48 gal
hr ^
= 3.5 ft
e. Required packing height = (NTU) (HTU)
= 4.64 x 3.5 ft
= 16.2 ft
Packing heights for other A/W ratios are calculated similarly. A plot of
packing height versus A/W ratio is made. Optimum A/W ratio can be estimated
from such a plot.
-7-
-------
VOC Emission Rate
PCE concentration in raw water (C.) = 650 ug/L
PCE concentration in treated water (C ) =10 ug/L
PCE Emission rate (Ibs/hr) =C.-C xVx 5
i e
io7
= (650 - 10) x 225 gpm x 5
io7
= 0.072 Ib/hr
Since OSHA regulations require exit air treatment only at emission rates
of above 0.1 Ib/nr, no additional treatment is required for the exit air at
this facility.
-8-
-------
•-ote :
PACKED COLUMN PROCESS DESIGN CRITERIA
3ASED ON PILOT STUDY
Maximum Flcwrate =
Hydraulic Loading Rate =
PCE Mass Transfer Coefficient, KTa =
j_i
A:W Ratio =
Air Flow
Packing Height
(1)
225 gpm
25 gpm/ft"
56 hr-1
40:1
1,200 cfia
20 feet
Based on 98.5% removal of PCE
7CC REMOVAL AT DESIGN CONDITIONS
VOC
TCE
PCE*
! ,1,1-Trichloroethane
*Denotes critical compound.
Design Percent
Removal
97
98.5
98
Actual Percent
Removal
)99
98.5
98
Design
Effluent
Concentrations
(ug/1)
5
10
10
Actual
Effluent
Concentration
(ug/1)
(5
10
10
uerine:
Treatment costs
-9-
-------
PRELIMINARY COST ESTIMATES FOR
PACKED COLUMN AERATION PROCESS
Based on cost curves presented in the workshop report:
Capital Cost (Figure 43) = $150,000
O&M Cost (Figure 44) = $ 8,000/year
Total Cost = $0.22/1,000 gallons
(20 year amortization period
and 10% interest rate)
-10-
-------
MINI-COLUMN TEST RESULTS
200
133
90
:0
70
_ 50
50
UJ
j± 40
~ 3
rt
O
CQ
CC
;3
I
TETRACHLOROETHYLENE —7
I I
A-TRICHLOROETHYLENE
,1.1-THJCHLOROETHANE
I
TETRACHLOROETHYLENE
TRICHLOROETHYLENE i
i. I-TRICHLOROETHANE::
6 8 10 20 40 60 30
INFLUENT VOC CONCENTRATION (ug/1)
200
400 GOO
Define:
Carbon Type
-II-
-------
MINI-COLUMN TEST RESULTS
E 30
xt
CXL
o
z
o
TETRACHLOROETHYLENE
TRICHLOROETHYLENE
s
1.1.1 -TRICHLORO
ETHANE
20
10
1000 2000 3000
WATER TREATED (ml)
4000
Define:
- Controlling VCC
Carbon Usage based on:
Carbon usage (lbs/1,000 gal) = W x 8461.5
V
whero 17 = weight of carbon (0.1 gm)
V = volume treated (ml)
- Preliminary Process Design
-12-
-------
GAG PROCESS DESIGN CRITERIA
1. Maximum Flowrate: 225 gpm
2. Carbon Usage Rate:
weight of carbon in the mini-column (W) = 0.1 gm
- volume treated based on tetrachloroethylene breakthrough at 10 ug/L
(V) = 1,500 ul
Carbon Usage (lbs/1,000 1) = W x 8461.5
V
= 0.1 gm x 8461.5 = 0.56
1,500 ml
3. Carbon Charge:
Assume EBCT = 10 minutes
3 3
Carbon volume = 225 gal x 10 min x _1 ft = 300 ft
min 7.48 gal
Carbon weight = 300 ft x 29 Ibs = 8,700 Ibs
ft3
4. Contactor Dimensions:
Assume hydraulic loading rate = 5 gpm/ft
Contactor diameter = 225 gpm x 4_ = 7.5 ft
•J 5 gpm/ft
Carbon bed depth = 300 ft = 7 ft
2 2
TT (7.5) ft
4
Contactor height = 15 feet
5. Replacement Frequency:
Replacement frequency = carbon charge (Ibs) x 0.7
carbon usage ( Ibs ) x flow (gpm)
1,000 gal
= 8,700 Ibs x 0.7 = 48 days
0.56 Ibs x 225 gpm
1,000 gal
-13-
-------
GAG PROCESS DESIGN CRITERIA
Define:
Maximum Flowrate
Hydraulic Loading Rate
Carbon Usage
EBCT
Number of Contactors
Contactor Dimensions
Depth of GAC Bed
Carbon Charge
Type of GAC
Carbon Life
= 225 gpm
2
= 5 gpm/ft
0.56 lbs/1,000 gallons
= 10 minutes
1
7.5 ft (diameter) x 15 ft (height)
7 ft
= 8,700 Ibs
Carborundum GAC 830
48 days
Treatment costs
-14-
-------
PRELIMINARY COST ESTIMATES FOR
GAG ADSORPTION PROCESS
Based on cost curves presented in the workshop report:
Capital Cost (Figure 31) = $160,000
O&M Cost (Figure 32) = $ 25,000/year
Total Cost ' = $.375/1,000 gallons
(20 year amortization period
and 10% interest rate)
-15-
-------
PROCESS SELECTION
DEFINE:
Recommended treatment process
CONSIDER:
Treatment efficiency
Operational factors
Treatment costs
-16-
-------
COMPARISON OF ALTERNATIVES
LITTLETOWN WELLFIELD
Packed Column
High removals
Continuous operation
- VOC emissions in exit air
less than OSHA limit
- No waste disposal
Costs:
Capital Cost = $150,000
O&M = $ 8,000/yr
Total = $0.22/1,000 gallons
GAG Adsorption
High removals
Requires down time for Carbon
replacement
No air emissions-
- Disposal of backwash water and
spent carbon
Costs:
Capital Cost = $160,000
O&M = $25,000/yr
Total = $0.375/1,000 gallons
-17-
-------
Technical Session: Inorganics
J. Edward Singley, Vice President, James W. Montgomery Consulting
Engineers, Inc., Gainesville, FL
VIII-1
-------
Emerging Technologies for Control
of Inorganic Contaminants
J. Edward Singley
I. Introduction
A. Regulations
1. Present
2. Proposed
B. Metals
1. Lead
2. Barium
3. Cadmium
4. Chromium
5. Mercury
6. Radium
7. Selenium
8. Silver
9. Strontium
C. Non-Metals
1. Arsenic
2. Fluoride
3. Nitrate
II. Lead
A. Lead leaching
B. Workmanship
C. Control techniques
D. Data needs
III. Barium
IV. Cadmium
V. Chromium
VI. Mercury
VII. Radium
A. Chemistry
B. Control techniques
1. Precipitative
softening
2. Ion exchange
3. Demineralization
C. Radioactivity
VIII. • Selenium
IX. Silver
X. Strontium
XI. Arsenic
XII. Fluoride
A. Activated alumina
adsorption
B. Ion exchange
XIII. Nitrate
XIV. Summary
-------
Emerging Technologies for Control
of Inorganic Contaminants
GENERAL
Of the 12 inorganic contaminants that are presently regulated by
primary agencies, the one most commonly present at levels in excess of
the primary MCL is lead, Pb. The others are less common but are present
in site or regionally specific cases. For this reason major attention
will be focused on Pb with limited attention to the others.
REGULATIONS
Present.
The present regulations on the inorganic contaminants under the
National Interim Drinking Water Regulations cover 12 different substances:
MCL (mg/L)*
Arsenic 0.05
Barium 1.0
Cadmium 0.010
Chromium (total) 0.05
Fluoride 4.0**
Lead 0.05
Mercury 0.002
Nitrate, as N 10
Selenium 0.01
Silver 0.05
Radium 5 (pCi/L)
Strontium-90 8 (pCi/L)
*as of August 1, 1987
**Secondary MCL of 2.0 mg/L
The significance of the twelve varies considerably since their
presence and concentrations vary so much. For example silver is rarely,
if ever, found unless a point-of-use water treatment device is used by
the consumer, whereas lead is found to be almost ubiquitous in samples
from the consumer's tap. Radium, strontium, cadmium, selenium, arsenic,
mercury, barium, nitrate, chromium and excess fluoride are found only in
-------
site-specific cases although their concentrations may vary widely. For
some of the above the problem of elevated concentrations is indigenous to
a section, or sections, of the country. Thus a treatment technique may
be of limited applicability. Despite this the need is real and must be
addressed because the MCLs must be met.
Proposed.
There is no doubt that additional inorganic contaminants will be
regulated nor that there will be changes in the MCLs of some of those
presently regulated. Some will increase and some decrease. The proposed
regulations are shown in Table 1.
Since silver is not found in the raw water, added in treatment or
by corrosion of the system, it will likely be dropped entirely. On the
other hand, the present lead MCL of 0.050 mg/L will certainly be reduced
to 0.020 mg/L or less, unless a treatment technique is specified. Other
changes in the regulations, as proposed can be observed in the Table.
The presence of lead, zinc, copper and cadmiun in water entering
the distribution system is rare. They occur because of reactions between
the water and the conveying system in transit from the plant to the tap
of the consumer. The particular contaminant(s) picked up depends upon
the composition of the pipes, valves, meters, etc., with which the water
comes in contact. The corrosion reactions that occur between the water
and metallic substances are electrochemical in nature and result in
converting the free metals, which are very insoluble in water, to their
ionic forms which are usually very soluble. The ionic solubility is
strongly influenced by the water quality. It may be stated safely that
"all metals corrode in direct contact with water." The rate is
determined primarily by the properties of the metal itself and
secondarily by the properties of the water, i.e., the water quality. The
most common metals used in transmission mains and plumbing systems are
steel (iron), galvanized steel, copper, lead, brass and bronze. The
presence of zinc and cadmium, and to a lesser extent lead, in water
samples comes from the zinc used in galvanizing pipe. The lead may come
from lead services, lead goosenecks, lead pipes (rare) and from
lead-based solders used to join copper pipes. In most parts of the U.S.
the latter is the major source of lead at the tap.
LEAD
It is of historical interest to note that lead pipes were used in
Rome to distribute the water from the Aquaducts to the homes of the
wealthy, which is cited by some authorities as the cause of the decline
of the Roman Empire.
Lead Leaching.
Lead based solders used to connect copper plumbing for many years
have been based on tin-lead combinations, the most common being 50:50 and
60:40 tin:lead. The recent ban on lead in such solders will strongly
affect lead content in water obtained from the taps of homes constructed
-------
after the states implement the ban but will have minimal effect on water
samples from homes and businesses constructed prior to the ban.
There is an extensive discussion of the rate of leaching of lead
from lead-based solders and fluxes in "Internal Corrosion of Water
Distribution Systems" published by the American Water Works Association
Research Foundation jointly with the Engler-Bunte Institute of the
University of Karlsruhe, West Germany. Examples from the U.S. have been
cited by Norman Murrell in testimony before Congress. The concentration
of lead varies with the amount of water that has been withdrawn both in
the sample itself and before the sample, as well as the time of contact
of the water with the household plumbing. Thus first draw samples give
the worst case examples and the smaller the sample the higher the
concentration measured. One study in Scotland by Lyon and Lenihan in
Glasgow in 1977, cited by Schock and Wagner in the reference above,
showed that as much as 20 ug of lead per fitting per hour could be picked
up for new connections in copper pipes even though they were "well
made." This declined to about 1.25 ug/fitting/hr in 4-5 weeks and
remained at that level for many years. Studies at the Illinois State
Water Survey by Neff and Schock also showed that lead can be leached from
chrome plated brass fixtures.
Workmanship.
There is no doubt that a major contribution to lead corrosion is
poor workmanship in the formation of lead-based soldered joints in copper
pipes. This leads to galvanic corrosion of the lead in contact with the
more cathodic copper and tin. It is likely that the lead concentration
in water in contact with poorly made joints is almost independent of the
water quality, thus there is little that the utility can do to reduce
lead levels significantly when such a mechanism predominates except to
provide a protective coating, such as CaC03-
Control methods.
Early studies in 1887 in Germany by Wolfhugel, identified oxygen,
carbon dioxide, water quality, age and quality of lead pipes, stagnation
time, flow velocity and temperature as contributing factors. These
include essentially all of the factors recognized today as important in
lead pick-up. These observations resulted from a serious leach problem
in Dessau and the effective solution was the addition of limestone
(calcium carbonate). Later soda ash (sodium carbonate) was added. These
treatments increased the pH thus reducing the free carbon dioxide and
increasing the carbonate alkalinity. Today, one hundred years later, we
use the same approach when treating the same water qualities that caused
the problems, i.e., low hardness, low alkalinity, and low pH.
Some of the classic cases in the U.S. were discussed and studied by
Karalekas, et al. starting in 1976 in Massachusetts. They showed that pH
adjustment to a value above 8.0 using caustic soda (NaOH) reduced the
pick-up of lead significantly. The use of zinc orthophosphate at levels
as high as 12 mg/1 showed little if any improvement. These effects
correlated well with the effects in Glasgow ninety years earlier.
Figure 1 shows the solubility of lead in water.
-------
The use of various corrosion inhibitors has been studied
extensively with the following results:
Carbonates and pH. These two are almost inseparable because of the
chemical interdependence of the carbonate system and pH. Whereas_ many
authors have concluded that a major water quality factor in a particular
corrosion situation was free carbon dioxide, none have shown carbon
dioxide, per se, as a participant in the corrosion reactions. Rather it
is the dependence upon the historical precedence of citing carbon dioxide
rather than pH because the term pH was not used or defined prior to 1909
when Sorenson simplified the discussion of the hydrogen ion concentration
by defining a new term "pH". Until shortly before that time it was not
possible to measure the hydrogen ion concentration in aqueous solution
accurately, and even then it was a laborious procedure. On the other
hand it was possible to analyze accurately for the "free carbon dioxide
in the middle 1880's.
The shift in the carbonate system from carbon dioxide to
bicarbonate and then carbonate as the pH is increased conforms to the
well known definition of the various forms of alkalinity. The relevant
reactions and equilibrium constants are:
C02 + H20 - H2C03,pK =1.5
H2C03 - H+ + HCOa'.pK, =6.3
HC03- = H+ + C03=,pK2 = 10.3
The conversion of free carbon dioxide to bicarbonate is complete at a pH
of about 8.3 and to carbonate above a pH of 10.3.
A comprehensive study of the effect of these variables on lead
corrosion was conducted by Schock, et al. at the EPA Laboratories in
Cincinnati. From these studies they developed chemical models of lead
solubility as functions of carbonate, pH, orthophosphate, sulfate and
chloride. They concluded that orthophosphate could be an effective
corrosion control inhibitor under a limited range of carbonate and
hydrogen ion concentrations.
Silicates. Although some authors and studies have suggested the
efficacy of silicates, no studies have supported the hypothesis.
Orthophosphates. The suggestions of the use of these to reduce
lead corrosion has not been borne out in many studies performed in the
U.S. and Europe. These studies showed that the solubility of lead was
greater than predicted by the models that had been developed. See
Figure 2.
Sulfate, Chloride and Nitrate. These have been shown repeatedly to
have little effect on lead solubility.
Organics. Some studies have shown that natural organics may
increase lead solubility.
Chlorine. The solubility of lead from lead-based solders has been
shown to be increased in the presence of chloramines as compared to free
chlorine.
-------
Polyphosphates. Recent studies have shown no advantage over the
simple orthophosphates. The addition of zinc sulfate did not improve the
effect.
Calcium Carbonate Films. One of the traditional methods of control
of corrosion is to interpose a barrier between the metal and the water.
This can be accomplished by reference to one of the indices of calcium
carbonate saturation, such as the Langelier Index or others.
Modifications of water quality that will initiate the precipitation of a
protective coating of calcium carbonate is generally based on pH control,
if adequate concentrations of alkalinity and calcium ion are present.
pH Control. It has been shown that pH values above 9.0 lead to
reduced lead concentrations because of the reduced availability of
electron acceptors. This has been one of the most effective methods of
control of lead corrosion. When pH control is practiced in conjunction
with alkalinity supplementation in low alkalinity waters, lead corrosion
is reduced. Waters of high alkalinity may require softening, to reduce
lead corrosion.
BARIUM.
Barium has chemical properties very similar to the other alkaline
earth elements calcium, magnesium and radium, though its removal from
drinking water parallels that of the other members of the family. Its
sulfate is less soluble than calcium sulfate but coagulation with ferric
sulfate or alum is relatively ineffective. BaS04 has a solubility of
only 2.2 mg/L at 25°C, which exceeds the MCL of 1.0 mg/L. The solubility
product constant is about lxlO~4. Barium carbonate has a solubility of
about 18 mg/L at 25°C but is removed, almost quantitatively, by lime
softening at pH values over the range of 9.5 to 10.5.
CADMIUM
This metal rarely, if ever, occurs in raw water. It is a
contaminant in zinc used for galvanizing, thus occurs when zinc
galvanizing corrodes. Its electrode potential is very close to that of
iron (-0.402v. vs -0.441v.), thus it would corrodes only slowly if used
alone to protect steel. In contrast, zinc's potential is much more
anodic (-0.76v.) and thus provides the desires cathodic protection.
The removal of cadmium from drinking water is rarely needed but the
prevention of its presence can be accomplished by corrosion control.
Figure 3 shows the effect of pH on cadmium solubility.
CHROMIUM.
Chromium exists in water as the +3 or +6 valence state. The
chromium +3 ion is relatively insoluble as the hydroxide, see Figure 4,
but the +6 state is very soluble as the chromate or dichromate ion. The
chromate ion varies with pH , as shown in Figure 5. The dichromate
ion is converted to chromate at pH values above 7. Its high solubility
prevents its removal by any technique but ion exchange (anionic) or
-------
demineralization (reverse osmosis or electrodialysis).
readily removed by coagulation or lime softening.
MERCURY.
Chromium III is
Although mercury occurs only rarely in untreated waters and then
more rarely above the present MCL of 0;002 mg/L, its high toxicity
It can occur as two mercury ions, +1 or +2. The +2
form is the more common but still contributes little to the total mercury
in natural water. The predominant forms are free metallic mercury above
organically complexed mercury. Surprisingly the free
even
justifies concern.
only effective methods for
above pH 5 and
metal has a solubility of 0.025 mg/L. The
removal is adsorption on activated carbon.
RADIUM.
Although the radium concentration in most of the major surface
supplies in the U.S. is low, as shown in Table 2, there are several areas
of the country where radium is a major problem, principally Illinois,
Iowa and Florida. Some examples of raw water quality in these areas are
shown in Table 3. Treatment of a water containing Ra-226 plus Ra-228 at
concentrations above the MCL of 5 pCi/L can be estimated from calculating
the amount that must be removed. This can be expressed as f=l-(5/RWR)
where f is the radium removal fraction required, and RWR is the raw water
radium concentration. This is shown graphically in Figure 6.
Chemistry.
It is to be noted from Table 3 that the waters containing radium
also have high total hardness (TH). The chemical properties of radium
are similar to those of calcium and magnesium; thus it is not unexpected
that those processes which remove hardness (primarily calcium and
magnesium salts) will also remove radium.
"The hardness of almost all water supplies is caused by the
presence in solution of calcium and magnesium ions. Other divalent ions
such as strontium, ferrous iron and manganese may contribute to the
hardness to a much lesser degree. The lime-soda process is a
precipitative softening process which uses the addition of lime
(CaO-quicklime, or Ca(OH)2-slaked or hydrated lime) to convert the
soluble bicarbonates of calcium and magnesium into insoluble calcium
carbonate and magnesium hydroxide. This is the removal of "carbonate
hardness", or the calcium and magnesium ions associated in solution with
the bicarbonate ion. Calcium and magnesium associated with the sulfate,
chloride or other ions, ("non-carbonate hardness"), are removed by the
addition of both lime and soda ash (Na2C03) which provides the carbonate
ion necessary for formation of calcium carbonate. Since magnesium
removal occurs only above a pH of about 11 at normal water temperatures,
excess lime sufficient to raise the pH to 11 must be added prior to
removal of magnesium as magnesium hydroxide.
The precipitated compounds are flocculated, settled, and removed as
sludge while the clarified effluent is usually filtered in order to
polish the effluent by removing residual floe particles.
-------
The chemistry of water softening is probably best explained or
illustrated by showing the chemical reactions that take place when lime
and soda ash are added to water containing calcium and magnesium salts.
The reactions in the lime-soda process, then are:
C02 + Ca(OH)2 = CaC03 + H20 (1)
Ca(HC03)2 + Ca(OH)2 - 2CaC03 + 2H20 (2)
Mg(HC03)2 + Ca(OH)2 = CaC03 + MgC03 + 2H20 (3)
MgC03 + Ca(OH)2 = CaC03 + Mg(OH)2 (4)
2NaHC03 + Ca(OH)2 = CaC03 + Na2C03 + 2H20 (5)
MgS04 + Ca(OH)2 - Mg(OH)2 + CaSO (6)
Na2C03 = CaC03 + Na2S04 (7)
These equations show all of the reactions taking place in softening
a water containing both carbonate and noncarbonate hardness by the
lime-soda process. It should be noted that, in Equation 1, the carbon
dioxide is not hardness as such, but in proportion to its content in the
water will consume lime and must therefore be considered in calculating
the amount of lime required. Similarly, in Equation 5, the sodium
bicarbonate or sodium alkalinity, if present, is not part of the hardness
but, since it is included in the total alkalinity, it will consume lime.
Equations 2 and 4 show the removal of carbonate hardness by lime.
Whereas only one molecule of lime is required for one molecule of calcium
bicarbonate, Equation 2, two molecules of lime are required for the
removal of one molecule of magnesium bicarbonate hardness, Equations 3
and 4. Equation 6 shows the removal of magnesium noncarbonate hardness,
shown as magnesium sulfate, by lime. No softening is effected by this
reaction because, for each molecule of magnesium noncarbonate hardness
removed, an equivalent amount of calcium noncarbonate hardness is
formed. Equation 7 shows the removal of calcium noncarbonate hardness,
shown as calcium sulfate, whether originally in the water or formed as
shown in Equation 6.
From these reactions it is apparent that the amounts of lime and
soda ash required to soften a water may be calculated from the
concentrations of free carbon dioxide, bicarbonate (usually the total
alkalinity), magnesium hardness, and noncarbonate hardness." (Singley,
et al. 1977).
Control Techniques.
Soluble radium, a divalent alkaline earth metal ion similar to
calcium and magnesium is also removed in the lime-soda softening
process. Table 4 shows the radium and total removal hardness
efficiencies of seven lime-soda softening plants in the United States.
The data from Iowa and Illinois are the most reliable, in that radium and
hardness analyses were performed on water samples taken at the same
time. The hardness data from Florida, on the other hand, were
-------
reconstructed from plant operating records on the same day that the
radium samples were taken by another agency. This means that the two
samples may have been taken as much as eight hours apart. It can be seen
from Table 4 that radium removal varied from plant to plant and ranged
from 59-96 percent removal, averaging 80 percent. Figure 7 shows a plot
of total hardness removal versus radium removal based on the data
presented in Table 4. If the two points from Florida plants are
discounted, the line y = x2-86 fits the data reasonably well ^nd serves
as the basis for a radium removal model in lime-soda plants." (Singley,
et al., 1977) Despite the fact the the data show some scatter, a model
based on the data in Table 4 for the fraction of hardness removal as
Total Hardness Removed Fraction = THR = f2-85
where f is the radium removal fraction, as before. The total hardness
removal fraction for achievement of a final radium concentration of
5 pCi/L can be estimated by combining the equations and getting
THR = 1 - (5/RWR) 2-86
This is plotted in Figure 8 for radium concentrations in the raw water
from 16 to 50 pCi/L. The lower limit for radium of 16 was chosen since a
lime-soda softening plant would rarely if ever reduce the total hardness
by less than 35%, i.e., from 150 to 100 mg/L, as CaC03. This would
compare to a radium removal fraction of about 70%, or from a RWR of
16 pCi/L down to 5 pCi/L. A correlation of radium concentrations with
other raw water quality parameters is shown in Table 3.
A major problem with any process for removal of radium is the
disposal of the waste stream(s). In the case of lime-soda softening
there are two streams, filter backwash and precipitated sludge.
Another process for the removal of ionic species, including
hardness, ion exchange, is also applicable to the removal of radium. In
this process the divalent radium ion is exchanged for sodium or hydrogen
ions. Simultaneously other divalent ions such as calcium and magnesium
are also exchanged for sodium ions to accomplish softening. Fortunately,
the ion exchange materials have a greater affinity for radium than either
calcium or magnesium since the concentration of radium is always much
lower than either calcium or magnesium. This also provides an additional
measure of protection since the hardness breakthrough precedes the radium
breakthrough. A study, sponsored by EPA, evaluated a radium selective
complexer (Dow Chemical Company, RSC) for removal of radium from brines.
The resin was shown to have a very high capacity for radium in water
having a TDS of 450 mg/L (51,000 pCi/g dry resin); Both weak and strong
acid exchange resins were shown to remove radium at over 96% efficiency
(Clark, 1987). Conventional cation exchange resins have exchange
capacities of 18,000 to 30,000 grains of hardness per cubic foot. A
diagram of a typical ion exchange unit is shown in Figure 9.
Since ion exchange for radium removal is almost 100% effective, the
flow fraction to be treated can be obtained from Figure 1. The actual
efficiencies for ion exchange in operating plants are shown in Table 5.
They are seen to average over 95% for those plants where no operational
-------
problems affected the results, thus the flow fraction can be modified to
accommodate the difference. Figure 10 shows the relationship between
radium and total hardness removal for the plants shown in Table 4. It
illustrates the greater affinity of the resins for radium than for
hardness ions.
Since the resins are greater than 95% effective for removal of both
hardness and radium, it is usually necessary to treat a fraction of the
water and blend with a by-passed untreated fraction. Assuming 95%
efficiency, the fraction to be treated can be calculated from: fraction
treated = Ft = 1.053(l-5/RWR).
As with lime-soda softening a major problem is disposal of the
waste stream, in this case the unused regenerate brine containing the
radium that was removed, diluted with the wash water.
Another method for removing radium is reverse osmosis. This method
is effective for ionic species as well as any molecular species which has
a size greater than the pores in the semi permeable membrane used. The
pressure required to produce a water of acceptable quality depends upon
overcoming the osmotic pressure of the raw water, the percent recovery
required, the properties of the membrane and the quality of the finished
water desired. Studies supported by EPA (Clark, 1987) showed that from
82-96% removal of radium could be accomplished.
Radioactivity. The control of radioactivity in drinking water is
summarized in Figure 11, from Aieta et al, (1987).
The only element presently regulated is radium, thus emphasis was
placed on it. Table 6 provides more detail as to the efficiency of the
various treatment technologies.
SELENIUM.
Selenium is chemically similar to arsenic and is generally present
in two valence states in water Se IV, selenite, and Se VI, selenate, both
as the oxides SeOs"2 and SeO^2. it is obvious that anionic exchange
resins will effectively remove both species. Coagulation with ferric
sulfate results in about 80% removal of Se IV but less than 10% Se VI.
Lime softening at high pH, i.e.>11.5, was only about 50% effective for
Se IV and less than 10% for Se VI.
SILVER.
The occurence of silver in untreated water is so rare that EPA has
proposed that it be eliminated from the regulated inorganics.
STRONTIUM.
Since strontium is an alkaline earth metal in the same chemical
family as calcium, barium and magnesium, its chemical properties are
similar. The technology for removal of strontium is the same as
discussed above for barium and radium.
-------
ARSENIC
Arsenic occurs in water primarily in the tri- and pentavalent form,
usually as anionic oxides, As03~3 and AsO/f3 see Figure 12 for effect Of
pH on arsenic species. The treatment techniques for arsenic V removal
using conventional coagulation with alum or - ferric salts show good.
removal below pH 8 by ferric sulfate (95+%) and about 90% by alum at pH
below 7. Arsenic III is only about removed 50% by ferric sulfate and
less than 15% by alum at the same pH values. Lime softening removes
arsenic V at pH values above 10.5 at about 99% efficiency but arsenic .III
at only about 80% above pH 10.5.
Although data are limited there is little doubt that reverse
osmosis and anion exchange will remove both forms effectively, if the
arsenic is present complexed by organics it would be removed effectively
by activated carbon. Figures 13 and 14 show the effects of ferric
sulfate coagulation and lime softening (Sorg and Logsdon, 1978b).
FLUORIDE.
The presence of fluoride in drinking water is regulated under both
the primary and secondary contaminant limits due to its health and
cosmetic effects. The anomalous addition of fluoride by many water
purveyors to reduce dental caries presents a conundrum. Excessive
fluoride must be removed but lower concentrations may require
supplementalion.
The most cost-effective control treatment method has been shown to
be the adsorption on activated alumina (Rubel, 1984). Several full-scale
plants have been operated in the U.S. using this process, which was
patented by Churchill in 1936. Studies at these plants and in the
laboratory have shown that capacities greater than 2,000 grains/cu. ft.
can be obtained by optimizing the pH at about 5.5 units. The major
sources of activated alumina is Alcoa but other sources are available.
It is possible to reduce the fluoride ion concentration concurrently with
TDS reduction by reverse osmosis. Rejections up to 90% have been
achieved. Lime softening can remove fluoride when high pH is used to
precipitate magnesium. Alum coagulation will reduce fluoride because of
the formation of insoluble aluminum fluoride. High concentrations of
alum are required, which minimizes the selection of this process unless
such alum concentrations are required for other purposes.
The design of an activated alumina plant is covered in detail by
Rubel (1984) in a manual written for the U.S. EPA.
NITRATE.
Nitrate occurs at levels above the MCL of 10 mg/L (expressed as N)
in many groundwater supplies in the U.S. Until recently there was little
alternative .to use of high nitrate water. Now there are two techniques
for removal, ion exchange and reverse osmosis, R.O. Both remove nitrate
so effectively that some untreated water usually can be by-passed and
blended. Nitrate removals of 75-95% have been experienced 1n RO
plants. (Clark, 1987).
10
-------
SUMMARY
Although many inorganic species occur in natural waters, there are
a limited number of applicable treatment processes for their removal.
The twelve that are presently regulated appear in the regulations because
of their toxicity and high incidence of occurence in raw waters used as
sources for drinking water. A brief summary of the best treatment
process(es) for each of the twelve is shown in Table 7. Although this is
not a comprehensive compilation of relative effectiveness, it does give
an overview.
11
-------
REFERENCES
Aieta, E.M., Singley, J.E., Trussell, A.R., Thorbjarnarson, K.W. and
McGuire, M.J. (1987), Radio nuclides in Drinking Water: An
Overview, Jour. AWWA 79:144.
AWWARF/EB1 (1985) Internal Corrosion of Hater Distribution Systems,
American Water Works Association Research Foundation, Denver.
Churchill, H.V., U.S. Patent 2,059,553 (November 3, 1936).
Clark, R.M. (1987), Treatment Techniques, Drinking Water Research
Division, U.S. EPA, Cincinnati (unpublished report).
Hursh, J.B. (1954), Radium Content of Public Water Supplies, Jour. AWWA
vol.46:43.
Rubel, F. Jr. (1984) Design Manual: Removal of Fluoride from Drinking
Water Supplies by Activated Alumina. EPA-600/2-84-134, Cincinnati.
Singley, J.E., Beaudet, B.A., Bolch, W.E., and Palmer. J.F. (1977),
Costs of Radium Removal From Potable Water Supplies, U.S. EPA,
EPA-600/2-77-073) Washington, D.C.
Sorg, T.A. (1978a) Treatment Technology to Meet the Interim Primary
Drinking Water Regulations For Inorganics: Part 1, Jour. AWWA
70:105.
Sorg, T.A. and Logsdon, G.S. (1978b). Treatment Technology to Meet the
Interim Primary Drinking Water Regulations For Inorganics: Part 2,
Jour. AWWA 70:379.
Sorg, T.A., Csanady, M., and Logsdon, G.S. (1978c) Treatment Technology
to Meet the Interim Primary Drinking Water Regulations For
Inorganics: Part 3, Jour. AWWA, 70:680.
Sorg, T.A. (1979) Treatment Technology to Meet the Interim Primary
Drinking Water Regulations For Inorganics: Part 4, Jour. AWWA,
71:454.
Sorg, T.A. and Logsdon, G.S. (1980) Treatment Technology to Meet the
Interim Primary Drinking Water Regulations For Inorganics: Part 5,
Jour. AWWA 72:411.
-------
TABLE 1. PROPOSED MCLG^) FOR INORGANIC
CONTAMINANTS
CONTAMINANT MCLG, mg/L
As 0.050
Asbestos 7.1
Ba 1.5
Cd 0.005
Cr 0.12
Cu 1.3
Pb 0.020
Hg 0.003
N03 10.0
N02 1.0
Sw 0.045
(!) MCL goal - as changed from RMCL by SDWA Amendments, 1986.
(2) MFL = million fiber/liter.
-------
TABLE 2. Ra CONTENT OF SELECTED RIVERS IN THE UNITED STATES
City Supply
Source
Atlanta, Ga.
Baltimore, Md.
Birmingham, Ala.
Bismarck, ND
Boston, Mass.
Charleston, S.C.
Charleston, W. Va.
Cincinnati, Ohio
Denver, Colo.
Detroit, Mich.
Indianapolis, Ind.
LaVerne, Calif.
Louisville, Ky.
Oklahoma City, Okla.
Omaha, Neb.
Philadelphia, Pa.
Phoenix, Ariz.
Pittsburg, Pa.
Portland, Ore.
Raleigh, N.C.
Richmond, Va.
Sacramento, Calif.
Salt Lake City, Utah
San Francisco, Calif.
St. Louis, Mo.
Tacoma, Wash.
Washington, D.C.
Ra Concentration
pCi/llter H-,0
Raw
Water
Chattahoochee R.
Gunpowder R.
Cahaba R. and L. Purdy
Missouri R.
Nashua R.
Edisto R.
Elk R.
Ohio R.
South Platte R.
Detroit R.
Fall Cr. and White R.
Colorado R.
Ohio R.
N. Canadian R.
Missouri R.
Delaware R.
Along Verde R.
Allegheny R.
Bull Run R.
Walnut Cr.
James R.
Sacramento R.
Cottonwood Cr.
Calaveras Res.
Mississippi R.
Green R.
Potomac R.
0.017
0.020
0.024
0.243
0.014
0.181
0.041
0.061
0.077
0.026
0.137
0.100
0.084
0.106
1.770
0.048
0.027
3.700
0.014
0.022
0.033
0.018
0.034
0.018
1.080
0.002
0.033
Hursh (1953)
-------
TABLE 3. CORRELATION OF RAW WATER QUALITY PARAMETERS WITH RADIUM
(Singley, et al. 1977)
DCI, IL
Peru, IL
Herschcr, IL
Elgin, IL
Lynwood, IL
Greenfield, IA
Adair, IA
Stuart, IA
Eldon, IA
Estherville, IA
Grinnel, IA
Hols^tein, IA
Webster City, IA
West Des Koines, IA
Sarasota, FL
Venice, FL
Englewood, FL
Mean
r*
ra = 0.05**
Ra2264
3.26
5.82
14.3
5.55
14.7
14.0
6.30
16.0
50.0
5.2
6.2
14.0
7.1
9.6
4.30
8.73
1.69
11.0
THb
286
298
404
246
849
630
710
640
400
915
368
920
530
376
460
570
360
527
0.057
0.606
Cab
153
188
226
142
620
400
450
375
246
600
205
600
275
210
335
0.036
0.532
Mgb
133
110
178
104
229
230 .
260
265
154
315
163
320
255
166
206
0.048
0.532
Fec
0.44
0.44
0.12
' 0,04
0.60
1.6
0.58
0.94
1.9
1.6
1.1
1.8
0.69
0.36
0.87
0.518
0.532
TDSC
1220
890
1426
350
1766
2160
1905
1770
1228
1350
784
1510
1010
1200
____ '•
_
1327
0.176
0.532
ALKb
286
318
259
303
215
190
158
182
252
367
298
288
294
260
____
•
262
0.252
0.532
Bac
0.13
0.13
0.10
•8.7
0.09
'<0.10
-------
TABLE 4. RADIUM226 AND TOTAL HARDNESS REMOVAL EFFICIENCIES LIME-SODA SOFTENING
(Singley, et al. 1977)
LOCATION
W. Des Moines, [A
8/1/74
Webster City, IA..
Clarlfier 11
8/13/74
Webster City, IA
ClarUier S2
3/13/74
Webster City, IA
Clarifier 111
2/20/75
Webster City, IA
Clarifier 11
2/20/75
IVru, IL
2/20/75
Peru, IL
2/25/75
Peru, IL 3/4/75
Mean
Elgin, IL 3/7/75
ElRin, II. 3/14/75
Elgin, IL 3/21/75
Mean
Venice, FL 8/7/75
Englewood, FL
8/7/75
CLARIFIKPS
Ra in
pCi/1
9.3
6.1
h - 1
7.8
7 . S
fi.49
5.41
5.48
5.82
7.45
5.7
3.51
5.55
8.73
1.69
Ra out
pCl/1
2.6
1.9
2.6
0.9
0.3
% RA
Remova 1
72
69
57
88
96
TH In
mg/l CaCO
376
507
507
482
482
329
278
236
289
246
243
242
244
570
360
TH Out
mR/1 CaC01
215
333
282
150
150
% Til
RFJ1
43
34
44
69
69
pH Treat-
ment
10.4
10.05
10.1
10.95
10.95
8.4
8.4
8.4
10.2
10. -2
10.2
9.7
8.5
FI1.TF.RF
RA In
pCI/l
2.6
1.9
2.6
0.9
0.3
RA Out
pCI/1
2.35
0.9
0.9
0.3
0.3
1 R.-,
P^mova I
10 '
53
65
Til In Tl! Out
ir.c/IC.-njO ! ma /I racn
215 190
X TH
REJI
12
331 262 ! 21
2RT. ' 262 7
67 j 150 ' 106
1 !
1
0
0.51
1.62
1.33
1.15
.75
.80
.71
.75
2.19
.69
150 ! 106
! 174
180
122
147
99
112
95
102
100
166
29
29
OVERALL
7. Ra
RP-I
?5
85
85
96
96
92
70
76
79
90
86
80
85
75
59
7, TH
REM
49
4S
48
78
78
47
35
57
49
60
54
61
59
82
5'.
-------
Table5. RADIUM REMOVAL IN ION EXCHANGE PLANTS
Plant
Eldon, la
Estherville, la
Grinnell, la
Holstein, la
Dwight Corn
Inst., II
Herscher, II
Lynwood, II
SarasoTa, Fl
Ra in
pCi/L
49
5.7
6.7
12
3.26
14.31
14.69
4.3
Ra out % Ra
pCi/L Rem. •
1.9
0.3
0.2
0.5
0.36
1.31
0.41
0.70±
96
95
97
96
89
91
97
• 84
TH in TH out
mg/L/CaC03 mg/L/CaC03
375
915
385
920
286
406
848
460
'0
46
11
18
43
60
78
159±
% TH
Rem. •
97
95
97
98
85*
85*
91
65**
• Removed
± Adjusted to take account of raw water blending.
* % Hardness and %Ra removals are somewhat low due to breakthrough
occurring prior to all samples being collected.
**%Hardness and % Ra removals are somewhat low due to incomplete
regeneration of media as 1/2 brine pumping capacity was down for
repair.
-------
TABLE 6. TREATMENT TECHNOLOGIES FOR REMOVING RADIONUCLIDES
(Aleta, et al. 1977)
Treatment
Technology
Coagulation-filtration
Lime-soda softening
Ion exchange
Adsorption
Aeration
Reverse osmosis
Radionuclide
Ra
U
Rn
Ra
U
Rn
Ra
U
Rn
Ra
U
Rn
Ra
U
Rn
Ra
U
Rn
Approximate
Process
Efficiency
percent
<25
95+
18-98
80-95
75-96
85
43-92
80
85-90
99
95+
99
99
90+
85-90
62-99
18
20-25
20-96
93+
87-96
87-98
95+
95+
Comments
High pH and Mg required
High pH (10+) and high
dosages of ferric chloride
or alum
Only accomplished in lab
studies with diatomaceous
earth filtration
N/A*
Best choice for large plants
Plant-scale results
Plant-scale results
pH 10.6-11.5
High pH, high Mg
N/A
Best choice for small plants;
cation exchangers
Brine disposal problem
Anion exchangers; largely
experimental
N/A
Experimental
Sand adsorption;
experimental
N/A
GAC adsorption
Adsorption on Fe(OH)a
Fe(OH)3 and sand
N/A
Depends on process
Depends on process
Plant-scale data
Based on eight plants
High-volume brine solution
for disposal
High-volume brine solution
for disposal
N/A
*N/A = not applicable
-------
TADLE 7. SUMMARY OF BEST TREATMENT PROCESSES FOR
REMOVAL OP INORGANIC CONTAMINANT
Contaminant Coagulation
Line
Softening
Exchange
Activated
Alumina
GAC R.O. (a)
..(c)
Bone char +
Da
Cd
Cr III
F
+pH>8
+pH 9.5-10.5
>10.5
+
+
Bone char +
Pb
Hg (Organic!
Ilg ( Inorganic )
(103
Se
(Se IV)
E pH 9.5-10.B +
Ho - reverse osmosis, also clectrodialysis assumed to be equivalent.
As III can bo oxidized to As V with C12.
•*• mcanB 90-1001 removal.
+ moans cation exchange; - moans anion exchange.
-------
200
*mg/L,as CaC03
8
PH
Figure 1. Solubility of Lead as a Function of pH
and Carbonate. (Adapted from AWWARF/EBI,
1985)
-------
P04
Concentration
minimum Pb
atpH=7.6
Figure2. Lead Concentration as Function of pH and
Phosphate Concentration at a Total Alkalinity
of 50mg/L (Adapted from Schock)
-------
JE
>N
.£>
13
I I I I I
.001
Figure3. CADMIUM SOLUBILITY AS A FUNCTION
OF pH,
-------
C
o
c
0)
o
c
o
o
o
1.0
0.1
0.01
Cr(OH),
MCL
1
I
6 7 8 9 10 11
PH
Figure 4. Solubility of Cr111 as a Function of pH
-------
TJ
CD
O
O
to
CD
"o
CD
Q.
c:
CD
CO
CD
°- .20 ~
0
Figures. Effect of pH on Speciation of Chromium VI
-------
.1OO r
CD
(T _J
_. \
> "^
o °-
E 10
^ •*-
CL o
II
O
o:
(D
£ °
o •»-
o
D
.80
,40
.20
10
20
30
40
50
60
Radium Level -Raw Water (pCi/L)
Figures. REQUIRED REMOVAL FRACTION FOR
RADIUM
-------
1.00
|.80
O
o
.60
O
QC
-------
O
-------
METER
INLET
ION-EXCHANGE
UNIT
EXCHANGE
MATERIAL.
BACKWASH
OUTLET
BACKWASH
INLET
OUTLET
RINSE
OUTLET
WASH-WATER
COLLECTOR
I'
PRESSURE WATER
EJECTOR
SUPPORTING
BED
REGENERANT TANK
TO WASTE
Figure 9. DIAGRAM OF TYPICAL ION EXCHANGE UNIT (3)
-------
1.00
.20
.40 .60 .80 1.00
Total Hardness Removal Fraction
FIGURE 10. RADIUM REMOVAL FRACTION
VS. TOTAL HARDNESS REMOVAL
FRACTION IN ION EXCHANGE PLANTS
(BEFORE BLENDING)
(Adapted fromS/ng/ey, eta I, 1977)
-------
4-
D
O
O)
"o
0)
Q.
•t—
c
-------
0.40
c
o
0.30
£ 0.20
o
c
o
o
o
£ 0.10
0.00
As*
Initial Concentration
0.4 mg/L
MCL
8
10
11
PH
Figure 13. Effects of pH on Arsenic Removal by
Lime Softening (Adapted from Sorg 8
Logsdon 1978)
-------
0.20
0.15
0.10
0.05
0.00
Asm0.3mg/L
Initial Concentration
As Shown
MCL
Figure 14. Effects of pH on Arsenic Removal by
Ferric Sulfate Coagulation (30mg/L)
(SorgB Logsdoml978)
-------
WORKSHOP ON EMERGING TECHNOLOGIES FOR
DRINKING WATER TREATMENT INORGANIC
CONTAMINANTS PROCESS,PLANNING OVERVIEW
The planning of appropriate treatment processes for the twelve regulated
inorganic contaminants is complex and thus difficult to adapt to a rigid check-
list format, although there are many elements in common. It is important to
keep in mind that there are many alternatives in each case, not only process-
wise, but compatibility with existing process streams and equipment. In many
cases the decision-making process may be considerably simplified by application
of common sense which cannot be programmed. Figure 1 shows many of the
elements that can be used sequentially - most of which are obvious.
-------
CONTAMINANT
-------
Inorganic Contaminants
Process Design Check List
I Contaminant Identification
A. Historical Records
B. Current Analyses
II MCL Compliance Evaluation
A. Present MCL
B. Proposed MCL
C. Treatment Required
III Plant and Process Evaluation
A. Physical Evaluation
1. Treatment Units
2. Appurtenances
3. On-site Space
B. Process Evaluation
1. Operating Parameters
2. Flexibility
IV Evaluation of Alternate Processes
A. Coagulation
1. Alum
2. Iron Salts
B. Lime-soda Softening
C. Ion Exchange
1. Cation
2. Anion
D. Reverse Osmosis
E. Activated Carbon
1. PAC
2. GAC
F. Activated Alumina
-------
V Optimization of Existing Processes
A. Feasibility
VI Evaluate Compatibility With Plant
A. Physical
B. Chemical
1. Control Contaminant(s)
2. Maintain Quality
VII Evaluate Process Charges in Existing Plant
VIII Evaluate Technical Alternatives
A. Waste Stream Disposal
1. Solids
a. Hazardous
b. Non-hazardous
2. Liquids
a. Hazardous
b. Non-hazardous
B. Cost
1. Capital
2. O&M
C. Energy Dependence
D. Reliability
E. Operational Compatibility
IX Final Selection
A. Process
B. Plant
1. Modifications to Existing Plant
2. New Plant
-------
PROCESS CHECK LIST
Evaluation of Alternative Processes
A. Contaminant(s) Present
1. Identification of those exceeding MCL
2. Historical record of contaminant(s)
3.
B. Candidate Process(es) Considered
1. Non-treatment
a. Blending
b. New source(s)
c. Purchase
1) Raw
2} Treated
d. Bottled water
e. Point-of-use treatment
2. Treatment
a. Coagulation
1) Alum
2) Iron salts
b. Softening, lime-soda
c. Ion exchange
1) cationic
2) anionic
3) pre-treatment
d. Reverse osmosis (electrodialysis)
1) high pressure
2) low pressure
3) pre-treatment required
e. Activated Alumina
f. Activated Carbon
1) PAC
2) GAC
C. Compatibility With Present Process
1. Process reoptimization
2. Facilities
3. Retrofitting
D. Waste Disposal
1. Solids
a. hazardous
b. non-hazardous
2. Liquids
a. hazardous
1) concentrate
2) neat
b. non-hazardous
1) .dry (concentrate solids)
2) recycle portion
3) sewer
4) return to source
-------
PROCESS CHECK LIST - Evaluation of Alternative Processes
Page 2
E. Cost Evaluation
1. Processing
a. Capital
b. 0 & M
2. Disposal
a. Capital
b. 0 & M
3. Environmental Permits
-------
Sample Ra Control Process
(Applicable to Any Other Contaminants)
I. Lime-Soda
Raw Finished
Parameter Mg/L Meq/L pCi/L Mg/L Meq/L PCi/L
TH,asCaCO^ 300
Ca, " J 250 12.50
Mg, " 50 4.17
Alk, " 200 4.00
TDS 400
CO 15 3.40
Ra 22 <5.0
A. Estimate the finished Total Hardness (TH)
B. Estimate the finished .a concentration
C. Approximate the pH of treatment
D. If the population of the service area was 15,000,
what would be
a. the approximate volume of filter backwash,
in gal/day?
b. the approximate dry weight of the sludge?
c. the radium content of the sludge (70% dry)?
d. the radium content of the backwash water?
II. Ion Exchange
Same water, same population
Assume 1) operating capacity of resin^
20 kgr hardness as CaC03/ft
2) regeneration with 10% brine
3) salt for regeneration =
0.30 pounds salt/kgr. hardness
removed (as CaCO.,) or
6 pounds/ft. .
4) 95% TH removal
5) 95% Ra removal
A. Calculate the fraction of the flow to be treated.
B. Estimate the amount of regenerant salt, per million
gallons of finished water.
D. Estimate the total volume of brine to be disposed
(backwash + rinse water) per MG.
E. Estimate the Ra concentration in the waste brine
(bw + rinse).
Ill. Reverse Osmosis
Same water, same population.
Assume 1) 95% removal of all ions
2) 95% removal of Ra
3) blending to 5 pCi/L Ra
-------
Sample Ra Control Process
(Applicable to Any Other Contaminants)
Page 2
A. Calculate the TH of the blended, finished water.
B. Calculate the fraction of water to be treated.
C. Calculate the reject brine Ra concentration.
IV. what would be the most compatible choice of treatment
process for the following utilities? What factors should
be most relevant?
A. Utility has only wells with chlorination.
B. Utility has lime-soda softening plant, reducing
Calcium hardness from 190 to 70 mg/L, as CaCO.,.
Its source is a river with average turbidity of 8 ntu.
C. Utility has an ion exchange plant for softening a well
supply from 200 to 100 mg/L as CaC03.
-------
SAMPLE CALCULATIONS
Lime-Soda Process
High Solids Raw Water
TH - 750 mg/1 @ CaCC>3
Ca++ - 500 mg/1 @ CaC03
Mg^ - 250 mg/1 <§ CaC03
ALK - 300 mg/1 @ CaC03
IDS - 2000 mg/1
C02 - 11.4 mg/1 @ CaC03
1 - Desired finished water for RaQ =7.5 pCi/1
Ra removal required - 33 percent or .33
TH removal required - Ra2-86
rem
= .332-86 = -042
Minimum TH considered practical in lime-soda process is 35 percent
TH removal required - .35 (750) = 262 mg/1 @ CaCC>3, or 5.25 mg/1
TH of finished water= 488 mg/1 @ CaC03 or 9.75 mg/1
1
2 - Chemical requirements
Parameter
CO 2
Ca"1"1"
Mg"^"
HC03
TOTALS
mg/1
5
200
60
366
Raw
AfJ
njgyl
0.23
10.00
5.00
6.00
Finished
meq/1
0
4.75
5.00
0
A
gxr
mgyl
\
0.23
5.23
0
6.00
Required
CaO
meq/1
0.23
5.25
'
0.75
6.23
Required
CaO
meq/1
_
-
-
-
Lime Required = i^-23^28) = 194 mg/1
Soda Ash Required = None
-------
IJJ4 Exchange
1 - Mass Balance Calculation
To determine fraction of raw water influent blended, and fraction treated,
assuming 95 percent ren .ival in I/E or R/O column.
For mass balance of Ra:
.05 Ra0x + RaQ (1-x) = (1) (5)
.05 Ra0x + Ra0 - Raox = 5
x Ran (.05 - 1) 4- Ra0 = 5
-.95 x Ra0 = 5 - Rao
X = ^pVo
x = Ra0 - 5
.95 Ra0
Blending allowed to attain minimum level of 5 pCi/1 Ra assuming 95 per-
cent Ra removal in I/E and R/0
Ra Level Fraction Blended Fraction Bicttderd
50 .053 .947
20 .211 .789
7.5 .649 .351
2 - Regenerant Chemicals Calculations
Assume: 1) operating capacity of resin = 20 kgr hardness as CaCOo/CF
2) regeneration is with a 10 percent brine concentration
3) salt required for regeneration is 0.30 Ib salt/kgr hardness
as CaCo3 removed (6 Ib salt/CF)
4) 95 percent TH removal in unit (5 percent leakage)
-------
Solids Levels:
High - TH = 750, TH_ = .05 (750) = 37.5
—— raw F
TH = 712.5 rag/1
rem &
Kilograins hardness removed
712. 5 mg/lxl grain/gal^ kgr
17.1 mg/l*1000 gr '
Salt regeneration requirement/MG
.0417 x 106 kgrx-30 Ib salt _ 2
rag kgr '
Volume of rinse water/MG, assume 30 gal/CF resin
41,700 kgrTHr CF 30 gal _
mg 20 kgr CF ~
Volume of regeneration brine solution/MG
- 12,510 = 112,590 lb H20
13,500 gal H20
Total volume of backwash + rinse water/MG
High Solids - 13,500 + 62,550 = 76,050 gal or 7.6 percent of total flow
-------
Reverse Osmosis Calculations
1 - Assumptions
High Solids - Initial TDSi = 2000 mg/1
Product TDSi = .05 (2000) = 100 mg/1
Calcium initial, Cai = 200 mg/1
2 - Brine to Product Ratio (BPR) =
RPR _ 1 - (TDSp/TDSj)
B K (900/Cai) - 1
_ i - (100/2000) _ ^9i _ 271
BPR ~ (900/200) - 1 " 3.5 " -271
3 - Brine Volume
Vb = Vp x BPR = .271 Vp Vp = Product Volume
4 - Feed Volume (Vj_)
Vi = Vp + .271 Vp = 1.271 Vp
-------
REMOVAL MECHANISMS
PRECIPTATION
CO-PRECIPTATION
ADSORPTION
COAGULATION
OXIDATION
COMPLEXATION
ION EXCHANGE
-------
ANIONS:
SPECIAT1ON
As (AsO^AsOi3)
Cr (CrOi2,CrO?)
F-
Se(SeO"2SeO;2)
CATIONS:
Ba+2
Gd+2
Pb+2
Ag
+1
COMPLEXES:
Hg (ORGANIC)
HgOH*
HgCL*
CdOH+
MC0°
-------
SPECIATION PARAMETERS
pH
SPECIES CONCENTRATION
COMPLEXANT CONCENTRATION
PARTICULATE CONCENTRATION
REDOX POTENTIAL
IONIC STRENGTH
TEMPERATURE
-------
EFFECT OF pH
IONIC SOLUBILITY:
HYDROXIDES EXCEEDS "REGS" ONLY BELOW:
Cd pH <11.0
Pb pH < 6.2
Hg pH < 7.0
ACIDIC SPECIES:
As03"—HAs Oa
As 0? - HAs 042- H2As 041- H3As 04
CrQ;2-HCr04--H2Cr04
Cr2a27-HCr20;-H2Cr207
SeOj2—HSe 0"3—H2Se 03
SeO^-HSe 04-H2Se 04
F-HF
-------
100
80
I 60
tu
a:
g 40
Ul
20
MCL for
0.4 mg/l
As 0.4 mg/l
0 As*3 chlorinated
5JC As*3 not chlorinated
Pilot plant tests
XAs*5
$£ As*3 not chlorinated
****
*.-
-*••-*
9 10
pH OF TREATED WATER
ARSENIC REMOVAL BY LIME SOFTENING
11
12
100 -
80
Q
Ul
o 60
ui
Q.
40
20
,,,,,,...
--
Lime softening
pH 10.9-11.1
A As*5
Ferric sulfate 30 mg/1
• As*5
OAs*3
Alum 30 mg/1
QAs*5
A As+3
I
I
I
0.10
0.5 1.0 5 10 20
ORIGINAL CONCENTRATION, mg/l
ARSENIC REMOVAL BY COAGULATION AND LIME SOFTENING
-------
CADMIUM SOLUBILITY AS A FUNCTION
OF pH.
-------
100
• 111»1111 • 111 • 11 n t • I • ® 111111 • « • • • • f}9M
Illl
80
2 60
MCLfor
0.03 mg/l
UJ
UJ
e_
40
20
Cd 0.03 mg/l
• Well water
X Pilot plant tests
I
I
9 10
pH OF TREATED WATER
CADMIUM REMOVAL BY LIME SOFTENING
11
12
100 -
80
Q
UJ
2 60
40
20
Cd 0.03 mg/l
River water
:J* Ferric su I fate
O Alum
Pilot plant tests
X Ferric sulfate
7 8
pH OF TREATED WATER
CADMIUM REMOVAL BY ALUM
COAGULATION
10
AND FERRIC SULFATE
-------
100
80
Q
UJ
6°
Ba 7-8 mg/l
I Hardness
_ X Pilot plant tests
UJ
te.
l-
UJ
y 40
HI
a.
20
MCLfor
8 mg/l
*
_L
:•*«,_—
W
\
tr
9 10
pH OF TREATED WATER
BARIUM REMOVAL BY LIME SOFTENING
11
12
100
80
UJ
oi
LJ
Ul
a.
60
40
20
MCL for
8 mg/l
Ba 7-8 mg/l
:£ Ferric sulfate 20-30 mg/l
O Alum 20-30 mg/l
BARIUM
COAGULATION
7 8
pH OF TREATED WATER
REMOVAL BY ALUM
10
AND FERRIC SULFATE
-------
1.00
•4—
o
o
.80 ~
.60
o
CD
O.
CO
S .40
4—
C
CD
CD
£ .20
O
5 6
PH
8
Effect of pH on Speciation of Chromium VI
-------
4—
O
-------
100 -
80
*•"
MCL for
0.15mg/l
UJ
I 60
UJ
C£.
UJ
40
Ul
Ag 0.15 mg/l
• Well water
20
I
I
9 10
pH OF TREATED WATER
SILVER REMOVAL BY LIME SOFTENING
11
12
100
80
o
UJ
tt
t
UI
a.
60
40
Cr*3 pH 7.3-7.6
Well water
Chlorine 2.2 mg/1
s|e 6 h contact time '
O 20 h contact time
20
0.10
0.5 1.0 :
ORIGINAL CONCENTRATION, mg/l
10
20
EFFECT OF PRECHLORJNATION ON Cr+3 REMOVAL BY ALUM
COAGULATION
-------
100
80
PERCENT REMOVED PERCENT REMOVEt
0 g £ 8 8 8 o8oS
9""ii,,
*****»*,
«•••
0
1
Se*4 0.03 mg/l
Ferric sulfate 25 mg/l
£ River water
jj: Well water
Alum 25 mg/l
*""••£,. MCL for D River water
"»,- 0.03 mg/l 0 Well water
•••.£<,, ^^ Pilot plant tests
x '^"Xj *tf*tt Gravel pit water
VN*. ' ***«>, X Ferric sulfate
^*^^ %^ River Water
^^^ **^ + Ferric Sulfate
1 i 1 M 1 i 1
6 7 8 9 10
pH OF TREATED WATER
Se+4 REMOVAL BY ALUM AND IRON COAGULATION
~" MCL for 0.1 mg/l
MCL for 0.3 mg/l
_ Se+4 Well water
~" • 0.1 mg/l
A 0.03 mg/l
X Pilot plant tests
~ Se*4 0.03 mg/l
1
8
^farjj%4fr**^**f\*~^
— i 1 1 1 i 1 i |
9 10 n i:
pH OF TREATED WATER
Se+4 REMOVAL BY LIME SOFTENING
-------
TOO
80
2 60
o
tu
ui
U
e£
40
20
Pbi
&**—*„*• i«
a*6
U •••€»•
I
6
• Illllllllflllll""1111
pH OF TREATED WATER
REMOVAL OF INORGANIC CONTAMINANTS BY ALUM COAGULATION
-------
100
tu
u
t£
80
S 60
o
LJ
40
20
r
7 8
pH OF TREATED WATER
REMOVAL OF INORGANIC CONTAMINANTS BY IRON COAGULATION
-------
100
80
Q
ui
O
ui
ui
Ui
O.
60
40
20
0*3
J
8 9 10 11
pH OF TREATED WATER
FIGURE 3 REMOVAL OF INORGANIC CONTAMINANTS BY LIME SOFTENING
12
-------
TREATMENT PROCESS
CONVENTION COAGULATION
PRINCIPAL APPLICATION
FOR WATER TREATMENT
CLARIFICATION OF SURFACE WATERS
HIGH
INORGANIC CONTAMINANT
TREATMENT CAPABILITY
EFFECTIVENESS*
MODERATE
LOW
Cd, Cr III, Cr VI, As V, Ag, Pb
As III, Se IV, Hg(0), Hg(l)
Ba, F, NO. Ra, Se VI
3
MOST PROBABLE
APPLICATION FOR
INORGANIC REMOVAL
REMOVAL OF Cd, Cr, As, Ag, OR Pb FROM SURFACE WATERS
HIGH - GREATER THAN 80 PERCENT MODERATE - 20 TO 80 PERCENT LOW - LESS THAN 20 PERCENT
-------
TREATMENT PROCESS
LIME SOFTENING
PRINCIPAL APPLICATION
FOR WATER TREATMENT
REMOVAL OF HARDNESS FROM
GROUND AND SURFACE WATER
HIGH
INORGANIC CONTAMINANT
TREATMENT CAPABILITY
EFFECTIVENESS*
MODERATE
LOW
Ba, Ra, Cd, Cr III, As V, Pb
Se IV, As III, Hg(l), F
Cr VI, NO , Se VI, Hg(O)
3
MOST PROBABLE
APPLICATION FOR
INORGANIC REMOVAL
REMOVAL OF Ba OR Ra FROM GROUNDWATERS;
REMOVAL OF Cd, Cr III, F, As V, OR Pb FROM HARD
SURFACE WATERS REQUIRING SOFTENING
HIGH - GREATER THAN 80 PERCENT MODERATE - 20 TO 80 PERCENT LOW - LESS THAN 20 PERCENT
-------
TREATMENT PROCESS
CATION EXCHANGE
ANION EXCHANGE
PRINCIPAL APPLICATION
FOR WATER TREATMENT
REMOVAL OF HARDNESS
FROM GROUNDWATERS
REMOVAL OF NITRATE
FROM GROUNDWATERS
HIGH
INORGANIC CONTAMINANT
TREATMENT CAPABILITY
EFFECTIVENESS*
MODERATE
LOW
Ba, Ra, Cd, Pb, Cr HI
NO , Cr VI, Se
3
Ba, Ra, Cd, Pb, Cr II!
As, Se, NO , F, Cr VI
O
MOST PROBABLE
APPLICATION FOR
INORGANIC REMOVAL
REMOVAL OF Ba OR Ra
FROM GROUNDWATERS
REMOVAL OF NO3
FROM GROUNDWATERS
HIGH - GREATER THAN 80 PERCENT MODERATE - 20 TO 80 PERCENT LOW - LESS THAN 20 PERCENT
-------
TREATMENT PROCESS
REVERSE OSMOSIS AND ELECTRODIALYSIS
PRINCIPAL APPLICATION
FOR WATER TREATMENT
DESALTING OF SEA WATER OR BRACKISH GROUNDWATER
HIGH
INORGANIC CONTAMINANT
TREATMENT CAPABILITY
EFFECTIVENESS*
MODERATE
LOW
As V, Ba, Cr, Pb, Cd, Se, Ag, F, Ra, Hg
NO , As III
3
MOST PROBABLE
APPLICATION FOR
INORGANIC REMOVAL
REMOVAL OF ALL INORGANICS FROM GROUNDWATERS
HIGH - GREATER THAN 80 PERCENT MODERATE - 20 TO 80 PERCENT LOW - LESS THAN 20 PERCENT
-------
TREATMENT PROCESS
POWDERED ACTIVATED CARBON
GRANULAR ACTIVATED
CARBON
PRINCIPAL APPLICATION
FOR WATER TREATMENT
REMOVAL OF TASTE AND ODORS
FROM SURFACE WATERS
REMOVAL OF TASTE,
ODORS, AND ORGANICS
HIGH
INORGANIC CONTAMINANT
TREATMENT CAPABILITY
EFFECTIVENESS*
MODERATE
LOW
, Hg(0),
, Hg(0), Cd
Cd
Ba, Ra, Cr III, F, NO , Ag
3
Ba, Ra, Cr 111, F, NO
MOST PROBABLE
APPLICATION FOR
INORGANIC REMOVAL
REMOVAL OF Hg FROM SURFACE
WATERS DURING EMERGENCY
SPILLS
REMOVAL OF Hg FROM
SURFACE OR GROUND-
WATERS
HIGH - GREATER THAN 80 PERCENT MODERATE - 20 TO 80 PERCENT LOW - LESS THAN 20 PERCENT
-------
TREATMENT PROCESS
ACTIVATED ALUMINA
PRINCIPAL APPLICATION
FOR WATER TREATMENT
REMOVAL OF FLOURIDE FROM GROUNDWATERS
HIGH
INORGANIC CONTAMINANT
TREATMENT CAPABILITY
EFFECTIVENESS*
MODERATE
LOW
F, As, Se
Ba, Ra, Cd
MOST PROBABLE
APPLICATION FOR
INORGANIC REMOVAL
REMOVAL OF F, As, OR Se FROM GROUNDWATERS
HIGH - GREATER THAN 80 PERCENT MODERATE - 20 TO 80 PERCENT LOW - LESS THAN 20 PERCENT
-------
PERFORMANCE SUMMARY FOR INORGANIC TECHNOLOGIES EXAMINED
Inorganic
Compound
Reverse
Osmosis
Removal Efficiency
Ion
Exchange
Aeration
Carbon
Adsorption
Nitrate
Radium
Uranium
Radon
-H-
-H-
•H-
•H-
+f= Excellent 70%-100%
% = Research Being Conducted By DWRD
-------
MOST PROBABLE APPLICATIONS
OF
WATER TREATMENT PROCESSES
FOR
INORGANIC CONTAMINANT REMOVAL
-------
Generally
effective tor
Alternatives for reducing radionuclide concentrations in
drinking water (all alternatives can include blending)
(Aieta, et al. 1987)
-------
.100 r
O)
cr
(D
.80
§5
E i?-60
Q: *o
D
,40
c
o
*4—
O
a
.20
) 10 20 30 40 50 60
Radium Level - Raw Water (pCi/L)
REQUIRED REMOVAL FRACTION FOR
RADIUM
-------
1.00
c
o
•M
O
1/1
(/I
d)
rci
-•->
O
16.0 20.0
30.0 40.0
Raw Water Radium pCi/1
50.0
Total hardness removal fraction as a function of raw water radium content
required to meet limit of 5.0 pCi/L finished water. (Singley, et al. 1977)
-------
1.00
1.80
u
o
§ .60
E
o>
rr
V)
X = West DesMoines, la
+ = Webster City, la W/0 Soda Ash
©= Webster City, la W Soda Ash
El55 Peru, i| (3dates)
A= Elgin, II (3 dates)
•X- = Englewood, Fl
0= Venice, Fl
.40
o
X
.20
.20 .40 .60
Radium Removal Fraction
.80
1.00
LIME SODA PROCESS, TOTAL HARDNESS
REMOVAL FRACTION VS RADIUM REMOVAL
FRACTION.
-------
Raw Water
Q
f, + fg- 1
f2Q
f,Q
Bypass
Water
Ion
Exchange
Unit
Q
Finished
Water
MASS BALANCE FOR DETERMINING FRACTION
OF RAW WATER TO BE TREATED
-------
1.00
c
o
'•5
D
-80
E
-------
WASH-WATER
COLLECTOR
OUTLET
REGENERANT TANK
DIAGRAM OF TYPICAL ION EXCHANGE UNIT
-------
1.00
c
o
o
o
.80
o
E
0)
a:
E .60
3
T3
O
o:
.40
.20
I
I
I
I
.40 .60 .80 1.00
Total Hardness Removal Fraction
RADIUM REMOVAL FRACTION VS. TOTAL HARDNESS REMOVAL
FRACTION IN ION EXCHANGE PLANTS (BEFORE BLENDING)
-------
you can't
SEE THEM
SMELLTHEM
or
TASTE THEM
and they can
Guess Who Has Started-Worryin
About Qualit -
*** «-W^ *ts«ra£
! a Ugt memoes. l£l£S%*fa ««,« trthalo.
^^.-^.fiSLaarEsjf*
• . »• Kunla also 'm^^. .. ' .
'?»'•• l-l pSdlitS*-?* ^thy Wr'
level of PracaSSrevHi',,?'' ta»*« « WE« metha
f""' I" be loldlKSJ^f' ^ "^ «™ tectf0'
ie«' <" ^^'ft'^er1'6 ' "i* "Dr.
5Wa5S«2SS Stft^s5te»g
V^SfftSS^r^ asaK^^jSrssaS
v - -*. «.fc«uus 'ft?iS53?s^aS
% ._" ' , i j/^t«..t_ ... "flflS*
Known1 • ' wulr
•»' • ' 2VI^i*«^^Sfl?*««i5
55«&a &;to <5top ^«"«
iWs. Conducted S _.S« Dr. Kunfa «»,„.„-,
-------
you can't
SEE THEM
SMELL THEM
or
TASTE THEM
and they can
KILL YOU!
The TRIHALOMETHANES (THM's) found In drink-
Ing water are organohalogen compounds that are named
derivative! of methane, formod when thrao of tha four hydro-
gen atomi have been replaced by three atom* of chlorine,
bromine or Iodine. Ten distinct compounds can be form-
ed from the possible combinations of three hydrogenated
atoms, one hydrogen atom, and one carbon atom. Current
analytical methodology applied to drinking water has thus far
led to the detection of chloroform (trlchloromethane),
bromodichloromethane, dlbrombchloromathane, bromoform
(trlbromomethane), and dlchlorolodiomathane. Monitoring
methods are currently available for the bromlneted and
chlorinated THM's, but not for the iodinated types because
of their chemical Instability. EPA has promulgated an Interim
standard of 0.1 mg/l (lOOppb) for THM's.
-------
Would You drink this?
\ ' •
,; CHLOROFORM
MAGNESIUM o o Q LIMESTONE
° .1 .'•''•' 0° COUFORM
IRON o ,0
•'• '"•' ' (
Q O O o
CHLORINE
„ » o O O.
ASBESTOS
But You maybe! v !
ft seems unbelievable, but it's true! EPA and GAO. Federal studies have
shown that most city water treatment plants do not effectively remove cheinK
^ cal and mineral contaminates from your water... the water you drink ...cook
[with... bathe with.,. launder with. The falnSoft Classic Apollo Phase Four
and treats every
want you to drlnkonly clean, sparkling, healthy.mbneysavingwater.
684-9411
PALMBEACHCOUNW
What water was meant to be.
-------
Guess Who Has Started 'Worrying
> * i ' > ''•••'••.' *—•* ••
About Quality of California Water
;>• By SCOT J. PALTROW , son" for Evian, concedes in a telephone in-
1 s«n//Reporter'o/THKWxij.STiucBT JOURNAL ' tervlpw that Beverly IlilJs water contains.
BEVERLY CHILLS, Calif.-Ttys wealthy fewer, than TOO parts-per million,of trihalo-
city has long prided itself on having'a high methanes, the level the Environmental Pro-
level of practically everything. And now it's tection ASen«y "as established as>safe.
about to be told that it may.also have a high Dr. KuiUn also concedes there hasn't
level of carcinogens in its water. ' been^any tafally conclusive proof of a. link
A campaign is underway to convince res- 'between:-the chemical and cancer. But he
idents that their water contains a suspected W» p"k£*J* °f on^u^,town, whose
carcinogen. But the campaign's booster isn't wate,r JJ,ad. levels of the chemical •signifl-
,a group of environmentalists .or concerned <*"«* higher,than 100 parts per million ex-
citizens It is Evian Waters of France Inc.- Perienced except onaljjr high levels of blad-
vhich is about to introduce its bottled Jer and Intestinal tract cancer. "Too little Is
.Tench Alps spring water to Southern Call- known- "^ he adds-
irnia. . • . /City's Advice
' Evian says it will hold "a news conference „ The Evian test/purportediy shows that
.today disclosing that "an independent lab Beverly HiHs has a higher level of the chem-
analysis" has found in Beverly Hills water ^ «£» °&er communities to the area. But
significant 'amounts of trlhalomethanes. Mel ™om< Beuver'y ffln» dl™*°* of ;P^»c |
which it says 'are "suspected carcinogens." f^ce- W» &<* W and ™f cpmmunitieir;
Evian says Dr. Samuel A. Kunin. a Los An- in &* region get. their water from the same
geles-area urologist and.surgeon, will pres- source,.and Its chemical content should be
ent and interpret the laboratory results. ^^ H« savs•$* water meets all testing
requirements, and he wouldn t advise resi-
"Too Little Is Known' dents "to 'stop drinking It with their
But Evlan's press conference invitation Scotch."
doesn't say that Evian itself commissioned But Dr. Kuntn says levels of the trihalo-
and paid for the analysis, -conducted by methanes can vary-and that Evian's water
Montgomery Laboratories of Pasadena. contains lower levels of the chemical than
Dr. Kunln, a paid "medical spokesper- several competing bottled waters.
-------
(Eh=0.875)
1.2
O
o
lO
CM
H
Q.
0)
10
O
UJ
II
X
ro
cvi
UJ
I!
1.0
0.8
0.6
0.4
0.2
LJ
(Eh=-0.356) 0
WATER OXIDIZED 02
H,0
Pb(H20)
Pb
WATER REDUCED H20
K
8
pH
(Eh=0.64)
(Eh«-0.59)
10
MODIFIED POURBAIX DIAGRAM FOR
LEAD IN WATER
-------
200
Solubility of Lead as a Function of pH
and Carbonate. (Adapted from AWWARF/EBI,
1985)
-------
1000
7
PH
Solubility of Lead in the Presence of
50mg/L (as CaC03) DIG. (Adapted from
Schock (AWWARF/EBIJ985).
-------
ORDER OF ACTIVITY
OF COMMON METALS
MAGNESIUM
ZINC
ALUMINUM
MILD STEEL
WROUGHT IRON
CAST IRON
LEAD
TIM MORE
TIN ACTIVE
BRASS
COPPER
STAINLESS STEEL
-------
2400
GO
300
LEAD
COLD WATER-12/13/82
5 ERNEST COURT
UNDER CONSTRUCTION
LEAD IN WATER
50>ig/L DRINKING
WATER STANDARD
i i i i i i \
0 10 20 30 40 50 60 70 80
TIME-SECONDS
-------
200 H
LEAD
KITCHEN SINK
OFFICE BUILDING-MELVILLE
PLUMBING-23 MONTHS OLD
f*-c.\jpi±f u.
1234
TIME (Minutes)
-------
0
USEPA/SHWD LEAD SOLDER STUDY
CADMIUM (pg/L)
TEST SITE (32W18)
20
40 60 80
Time-(Seconds)
100
120
-------
LOCATION
f
(
DRAW
NASSAU COUNTY HEALTH DEPT. TEST
LEAD
PH
HARDNESS
mg/L
% LEAD
SOLDER
AGE
LOCUST VALLEY
17000
7.0
41
60.3
PORT WASHINGTON
4400
6.8
49
61.7
MANORHAVEN
3500
6.8
49
47.9
WOODBURY
2900
7.3
64
58.4
NORTH PORT WASH.
930
7.0
49
50.3
NORTH HILL
750
6.7
23
56.2
NORTH HILL
530
6.7
23
60.0
-------
MINNESOTA
DEPT. OF HEALTH
46 COMMUNITIES
LESS THAN 3 YEARS OLD
COMMUNITY
Winona
Randolph
Mankato
New Trier
New Germany
Prinsburg
Red Wing
Madelia
Norwood
Watertown
Minnetonka Beach
Barnum
Waconia
Young America
Spring Lake Park
Fridley
Becker
first draw lead
ug/L
1500
860
790
300
270
260
200
140
130
120
110
95
80
63
61
57
54
25 COMMUNITIES
NEW CONSTRUCTION
COMMUNITY
Mankato
Kinckley
Rice
Park Rapids
Northfild
Eveleth
Winsted
St Peter
Brainerd
Waconia
Rushford
Willmar
Zumbrota
Elk River
New Ulm
Backus
Rochester
first draw lea
ug/L
33000
15000
8800
4600
4100
3800
2400
1900
1100
1100
640
450
400
350
310
290
150
VS 50 ug/L DRINKING WATER STANDARD
-------
PERCENTAGE OF TEST SITES WITH LEAD IN DRINKING WATER
GREATER THAN 20 ug/L AT LOW pH (6.4 & LESS)
AGE OF
TEST SITE
(Years)
0-1
1 -2
2-3
3-4
4-5
6-7
9-10
15-16
20 a Older
FIRST
DRAW
100%
100%
86%
100%
86%
78%
71 %
57%
86%
10
SEC
100%
, 71 %
86%
86%
57%
44%
29%
14%
27%
20
SEC
100%
86%
57%
100%
29%
33%
14%
14%
29%
30
SEC
100%
57%
57%
71%
43%
33%
14%
14%
0%
45
SEC
100%
57%
43%
71 %
43%
11 %
14%
14%
14%
60
SEC
86%
29%
43%
71%
43%
11 %
14%
14%
0%
90
SEC
86%
43%
43%
29%
14%
11%
0%
14%
14%
120
SEC
88%
14%
29%
29%
0%
0%
0%
14%
0%
-------
PERCENTAGE OF TEST SITES WITH LEAD IN DRINKING WATER
GREATER THAN 20 ug/L AT MEDIUM pH (7.0-7.4)
AGE OF
TEST SITE
(Years)
0-1
1 -2
2-3
3-4
4-5
6-7
9-10
15-16
20 a Older
FIRST
DRAW
100%
80%
40%
5O%
30%
10%
20%
40%
20 %
10
SEC
90%
60%
20%
20%
10%
0%
0%
20%
0%
20
SEC
90%
30%
10%
20%
10%
0%
0%
20%
0%
30
SEC
60%
10%
10%
30%
0%
0%
0%
10%
0%
45
SEC
30%
20%
10%
20%
10%
0%
0%
0%
10%
60
SEC
20%
0%
0%
30%
0%
0%
0%
0%
0%
90
SEC
10%
10%
0%
30%
0%
0%
0%
0%
0%
120
SEC
10%
0%
0%
20%
0%
£>%
0%
0%
0%
-------
PERCENTAGE OF TEST SITES WITH LEAD IN DRINKING WATER
GREATER THAN 20 ug/L AT HIGH pH (8.0 a GREATER)
AGE OF
TEST SITE
(Years)
0-1
1 -2
2-3
3-4
4-5
6-7
9-10
15-16
20 a Older
FIRST
DRAW
100%
67%
30%
25%
30%
20%
10%
33%
20%
10
SEC
100%
22%
10%
0%
10%
0%
0%
22%
0%
20
SEC
60%
11%
10%
0%
0%
0%
10%
11%
0%
30
SEC
10%
11%
0%
0%
0%
0%
0%
11%
0%
45
SEC
20%
11%
0%
0%
0%
0%
0%
0%
0%
60
SEC
10%
0%
0%
0%
0%
0%
10%
0%
0%
90
SEC
20%
11%
0%
0%
0%
0%
0%
0%
0%
120
SEC
0%
0%
0%
13%
0%
0%
10%
0%
0%
-------
RESULTS FROM HOUSEHOLD LEAD STUDY FIRST DRAW SAMPLES
pH % of Homes Exceeding 50ug/L
6.2 58.7
6.2-8.2 26.6
78.2 24.4
-------
SOLDER TEMPERATURES
NORMAL COMPOSITION %
TIN LEAD ANTIMONY SILVER
95 — 5 —
95 — — 5
30 70 — —
35 65 — —
40 60 — —
45 55 — —
50 50 — —
60 40 — —
MELTING
SOLIDUS°F
452
430
361
361
361
361
361
361
RANGES
LIQUIDUS°F
464
473
491
471
460
441
421
374
-------
SOLDER
Common Usage
Type
*April 1984
Percentage Liquid Temp. Range Solid Temp. Wholesale Cost
Tin/Lead
50/50
421° F 60°F 361° F
$4.5071 b.
Alternates
Tin/Antimony 95/5
Tin/Silver
95/5
464° F
12° F 452° F
$8.00/lb.
473° F 43° F 430° F $20,00/lb.
ASTM Standards - Part 8-B32
*estimated by plumbers at less than 1 pound SOLDER required per residence
-------
US EPA/SEATTLE STUDY
Tin Antimony Solder
UNIVERSITY OF WASHINGTON BUILDINGS
BUILDING
YEAR
OCCUPIED
SAMPLE
ANTIMONY
OCEANOGRAPHY
TEACHING
1969
MECH. ROOM
STANDING
<0.6
<2.0
SOUTH CAMPUS
CENTER
1969
MECH. ROOM
STANDING
<0.6
<0.6
HEALTH SCIENCES
WING T
1973
MECH. ROOM
STANDING
<0.6
<0.6
CONDON
1974
MECH. ROOM
STANDING
<0.6
<0.6
WATER FRONT
ACTIVITIES
1977
MECH. ROOM
STANDING
<0.6
<0.6
HALL HEALTH
CENTER
1978
MECH. ROOM
STANDING
<0.6
<0.6
LOEW
1979
MECH. ROOM
STANDING
<0.6
<0.6
ENGINEERING
LIBRARY
1969
MECH. ROOM
STANDING
<0.6
<0.6
-------
COPPER LEACHING IN FOUR PIPE LOOPS
pH SOLDER COPPER
(mg/L)
5.5 TIN/SILVER 3.80
5.5 TIN/ANTIMONY 4.28
5.5 TIN/COPPER 4.50
5.5 TIN/LEAD 4.28
-------
• AGE OF THE PLUMBING
CORROSIVE WATERS
GALVANIC CORROSION
FLUX USED
-------
CORROSION CONTROL
pH ADJUTMENT ALONE
ADJUST ALKALINITY AND pH
STABILIZE WATER (CALCIUM CARBONATE BALANCE)
ADD CORROSION INHIBITORS
-------
SEATTLE CORROSION CONTROL PROGRAM
a SERVICE LIFE ECONOMICS
oLOSS $7.8 MILLION/YEAR
o AESTHETICS
o RED a BLUE STAINS
o METALLIC TASTE COMPLAINTS
a HEALTH
oLEAD ABOVE MCL
(Courchene, J.E. & Hoyt, B.R "Benefits of Corrosion Treatment
in Seattle" CA/NV AWWA Seminar, 1985)
-------
SEATTLE CORROSION CONTROL PROGRAM
INSTITUTED MATERIAL SELECTION PROGRAM
n PL AST 1C PIPE
o ALLOWED PB AND CPVC IN BOTH
HOT AND COLD
o REQUIRED DIELECTRIC INSULATORS
o ENCOURAGED TYPE K AND L COPPER
PIPE
o DISCOURAGED GALVANIZED PIPE
o < 0.2% LEAD SN SOLDERS
(Courchene, J.E. 8 Hoyt, B.P "Benefits of Corrosion Treatment
in Seattle" CA/NV AWWA Seminar, 1985)
-------
SEATTLE CORROSION CONTROL PROGRAM
WATER TREATMENT PROGRAM
TOLT
2mg/L CaO
9mg/L Na2C03
CEDER
2 mg/L Ca 0
(Courchene, J.E. 8 Hoyt, B.R "Benefits of Corrosion Treatment
in Seattle" CA/NV AWWA Seminar, 1985)
-------
SEATTLE CORROSION CONTROL PROGRAM
RWQM* OVERNIGHT STANDING SAMPLES,
ALL CEDAR AND TOLT GROUPS
Parameter
Pb,pg/L
Cd,jjg/L
Cu, mg/L
Fe, mg/L
Zn, mg/L
Average Concentration
Before
10.6
0.76
0.28
1.21
0.60
After
4.0
0.41
0.09
0.99
0.33
^Reduction
63
46
68
18
45
%>MCL
Before
5
0
1 1
75
0
After
0.4
0
0
67
0
# Residential Water Quality Monitoring-300 residences
(Courchene, J.E. B Hoyt, B.R "Benefits of Corrosion Treatment
in Seattle" CA/NV AWWA Seminar, 1985)
-------
SEATTLE CORROSION CONTROL PROGRAM
LEAD LEVELS
% BEFORE
% AFTER
SOURCE
>20
>50'
>20'
>50:
TOLT
12.5
5.9
3.9
0.6
CEDAR
9.8
4.3
0
* Jjg/L"
(Courchene, J.E. 8 Hoyt, B.R "Benefits of Corrosion Treatment
in Seattle" CA/NV AWWA Seminar, 1985)
-------
SEATTLE CORROSION CONTROL PROGRAM
WATER QUALITY CHANGES SINCE TREATMENT
Water Characteristic
pH
Tot. Alkalinity, mg/L CaC03
Carbon Dioxide, mg/L
Hardness, mg/L CaC03
Sodium, mg/L
Tot. Diss. Solids, mg/L
Spec. Conductance, pmhos
Larson's Ratio
Cedar
Before
7.2
16
2.7
21
1.8
46
55
0.65
After
8.2
19
0
24
1.8
48
59
0.55
Tolt
Before
6.0
2.5
6.0
8
1.0
25
27
3.9
After
8.2
13.5
0
11
4.8
36
42
0.8
(Courchene, J.E. 8t Hoyt, B.R "Benefits of Corrosion Treatment
TA/MW AWWA
n
-------
INLET
OUTLET-
I u u.
r
t_
u
il
u
u
II
li
n
U
II
II
!S!
l^f M
T ~i
u i
u i
11 i
11
ti i
n i
u l
!! i
IIOUT,
II ff^ 1
J
PLAN VIEW
INLET/OUTLET ELEVATION
SECTION
LIMESTONE CONTACTOR
------- |