EPA/600/R-14/236 | September 2014 | www.epa.gov/research
United States
Environmental Protection
Agency
   Removal of Fluoride from
   Drinking Water Supplies by
   Activated Alumina
 Office of Research and Development

-------
                                                  EPA/600/R-14/236
                                                       March 2014
                  Design Manual

Removal of Fluoride from Drinking Water Supplies
               by Activated Alumina
                    Prepared by

                  ALSA Tech, LLC
              North Bethesda, MD 20852

                        and

                      Battelle
              Columbus, OH 43281-2693
         Under EPA Contract No. EP-C-11-038
                        for
                   Thomas J. Sorg
                Task Order Manager

       Water Supply and Water Resources Division
     National Risk Management Research Laboratory
          Office of Research and Development
         U.S. Environmental Protection Agency
                Cincinnati, Ohio 45268

-------
                                       DISCLAIMER
The information in this document has been funded by the United States Environmental Protection Agency
(U.S. EPA) under Task Order (TO) No. 0012 of Contract No. EP-C-11-038 to Battelle.  It has been
subjected to the Agency's peer and administrative reviews and has been approved for publication as an
EPA document.  Mention of trade names or commercial products does not constitute an endorsement or
recommendation for use.

-------
                                         ABSTRACT
This document is an updated version of the Design Manual: Removal of Fluoride from Drinking Water
Supplies by Activated Alumina (Rubel, 1984). The manual is an in-depth presentation of the steps
required to design and operate a fluoride removal plant using activated alumina (AA), which is a reliable
and cost-effective process for treating excess fluoride from drinking water supplies.

For the design of an AA system, a list of critical items is provided as follows:

       •   Perform raw water analysis.
       •   Determine required/optimum system flowrate and utilization rate, maximum daily demands,
           as well as redundancy requirement, if any.
       •   Determine system configuration.
       •   Determine number of adsorption vessels and vessel dimensions (use an empty bed contact
           time [EBCT] of at least 5 min and a bed depth of 3 to 6 ft).
       •   Determine acid storage and feed subsystem requirements.
       •   Determine wastewater disposal needs.

For the startup and operation of an AA system, a list of critical items is provided as follows:

System Startup
       •   Half-fill adsorption vessels with water prior to placing underbed material and AA media.
       •   Thoroughly backwash vessels after leach lift of media loading.
       •   Measure free board in each vessel after placement of underbed material and media.
       •   Calibrate in-line pH meters.

System Operation - Treatment Mode
       •   Place one vessel in operation at a time.
       •   Adj ust pH of influent water to pH 5.5 to 6.0.
       •   Raise pH of effluent water to no higher than 8.5.
       •   Extend a treatment run by blending.
       •   Take vessels offline for regeneration at a set breakthrough level; use staggered regeneration
           to minimize system downtime.

System Operation - Backwash Mode
       •   Drain bed for most effective backwash.
       •   Watch bed expansion and adjust backwash rate to avoid backwashing media out of vessels.
       •   Avoid excessive backwash.

System Operation - Regeneration Mode
       •   Use 1% NaOH at pH 13.0 for media regeneration.
       •   Adopt a one-step (upflow or downflow) or a two-step (upflow and then downflow) process
           for regeneration.
       •   Volume of caustic solution required per regeneration step is approximately 30 gal/ft3 of
           media.
       •   Regeneration flowrate is 2.5 to 3.0 gpm/ft2 and regeneration time is 50 to 60 min per step.

System Operation - Neutralization mode
                                              in

-------
       •   Use sulfuric acid for media neutralization; pH of acidic water should be no lower than 2.5.
       •   Measure free board prior to replacing regenerated vessel into service; replenish media if
           needed.

A number of other treatment technologies are also available for removing fluoride from drinking water to
levels below the maximum contaminant level (MCL) of 4 mg/L. These technologies are briefly reviewed
in Section 2 of this design manual.
                                               IV

-------
                                         FOREWORD
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the Nation's
land, air, and water resources. Under a mandate of national environmental laws, the Agency strives to
formulate and implement actions leading to a compatible balance between human activities and the ability
of natural systems to support and nurture life. To meet this mandate, EPA's research program is providing
data and technical support for solving environmental problems today and building a science knowledge
base necessary to manage our ecological resources wisely, understand how pollutants affect our health,
and prevent or reduce environmental risks in the future.

The National Risk Management Research Laboratory (NRMRL) is the Agency's center for investigation
of technological and management approaches for preventing and reducing risks from pollution that
threaten human health and the environment. The focus of the Laboratory's research program is on
methods and their cost-effectiveness for prevention and control of pollution to air, land, water, and
subsurface resources; protection of water quality in public water systems; remediation of contaminated
sites, sediments, and ground water; prevention and control of indoor air pollution; and restoration of
ecosystems. NRMRL collaborates  with both public and private sector partners to foster technologies that
reduce the cost of compliance and to anticipate emerging problems. NRMRL's research provides
solutions to environmental problems by: developing and promoting technologies that protect and improve
the environment; advancing scientific and engineering information to support regulatory and policy
decisions; and providing the technical support and information transfer to ensure implementation of
environmental regulations and strategies at the national, state, and community levels.

                                                                  Cynthia Sonich-Mullin, Director
                                                    National Risk Management Research Laboratory

-------
                                  ACKNOWLEDGMENTS
This manual has been revised from its 1984 version written by Frederick Rubel of Rubel and Hager, Inc.
This revision also incorporated materials from two other design manuals for arsenic removal using
adsorptive media (EPA/600/R-03/019) and iron removal processes (EPA/600/R-06/030).  This manual
was prepared with input from Thomas J. Sorg of U.S. EPA, NRMRL. The manual was reviewed by the
following people and their suggestions and comments were of valuable assistance in preparation of the
final document:

Rich Dennis, Severn Trent Services
Cindy Klevens, New Hampshire Department of Environmental Services
Ray Kolisz, Twentynine Palms Water District
Bill Reid, Axens North America, Inc.
Mike Wright, Twentynine Palms Water District (Retired)
                                             VI

-------
                                       CONTENTS

DISCLAIMER 	ii
ABSTRACT   	iii
ACKNOWLEDGMENTS	vi
APPENDICES 	ix
FIGURES     	ix
TABLES      	ix
ACRONYMS AND ABBREVIATIONS	x

1.0: INTRODUCTION	1
    1.1  Purpose and Scope	1
    1.2  Fluoride Regulation	1
    1.3  Fluoride in Water Supplies	2
    1.4  Available Treatment Technologies	2

2.0: TREATMENT METHODS FOR FLUORIDE REMOVAL	3
    2.1  Introduction	3
    2.2  Non-Treatment Options	3
    2.3  Centralized Treatment Options	4
         2.3.1 Adsorption/Ligand Exchange Processes	4
              2.3.1.1 Granular Activated Alumina	4
              2.3.1.2 Bone, Bone Char, and Apatite-Based Materials	7
              2.3.1.3 Ion Exchange Resins	8
              2.3.1.4 Other Adsorbents	8
         2.3.2 Membrane Processes	9
              2.3.2.1 Reverse Osmosis and Nanofiltration	9
              2.3.2.2 Electrodialysis and Electrodialysis Reversal	11
         2.3.3 Other Centralized Treatment Methods	11
              2.3.3.1 Conventional Coagulation/Filtration	11
              2.3.3.2 Lime Softening	12
              2.3.3.3 NalgondaTechnique	12
         2.3.4 Point of Use/Decentralized Treatment Technology	12

3.0: DESIGN OF CENTRALIZED ACTIVATED ALUMINA PLANT	14
    3.1  General Plan	14
         3.1.1 General Considerations	14
         3.1.2 Conceptual Design	18
    3.2  Preliminary Design	20
         3.2.1 Basis of Design	20
              3.2.
              3.2.
              3.2.
              3.2.
              3.2.
              3.2.
              3.2.
              3.2.
.1  General	20
.2  Project Scope	21
.3  Process Design Data Summary	21
.4  Site	21
.5  Layout of Structure	21
.6  Structural	22
.7  Mechanics	23
.8  Electrical	23
         3.2.2 Treatment Equipment	24
              3.2.2.1  Manual or Automatic Operation	24
                                            vn

-------
              3.2.2.2  Treatment Equipment Preliminary Design	27
         3.2.3 Preliminary Treatment Equipment Arrangement	32
         3.2.4 Preliminary Cost Estimate	33
         3.2.5 Preliminary Design Revisions	33
     3.3  Final Design	33
         3.3.1 Treatment Equipment Final Design	34
              3.3.
              3.3.
              3.3.
              3.3.
              3.3.
              3.3.
.1  Treatment Bed and Vessel Design	34
.2  Pipe Design	36
.3  Instrument Design	37
.4  Acid Storage and Feed Subsystem	37
.5  Caustic Soda Storage and Feed System	38
.6  Regeneration Wastewater Surge Tank	38
         3.3.2 Final Drawings	38
         3.3.3 Final Capital Cost Estimate	38
         3.3.4 Final Design Revisions	38

4.0: ACTIVATED ALUMINA PLANT OPERATION	40
    4.1  Introduction	40
    4.2  Initial System Startup	40
    4.3  System Operations	44
         4.3.1 Treatment Mode	44
         4.3.2 Backwash Mode	46
         4.3.3 Regeneration Mode	47
         4.3.4 Neutralization Mode	48
    4.4  Operator Requirements	48
    4.5  Laboratory Requirements	49
    4.6  Operating Records	49
         4.6.1 Plant Log	50
         4.6.2 Operation Log	50
         4.6.3 Water Analysis Reports	50
         4.6.4 Plant Operating Cost Records	50
         4.6.5 Correspondence Files	50
         4.6.6 Regulatory Agency Reports	50
         4.6.7 Miscellaneous  Forms	50
    4.7  Treatment Plant Maintenance	50
    4.8  Equipment Maintenance	51
    4.9  Treatment Media Maintenance	51
    4.10 Treatment Chemicals Supply	51
    4.11 Housekeeping	51

5.0: CENTRAL TREATMENT PLANT CAPITAL AND OPERATING COSTS	54
    5.1  Introduction	54
    5.2  Capital Costs	55
         5.2.1 Discussion of Cost Variables	55
              5.2.
              5.2.
              5.2.
              5.2.
              5.2.
              5.2.
              5.2.
.1  Water Chemistry	55
.2  Chemical Supply Logistics	55
.3  Manual Versus Automatic Operation	56
.4  Backwash and Regeneration Disposal Concept	56
.5  Ambient Conditions	56
.6  Existing and Planned (Future) Water System Parameters	57
.7  Financial  Considerations	58
                                             vin

-------
         5.2.2 Capital Costs of Central Fluoride Removal Plants Based on Flowrate	58
     5.3  Operating Costs	59
         5.3.1 Discussion of Cost Variables	59
              5.3.1.1 Treatment Chemical Costs	59
              5.3.1.2 Operating Labor Costs	63
              5.3.1.3 Utility Costs	64
              5.3.1.4 Media Replacement Costs	65
              5.3.1.5 Replacement Parts and Miscellaneous Material Costs	66
         5.3.2 Operating Cost Summary	66

6.0:  REFERENCES	67
                                       APPENDICES

Appendix A: Summary of Subsystems Including Components
Appendix B: Treatment System Design Example
Appendix C: Discussion of Acid Consumption Requirements for pH Adjustment of
            Raw Water
Appendix D: Activated Alumina Plant Search and Visits
Appendix E: English to Metric Conversion Table
                                         FIGURES

Figure 3-1.  Project Development Process	17
Figure 3-2.  An Example Parallel Flow Diagram	19
Figure 3-3.  Treatment System Plan for an Activated Alumina Plant	25
Figure 3-4.  Treatment Bed and Vessel Design Calculations	Error! Bookmark not defined.
Figure 3-5.  Treatment Vessels Piping Isometric	39
Figure 4-1.  An Example Valve Number Diagram	41
Figure 4-2.  Example Operating Mode Flow Schematics	45
Figure 4-3.  Fluoride Removal Water Treatment Plant Operation Log	52
Figure 4-4.  5,000-gal Chemical Storage Tank - Liquid Volume	53
Figure 5-1.  Curve Illustration Rule of Thumb for Volume of Water Treated per
            Cycle vs. Raw Water Fluoride Level	61
                                          TABLES

Table 2-1. AA System Run Length Results	6
Table 2-2. Fluoride Removal Efficiencies (Rejection Rates) by RO	10
Table 3-1. Example of Fluoride Removal Plant Water Analysis Report	16
Table 4-1. An Example Valve Operation Chart for Treatment Vessels in Media
            Regeneration Modes	42
Table 4-2. Calculated Activated Alumina (-28, +48 Mesh) Downflow Pressure Drop
            Data	43
Table 4-3. Example Two-Step Process Conditions for Regeneration of an Activated
            Alumina Treatment System	49
Table 5-1. Price for Typical -28, +48 Mesh Activated Alumina	65
Table 5-2. Operating Cost Breakdowns for an Activated Alumina Plant	66

-------
                           ACRONYMS AND ABBREVIATIONS
AA        activated alumina
ADA      Americans with Disabilities Act
AIX       anionic exchange
ASME     American Society of Mechanical Engineers
AWWA    American Water Work Association

BTGA     best technologies generally available
BV        bed volumes

CDC      Centers for Disease Control and Prevention
CO2       carbon dioxide
CPVC     chlorinated polyvinyl chloride
CWS      community water system

EBCT     empty bead contact time
ED        electrodialysis
EDR      electrodialysis reversal
EPA       United States Environmental Protection Agency
EPDM     ethylene propylene diene monomer
EPTDS     entry point to the distribution system

FRB       fiberglass and fiber re-enforced plastic
FRP       fiber-reinforced plastic

GAC      granular activated carbon
gpd        gallon per day
gpm       gallon per minute

HOPE     high density polyethylene
HHS      Health and Human Services
HVAC     heating, ventilating, and air conditioning

1C         ion chromatography
IX         ion exchange

MCL      maximum contaminant level
MCLG     maximum contaminant level goal
MGD      millions of gallons per day

NEC      National Electrical Code
NF        nanofiltration
NRC      National Research Council

O&M      operating and maintenance
OSHA     Occupational Safety and Health Administration

PAC       powder activated carbon
PHS       U.S. Public Health Service

-------
PLC       programmable logic controller
POE       point-of-entry
POTW     publicly-owned treatment works
POU       point-of-use
PVC       polyvinyl chloride

RO        reverse osmosis

SDWA     Safe Drinking Water Act
SMCL     secondary MCL
SSCT      small system compliance technologies

TCLP      Toxicity Characteristic Leaching Procedure
TDS       total dissolved solids
TOC       total organic carbon

WET      Waste Extraction Test

ZPC       zero point of charge
                                             XI

-------

-------
                                    1.0:  INTRODUCTION
1.1        Purpose and Scope

The United States Environmental Protection Agency (EPA) published a fluoride design manual in 1984 to
present the design, operation and cost information of central treatment plants for fluoride removal from
water supplies using activated alumina (AA) (Rubel, 1984). While the manual continues to be a very
useful document to small and large water systems containing high fluoride, improvements in the design
and operation of the AA process have been made during the past 30 years.  The capital and operating
costs of AA systems have also changed significantly since then.

The information provided in the original and this revised design manual applies primarily to central
treatment plants with capacities ranging from 30,000 to 1,000,000 gallon per day (gpd). However, the
treatment information, for the most part, can be adapted to both larger and smaller systems.  For very
small systems with capacities of less than 30,000 gpd (-20 gal/min [gpm]), less expensive tanks such as
fiber-reinforced plastic (FRP) canisters and valves likely would be used.

Several other treatment methods have also been used for removal of excess fluoride, but none with the
cost-effectiveness and process efficiency of the AA method. AA and these alternative treatment methods
and their limitations are reviewed in Section 2 of this revised design manual.

When excess fluoride is present in combination with quantities of other organic and/or inorganic contami-
nants, the AA process may not be the most effective method for fluoride removal. Those water supplies
should be evaluated on a case-by-case basis for selection of the appropriate treatment method, or combi-
nation of methods for contaminant removal.

For the most part, this manual is prepared for engineering firms that design AA systems, but certainly can
be used by others. For the very small systems, they would likely just ask for proposals from system
suppliers without the use of an engineering design firm.

1.2        Fluoride Regulation

The Safe Drinking Water Act (SOWA) of 1974 mandated that EPA identify and regulate drinking water
contaminants that may have an adverse human health effect and that are known or anticipated to occur in
public water supply systems.  In 1975, EPA proposed an interim primary drinking water regulation for
fluoride of 1.4 to 2.4 milligrams per liter (mg/L) (temperature-dependent) to prevent the occurrence of
objectionable enamel fluorosis, mottling of teeth that  can be classified as mild, moderate, or severe. In
1986, EPA established a maximum contaminant level goal (MCLG) and a maximum contaminant level
(MCL) for fluoride at a concentration of 4 mg/L (to protect against crippling skeletal fluorosis) and a non-
enforceable secondary MCL (SMCL) of 2 mg/L (to protect against objectionable dental fluorosis). These
guidelines are restrictions on the total amount of fluoride allowed in drinking water (and should not be
confused as recommendations about adding fluoride to drinking water to protect the public from dental
caries). In early 1990s at the request of EPA, the National Research Council (NRC) independently
reviewed the health effects of ingested fluoride and the scientific basis for EPA's MCL and concluded in
1993 that the MCL was an appropriate interim standard but that further research was needed to fill data
gaps on total exposure to fluoride and its toxicity (NRC,  1993).

Because new research  data on fluoride became available since then, EPA, upon completion of its first six
year review of the relevant drinking water standards in 2003, requested that NRC again review the
adequacy of its MCLG and SMCL for fluoride. In response, NRC convened a committee and published a

-------
report in 2006, in which it concluded unanimously that the present MCLG of 4 mg/L should be lowered
to cope with the risk of developing enamel fluorosis, a condition associated with enamel loss and pitting
for children (NRC, 2006). The majority of the committee also concluded that the MCLG is not likely to
be protective against bone fractures.

Although no revision was established after the release of the 2006 NRC report and completion of its
second six year review in 2010, EPA announced in a joint press release with the Health and Human
Services (HHS) on January 7, 2011, steps to ensure protection of Americans, especially children, by the
Fluoride Regulation, including plans to review the current 4.0 mg/L standard. Soon after, HHS proposed
via Federal Register (2011) that the recommended optimal fluoridation level to prevent dental caries be
set at 0.7 mg/L, the lowest end of the 0.7 to 1.2 mg/L range set previously by HHS.

1.3        Fluoride in Water Supplies

Fluoride occurs naturally in groundwater via weathering/leaching of rocks and soils containing fluoride-
bearing minerals such as fluorspar (CaF2), cryolite (NasALFe) and fluorapatite (3Ca3[PO4]2'Ca[F,Cl]2)
(Mackay and Mackay,  1989; Cotton and Wilkinson, 1988). Atmospheric deposition of fluoride-
containing emissions from coal-fired power plants and other industrial sources also contributes to the
presence of fluoride in surface water. Fluoride exists in water as fluoride ion (F~). According to a survey
performed by Centers for Disease Control and Prevention (CDC, 1993), over  170 water systems in the
U.S. were supplied by water containing over 4 mg/L of fluoride. Based on sampling conducted in 16
states in 2002, EPA estimated that a total of 106 water systems had a system mean concentration
exceeding the threshold of 4 mg/L and a total of 603 water systems exceed the threshold of 2 mg/L (EPA,
2003).  The number of water utilities containing more than 2.0 mg/L of fluoride in the finished water was
much higher at 1,200 as reported by the U.S. Public Health Service (PHS, 1969).  Higher fluoride
concentrations were usually found in western and southern states, such as Idaho (e.g.,  15.9 mg/L), New
Mexico (e.g., 13.0 mg/L), Oklahoma (e.g., 12.0 mg/L), Colorado (e.g., 11.2 mg/L), Texas (e.g.,
8.8 mg/L), and Arizona (e.g., 7.4 mg/L).

1.4        Available  Treatment Technologies

Activated alumina has  been shown, in the field, to be effective in fluoride removal, especially by paying
close attention to pH control during service cycles. Upon exhaustion, the spent media can be effectively
regenerated by means of pH adjustment using caustic. Other adsorptive media have also been developed
more recently but most, if not all, have been tested only in the laboratories with little or no full-scale
application data. Membrane technologies, such as reverse osmosis (RO), nanofiltration (NF), and
electrodialysis (ED), have also been utilized to separate and remove fluoride along with other
contaminants such as total dissolved solids (TDS), arsenic and uranium. In general, these technologies
are more complex to operate and more costly.  Section 2 of this manual provides detailed reviews of
available treatment technologies for fluoride removal.

-------
                 2.0: TREATMENT METHODS FOR FLUORIDE REMOVAL
2.1        Introduction

According to the 1986 Fluoride Regulation, community water systems (CWSs) that use source waters
containing excess fluoride must reduce its level to below the MCL at each entry point to the distribution
system (EPTDS). CWSs in the U.S. are defined as systems with at least 15  service connections or that
serve 25 or more persons year-round.  EPA provides guidance on available treatment technologies,
residuals produced by each technology, disposal options, and relevant regulations governing these options
(EPA, 2003; 1986). For each EPTDS exceeding the MCL, a system must consider a mitigation strategy
such as existing treatment optimization/modifications or new treatment installation if it cannot achieve
compliance through a non-treatment option.

EPA recommends that a system first determine if its existing treatment technology is capable of removing
fluoride even though it was not originally designed to do so. If non-treatment options or treatment
modifications are ruled out, treatment may be accomplished either at the wellhead with a centralized
treatment system, or at point-of-use (POU) or point-of-entry (POE) locations within a building or entering
a building.

2.2        Non-Treatment Options

Non-treatment options do not produce residuals, usually are less expensive,  and do not require any
additional operator training.  Typical non-treatment options include switching to a new or better source,
blending, and interconnecting with and/or purchasing water from another water system (EPA, 2012).

Switching to another source may involve drilling a new well in an aquifer containing low fluoride levels,
sealing off water producing zones containing high fluoride levels, or finding an uncontaminated surface
water source. Before switching, attention must be given to water quality of the new source so that it does
not interfere with the existing treatment process(es). Significant changes in water quality may require
new treatment processes, impact the distribution system, and cause other compliance issues. Switching to
another source also may be limited by the availability of new sources, existing water rights, and/or costs
for transporting the new source water to the treatment plant.

Blending involves diluting fluoride  concentrations of a contaminated source with another source
containing low or no fluoride. To minimize piping required to carry the sources to a common mixing
point, it would be ideal for the sources to be close to each other.  Mixing usually occurs in a storage tank
or a common header with resulting fluoride concentrations below the MCL or certain percentages (as
safety factors) of the MCL set by authorized regulatory agencies.  Similar to switching to another source,
care should be taken to any changing water quality.

Characterization of water quality must be carried out to  ensure that changes in water quality resulting
from blending are assessed and that potential impacts to the existing treatment processes and distribution
system are determined (EPA, 2012). Also any change in water quality should not cause other compliance
issues.  Blending may require specific regulatory approval.

Interconnecting with another water  system must consider if there is a nearby CWS meeting the
requirements of the Fluoride  Regulation, if the system is willing to interconnect or consolidate, and if the
system can  handle increased  demand from additional customers. Costs and the impact of interconnection
on water quality and the distribution system also need to be considered in the decision-making process.
In general, as the distance increases, the rise in elevation increases and the existence of physical barriers

-------
occurs, the cost of installing a delivery system and delivering the water becomes increasingly
unfavorable.

Use of bottled water is also an option. The reliability, cost and assurance that the consumers will use that
source are some of the deterrents to be considered.

2.3        Centralized Treatment Options

Common centralized treatment options for fluoride removal include AA and RO, both of which are
identified by EPA as the best technologies generally available (BTGAs) for fluoride removal based on
technological efficiency and economic accessibility (EPA, 2003; 1986). Centralized AA and RO
treatment and POU RO are also identified by EPA as small system compliance technologies (SSCTs)
(EPA, 2003; 1986). For water systems requiring only fluoride removal, AA is the preferred treatment
method. RO is the recommended process when, in addition to fluoride, TDS and other contaminants also
need to be removed.

Removal of fluoride by AA relies on exchange of fluoride ions (and some anionic contaminant ions such
as arsenic) for hydroxides on alumina surface, a process that is generally known as adsorption (although
ligand exchange may be a more appropriate term for the process involved [Clifford, 1999; Stumm,
1992]).  In addition to AA, other adsorbents such as bone, bone chars, and a variety of more recently
developed materials also utilize adsorption/ligand exchange to achieve fluoride removal from water
supplies.

RO utilizes a semi-permeable membrane and an applied pressure that overcomes the osmotic pressure to
achieve fluoride removal.  In addition to fluoride, many types of molecules and ions are also separated
and removed. Several other membrane processes have also been used for fluoride removal. These
include NF, ED, and electrodialysis reversal (EDR).

Conventional coagulation/filtration, lime softening, and somewhat modified methods, such  as the
Nalgonda technique, also have been suggested for fluoride removal.  These  processes rely on adsorption
and co-precipitation of fluoride and other contaminants such as turbidity, color, and/or hardness during
the formation of aluminum floes or magnesium hydroxide.

Methods to remove fluoride from water supplies have been reviewed by many researchers and
professional/government organizations both in the U.S. and abroad. Among the example articles and
reports are American Water Work Association (AWWA) (2011, 2004), National Health and Medical
Research Council (NHMRC, 2011), Health Canada (2010), Mohapatra et al. (2009), Shrivastava and Vani
(2009), Cooperative Research Center (CRC, 2008), Onyango and Matsuda (2006), Fawell et al. (2006),
Meenakshi and Maheshwari (2006), Pickard and Bari (2004), EPA (1998), and Sorg (1978).

2.3.1       Adsorption/Ligand Exchange Processes

2.3.1.1     Granular Activated Alumina. AA has been used since the 1940s as an effective adsorptive
media for fluoride removal (Sorg, 1978; Maier, 1947, 1953). It is a mixture of amorphous and gamma
aluminum oxide (y-AhOs) prepared by low-temperature (300 to 600QC) dehydration of A1(OH)3
precipitates. The material is highly porous and has a high average surface area per unit weight of 350
m2/g of media.  Adsorption occurs at both internal and external surfaces of AA.

The adsorptive capacity for fluoride by AA is pH-dependent, with fluoride best adsorbed below a pH of
8.2 - a typical zero point of charge (ZPC) - where alumina surface has a net positive charge. The
maximum removal capacity generally is achieved at pH between 5.5 and 6.0 (Clifford, 1999; Schoeman

-------
and Leach, 1987; Rubel and Williams, 1980; Choi and Chen, 1979; Rubel and Woosley, 1978), although
other optimum pH values also have been reported (e.g., at 4.4 [Goswami and Purkait, 2012], 7.0 [Ghorai
and Pant, 2004], and 5.0 to 8.0 [Choi and Chen, 1979]).  Therefore, adjusting the pH of source water
provides removal capacity advantages.  As the pH values deviate upwardly from the 5.0 to 6.0 range, the
adsorptive capacity for fluoride decreases at an increasing rate. Over-adjusting the pH to below 5.0,
however, can result in gradual dissolution of the media (Hao and Huang, 1986).

A common AA mesh size is 28  x 48 (Tyler Equivalent). Larger mesh sizes can be used, but their
capacities for fluoride are lower (Clifford, 1999).  A larger mesh size such as 14 x 28 is also
recommended by some media suppliers as a starting media size distribution, which is reduced with
regeneration to effectively become an operating mesh size of 28 x 48 over time. The use of larger mesh
sizes can extend the media life before losing the media as fines to the backwash cycles. Finer mesh
materials have not been used for this application other than in laboratory bench-scale work (AWWA,
2011, 2004) because they cause pressure-drop and backwash problems.

A typical AA system consists of two or more packed-bed adsorbers configured in parallel or in series.  In
general, the in-series configuration yields the highest fluoride loading on the media and the lowest
fluoride level in the treated water. The parallel configuration provides greater flexibility in treatment
flowrate and system operations. Because it uses two large tanks to accommodate the design flowrate
through each tank, an in-series system generally is more expensive. Downflow is a typical mode of
system operation, although an upflow fluidized system may produce a greater fluoride  removal capacity
than does a downflow system (Bishop and Sansoucy, 1978). Media attrition in an upflow system can
result in excessive media fines and cause premature media replacement especially under high loading
rates. Most media and system suppliers recommend an empty bed contact time (EBCT) of 5 min
(equivalent to 1 ft3 of media per 1.5 or 1.0 gpm of treated water flowrate) across each of two lead-lag
beds. A more conservative EBCT of 7.5 min was referenced in the 1984 design manual when the beds
are configured in parallel. For a given EBCT and a system design flowrate and a bed depth of 3 to 5 ft, a
vessel diameter can be calculated.

Typically, a fluoride breakthrough curve from an AA bed is not as sharp as that from an ion exchange
(IX) resin bed (Clifford, 1999), thereby making blending more difficult to implement.  To maximize the
media run length, one or more product water storage tanks can be utilized to equalize effluent fluoride
concentrations  and achieve a target effluent level through the service cycles.

The useful AA bed life is usually measured by the number of bed volumes (BV) of water treated before
breakthrough of fluoride at a target level (e.g., 1.5 mg/L). Depending on the water quality and operating
conditions, an AA bed can last from several hundred to a few thousand BV before media regeneration is
required (AWWA, 2011; Schoeman et al., 2006; Schoeman, 2009; EPA, 1980).  Table 2-1 presents a few
run length test results reported in the literature. As expected, influent water pH significantly affects the
system run length.

Spent AA can be regenerated using caustic soda.  Fluoride along with other contaminants such as arsenic,
selenium and organic molecules also adsorbed during the service cycle are removed during the process.
A 1% NaOH solution is effective in removing fluoride, but not as effective in removing more highly
charged contaminant ions, such as arsenic. If desired, higher strengths of caustic can also be used to
regenerate AA, but will result in more media dissolution. Regeneration usually involves steps  such as:

           •  Backwash to remove media fines and sediments and to fluff the media bed

           •  Drain the media vessel

-------
           •   Apply caustic upflow through the vessel with a slow flowrate (some facilities use both
               upflow and downflow during this caustic wash step)

           •   Apply rinse water and then acidic water at pH 2.5 and then 4.0 (using sulfuric acid)
               downflow through the vessel
                           Table 2-1. AA System Run Length Results
Study
Type
Bench-Scale
POU Devices
Full Scale
Pilot Scale
Pilot Scale
Influent
Fluoride
Level
(mg/L)
10.5
5.3-10.8
2.0
3.0
4.0
2.4-3.3
Influent
pH
(After pH
Adjustment)
(S.U.)
7.9-8.0
6.6
5.5-6.0
5.5-6.0
Run
Length (BV)/
Target
Fluoride
Breakthrough
Level
(mg/L)
167/1.5
566-780/
1.9-2.1
728/1.5
2,300/1.4
1,200/1.4
1,150/1.4
1,900^/1. l(c)
3,800^/1. l(c)
Other
Relevant
Water
Quality
Parameters
Alkalinity 325 mg/L(a)
Hardness 360 mg/L (a)
Sulfate 55 mg/L
Hardness 263 mg/L(a)
Sulfate 139 mg/L
TDS 810 mg/L
TDS 1,350 mg/L
TDS 1,210 mg/L
As 0.10 to 0.17 mg/L
Hardness 105-108 mg/L(a)
Sulfate 124-147 mg/L
Silica 45 mg/L(e)
Reference
Chauhan
etal.,2007
Schoeman
etal.,2006;
Schoeman,
2009
AWWA, 1999
EPA, 1980
(a) asCaCO3.
(b) Following lead column.
(c) Average value.
    (d) Following lag column.
    (e) asSiO2.
After media regeneration, the freeboard in the media vessel is often measured to determine if media
replenishment is required before returning the vessel to the service cycle (an AA plant adds media prior to
regeneration to take advantage of the backwash step during media regeneration). A marginal decrease in
adsorptive capacity (e.g., 5% loss after five regeneration cycles [Ghorai and Pant, 2004]) can occur after
each regeneration cycle. Much higher losses of adsorptive capacity (e.g., 30 to 40% [Fawell et al., 2006];
CRC, 2008) and media (e.g., 5 to 10% [Fawell et al., 2006; CRC, 2008]) have  also  been reported in the
literature, and require replacement of spent media after only a few regeneration cycles. These unusually
high losses are caused mainly by higher strengths (e.g., 4%) of caustic used.

Reduction in media capacity also may be caused by media fouling due to entrapped suspended solids,
metal hydroxides, carbonates and adsorbed silica. The media capacity can be restored by using
backwashing and air scouring with repeated rinsing and/or periodic acid treatment to remove silicate and
hardness, as demonstrated by an investigation of the performance of two AA plants in South Africa
(Schoeman and Leach, 1987).

Waste streams produced by media regeneration include spent caustic and alkaline rinse water. The most
common disposal  methods include  direct discharge to a publicly-owned treatment works (POTW) via

-------
sewer and to a lined evaporation pond (especially in arid areas). Some plants may opt to treat their
regeneration waste streams to a clarifier where calcium chloride is added to form calcium fluoride
precipitates, which are then processed through a filter press to form a sludge cake before final disposal.

Some small and very small systems may operate their AA plants without pH adjustment and media
regeneration to avoid needs for strong acid/base storage and handling and waste stream disposal (EPA,
1998; Fox and Sorg, 1987).

Key advantages and disadvantages of using AA for fluoride removal include:

           •   AA is  somewhat specific for fluoride; its capacity is not significantly affected by sulfate
               and chloride concentrations of the water. However, arsenic, selenium, and other
               inorganic ions (such as phosphate) and organic molecules (such dissolved organic matter)
               can compete with fluoride for available adsorption sites.

           •   Upon exhaustion, spent AA can be effectively regenerated using caustic followed by acid
               neutralization and a water rinse. The use of acid and caustic, often at the industrial
               strengths of 93% and 50%, respectively, is  also the major disadvantage to the use of AA.

           •   AA is  a relatively low cost media comparing to the costs of a synthetic anionic resin and
               bone char (Sorg, 1978).

Some of the earliest and longest-operating AA plants include:

           •   Bartlett,  TX: The system was designed to remove  8 mg/L of fluoride initially. After a
               new well was developed with only 3 mg/L  of fluoride, the system continued to operate
               with essentially the original media from 1952 through  1977 when the system was
               abandoned (Sorg, 1978; Maier, 1953).  The raw water pH was adjusted to 7.0 during the
               entire operating period.

           •   Desert Center, CA: This 1968-built,  1,100-gpm manually-operated plant has  two 14 ft x
               16 ft cylindrical vessels configured in parallel. Due to corrosion, the two vessels had to
               be replaced with a carbon steel and a stainless steel vessel in  1995 and 2001,  respectively.
               Currently, the gravity-flow system continues to operate with one tank at a time on a much
               reduced  schedule (i.e., one day per week).

           •   Gila Bend, AZ: Built in 1978, this 900-gpm plant  continued to operate through early
               2000 before it was replaced by a three-train RO system for fluoride and arsenic removal.

The largest AA plant ever built in the U.S. is the Twentynine Palms Fluoride Removal Plant in
California's Morongo Basin.  Started in 2003, the 2,100-gpm (or 3 millions of gallons per day [MGD])
plant consists of three parallel treatment modules, each having two 11 ft x 12 ft vessels configured in
series. The plant operates at an influent pH of 5.0 and an effluent pH of 8.0 after respective acid and
caustic additions.  The media beds are regenerated on a frequency  of approximately one bed every three
days at the current production rate of 1.25 MGD. The regeneration uses both upflow and downflow
modes during the caustic wash step. Following the free-board measurement and media replenishment
(media is currently added to the vessel prior to regeneration to utilize the backwash step during the
regeneration process), the newly-regenerated vessel is switched to  the lag position for continuing service.
An average of 25% blending ratio is implemented to meet the statutory limit of 2.0 mg/L in California.

2.3.1.2     Bone, Bone  Char, and Apatite-Based Materials.  Bone was one of the first materials
suggested for fluoride removal due to its known affinity for fluoride. Through ligand exchange (i.e., F~

-------
for COs) on its main ingredient, carbonate apatite (CagfPO^e'CaCOs), a degreased, caustic-and-acid
treated bone material was found effective in reducing fluoride concentrations from 3.5 to < 0.2 mg/L
(Sorg, 1978). Because of its high cost, this material has not been widely used.

Another bone material, bone char, has been tested and used in several full-scale applications for fluoride
removal. Originally developed for decoloring cane syrups in sugar manufacturing, bone char is ground
animal bones charred to remove all organics and consists essentially tricalcium phosphate (CasfPO/^) and
carbon. Similar to AA, bone char also removes competitive anions, such as arsenic, and can be
regenerated with caustic upon exhaustion. Unlike AA, arsenic cannot be stripped off during normal
caustic regeneration,  thus rendering the material useless after a number of regeneration cycles (Sorg,
1978). The use of bone char can slough phosphate.  Because it is more soluble in acidic water and more
expensive than AA and because the treated water has a bad taste in some cases, bone char is not an
economically viable option for fluoride removal.

Synthetic bone materials such as granular/powder tricalcium phosphate have been developed and used at
a few full-scale plants in the 1940s. Due to less than satisfactory results (i.e., high attrition losses and
diminishing capacities due to the presence of sulfate ions), these plants have been abandoned after several
years of service (Sorg, 1978).  More recently, there have been renewed interests in using synthetic bone
materials, such as hydroxylapatite (CaiofPC^MOHh), as low-cost adsorbents for fluoride removal in low-
income countries (Sternitzke et al., 2012; Fan et al., 2003).  Tests performed for these materials, however,
have been limited mainly to material properties in the laboratories.

2.3.1.3     Ion Exchange Resins.  IX is a common water treatment process that removes ionic
contaminants via displacement of weaker binding, exchangeable ions on a resin. Removal of fluoride has
been tested using strongly basic anionic exchange (AIX) resins containing quarternary ammonium
functional groups, on which chloride ions are replaced by fluoride  ions.  Upon saturation, resins are
regenerated with brine solutions to recharge the functional groups with chloride ions. Because of
fluoride's low  selectivity, the efficiency of the resins is significantly reduced in the presence of competing
anions such as phosphate, arsenate, sulfate, carbonate, and alkalinity (Clifford, 1999).  Due to this and
other factors such as  pretreatment requirements (to prevent resin fouling) and high costs, IX is not an
economically viable treatment option for fluoride removal.

Special-purpose resins have been developed and tested recently in  the laboratories for fluoride removal.
The materials tested include La(III)-loaded polymethyl acrylate (PMA) resin, Zr(IV)-loaded Amberlite
XAD-7 resin, La(III)-AFB resin, Pr(III)-AFB resin, A1(III)-AMPA resin, La(III)-impregnated silica gel,
cross-linked pectic acid gel, phosphorylated cross-linked orange juice gel, and La(IIII)-loaded 200CT
resin (Onyango and Matsuda,  2006). Except for the batch experiment results, not much is known about
their long-term stability or large-scale operation. These materials are also expected to be expensive.

2.3.1.4     Other Adsorbents. In addition to AA and bone/apatite-based materials, a number of one-time
use or regenerable novel adsorbents have been developed in recent years  and are claimed, based mostly
on laboratory test results, to be effective in reducing fluoride to below the relevant statutory limits.
Behaving somewhat similarly to AA (e.g., being pH-dependent and affected by certain competing
anions), many  of these materials are found to  possess greater adsorptive capacities, achieve more rapid
uptake kinetics or be  more economically affordable. Attempts were made to tabulate and compare test
results of these materials against AA, but the tabulated entries are not necessarily comparable to one
another due to  significantly different test conditions used by various researchers. Further, by using
mostly batch and some column experiments, most researchers reported only initial/influent and treated-
water fluoride  concentrations along with percent removal  results (some also reported isotherm data).
While being useful for demonstrating the materials' potential to become robust adsorbents, these data
have rather limited values to reflect the materials' usefulness under the real world settings. As such, the

-------
test results of these novel adsorbents are not discussed in this subsection. Those interested in any material
are encouraged to read the specific articles as referenced below and Bhatnagar et al. (2011), which review
the performance of over 100 novel adsorptive materials for fluoride removal under various conditions.

Examples of the newly developed novel adsorbents and existing low-cost materials tested include:

           •   Modified alumina such as manganese (di)oxide-coated AA (Teng et al., 2009; Tripathy
              and Raichur, 2008; Maliyekkal et al., 2006) and magnesia-amended AA (Maliyekkal et
              al., 2008)

           •   Nanomaterials such as nanomagnesia (Maliyekkal et al., 2010)

           •   Layered mixed metal oxides/hydroxides (or hydrotalcite-like compounds/anion clays),
              consisting of brucite-like  hydroxide sheets with a general formula:

                                  [Mni.xMinx(OH)2]x+[(An-)x/n • mH20]x

              where Mn is a divalent cation like Mg2+, Zn2+, Cu2+> etc., Mm is atrivalent cation like
              A13+, Cr3+, Fe3+, etc., and An~ is an anion (Sujana and Anand, 2010; Onyango and
              Matsuda, 2006)

           •   Geo-materials such as bauxite (Das et al., 2005); limestone/calcite (Turner et al., 2005;
              Fan et al., 2003); fluorspar, quartz, and iron activated quartz (Fan et al., 2007);
              lanthanum, magnesium and manganese-impregnated bentonite clay (Kamble et al., 2009);
              and other low-cost clays and soils (Mohapatra et al., 2009; Onyango and Matsuda, 2006)

           •   Zeolites and surface-tailored zeolites (Onyango and Matsuda, 2006)

           •   Biopolymers such as chitin, chitosan, and Al or La-modified chitosan (Jagtap et al., 2011;
              Miretzky and Cirelli, 2011;  Swain et al., 2009; Kamble et al., 2007) and Zr-impregnated
              collagen fiber (Liao and Shi, 2005)

           •   Carbonaceous materials such as A1-, Ti- and La-impregnated granular activated carbon
              (GAC; Jing et al., 2012; Onyango and Matsuda, 2006); calcium compounds-containing
              charcoals (Tchomgui-Kamga et al., 2010); biochar (Mohan et al., 2012) and other coal-
              based sorbents (Sivasamy et al., 2001)

           •   Waste-derived adsorbents such as alum sludge (Sujana et al., 1998); red mud (Mohan and
              Pittman, 2007); and waste residue from alum manufacturing process (Nigussie et al.,
              2007)

           •   Other low cost materials,  such as ragi seed powder, horse gram seed powder, orange peel
              powder, chalk powder, pineapple peel powder, and multhani matti (Gandhi et al., 2012).

2.3.2       Membrane Processes

2.3.2.1     Reverse Osmosis andNanofiltration.  RO, a process that moves clean water (or permeate)
across a membrane against the concentration gradient as a pressure that is higher than the osmotic
pressure, is exerted on the side with the concentrated solution.  The remainder of the feedwater along with
rejected contaminants (or reject) is discharged as a concentrated waste stream.  The effectiveness of the
process and the amount of permeate produced (referred to as recovery rate) are heavily dependent on
water quality, size and charge of contaminant ions, membrane properties, as well as system operating
pressure, temperature and flowrate.

-------
In general, larger and charged species are retained more easily by the membrane than smaller and
uncharged species. "Tighter" membranes allow less solute passage than "looser" membranes. Increasing
the operating pressure will force more water through the membrane, thus increasing the solute rejection
rate. However, increasing the feed flowrate under a constant pressure, will reduce the system recovery
and solute rejection rates. A higher feedwater temperature will increase the permeability of the
membrane, thus lowering the operating pressure and solute rejection rate.

During the past decade, membrane manufacturers have developed RO membranes with pore sizes and
operating pressures that lie between those of traditional RO and NF membranes. As such, RO and NF are
on a continuum of membrane characteristics, rather than two distinct technologies.  Due to larger pores,
NF membranes offer less resistance to passage of both solvent (like water) and solutes, thus requiring less
pressure to operate the systems. The low-pressure RO membranes are those with operating pressures that
are close to those of traditional NF membranes.

When contaminants other than fluoride  also need to be removed from source water, RO can be a good
option for fluoride removal.  The removal efficiency (or rejection rate) is high, ranging from 83 to 98%
based on results of a pilot study that evaluated five commercially available RO membrane elements (EPA,
1988). A more recent study carried out in India also achieved a 95% fluoride rejection rate (Arora et al.,
2004). Operating at an 80% recovery rate, a 1.6 MOD, extra low energy RO plant in Finland achieved a
permeate fluoride concentration of <0.03 mg/L prior to 70% blending (Sehn, 2008). After more than
three years of system operation, the plant maintained a salt and fluoride  rejection rate as high as what was
obtained during the initial system startup. Table 2-2 summarizes some reported fluoride removal results
by either RO or NF.
                 Table 2-2. Fluoride Removal Efficiencies (Rejection Rates) by RO
Study
Type
RO-LS
RO-LS
NF-BS
NF-PS
NF-PS
NF-PS
NF-PS
NF-PS
RO-PS
NF-PS
RO-FS
RO-FS
Influent
Fluoride
Level
(mg/L)
1.4-6.6
4.2-9.3
4.0-6.8
1.8-20.0
2.3-22.3
13.5
5.0
2.8
5.3-14.5
4.7
1.3-1.8
3.3-5.6
Treated
Water
Fluoride
Level
(mg/L)
0.17-0.78
0.32-0.88
<0. 19-1.67
0.07-2.79
0.05-4.0
0.7
NR
0.6
NR
0.03-0.06
0.03
0.33-0.56
Removal
Efficiency
(or Rejection
Rate)
(%)
88-89
91-92
62-97
50-99.5
74-99
94
78-95
79
83-98
98-99
>97->98
90
Reference
Arora etal., 2004
Meenakshi and Maheshwari, 2006
Cohen and Conrad, 1998
Tahaikt et al., 2007
Tahaikt et al., 2008
Lhassani etal., 2001
Diawara etal., 2005
Health Canada, 2010
Huxstep and Sorg, 1988
Cohen and Conrad, 1998
Sehn, 2008
Cohen and Conrad, 1998
BS = bench scale;
osmosis
FS = full scale; LS = laboratory scale; NF = nanofiltration; PS = pilot scale; RO = reverse
Limitations of the RO process include possible membrane scaling, fouling, and failure as well as higher
energy and capital costs. Calcium, magnesium, and silica can cause scaling and decrease membrane
efficiency.  Colloids and bacteria can also cause fouling.  Both fouling and scaling will increase pressure
drop, thus decreasing membrane life and increasing energy costs. Pre-treatments such as softening and
cartridge filtration and/or membrane cleaning can help obtain acceptable membrane run times.
                                               10

-------
Membrane failure can allow contaminants to pass through to the finished water. Membrane integrity
testing through either pressure drop or markers or measuring parameters in the effluent such as TDS,
turbidity, or particle counts can help alleviate the concern.  Chlorine can damage RO membranes and
should be quenched using de-chlorination chemicals or GAC.

Due to high pressure, full-scale RO is a more expensive treatment option for removing fluoride. Also, a
RO unit can produce a large amount of reject water, e.g., from 10 and 70% depending on pressure drop
and pore size. Increasing the number of membrane stages can increase the percent recovery and decrease
the waste production.

Because RO removes alkalinity in water, it will lower product water pH and increase its corrosivity.
Therefore, the product water pH must be adjusted to avoid simultaneous compliance issues in the
distribution system such as elevated lead and copper.

During the past decade, a number of large, full-scale RO systems have been installed for the removal of
fluoride and other contaminants.  Among the example plants are Gila Bend, AZ (to replace an AA plant
installed in late 1970s);  Andrews, TX (to remove fluoride and arsenic); Wolfforth, TX (to remove
fluoride and arsenic); Blythe, CA (this, as the main plant, and an AA plant supply water to a large state
correction facility); and Maude, OK.

2.3.2.2     Electrodialysis andElectrodialysis Reversal.  Originally developed for demineralizing water
from brackish water sources, ED and EDR use an electrical current to separate ionic contaminants from
the feedwater through semi-permeable membranes. In EDR, the polarity of electrodes is reversed
periodically on a preset time cycle, thus causing a reversal  in ion movement. The change in the direction
of ion movement helps reduce scaling and eliminate the need for chemical conditioning. Because water
does not physically pass through the membrane, particulate matter is  not removed.

A basic EDR unit, commonly referred to as a membrane stack, consists of several hundred cell pairs
bound together with electrodes on the outside. Feedwater passes simultaneously through the cells to
provide  a continuous, parallel flow of desalted product water and brine that emerge from the  stack. A
single pass EDR unit has 20 to 30% water loss; a sequential EDR unit can reduce water loss to 5 to 10%.

Results of a laboratory study showed that fluoride concentrations could be reduced to 0.67 mg/L when
applying a 10 V current to a brackish water containing 3 mg/L of fluoride and 3,000 mg/L of TDS (Amor
et al., 1998). Increasing the voltage  to 15 V further reduced the concentration to 0.21  mg/L.  A pilot
study on a brackish water reported a reduction in fluoride concentration from 1.8 to 0.5 mg/L (Tahaikt et
al., 2006).

Because of system complexity and costs, only a few full-scale ED/EDR systems are installed to date.  A
3.8-MGD EDR system installed in Virginia reduced fluoride concentration from 4.8 to 1.2 mg/L
(Thompson and Robinson, 1991). Another EDR system was installed in Thunderbird Farms  Domestic
Water Improvement District in Arizona (De Haan, 2011), but the relevant system performance data were
not available at the time of this manual revision.

2.3.3       Other Centralized Treatment Methods

2.3.3.1     Conventional Coagulation/Filtration. Conventional coagulation/filtration is one of the most
common water treatment processes for turbidity and color removal from water supplies. With the
addition of an aluminum salt (i.e., alum), the process  can remove fluoride (e.g., reduced from 3.6 to 1.0
mg/L), but the removal requires large doses (e.g., 350 mg/L), as demonstrated by early laboratories
studies performed in the 1940s (Sorg, 1978). A more recent study used an alum and powder activated
                                              11

-------
carbon (PAC) slurry to reduce fluoride concentrations from 15 to 1.5 mg/L. High alum doses (i.e., up to
850 mg/L as AhCSO^s 16H2O), again, were required to achieve the intended result (Mekonen et al.,
2001). Similarly large doses (i.e., 80 mg of Al2(SO4)3 per 1 mg of F") were used to remove fluoride from
an Estonia water. Because these doses are much greater than those commonly used for turbidity and color
removal, alum coagulation has not been considered a practical solution for fluoride  removal.

2.3.3.2     Lime Softening. Lime softening is a precipitative process that removes calcium and
magnesium ions from hard water.  Lime and, perhaps, soda ash and/or magnesium carbonate are added to
raise pHto up to  9.5 and greater than 10.6, respectively, resulting in calcium carbonate and magnesium
hydroxide precipitates. Lime softening can remove fluoride from water, and the removal is a function of
the amount of magnesium removed, according to some early laboratory studies performed in 1930s (Sorg,
1978). The mechanism involved is believed to be coprecipitaiton with magnesium hydroxide.  Similar to
alum coagulation, large quantities of chemical(s) are required for fluoride removal.  Therefore, this
method is adaptable to only low-fluoride-high-magnesium water requiring softening.

2.3.3.3     Nalgonda Technique. Considered a simple and economical method, the Nalgonda technique
utilizes alum and lime followed  by rapid mixing, flocculation, sedimentation and filtration to remove
fluoride.  The process involves dissolution of alum and development of aluminum hydroxide micro-
particles to remove  fluoride via electrostatic attachment during flocculation.  Because the alum solution is
acidic, simultaneous addition of lime is required to ensure a neutral pH and complete precipitation of
aluminum.  Compared with the conventional coagulation process, the Nalgonda technique requires a
much larger dosage of alum and the fluoride captured in the aluminum hydroxide floes can be released
slowly back to water (Shrivastava and Vani, 2009). The process has been used for fluoride removal from
water supplies in some parts of India (Nawlakhe and Rao, 1990; Bulusu et al., 1983; 1979).

2.4        Point of Use/Decentralized Treatment Technology

The POU technology uses devices that sit on a counter, attach to a faucet, or are installed under a sink to
treat small amounts of drinking water at homes or other locations such as schools. They differ from POE
devices, which are installed on the water line as it enters the home or a building and treat all the water in
the building. POU RO is identified by EPA as an SSCT (EPA, 2003; 1986). A typical POU RO unit is
composed of a prefilter (to remove suspended solids), a GAC prefilter (to remove chlorine), a RO
membrane module, and, perhaps, a GAC postfilter before the outlet faucet.

Fox and Sorg (1987) investigated a commercial POU RO system in an EPA laboratory using Cincinnati
tap water spiked with 5.95 mg/L of fluoride. Over 98% removal was  observed during a four-day
intermittent system operation.

In a guidance document published by EPA (2006) for small drinking water systems, four POU RO case
studies, including one for a school and three for homes, were discussed for fluoride removal. All four
units treated fluoride from mean influent levels of up to 6.1 mg/L to levels below the MCL. In the three
household units, there were signs of bacterial growth in GAC postfilters.  The issue in two of the studies
was addressed by flushing the units, consumer education for more frequent use of the units, or flushing
the distribution system to ensure a chlorine  residual of at least 1.5 mg/L (as Ch) in the feed water.

Numerous POU RO devices are commercially available, for which NSF International has listed, as of
June 2013, 30 different manufacturers as certified providers for fluoride removal under NSF/ANSI
standard  058 (NSF International, 2013a). The  standard requires certified devices to reduce fluoride from
an average influent level of 8.0 mg/L to an effluent level of less than 1.5 mg/L (NSF International,
2013b). In general, these devices operate at rather low recoveries (e.g., 40%).
                                              12

-------
Although not as popular as POU RO devices, some POU units using AA or other adsorbents also are
commercially available for fluoride removal (Natural Health Enterprises, 2013; National Fluoridation
Information Service, 2012).  These products, however, are not NSF International-certified.

Depending on the technology used, POU units can be more cost effective for very small communities that
do not have the resources required to build a centralized treatment facility.  Some states do not allow POU
units for MCL compliance, primarily due to the challenge to develop an acceptable monitoring program
for homeowners and systems that install POU units.
                                              13

-------
              3.0:  DESIGN OF CENTRALIZED ACTIVATED ALUMINA PLANT
When designing a centralized fluoride removal plant, a design engineer typically divides the project into
three phases:

           (1)  General Plan - The general plan encompasses the conceptual design with basic design
               information and is often required for regulatory agency review.
           (2)  Preliminary Design Plan - The preliminary design plan typically includes the completion
               of 30% of system design drawings, which are used to establish a cost estimate and select
               potential major equipment suppliers.
           (3)  Final Design Plan - The final design plan is the completion of the contract documents,
               which are used to bid and construct the treatment plant, subject to regulatory agency
               review and approval.

3.1        General Plan

3.1.1       General Considerations. The general plan is prepared to provide background information
on the project and outline specific issues that must be addressed to treat the source water. The plan
summarizes the basis of design for all elements of the project and evaluates those against any regulatory
standards to make sure that regulatory compliance will be met. Key elements of the plan include an
analysis of the source water, reliability of supply, evaluation of the appropriate treatment process,
establishment of design data in accordance with regulatory requirements, and conceptual layout. Budget
cost estimates are derived using general guidelines with conservative contingencies provided for unknown
items, which may be determined during the preliminary and final design.

An analysis of the raw or source water is perhaps the most critical consideration during this phase of
system design.  Comprehensive raw water analyses of all inorganic, organic, radionuclide, and
bacteriological contaminants can help verify that the AA process is applicable for fluoride removal. The
data from the source water analysis will impact all aspects of system design, from selection of a treatment
system to materials costs and labor. An example  of the different types of information required for a raw
water analysis is provided in Table 3-1.

Another major consideration at this phase is siting of the treatment plant. The most practical approach is
to install the plant in such a location that expensive improvements do not need to be made in order to
convey finished water to the customers of the plant. In some cases, existing well pumps can provide
adequate flow and pressure through the plant. If the existing well pump is oversized (i.e., at a much
higher flowrate than the peak requirement), it should be resized to deliver only slightly more (e.g., 125%)
than the peak flowrate.  Because the flowrate dictates the treatment equipment size and capital cost, the
design flowrate should be minimized to the extent possible to ensure that the capital cost of the treatment
system is minimized. The AA volume is a function of flowrate and EBCT. The treatment vessels, pipe
sizes, and chemical feedrates all increase as the flowrate increases. A well-matched pump should be able
to handle additional head loss associated with the treatment system without a significant drop in pump
efficiency.  If the additional head loss cannot be met with the existing pump, several options exist:
increasing the size of the motor, increasing the size of the impeller, or replacing the pump. Reducing
flowrate for an oversized pump can result in excessive equipment wear and energy cost.
                                               14

-------
Other items that need to be determined in the general plan may include the following:

           •  Present water consumption and the projected average and maximum daily demands.

           •  Fire flows that meet the recommendations of the insurance service office or other similar
              agencies for the service  area involved

           •  Hours of operation, that is related to the system utilization rate.

           •  Redundancy requirements per relevant state regulations.

           •  Automatic or manual system operations. With manual operation, personnel must be
              available or on site during operation of the plant.  Automatic operation can save labor
              costs if designed properly.

           •  Water storage facilities, which must be evaluated to balance the hours of operation
              against the sizing of the plant. In general, storage should be provided to contain a mini-
              mum of one half the maximum daily consumption requirement. This is based on the
              premise that maximum consumption takes place during 12 hr of the day.  Then, if the
              system operates during the entire 24 hr, storage drawdown occurs during 12 hr and
              recovers during the remaining 12 hr.

           •  Construction materials that comply with Occupational Safety and Health Administration
              (OSHA) standards, local building codes, and health department requirements. Materials
              also must be suitable for the pH range of the water and be compatible with the use of pH
              adjustment and regeneration chemicals. Consideration for oxidants being used will
              determine the types of materials and ventilation system used in the treatment facilities.
              All chemicals used must be properly stored and all chemical storage facilities must meet
              all safety and containment requirements. Both drinking water chemicals  and system
              components should comply with NSF/ANSI Standards 60 and 61, respectively.

           •  Protection of treatment system equipment from ambient weather. It is recommended that
              the system be housed within a treatment building, although housing is not mandatory in
              some locations.

           •  Options of wastewater disposal.  This should be carefully evaluated in the design of any
              centralized water treatment plant. Wastewater resulting from backwash and regeneration
              of spent AA media must be disposed of in a manner permitted by state  and/or local
              regulatory agencies. Separate local and state regulatory reviews may be required for
              wastewater disposal.  Quantifying the backwash and regeneration waste and determining
              the disposal requirements also should be outlined.
A general plan report containing all of this information as well as a preliminary project estimate and
schematic drawings should be submitted for review and approval by the appropriate authorities. This
document can be used to establish funding requirements for the project. A determination of what funding
is available should be made before the project is authorized for preliminary and final design.  If the
preliminary estimate of project costs exceeds the available funds, adjustments should be made to increase
the funding or reduce the scope of the project.  Figure 3-1  illustrates the steps of the project development
process from project authorization through final design.

This design manual is applicable when fluoride removal is the only treatment required.  Removal of other
contaminants such as iron/ manganese, suspended solids, hardness, organics, and radionuclides also may
be required.  In those cases, alternative treatment processes and/or additional treatment  processes must be
evaluated.
                                               15

-------
The sequence of other treatment steps should be compatible with fluoride removal using AA. Removal of
iron/manganese, suspended solids, hardness, and organics should take place upstream of the fluoride
removal process. Disinfection with chlorine should take place after fluoride removal because it has been
noted that chlorine can degrade the performance of activated alumina (Rubel, 1984). Although no known
investigation has determined the amount of chlorine that can be tolerated by AA, process degradation has
been eliminated on projects where prechlorination was terminated (Rubel, 1984). Other treatment
processes may be required upstream of the fluoride removal process, but that decision will be made on a
case-by-case basis.
              Table 3-1. Example of Fluoride Removal Plant Water Analysis Report
                     Name and Address:                   Source of Water:
                     Container:
                     Sampling Date:
              	Sampled by:	
Analysis No.
Sodium (mg/L)
Calcium (mg/L)
Magnesium (mg/L)
Total Iron (mg/L)
Soluble Iron (mg/L)
Total Manganese (mg/L)
Soluble Manganese (mg/L)
Aluminum (mg/L)
Chloride (mg/L)
Fluoride (mg/L)
Nitrate (mg/L as N)
Nitrite (mg/L as N)
Ammonia (mg/L as N)
Sulfate (mg/L)
Silica (mg/L) (mg/L as SiCh)
Arsenic ((ig/L)
Phosphate (mg/L as P)
Total Alkalinity (as CaCO3)
Total Hardness (as CaCOs)
Total Dissolved Solids (mg/L)
Total Organic Carbon (mg/L as C)
Turbidity (NTU)
Color (Units)
pH
Temperature (°F)
Specific Conductance (micro-mhos)
#1


























#2


























#3





















































                                              16

-------
       Project Authorized
           Pre-Planning
               Budget

              Milestones

               Schedule
        Preliminary Design

       Surveying
       Geotechnical Services
       Client Meetings
       Basis of Design
       Drafting of Existing Treatment
       Facilities
       Design Sketches
       Equipment Information
       Hydraulic Profiles
       Electrical/Mechanical Data
       Operational Description
       Estimate of Costs
               Review
     Client Comments
     Resolve Potential Regulatory Issues
            Final Design

     Detailed Drawings
     Detailed Specifications
     Quality Control Review
     Agency Reviews
     Client Reviews
     Estimate of Cost
     Revisions
     Final Contract Documents Completed
Figure 3-1. Project Development Process
                  17

-------
3.1.2       Conceptual Design. Several conceptual design issues are to be considered in the general
plan.  These design considerations provide a conceptual delineation of the process, but do not provide
specific details in equipment size, arrangement or material selection. There are four basic options from
which a conceptual design can be selected.  While every combination of the operation can perform the
process but, under a select set of conditions, a certain combination may be preferred.  The options are as
follows:
       (1) Gravity or pressure flow
       (2) Single or multiple AA vessel(s)
       (3) Up- or downflow direction
       (4) Series or parallel vessel arrangement.

The most cost-effective AA configuration is a downflow multiple-bed pressure system arranged in
parallel. However, some large plant operators and media suppliers prefer the use of multiple lead-lag
trains. The lead-lag configuration yields the highest fluoride loading on AA and the lowest treated water
fluoride concentrations, but requires more capital investment.  The single treatment unit configuration is
less efficient unless there is an exceptionally large treated water storage capacity. In that case, the
economy of treated water blending can take place in storage.  Because of the space and capital
requirements, this is not an economic option.

A gravity flow system does not provide the economics of a pressure system; the treatment flowrate is
lower; re-pumping of treated water is always required, and capital costs are higher.  Because free carbon
dioxide (CCh) is released to the atmosphere in a gravity system utilizing pH adjustment, it is easier to
control pH in a pressure system. Downflow treatment has consistently yielded higher fluoride removal
efficiency than upflow.  Because downflow utilizes a packed bed, the flow distribution is superior. If
upflow beds are restrained from expanding, they would in effect also be packed.  However, they would
forfeit the necessary capability to backwash. Once the bed configuration is defined, a basic  schematic
flow diagram is prepared (see Figure 3-2). This diagram presents all of the  subsystems required for pH
adjustment and media regeneration. A summary of subsystem components  is presented in Appendix A.

For systems in which the raw water fluoride concentration is slightly above the fluoride MCL, bypassing
and blending a fraction of raw water with AA-treated water should be evaluated.  This option saves
treatment chemicals, extends treatment media cycle life, and reduces operating costs.  If bypassing and
blending is found to be feasible, the treatment system can be sized to treat less than 100% of the total
flow.  Plants with less budget constraints can still size their systems to 100% of the total flow to allow for
blending volume adjustments.

Treatment of regeneration wastewater can be handled by several different processes, and therefore, is
beyond the scope of this design manual. For the purpose of this manual, a lined evaporation pond is used
for disposal of regeneration wastewater, but is applicable in arid climates where evaporation rates are high
and land required for the basins is available at low costs. In regions where evaporation rates are low,
backwash and regeneration wastewater can be neutralized and contained in  a surge tank from which slow
discharge to a sewer system is permissible.  This latter disposal method can only be used when local
regulatory agency approval is provided.

Prior to proceeding with the preliminary design, financial feasibility should be determined.  Funding
limits for the project should be defined.  A determination that funding is available to proceed with the
project should be made; this requires a preliminary rough project estimate with an accuracy of ±30%. If
the preliminary rough estimate exceeds the available funds, adjustments should be made to increase
funding or reduce the project scope.
                                               18

-------
       Raw Water
                                                               .Vent
                      _Oi i


                      as
                                                           CAUSTIC STORAGE
                                             CAUSTIC DAY
                                                 TANK
                                                      n
                                         ACID-

                                        PUMPS
                                                   TANK


                                                 tVont
                                               ACID DAY

                                                 TANK
                                               ^Future Conn.
                                                             ACID STORAGE

                                                                 TANK
    PH

SENSOR
                                                I Future Conn,
Treated Water To

Indicator & Alarm
                     AUTOMATIC
                        VENT
                             AUTOMATIC
                                VENT
               TREATMENT


               UNIT NO.  1
                                   TREATMENT


                                    UNIT NO. 2
                                               (Future  Conn.
                                                   ^Backwash  Wastewater
                                        Regeneration Wastewater
                            LINED EVAPORATION  POND
                 Figure 3-2. An Example Parallel Flow Diagram
                                     19

-------
3.2        Preliminary Design

After completion and approval of the conceptual design by the client, the regulatory agency(s), and any
other affected party, preliminary design can begin.

3.2.1      Basis of Design.  The basis of design is a document, outline, or strategic plan that is
developed early in a water treatment system project to record and summarize decisions that have a major
and extensive impact on project design and implementation. The basis of design also helps minimize late
changes, additions, or modifications to the project, as well as the high expenses commonly associated
with late changes.  The development of the basis of design should not be performed solely by the project
manager; the owner of the water treatment system must have opportunity to review and comment on the
content of each design element. The following subsections discuss elements that should be addressed in a
basis of design.

3.2.1.1     General

          (1)  State the purpose of the project (i.e., what problem is the project designed to correct?).
          (2)  Identify areas of new or unique design and provide criteria.
          (3)  Identify areas where evaluation of alternatives must first be completed before initiating
               final design.  Identify alternatives to be evaluated.
          (4)  Identify critical structures, processes, or complex areas that require early engineering and
               design effort to avoid later delays.
          (5)  State major constraints such as maximum construction cost, and court-imposed or client-
               imposed deadlines.
          (6)  Note availability of prior drawings and dates when previous on-site project work was
               done.
          (7)  Note major potential trip-up items (i.e., power availability, flood plain location, historic
               register, property or easement availability, financing, etc.).
          (8)  Identify provisions to be made for future construction and expansion, beyond present
               scope, for sizing of or location of structures or equipment.
          (9)  Note who has jurisdiction for permit approvals (i.e., plumbing, electrical, building,
               elevator, elevated tank, groundwater protection, EPA, etc.).

          (10) Identify unusual situations that will affect design (i.e., rock, unstable soil, high
               groundwater, corrosion).

          (11) List specific points where client has expressly requested to be advised of design
               decisions, or where client will require involvement of staff in decision-making.
          (12) Identify hazards or hazardous areas (i.e., asbestos, windowless building story, confined
               space, fire, National Electrical Code (NEC) explosion areas, corrosion, fumes, dust,
               odor).  For asbestos, determine responsibility for discovery, arrange testing, and
               determine level of abatement required.
          (13) Identify large or complex structures that will require special building code compliance
               review prior to initiating final design.
                                                20

-------
3.2.1.2     Project Scope
           (1) Provide a schematic process flow diagram (i.e., show such items as water or wastewater
               flow, chemical feed, site sanitary sewer, and drain piping).
           (2) Provide a list of building, structures, and equipment.
           (3) Based on client's input, identify major equipment or brands of equipment to be used or
               not used.
           (4) Prepare tentative list of plan sheets.
3.2.1.3     Process Design Data Summary
           (1) List design data summary.  Note average, maximum, and peak hydraulic flowrate
               capacities.  Define concentrations to be removed or treated.
           (2) Identify "design parameters" and "units furnished" for each unit process or major
               equipment item.
3.2.1.4     Site
           (1) Provide a simple site plan with locations of existing and new structures, including
               sanitary and storm-water pumping  stations as applicable.
           (2) Note any special consideration related to design (i.e., location in flood plain, dike
               construction, location to adjacent residential areas or parks, requirements for site clearing,
               major underground facilities that will affect location of new improvements).
           (3) Summarize concept for removing stormwater from site.
           (4) Identify any site constraints (i.e., required area set aside for future expansion, other client
               land uses).
           (5) Identify structures to be demolished.
           (6) Determine general fencing requirements and whether motorized gates are desired.
           (7) Identify extent of landscaping if desired by client.
           (8) Identify 100-year flood plain elevation  if applicable.
3.2.1.5     Layout of Structure
           (1) Identify approximate structure size and preliminary location of rooms and/or major
               equipment on a floor plan.
           (2) Determine building(s) use group, fire resistance ratings, ceilings,  stairwells, height and
               area restrictions, special fire and life safety requirements, and means of egress strategy to
               at least the level that they will affect preliminary building layouts and costs. Address
               requirements of the Americans with Disabilities Act (ADA).
           (3) Coordinate location and layout of chlorine, acid and base rooms.
           (4) Identify particular client preferences early for architectural details.
           (5) Determine architectural style and requirements, with consideration to insulation
               requirements:
                                                21

-------
               (a) Wall construction (i.e., brick and block, concrete block, glazed structural block,
                   sound block, metal siding, pre-engineered, aggregate panels).
               (b) Roof construction (i.e., pre-cast concrete, poured-in-place concrete, steel deck and
                   bar joists, wood trusses). Consider type of structure and its interior use (i.e., wet
                   areas, chemical feed area, etc.).
               (c) Roofing materials (i.e., single-ply ballasted or adhered membrane, built up, shingles,
                   metal).
               (d) Windows (i.e., natural light, ventilation, aesthetics). Match or replace existing
                   windows: material (i.e., aluminum, steel, wood, vinyl) and/or finish (i.e., anodized,
                   painted, primed).
               (e) Doors. Match or replace existing doors: material (i.e., hollow metal, aluminum, FRP,
                   stainless steel, wood, acoustical).
               (f) Overhead and/or roll-up doors.  Identify electric operator versus manual lift doors.

           (6) Provide room finish schedules based on client input. Items to include are listed as
               follows:
               (a) Interior wall construction (non-load bearing); material (i.e., concrete block, glazed
                   block, steel or wood stud walls); finishes (i.e., unfinished, painted, gypsum board,
                   wallpaper, paneling, chair railing, molding at ceiling and floor).
               (b) Flooring. Unfinished or sealed concrete, seamless floor covering, vinyl, carpeting,
                   tile (i.e., thin-set or thick-set), terrazzo, applied composite material with urethane
                   overcoats, embedded steel mats where heavy steel wheel loads  are anticipated (i.e.,
                   dumpster containers).
               (c) Ceilings. Material and finishes.
           (7) Identify stair type (i.e., concrete pan, metal, cast  in place).
           (8) Identify method of removing rainwater from roofs of each building and point of
               discharge (i.e., roof drains, gutters and downspouts, roof scuppers discharging to ground,
               or storm sewers).
           (9) Identify locations of rest rooms (for both genders) in building.
           (10)  Identify locations of drinking fountains and coolers.
           (11)  Identify areas where service sinks or portable sampler wash down basins will be
                 provided.
           (12)  Specify grating material (i.e., aluminum,  steel, FRP such as in certain chemical feed
                 and fill areas).
           (13)  Determine extent of laboratory improvements.
           (14)  Identify any existing structures to be re-roofed or repainted.
           (15)  Write preliminary outline of requirements for OSHA (i.e.,  signing, color coding, fire
                 extinguishers) and ADA.
3.2.1.6     Structural
           (1) Identify local code requirements for seismic design, frost depths, wind loads, and snow
               loadings.
           (2) Identify design of live load requirements for stairway, office, and corridor floors. Also
               floor loadings for operating and storage areas.
           (3) Identify design for water, earth, and live load requirements for foundation walls.
                                                22

-------
           (4) Identify likely areas where peripheral drains and hydrostatic pressure relief valves will be
               necessary to prevent flotation and reduce exterior pressures (if high groundwater
               conditions are known to exist prior to obtaining soil boring data).
           (5) Identify requirements for protection of existing adjacent structure foundations that could
               be damaged during excavation.

           (6) Identify any material handling that is required (monorails, crane, davit, dock access,
               eyebolts) and approximate lifting capacities.
           (7) Identify major equipment and provide approximate weights (i.e., pumps, blowers,
               generators, engines).
           (8) Note any structural repairs required in existing buildings or any new or enlarged wall or
               floor openings. Note any concrete repairs or masonry rehabilitation and coordinate with
               client.
           (9) Identify design strength criteria for reinforced concrete and steel.

3.2.1.7     Mechanics

           (1) For heating, ventilating, and air conditioning (HVAC) and other mechanical building
               systems, identify any special or specific expectations or the client.
           (2) Identify energy source(s) to be used for providing building heat (i.e., natural gas or
               electric) and supplier(s).
           (3) State method of providing heat to each structure, building, or section of building such as
               a lab or office area. Identify preliminary location of central heating and cooling facilities.
           (4) Identify ventilation method for each building and preliminary location of exhaust fans,
               louvers, air handling systems and ventilation rate criteria (air changes, cfm/ft2,
               cfm/person).

           (5) Provide conceptual strategy for dealing with dust control, explosion resistance, fire
               protection, humidity control, emergency showers and/or eyewash, and hazard detection
               interlocks with ventilation. Describe equipment to be provided.
           (6) Identify mechanical building system requirements for generator and engine rooms
               (ventilation, combustion air, cooling system strategy, fuel system and storage, and
               drainage).
           (7) Identify areas to be air conditioned or de-humidified.

3.2.1.8     Electrical

          (1)   Provide any special or specific expectations of the client. Note any problems with
               existing equipment, if applicable, or certain manufacturer's equipment to be used or not
               used.
          (2)   Identify power supply source.
          (3)   Identify source and location of emergency power generator if required.
          (4)   Provide general control descriptions that will be used to develop loop descriptions for
               automatic controls.
          (5)   Complete an "Equipment and Controls Listing" as completely as possible.
                                                23

-------
         (6)   Confirm instrumentation and control philosophy with the client (i.e., completely manual,
               data acquisition and logging with manual control, automated control of specific
               equipment or processes, or completely automated).
         (7)   Identify work required at remote site from the project site (i.e., lift stations, well sites,
               booster stations, elevated tanks, other plants).
         (8)   Identify equipment that is to be driven by variable speed systems.
         (9)   Determine whether plant power distribution is to be overhead and/or underground.
         (10)  Identify if existing lighting is to be revised with the client.
         (11)  Identify method of providing outdoor lighting (i.e., high mast lights, pole-mounted street
               lights, or wall-mounted exterior building lights).
         (12)  Identify whether Process and Instrumentation Diagram (P&ID) drawings are required and
               how many there will be.
         (13)  Identify areas where electrical equipment including computers must be located in rooms
               with special temperature or humidity environments.
         (14)  Identify pumps requiring seal water systems with solenoid valves, pressure switches, and
               controls for alarm/lockout.

3.2.2       Treatment Equipment. This stage of the preliminary design includes determining the
system operating mode (i.e., manual or automatic), sizing the equipment, selecting materials of
construction, determining an equipment layout, and upgrading the preliminary capital cost estimate to a
±20% accuracy.  Key deliverable items include, but not limited to:

           (1) Schematic flow diagram (see Figure 3-2 for example)
           (2) Preliminary process equipment arrangement drawings (see Figure 3-3 for examples)
           (3) Outline specifications
           (4) Preliminary capital cost estimate.

3.2.2.1     Manual or Automatic Operation. AA fluoride removal systems can be operated either
manually or automatically. In a manual operation, the  treatment plant operator personally performs all of
the operating functions and makes all operating decisions. The treatment plant equipment does not
accomplish any function independent of the operating personnel. The equipment is simple and performs
the basic functions that the operator implements.  The manual operation includes the following:

           (1) Motors (pumps, chemical pumps, etc.) with manual start/stop controls. Some motors
               have manual speed adjustment capability. Chemical pumps have manual speed and
               stroke length adjustment capability.
           (2) Valves with manual handle, lever, handwheel, or chainwheel operators.
           (3) Instrumentation sensors with indicators. Instrumentation is installed in-line where
               operating data (flowrate, total flow, pressure, pH, and liquid levels) are indicated. In-line
               pH sensors, magmeters, ultrasonic level sensors are other instruments that require electric
               service.

In an automatic operation, the treatment plant is operated by a Programmable Logic Controller (PLC),
which initially is programmed by the operator, the computer supplier, or an outside specialist.  If pro-
grammed by someone other than the plant operator, the operator should be trained by that individual to
adjust program variables and, if necessary, modify the program. The operator interface and printer are the
                                               24

-------
                             ACID
                         STORAGE TANK
                                                      o
   CAUSTIC
STORAGE TANK
                                                                            CAUSTIC DAY TANK
                                                                           FOR TREATED WATER
                                                                           Ph ADJUSTMENT AND
                                                                              SPENT MEDIA
                                                                              REGENERATION
                                                                             (OPTIONAL)
                                                                              BYPASS
                                                                             REGENERATION
                                                                          *•   WASTEWATER
                                                                            TO SURGE TANK
                                                                              (NOT SHOWN)
                                                                              SAMPLE
                                                                               PANEL
Figure 3-3. Treatment System Plan for an Activated Alumina Plant
                                 25

-------
equipment items the operator uses during the performance of treatment plant functions.  In addition, the
operator should calibrate and check all of the components of the automatic operating equipment system
on a routine periodic basis. Finally, the treatment plant operator or a designated instrumentation and
control specialist should be capable of performing emergency maintenance and/ or repair of all
components.

Every function included in an automatic system should be capable of manual operation.

The automatic equipment is more sophisticated and costly than that used in a manual operation. When
functioning normally, an automatic operation can function continuously with minimal operator attention.
This is recommended for treatment systems in  remote areas or areas that are difficult to access, and
systems for which operator availability is limited. The automatic  operation includes the following:

           (1)  Motors for pumps, chemical pumps, air compressors, etc., are automatically turned on
               and off and may have speed adjustment controls.  Chemical pumps may have a manual
               stroke length adjustment but can be paced by the  flow and on/off operation of the plant.
           (2)  Valves with either pneumatic/hydraulic or electric operation are required on the
               equipment.  Valves require manual overrides during startup, power failure or compressed
               air failure. Valves should have opening and closing speed controls to prevent water
               hammer during automatic operation, especially on pump systems.
           (3)  Automatic instrumentation may be electronic, pneumatic, or a combination of both. The
               instruments and controls should always be capable of transmitting and receiving
               electronic information to and from the PLC. In a fully automatic system all of the
               control, monitoring, and alarm functions are monitored and controlled by the PLC.
               Backup manual instruments (e.g., flowrate indicators, pressure indicators, pH indicators,
               and liquid level indicators) are recommended to provide verification of automatic instru-
               mentation if treatment plant budget is available. Comprehensive automatic alarms that
               notify operators and/or shut down increments or the entire treatment system relating to
               every type of system malfunction at the moment such events occur is a necessary
               function that should be incorporated in all applicable instrumentation components.

A semiautomatic operation which employs individual controllers to automatically start/stop or adjust
some, but not all, of the operational items in the system can contribute significantly to the treatment
system operation without  computer control of the entire operation. These semiautomatic functions should
include alarms that will notify operators of process functions exceeding limits established for effective
and/or safe operation.  Alarm events can be staged at single (e.g.,  high) or dual (e.g., high-high) levels. In
a dual-level alarm, the first level notifies the operator that the performance is out of tolerance, and the
second level shuts down either a single process function (e.g., a pump) or the entire process. Examples of
semiautomatic operational functions include, but are not limited to, the following:

           (1)  Flow control loop includes an  electronic flow sensor with totalizer (e.g., magnetic
               flowmeter) that sends an electronic signal to an electronic flow controller (with high and
               low flowrate alarms), which in turn sends an electronic signal to a flow control valve
               (butterfly valve or ball valve) with an actuator and electronic positioner.  The plant
               operator designates the required flowrate at the flow controller. The controller receives
               the flowrate measurement from the flow sensor and transmits signals to the flow control
               valve positioner to adjust the valve position until  the flowrate matches that required by
               the process.  If the flowrate deviates from the limits established for the process, then a
               high flowrate or low flowrate alarm will be issued.
                                               26

-------
           (2) Pressure control loop includes an electronic pressure transmitter that sends an electronic
               signal to an electronic pressure controller (with high and low pressure alarms), which in
               turn sends an electronic signal to a pressure control valve with an actuator and electronic
               positioner.  The plant operator designates the required pressure at the pressure controller.
               The controller receives the pressure measurement from the pressure transmitter and
               transmits signals to the pressure control valve positioner to adjust the valve position until
               the pressure matches that required by the process. If the pressure deviates from the limits
               established for the process, then a high pressure or low pressure alarm should be  issued.
           (3) pH control loop includes an electronic pH sensor which transmits a pH signal to a pH
               analyzer (with high and low level alarms) which in turn sends an electronic signal to a
               converter which transmits a pulse signal to a chemical feed pump (acid or caustic) to
               adjust the feed pump stroke speed. The plant operator designates the required pH at the
               pH analyzer.  The pH analyzer receives the pH measurement from the pH sensor and
               transmits signals to the chemical feed pump (via the converter) to adjust the pump stroke
               speed until the pH matches that required by the process. If the pH deviates from the
               limits established for the process, then a high pH or low pH alarm should be issued.
           (4) Liquid level control loop includes an electronic liquid level  sensor (e.g., ultrasonic level
               sensor) which transmits an electronic liquid level signal to a level controller which
               indicates the liquid level and transmits  an electronic signal to one or more motors (pump,
               mixer, etc.) to start or stop.  At the level controller the plant operator designates the
               required liquid levels at which motors are to start and stop.  The level controller receives
               the liquid level measurement from the liquid level sensor and transmits signals to the
               motor(s) to start or stop. If the liquid level deviates from the limits established for the
               process, then a high or low liquid level alarm should be issued.

Many other process functions are performed automatically by means of relays and other electrical
devices.  An example is the electrical interlock of chemical feed pumps with raw water pumps, which
prevents chemical feed into the process without the flow of process water. Another example is the use of
a flow switch in a pressure relief valve discharge pipe, which, upon detection of water flow, issues an
alarm and stops the process feed pump. The list of individual failsafe automatic functions can be
extensive.  All applicable codes, standards, and OSHA requirements should  be reviewed to determine
which requirements are applicable to the project. Then based upon sound judgment,  available budget,
treatment plant operator capability, and availability, a decision  should be made as to whether a given
function should be automatic or manual.

3.2.2.2     Treatment Equipment Preliminary Design. This stage of the preliminary design
encompasses sizing of equipment items and selecting materials of construction. An example with the use
of dual vessels in parallel each with a conservative EBCT of 7.5 min in a downflow mode is provided in
Appendix B (comparing this EBCT with the EBCTs of several existing AA plants on Table D-l, which
range from 5.7 to 11.5 min). For automatic or semiautomatic operation the system basic design does not
change; however, equipment material and installation costs will vary.

Media Bed and Vessel Design. In accordance with the discussion presented in Section 3.1.2, the
recommended treatment concept is based on the use of two pressure vessels piped in parallel using the
downflow treatment mode. Treatment vessel piping is configured to provide for media backwashing
(upflow) and spent media regeneration. Materials of construction employed in the design example
presented in Appendix B are carbon steel (grade selection based on cost-effective availability) that complies
with American Society of Mechanical Engineers (ASME) Code  Section VIII, Division 1. The interior of the
vessels is lined with abrasion-resistant vinyl ester or epoxy coating.  Interior lining material is NSF
International-certified for potable water applications, and suitable for pH range 2.0-13.5. Vessel pressure
                                               27

-------
rating is 50 psig or the minimum necessary to satisfy system requirements.  Other vessel materials of
construction (e.g., fiberglass and fiber re-enforced plastic [FRB]), internal lining materials (e.g., abrasion
resistant epoxy, rubber, etc.), and stainless steel without lining, may also be used.

Prior experience with AA indicates that the volume of the media (V) in each of the two parallel vessels is
1 ft3 per gpm of process water flowrate, which provides a conservative EBCT of 7.5 min. Actual
residence time is approximately half the EBCT, because the space between the grains of media is
approximately 50% of the total bed volume.  When multiple beds are used, the volume of treatment media
per unit is equal to the total treatment flowrate divided by the number of treatment beds (N). (Note:
When raw water is bypassed and blended back with treated water, only the treated water is included in
sizing the treatment media volume.) In order to prevent "wall effects", bed diameter (d) should be equal
to or greater than the  bed depth (h). Good practice indicates that the bed depth should be a minimum of 3
ft and a maximum of 6 ft. At less than minimum depth, distribution problems may develop; and, at
greater than maximum depth, fine material removal and pressure loss becomes a problem.  For very small
systems using tanks of 1-2 ft in diameter, the bed depths could be as low as 2 ft. A typical treatment bed
and vessel design is illustrated in Figure 3-4. A summary of AA plant design and operation surveyed
during the initial stage of this manual revision project is presented in Appendix C.

For fluoride removal, AA manufacturers commonly recommend  a 5-min EBCT across each of two lead-
lag beds (Reid, 2013). A more conservative EBCT of 7.5 min is  also used to design adsorption vessels
configured in series (Rubel, 1984). As the EBCT decreases below the recommended value, two
undesirable outcomes may occur. First, the treatment is less efficient, resulting in treated water fluoride
concentration not reaching a low enough concentration. Second, regeneration frequency increases,
requiring more chemicals, operating cost, operator attention, and proportionately more downtime.
Conversely, raising the EBCT above the recommended level increases the size of the treatment beds and
their vessels, thereby increasing capital cost and space requirements.

Pressure vessel fabrication is standardized by diameter in multiples of 6-in outside diameter increments.
Tooling for manufacture of pressure vessel dished heads is set up for that  standard. Design dimensions
differentiate between pressure vessel and treatment bed diameters.  The vessel outside diameter (D) is
approximately 1 in. greater than the bed (or vessel inside) diameter, which conservatively provides for
both vessel walls with lining as well as fabrication tolerances.  If the pressure is high (100 psig or
greater), the 1 in. thickness will increase to reflect the increased vessel wall thickness.

Although many methods are available for distributing the water flow through a treatment bed, the
following method has been successfully used in adsorptive media plants that are presently in operation.
The water is piped downward into the vessel through an inlet diffuser.  This diverts the flow into a
horizontal pattern.  From there it radiates  in a horizontal plane prior to starting its downward flow through
the adsorptive  media  bed.  The bed, in turn, is supported by a false flat bottom, which is supported by the
bottom head of the pressure vessel by means of concentric rings.  The false flat bottom also supports the
horizontal header and plastic fabric sleeved perforated lateral collection system. Treatment media are
placed in the vessel through circular manway(s) with hinged cover(s) in the top head of the  vessel.
                                               28

-------
FREEBOARD

50% BED
EXPANSION
TREATMENT
MEDIA
            h/2
                                TREATMENT VESSEL
                                                                   SS
                                                              Minimum clearance
                                                              required  between
                                                              bottom  of  vessel  and
                                                              concrete pad
    q
    d
    h
    V
    wd
    Mw
    D
    dn
    H
    SS
GIVEN
    d >
    H =
    D =
   SYMBOLS
- TREATED WATER  FLOW  RATE  (gpm)
- TREATMENT BED  DIAMETER  (ft.),   d  =
- TREATMENT BED  DEPTH  (ft.)
- TREATMENT BED  VOLUME - jdii.  (ft.3 )
- DENSITY OF TREATMENT MEDIA  (Ib./ft.3 )
- WEIGHT OF MEDIA  (Ibs.)
- OUTSIDE DIAMETER OF  TREATMENT VESSEL  (ft.)
- DEPTH OF DISHED PRESSURE HEAD  (ft.)
- OVERALL HEIGHT OF SKID MOUNTED TREATMENT VESSEL
- STRAIGHT SIDE   (ft.)
          h/2,  3'-0" < h <  6'-0"
          2 dH +  h  + h/2 + 6" +  1"
          d +  1"
            48 Ib/tt3   (VARIES WITH MEDIA  IN VESSEL)
          M. x v  =  t5V   (Ib.)
                                                           (ft.)
                Figure 3-4. Treatment Bed and Vessel Design Calculations
                                            29

-------
Pipe Design. Because AA systems require process water pH adjustment and spent media regeneration,
pipe material should be suitable for ambient temperature, pH range of 2.0 to 13.5, system pressure, and
potable water service. At a low pH, carbon steel is not acceptable unless interior lining is included.
Stainless steel is acceptable; however, it may be too costly. Plastic materials such as polyvinyl chloride
(PVC), polypropylene, and high density polyethylene (HDPE) are satisfactory.  PVC is usually the best
selection based on its availability, NSF certification for potable water service, low cost, and ease of
fabrication and assembly.  The drawbacks to the PVC materials are their loss of strength at elevated
temperatures (above 100°F); their coefficients of thermal expansion; their external support requirements;
their deterioration from exposure to sunlight; and their vulnerabilities to damage from impact. Neverthe-
less, these liabilities are outweighed by the low cost and suitability for the service.  The piping can easily
be protected from all of the above concerns, except elevated ambient and/or water temperatures. If
elevated temperature exists, the use of FRP pipe is recommended. This material provides the strength and
support that is lacking in the pure plastic materials.

The piping system should be economically sized to allow for delivery of design flow without excessive
pressure losses.  If water velocities present conditions for water hammer (due to fast closing valves, etc.),
shock-preventing devices should be provided.

Isolation and process control valves should be wafer style butterfly type, except in low flowrate systems
where small pipe size dictates the use of true union ball valves. The use of inexpensive, easily maintained
valves that operate manually provides minimum capital cost.  The valves are automated by the inclusion
of pneumatic or electric operators.

Pressure regulator and rate of flow control  valves are recommended for safe operation of manually
controlled treatment systems.

See Appendix B for pipe size  design using the example employed for vessel and treatment media design.

Instrumentation Design. System functional requirements that are adapted to commercially available
instruments should be specified. Included are:

Instrument                  Range        Accuracy
1.  Flow sensor                 Varies(a)    ±2%
(indicator/totalizer)
2.  Pressure indicator            Varies(a)    ±1%
3.  pH sensor/analyzer/alarm     0-14        ±0.1
4.  Level sensor/indicator        Varies(a-1    ±1%
5.  Temperature indicator        30-120°F    ±1%
(optional)
(a) Range to be compatible with application, maximum measurement not to exceed 90% of range.

Acid Storage and Feed Subsystem.  Acid feed and storage  subsystems are included for pH adjustments of
inlet water.  The acid storage tank should be sized to contain tank truck bulk delivery quantities of
concentrated sulfuric acid. For water systems that are not permitted to increase the sulfate concentration
of the water, hydrochloric acid can be substituted. However, this acid is more costly, more difficult to
handle, and results in highly corrosive treated water; therefore hydrochloric acid is not recommended.
Bulk  delivery provides the lowest unit price for the chemical.  In  small plants, acid consumption may not
be enough to justify large volume purchase of chemicals. In the smaller plants, drums or even carboys
may be more practical; therefore, for that type of operation, the requirement for a storage tank is
eliminated.  A 48,000-lb tank truck delivers 3,100 gal of 66°B' H2SO4 (15.5 Ib/gal). A 5,000-gal tank
                                               30

-------
provides a 50% cushion. The example in Appendix B illustrates the method of designing the components
of this system.

The sulfuric acid storage tank can be constructed of lined carbon steel or chlorinated polyvinyl chloride
(CPVC). The storage tank should be protected from the elements and include a containment basin located
outside of the treatment building. Typically, the containment basins are sized for 110% of the capacity of
the storage tank. The 66°B' H2SO4 freezes at -20°F. Therefore, unless the treatment plant is located in
an extremely cold climate, no freeze protection is required.  All piping is to be 2-in. carbon steel with
threaded cast iron fittings and plug valves.  Elastomer seals, seats and gaskets should be Viton®.

The acid pumps are standard diaphragm models with materials of construction suitable for 66°B' H2SO4
service. Standard sulfuric acid service pumps should be specified.  In the preliminary design, the sizing is
determined by field test or theoretical calculation (see Appendix B). Acid feedrate varies with total alka-
linity and free €62 content of the raw water.  The feedrate is accurately determined experimentally by
adjusting a raw water sample pH to 5.5 by acid titration. In a manual treatment plant operation, the
operator should check the pH periodically and maintain it at 5.5. The pump stroke speed  and length
should be adjustable to  accommodate these variations.  An in-line static mixer should be installed
immediately downstream of each acid injection point. This provides thorough mixing of the acid, which
results in an accurate pH measurement by a pH sensor located at the discharge end of the  mixer. The pH
probes that are used to control pH should be calibrated against standard buffers at least once per week.

Immediately after spent media regeneration, an acid feed is required to lower pH of water for neutraliza-
tion of a treatment bed prior to placing that bed back into treatment service. Neutralization pH is initially
set at 2.5 and increases  in steps until the treatment pH of 5.5 is achieved.  A rinse step using the source or
treated water can be included ahead of the neutralization step to remove residual fluoride  and silica along
with the dissolved alumina. Finally wastewater from the regeneration of an AA bed is collected in a
surge tank where the pH is adjusted to near neutral.  An additional acid feed pump is required to feed acid
to the wastewater.

Caustic Soda Storage and Feed Subsystem. The caustic soda storage tank also is sized to contain tank
truck bulk delivery quantities of 50% or 25% sodium hydroxide. A 48,000-lb tank truck delivers 3,850
gal of 50% NaOH which provides a 25% cushion in a 5,000-gal storage tank. 50% NaOH freezes at
55°F; 25% NaOH freezes at 0°F. Therefore, 50% NaOH, which is preferable because of price, requires
heating to prevent freezing.

The caustic is used for treatment bed regeneration and pH adjustment of treated water.  Regeneration
frequency is a function  of raw water fluoride concentration, flowrate and AA fluoride capacity. The
amount of caustic required to neutralize the treated water, that is to raise the pH from 5.5 to the pH
required for corrosion protection for the water system, is a function of the water chemistry at each
installation.  The actual caustic feedrate is easily determined experimentally by readjusting the treated
water pH by titrating a sample with caustic until the desired pH is achieved. If a fraction of the raw water
bypasses treatment and is blended with treated water, then the chemical required for pH adjustment is
reduced.

In raw water with high alkalinity the lowering of pH produces high levels of dissolved CO2. In those
waters, removal of the CO2 by aeration raises the pH (prior to blending), providing a less expensive
treatment due to reduction of caustic required to raise the pH of the treated water. In low alkalinity water,
the chemical addition is less expensive. The carbon steel caustic storage tank is covered in Appendix B.
This vessel should be heat-treated to stress relieve welds. The carbon steel does not require an interior
lining; however, it does require sandblasting and vacuum-cleaning prior to filling. All piping is to be 2-
                                               31

-------
inch carbon steel with threaded cast iron fittings and plug valves.  Elastomer seals, slots and gaskets
should be ethylene propylene diene monomer (EPDM).

Because 50% NaOH freezes at 55°F, it should maintain a minimum temperature of 70°F. This can be
handled by a temperature-controlled electrical immersion heater.  Insulated tanks with wall heaters are
also available to store and heat the caustic soda. Twenty-five percent sodium hydroxide freezes at 0°F;
therefore, unless it is located in an extremely cold climate, freeze  protection is not required. The storage
tank should be placed in a containment basin inside of an enclosure outside of the treatment building.

A pump is required to feed caustic into the effluent main through  an in-line static mixer where the treated
water is neutralized. For regeneration, a larger caustic feed pump is required for pumping the caustic
through a static mixer in the regeneration feed pipe.  There the caustic is diluted to the 1% (by weight)
concentration required to regenerate the adsorptive treatment media.

Wastewater Lined Evaporation Pond. In the example used in Appendix B, it is assumed that the most
cost-effectively and preferred wastewater disposal option is a lined evaporation pond. This method,
however, can be used only in arid regions in the desert southwest. It is not a viable method in the humid
southeast or cold climate of the northern tier of states. In those areas a viable disposal option is to
neutralize the regeneration wastewater with acid as it leaves the treatment vessel and collect the entire
regeneration wastewater batch in a surge tank. The neutralized wastewater is then bled at a controlled
flowrate to the sanitary sewer. In the sewer, it blends with the defluorinated water that has  been
discharged to waste.

To size the lined evaporation pond, the basic information required is the average annual volume of
regeneration wastewater to be evaporated and the average annual  evaporation rate. The former can be
estimated and the  latter can be obtained from the national weather bureau.  The treatment plant production
is normally much  higher in summer than winter, and evaporation  rate is also  higher in summer.  The
ponds have sloped sides, pond depth to be 8ft minimum. Ponds are to be lined with 30 mil  reinforced
hypalon, a material that is not vulnerable  to ultraviolet radiation deterioration or exposure to pH 12. The
dissolved solids will concentrate and precipitate in the pond.

3.2.3      Preliminary Treatment Equipment Arrangement.  Once all of the major equipment size
and configuration information is available, a layout (arrangement  drawing) is prepared.  The layout
provides sufficient space for proper installation, operation and maintenance for the treatment system as
well as each individual equipment item. OSHA standards should  be applied to these decisions during the
equipment arrangement design stage.  These requirements may be supplemented or superseded by state or
local health and safety regulations, or, in some cases, insurance regulations.  A compact arrangement to
minimize space and resulting cost  requirements is recommended.  Figure 3-3 illustrates typical
preliminary arrangement plans. These arrangements provide no frills, but do include ample space for ease
of operation and maintenance. Easy access to all valves and instruments reduces  plant operator effort.

The type of building used to protect the treatment system (and operator) from the elements  depends on the
climate. Standard pre-engineered steel buildings are low-cost, modular units. Concrete block or other
material also may be used. Standard building dimensions that satisfy the installation, operation, and
maintenance space requirements for the treatment system should be selected. The building should
provide access doors, lighting, ventilation, emergency shower and eye wash, and  a laboratory bench with
sink. All other features are optional.

When the arrangement is completed, the preliminary cost estimate is prepared.
                                               32

-------
Manual operation is the method employed in the design example in Appendix B. The basic process
requirements should be reviewed at each stage of design to assure that every item required to operate the
process is included. Although detail design occurs during the final design phase, provision for operator
access for every equipment item should be provided. Automatic operation does not require total
accessibility; access for maintenance functions for which ladder or scaffold access will suffice. The extra
equipment items required solely for automatic operation (including but not limited to PLC, and operator,
interface) occupy minimal space and are located in positions that are most accessible to the operator.

3.2.4       Preliminary Cost Estimate. At completion of the preliminary design,  the preliminary cost
estimate is prepared based upon the equipment that has been selected, the equipment arrangement and the
building selection. This estimate should be based on the material equipment quantities, unit prices to
labor and material, and finally summarized in a format that is preferred by the owner. This estimate
should have an accuracy of ±20%. To assure sufficient budget for the project, it is prudent to estimate on
the high side at this stage of design. This may be accomplished by means of a contingency to cover
unforeseen costs, and an inflation escalation factor.

3.2.5       Preliminary Design Revisions.  The preliminary design package (described above) then is
submitted for approval prior to proceeding with the final design. This package may require the approval
of regulatory authorities, as well as the owner. Requested acceptable changes should be incorporated and
resubmit for approval. Once all requested changes are implemented  and preliminary design approval is
received, the final design can proceed.

3.3        Final Design

After completion and approval of the preliminary design by the client, the final design proceeds. This
includes detail design of all of the process equipment and piping, complete process system analysis,
complete detail design of the building including site work, and a final capital cost estimate accurate to
within 10%. The deliverable items are:

           (1) Complete  set of construction plans and specifications
           (2) Final capital cost estimate.

The final design starts with the treatment system equipment (if applicable, including the wastewater surge
tank); continues with the building (including concrete slabs and foundations, earthwork excavation/back-
fill/compaction, heating, cooling, painting, lighting, utilities, laboratory, personnel facilities, etc.); and
finishes with the site work (including utilities, drainage, paving and landscaping). The latter items apply
to every type of treatment plant; although they are integral with the treatment system, they are not
addressed in this manual. The only portions of the final design that should be addressed are the pertinent
aspects of the treatment equipment which were not covered in the preliminary design.  During the
conceptual  design and preliminary design, the basic equipment that accomplishes the required functions
were selected,  sized, and arranged in a compact, efficient layout. The decision was cost-conscious, using
minimum sizes (or standard sizes) and the least expensive materials that satisfied the service and/or
environment. However, in the final design, this effort can be defeated by not heeding simple basic cost
control principles.  Some of these are:

           (1) Minimize detail (e.g., pipe supports-use one style, one material, and components
               common to all sizes).
           (2) Minimize the number of bends in pipe runs (some bends are necessary-those that are
               optional only increase costs).
                                               33

-------
           (3) Minimize field labor; shop fabricate where possible (e.g., access platforms and pipe
               supports can be mounted on brackets that are shop fabricated on vessel).
           (4) Skid-mount major equipment items (skids distribute weight of vessels over small mat
               foundations in place of piers and spread footings, thereby costly foundation work is
               eliminated).
           (5) Use treatment vessels as a heat sink to provide insulated building cooling or heating or
               both (eliminates heating and/or cooling equipment in addition to reducing energy cost).
               Consideration must be given, however, to humid climates where cold tanks will result in
               sweating problems.
           (6) Simplify everything.

All subsystems should be analyzed  (refer to schematic flow diagrams in Figure 3-2) to account for all
components in both equipment specifications and installation drawings. The specifications and drawings
should provide all information necessary to fabricate and install the equipment. Extra effort to eliminate
ambiguity in detail and/or specified requirements should be exercised. All items should be satisfactory
for service conditions besides being able to perform required functions. Each item should be easy to
maintain; spare parts necessary for continuous operation should be included with the original equipment.
All tools required for initial startup  as well as operation and maintenance should be furnished during the
construction phase of the  project. After all components in each of the subsystems have been selected,
hydraulic analysis should be made to determine velocities and pressure drops through the system. Calcu-
lations should be run for normal treatment flow and backwash flow. The latter is more  severe, but of
short duration.  If pressure losses are excessive, the design should be modified by decreasing or
eliminating losses (e.g., increase pipe size, eliminate bends or restrictions, etc.).

Upon completion of system installation, functional checkout requirements should be accomplished. All
piping should be cleaned  and pressure tested prior to startup. All leaks should be corrected and retested.
Recommended test pressure is 150% of design pressure. Potable water piping and vessels should be
disinfected prior to startup.  Disinfection procedures should be  in compliance with regulatory agency
requirements and material manufacturer's disinfection requirements/limitations.  All electrical systems
should satisfy a functional checkout. All instruments should be calibrated; if accuracy does not meet
requirements stated under instrumentation design in Section 3.2.2.2, the instruments are to be replaced.
When the plant operation begins, a  check on actual system pressure drop is required. If there is a discrep-
ancy between design and actual pressure drop, the cause should be determined (obstruction in line, faulty
valve, installation error, design error, etc.) and rectified.  Pressure relief valves should be tested; if not
accurate, they should be adjusted or replaced. Although this activity takes place during treatment plant
startup, it should be incorporated as a construction document requirement.

3.3.1       Treatment Equipment Final Design

3.3.1.1     Treatment Bed and Vessel Design.  The required AA media volume is determined by bed
dimensions and resulting  weight in  the preliminary design. It is recommended that a minimum of 10%
extra media be ordered. For lowest price and ease of handling, the media should be ordered in fiber
drums (approximately 5 to 8 ft3) on pallets or in supersacks (approximately 40 ft3 per supersack). The
media should be NSF-certified for potable water application and demonstrate fluoride removal capacity.
The commonly available AA products on the market include Alcan 400G and Alcoa CPN.

Each media vessel  should have a support system to transfer its loaded weight to the foundation and
ultimately to the soil. The loaded weight includes the media, the water, attached appurtenances (platform,
pipe filled with liquid, etc.), the vessel, and applicable seismic and/or wind loads. The support legs
                                               34

-------
should be as short as possible to reduce head room requirements as well as cost. If the equipment is skid-
mounted, the vessel legs should be integral with the skid to distribute the weight over an area greater than
the dimension of the vessel.  This distribution eliminates point loads of vessel support legs, so costly
piers, footings, and excavation requirements are eliminated. The skid should have provisions for anchor-
age to the foundation. Exterior brackets (if uniform and simply detailed) are not costly and provide
supports that eliminate need  for cumbersome costly field fabrications. Conversely, interior brackets,
though required to anchor (or support) vessel internal distribution or collection systems, should be held to
a bare minimum because they are costly to line. Epoxy (or rubber) linings with abrasion resistance qual-
ities  are recommended. Vessel interior lining should extend through vessel opening out to the outside
edge of flange faces. Alternatively, vessels may be constructed of stainless steel (no lining required).
Openings in the vessels should be limited to the following:

           (1) Influent pipe - enters vertically at center of top head.
           (2) Effluent pipe - exits horizontally through vertical straight side immediately above false
               flat bottom in front of vessel, or vertically at the center of the  bottom head.
           (3) Air/vacuum  valve (vent) - mounts vertically on top head adjacent to influent pipe.
           (4) Media removal - exits horizontally through vertical straight side immediately above false
               flat bottom at orientation assigned to this function. Some AA plants do not remove
               media, which attrites due to backwash and  regeneration.  These plants routinely top off
               the vessels with virgin AA.
           (5) Manway - 16-in diameter (minimum) mounted on top head with center line located
               within 3 ft of center of vessel and oriented  toward work platform.  Manway cover to be
               hinged or davited.

It is recommended that pad flanges be used for pipe openings in place of nozzles.  Pad flanges are flanges
that are integral with the tank wall. The exterior faces are drilled and tapped for threaded studs. The pad
flanges save the cost of material and labor, and are much easier to line;  they also reduce the  dimensional
requirements of the vessel. The vessel also requires lifting  lugs suitable for handling the weight of the
empty vessel during installation. Once installed, the vessel should be shimmed and leveled. All space
between the bottom surface of the skid structure and the foundation should be sealed with an expansion-
type  grout; provisions should be included to drain the area under the vessel.

The type of vessel internal distribution and collection piping used in operational fluoride removal plants is
defined in the preliminary design. Because there are many  acceptable vessel internal  design concepts,
configuration details are left  to sound engineering judgment.  The main points to consider in the design
are as follows:

           (1) Maintain uniform distribution

           (2) Provide  minimum pressure drop through internal piping (but sufficient to assure uniform
               distribution)
           (3) Prevent wall effects and channeling
           (4) Collect treated water within 2 in. of bottom of treatment bed
           (5) Anchor internal piping components to vessel to prevent any horizontal or vertical
               movement during operation
           (6) Ensure that construction materials are suitable for pH range of 2.0 to  13.5 (PVC, stainless
               steel are acceptable).
                                                35

-------
Underdrain failures create significant problems; treatment media loss, service disruption and labor to
repair problems are very costly. A service platform with access ladder or a roof hatch is required for use
in loading AA media into the vessel.  Handrail, toe plate, and other OSHA-required features should be
included.

3.3.1.2     Pipe Design.  Each piping subsystem should be reviewed to select each of the subsystem
components.  Exclusive of the chemical subsystem, five piping subsystems and two optional subsystems
are listed in the general plan; they are:

           (1) Raw water influent main
           (2) Intervessel pipe manifold
           (3) Treated water effluent main
           (4) Raw water bypass main
           (5) Backwash regeneration feed main (optional)
           (6) Wastewater main
           (7) Sample panel (optional).

The detail design now proceeds for each of those subsystems. First, the equipment specification for each
equipment component in each subsystem should be defined. This is followed by a detailed  installation
drawing, which locates each component and provides access for operation and maintenance. As each
subsystem nears completion, provisions for pipe system support and anchorage, as well as for thermal
expansion/contraction, should be incorporated in the detail design.

The interface where the concentrated chemical and treatment unit branch piping join is designated as a
chemical injector detail. The chemical injector detail should include provisions to protect materials of
construction from the heat of dilution of concentrated corrosive chemicals. The key factor is to prevent
flow of concentrated  chemical when raw water (dilution water) is not flowing.  The  dilution water should
dissipate the heat.  The actual injection should take place in the center of the raw water pipe through an
injector that extends from the concentrated chemical pipe.  The injector material should be capable of
withstanding the high heat of dilution that develops specifically with sulfuric acid and to a lesser degree
with caustic soda.  Type 316 stainless steel and Teflon® are satisfactory.  It also is very important that the
concentrated chemical be injected upward from below; otherwise concentrated chemicals with specific
gravities greater than that of water will seep by gravity into the raw water when flow stops.  As described
previously, the chemical pumps are to be de-energized when the well pump (or other feed pump) is not
running.

The treated water pH should be monitored carefully. A pH sensor installed in the treated water main
indicates the pH at an analyzer. This analyzer should be equipped with adjustable high and low level pH
alarms.  The alarms should be interlocked with the well pump (or other feed pump)  control  (magnetic
starter), shutting it down when out-of-tolerance pH excursions occur.  A visual and/or audio alarm should
be initiated to notify the operator regarding the event.

A chemical injector detail similar to that used for acid in the treatment unit branch piping should be used
in the treated water main to inject caustic in order to raise pH in the treated water. If aeration for removal
of CO2 is used in place of or in combination with caustic soda injection for raising treated water pH, then
system pressure will be dissipated and the treated water will be repressurized.  If the water utility has
ground level storage tanks, the aeration-neutralization concept can be accomplished without the need for a
clearwell and repressurization. The aerator can be installed at an elevation that will permit the neutralized
treated water to flow  to storage via gravity.
                                               36

-------
Easy maintenance is an important feature in all piping systems. Air bleed valves should be installed at all
high points; drain valves should be installed at all low points. This assists the plant operator in both
filling and draining pipe systems. Air/vacuum valve and pressure relief valve discharges are to be piped
to drains. This feature satisfies both operator safety and housekeeping requirements. Bypass piping for
flow control, pressure control, flowmeter, and other in-line mechanical accessories is optional. Individual
equipment item bypass piping is costly and requires extra space. However, if continuous treatment plant
operation is mandatory, bypass piping should be included.

3.3.1.3     Instrument Design. Ease of maintenance is very important. Instruments require periodic
calibration and/or maintenance. Temperature indicators (optional) require thermal wells installed
permanently in the pipe. Pressure indicators require gauge cocks to shut off flow in the branch to the
instrument. pH sensing probes require isolation valves and union type mounting connections (avoids
twisting of signal cables). Supply of pH standard buffers (4.0, 7.0, and 10.0) should be specified for pH
instrument calibration. A laboratory bench should be located adjacent to the  sample panel. The sample
panel receives flow directly from sample points located in the process piping. The sample panel consists
of a manifold of PVC or polyethylene tubing with shutoff valves, which allows the plant operator to draw
samples from any point in the process at the laboratory bench. Laboratory equipment should be specified
to include wall cabinet, base cabinet with chemical  resistant counter top and integral sink, 115V/lN/60Hz
20-amp duplex receptacle, laboratory equipment/glassware/reagents for analysis of pH, arsenic, and other
ions. A deionized water capability for cleaning glassware and dilution of samples should be included.

3.3.1.4     Acid Storage and Feed Subsystem.  Operator safety for work within close proximity of highly
corrosive chemicals takes priority over process functional requirements. Emergency shower and eyewash
must be located within 20 ft of any work area at which operator exposure to acid or caustic soda exists.
Protective clothing should be specified. Neutralization materials (e.g., sodium carbonate) should be
provided to handle spills.  Potential spill areas must be physically contained.  Containment volumes
should be sufficient to completely retain maximum  spillage.

Chemical bulk storage tanks are covered in the preliminary design.

To minimize corrosion of acid pipe material, acid flowrate  is recommended to be less than 0.1 ft/sec.
Threaded pipe and fittings are not recommended; tubing and Swagelok fittings are recommended. CPVC
or Teflon® are satisfactory except for their vulnerability to damage from external impact forces.
Therefore protective clear reinforced plastic tubing  completely containing the plastic chemical lines is
recommended.  The use of double containment piping systems can also be considered.  Positive backflow
prevention should be incorporated in each chemical feed line. Day tanks should be vented to the
atmosphere, have a valved drain, and have a fill line float valve for failsafe backup control to prevent
overflow.  For treatment systems that use HC1 instead of FfcSC^ for pH adjustment, it is recommended
that references on materials acceptable for the handling and storing of this acid be consulted.

One acid feed pump is required for influent water pH adjustment. Acid feed  pumps are also required to
adjust pH during neutralization following a regeneration, and to neutralize regeneration waste water in the
wastewater surge tank.  Though preferable to use separate pumps for each function, it is feasible to
accomplish all three functions with a single pump.  The pump should be sized for a minimum of 110% of
the  maximum flowrate that it will provide; it should have a turndown limit no greater than 50% of the
minimum required flow.  Acid pump power should  be interlocked with the well pump (or other feed
pump) so that the acid pump is de-energized when that pump is not running.  If the chemical feed pump is
mounted above the day tank, a foot valve is required in the suction tube. Antisiphon provisions should be
included in the system. Because considerably more acid (approximately 1 gal/ft3 of activated alumina) is
consumed during the regeneration of an activated alumina bed than during routine treatment operation, a
day tank will  need to be refilled several times during the neutralization phase of the regeneration. The
                                               37

-------
day tank should be sized for a minimum of 200% of the daily acid consumption for the treatment process
pH adjustment requirement. The day tank should be translucent with gallon calibration on the tank wall.
The day tank should be set in an open-top, acid-resistant containment basin. All relevant regulatory
authorities should be consulted to ensure compliance with all safety regulations.

3.3.1.5     Caustic Soda Storage and Feed System. The safety requirements stated for acid also apply to
caustic soda. Vinegar should be provided to neutralize caustic spills.

The day tank and pump design features recommended for acid systems also apply to caustic.  Two caustic
pumps and day tanks are required. The process pH adjustment pump should be sized to pump 110% of
the maximum process  required. The rule of thumb for sizing the caustic soda regeneration feed pump
requires provisions of 2 gal of 50% NaOH/ft3 of activated alumina for activated alumina systems per
hour.  Depending upon the size of the system, a centrifugal pump or an air-operated diaphragm pump are
feed pump options. The process pH adjustment day tank should be sized for 200% of the maximum daily
consumption. The regeneration day tank should be the next standard tank size greater than the
requirement for one regeneration. Both tanks can be set in  one containment basin, sized for the  largest
tank.  Alternatively, a  common day tank can be sized appropriately and utilized for both process pH
adjustment and regeneration; this is more cost effective  due to the cost of day tanks with heat trace
installed. The regeneration pump can be calibrated by means of timing the flow and adjusting as
necessary to arrive at the design flowrate.  Carbon steel  threaded pipe or PVC pipe is suitable for the
service. All relevant regulatory authorities should be consulted to ensure compliance with all safety
regulations.

3.3.1.6     Regeneration Wastewater Surge Tank. Although treatment and disposal of regeneration
wastewater are not included in this design manual, a surge tank to receive the  wastewater is indicated.
The wastewater surge  tank should receive the entire batch of regeneration wastewater from the start of
backwash to the completion of treatment bed neutralization. To provide adequate capacity for
containment of the entire batch of regeneration wastewater, this tank should be sized to contain  400 gal/ft3
for AA systems.  This  tank can be either a ground-level  atmospheric tank or a tank positioned below the
discharge point of the  vessels to allow for gravity-transferring of portion of the waste stream. The tank is
constructed of carbon  steel or  PVC and should include a carbon steel floor and roof and an interior  epoxy
lining. The tank should include a reinforced concrete containment structure.  The tank should include fill,
chemical feed, drain overflow vent, multiple discharge,  and multiple sample pipe connections. The tank
should include one ground-level manway and one roof manway (with  safety ladder and handrails),
provisions for a liquid level indicator, for an ultrasonic liquid sludge level sensor, liquid level controller,
and a side entry mixer.

3.3.2       Final Drawings.  All of the information required for complete installation of a fluoride
removal plant should appear in the final construction drawings and specification package.
Isometric drawings for clarification of piping subsystems are recommended; these views clarify the
assembly for the installer (see  Figure 3-5). Cross-referencing drawings, notes, and specifications also is
recommended.

3.3.3       Final Capital Cost Estimate. Similar to the preparation of the preliminary cost estimate, the
final cost estimate is prepared  based on a take off of the installed system. The estimate is now based upon
exact detailed information rather than general information which was used during the preliminary
estimate. The estimate is presented in the same format and is to be accurate within ±10%. Because
financial commitments are determined at this stage, this degree of accuracy is required.

3.3.4       Final Design Revisions.  Upon their  completion, the final construction drawings and
specifications are submitted for approval to the owner and the regulatory authorities. If changes or addi-
                                               38

-------
tional requirements are requested, they should be incorporated and resubmitted for approval.  If communi-
cation with the approving parties has taken place during the course of the design, then time-consuming
resubmittals should not be necessary.  Upon receipt of approval, the owner, with assistance from the
engineer, solicits bids for the construction of the arsenic removal water treatment plant.
                    A/V
        TREATMENT  VESSEL
               TYP.
                                                                    REGEN. WASTE
                    Figure 3-5. Treatment Vessels Piping Isometric
                                               39

-------
                      4.0: ACTIVATED ALUMINA PLANT OPERATION
4.1        Introduction

Upon completion and approval of the final design plans, the owner (client) proceeds to advertise for bids
for construction of the treatment plant.  The construction contract normally is awarded to the firm
submitting the lowest bid.  Occasionally, circumstances arise that disqualify the low bidder, in which case
the lowest qualified bidder is awarded the contract. Upon award of the construction contract, the engineer
may be requested to supervise the work of the construction contractor.  This responsibility may be limited
to periodic visits to the site to assure the client that the general intent of the design is being  fulfilled; or it
may include day-to-day inspection and approval of the work as it is being performed.  The engineer
should review and approve all shop drawings and other information submitted by the contractor and/or
subcontractors and material suppliers. All acceptable substitutions should be approved in writing by the
engineer.  Upon completion of the construction phase of the project, the engineer normally  is requested to
perform a final inspection. This entails a formal approval indicating to the owner that all installed items
are in compliance with the requirements of the design. Any corrective work required at that time is
covered by a punch list and/or warranty. The warranty period (normally one year) commences upon final
acceptance of the project by the owner from the contractor.  Final acceptance usually takes  place upon
completion of all major punch list items.

Preparation for treatment plant startup and operator training may or may not be included in the construc-
tion contract. Although this area of contract responsibility is not germane to this manual, the activities
and events that lead up to routine operation are.  This chapter discusses those steps in the sense that the
operator is performing them. The operator could be the contractor, the  owner's representative, or an
independent third party.

System operating supplies, including treatment chemicals, laboratory supplies, and recommended spare
parts should be procured, and stored on site. The treatment plant operating and maintenance instructions
(O&M manual) should be available at the project site. Included in the O&M manual are the valve
number diagram which corresponds to brass tags on the valves (see Figure 4-1), a valve directory
furnished by the contractor, and a valve operation chart (see Table 4-1).

The media vessels and piping should be disinfected in accordance with AWWA standard procedures.
Activated alumina then is placed in the vessels. After backwashing to remove fines, the AA plant is ready
to start operation.

There are four basic modes of operation: treatment, backwash, regeneration, and acid neutralization.
Operating details for each of these modes are covered in this section. It is important to note that each of
the above modes uses raw water during the  operation.

4.2        Initial  System Startup

The operator should thoroughly review the O&M manual, become familiarized with every component of
the plant, and resolve any questions that arise.

Proper placement of AA in a treatment vessel is critical to future system performance.  Dry media usually
is delivered in drums or sacks.  The media volume is determined on a dry weight basis. The actual  den-
sity varies with the degree of packing of the bed.  Unless instructed otherwise by the manufacturer,
48 lb/ft3 is a suggested media density for use in weight calculations for AA. The virgin granular AA
material is "coated" with caustic. During media loading, a small amount of fines can become airborne
                                               40

-------
txj
s:
ixi
   LEGEND
Shut-off Valve
Butterfly Valve
Check Valve
Pressure Control
Valve
Expansion Joint
Pressure Indicator
     Pressure Indicator/
     TotalIrer
                                                        TREATMENT
                                                        UNIT NO.  2
                                                                          Vent
                                                                              DrainNfc
                                                                  CAUSTIC  STORAGE TANK
                                                                       OVent
                                                                    ACID STORAGE TANT
                                                             Backwash Wastewater
                                                    Regeneration  Wastewater
                                LINED  EVAPORATION  POND
                   Figure 4-1. An Example Valve Number Diagram
                                                41

-------
                       Table 4-1. An Example Valve Operation Chart for Treatment Vessels in Media Regeneration Modes(a)
to
Valve
Function
Unit No. 1
Treatment
Drain
Backwash
Drain
Upflow Regen.
Upflow Rinse
Drain
Down flow Regen.
Drain
neutralization
Treatment
Treatment
Shu toff
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Unit No. 1
Operation
Unit No. 2
Treatment
Treatment
Shu toff
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Drain
Backwash
Drain
Upflow Regen*
Upflow Rinse
Drain
Down flow Regen.
Drain
Neutralization
Treatment
11
0
X
0
X
0
0
X
0
X
0
0
0
X
0
Q
0
0
0
0
0
0
12
0
X
X
X
X
X
X
X
X
X
0
0
X
0
0
0
0
0
0
0
0
13
0
X
X
X
X
X
X
0
X
0
0
0
0
0
0
0
0
0
0
0
0
14
X
X
0
X
0
o
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
15
X
X
0
X
0
0
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
16
X
0
X
0
X
X
0
0
0
0
X
X
X
X
X
X
X
X
X
X
X
21
0
0
X
0
0
0
0
0
0
0
0
X
0
X
0
0
X
0
X
o
0
Unit No. 2
Opera t ion
22
0
0
X
0
0
0
o
0
o
0
0
X
X
X
X
X
X
X
X
X
0
23
0
0
0
0
0
o
0
0
0
0
0
X
X
X
X
X
X
0
X
0
0
24
X
X
X
X
X
X
X
X
X
X
X
X
0
X
0
0
X
X
X
X
X
25
X
X
X
X
X
X
X
X
X
X
X
X
0
X
0
0
X
X
X
X
X
26
X
X
X
X
X
X
X
X
X
X
X
0
X
o
X
X
o
0
0
o
0
1
0
o
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
o
0
0
Numbers
System
Isolation
2
0
0
O
0
0
0
o
o
0
o
0
0
0
0
0
0
o
0
0
o
0
3
X
0
0
0
0
0
0
0
0
0
X
0
0
0
0
o
o
0
o
0
X
4
X
0
0
0
X
X
X
X
X
X
X
I)
0
0
X
X
X
X
X
X
X
5
X
X
X
X
0
o
0
0
0
0
X
X
X
X
0
0
0
0
0
0
X
6
p
X
X
X
X
X
X
X
X
X
F
X
X
X
X
X
X
X
X
X
•p
7
X
X
o
X
p
X
X
p
X
p
X
X
0
X
p
p
X
p
X
p
X
Sample
17
P
X
X
X
X
X
X
X
X
0
p
p
X
X
X
X
X
X
X
X
p
18
p
X
X
X
X
X
X
X
X
p
p
p
X
X
X
X
X
X
X
X
t
27
p
p
X
X
X
X
X
X
X
X
p
X
X
X
X
X
X
X
X
0
p
28
P
P
X
X
X
X
X
X
X
X
p
X
X
X
X
X
X
X
X
p
p
31
o
X
X
X
X
X
X
X
X
0
0
0
X
0
0
o
0
0
0
0
0
Chemical
Shutoff
32
0
O
X
0
0
o
0
0
0
0
0
X
X
X
X
X
X
X
X
o
0
41
X
X
X
X
0
X
X
0
X
X
X
X
X
X
X
X
X
X
X
X
X
42
X
X
X
X
X
X
X
X
X
X
X
X
X
X
0
X
X
0
X
X
X
43
0
0
X
0
0
0
o
0
o
0
0
o
X
0
0
0
0
0
0
0
0
               Legend:  0 — VaIve Open
                       X - Valve Closed
                       P - Periodic Sample

                (a) Refer to Figure 4-1 for valve location.

-------
and are irritating to the personnel who are handling them. Eye, skin, and inhalation protection are recom-
mended during vessel loading activity.

The vessel should be half-filled with water prior to placing underbed materials and AA media through a
manway in the top of the vessel. Upon placing AA into the vessel, heat is generated by the wetting of the
caustic "coating" on the alumina grains.  The water in the tank dissipates the heat, thereby preventing
cementing of the bed. The water in the vessel also separates fines from granular materials, protects the
underdrain assembly from impact, and initiates stratification of the bed.  It is recommended that the bed
be placed in two or three lifts. In a two- or multiple-bed treatment system, alternate placing of media and
backwashing can be worked together among the treatment units. The beds should be thoroughly back-
washed with raw water after each lift. The backwash flowrate should be adjusted to provide 30% to 50%
bed expansion, which is typically 9 gpm/ft2 or higher, For extremely warm or cold water, backwash
flowrates may have to be adjusted up or down, respectively. To remove all media fines from the bed,
each backwash should last for at least 30 min. If the fines remain in the beds, potential problems such as
channeling, excessive pressure drop, or even cementing can develop. The extra backwashing effort
during bed placement permits fines at the bottom of the beds to work their way up and out to waste.
Because the lower portions of the beds (which contain the largest particles) do not expand during
backwash, fines not backwashed out of the bed at that stage may be permanently locked into the beds.
The backwash wastewater should be directed to waste.

To accurately estimate the bed depth in each vessel, it is recommended that freeboard be measured after
the placement of the underbed materials and after the placement, backwashing, and downflow rinse of the
AA media. The downflow rinse helps settle a freshly backwashed bed. The  freeboard is measured using
a tape measure from the top of the underbed materials or the AA bed to a set point on the manway used to
load the materials. The difference of the measurements is the bed depth, which can be used to calculate
the bed volume and confirm if sufficient media has been placed into the vessel.

The pressure loss checkout mentioned in Section 3.3, final design, should be accomplished at this point.
Table 4-2 presents calculated pressure drop through the AA media.  If there is a pressure loss problem, it
should be corrected prior to system startup.
                     Table 4-2.  Calculated Activated Alumina (28x48 Mesh)
                                 Downflow Pressure Drop Data
Water
Flowrate
(gpm/ft2)
2.0
3.0
4.0
5.0
6.0
7.0
Pressure Drop
in psi per Foot of
Bed Depth
0.009
0.018
0.028
0.040
0.053
0.068
Modified
Reynolds
Number
2375
3555
4735
5900
7111
8291
The acid and base used for pH adjustment should now be placed in or carefully transferred to the designed
containers. The online pH meters should be calibrated. For AA, the optimum pH values are in the range
of 5.0-6.0.  Because acid feedrates are a function of raw water alkalinity, they vary from one water to
another. When the alkalinity of the raw water is extremely high or the cost of acid is very high, it can be
more cost-effective to operate in a pH range of 6.0 to 6.5 to reduce the acid consumption (even though
fluoride removal efficiency is also reduced).
                                              43

-------
Before system startup, the plant should be cleaned up.  Airborne fines that form a dust-like coating on
piping and equipment should be removed. Good housekeeping should begin immediately and be
continued on a permanent basis.

4.3        System Operations

4.3.1       Treatment Mode. The downflow treatment for the virgin run can now begin. It is
recommended that one vessel be placed in operation at a time. This will allow the operator to concentrate
on initial raw water pH adjustment on one treatment unit until it is in stable operation; the operator then
can devote his/her full attention to the second treatment unit. The basic operating mode flow schematic is
illustrated in Figure 4-2.

With AA, the initial effluent pH is high with no fluoride removal. After a short period, both pH and
fluoride in the treated water drop to anticipated levels.  At that time, the treated water can be directed to
storage and/or distribution.  The freshly installed media can also be neutralized with sulfuric acid similar
to the final step of acid neutralization prior to putting back the vessels into service following media
regeneration. Depending on the requirements of the state or local regulatory authorities, samples may
have to be analyzed at a certified testing laboratory prior to approval of treated water entering the
distribution system.

In the parallel operation utilizing two treatment vessels, the fluoride removal process takes place in a
treatment band (or adsorption zone) contained in each of the two vessels. After a treatment period, the
adsorption media at the top of the adsorption zones becomes saturated and the adsorption zones begin to
simultaneously migrate downward through the vessels until fluoride starts to break through.
Breakthrough is defined as the first detectable amount of fluoride appearing in the effluent of a vessel.

As treatment progresses, the adsorption zones in both vessels continue to move downward through the
vessels until fluoride  concentrations in the vessel effluent have reached a set breakthrough level (e.g.,
80% of MCL).  Both  vessels are then taken offline for regeneration. To minimize system downtime,
regeneration can be staggered with a vessel in the regeneration mode while the other continues to be in
the service mode. Until both vessels have been regenerated, the system resumes to be in the normal
service mode.

In the vessel that has  completed the regeneration process, the treated water pH gradually drops to the
adjusted raw water pH level where it remains through the duration of the run. Because the pH of the
treated water is lower than the normally accepted minimum pH of 6.5, it should be raised either by chem-
ical addition, aeration, and/or blending with raw water.  Regardless of the method of pH adjustment, it
should take place and be stabilized at the desired level prior to delivering the treated water into the
distribution system.

High pH in the treated water is also a concern.  Normally the maximum allowable pH is 8.5; however,
there are exceptions where pH 9.0 may be permitted. If desired, the treated water can flow past a failsafe
pH sensor with high and low level alarms. If there is a pH excursion exceeding the allowable limits, an
interlock (incorporating the  pH alarms with the well pump(s) magnetic starter) de-energizes the well
pump(s).  Simultaneously, the chemical pumps shut down as their controls are interlocked with the well
pump(s) power circuitry. The failsafe pH override automatically prevents any treated water for which pH
is out of tolerance from entering the distribution system. In the event of such an excursion, the operator
manually controls the well pump(s) to divert the unacceptable water to waste, determine the cause of the
deviation, and make corrections prior to placing the treatment system back on line. Probable causes for
treated water pH deviations  are: changes in water flowrate, acid flowrate, caustic flowrate, and raw water
chemistry.
                                               44

-------
            Raw Water
                    AcW
           Treated Water
          TREATMENT
             UNIT
TREATMENT AND DOWNFLOW
            RINSE
              Raw Water
         TREATMENT
            UNIT
                                                           T
                                                            *-
                   Waste

BACKWASH AND UPFLOW RINSE
               Raw Water
          TREATMENT
             UNIT
                             J
                           .Caustic
                    Waste
  UPFLOW  REGENERATION
  NOTE:  For Clarity Only Relevant
         Pipes And Shutoff Valves
         Are Shown.
              Raw Water
          TREATMENT
            UNIT
                                                               Caustic
                     Waste
 DOWNFLOW  REGENERATION
          Figure 4-2. Example Operating Mode Flow Schematics
                                45

-------
A treatment run can be extended by blending treated water in which the fluoride level exceeds the MCL
with treated water with a low fluoride level. This can be done either in the effluent main leaving the
treatment plant, in the storage reservoir or bypassing raw water to blend with treated water. During a
treatment run, there is a period when the fluoride level of the treated  water is well below one half of the
MCL.  As breakthrough occurs, there is a period of slowly increasing fluoride concentration in the treated
water.  Blending in the effluent main  entails staggering the treatment cycles of two or more treatment
units. This can be accomplished by continuing treatment in one unit after its increasing fluoride level has
surpassed the MCL and blending it with low fluoride effluent from one or more units that is in the early
stage of a treatment cycle. The operator can extend the run until the  fluoride level reaches at least twice
the MCL before terminating the run.  As the fluoride level gets higher the operator must reduce the
flowrate to maintain the combined high and low fluoride levels at an acceptable average. The same
processes take  place in the storage reservoir using one or more treatment  units.

These blending practices can significantly increase the fluoride loading on the alumina and result in lower
operating costs. As noted by in the original fluoride design manual (Rubel, 1984), the increase in fluoride
loading can achieve 25 to 50% or even 100%. It should be noted that the higher the raw water fluoride
level, the greater the adsorption capacity. Since there are many other factors that can affect this capacity,
the precise amount is difficult to predict. The operator must be cognizant of the fact that the more water
treated during a run, the lower the operating costs.

In raw waters where the fluoride level does not exceed two times the MCL, part of the raw water can
bypass treatment and be blended back with the treated water. Water  with even higher fluoride levels can
also profit from bypassing, but the economic benefits rapidly diminish.

The operator can also reduce chemical consumption by blending high pH with low pH treated waters.
This is accomplished during the period when one treatment unit has recently been regenerated and treated
water pH is still high. A skilled operator develops many techniques such as this to minimize operating
costs.

High iron and manganese contents in raw water can cause problems to the AA treatment process. Iron
oxidizes, precipitates, and is filtered from water by AA.  This results in increased pressure drop and
shortened treatment runs. Raw water iron content greater than 0.3 mg/L is a cause for concern.  However,
if the iron  concentration is above 0.3 mg/L, the secondary MCL, an iron removal process should be
considered as the treatment process for fluoride removal in place of the adsorptive media process because
of the capability of the process to remove fluoride.

4.3.2      Backwash Mode. It is important that the bed be backwashed with raw water after each
treatment run prior to regeneration for two reasons.  First, any suspended solids that have been filtered by
the bed tend to blind the bed. Therefore, these particles should be removed from the bed.  Second, even
though filtration may have been negligible, the  downward flow tends to pack the bed. An upflow
backwash  will  expand the bed, and break up any tendency towards wall effects and channeling. A
backwash  rate  of 7 gpm/ft2 will expand the 28 x 48 mesh AA bed approximately 50%. The operator
needs to carefully watch bed expansion and adjust backwash rate to avoid backwashing granular bed
material out of the vessel. Normally backwashing lasts 10 min or until all suspended solids are  removed
from the media. Readers are referred to  Table 4-1 for valve positions for the backwash mode.  The basic
flow schematic for the backwash mode of a lead-lag system is shown in Figure 4-2.

For most effective backwash, it is recommended that the vessel be drained prior to backwash. As
backwash  water flows into a drained bed, it lifts the entire bed approximately 1 ft prior to the bed
fluidizing.  This action provides an efficient scouring action without  excessive abrasion to the adsorptive
media grains.  Backwash wastewater  samples should be monitored to determine that suspended solids are
                                              46

-------
being removed and AA media is not being washed out of the bed. Excessive backwash causes abrasion
that wears down activated alumina grains, and also wastes raw water and increases the wastewater
disposal volume. Therefore, backwash volume should be minimized.

It is prudent to periodically inspect the media level of each treatment bed to determine whether bed
volume has changed.  If desired, the treatment vessels can be replenished with virgin AA and backwashed
prior to being returned to the service mode.

4.3.3       Regeneration Mode.  Upon completion of a treatment run, the AA beds can be regenerated
using a two-step (upflow and then downflow) or a one-step (upflow or downflow) regeneration process.
The two-step process involves an upflow regeneration step following draining of the bed after the back-
wash mode. The upflow regeneration is followed by an upflow rinse. The vessel is then drained to the
top of the treatment bed prior to a downflow regeneration step.  The one-step process utilizes only upflow
or downflow regeneration to complete the process. Both processes use a 1% (by weight) NaOH solution.

The objective of regeneration is to remove as much fluoride ions (and other co-contaminants) as possible
from the media before returning the bed to the treatment mode.  Fluoride ions lose their attraction
(adsorptive force) and become repelled by the  alumina when the pH rises above 10.5.  The higher the pH,
the faster and more efficient the regeneration.  However, too high a pH not only costs more (because of
higher caustic consumption for media regeneration and higher acid consumption for media
neutralization), but is also increasingly more aggressive to the AA material and causes the media to
dissolve.  The 1% NaOH solution (at pH 13.0) is sufficient to remove most fluoride ions while keeping
the media dissolution to the minimum. When other co-contaminants such as arsenic also exist, the caustic
concentration must be increased to as high as 4% to effectively remove arsenic (and fluoride) from the
media. As described in Section 3.0, the dilution of the caustic takes place at an injector in the  regenera-
tion water piping. Both the raw water and the  50% NaOH are metered prior to injection into the
regeneration main.  The accuracy of the metering ranges from ±2% to ±5% depending on the type of flow
instrumentation.

The rule of thumb for the volume of 1% caustic solution required per AA regeneration step is 30 gal/ft3 of
treatment media. For the two-step process, the actual regeneration time exclusive of draining, flushing
and neutralization is 100 min, i.e., 50 min per step for the solution to flow through the bed.  For a 5-ft-
deep treatment bed, a flow of 3.0 gpm/ft2 for a period of 50 min for each regeneration step is sufficient.
This equates to approximately 0.4 gal 50% NaOH per cubic foot of media for each regeneration step
(upflow and downflow).  Table 4-1 shows the valve positions during the two-step regeneration process.
The basic flow schematics for the regeneration modes are illustrated in Figure 4-2. Plant A in
Appendix C follows a regeneration regime modified from this two-step process.

After backwash and prior to caustic regeneration, the bed should be drained to remove water, which
dilutes the caustic concentration. Upon completion of the 50-min upflow regeneration, the caustic feed
pump is turned off and the caustic soda day tank refilled. The raw water continues to flow for 60 min at
2.5 gpm/ft2 flowrate upward through the bed, flushing out the fluoride.  After this rinse step is completed,
the vessel is drained to the top of the treatment bed, again to remove dilution water. The downflow
regeneration then takes place for 50 min.  The  downflow regeneration is followed by draining fluid down
to the top of the bed prior to the start of the acid neutralization step.

For the one-step upflow regeneration process, the  bed is backwashed and drained prior to applying caustic
upflow similar to the first part of the two-step regeneration process.  The upflow regeneration is followed
by draining fluid down to the top of the bed prior to the start of the neutralization step. For the one-step
downflow regeneration process, the bed is backwashed and drained to the top of the bed prior to the
                                              47

-------
commencement of downflow regeneration.  The downflow regeneration is then followed by draining to
the top of the bed prior to the start of acid neutralization.

The one-step regeneration process appears to be as effective as its two-step counterpart. As shown in
Appendix C, Plants B utilizes a one-step downflow regeneration process, which draws 38.6 gal/ft3 of 1%
caustic through the 420-ft3 bed at 270 gpm (or 2.4 gpm/ft2) for 60 min. This gives an EBCT of 11.6 min
through the 3.7-ft-deep AA bed (versus 12.5 min through the 5-ft-deep bed described above). This one-
step process with a shorter regeneration time (60 vs. 100 min) is also sufficient to completely regenerate
the bed for sustainable system operation, according to the plant operator.

4.3.4       Neutralization Mode. Following caustic regeneration, the bed must be treated with acid
prior to returning it to the treatment mode as rapidly as possible.  The pH of the AA media after regen-
eration is about 13, which must be adjusted  down to pH 5.5 for effective fluoride removal during the
following treatment run. The minimum pH that can be safely exposed to the granular AA is 2.5. A pH
lower than that can dissolve media and cause cementing of the bed (especially if the acidic water is
allowed to stand in the bed for any period of time).

At the start of the downflow neutralization mode, the valves are positioned according to Table 4-1, and
the flowrate  is  adjusted to that of the normal treatment mode.  The basic flow schematic for the
neutralization mode is illustrated in Figure 4-2. After the acid pump is started, the pH of the raw water is
adjusted to 2.5. The acid feedrate again varies with the alkalinity of the raw water. The raw water flow-
rate may have to be reduced to achieve pH 2.5 at the maximum acid pump feedrate.

As the neutralization step proceeds, the pH of the bed effluent gradually drops below  13. The rate of pH
reduction increases at an increasing rate. As the effluent pH drops below 10, the effluent fluoride level
begins to drop  below that of the raw water.  At the point where the fluoride level drops below the MCL,
the water becomes usable and can be directed to storage.  When the effluent pH drops to 8.0, the raw
water pH is adjusted up to 4.0 as the bed rapidly neutralizes.  When the effluent pH drops to 6.5, the raw
water pH is adjusted up to 5.5 where it remains through the duration of the treatment cycle. The
regenerated bed now starts the next cycle in the treatment mode.

Prior to placement of the regenerated treatment unit into service,  the operator should open the manway in
the top head of the vessel to check the level of the treatment media. The operator may replace the lost AA
by adding a proper amount to bring the bed  back to the original level.

The operator should then backwash the bed with water adjusted to pH 5.5 for 30 min. The regenerated
treatment unit will then be placed into service.  An example of the two-step regeneration process is shown
in Table 4-3.

The volume  of wastewater produced during the regeneration of a treatment bed varies with the physi-
cal/chemical characteristics of the raw water. A rule of thumb that can assist the operator in his logistical
handling is that 300-400 gal of wastewater is produced per cubic  foot of AA during each regeneration.
Typical volumes of wastewater generated per cubic foot of AA during each regeneration step for a hypo-
thetical treatment bed are shown in Table 4-3. Operational experience at a specific treatment plant will
present deviations from these quantities.

4.4        Operator Requirements

A qualified operator for a fluoride removal plant should have thorough fluoride removal process training,
preferably at an existing treatment plant. The operator should be able to service pumps, piping  systems,
                                              48

-------
instrumentation, and electrical accessories. The operator should be fully informed about the safety
requirements and physical/chemical characteristics of both acid and caustic in all concentrations.
        Table 4-3. Example Two-Step Process Conditions for Regeneration of an Activated
                                  Alumina Treatment System(a)
Step
No.
1
2
3
4
5
6
7
Step
Backwash
Regeneration
Rinse
Regeneration
Neutralization
Neutralization
Neutralization
Process
Solution
Raw water
!%NaOH
Raw water
!%NaOH
Raw Water adjusted to
pH2.5
Raw water adjusted to
pH4.0
Raw water adjusted to
pH5.5
Flow
Direction
Upflow
Upflow
Upflow
Downflow
Downflow
Downflow
Downflow
Flowrate
(gpm/ft2)
7
1.2
2.5
1.2
Varies
Varies
Varies
Time
(min)
10
50
60
50
Time to achieve
pHofS.O
Time to achieve
pHof6.5
Time to achieve
pHof5.5
TOTAL
Wastewater
(gal/ ft3)
60
15
30
15
180
300
(a) Consult media or system suppliers for specific media regeneration/neutralization steps.
Corrosive chemical safety requirements as to clothing, equipment, antidotes, and procedures must be
thoroughly understood.  The operator should be thoroughly trained to run routine water analyses
including the method for determining fluoride levels. The operator should be well grounded in mathe-
matics for operation cost accounting and treatment run recordkeeping.  The operator, above all, should be
dependable and conscientious.
4.5
Laboratory Requirements
In addition to the O&M manual, the treatment plant should have the latest edition of Standard Methods
for the Examination of Water and Wastewater prepared jointly by the American Public Health
Association-American Water Works Association-Water Environment Federation (1995). This manual
supplies the plant operator with necessary information for acceptable methods for analyzing water. A
recommended list of items for analysis is illustrated in Table 3-1. The primary requirement is accurate
analysis for fluoride and determination of pH. As long as pH meters are calibrated and cleaned regularly,
high precision measurements are easily obtained.  Care should be exercised to prevent contamination of
pH buffers.

Fluoride can be effectively preserved in field samples and analyzed by several analytical methods down
to the MCL of 4 mg/L or less.  Preservation of fluoride is accomplished by acidifying the sample to pH
<2. Fluoride may be analyzed using several EPA approved analytical methods: ion chromatography (1C),
selective fluoride electrode, etc.
4.6
Operating Records
A system of records should be maintained on file at the treatment plant covering plant activity, plant
procedures, raw water chemical analyses, plant expenditures, and inventory of materials (spare parts,
tools, etc.). The plant operator should have the responsibility of managing all aspects of the treatment
                                               49

-------
plant operation. The operator is accountable to the water system management. The recommended record
system should include, but not be limited to, the items described in the following subsections.

4.6.1       Plant Log. A daily log should be maintained in which the plant operator records daily acti-
vities at the plant. This record should include a listing of scheduled maintenance, unscheduled
maintenance, plant visitors, purchases, abnormal weather conditions, injuries, sampling for state and other
regulatory agencies, etc. This record should also be used as a tool for planning future routine and special
activities.

4.6.2       Operation Log.  The operator should maintain a log sheet for each treatment run for each
treatment unit. Thereby, a permanent plant performance record will be on file. Figure 4-3 illustrates a
copy of a suggested condensed form.

4.6.3       Water Analysis Reports.  It is recommended that the plant operator run an analysis of raw
and treated fluoride levels once each week for each unit, and should run a total raw water analysis once
per month. Changes in raw water may necessitate changes in the treatment process. Raw water changes
that can impact the treatment process include, but are not limited to, pH, alkalinity, iron, manganese,
hardness, phosphate, silica, sulfate, sodium, TDS, and turbidity.  Table 3-1 illustrates a copy of a
suggested form. A permanent file of these reports can be a valuable tool.

4.6.4       Plant Operating Cost Records. Using accounting forms supplied by the water system's
accountants, the plant operator should keep a complete record of purchases  of all spare parts, chemicals,
laboratory equipment and reagents, tools, services,  and other sundry items.  This should be  supplemented
by a file of up-to-date competitive prices for items that have been previously purchased.

4.6.5       Correspondence Files. The plant  operator should retain copies of all correspondence
pertaining to the treatment plant in chronological order.  Included would be intradepartmental notes and
memos, in addition to correspondence with other individuals and/or organizations.

4.6.6       Regulatory Agency Reports. The plant operator should maintain a complete file of copies of
all reports received from state, county, or other regulatory agencies pertaining to the treatment plant.

4.6.7       Miscellaneous Forms. The operator should have an adequate supply of accident and
insurance forms.

4.7        Treatment Plant Maintenance

The maintenance concept for the fluoride removal water treatment plant is to isolate the equipment to be
serviced by means of shutoff valves, vent and drain lines (as required), repair or replace equipment, fill
lines, open valves, and start service. To accomplish this, all equipment items  are equipped with isolating
valves, and all piping systems have vents at high points and drains at low points.

Equipment manufacturers' recommended spare parts  should be stocked at the treatment plant to avoid
lengthy maintenance shutdowns.

If the entire treatment plant needs to be shut down and the plant has bypass  and if the bypass piping is
large enough to handle the full flow, the plant itself can be bypassed.  This can be done by closing the
butterfly valves in the raw water and treated water line and then opening the butterfly valve in the bypass
line. This would result in untreated water with excessively high fluoride being pumped to distribution, an
event that should not occur without the approval of the water system manager and the regulatory agency.
                                               50

-------
4.8        Equipment Maintenance

Equipment manufacturer's maintenance instructions should be included in the Suppliers Equipment
Instructions section of the O&M Manual.

4.9        Treatment Media Maintenance

The plant operator should inspect the surface of each treatment bed at least once a month. If the level of a
bed lowers more than 8 in., makeup adsorptive media should be added. Makeup adsorptive media should
be evenly distributed. There should be a minimum depth of 2.0 ft of water above the surface of the
existing bed through which the makeup adsorptive media will be added.  The vessel should be closed
immediately and backwashed at 7 gpm/ft2 (or at rate recommended by the manufacturer) for at least
30 min. It is very important to flush the fines out of the virgin activated alumina as soon as it is wetted.

It is important that the treatment beds should not remain in the drained condition for more than one hour.
Treatment units not in use should remain flooded.

4.10       Treatment Chemicals Supply

The operator should carefully monitor the consumption of liquid chemicals and reorder when necessary.
The operator should have a method of determining the depth of liquid in the storage tank (e.g., dipstick)
and equating that to the volume of liquid in the tank. Figure 4-4 illustrates a liquid depth versus volume
curve for a 5,000-gal horizontal cylindrical tank with dished head.

4.11       Housekeeping

The plant operator should wash down all equipment at least once per month. Floors should be swept.
Bathroom  and laboratory fixtures should be cleaned once per week.  All light bulbs should be replaced
immediately upon failure. Emergency shower and eyewash should be tested once per week. Any
chemical spill should be neutralized and cleaned up immediately.  Equipment should be repainted at least
once every five years.
                                              51

-------
FLUORIDE REMOVAL WATER TREATMENT PLANT
OPERATION LOG
Unit # Run #
Service to Reservoir
Meter End Meter Start
Bypass to Reservoir
Meter End Meter Start
Backwash to Sewer
Meter End Meter Start
Regeneration Solution to Waste
Upflow:
Meter End Meter Start
Downflow:
Meter End Meter Start
Rinse to Waste
Meter End Meter Start
Neutralization Rinse to Waste
Meter End Meter Start
Total Wastewater Summary
Total K-aal.

Date Start
Total Treated
Total Treated
Total Treated
Total Treated
Total Treated
Total Treated
Total Treated
Percent Waste
TREATED WATER LOG
Date End
K-aal.
K-aal.
K-aal.
K-aal.
K-aal.
K-aal.
K-aal.
%

Date










Meter
Reading
(K-gal)










A
(K-gal)










Total
Q
(K-gal)










RawF
(mg/L)










Treated
F*
(mg/L)










AF
(mg/L)










F
Removed
(mg)










IF
Removed
(mg)










* Average treated water fluoride.
        Figure 4-3. Fluoride Removal Water Treatment Plant Operation Log
                                         52

-------
o
                                                        HORIZONTAL CYLINDRICAL VESSEL
                                                        8'-0"«l x 12'-0" S/S WITH  FLANGED
                                                        AND DISHED HEADS
         6    12   18   24    30   36   42   48   54    60   66   72   78    84   90   96

                                         DEPTH OF LIQUID  (INCHES)

                     Figure 4-4.  5,000-gal Chemical Storage Tank - Liquid Volume
                                                    53

-------
          5.0: CENTRAL TREATMENT PLANT CAPITAL AND OPERATING COSTS
5.1        Introduction

The primary objective of the central treatment plant design is to provide the client with a least expensive
AA system that can remove excess fluoride and meet the peak water demand year round.  The system not
only should work efficiently and reliably but also be simple and inexpensive to operate. Therefore, the
economic feasibility evaluation should include initial capital and O&M costs.

In addition to the system flowrate, many other factors can impact capital costs. These include, but are not
limited to, the following:

           (1)  Raw water quality, including but not limited to pH, fluoride concentration; presence of
               competing anions such as arsenic, silica and phosphate; and amounts of alkalinity, total
               organic carbon (TOC), hardness, iron and manganese
           (2)  Chemical supply logistics
           (3)  Manual vs. automatic operation
           (4)  Backwash and regeneration wastewater disposal requirements

           (5)  Ambient conditions including climate (temperature, precipitation, wind, etc.), soil
               condition, seismic zone, and 100-year flood plain

           (6)  Existing and planned (future) potable water system parameters
               (i)  Number and location of wells
               (ii) Water storage (amount, elevation, and location)
               (iii) Distribution (location, peak flows, total flow, pressure, etc.)
               (iv) Consumption (daily and annual)
           (7)  Financial considerations (cost trends, capital financing costs, cash flow, labor rates,
               utility rates, chemical costs, etc.).
Operating costs are normally passed directly  on water customers in their monthly water bills. These costs
include the following:

           (1)  Treatment chemical costs
           (2)  Operating labor costs
           (3)  Utility costs
           (4)  Replacement media costs
           (5)  Replacement parts and miscellaneous materials costs

As the consumer's water bill normally is based on metered water consumption, the costs for treatment are
prorated on the unit of volume measurement, which is usually 1,000 gal (i.e., 0/1,000 gal). Some systems
do not meter consumption; instead, they charge a flat monthly rate based upon the size of the branch
connection to the water main.  Although this  latter mode of distribution saves the cost of meters as well as
the reading of meters, it does not promote water conservation. Therefore, far more water is pumped,
treated, and distributed, resulting in a net increase in operating cost.

The common denominator that applies to both the operating cost and the bill for water consumption is the
unit of volume,  1,000 gal. Each operating cost factor can be reduced to cost/1,000 gal. The sum total of
the annual operating costs based on total water production yields the cost per 1,000 gal.
                                              54

-------
5.2        Capital Costs

5.2.1       Discussion of Cost Variables.  All cost variables mentioned above have direct impact on the
costs of a central treatment system. Ideally, conditions exist in which a minimum cost system can be
designed and fabricated. A hypothetical example of an ideal situation would resemble the following:

           (1) Raw water quality presents few or no challenges to AA, i.e., with moderate temperature
               and contains low alkalinity, low concentrations of competitive ions, etc.
           (2) Acid and caustic stored in large quantities onsite for other purposes
           (3) Manual operation by labor that is normally at the site with sufficient spare time
           (4) Existing wastewater disposal capability adjacent to treatment site (e.g., a large tailings
               pond at an open pit mine)
           (5) Warm moderate climate with minimal precipitation and no freezing, no high temperature,
               and no high wind; existing concrete pad located on well compacted, high-bearing
               capacity soil; no seismic requirements.
           (6) Single well pumping to subsurface storage reservoir with capacity for peak consumption
               day
           (7) Funding, space, etc., available.
This ideal situation never exists in reality. Occasionally one or more of the ideal conditions occur, but the
frequency is low.  If the final estimate for the example used in Appendix B is revised to incorporate the
above ideal conditions, the cost estimate would be reduced about 60% (Rubel, 1984). Conversely, ad-
verse conditions could accumulate, resulting in a cost far in excess of that for the typical treatment system
for the same treatment capability.  The following subsections provide the basic insight needed to benefit
from the  above variables.

5.2.1.1     Water Chemistry. The water chemistry can affect capital as well as operating costs. With a
clear picture of the raw water quality, its possible variations and its adverse characteristics, the effect on
the capital cost can be determined readily. Higher water temperature (greater than 100°F) requires higher
cost piping material and/or pipe support. Varying water temperature requires inclusion of special provi-
sions for thermal expansion and contraction. Very high fluoride may require larger media vessels to
reduce regeneration frequency. Higher alkalinity requires higher acid consumption for pH adjustment
resulting in the needs for larger feed pumps, day tank, piping, etc.  This may result in an aeration step for
post treatment pH adjustment in place of caustic addition. High turbidity, iron, manganese, suspended
solids, and/or other contaminants can require the addition of pretreatment steps to accomplish removal
prior to fluoride removal, or the implementation of a different fluoride removal treatment method.

Each of the physical and chemical characteristics of the raw water should be evaluated. The technical as
well as the economical feasibility for the entire project could hinge on these factors.

5.2.1.2     Chemical Supply Logistics. Sulfuric acid (normally 66°B' fhSC^) and caustic soda (normally
50% NaOH) are commercially available and are usually the least expensive chemicals to use for pH
adjustment.  Other chemicals such as hydrochloric acid and caustic potash (KOH) are technically
acceptable, but almost always more costly, and therefore are not commonly used. The acid and caustic
are much cheaper when purchased in bulk quantities, usually 48,000-lb tank trucks.  In  very small plants,
the cost of storage tanks for those volumes is not justified and therefore, drums and carboys with smaller
volumes but higher unit prices are procured. In very large treatment plants, cost can be lowered by
procuring the chemicals via 200,000-lb railroad tank cars. However, this approach requires a rail siding
and rail unloading facility; nevertheless, it does present an option of lowering the overall cost.
                                               55

-------
A chemical unloading rail terminal presents another intriguing option for facilities with multiple treatment
plants. In this approach, smaller site storage tanks are supplied via "mini tank trucks" relaying chemicals
to the treatment site from the rail terminal.  This brings down the size (and cost) of chemical storage tanks
at each site.  However, this could increase the truck traffic of corrosive chemicals through populated
areas, a risk that may not be acceptable.

5.2.1.3     Manual Versus Automatic Operation. Automatic operation is technically feasible. However,
the periodic presence of an operator is always required. The capital costs of automation for computer
hard-ware/software, valve operators, controls, instrumentation, etc., as well as maintenance costs may
exceed budget limits that the client will accept.  Therefore, either manual or semiautomatic operation is
normally furnished. The  advantages and disadvantages of manual, automatic and semiautomatic  opera-
tion require careful evaluation prior to determination of the proper selection.

5.2.1.4     Backwash and Regeneration Disposal Concept.  Regeneration wastewater and waste solids
processing and disposal is not included in the scope of this document. Depending on wastewater
discharge limits established by the U.S. EPA, state and local regulatory authorities, wastewater disposal
can be a significant cost item in the capital  (and operating) cost projection. Requirements can vary from
zero discharge to discharge into  an available existing receiving facility.  Disposal and/or discharge can be
accomplished by chemical precipitation of calcium fluoride or aluminum hydroxide with  subsequent
dewatering of precipitated solids and adjustment of pH. The dewatered solids should pass the U.S. EPA
Toxicity Characteristic Leaching Procedure (TCLP) test and/or California Waste Extraction Test (WET)
test (if in the state of California). The wastewater, though containing low levels of fluoride, will contain
elevated levels of TDS, sodium, and sulfate. If regulatory agency permits disposal by conventional
methods such as surface discharge, percolation, and evaporation pond, the disposal costs are not large.
The total volume of wastewater regeneration generally is  300 to 400 gal/ft3 of AA media.

In the event a zero discharge of wastewater is required, the wastewater supernatant and filtrate (from
solids dewatering) should be fed back to the head of the treatment plant and very slowly added to the raw
water.  This concept, however, has not been incorporated in a full-scale treatment plant.

5.2.1.5     Ambient Conditions

Climate:  Temperature extremes, precipitation, and high wind will necessitate a  building to house the
treatment system equipment. High temperature along with direct sunlight adversely affects the strength
of plastic piping materials. Freezing is obviously damaging to piping and in some extreme cases also to
tanks. Temperature variation introduces requirements for special thermal expansion/contraction
provisions. A building with heating and/or cooling and adequate insulation will eliminate these problems
and their costs, but will introduce the cost of the building. The building cost should reflect wind loads as
well as thermal and seismic requirements.  Operator comfort in place of economic considerations may
dictate building costs.

The installation costs for the buildings and  regeneration wastewater surge tank along with their associated
civil work becomes a major portion of the overall capital  cost.  Care in interpreting the climatological
conditions and their requirements are necessary.

Soil Conditions: Unless soil-boring data are already available for the treatment  system site, at least one
boring in the location of the foundation for each heavy equipment item (treatment vessels, chemical
storage tanks, and regeneration wastewater surge tank) is  required.  If the quality of the soil is
questionable (fill, or very poor load-bearing capacity), additional soil borings should be obtained.  Poor
soil may require costly excavation/backfill  and foundations.
                                               56

-------
Combinations of poor soil with rock or large boulders can make foundation work more complex and
costly. Rock and boulders in combination with extreme temperatures can result in very high installation
costs for subsurface raw, treated, and wastewater pipe mains.

Seismic Zone: Compliance with the seismic design requirements of the local building codes can impact
capital costs. Buildings and tall slender equipment are vulnerable to seismic loads.  The magnitude of
seismic design requirements  should be determined. In zones of extreme seismic activity, low profile
equipment and buildings are  recommended.

100-Year Flood Plan: For water treatment facilities located within a 100-year flood plain, the entire site
should be relocated to another site outside of the 100-year flood plain, be elevated 3 ft above the 100-year
flood plain level, or be protected on all sides by an armored berm that extends a minimum of 3 ft above
the 100-year flood plain level.

5.2.1.6     Existing and Planned (Future) Water System Parameters. Many existing and planned
(future) facility configurations can either significantly increase or decrease the capital cost. The most
important factors are discussed in this section.

Number and Location of Wells: When only one well requires treatment, the removal of fluoride should be
accomplished prior to entering the distribution system.  Theoretically, treatment can occur before or after
entering storage. Practically speaking, treatment prior to entering storage is much easier to control because
the treatment plant flowrate remains the same as the pump flowrate or will be constant.  If treatment takes
place after storage, or if there is no storage, flowrate is intermittent and variable, and pH control is only
achievable for a sophisticated automatic pH control/acid feed system.

When more than one well requires treatment, a decision is required regarding whether a single treatment
plant treating water from all wells manifolded together or individual treatment plants at each well present
would be a more efficient and cost-effective concept.  Factors such as distance between wells, distribution
arrangement, system pressure, and variation in water quality should be evaluated in that decision.  If all of
the wells are in close proximity and pump similar quantity and quality of water, a single treatment plant
serving the entire system is preferable. When wells are widely dispersed, manifolding costs become
prohibitively expensive, thus dictating implementation of individual treatment plants at each well.
Frequently, the distances may be such that the decision is not clear cut; then other variables such as water
quality, system pressure, distribution configuration, land availability should be evaluated.

Systems that require multiple treatment plant installations can achieve cost savings by employing an
identical or similar system at each location. This results in an assembly line approach to procurement,
manufacture, assembly,  installation, and operation. Material cost savings, labor reduction and
engineering  for a single  configuration will reduce the cost for the individual plant.

Potable Water Storage Facilities: Similar to the wells, the number, size, and location of storage tanks can
affect treatment plant size (flowrate) and capital costs. If there is no storage capacity in the system, the
well pump should be capable of delivering  a flowrate equal to the system momentary peak consumption;
this can be many times the average  flowrate for a peak day. Therefore, if no storage capacity exists, a
storage tank should be added with the treatment  system for treated water storage.  Otherwise, automatic
pH instruments and controls  will be required to pace pH adjustment chemical feedrates to the varying
process water flowrate.

Most systems have existing storage capacity. The storage may be underground reservoirs, ground level
storage tanks, or elevated storage tanks (located  on high ground or structurally supported standpipes).
The first two require repressurization; the latter does not. The elevated storage tanks apply a
                                               57

-------
backpressure on the ground level treatment system requiring higher pressure (more costly) construction of
treatment vessels and piping systems. If aeration of treated effluent for pH adjustment is selected with an
elevated storage tank, the treated water should be contained in a clearwell and repumped to storage.
However, the treatment system vessels and piping may be low-pressure construction.  When storage is at
or below ground level storage, loss of system pressure is not a factor.

The amount of storage capacity also affects treatment system cost. The larger the storage capacity (within
limits), the lower the required treatment plant flowrate (and resulting cost). A minimum storage capacity
of one-half of system peak day consumption is recommended.

Distribution and Consumption: The factors that determine the sizing of the treatment system are the well
(or feed) pump flowrate, EBCT, storage capacity, and system consumption characteristics.  Those
features should be coordinated to provide a capacity to deliver a peak treated water supply to satisfy all
possible conditions of peak consumption.  If there is adequate storage capacity, the momentary peaks are
dampened out.  The peak day then defines the system capacity. The well (or feed) pump then is sized to
deliver the  peak daily requirement. The treatment system in turn is sized to treat a minimum of what the
well (or feed) pump delivers.

The distribution system may anticipate future growth or increased consumption. The well (or feed) pump
then should either pump a flow equal to or greater than the maximum anticipated peak daily flows, or be
able to adjust to future increased flowrate. The treatment plant in turn should incorporate capacity to treat
the ultimate peak flowrate or include provisions to increase the treatment capacity in the future.

5.2.7.7     Financial Considerations. Many financial factors should be considered by the designer and
the client.  The client can superimpose financial restrictions (beyond any of the technical factors men-
tioned above) which result in increased (or decreased) capital cost.  These include, but are not limited to,
the following: inflationary trends, interest rates, financing costs, land costs (or availability), cash flow,
labor rates, electric utility rates, and chemical costs. All or some of these factors could affect the capital
investment with reduced operating cost because interest rates are low, inflation  is anticipated, cash is
available, and labor and electric utility rates are high; or the opposite can be true.  The varying
combinations of factors that could develop are numerous; each one will affect the ultimate capital cost.

5.2.2       Capital Costs of Central Fluoride Removal Plants Based on Flowrate.  Capital costs of
central AA plants can be broken down into several categories such as process equipment (including
treatment vessels, AA media, process piping/valves, instrument/controls, chemical storage
tanks/containments, chemical pumps/piping, etc.), process equipment installation (including mechanical,
electrical, painting, etc.), facility infrastructures (including treatment building/concrete, site work,
wastewater pond, etc.), engineering fees and contingency. Capital investment costs can vary significantly
especially with the inclusion of items such as facility infrastructures and various disposal options. For
example, Plant B with a 900-gpm maximum flowrate invested $900,000 for treatment vessels, media, face
piping/valves (not including the bulk piping which was done by a separate contractor),
instrumentation/controls,  spent regenerate treatment (CaF2 precipitation), and chemical
storage/pumps/piping. Plant D also with a 900-gpm maximum flowrate invested over $11,000,000 for
everything stated above under Plant B plus the treatment building, a distribution pumping station, two
storage tanks, and a sand separation device.
                                               58

-------
5.3        Operating Costs

5.3.1       Discussion of Cost Variables. Similar to capital costs, many variables affect operating costs.
The following subsections discuss each of the operating costs previously listed.

5.3.1.1     Treatment Chemical Costs. The treatment chemicals discussed are limited to sulfuric acid
(H2SO4) and caustic (NaOH). Both are highly corrosive, hazardous liquid chemicals that require com-
patible materials of construction, containment provisions, safety provisions, weather protection, and
operator training. Although special precautions and training are required, they are routinely
accomplished. Other acids and bases can be substituted for those chemicals, but they are usually more
costly and therefore rarely considered.  Other chemicals also are used for other requirements such as
corrosion inhibition, precipitation of regeneration wastewater solids, dewatering of precipitated solids in
wastewater, and disinfection; however, these are site-specific requirements that are not covered in this
manual.

The chemicals used for treatment of water for public consumption require NSF/ANSI STD 60
certification by most state regulatory agencies. It is also recommended that chemical suppliers be
required to certify that containers used to store and deliver the chemicals have not been used for any other
chemical; or if they have, that they have been decontaminated according to procedures  required by
governing regulatory agencies.

Chemical costs are variable and have been volatile. Like all commodities, they are sensitive to the supply
and demand fluctuation of the marketplace.  The geographic location of a plant site in relation to that of a
supplier has an impact on the delivered cost.  In some cases, delivery costs  are greater than chemical
costs. The conceptual design evaluates the chemical logistics and determines the  most  cost-effective
mode of procurement as well as whether chemicals for pH adjustment are economically feasible.

Chemical costs are sensitive to the volume and containment mode of the commodity purchased. Because
commodity handling is minimized, bulk tank truck quantities entail the least cost.  Tank truck quantities
are normally 48,000 Ib.  Bulk deliveries require chemical storage tanks within containment basins located
at the plant site with necessary safety provisions and weather protection. The same commodities can be
routinely purchased in drums (55-gal or 30-gal), totes, carboys, gallon jugs, etc. These packaged
quantities result in much higher unit prices than bulk quantities. Drum and other  small container prices
also depend on the quantities procured at one time.  Small containers also introduce additional handling
requirements for the treatment plant operator.  For very small treatment systems, bulk procurement and
storage is not justified unless the feedwater fluoride and alkalinity concentrations  are extremely high. In
special low flowrate systems where high fluoride and high alkalinity concentrations are present in the
feedwater and drum quantity costs are significantly higher than bulk quantity costs, the increased
chemical consumption could justify bulk purchase.

The chemistry of the raw water to be treated is the most significant factor affecting treatment chemical
consumption and costs.  Fluoride and alkalinity are the key ions in the raw water;  the higher the
concentration of either ion, the higher the chemical consumption and cost per 1,000 gal of treated water.

Acid Costs: The most cost-effective, commercially available chemical for lowering pH is concentrated
sulfuric acid.  Hydrochloric acid is also applicable, but it is more difficult to handle, increases chlorides
(i.e., is corrosive), and usually is more costly.  The chemical designation of commercially available
sulfuric acid is 66°B' H2SO4. Its concentration is 93.14%.  The remaining 6.86%  is water (plus other
ions). The other ions that could be present should be evaluated and could result in a slight increase in
their concentration in the treated water. Frequently, small quantities of iron and trace amounts of heavy
metals are present. For water treatment service, there are stringent limits on the levels of contaminants in
                                               59

-------
the acid which will be rigidly enforced.  NSF certification of the acid for use as an additive in drinking
water is required.

The most economical method of procuring acid is in bulk tank truck quantities (48,000 Ib), which are
3,100 gal each. The tank trucks are loaded at each acid manufacturer's site or at a distribution storage site
and delivered directly to the treatment plant where the acid is transferred to the acid bulk storage tank.
Transfer is accomplished by means of compressed air, which is provided by an air compressor on the
truck (unless the treatment plant can provide the compressed air). In addition to the lower commodity
price resulting from minimum handling and storage of the chemical, there is minimum chance of
contamination.  At large treatment plants where there is potential for high acid consumption, rail tank car
quantity (200,000 Ib) delivery, which is cheaper, may be justified.  Capital expenditures for a 16,000-gal
(minimum) storage tank and a rail spur with unloading equipment then are required.

The delivered costs of bulk tank truck quantities of sulfuric acid increased significantly during the last
years.  In 2003, the bulk delivery cost ranged from 4.5 to 60/lb (or $90 to $120/ton) depending on the
geographic location of the treatment plant. (Drum quantity costs ranged from 10 to 120/lb  or higher.)  In
2013, the delivered cost to Plant A is 12.50/lb (or  $250/ton), more than twice a plant would pay in 2003.

The acid is consumed in three phases of the treatment process at every fluoride removal plant. First, it is
used to adjust the raw water pH to the treatment requirement (5.5); second,  it is used to rapidly neutralize
the treatment bed immediately after regeneration;  finally, it may be used for pH adjustment of the
regeneration wastewater for discharge to sewer or other receiving facilities. This final application does
not apply to treatment systems that discharge regeneration wastewater to lined evaporation ponds. The
raw water alkalinity dictates the amount of acid required for the pH adjustment step. The AA fluoride
removal process has been employed on natural waters  with alkalinity values ranging from 10 to 1,500
mg/L (as CaCO3) (Rubel, 1984).

The acid consumption for pH adjustment can be accurately projected by running a titration on a raw water
sample. The cost of acid required for pH adjustment is then determined by  extending the acid addition in
mg/L to the weight (Ib) required per 1,000 gal and multiplying by the  commercial cost for the acid.

For the design example presented in Appendix B,  a hypothetical feedwater analysis includes the
following:

                              Total alkalinity (M) = 220 mg/L (as CaCO3)
                                           pH     = 8.0.

Based upon titration results, the quantity of 66°B'  H2SO4 required to adjust the pH to 5.5 is 205 mg/L.
The amount of acid required per 1,000 gal treated  water is as follows:

                       205mg   1(T6  kg       Ib                  3.785L
                      	-x	-x	x 1000 gal  x	=
                         L       mg    0.4545kg              gal
                                        1.7 lib/1,000 gal

Therefore, for an acid bulk quantity price of 12.50/lb, the acid cost per 1,000 gal treated water is 21.250.
If the acid is procured in drum quantities at 400/lb, the resulting cost is 67.50/1,000 gal. Conversely, if
the feedwater total alkalinity had been 100 mg/L as CaCOs and the pH 7.5,  then the resulting acid
required to adjust pH to 5.5 would be 92.4 mg/L.  That equates to 0.77 lb/1,000 gal, or 9.750/1,000 gal
(for acid bulk quantity price of 12.50/lb). The acid requirement for pH adjustment  in Appendix B is
9.690/1,000 gal.
                                               60

-------
The acid consumption for neutralization after regeneration is a function of the caustic concentration
employed during regeneration and the raw water alkalinity.  This quantity varies from plant to plant.  The
consumption also is a function of the raw water fluoride level, which dictates the regeneration frequency,
and the volume of water over which this cost is distributed.  The higher the fluoride level, the fewer
gallons treated per treatment cycle. When projecting chemical costs and volumes, a rule of thumb is
10,000 gal of treated water per cycle per ft3 of AA with 6 mg/L raw water fluoride (this decreases to
4,000 gal/ft3 at 20 mg/L fluoride and increases to 16,000 gal/ft3 at 3 mg/L fluoride) (Figure 5-1). The
weight of acid required for neutralization after regeneration  is normally in the range of 1 to 2 lb/ft3 of the
AA media.

Using acid bulk quantity price of 12.50/lb, the  acid costs for neutralization after regeneration will fall in
the range of $1.44 to 2.880/1,000 gal of treated water.
    14
    12
              \
 UJ
 >
 til
 UJ
 Q
 C  10
 o
 <£
 UJ
                       6        a       10       12       14       16       18
                        THOUSAND GALLONS  OF TREATED WATER / cu ft Media
                                                                                  20
          Figure 5-1.  Curve Illustration Rule of Thumb for Volume of Water Treated per
                               Cycle vs. Raw Water Fluoride Level
Caustic Costs: Caustic (NaOH) can be procured in either solid (100% NaOH) or liquid (50% NaOH or
lower).  The 50% NaOH is the standard concentration that is handled and applied to water treatment
applications.  That concentration is a byproduct of the chlorine manufacturing process. Therefore, it
requires minimum handling to place it into a 48,000-lb bulk tank truck (3,850 gal).  One challenge with
50% NaOH concentration is that it freezes at 55°F; it is also very viscous and difficult to transfer at
temperatures below 70°F. Therefore, it normally requires heating. Also, because it is 50% water by
weight, the freight is a cost factor. Solid caustic in bead or flake form is also readily available in drums or
                                               61

-------
bulk.  Its freight cost is roughly half that of the liquid, but getting it into solution is difficult and danger-
ous. Regardless of the economics, solid caustic is not recommended for this application. Commercially
available caustic in the 25% NaOH concentration has a freezing point of 0°F; however, freight costs for
shipping this material  are high (75% water).  Capital costs for larger storage and pumping requirements
are also increased. Even though heating and temperature protection are required, the 50% NaOH is
recommended.  Transferring caustic from tank trucks to storage tanks is accomplished with compressed
air similar to the method for acid.

Similar to sulfuric acid, delivered costs of bulk tank truck quantities of 50% NaOH increased significantly
in the last 10 years.  In 2003, the bulk delivery costs ranged from 10 to 150/lb (or $200 to $300/ton)
depending on the geographic location of the treatment plant.  In 2013, Plant A paid 210/lb ($420/ton) for
the bulk delivery.

The caustic is consumed at two phases of the treatment process. First, it is used to raise the pH of the
treated water to the level desired for distribution; second; it is used to raise the pH of the raw water to the
level required for  spent media regeneration.  The first requirement may be reduced or replaced by aeration
of the treated water to  strip free CO2 from the treated water.

The volume of 50% NaOH required for a 1% NaOH concentration regeneration (includes upflow and
downflow requirements) is 0.4 gal (5 Ib) per ft3 per regeneration cycle. As with the acid required for
neutralization, the caustic consumption is a function of the raw water fluoride level, which  dictates the
regeneration frequency and the volume of water over which this cost is distributed. This varies from
treatment system to treatment system.

The caustic consumption for treated water pH adjustment is also a function  of raw water alkalinity and the
desired treated water pH. The concentration of free CO2 in the water after the initial pH adjustment with
sulfuric acid will determine the caustic requirement. High CO2 concentration or community objection to
addition of sodium to the water supply could dictate the aeration method for pH adjustment. In general,
when cost dictates the  method, caustic pH adjustment is recommended when alkalinity is less than 100
mg/L and aeration is recommended when alkalinity is over 200 mg/L. In the alkalinity range of 100 to
200 mg/L, a general recommendation is difficult; other factors such as storage tank elevation must be
considered.  If caustic  is used to raise the pH of the treated water, the quantity will be small. The con-
sumption requirement can again be accurately determined by continuing the original titration required for
acid to lower the pH to the treatment level of 5.5; then adding the 50% NaOH required to raise the pH to
the desired level (e.g.,  7.5).  The cost of caustic required then is determined by extending the caustic
addition in mg/L to the weight required  per 1,000 gal and multiplying by the commercial price for the
delivered caustic.  The actual caustic cost will normally fall in the range of $0.02 to $0.12/1,000  gal of
treated water.

For the design example presented in Appendix B for which the feedwater pH had been adjusted to 5.5 for
treatment, the treated water pH is readjusted back to a desired level (e.g., pH 7.7).  The 50% NaOH
requirement determined by titration is 210 mg/L. The required quantity of 50% NaOH per 1,000 gal
treated water is  as follows:

          (210 mg/L  x 1,000 gal x 3.7854 L/gal)/(1000 mg/g x 453.6 g/lb) = 1.75 lb/1,000 gal

Therefore, at a caustic bulk quantity price of 210/lb, the caustic cost per 1,000 gal is 36.80/1,000 gal. The
caustic used in the estimated operating cost example in Appendix B is 36.60/1,000 gal.
                                               62

-------
Using the same activated alumina fluoride capacity (129,800 mg/ft3 [2,000 grains/ft3]) and volume of
water treated per treatment cycle (2,700,000 gal/vessel) discussed in Section 4.a of Appendix B, the cost
of caustic soda is as follows:

     (0.4 gal/ft3/regeneration cycle.x 312 ft3/vessel xi2.9 Ib/gal of 50%NaOH x 21e71b)/(2,700,000
                           gal/vessel/regeneration cycle) = 12.50/1,000 gal

Therefore, for the example provided in Appendix B, caustic soda cost is 49.10/1,000 gal of water treated.

5.3.1.2     Operating Labor Costs.  Operating labor costs are difficult to quantify. The operator is
required to be dependable and competent; however, the position is not always full-time. Depending on
the size of the system and the other duties available for the operator, the operator's time should be
distributed over several accounting categories. Except for days when  spent media regeneration takes
place, the treatment plant normally requires less than 2 hr per day of operator attention.  During regenera-
tion, the operator may be required to spend approximately 8 hr over a  12-hr period.

On routine operating days, the operator checks the system to see that pH is being controlled, takes and
analyzes water samples, checks instruments (flow, temperature, pressure), and makes entries in daily logs.
The only exceptions to the normal  routine include special activities such as media regeneration, equip-
ment maintenance and chemical tank truck deliveries.  During the remainder of the time, the operator is
able to operate and maintain other  systems (distribution, pumps, storage, etc.), read meters, or handle
other municipal responsibilities (e.g., operate  sewage treatment plant). There should always be a second
operator available to take over in case of an emergency; that individual should be well versed in the
operation of the plant.

Using the example treatment plant presented in Appendix B, the cost of operational labor will be as
follows (it is assumed that the hours not used  for treatment plant operation will be efficiently used on
other duties):

   Given:
   Flowrate =  600 gpm
   Annual average utilization  =       40%
   Number of regenerations per year   =       52
   Number of hours daily      =       1.5hr
   Operator annual salary      =       $35,000
   Overhead and fringe benefits        =       30%
   Available hours per year   =       2,000/man

   Then:
   Number of hours on
     regeneration/year 52  x 8 =       416 hr
   Number of hours on routine
     operations/year   1.5 x (365-52) =       469.5 hr
   Total plant operator time   =       885.5 hr

   Operator hourly rate     35,000/2,000       =       $17.50/hr
   30% (overhead and fringe benefits)  =       $ 5.25/hr
   Operator Rate       $22.75/hr
      Total operator cost: 885.5 hr/year x $22.75/hr =
      $20,145/year
                                               63

-------
      Total gallons water produced:
      0.4(600 gpm) x 1,440 min/day x 365 days/year =
      126,144,000 gal/year

      Labor cost/1,000 gal $20,145/126,144 (1,000 gal) =
      $0.16/1,000 gal.

If the operator had no other responsibilities and the operator's entire salary were expended against this
treatment plant operation, the operating labor cost would become $0.36/1,000 gal.  Obviously, there are
many variables, which can be controlled in different ways. Depending on the operational philosophy of
the designer/planner/manager, the operating labor cost can be minimized or maximized over a very broad
range. In the case of a very high production plant, the operating labor requirement is not significantly
larger than that for a very small treatment plant.  Therefore, depending on relative salaries, the resulting
costs per 1,000 gal can range from a few cents to more than a dollar.  In proper perspective, the operating
labor cost should fall in the $0.06 to $0.20/ 1,000-gal range.

If the treatment plant in the example in Appendix B had used automatic operation in place of manual
operation, the operating labor costs might be lower. However, because a higher skilled operator is
required to maintain and calibrate  the more sophisticated instrumentation and control equipment, the
operating labor cost may not be lower. Therefore, no reduction of operating labor cost is assumed for
systems with automatic operation.

5.3.1.3      Utility Costs.  Utility costs are normally electricity bills. However, there can also be
telephone and natural gas (or oil) utility costs. Telephone service to the treatment building is recom-
mended as a safety precaution in case of accident as well as operator convenience.  Costs for that service
should be the minimum available monthly rate. Depending upon the local climate, the costs for heating
can vary. The purpose of the building is to protect the equipment from elements (primarily freezing), not
for operator comfort. Normally the treatment units act as heat sinks, maintaining an insulated building at
a temperature near that of the raw  water. In cold climates, the building should have an auxiliary heat
source to prevent freezing of pipes in the event that the water is not flowing.  If the client determines that
the treatment building is to serve additional functions, heating to a comfort temperature could be an
additional required cost.

Electric power is needed for the following functions:

            (1) Chemical pumps
            (2) pH controls
            (3) Caustic storage tank immersion heater
            (4) Lighting

            (5) Convenience receptacle
            (6) Aeration unit blower (optional)
            (7) Repressurization pump (optional)
            (8) Extra load on well pump for regeneration/backwash wastewater, and loss of head through
               the treatment system.

Items 1, 2, 4, and 5 are negligible. Item 3 is a function of the climate and the heat losses through the
insulation.  Provisions to conserve energy for this function should be incorporated.  Item 6 is a relatively
small load (1-3 hp blower motor).  Item 7 is potentially the biggest electrical load.  This requirement only
                                               64

-------
exists when aeration is used to adjust treated water pH, and the water is pumped to an elevated storage
tank.  This electrical load can be equal to the well pump motor load. However, when repressurization is a
requirement, then the well pump should be modified to reduce its discharge pressure capability to only
that which is required to pump the raw water through treatment into the clearwell in place of the pressure
to pump to the elevated storage tank. Then the net increase of electrical energy consumption is nearly
negated.  Item 8 amounts to 3-5% of the well pump electrical energy consumption.

The electrical utility rates also vary considerably from one geographic location to another.  According the
U.S. Energy Information Administration, industrial electricity rates in April 2013 varied from $0.05 to
$0.26/kWh (http://www.eia.gov/electricitv/monthly/epm table grapher.cfm?t=epmt 5  6 a).  Under
normal operating conditions, the electrical utility costs can range from $0.01 to $0.10 per 1,000 gal of
water treated.

5.3.1.4     Media Replacement Costs.  The consumption of treatment media per regeneration for a
system with process water pH adjustment and spent media regeneration in a well-operated AA removal
water treatment plant is about 1% of the bed volume (Plant A reported a loss of 2.6% per regeneration
cycle). The loss of media is caused by dissolution of a small amount of aluminum  during the regeneration
and neutralization steps where excessively high and/or low pH solutions are in contact with AA. If the
pH of the regeneration solution exceeds 1.5% NaOH, the solution becomes increasingly aggressive to
AA. Similarly, if the pH of the neutralization solution is lower than pH 2.0, a more drastic AA
dissolution takes place.

There are additional ways in which the  media can be lost. Backwash, if conducted carelessly, can result
in media carry over. An excessive backwash rate can expand the AA bed by an amount that carries the
media out of the vessel.  Monitoring the backwash water will detect and provide prevention of that. If
backwash water flows into the wastewater surge tank, the lost media can be recovered.

Another way to lose media is through the effluent underdrain (collection system) within the vessel.  If
media grains ever appear in the treated effluent, the treatment unit should be immediately taken out of ser-
vice for inspection (and repair) of the collection system.

Media replacement costs are difficult to predict. Significant media replacement can occur at a treatment
plant where backwash at an excessive rate for an extensive period has been required to remove filtered
solids from the media. A plant in which suspended solids in the raw water require  frequent extended
backwashing is vulnerable to media loss problems.  For systems encountering such conditions an
upstream filter (e.g., bag filter) should be evaluated.

A typical pricing structure for a representative AA product in 2001 is provided in Table 5-1. In 2013,
Plant A pays $1.73/lb for 37,000 Ib of media in 20 supersacks (each containing 1850 Ib).
                Table 5-1. 2001 Price for Typical -28, +48 Mesh Activated Alumina
Quantity
2,000-10,000 Ib
12,000-20,000 Ib
22,000-38,000 Ib
40,000 Ib and over
Price
$1.00/lb
0.90/lb
0.75/lb
0.70/lb
                                               65

-------
Assuming a 1% loss per regeneration cycle and 24 regeneration cycles per year, a conservative bed media
replenishment is 24% per year. In the example in Appendix B where two 312 ft3 beds are used, the media
replacement will be:

                       2 x 312 ft3 x 48 lb/ft3 x $1.73/lb x 0.24 = $12,436/year

$12,4367(2,700,000 gal/regeneration cycle/vessel x 2 vessel x 24 regeneration cycles) = $0.096/1,000 gal.

5.3.1.5     Replacement Parts and Miscellaneous Material Costs. This is a very small operational cost
item.  Replacement parts (e.g., chemical, pump diaphragms, seals and replacement pump heads) should be
kept in stock in the treatment plant, to prevent extended plant shutdown in the event a part is required.
Also included are consumables such as laboratory reagents (and glassware), and recordkeeping supplies.
An operating cost allowance of $0.02/1,000 gal of treated water is conservative.

5.3.2       Operating Cost Summary. Table 5-2 summarizes the total operation cost based on the
example provided in Appendix B. The $0.94/1,000 gal cost is very close to  the $1.23/1,000 gal annual
operating cost reported by Plant A. As has been pointed out, the operating costs can vary significantly
from plant to plant.

For AA fluoride  removal plants in which flowrates, raw water arsenic concentration, raw water analyses
(pH, alkalinity, silica, sulfate, etc.), adsorptive media, labor rates, and utility rates vary from the values
used in the example in Appendix B, the operating costs will deviate from those indicated in Table 5-3.
The information  included in this subsection provides a method for the determination of an operating cost
estimate for any AA fluoride removal plant.
             Table 5-2.  Operating Cost Breakdowns for an Activated Alumina Plant(a)
                   Operating Cost Items	$/1,000 gal Treated Water
     Treatment Chemicals - acid                                          0.12
                        - caustic                                       0.49
     Operating Labor                                                   0.16(b)
     Utility                                                            0.05
     AA Media Replenishment                                           0.10
     Replacement Part and Misc. Material                                  0.02
                                            TOTAL                   0.94
     (a)  System flow rate 600 gpm; manual operation; wastewater and waste solids, processing and
         disposal not included.
     (b)  Applicable to automatic operation.
                                               66

-------
                                     6.0: REFERENCES
American Public Health Association-American Water Works Association-Water Environment Federa-
       tion. Standard Methods for the Examination of Water and Wastewaters, 19th ed. A.E. Greenberg
       (Ed.), American Public Health Association, Washington, DC.

American Water Works Association (AWWA) .2011. Water Quality and Treatment: A Handbook of
       Community Water Supplies. Chapter 9. Ion Exchange and Inorganic Adsorption. 6th ed., Denver,
       CO.

American Water Works Association (AWWA). 2004. Manual of Water Supply Practices -M4. Water
       Fluoridation Principles and Practices. Chapter 5. Defluoridation. 5th ed., Denver, CO.

Amor, Z., S. Malki, M. Taky, B. Bariou, N. Mameri and A. Elmidaoui. 1998. "Optimization of Fluoride
       Removal from Brackish Water by Electrodialysis." Desalination, 120: 263-271.

Arora, M., R.C. Maheshwari, S.K. Jain and A. Gupta. 2004. "Use of Membrane Technology for Potable
       Water Production." Desalination, 170(2): 105-112.

Bhatnagar, A., E. Kumar and M. Sillanpaa.  2011. "Fluoride Removal from water by adsorption - A
       Review." Chem. Engrg. J.,171: 811-840.

Bishop, P.L. and G. Sansoucy.  1978. "Fluoride Removal from Drinking Water by Fluidized Activated
       Alumina Adsorption." JAWWA, 70: 554-559.

Bulusu, K.R, Y.S. Murty, B.N. Pathak, W.G. Nawlakhe and D.N. Kulkarni. 1983. "Performance of
       Defluoridation Plant at Kadiri." J Inst. Engs.  (Indian), 64: 35-40.

Bulusu, K.R, B.B. Sunderasan, B.N. Pathak, W.G. Nawlakhe, D.N. Kulkarni and V.P. Thergaonkar.
       1979. "Fluoride in Water, Defluoridation Methods and Their Limitations." J. Inst. Engs. (Indian),
       60: 1-25.

Centers for Disease Control and Prevention (CDC). 1993. Fluoridation Census 1992 - Vol. 1. U.S.
       Department of Health & Human Services, Public Health Service, Centers for Disease Control and
       Prevention Services, National Center for Prevention Services, Atlanta, GA.

Chauhan, V.S., P.K. Dwivedi and L. lyengar. 2007. "Investigations on Activated Alumina Based
       Domestic Defluoridation Units." J.  Haz. Mater., 139:103-107.

Choi, W.-W. and K.Y. Chen. 1979. "The Removal of Fluoride from Waters by Adsorption." JAWWA,
       XX(S):562-57Q.

Clifford, D.A. 1999. Ion Exchange and Inorganic Adsorption. Chapter 9  in Water Quality and Treatment.
       A Handbook of Community Water Supplies. 5th ed., AWWA, McGraw-Hill, New York, NY.

Cohen, D. and H. Conrad. 1998."65,000 GPD Fluoride Removal Membrane System in Lakeland,
       California, USA" Desalination, 777:19-35.

Cooperative Research Center (CRC). 2008. Investigation of Defluoridation Options for Rural and Remote
       Communities. Research Report 41. The Cooperative Research Center for Water Quality and
                                              67

-------
       Treatment, Salisbury SA, Australia.

Cotton, F.A. and G. Wilkinson. 1988. Advanced Inorganic Chemistry. John Wiley & Sons, New York,
       NY.

Das, N., P. Pattanaik and R. Das. 2005. "Defluoridation of Drinking Water Using Activated Titanium
       Rich Bauxite." J. Colloid Interface Sci., 292:311-323.

De Haan, M. 2011. Projections on New Fluoride Regulatory Activities and Treatment Options. Presented
       atAWWA  Water Quality Technology Conference, Phoenix, AZ.

Diawara, C.K., L. Paugam, M. Pontie, J.P. Schlumpf, P. Jaouen and F. Quemeneur. 2005. "Influence of
       Chloride, Nitrate and Sulphate on the Removal of Fluoride Ions by Using Nanofiltration
       Membranes."^. Sci. & Technol.40:3339-3341.

Environmental Protection Agency (EPA). 2012. Radionuclides in Drinking Water, Compliance Options:
       Treatment Technology Descriptions.
       http://cfpub.epa.gov/safewater/radionuclides/radionuclides.cfm

Environmental Protection Agency (EPA). 2006. Point-of-Use or Point-of-Entry Treatment Options for
       Small Drinking  Water Systems. EPA-815-R-06-010. U.S. EPA, Office of Water, Washington,
       D.C.

Environmental Protection Agency (EPA). 2003. Occurrence Estimation Methodology and Occurrence
       Findings Report for Six-year Review of National Primary Drinking Water Regulations. EPA/815-
       R-03-006. U.S. EPA, Washington, D.C.

Environmental Protection Agency (EPA). 1998. Small System Compliance Technology: List for the Non-
       Microbial Contaminants Regulated Before 1996. EPA 815-R-98-002. U.S. EPA, Office of Water,
       Washington D.C.

Environmental Protection Agency (EPA). 1986. National Primary and Secondary Drinking Water
       Regulations. Fluoride. Final Rule. U.S. EPA, 40 CFR Parts 141, 142, and 143, Federal Registry,
       Vol. 51, No. 63. P. 11396-11412.

Environmental Protection Agency (EPA). 1980. Pilot Study of Fluoride and Arsenic Removal from
       Potable Water. Municipal Environmental research Laboratory. U.S. EPA, Cincinnati, OH.
Fan, J., Z. Xu and  S. Zheng, S. 2007. "Comment on Factors Influencing the Removal of Fluoride from
       Aqueous Solution by Calcined Mg-Al-CO3 Layered Double Hydroxides." J. Hazard. Mater.,
       139: 175-177.
Fan, X., D.J. Parker, and M.D. Smith. 2003. "Adsorption Kinetics of Fluoride on Low Cost Materials."
       Wat. Res., 37: 4929-4937.
Fawell J., K. Bailey, J. Chilton, E. Dahi, L. Fewtrell and Y. Magara. 2006. Fluoride in Drinking Water.
       World Health Organization (WHO), London, UK.

Federal Register. 2011. Vol. 76, No. 9, January 13.

Fox, K.R. and T. J. Sorg. 1987. "Controlling Arsenic, Fluoride, and Uranium by Point-of-Use
       Treatment." JAWWA, 79(1): 81-84.
                                             68

-------
Gandhi, N., D. Sirisha, K.B. Chandra Shekar and S. Asthana. 2012. "Removal of Fluoride from Water
       and Waste Water by Using Low Cost Adsorbents." Int'lJ. ChemTech Res., 4(4):  1646-1653.

Ghorai, S. and K.K. Pant. 2004. "Investigations on the Column Performance of Fluoride Adsorption by
       Activated Alumina in a Fixed Bed." Chem. Engrg. J. (India), 98: 165-173.

Goswami, A. and M.K. Purkait. 2012. "The Defluoridation of Water by Acidic Alumina." Chem. Engrg.
       Res. And Design, 90(12): 2316-2324.

Hao, O.J. and C.P. Huang. 1986. "Adsorption Characteristics of Fluoride onto Hydrous Alumina." J. Env.
       Engr., 112(6): 1054-1069.

Health Canada. 2010. Guidelines for Canadian Drinking Water Quality: Guideline Technical Document-
       Fluoride. Water, Air and Climatic Change Bureau, Health Canada, Ottawa, Ontario.

Huxstep, M.R. and T.J. Sorg.  1988. Reverse Osmosis Treatment to Remove Inorganic Contaminants from
       Drinking Water. EPA/600/S2-87/109. U.S. EPA, Water Engineering Research Laboratory.
       Cincinnati, OH.

Jagtap, S., M.K. Yenkie, S. Das and S. Rayalu. 2011. "Synthesis and Characterization of Lanthanum
       Impregnated Chitosan Flakes for Fluoride Removal in Water." Desalination, 273(2-3): 267-275.

Jing, C., J. Cui, Y. Huang and A. Li.  2012. "Fabrication, Characterization, and Application of a
       Composite Adsorbent for Simultaneous Removal of Arsenic and Fluoride. "^4/>/>/. Mater.
       Interfaces, 4(2): 714-720.

Kamble, S.P., P. Dixit, S.S. Rayalu and N.K. Labhsetwar. 2009. "Defluoridation of Drinking Water Using
       Chemically Modified Bentonite Clay." Desalination, 249(2):6&7-693.

Kamble, S.P., S. Jagtrap, N.K. Labhsetwar, D. Thakare, S. Godfrey, S. Devotta and S.S. Rayalu. 2007.
       "Defluoridation of Drinking Water Using Chitin, Chitosan and Lanthanum-modified Chitosan."
       Chem. Engrg. J., 729:173-180.

Lhassani, A., M. Rumeau, D. Benjelloun, and M. Pontie. 2001. "Selective Demineralization of Water by
       Nanofiltration Application to the Defluoridation of Brackish Water." Wat. Res., 35(13):3260-
       3264.

Liao, X.P. and B. Shi. 2005. "Adsorption of Fluoride on Zirconium (IV)-Impregnated Collagen Fiber."
       Environ. Sci. Techno!., 39(12):4628-4632.

Mackay, K.M. and R.A. Mackay. 1989. Introduction to Modern Inorganic Chemistry. 4th Ed. Prentice
       Hall, Englewood Cliffs, New Jersey.

Maier, F.J. 1953. "Defluoridation of Municipal Water Supplies." JAWWA, 45(8): 879.

Maier, F.J. 1947. "Methods of Removing Fluoride from Water. "Am. J. Public Health, 37(12): 1559.

Maliyekkal, S.M., K.R. Anshup and T.P. Antony. 2010. "High Yield Combustion Synthesis of
       Nanomagnesia and Its Application for Fluoride Removal." Sci.Total. Env., 408: 2272-2282.
                                             69

-------
Maliyekkal, S.M., S. Shukla, L. Philip and M.N. Indumathi. 2008. "Enhanced Fluoride Removal from
       Drinking Water by Magnesia-Amended Activated Alumina Granules. " Chem. Eng. J., 140: 183-
       192.

Maliyekkal, S.M., A.K. Sharma and L. Philip. 2006. "Manganese-Oxide-Coated Alumina: A Promising
       Sorbent for Defluoridation of Water." Water Res., 40:3497-3506.

Meenakshi, R.C. and S.K. Maheshwari. 2006. "Fluoride in Drinking Water and Its Removal." J. Haz.
       Mat., 137:456-463.

Mekonen, A., P. Kumar and A. Kumar. 2001. "Integrated Biological and Physiochemical Treatment
       Process for Nitrate and Fluoride Removal." Wat. Res., 35(13): 3127-3136.

Miretzky, P. and A.F. Cirelli. 2011. "Fluoride Removal from Water by Chitosan Derivatives and
       Composites: A Review." J. Fluorine Chem., 132(4):23l-240.

Mohan, D. and C.U. Pittman Jr. 2007. "Arsenic Removal from Water/Waste Water Using Adsorbents - A
       Critical Review." J. Haz. Mater. 142: 1-53.

Mohan, D., R. Sharma, V.K. Singh, P. Steele, and C.U. Pittman. 2012. "Fluoride Removal from Water
       Using Bio-Char, a Green Waste, Low-Cost Adsorbent: Equilibrium Uptake and Sorption
       Dynamics Modeling." Ind. Eng. Chem.  Res., 51(2): 900-914.

Mohapatra, M., S. Anand, B.K. Mishra, D.E. Giles and P. Singh. 2009. "Review of Fluoride Removal
       from Drinking Water." J. Environ. Mamgt, 97:67-77'.

National Research Council (NRC). 1993. Health effects of ingested fluoride. National Academy Press,
       Washington, D.C.

National Research Council (NRC). 2006. Fluoride in Drinking Water: A Scientific Review of EPA 's
       Standards. National Academy Press, Washington, D.C.

Natural Health Enterprises. 2013. Fluoride Removal Water Systems at:
       http: //www. waterus .com/fluoride removal .htm.

Natural Resource Management Ministerial Council (NHMRC). 2011. Australian Drinking Water
       Guidelines. National Health and Medical Research Council, Natural Resource Management
       Ministerial  Council, Australian Government.

Nawlakhe, W.G. and A.V.J. Rao. 1990. "Evaluation of Defluoridation Plant at Tartatur." J. Indian Water
       Works Assoc., 13(3): 287-290.

Nigussie, W., F. Zewge and B.S. Chandravanshi. 2007. "Removal of Excess Fluoride from Water Using
       Waste Residue from Alum Manufacturing Process." J. Haz. Mat.,  147: 954-963.

NSF International. 2013a. NSF Certified Products: Drinking Water Treatment Units. Web Search for
       Point-of-Use Reverse Osmosis Devices Certified for Fluoride Removal. Available online at:
       http://www.nsf.Org/Certified/D WTU/Listings.asp?ProductFunction=058%7CFluoride+Reduction
       &ProductTvpe=&submit2=SEARCH. Accessed June 10, 2013.
                                             70

-------
NSF International. 2013b. NSF Consumer Information: Contaminant Testing Protocols. Available at:
       http://www.nsf.org/consumer/drinking water/dw contaminantjrotocols.asp?program+WaterTre

National Fluoridation Information Service. 2012. Household Water Treatment Systems for Fluoride
       Removal. National Fluoridation Information Service. At
       http://www.rph.org.nz/content/cc35652b-aeOb-4deO-a200-4f617206495f.cmr.

Onyango, M.S. and H. Matsuda. 2006. Advances in Fluorine Science, Chapter 1. Fluoride Removal from
       Water Using Adsorption Technique. Elsevier, New York, NY.

Pickard, B. and M. Bari. 2004. "Feasibility of Water Treatment Technologies for Arsenic and Fluoride
       Removal from Groundwater." Presented at AWWA WQTC, San Antonio, TX.

Public Health Service (PHS).  1969. Natural Fluoride Content of Community Water Supplies. DHEW
       Public Health Service, NIH, Bethesda, MD.

Reid, B. 2013. E-mail communication on December 30, 2013.

Rubel, F. 1984. Design Manual: Removal of Fluoride from Drinking Water Supplies by Activated
       Alumina. EPA-600/2-84-134. U.S. Environmental Protection Agency, Municipal Environmental
       Research Laboratory, Cincinnati, OH

Rubel, F., and F. Williams.  1980. Pilot Study of Fluoride and Arsenic Removal from Potable Water.
       EPA-600/2-80-100. August.

Rubel, F and R.D. Woosley. 1978. Removal of Excess Fluoride from Drinking Water. EPA 570/9-78-001.
       Office of Water Supply, U.S. Environmental Protection Agency, Washington B.C.

Schoeman, J.J. 2009. "Performance of a Water Defluoridation Plant in a Rural Area in South Africa."
       Water SA, 35(1): 97-101.

Schoeman, J.J. and G.W. Leach. 1987. "An Investigation of the two Newly Installed Defluoridation
       Plants in South Africa and Some Factors Affecting Their Performance." Water Sci. & Technol,
       19(5-6):953-965.

Schoeman, J.J., V.R. Oppel, and O. Novhe. 2006. "Commissioning and Performance of a Water
       Defluoridation Plant in a Rural Area in South Africa." Presented at WISA  2006 Biennial
       Conference, Durban, South Africa.

Sehn, P. 2008. "Fluoride Remvoal with Extra Low Energy Reverse Osmosis Membranes: Three Years of
       Large Scale Field Experience in Finland." Presented at Conf. on Desalination and the
       Environment, Halkidiki, Greece.

Shivastava, B.K. and A. Vani. 2009. "Comparative Study of Defluoridation Technologies in India." Asian
       J. Exp. Sci., 23(1): 269-274.

Sivasamy, A., K.P. Singh, D. Mohan and M. Maruthamuthu. 2001. "Studies on Defluoridation of Water
       by Coal-Based Sorbents" J.Che m. Technol. & Biotechnol, 76(7):  717-722.

Sorg, T.J. 1978.  "Treatment Technology to Meet the Interim Primary Drinking Water Regulations for
       Inorganics." JAWWA, 700:105-112.
                                             71

-------
Sternitzke, V., R. Kaegi, J.-N. Audinot, E. Lewin, J.G. Hering, and C.A. Johnson. 2012. "Uptake of
       Fluoride from Aqueous Solution on Nano-Sized Hydroxyapatite: Examination of a Fluoridated
       Surface Layer." Environ. Sci. Technol., 46: 802-809.

Sujana, M.G. and A. Anand. 2010. "Iron and Aluminum Based Mixed Hydroxides: A Noval Sorbent for
       Fluoride Removal from Aqueous Solutions. "^4/>/>/. Surface Sci., 256(23): 6956-6962.
Sujana, M.G., R.S. Thakur and S.B. Rao. 1998. "Removal of Fluoride from Aqueous Solution by Using
       Alum Sludge." J. Colloid Interface Sci., 206:94-101.

Swain, S.K., RK. Dey, M. Islam, RK. Patel, U. Jha, T. Patnaik, and C. Airoldi. 2009. "Removal of
       Fluoride from Aqueous Solution Using Aluminum -Impregnated Chitosan Biopolymer." Sep. Sci.
       andTechnol, 44: 2096-2116.

Tahaikt, K., I. Achary, M.A. Menkouchi, Z. Amor, M Taky, A. Alami, A. Boughriba, M. Hafsi and A.
       Elmidaoui. 2006. "Defluoridation of Moroccan Groundwater by Electrodialysis: Continuous
       Operation." Desalination, 189(l-3):2\5-22Q.

Tahaikt, K., R.E. Habbani, A.A. Haddou, I. Achary, Z. Amor, M Taky, A. Alami, A. Boughriba, M. Hafsi
       and A. Elmidaoui. 2007. "Fluoride Removal from Groundwater by Nanofiltration." Desalination,
       212: 46-53.

Tahaikt, K., A.A. Haddou, R.E. Habbani, Z. Amor, F. Elhannouni, M Taky, M. Kharif, A. Boughriba, M.
       Hafsi and A.  Elmidaoui. 2008. "Comparison of the  Performances of Three Commercial
       Membranes in Fluoride Removal by Nanofiltration. Continuous Operation." Desalination, 225:
       209-219.

Tchomgui-Kamga, E., E. Ngameni and A Darchen. 2010. "Evaluation of Removal Efficiency of Fluoride
       from Aqueous Solution Using New Charcoals that Contain Calcium Compounds." J. Colloid and
       Interface Sci., 346: 494-499.

Teng, S-X, S.-G. Wang, W.-X. Gong, X.-W. Liu and B.-Y.  Gao. 2009. "Removal of Fluoride by Hydrous
       Manganese Oxide-Coated Alumina: Performance and Mechanism." J. Haz. Mater. 168: 1004-
       1011.

Thompson, M.A. and M.P. Robinson. 1991. Suffolk Introduces EDRto Virginia. AWWA Membrane
       Processes Conference, Orlando, Fl.

Tripathy, S.S. and A.M. Raichur. 2008. "Abatement of Fluoride from Water Using Manganese Dioxide-
       Coated Activated Alumina." J. Haz. Mater.,  153: 1043-105 1 .

Turner, B.D., P. Binning, and S.L.S Stipp. 2005. "Fluoride Removal by Calcite: Evidence for Fluorite
       Precipitation and Surface Adsorption." Environ. Sci. Technol., 39: 9561-9568.
                                             72

-------
               Appendix A




Summary of Subsystems Including Components

-------

-------
The items that are designated as "optional" are not mandatory requirements. Some of those items may
already be included in systems other than treatment and therefore, would be redundant. Other items,
though desirable, are not mandatory. Automatic and semiautomatic operation is optional. Therefore, for
each instrument and control item, though not indicated for clarity, there is an automatic option.

1.          Raw Water Influent Main
           a.  Flow Control
           b.  Flowrate measurement, flow total
           c.  Acid injection for pH adjustment
           d.  In-line static mixer
           e.  pH measurement, indicator, alarm, and fail-safe control
           f.   Pressure indicator
           g.  Pressure control (optional)
           h.  Backflow preventer
           i.   Sample before pH adjustment piped to sample panel (optional)
           j.   Sample after pH adjustment piped to sample panel (optional)
           k.  Isolation valve
           1.   Temperature indicator  (optional)

2.          Intervessel Pipe Manifold
           a.  Process control valves
           b.  Pressure indicators
           c.  Sample piped to sample panel (optional)

3.          Treated Water Effluent Main
           a.  Caustic injection for pH adjustment
           b.  In-line static mixer
           c.  pH measurement, indicator, alarm and fail-safe control
           d.  Sample after pH adjustment piped to sample panel (optional)
           e.  Pressure indicator
           f.   Aeration subsystem(optional)
               i.   Air blower (optional)
               ii.  Clear well (optional)
           g.  Booster or repressurization pump (optional)
           h.  Disinfection injection (optional)
           i.   Isolation valve

4.          Raw Water Bypass Main
           a.  Flow control
           b.  Flowrate measurement, flow total
           c.  Backflow preventer
           d.  Isolation valve

5.          Backwash/Regeneration Feed Main
           a.  Flow control
           b.  Flowrate measurement, flow total
           c.  Caustic injection for pH adjustment
           d.  Acid inj ection for pH adj ustment
           e.  In-line static mixer
           f.   pH measurement
           g.  Sample after pH adjustment piped to sample panel (optional)
                                              A-l

-------
           h.   Backflow preventer
           i.   Isolation valve

6.          Waste water Main
           a.   Backflow preventer
           b.   Process isolation valves
           c.   Acid injection for pH adjustment (optional)
           d.   Chemical injection (optional)
           e.   In-line static mixer (optional)
           f.   Sample after chemical injection piped to sample panel (optional)

7.          Treatment Unit
           a.   Pressure vessel
           b.   Treatment media
           c.   Internal distribution and collection piping
           d.   Pressure relief valve
           e.   Air/vacuum valve
           f.   Operating platform and/or ladder (optional)

8.          Sample Panel (optional)
           a.   Manifolds
               i.  Influent manifold (influent main sample and raw water samples from each treatment
                  vessel after pH adjustment)
               ii.  Effluent manifold (effluent main sample after pH adjustment, treated water samples
                  from each treatment vessel and wastewater manifold sample after pH adjustment and
                  chemical injection)
               iii. pH indicator (influent sample manifold and effluent sample manifold)
               iv. Sample collection spigots with drain
           b.   Wet chemistry laboratory bench with equipment, glassware, reagents, etc.

9.          Acid Storage and Feed Subsystem
           a.   Emergency shower and eyewash, signage
           b.   Acid storage tank (outside treatment building)
               i.  Fill, discharge, drain, vent, and overflow piping
               ii.  Liquid level sensor (optional)
               iii. Desiccant air dryer in vent (optional)
               iv. Weather protection
               v.  Containment basin
           c.   Acid day tank (inside treatment building)
               i.  Fill pipe float valve
               ii.  Drain valve
               iii. Containment basin
           d.   Acid pumps
               i.  Treatment unit pH adj ustment
               ii.  Neutralization pH adjustment
               iii. Wastewater pH adjustment (optional)
           e.   Acid piping (interconnecting piping)
               i.  Between storage tank and day tank
               ii.  Between feed pumps and raw water injection point
               iii. Between feed pumps and regeneration feed and wastewater mains injection points
                  (optional)

                                              A-2

-------
               iv. Backflow prevention

10.         Caustic Storage and Feed Subsystem
           a.   Emergency shower and eye wash, signage
           b.   Caustic storage tank (outside treatment building)
               i.  Fill, discharge, drain, vent, and overflow piping
               ii.  Liquid level sensor (optional)
               iii. Immersion heater with temperature control
               iv. Weather protection
               v.  Containment basin (optional)
           c.   Caustic day tank (inside treatment building)
               i.  Fill line float valve
               ii.  Drain valve
               iii. Containment basin (optional)
           d.   Caustic piping (interconnecting piping)
               i.  Between storage tank and day tank
               ii.  Between feed pump and, regeneration feed main injection point (optional)
               iii. Between feed pump and treated effluent main injection point (optional)
               iv. Backflow prevention

11.         Backwash Water Disposal System (optional)
           a.   Surge tank (optional)
           b.   Lined evaporation pond (optional)
           c.   Unlined evaporation pond (optional)
           d.   Sewer (optional)
           e.   Drainage ditch (optional)
           f.   Other discharge  method (optional)
12.         Toxic Regeneration Wastewater Disposal System
           a.   Surge tank (optional)
           b.   Lined evaporation pond (optional)
           c.   Wastewater reclamation system (optional)
           d.   Other discharge method (optional)
                                              A-3

-------
          Appendix B




Treatment System Design Example

-------
This design example is applicable to a specific manually operated fluoride removal AA treatment system
with two parallel vessels.

    Given:
    q (flowrate) = 600 gpm
    N (number of treatment vessels) = 2
    Raw water fluoride concentration = 5.0 mg/L
    Treated water fluoride concentration =1.0 mg/L (max)
    AA fluoride removal capacity = 4,580 g/m3 (2,000 grains/ft3)
    (Note: Indicated capacity applies only to system with a raw water fluoride concentration of 5.0 mg/L
      and a process water pH value of 5.5)
    Md (media packing density) = 48 lb/ft3
    EBCT = 7.5 min (1 ft3/gpm)
    Pipe material-Type I Schedule  80 PVC,
    v (pipe velocity) = 5 ft/second (max.)
    p (system pressure): 50 psig (max.)
    T (ambient temperature): 95°F (max.)
    Tw (water temperature): 85°F (max.)

1.   Vessel and Treatment Bed Design (reference: Figure 3-4)

    Solve for:  h (treatment bed depth)
              d (treatment bed diameter)
              A (treatment bed horizontal surface area)
              V (treatment bed volume)
              Mw (total weight of treatment media)
              D (vessel outside diameter)
              H (vessel overall height)

    q/N = 600 gpm/2 treatment beds = 300 gpm/bed
    Then, using 1 ft3 of media per gpm of flowrate, we require 300 ft3 of A media per treatment bed
    (equivalent to an EBCT of 7.5 min), or

                                      V=300ft3 = Ah = 7id2h/4

    Then, when h = 5 ft,

                                         A = 60 ft2 = 7id2/4

    Then, d = 8.74 ft = 8 ft 8.8 in

    Then, D = d+ I" = 8ft9.8in, therefore use D = 9 ft 0 in (note: D must employ the next even multiple
    of 6 in)

    Then, d = D-lin = 8ftllin
    Then, A = 7id2/4 = 7i(8.92)2/4 =  62.5 ft2
    Then, V = 7id2h/4 = 7i(8.92)2(5)/4 = 312 ft3
    Then, Mw = 2 vessels x 312 ft3  x 48 lb/ft3 = 29,952 Ib

    Because the media quantity is almost 30,000 Ib, it is prudent to procure media in supersacks.
                                             B-l

-------
    Then the vessel height (H) (Figure 3-4) is:

                                 H = h + h/2 + 6 in + (2)D/4 + 1 in =

                           60 in + 30 in + 6 in + 2 (108 in/4)+ 1 in = 151 in =
                                              12 ft 7 in

2.   Pipe Sizing

    Solve for:

    a.      Sizes of raw and treated water pipe mains
    b.      Sizes of treatment unit branch piping

           a.   Mains:
        Q = 600 gpm (max.)
        Try 6 in, v = 6.5 in/sec > 5 in/sec, therefore NG
        Try 8 in, v = 3.6 in/sec < 5 in/sec, therefore OK
        Use 8 in Schedule 80 PVC

           b.   Branches:
        Qe  = q/w = 3 00 gpm
        Try 4 in, v = 7.7 in/sec > 5 in/sec, therefore NG
        Try 6 in, v = 3.4 in/sec < 5 in/sec, therefore OK
        Use 6 in Schedule 80 PVC

    Backwash rate is not to exceed rate required for 50% treatment bed expansion.

    Then, backwash rate = A x 7 gpm/ft2 = 62.5 ft2 x 7 gpm/ft2 = 437.5 gpm < 600 gpm, therefore OK.
    (Note: The backwash rate is  sensitive to water temperature.)

3.   Acid Subsystem Design

    a.   Storage Tank Size

        Storage tank size is based upon logistical requirements, which are a function of treatment plant
        acid consumption rate and bulk tank truck deliveries of acid. The tank truck can deliver up to
        48,000 Ib of 66°B' H2SO4. The density of this liquid is 15.5 Ib/gal. Therefore, a delivery contains
        3,100 gal.

        In this example the peak treatment flow is 600 gpm, and it is assumed that the acid consumption
        as determined by titration is 0.05 gal/1,000 gal of treated water.  Then the acid  consumption is 1.8
        gal/hr and a tank truckload would  supply a minimum of  1,720 hr of treatment operation. Acid
        consumption for raw water pH reduction, which is a function of total alkalinity and free CO2, is
        discussed in Appendix D.

        A 5,000-gal acid storage tank provides capacity for more than \1A bulk tank truckloads of 66°B'
        H2SO4.  Therefore, when half a truckload has been consumed (providing capacity for the next
        truckload delivery), there is a minimum of a 860-hr acid supply available in storage before the
        acid supply is  exhausted.
                                              B-2

-------
    b.  Day Tank Size

       The storage tank supplies a polypropylene day tank located inside of the treatment building. A
       100-gal day tank will satisfy more than 200% of the maximum treatment process pH adjustment
       acid requirements (i.e., 43.2 gal/day) for a maximum daily treatment flow of 864,000 gal.

    c.  Acid Pump Size

       The acid feedrate required for the treatment process pH adjustment function is: 600 gpm x
       60 min/hr x 0.05 gal acid/1,000 gal water =1.8 gph, which can be satisfied by a positive
       displacement diaphragm pump that has a maximum flowrate of 2.5 gph @ 50 psig with a 1,000:1
       turndown capability (materials of construction to be recommended for 66°B' H2SO4 service).

       For neutralization of the treatment bed after completion of regeneration and the regeneration
       wastewater flowing from the treatment vessel to the regeneration wastewater surge tank, two
       additional acid feed pumps are required . The rule of thumb relating to the volume of acid
       required to be applied to accomplish both functions is 1 gal/ft3 of AA, or 312 gal/regeneration.
       The acid feed for these two functions will take place over a period of 4 to 6 hr. The first pump
       feeds acid into the regeneration feedwater main to adjust the pH initially to 2.5, then to 4.0, and
       finally at completion of the neutralization to 5.5. The second pump feeds acid into the wastewater
       main at a rate required to adjust the pH of the entire wastewater batch to a range of 6.0 to 6.5.
       This latter acid feed requirement can take place at a constant rate that will provide the necessary
       wastewater pH for the volume of the entire wastewater batch (thoroughly mixed in the
       wastewater surge tank) at the conclusion of the regeneration process. The two acid feed pumps
       required for the two functions can be identical air-operated diaphragm pumps with maximum
       flowrate of 2  gpm at 50 psig with a 100:1 turndown capacity (materials of construction to be
       recommended for 66°B'  H2SO4 service).

       A 5-hp air compressor with a 60-gal receiver capable of supplying 14.7 cfm at 175 psig com-
       pressed air. The air compressor will supply compressed air for both air-operated diaphragm acid
       feed pumps, the air-operated diaphragm caustic soda feed pump, and (for automatic operation)
       the pneumatic-operated process control butterfly valves. If there is a wastewater sludge dewater-
       ing system, the air compressor will be available to operate the air-operated diaphragm pump (for
       sludge transfer) and the plate and frame filter press.

4.   Caustic Subsystem Design

    a.  Storage Tank Size

       Storage tank  size is based upon logistical requirements, which are a function of treatment plant
       caustic consumption rate and bulk tank truck deliveries of caustic. A tank truck can deliver up to
       48,000 Ib of 50% NaOH. The density of this liquid is 12.9 Ib/gal. Therefore, a delivery contains
       3,700 gal.

       In this example the peak treatment flow is 600 gpm, and it is assumed that the caustic con-
       sumption as determined by titration is 0.135 gal/ 1,000 gal treated water (Rubel F Manual uses
       0.02). Then the caustic consumption is 4.86 gal/hr. Then, a tank truckload would supply a
       minimum of 760 hr of treatment operation.

       A 5,000-gal caustic storage tank provides capacity for more than 1% bulk tank truckloads of 50%
       NaOH. Therefore, when 75% of a truckload has been consumed (providing capacity for the next
                                              B-3

-------
    truckload delivery), 925 gal remains, which provides a 190-hr (8-day) caustic supply available in
    storage before the caustic supply is exhausted. Note: When the supply remaining in the storage
    tank provides capacity for a bulk tank truck delivery, spent media regeneration (if applicable) will
    be deferred until after caustic delivery.

b.   Day Tank Size

    The storage tank supplies a polypropylene day tank located inside of the treatment building. A
    500-gal day tank will satisfy more than 200% of the maximum treatment process pH adjustment
    caustic requirements (117 gal/day) for maximum treatment flow of 864,000 gal for one day as
    well as the requirement (122 gal, see caustic pump size below) for one two-step regeneration of
    spent media.

c.   Caustic Pump Size

    The caustic feedrate required for the treatment process pH adjustment function is: 600 gpm x  60
    min/hr x 0.135 gal caustic/1,000 gal water = 4.86 gph.

    The caustic feedrate required for the treatment process pH adjustment function (4.86 gph) is satis-
    fied by a positive displacement diaphragm pump that has a maximum flowrate of 5 gph @ 50
    psig with a 1,000:1 turndown capability (materials of construction to be recommended for 50%
    NaOH service).

    For frequency of regeneration, the following is calculated:

    Given:
    Raw water fluoride concentration = 5.0 mg/L
    Treated water fluoride concentration =1.0 mg/L
    AA media capacity = 2,000 grains/ft3
    Bed volume = 312 ft3
    Density of 50% NaOH = 12.9 Ib/gal

    Find: Regeneration frequency

    Amount of fluoride removed = 5.0 - 1.0 = 4.0 mg/L = 4.0 x 0.058 = 0.23 grains/gal

    Amount of water treated/treatment run/per vessel = (2,000 grains/ft3 x 312 ft3)/(0.23 grains/gal) =
    2,700,000 gal

    Therefore, during maximum treatment flow continuous operation, minimum regeneration
    frequency would be six days per bed. Using the two bed system in this example, the maximum
    regeneration  frequency could be as often as once every three days.

    For regeneration of the AA treatment bed two regeneration steps are required utilizing 15 gal of
    1% NaOFi/ft3 of media per step. The following calculations provide the volume and flowrate of
    50% NaOH required per regeneration.

    Given:
    di = density of 1% NaOH = 8.4 Ib/gal
    d2 = density of 50% NaOH = 12.9 Ib/gal
    vi = volume of 1% NaOH/regeneration step-ft3 = 15 gal/step-ft3

                                          B-4

-------
       n = number of steps = 2 (upflow and downflow)
       V = 312ft3

       Find:
       wi = weight of 1% NaOH/step
       V2 = volume 50% NaOH required/step

       Then, wi = vi(V)(di) = 15 gal/step-ft3 x 312 ft3 x 8.4 Ib/gal = 39,312 Ib/step

       Then, weight of 50% NaOH/step = 39,312 lb/50 = 786 Ib/step

       Then, v2 = 786 Ib/d2 = 786 lb/(12.9 Ib/gal) = 61 gal

       Then: If, step duration is 50 min,

       50% NaoH flowrate = V2/50 min = 61 gal/50 min =1.2 gpm

       Then, total 50% NaOH required per regeneration = V2 x n = 61 gal/step x 2 steps =122 gal.

       The caustic feed pump required for this function will be a 2 gpm metering pump with materials of
       construction suitable for 50% NaOH service. Each regeneration step (upflow and then downflow)
       requires 61 gal of 50% NaOH to be fed into the mixing tee  where it is diluted to 1% NaOH. Each
       regeneration step is designed to last for 50 min.

5.   Regeneration Wastewater Surge Tank Design

    Given:
    Maximum volume of regeneration wastewater per cubic foot media = 400 gal/ft3
    Number of cubic feet of media per regeneration = 312 ft3
    Tank construction - epoxy interior lined carbon steel

    Find:
    Volume of wastewater per regeneration = 400 gal/ft3  x 312 ft3 = 124,800 gal = 16,700 ft3

    Dimensions of surge tank (use height = 20 ft)

    Then, (diameter)2 = 4 x 16,700 ft3/7i x 20 ft = 1,063 ft2
    Then, diameter = 33 ft

    Suggested Containment Basin Dimensions: length 72 ft, width  64 ft, height 4 ft; volume = 18,400 ft3
    = 137,600 gal > 124,800 gal.

6.   Lined Regeneration Wastewater Disposal Evapora-tion Pond Design

    a.      It is assumed that the average plant utilization rate is 40% and that the annual average net
       evaporation rate is 6 ft 0 in.

    b.      Find: Evaporation pond size
                                             B-5

-------
Total annual volume of water treated = 40% x 600 gpm x 1440 min/day x 365 day/yr = 126 x 106
gal/yr

Amount of water treated/treatment run/vessel = 2. 7 x 106 gal (see Section 4.c in Appendix B)

Then, number of regeneration cycles per year = 24.

Amount of wastewater per cubic foot of treatment media = 300 gal/ft3, then each regeneration cycle
yields 300 gal/ft3 x 312 ft3 x 2 = 187,200 gal of wastewater.
Then, total amount of wastewater produced per year = 187,200 gal x 24 = 4.5 x 1Q6 gal/yr = 601,000
fWyr

Using an average annual evaporation of 6 ft 0 in and deducting 1 ft 0 in for deviation from average, 5
ft 0 in is the net minimum evaporation rate per year.

Therefore, the required pond area can be calculated by dividing the total annual wastewater produced
by the net minimum evaporation rate.

Pond area =601,000 ft3/5 ft = 120,200 ft2 with 8 ft minimum pond depth.
                                          B-6

-------
                Appendix C

Discussion of Acid Consumption Requirements for
         pH Adjustment of Raw Water

-------
The practical method described in the text that is used to determine the acid feed requirement for lowering
the raw water pH to 5.5 is acid titration. However, this can also be accomplished theoretically when a raw
water analysis is available and raw water samples are not. This method requires pH, total alkalinity, and/or
free carbon dioxide from the raw water analysis in addition to the graph illustrated in Figure C-l.  If only
two of the three raw water analysis items are available, the third is determined by the graph. The pH curves
illustrated in Figure C-l were developed from theoretical chemical formulae, which integrate the
relationship between pH, alkalinity and free carbon dioxide. Trial and error usage of these curves can lead
the designer to the acid feed requirement for the desired pH adjustment. The designer should be aware of
the fact that the reduction in alkalinity coincides with the corresponding increase in free carbon dioxide.
The following examples best illustrate this method:
      1000
                                    10                       100

                               TOTAL  ALKALINITY  AS CaCO-j-p.p.m.
1000
         FIGURE C-1  GRAPH OF pH AS A FUNCTION OF TOTAL  ALKALINITY AND
                        FREE CARBON DIOXIDE
                                              C-l

-------
Example 1:
Given:     Raw Water pH = 8.0
           Raw Water M = 220 ppm (as CaCO3)
           Raw Water CO2 = 4 ppm

Find:       a) M and free CO2 for pH adjusted to 5.5
           b) 66 degree Sulfuric acid required feed rate to adjust pH to 5.5

a) Try reducing M by 200 ppm (as CaCOs) to 20 ppm (as CaCOs)
Then, increase in free CO2 = 200 xO.88 = 176 ppm
    Then, total free CO2 = 176 + 4 = 180 ppm
    Then using the graph we find the adjusted pH to be 5.4 when:
       1) M = 20 ppm (as CaCO3)
       2) Free CO2= 180 ppm
Therefore, too much alkalinity was removed, try reducing M by 196 ppm (as CaCOs) to 24 ppm (as
CaCO3).

Then, increase in free CO2 = 196 x 0.88 = 172.5 ppm
    Then, total free CO2 = 172.5 + 4 = 176.5 ppm
    Then using the graph we find the adjusted pH to be 5.5 when:
       1) M = 24 ppm (as CaCO3)
       2) Free CO2= 176.5 ppm

b) For each  100 ppm (as CaCOs) reduction of total alkalinity, 105 ppm 66 degree sulfuric acid must be
added. Therefore, reduce M by 196 ppm (as CaCOs) by feeding 1.96 x 105 ppm = 205.8 ppm 66 degree
sulfuric acid to adjust raw water pH to 5.5. If we desire to find what acid feed rate would be required per
1,000 gallons of treated water, we find that:

Feed rate = (205.8 x 10'6 ppm) x (1,000 gal x 8.34 lb/gal)/15.5 Ib/gal) = 0.11 gal H2SC>4/1,000 gal water

Example 2:

Given:     Raw Water M - 100 ppm (as CaCO3)
           Free CO2 = 6 ppm

Find:       a) Raw Water pH
           b) M and free CO2 for pH adjusted to 5.5
           c) 66 degree Sulfuric acid required feed rate to adjust pHto 5.5

a)  From graph we find raw water pH to be 7.5
b)  Try reducing M by 80 ppm (as CaCOs) to 20 ppm (as CaCOs)
    Then, increase in free CO2 = 80 xQ.88 = 70.4 ppm
    Then, total free CO2 = 70.4 + 6 = 76.4 ppm
    Then using the graph we find the adjusted pH to be 5.75 when:
       1) M = 20 ppm (as CaCO3)
       2) Free CO2 = 76.4 ppm
Therefore, too little alkalinity was removed, try reducing M by 87 ppm (as CaCOs) to 13 ppm (as CaCOs).
Then, using the graph we find the adjusted pH to be 5.55 when:

-------
        l)M=13ppm(asCaCO3)
        2) Free CO2 = 82.5 ppm
Therefore, too little alkalinity was removed, try reducing M by 88 ppm (as CaCOs) to 12 ppm (as CaCOs).

Then, increase in free CC>2 = 88 x 0.88 = 77.5 ppm
     Then, total free CO2 = 77.5 + 6 = 83.5 ppm
     Then using the graph we find the adjusted pH to be 5.5 when:
        1) M = 12 ppm (as CaCO3)
        2)FreeCO2 = 83.5ppm

c)   Therefore, reduce M by 88 ppm (as CaCOs) by feeding 0.88 x 105 = 92.4 ppm 66 degree sulfuric acid
to adjust raw water pH to 5.5.

     Acid feed rate = (92.4 x 1Q-6 ppm) x (1,000 gal x 8.34 lb/gal)/(15.5 Ib/gal)  = 0.05 gal H2SC>4/1,000 gal
     water
                                             C-l

-------
              Appendix D




Activated Alumina Plant Search and Visits

-------

-------
To facilitate the revision of the 1984 fluoride design manual and better understand the evolution of AA
system design and operations,  a list of active AA plants in the U.S. was  identified and their design and
operation information was collected. The search for the plants was conducted via a brief internet and
literature search, followed by email and telephone communications with individuals such as state drinking
water officials, technology vendors, and AA media manufacturers and suppliers. The time period covered
for the search spanned from early  1960s when some of the earliest AA plants were first installed through
2013 when a 900-gpm newly designed and installed AA plant awaits system startup. The search focused
on states such as Arizona, California, Idaho, New Mexico, Oklahoma,  and Texas where  significantly
elevated levels of natural fluoride in drinking water supplies have been reported (CDC, 1992).

System information collected  included, but not  limited to, treatment  type, technology/media vendors,
system size, system footprint, system installation time, source water quality (i.e., fluoride level and pH at
a   minimum),   process/design   parameters,   system   operation    and   maintenance   (O&M),
performance/compliance monitoring, media regeneration, residual handling, and costs.  The pieces  of
information were collected mostly through emails and phone calls, with some obtained via site visits (to
three plants). Table D-l  summarizes the plants identified and design and operation information collected.
Obviously, the plants shown in Table 3-1  does not represent an exhaustive list of AA plants currently
operating in the U.S. Rather, they represent only those having relevant information available during the
information collection period of this manual revision project. The plants are identified as Plants A through
F.

While a number of variations exist in system design and operations (see  Table C-l), the basic parameters
remain essentially unchanged among the plants. For example, depending on the system flowrate, a plant
comprises a minimum of two parallel vessels or multiple parallel lead-lag modules to accommodate the
maximum daily demand. Each vessel is sized to contain a volume of AA that provides a minimum of 5
min EBCT. The depths of AA beds range between 3 to 6 ft and the straight side of a vessel will allow at
least 50% AA bed expansion plus 6 in during media backwash. The materials of construction is suitable
for exposure to acid and  caustic used for pH adjustments during service and media regeneration. The
media regeneration generally includes a backwash, a 1% caustic wash, and a sulfuric acid neutralization
step. The 1984 design manual contains information consistent with all of these design and operating bases
and therefore is considered still valid for use as guidelines for AA fluoride  removal plant design and
operations.
                                              D-l

-------
                                           Table D-l. Activated Alumina Plant Design and Operational Parameters

General
Adsorption
Vessels/Media
Service
pH Adjustment
Regeneration
Parameter
Location
System Designed by
System Constructed by
Time System On-line
Number of Vessels
Vessel Configuration
Vessel Diameter
Vessel Height
Vessel Straight Height
Vessel Construction
Vessel Pressure Rating
Media Type
Media Volume
Media Weight
Media Bed Depth
Free Board Depth
System Design Daily Flowrate
System Design/Max. Pumping Rate
System Operating Flowrate
System Bypass Flowrate
Vessel Hydraulic Loading Rate
Vessel EBCT
Raw Water pH
Target Treatment pH
Effluent pH
Acid Used
Regeneration Frequency
Chemical Used
Flow Pattern
Acid Neutralization
Disposition of Waste
Unit






Ft
Ft
Ft

Psi

ft3/vessel
Ib/vessel
Ft
Ft
MGD
Gpm
Gpm
Gpm
gpm/ft2
Min
S.U
S.U
S.U.






Plant A
29 Palms, CA
Fred Rubel Engineering
Pro Contracting/District
Mar 2003
6
3 Parallel Modules
11
12
8
Epoxy-coated carbon steel
900
Alcoa CPN
475
NA
5
4.5
3
2,100
1,650
550
5.8
6.5
8
5.0
7.5-8.4
Sulfuric
1 vessel/day
!%NaOH
Upflow/Downflow
Yes
Evaporation Pond
Plant B
Lordsburg, NM
Severn Trent
( SORB 09™ Process)
Severn Trent
2009
0
Parallel
12
NA
6
Carbon steel with NSF internal
painting
NA
Alcan AA400G
420
20,160
3.7
NA
1.3
900
720
19%
3.2
8.7
8.5
6
7.8
Sulfuric
-10 day (-3.2 MG throughput)
!%NaOH
Downfiow
Yes
F ppt post treatment in place but
not used
Plant C
Desert Center, CA
NA
NA
1968 with one vessel
replaced in 1995 and the
other in 2001 or 2002
2
Parallel
14
16
16
Original - carbon steel with
epoxy coating, replacement
- one carbon steel with
epoxy coating and one
stainless steel
Atmospheric (by gravity)
Alcan AA400G
770
36,960
5
NA
NA
NA
500
NA
3.2
11.5
8.3
5.6
7.3-7.5
Sulfuric
When F level is over 2
mg/L
!%NaOH
Downfiow
Yes
Evaporation Pond
Plant D
8 miles West of Blythe,
CA
Purefiow
Purefiow
2013
(Still to be Started up)
2
Series, but system runs
with one tank at a time
11
12
10
Carbon steel lined with
epoxy
75
NA
475
22,800
5
NA
1.3
900
900, but will run only at
500 gpm
NA
5.3
7.1
7.4
5.5
7.5
Sulfuric
TBD
!%NaOH
Downfiow
Yes
Evaporation Pond
Plant E
Blythe, CA
NA
Third Tank by USFilter
NA
3
Parallel
12
12
NA
Steel
NA
Alcan AA400G
530
25,500
4.7
NA
NA
NA
2,100
NA
6.2
5.7
8.2
5.5
6.0-8.5
Sulfuric
Based on service
throughput
!%NaOH
Upflow/Downflow
Yes
To wastewater Treatment
Plant
Plant F
Ajo, AZ
Malcomb Pirnie (Arcadia)
A Construction Firm in
Phoenix
Apr 20 11
2
Parallel
8
NA
10
Carbon Steel with Epoxy
Coating
120
Alcan AA400G 14X28
300
NA
6
3.5
0.45
300
300
30
6
NA
7.4-7.8
8.3
NA
H2SO4
Once per week per vessel
!%NaOH
Downfiow
Yes
Disposed to Sewer
o
to
           NA - information not available

-------
           Appendix E




English to Metric Conversion Table

-------

-------
English
inch
inch2
inch3
feet (ft)
ft2
ft3
gallon (gal)
gal
gal
grains (gr)
gr/ft3
pounds (Ib)
lb/inches2 (psi)
lb/ft2 (psf)
c/1,000 (gal)
Multiply by
0.0254
0.000645
0.000016
0.3048
0.0929
0.0283
0.2642
0.0038
0.0038
0.0649
2.2919
0.4545
0.00689
4.8922
0.2642
Metric
meter (m)
m2
m3
m
m2
m3
liter (L)
m3
kiloliter (kL)
gram (g)
g/m3
kilogram (kg)
megapascals (MP)
kg/m2
c/1,000 L
D-l

-------