United States
Environmental Protection
Agency
Office of Water
4607
EPA815-D-99-004
May 1999
oEPA TECHNOLOGIES AND COSTS FOR
THE REMOVAL OF RADON
FROM DRINKING WATER
PUBLIC COMMENT DRAFT
-------
DISCLAIMER
The mention of trade names, companies, organizations, or products does not constitute and
should not be interpreted as an endorsement, approval or recommendation for use or application.
May 1999
-------
This Technology and Costs for the Removal of Radon From Drinking Water document was prepared
by Science Applications International Corporation (SAIC), under the auspices of Mr. William
Labiosa (Work Assignment Manager) of the U.S. Environmental Protection Agency's (USEPA)
Office of Ground Water and Drinking Water (OGWDW). SAIC wishes to gratefully acknowledge
the support of the EPA WAM. SAIC Contract No. 68-C6-OQ59, Work Assignment No. 1-22, SAIC
ProjectNo. 01-0833-08-3566-000. SAICcontractandworkassignmentmanagersare: James Parker,
Contract Project Manager; and Faysal Bekdash, Work 'Assignment Manager. Individuals who
prepared this T&C document are: Faysal Bekdash, Ph.D., Senior Environmental Engineer; Pravin
Rana, Environmental Engineer, Tracy Scriba Environmental Scientist, Pat Ransom, Environmental
Engineer, Mary Waldron, Environmental Scientist, Abhjit Palit, Jr. Environmental Engineer, and
Djamel Benelmouffok Ph.D. QA reviewer.
May 1999
-------
-------
PRELUDE
The purpose of this document is to assist the U.S. Environmental Protection Agency (EPA) in
updating the radon-relevantportions of its "Technologies amd Costs" (T&C) documents1-2 developed
to support the 1991 proposed radionuclides rule. Therefore, much of the background information,
description of technologies, and valid design parameters sections were taken from the above
referenced T&C documents. Subsequent to this Prelude, this document is organized as follows:
• List of Abbreviations
• List of Definitions and Conversion Units
• Table of Contents
• List of Figures
• List of Tables
• Chapter 1. Introduction: Presents general information on the chemical and physical
properties of radon and the regulatory background, and defines the objectives of this
document.
• Chapter 2. Removal of Radon from Drinking Water: Discusses technologies and
techniques for removing radon from drinking water sources. Technologies presented
include various aeration technologies, granular activated carbon (GAG) treatment, and
other low technology radon removal techniques. Also, this chapter discusses pre- and
post treatment requirements for selected radon removal technologies.
• Chapters. Best Management Practices: Presents an overview of best management
practices (BMPs) for removing moderate amounts of radon from drinking water sources.
• Chapter 4. Radon Removal Costs: Provides information on available cost models used
for costing various radon removal technologies and presents costs for various system size
categories and types.
• Appendix A. Provides supporting information for cost analysis and possible off-gas
emissions regulations.
1 U.S. EPA. (1987). Technologies and Costs for the Removal of Radon from Potable Water Supplies (Fourth
Draft). Prepared by Malcolm-Pirnie (January 8, 1987).
2 U.S. EPA. (1992) Technologies and Costs for the Removal of Radionuclides from Potable Water Supplies.
Prepared by Malcolm-Pirnie (July, 1992).
May 1999
-------
ABBREVIATIONS LIST
A: W air: water ratio (quantity of air provided compared to the quantity of water)
atm atmosphere
atm-m3/M atmosphere-meter cubed per mole
avg average
°C degrees Celsius
cc cubic centimeters
cm centimeter
C02 carbon dioxide
CT concentration time
DBA diffused bubble aeration
EPA U.S. Environmental Protection Agency
°F degrees Fahrenheit
Fe iron
ft. feet
ftVmin. cubic feet per minute
gal. gallon
gm/M grams per mole
gpd gallons per day
gpm gallons per minute
gprn/sq ft. gallons per minute per square foot
hp horsepower
hr hour
in. inch
K Kelvin
K$ $1,000
L/hr. liters per hour
Mn manganese
MCL maximum contaminant level
ma/s square meters per second
m3K/M cubic meter Kelvin per mole
mg/L milligrams per liter
MGD millions of gallons per day
mW/cm2 milliwatts per square centimeter
mWs/cm2 milliwatts-seconds per square centimeter
MSBA multi-stage bubble aeration
NOM natural organic matter
PCE Perchloroethylene
PTA packed tower aeration
ppm parts per million
pCi/L picocuries per liter
pE electron activity; -log of the electrons exchanged in a redox reaction; -log[e]
pH negative logarithm of the hydronium ion concentration in mol/L; -log [H3Cr]
n
May 1999
-------
POE
POU
psi
SDWA
s
sf
scfin
VOC
point of entry
point of use
pounds per square inch
Safe Drinking Water Act
second
square foot
standard cubic feet per minute
micrograms per liter
volatile organic compound
111
May 1999
-------
DEFINITIONS AND CONVERSION UNITS
Aerosol is a suspension of solid or liquid in air.
Alpha particles are positively charged particles, consisting of two protons and two neutrons.
Emitted by decaying radon."
Becquerel (Bq) is a unit of activity in the international system of units (SI) that is equal to one
disintegration per second. 1 Bq is equivalent to about 27 pCi.
Concentration Time (CT) is the product of the residual disinfection concentration in mg/L © and
the disinfectant contact time in minutes (T). Disinfectant contact time is the time needed for water
being treated to flow from the point of disinfectant application to a point before or at the first
customer during peak hourly flow (U.S. EPA, 1997).
Curie is a unit of activity that equals 3.7><1010 disintegrations/second. 6.48 mg of radon has an
activity of one curie.
Dose equivalent is the product of the absorbed dose from ionizing radiation and such factors which
account for differences in biological effectiveness due to the type of radiation and its distribution in
the body as specified by the International Commission on Radiological Units and Measurements
(ICRU).
Henry's Law is p = kc, where p is the partial pressure (in atm) of the gaseous solute above the
solution, c is the concentration (in mol/L) of the dissolved gas, and k is a constant (in L-atm/mol)
that is characteristic of the particular solution.
Henry's Constant is an indicator of the transfer efficiency of a gas from a liquid solution. It is
expressed in atmosphere (atm) or in m3atm/mole. To convert Henry's Constant from atm to
m3atm/mole, the following equation applies Hatm=Hm3atm/m0|e x P/RT, where P is pressure in
atmosphere, T is temperature in Kelvin (K), and R is the universal gas constant 8.205 x IQ-5
m3atm/mole.
1 Picocurie = 10"1- curies, which is approximately two disintegrations per minute.
Radiation Absorbed Dose (Rad) is a unit representing deposition of energy in matter. One rad
equals the deposition of 100 ergs per gram of irradiated material. 100,000 rads are equal to one watt.
Radioactivity is the nuclear transformation of a radioactive substance which occurs in a specific
time interval.
Roentgen Equivalent in Man (Rem) is the dose equivalent from ionizing radiation to the total body
or any internal organ or organ system that will produce the same biological effect as one rad of high-
IV
May 1999
-------
penetrating x-rays. It is equal to the absorbed dose in rads multiplied by a quality factor (rem = rad
x Q). The quality factor is a measure of the relative biological effectiveness, which depends on the
part of the body irradiated and the type of radiation.
Sievert (sv) is an SI unit of dose equivalent from ionizing radiation to the total body or any internal
organ or organ system, and is equal to 100 rem.
1 Standard atmosphere = 1 atm = 760 mmHg = 101,325 Pascal = 14.7 lb/in2
May 1999
-------
TABLE OF CONTENTS
Page
PRELUDE *
ABBREVIATIONS LIST u
DEFINITIONS AND CONVERSION UNITS iv
1. INTRODUCTION 1-1
1.0 INTRODUCTION 1-1
1.1 PURPOSE OF THIS DOCUMENT 1-3
2. REMOVAL OF RADON FROM DRINKING WATER 2-1
2.0 INTRODUCTION TO TREATMENT TECHNOLOGIES 2-1
2.1 AERATION 2-4
2.1.1 Process Description 2-4
2.1.1.1 Packed Tower Aeration 2-7
2.1.1.2 Diffused Aeration 2-10
2.1.1.3 Spray Aeration 2-12
2.1.1.4 Tray Aeration 2-13
2.1.1.5 Point of Entry (POE)/Point of Use (POU) Devices 2-14
2.1.1.6 Cone Aeration 2-16
2.1.1.7 Gas-Permeable Membrane Aeration 2-17
2.1.1.8 Air Sparging 2-17
2.1.2 Removal Efficiency and the Effect of Key Design Criteria 2-18
2.1.2.1 Packed Tower Aeration 2-27
2.1.2.2 Diffused Bubble Aeration 2-29
2.1.2.3 Spray Aeration 2-30
2.1.2.4 Tray Aeration 2-30
2.1.2.5 Point of Entry Devices 2-31
2.1.2.6 Comparison of Technologies 2-31
2.1.3 Pretreatment 2-32
2.1.3.1 Iron and Manganese 2-32
2.1.3.2 Other Factors 2-35
2.1.4 Post Treatment 2-36
2.1.4.1 Disinfection Following Aeration 2-36
2.1.4.2 Water Pump Modifications 2-43
2.1.5 Off-Gas Emissions 2-43
2.1.5.1 Worker Radiation Exposure 2-43
2.1.5.2 Air Emissions 2-44
2.1.6 Treatability/Case Studies 2-45
2.1.6.1 Packed-Tower Aeration 2-45
2.1.6.2 Diffused-Bubble Aeration 2-51
2.1.6.3 Spray Aeration 2-52
VI
May 1999
-------
2.1.6.4 Slat or Cascade Tray Aeration 2-53
2.1.6.5 Point of Entry (POE) Devices 2-54
2.2 LOW-TECHNOLOGY AERATION METHODS 2-57
2.2.1 Process Description 2-57
2.2.1.1 Free-Fall Aeration 2-58
2.2.1.2 Low Technology Spray Aeration 2-58
2.2.1.3 'Low Technology Bubble Aeration ..... 2^58
2.2.1.4 Venturi Aeration 2-58
2.2.1.5 Mechanical Surface Aeration 2-59
2.2.2 Removal Efficiency 2-59
2.2.3 Treatability/Case Studies 2-59
2.3 GRANULAR ACTIVATED CARBON (GAC) 2-61
2.3.1 Process Description 2-62
2.3.2 Removal Efficiency 2-70
2.3.3 Pretreatment 2-70
2.3.4 Post Treatment 2-73
2.3.5 Operational Considerations 2-73
2.3.5.1 Gamma Emissions 2-74
2.3.5.2 Spent GAC Disposal 2-77
2.3.6 Case Studies 2-80
3. BEST MANAGEMENT PRACTICES 3-1
3.0 INTRODUCTION TO BEST MANAGEMENT PRACTICES 3-1
3.1 DESCRIPTION OF PRACTICES 3-1
3.1.1 Geologic Controls (Siting)/Alternate Sources 3-1
3.1.2 Regionalization 3-1
3.1.3 Extended Atmospheric Storage 3-4
3.1.4 Blending 3-5
3.1.5 Limitations of BMPs 3-5
3.2 REMOVAL EFFICIENCY 3-6
3.3 DESIGN CRITERIA 3-6
3.4 TREATABILITY/CASE STUDIES 3-6
4. TREATMENT COST ANALYSIS FOR RADON REMOVAL TECHNOLOGIES . 4-1
4.0 INTRODUCTION 4-1
4.1 DESCRIPTION OF COST ESTIMATING APPROACH 4-1
4.1.1 Description of the Approach Used for Estimating and Validating PTA and
GAC Costs 4-2
4.1.2 Description of the Cost Estimating Models For PTA and GAC 4-3
4.1.2.1 The PTA-COST Model 4-3
4.1.2.2 The GAC-COST Model 4-4
4.1.2.3 Other EPA Models 4-5
4.1.3 Case Studies 4-7
4.1.4 Costing of PTA and DBA Units Using Direct Engineering Costing
Methods 4-8
4.1.5 Assumptions for Engineering Cost Factors and Other Costing Inputs 4-9
vn
May 1999
-------
4.1.6 Summary of Design and Cost Assumptions Used for Each Model 4-20
4.1.8 Summary of Assumptions Used for Alternative DBA 4-21
4.2 CAPITAL AND O&M COSTS AND EQUATIONS 4-24
4.2.1 Capital and O&M Costs for PTA and DBA 4-25
4.2.2 Capital and O&M Costs for GAC 4-28
4.3 INTERCONNECTION (REGIONALIZATION) COSTS .'!!.'.' 4-31
4.4 CENTRALIZED TREATMENT FOR SYSTEMS WITH LESS THAN 10 000
GPD 4_31
4.5 COMPARISON OF PTA CAPITAL COSTS WITH CASE STUDIES 4-32
4.5.1 Comparison of O&M Costs With All Case Studies 4-34
4.6 SUMMARY OF EQUATIONS FOR AERATION AND GAC
TECHNOLOGIES 4.35
REFERENCES R_!
APPENDICES
Appendix A-0 Conceptualized Diagrams for PTA Configurations Assumed
in the PTA-Cost Model and the Direct Engineered Approach A-0-1
Appendix A-l Detailed Breakdown of Estimated Costs for PTA and Cost
Curves Based on PTA-Cost Model A-l-1
Appendix A-2 Raw Design and Cost-Estimating Data and Cost Curves
for Direct Engineered PTA A-2-1
Appendix A-3 Detailed Breakdown of Estimated Costs for GAC and Cost
Curves Based on GAC-Cost Model A-3-1
Appendix A-4 Investigation of Possible Off-Gas Emissions Regulations A-4-1
Vlll
May 1999
-------
LIST OF FIGURES
Figure 2-1. Practical Radon Removal with Increasing Air: Water Ratio for PTA 2-28
Figure 2-2. Theoretical Radon Removal with Increasing AinWater Ratio for PTA 2-28
Figure 3-1. Regionalization 3-3
Figure 4-1. Percent Breakdown of Capital Costs for PTA, STA, and MSB A Case
Studies ..."::-.: ~ ... ~:v.... / • - •' ........... .T.-4-11
Figure 4-2. Percent Breakdown of Capital Costs for GAC Case Studies 4-13
Figure 4-3. Mean Percent Breakdown of Various Indirect Cost Compounds 4-17
Figure 4-4. Comparison of PTA-COST Model Capital Costs to Case Studies 4-32
Figure 4-5. Comparison of PTA-COST Model Capital Costs with Other EPA Models .. 4-35
Figure 4-6. Comparison of PTA-COST Model with O&M Costs with Case Studies and
Other Models 4~36
LIST OF TABLES
Table 1-1. Physical Properties of Radon 1-1
Table 2-1. Summary of Technologies for Radon Removal and Removal Efficiencies ... 2-3
Table 2-2. Henry's Law Constants for Selected Compounds (20°C) 2-5
Table 2-3. Removal Efficiency Data for Aeration Pilot Tests 2-19
Table 2-4. Radon Removal for Spray Aeration Pilot Tests 2-30
Table 2-5. Correlation of Occurrence of Fe and Mn with Radon 2-32
Table 2-6. Treatment Levels for Iron and Manganese 2-33
Table 2-7. Packed-Tower Aeration 2-47
Table 2-8. Cascade Tray Aerator • 2-54
Table 2-9. Low-Technology Aeration Removal Efficiencies Observed 2-60
Table 2-10. Radon Removals for Low Technology Techniques 2-61
Table 2-11. Freundlich Isotherm Data and Relative Ranking for Six Activated Carbons
and Radon at 10°C 2-65
Table 2-12. GAC Kr Constant by Carbon Type 2-67
Table 2-13. GAC Kr Constant by Carbon Type as Reported From Different Sources 2-68
Table 2-14. Radon Removal Efficiencies by GAC •• • • 2-71
Table 2-15. Turbidity, Iron, and Manganese Levels 2-72
Table 2-16. Gamma Emissions from GAC Contactors 2-75
Table 2-17. EPA Guidelines for Disposal of Radioactive Water Treatment Plant
Residuals 2-78
Table 2-18. Summary of Hodsdon's 1993 Survey of GAC Treatment Facilities 2-82
Table 2-19. Categorization of POE GAC Units by State 2-83
Table 2-20. Relative Use of Different Sized GAC Units 2-83
Table 3-1. Removal Efficiencies for BMPs 3-6
Table 3-2. Bench Studies 3-7
Table 4-1. Summary of Percentages Recommended by the TDP 4-9
Table 4-2. Percentages Used in the PTA-COST and GAC-COST Models 4-14
IX
May 1999
-------
Table 4-3. Co-Occurrence Data for Iron and Manganese Versus Radon , 4-19
Table 4-4. Summary of Design Inputs for Estimating PTA Costs 4-22
Table 4-5. Summary Design Inputs for Estimating GAC Costs 4-22
Table 4-6. Cost Indices and Other Factors for Models 4-23
Table 4-7. Design and Cost Assumptions for Alternative PTA Configuration 4-24
Table 4-8. Design and Cost Assumptions for Direct Engineered DBA 4-25
Table 4-9a. Capital and O&M Costs for Packed Tower Aeration (PTA) 4-26
Table 4-9b. Costs for Indirect Items Potentially Associated with Aeration 4-27
Table 4-9c. Capital and O&M Costs for Direct-Engineered PTA 4-28
Table 4-9d. Capital and O&M Costs for DBA 4-28
Table 4-10a. Capital and O&M Costs for GAC 4-29
Table 4-1 Ob. Costs for Indirect Items Potentially Associated with GAC 4-30
Table 4-11. Summary Table of Cost Equations for Radon Removal Technologies 4-37
May 1999
-------
Introduction
1.0 INTRODUCTION
Radon-222 (radon) is a noble gas that is formed by the radioactive decay of the immediate
parent element radium-226. Noble gases (Periodic Group 8A) are inert, odorless, and colorless.
Radon-222 undergoes further radioactive decay emitting alpha particles in the process. The half life
of radon is about 3.82 days. The decay products of radon, called radon progeny or radon daughters,
are short half-life radioactive isotopes that emit alpha and beta particles, and gamma radiation. The
concentration of radon dissolved in water is extremely small in comparison to its activity. For
example, a volume of water containing 6.48 xlO'10 mg/L of radon gas contains 100,000 pCi/L.
The physical properties of radon are listed in Table 1-1.
Table 1-1. Physical Properties of Radon(1)
Molecular Weight
Boiling Point
Melting Point
Solubility in Water
Air Diffusion Coefficient
Water Diffusion Coefficient
222 gm M-!
211 K(-62°C)
202K(-71°C)
230cm3/L@20°C
1.2x 10-5m2Sec-'
1.2 x IQ-'nrSec-1
(1) Chemical Engineer's Handbook 6th Edition (Perry, 1984) and The Merck Index Eleventh Edition (Budavari, 1989).
The rate and amount of gas that transfers in and out of water is greatly impacted by its
solubility. Gases either react with water or do not chemically react with water. For gases such as
radon which do not react with water, the attraction that water molecules have to themselves opposes
solubility since a gas must be more attracted to the water than are other water molecules in order for
it to solubilize. Since radon does not bond to water molecules, it is not solubilized. Radon's low
solubility and its high vapor pressure mean that it strongly partitions into the air by diffusion.
1-1
May 1999
-------
Since it easily transfers from water to air, radon is rarely found in surface waters and is
primarily an issue in ground waters. Radon enters drinking waters supply sources from the decay
of naturally occurring radium-226 in the rock and soil matrix. Radon levels can vary greatly from
one region to the next because of differences in the local geology. Radon in well water also varies
due to local, site specific factors such as well depth, distance from the radon source, pumpage
patterns, and the characteristics of the radon source. For example, the relationship between granite
bedrock and high radon levels has been observed in sections of the United States and other parts of
the world (Michel, 1990; Land and Water Resource Center, 1983; Castren, 1977; Sasser and Watson,
1978). In addition to the relationship between granite bedrock and the occurrence of radon, radon
has been detected in thermal springs at concentrations of 100 to 30,000 pCi/L and in regions of
phosphate mining (Hess, et al., 1985; Partridge et al., 1979; Smith et al., 1961).
The National Inorganics and Radionuclides Survey (NIRS) conducted by EPA in 1988
indicated that the concentration of radon hi ground water supplies ranged from the minimum
reporting level of 100 pCi/L to 25,700 pCi/L (Longtin, 1988). Levels of radon in ground water
supplies were in the range of 100 to 1,000 pCi/L for 61.5 percent of the 978 sites sampled in the
NIRS. The highest levels of radon observed in the NIRS were in small system supplies serving
fewer than 500 people. Atoulikian, et al. (1995) estimates that about 83 percent of ground water
systems have aradon concentration of less than 500 pCi/L.and that about 10 percent of ground water
systems have a radon concentration between 500 and 1,000 pCi/L.
The concentration of radon in drinking water may increase or decrease in the distribution
system as it travels from the treatment plant to customers. The decay of radon during transit or
storage in the distribution system has been shown to generally reduce radon levels by 10-20% (NRC,
1998). However, radon levels in the distribution system can also increase due to the decay of radium
that has accumulated in iron-based pipescale (NRC, 1998; Valentine and Stearns, 1994).
Once radon hi water supplies reaches consumers, it may produce human exposure via two
routes: inhalation and direct ingestion. Radon in water transfers into the air during normal water
uses such as showering, flushing toilets, washing dishes, and washing clothes. For inhalation, the
1-2 May 1999
-------
main risk from exposure to radon gas is not from the gas itself, but the radioactive progeny it
produces. This is because radon is an inert gas while the progeny are chemically active-and associate
readily with aerosols (suspension of solid or liquid in air). The aerosols tend to deposit in the lungs
where they release radiation that has been shown to increase the likelihood of lung cancer. Radon
is second only to cigarette smoking as a leading cause of lung cancer in the United States (U.S. EPA,
1994a).
Some of the radon and its progeny also reach body tissues through ingestion, resulting in
radiation exposure to the internal organs. Ingested radon is thought to move from the gastrointestinal
tract to the bloodstream, and from there is carried to the liver, lungs, and general body tissue
(Crawford-Brown, 1990). Crawford-Brown (1990) notes that studies have shown that radon is
generally retained in the body with a half-life of 30 to 70 minutes, and leaves the body mostly
through exhalation from the lungs. Ingested radon is believed to increase the risk of stomach cancer
and, to a smaller degree, the risk of other cancers, but this is based on indirect evidence (Mills,
1990).
A study prepared for the American Water Works Association Research Foundation (Deb,
1992) examined the effect that reducing waterborne radon concentrations had on indoor air radon
concentrations. The study found that a reduction of 1.3x10'4 pCi/L of indoor air radon occurred for
every 1 pCi/L reduction in waterborne radon (Deb, 1992). For example, a reduction in waterborne
radon concentration from 2000 pCi/L to 200 pCi/L (1800 pCi/L reduction, which is 90 percent)
would result in a reduction of 0.234 pCi/L in the airborne radon concentration in a home. This
relationship corresponds well with those found in other research (Deb, 1992).
1.1 PURPOSE OF THIS DOCUMENT
The objective of this document is to support the EPA Office of Ground Water and Drinking
Water (OGWDW) in its preparation of requisite technology and cost documentation for rule
development and regulatory impact analysis for the radon rule. This document presents information
on various radon removal technologies and techniques and provides information on expected
removal efficiencies based on peer reviewed literature and documented case studies. The
1-3
May 1999
-------
technologies and techniques discussed in this document include various aeration technologies,
granular activated carbon (GAC), and storage and other best management practices as means to
remove and reduce radon in drinking water. This document also provides unit treatment cost
estimates for a range of plant sizes, and cost estimate for regionalization,, In addition, it provides
guidance on how regional cost variations may affect unit treatment cost estimates.
1-4
May 1999
-------
Removal of Radon from Drinking Water
2.0 INTRODUCTION TO TREATMENT TECHNOLOGIES
The feasibility of technologies for radon removal from water is largely determined by
radon's chemistry. Other factors include secondary risks from treatment and site-specific
considerations (e.g., physical space constraints). Radon is virtually inert, has a short half-life (3.82
days), and is a soluble gas at normal temperature and pressure (20°C, 1 atm). Because of its short
half-life, 2 days of storage removes about 30 percent of the initial mass and radioactivity of radon
in water by decay alone.
Henry's Law states that the amount of gas that dissolves in a given quantity of a solution,
at constant temperature and total pressure, which is directly proportional to the partial pressure of
the gas above the solution (Zumdahl, 1989). Henry's Law is expressed by the following equation:
HC
where:
p = mole fraction of gas in air = mol gas/mol air
C = mole fraction of gas in water = mol gas/mol water
H = Henry's Law constant = atm
PT = total pressure = atm (usually =1).
Since PT is usually defined as 1, the equation becomes p = HC and H becomes unitless. Thus,
H=p/C, and the larger Henry's constant is, the larger the contaminant concentration in air is at
equilibrium. When a contaminant is at saturation in both the liquid and vapor phase, the partial
pressure of the contaminant is equal to the vapor pressure of the pure material and Henry's Law
constant is proportional to PV/S (where Pv is the vapor pressure of the liquid and S is the solubility
of the contaminant in water). This means that a contaminant with lower solubility and/or higher
2-1
May 1999
-------
volatility (i.e., higher vapor pressure) will have a higher Henry's Law constant. (Faust and Aly,
1998)
The Henry's Law constant for radon in water at 20 °C is 2.26x 103 ATM, or 40.7 L-atm/mole
which is equivalent to 5.09x 1017 pCi/L-atm3 (Hess et al., 1983). Because of this large Henry's Law
constant, radon easily transfers into ah- above water. At 20°C, ammonia (NH3) has a Henry's Law
constant of 0.76 atm, while carbon dioxide (CO2) has a Henry's Law constant of 1.51 xio3 atm
(AWWA and ASCE, 1998). Radon's relatively high Henry's Law constant indicates that it can
transfer from water into the air faster than both ammonia and carbon dioxide, which are readily
strippable gases.
If a water storage tank is left open to the atmosphere and undisturbed, the radon-
contaminated water will lose virtually all radon through diffusion and decay. Dixon and Lee (1987)
noted that, while filling a stand pipe, volatilization and seepage (diffusion) of radon into the air is
a far more important factor than the decay of radon. Data from the 2-day experiment show that
radon levels in the 0.032 MG steel stand pipe effluent were 15 percent less than'the influent radon
levels about one hour after pumping well water into the tank. The well water radon levels were in
the 4,600 pCi/L range. Aeration hastens the diffusion process by providing a larger air/water surface
area and a higher degree of turbulence.
Because of the physico-chemical characteristics of radon and natural processes (e.g., natural
diffusion and decay, turbulence), radon levels in surface waters are typically much lower than those
found in ground water. Since radon has the above-mentioned properties, options for removing
radon from drinking water sources include aeration, adsorption onto another media (e.g., GAC), and
storage.
3 6.48 mg of radon has an activity of 1 Curie. For conversion of Henry's Constant from atm to L-atm/mole
refer to page iv. '
2-2 May 1999
-------
Table 2-1. Summary of Technologies for Radon Removal and Removal Efficiencies
Treatment Method
Packed Tower Aeration
Diffused Bubble Aeration
Point of Entry
Diffused Bubble Aeration
Spray Aeration
Point of Entry
Spray Aeration
Slat Tray Aeration
Low Technology Aeration (2)
Granular Activated Carbon
Percent Removals'1'
78-99.9
71-99.9
92-99.9
35-99
82-93
70-94
10-96
70-99
Comments
- Proven technology
- Low maintenance
- Pretreatment may be required
- . Potential emissions concerns
- Potential temperature concerns
- Potential aesthetic concerns
- Proven technology
— Low maintenance
- Low profile and compact
- Pretreatment may be required
- Potential emissions concerns
- Potential temperature concerns
- Multiple passes required for high removals
Operational problems in cold conditions
- Pretreatment may be required
- Potential temperature concerns
- Footprints maybe larger than those needed for
other technologies
- Potential temperature concerns
- EBCT of 30-1 30 minutes (longer than that
needed for the removal of taste and odor and
volatile organic compounds)
- Radiation concerns.
(1) Removals as high as these ranges have been reported in literature. .
2) Low technology processes include relatively simple techniques such as the use of free-fall aeration, spray nozzles, or Ventun
laboratory devices to deliver influent to an atmospheric storage tank, or mechanical surface aeration to agitate the water in a tank
or basin.
Table 2-1 shows various technologies available for the removal of radon. These
technologies are water treatment processes within the technical and financial capability of most
public water systems. Prior to implementing a technology, site specific engineering studies of the
methods identified to remove radon should be performed. The engineering study should evaluate
technically feasible and cost effective methods for the specific location where radon removal is
required. In some cases a simple survey may suffice, while in other cases, extensive chemical
analysis, design, and performance data will be required. The study may include laboratory tests
and/or pilot-plant operations to cover seasonal variations, preliminary designs, and estimated capital
and operation costs for full-scale treatment. The evaluation of other options, such as point of
use/point of entry devices and spraying in storage tanks, as well as best management practices
2-3
May 1999
-------
(BMPs) such as extended atmospheric storage, may be included. Cost estimates for such
engineering studies will vary based on factors such as raw water quality, system size, and the
number of options.
Radon removal techniques can be divided into three categories:
• Aeration
• Granular Activated Carbon (GAC)
• Simple techniques and best management practices.
The sections that follow in this chapter contain a description of these technologies, discussion of
removal efficiencies achieved, issues related to pretreatment, post treatment, and off-gas emissions,
and information gathered from treatability/case studies. Simple techniques and best management
practices are discussed in Chapter 3.
2.1 AERATION
2.7.7 Process Description
Aeration may be described as the process of bringing air and water into close contact with
each other for the purposes of transferring undesirable water constituents to air, oxidizing some
natural organic matter (NOM), and improving the treatability of water. Aeration has been used
effectively in water treatment to reduce the concentration of taste and odor-producing compounds
such as hydrogen sulfide and certain synthetic volatile organic compounds (VOCs), to remove
carbon dioxide to reduce corrosivity and lime demand in lime softening treatment, and to oxidize
iron or manganese. However, the use of aeration solely for the purpose of controlling radon is a
relatively new concept in the drinking water industry.
The driving force for mass transfer of radon from water to air is the difference between the
actual concentration in water and the concentration associated with equilibrium between the gas and
liquid phases. The equilibrium concentration of a solute in air is directly proportional to the
concentration of the solute in water at a given temperature according to Henry's Law. Henry's Law
(p=Kc) states that the amount of gas that dissolves in a given quantity of liquid (c), at constant
2-4
May 1999
-------
temperature and total pressure, is directly proportional (K) to the partial pressure of the gas above
the solution (p). Thus, the Henry's Law constant (K) can be considered a partition coefficient which
describes the relative tendency for a compound to separate, or partition, between the gas and liquid
phases at equilibrium.4 Aeration is used to increase the speed of the natural process of moving
toward equilibrium between dissolved, volatile substances in the water and the same substances in
the air to which the water is exposed. Aeration also enables more of the dissolved, volatile
substances to move from water to air by exposing the water to a fresh source of air that has lower
concentrations of the substances.
Equilibrium constants for radon and several other compounds which have been found in.
ground water supplies are presented in Table 2-2. A Henry's Law constant is a measure of the
relative escaping tendency of a compound; a compound with a high vapor pressure and a low
aqueous solubility tends to volatilize more readily. Thus, high Henry's Law constant indicates
equilibrium favoring the gaseous phase; i.e., the compound generally is more easily stripped from
water than one with a lower Henry's Law constant. As shown in Table 2-2, radon has a larger
Henry's Law constant than carbon dioxide and trichloroethylene which are known to be easily
removed by air stripping.
Table 2-2. Henry's Law Constants for Selected Compounds (20°C)(1)
Compound
Vinyl Chloride
Oxygen
Radon
Carbon Dioxide
Tetrachloroethylene
Trichloroethylene
Ammonia
Henry's Law Constant12'
(atm-m3)
mole
6,295 x 10°
773 x lO'3
40.7 x lO'3 .
27.2 x ID'3
19.8 x lO'3
9.89 x lO'3
0.0137 x lO'3
Henry's Law Constant*2' atm
3.5 x 10s
4.3 x 10"
2.26 x 103
1.51 x 103
1.1 x 103
5.5 x 102
0.76
(1) To convert from atm- mVmole to atm, the following equation applies: H (atm- mVmole) x p/RT=H(atm), where P is pressure
in atmosphere, T is temperature in Kelvin, and R is the universal gas constant (8.205 x 10'5 atm-mVmole).
(2) Hess et al., 1983.
4 Henry's Law applies to most gases, particularly those that are slightly soluble and do not react with the
solvent (e.g., dilute solutions like radon in ground water).
2-5
May 1999
-------
2-6
May 1999
-------
The following factors are key elements in controlling the transfer of volatile substances from
water to air and must be considered in the design of aeration systems:
• Contact time (time of exposure)
• Area to volume ratio (available area for mass transfer, air to water ratio)
• Proper dispersal of waste gases into atmosphere (gas transfer resistance, particularly due
to liquid film and gas film resistance at the air-water interface; partial pressure of gases
in the aerator atmosphere; turbulence in gaseous and liquid phases)
• Physical chemistry of the contaminant
• Influent concentration of the contaminant
• Water and surrounding air temperatures.
The first three factors are aeration unit dependant, while the last three are contaminant and site
specific.
Aeration may also have other effects besides radon and VOC removal. These secondary
effects may be either beneficial or adverse and may include the following:
Beneficial
Adverse
Removal of hydrogen sulfide and other taste and odor-causing compounds.
Removal of some carbon dioxide which results in increased pH and lower corrosivity.
Potential reduction in the amount of chlorine needed to treat water. Since aeration
removes sulfide, it can significantly reduce the amount of chlorine needed to oxidize
sulfide (Dell'Orco, et al., 1998). However, there may be no net reduction in chlorine
dose since aeration also increases pH and thus increases chlorine requirements.
Partial oxidation of iron and manganese that may be removed by subsequent filtration.
Permitting procedures may be required for off-gas emissions containing radon in some
urban locales; although, a properly designed system would not pose a significant risk to
the public due to the dispersion of gases containing radon and its progeny (as discussed
in Section 2.1.5.2 of this report).
2-7
May 1999
-------
• Increased potential for scaling in the distribution system due to the increase in pH.
• Increased corrosivity due to higher dissolved oxygen levels.
• Need to disinfect treated water and aeration equipment.
• Need to prevent deposition of iron and manganese in the distribution system.
Aeration technologies can be divided into four basic categories:
• waterfall aerators,
• diffusion or bubble aerators,
• mechanical aerators, and
• pressure aerators.
Some of the more common types of waterfall aerators are packed tower/column, spray, tray, cone,
and cascade aerators (AWWA and ASCE, 1998). Several aeration technologies can be applied both
at water treatment plants to treat full water supplies and at homes as point of entry (POE) devices.
Technologies hi the first two categories, including their application as POE devices, and the
emerging technologies of gas-permeable membranes and sparging are presented in this section.
Some of the technologies in the first two categories, such as spray aeration arid cascade aeration, can
be applied using simpler structures in what can be considered a low technology manner. Similarly,
technologies in the third category can be classified as low technology. These lower technology
aeration techniques are described in Section 2.2. Pressure aerators, used to aerate water that is under
pressure, are available in two types. One type sprays water into the top of a closed tank while the
tank receives a continuous supply of compressed air; aerated water leaves from the bottom of the
tank. With the second type, compressed air is injected directly into a pressurized pipeline to add air
bubbles to the flowing water. Pressure aerators are applied in iron and manganese oxidation, but are
not used for radon removal, so they are not discussed further in this document.
2.1.1.1 Packed Tower Aeration
2-8
May 1999
-------
Radon is readily volatilized from water and thus is easily stripped like many VOCs. Packed
towers have been shown to be the most efficient form of aeration for VOC removal, therefore packed
towers have been applied for radon removal. In countercurrent flow packed towers, packing
materials are used which provide high void volumes and. high surface area. The water flows
downward by gravity while air is forced upward. The untreated water is usually distributed on the
top of the packing with sprays or distribution trays and the air is blown up the column by forced or
induced draft. This design results in continuous and thorough contact of the water with air and
minimizes the thickness of the water layer on the packing, thus promoting efficient mass transfer.
The design of air stripping equipment has been extensively developed in the drinking water industry
for VOC and hydrogen sulfide removal and in the chemical engineering industry for stripping
concentrated organic solutions.
The removal of radon using packed tower aeration is determined by the following factors:
' s
• Air to water (A:W) ratio
• Contact time
• Available surface area for mass transfer
• Surface loading rate
• Physical and chemical characteristics of radon (an inert gas with a high Henry's Law
constant)
• Radon concentrations hi the influent water and air
• Temperature of the water and the air.
The design of a packed tower aerator plays a major role in establishing the effects of the first
four factors, while the last three factors are established by the contaminant, source water, and
location of the tower (AWWA and ASCE, 1998).
The air flow requirements for a packed tower depend on the Henry's Law constant for the
particular compound(s) to be removed from the water. In a perfect aeration system, the minimum
A:W ratio which will achieve complete removal of a contaminant is dependent on Henry's Law
2-9
May 1999
-------
constant. The greater the Henry's Law constant, the less air is required to remove the compound
from water. Because aeration systems are not perfect and the contaminant concentration in the feed
air may not be zero, actual A:W ratios to achieve a given removal efficiency are greater than the
ideal or theoretical relationship between radon removal and the A:W ratio (Spencer and Brown,
1997).
The contact time is a function of the depth and type of the packing material. An increase in
the depth of packing material results in a greater contact time between the air and the water, and
consequently, higher removals are achieved. The depth of the packing material is determined by the
height of the packing in the tower.
The available surface area for mass transfer is a function of the packing material. Various
sizes and types of packing material are available including '/4-inch to 3-inch sizes and metal, ceramic
and plastic materials. In general, the smaller packing materials provide a greater available area for
mass transfer per volume of material thus increasing the mass of contaminant removed. However,
the resulting increased pressure drop for air passing through the column must also be considered.
The surface loading rate is the amount of water that passes through the tower and is largely
a function of the diameter of the tower and the system design flow. The surface loading rate
typically ranges from 25 to 30 gpm/ft2 (AWWA and ASCE, 1998).
Temperature affects the solubility of radon in water and its Henry's Law constant. As the
temperature increases, radon's solubility in water decreases. However, radon, as an inert gas, is not
expected to show a large margin of difference in solubility between near freezing temperature and
20°C. Although removal efficiencies usually increase as water temperature increases for packed
tower aerators, heating influent water is generally not cost effective (AWWA and ASCE, 1998).
Packed tower aeration can generally be used with systems of all sizes. A typical packed
tower installation consists of the following:
2-10
May 1999
-------
Packed Tower—Either metal (stainless steel or aluminum), fiberglass-reinforcedplastic,
or concrete construction. Internals (packing, supports, distributors, mist eliminators) are
generally made of metal or plastic. Packing can be random or structured.
Blower—Typically centrifugal type, either metal or plastic construction. Noise control
may be required depending on the size and system location.
Effluent Storage—Generally provided as a concrete clearwell (also called airwell)
below the packed tower. Typical storage time is 5-15 minutes of design flow with the
higher storage capacity at small systems (U.S. EPA, 1979).
Effluent Pumping—Generally required since effluent is at atmospheric pressure.
Vertical turbine pumps mounted on clearwell are typical.
Packed towers are often installed outdoors, potentially creating temperature, aesthetic, and
noise concerns. In cold climates, piping should be protected from freezing, especially during low
flows that occur during periods of lower demand for water. Fog and surface icing may also be cold
weather concerns. Aesthetic problems due to the height and appearance of a packed tower may
necessitate special artistic touches and architectural designs. Some large outdoor facilities may need
to locate the blowers in a building when noise is a concern. In addition, public perception about off-
gas emissions may require a public relations/outreach program.
2.1.1.2 Diffused Aeration
Aeration is accomplished in the diffused-air type equipment by injecting bubbles of air into
the water by means of submerged diffusers. Diffusers are usually either porous plates or tubes, or
perforated pipes. The older, more traditional applications included a deep tank. The more recently
developed diffused-bubble aeration systems include a shallow depth tank. Ideally, diffused aeration
is conducted counterflow with the untreated water.5 The untreated water enters the top of the basin
and exits from the bottom treated, while the fresh air is blown from the bottom and is exhausted from
the top. The air bubbles produced by the diffusers rise through the water, creating turbulence and
providing an opportunity for the transfer of volatile materials. Gas transfer can generally be
improved by increasing basin depth, producing smaller bubbles, improving contact basin geometry,
and by using a turbine to reduce bubble size and increase bubble holdup (EPA, 1992).
Since some mixing results from aeration, flow is rarely this straightforward.
2-11
May 1999
-------
Diffused aeration generally provides less interfacial area for mass transfer but greater liquid
contact time when compared to packed towers (AWWA and ASCE, 1998). Diffused aeration
provides an optimum treatment system for the dissolution of a soluble gas in the water (i.e.,
oxygenation or ozonation), whilepacked tower aeration provides an optimum system for the removal
of volatile contaminants from the water.
A viable option for small and medium sized systems is a variation of diffused aeration
technology called multi-stage bubble aeration (MSBA). MSBA units are available commercially.
Typical commercial units consist of a high-density polyethylene vessel partitioned into multiple
stages with stainless steel and polyethylene divider plates. Each stage is provided with an aerator.
Individual aerators are connected to a supply manifold. The units are compact and low profile.
Water depths are shallow for MSBA, with sidewater depths typically less than 1.5 ft (compared to
depths of 10 to 20 ft for typical aeration basins).
Diffused aeration may be adapted to existing storage tanks and basins. The air diffusers may
be placed on the side of the tank to further induce turbulence and assist in gas transfer. When porous
plates are used, they are located at the bottom of the tank. If porous tubes or perforated pipes are
used, they may be suspended at about one-half depth of the tank to reduce compression heads.
Diffusers are designed to produce bubbles of certain sizes. Smaller bubbles create more total area
for mass transfer, thus increasing the exchange of volatile substances. When porous diffusers are
used, incoming air should be filtered carefully through an electrostatic unit or a filter of metal wool
or glass in order to minimize clogging. Static tube aerators have also been used in a variety of
applications and have provided adequate aeration when properly designed.
The design of diffused aeration equipment has been developed extensively in the chemical
processing industry for handling concentrated organic solutions. The procedures found in the
chemical engineering literature can be applied to water treatment for radon. The rate at which radon
is removed from water by diffused aeration depends upon many of the same factors as for packed
tower aeration:
• Temperature of the water and the air
2-12
May 1999
-------
" Physical and chemical characteristics of radbn
« Radon concentrations in the influent air and water
«. A:W ratio
» Contact time (e.g., flow rate)
» Available area for mass transfer (e.g., bubble fineness).
The first three factors are fixed by the liquid stream and the contaminant; the last three are dependent
upon the equipment and operating conditions and can be evaluated in a pilot testing program.
This technology has a number of advantages and disadvantages relative to packed tower
aeration (PTA). The advantages include the potential for modifying an existing basin or storage tank
with diffused aeration, and marginal savings due to no packing costs, reduced pumping costs, and
generally lower energy costs. MSB A in particular offers the advantage of being compact and thus
is favorable for aesthetic reasons and often involves lower building costs. The disadvantages include
the requirement of increased contact time (which could rule out the use of a given modified basin
or storage tank), the possibility of needing a greater A:W ratio, and overall less efficient mass
transfer. MSBA is also limited in treating larger flows.
2.1.1.3 Spray Aeration
Spray aerators direct water upward, vertically, or at an inclined angle, in such a manner that
the water is broken into small drops. Installations commonly consist of fixed nozzles on a pipe grid.
The small droplets formed expose a large interfacial surface area through which the radon migrates
from the liquid phase to the gaseous phase.
Design factors that impact the effectiveness of spray aration include:
• Nozzle design and operating pressure (e.g., velocity of spray)
• Nozzle orientation (e.g., size, number, and spacing of multiple spray nozzles; nozzle
trajectory)
• Distance of water droplet free fall
• Water droplet size
2-13
May 1999
-------
Amount of ventilation (including the effects of wind on the movement of rising and
falling water droplets).
Although the use of many small spray nozzles .that each produce very small water droplets
may provide the greatest area-volume ratio (i.e., most available area for mass transfer), these small
nozzles tend to clog and require high maintenance. Spray aerator nozzles generally have a diameter
of 1.0-1.5 in., discharge ratings of 75-150 gpm (at about 10 psi), and are installed every 2-12 feet
apart. (AWWA and ASCE, 1998)
Spray aeration, like diffused aeration, has a number of advantages and disadvantages in
comparison to other aeration technologies. The advantages include the capability of achieving
efficient mass transfer due to the small water droplets created by the fixed nozzles, the lack of any
packing costs, and potentially lower maintenance costs. The disadvantages include the need for a
large operational area, which translates into increased building construction costs, potentially
increased operating problems during the cold weather months when the temperature is below the
freezing point, short exposure time between air and water, and high pressure requirements (AWWA
and ASCE, 1998).
2.1.1.4 Tray Aeration
A slat tray or multiple tray aerator consists of a series of trays (usually 3—9 trays spaced
12-30 in. apart) equipped with slats, or perforated or wire-mesh bottoms, over which water is
distributed and allowed to fall to a collection basin at the base of the unit. Distribution of the water
over the entire tray area is important from an efficiency standpoint. In many tray aerators, coarse
media such as coke, stone, or ceramic balls ranging from 2 to 6 inches in size are placed in the trays
to improve the efficiency of gas exchange and distribution. Radon removal occurs as the water
falling through the trays contacts air and radon is transferred from the water to the air. Air can be
supplied to tray aerators either with a natural draft or a forced draft from a blower.
A cascade tray aerator is essentially a slat tray aerator without a blower to force air through
the aerator. Instead, the water is allowed to flow over the trays and fall to a basin through naturally
induced air.
2-14
May 1999
-------
If tray aerators are placed in a poorly ventilated building, performance will be impaired and
operator safety may be of concern. Artificial ventilation is a requirement under these conditions.
Artificial ventilation is provided in certain types of tray aerators by enclosure and provision of a
forced draft. Such aerators generally employ the countercurrent flow principle. The air is supplied
at the bottom of the aerator enclosure by a blower and travels upward through the aerator counter
to the downward flow of the water. The counterflow of air and water is advantageous, and such
aerators have shown excellent oxygen absorption and carbon dioxide removal capabilities. More
research is needed to determine how the use of counterflow impacts radon removal capabilities.
Another variation of this technology is shallow tray aeration, where the primary component
is a shallow tray module that has one to six compartments or stages of limited depth (e.g., 18 to 30
inches). Water is pumped through the module as air is pumped in through diffusers at the bottom.
Since these units are generally modular, they are compact (yielding a smaller footprint) and are
considered relatively simple to install for both retrofits and new construction.
A type of multiple tray aerator, the crossflow tower, has been extensively utilized in cooling
applications. In this system, water is allowed to fall over the tray area while air is forced or induced
to flow across the slats perpendicular to the water path. In some crossflow columns, air is drawn in-
from the sides and expelled out the top of the aerator. This type of column is actually a hybrid of
the countercurrent and crosscurrent units. The reviewed literature did not provide data on radon
removal rates for crossflow tower tray aerators.
Tray aerators can generally be used with systems of all sizes. One disadvantage of tray
aerators is that slime and algae can grow on the trays, possibly necessitating the addition of copper
sulfate or chlorine to control growth (NRC, 1997).
2.1.1.5 Point of Entry (POE)/Point of Use (POU) Devices
Point of entry (POE)TPoint of use (POU) treatment devices are installed at residences where
only household water is treated. POU systems generally treat only water used for drinking and
cooking, while POE systems treat the entire supply entering a household. POU treatment is provided
at the tap by devices that can be faucet mounted or with line-bypass units. As the name implies,
2-15
May 1999
-------
faucet-mountedunits are devices which attach to an existing faucet. Line-bypass units supply treated
water to a special tap or third faucet which is usually located at the kitchen sink. These treatment
devices themselves are generally located under the sink and are usually larger and treat a greater
volume of water than faucet-mounted units. POE devices can be installed in the basement or outside
the home. The health risk associated with airborne radon would necessitate that all of the water
entering a house be treated for radon removal. Therefore, a POE system would be more appropriate
than a POU treatment device from an exposure standpoint.
Diffused bubble, bubble plate, and spray aeration are adaptable to individual family homes
or, hi some cases, buildings serving more than one family, or a small community. Packed tower
aeration is not generally used in POE systems because of its greater cost and aesthetic and
installation concerns due to tower height. MSBA and shallow tray aeration can be installed as POE
systems, but may be cost prohibitive. POE devices may require pretreatment and post treatment to
avoid operational problems.
POE treatment devices provide a technically feasible alternative to centralized treatment
facilities, and may be more economical than a central treatment system. POE treatment may be most
economically competitive in very small water systems where the economies of scale for the larger
centralized facilities are not apparent. Another advantage of POE treatment is the small capital
investment relative to centralized treatment for many technologies. In some cases, POE treatment
may provide more effective contaminant reduction than the use of a central treatment system.
However, even though as a treatment technology POE devices may be capable of removal
efficiencies equal to those achieved with full scale systems, this treatment alternative has some
disadvantages. These include:
Increased complexity of controlling treatment, monitoring, maintenance, and regulatory
oversight of the devices.
Concern for possible bacterial colonization on the treatment devices.
Potential for high operation, monitoring, and maintenance costs.
Standards have only been developed for a limited number of POE devices.
2-16
May 1999
-------
« Potential for radiation hazards at homes.
« Liability associated with failure of the devices.
Responsibility for the use of POE devices rests with the public water system. The Safe
Drinking Water Act (SDWA), as amended in 1996, specifies that POE and POU treatment units
"shall be owned, controlled and maintained by the public water system or by a person under contract
with the public water system to ensure proper operation and maintenance and compliance with the
maximum contaminant level or treatment technique and equipped with mechanical warnings to
ensure that customers are automatically notified of operational problems." (SDWA, Section
Standards and a certification program for some POE devices have been developed by the
National Sanitation Foundation (NSF International). Initially prompted by State drinking water
administrators (and later by U.S. EPA), NSF began developing standards and a certification program
for the contaminant reduction claims (both health and aesthetic) made by manufacturers. NSF is
currently working on standards for G AC POE devices. Manufacturers can submit their products and
product information to NSF to apply for certification. (NSF, 1998)
The Water Quality Association has also developed a certification program, however, its
equipment assessments are not independent since it is a trade association for POE and POU
equipment manufacturers. Regulations for certifying POE devices have been developed by several
States. California only allows POE devices to be installed when other alternatives have been
evaluated and determined to be infeasible. (NRC, 1997)
2.1.1.6 Cone Aeration
Cone aerators consist of a stack of pans, with inverted cone-shaped protrusions extending
from the bottom of each pan. Water fills the top pan, then drains down through the inverted cones
to cascade down to the succeeding pan. As water flows through the inverted cones (wider at the top,
^narrower at the bottom), it increases in velocity and splashes down onto the pan below, generating
2-17
May 1999
-------
bubbles and greater contact with the air. A review of literature did not find any studies using cone
aerators for radon removal.
2.1.1.7 Gas-Permeable Membrane Aeration
Gas-permeable membranes have the potential to achieve the highest removal efficiencies,-
but their long-term performance has not been sufficiently evaluated to consider this technology more
than emerging. For this technology, water flows through tubes made of highly porous fiber
membranes that allow gases but not liquids to escape. The membranes provide a very large area for
air and water contact compared to other aeration systems of similar size and thus may prove to be
very efficient for removal of both semivolatile and volatile organic chemicals. (NRC, 1997) At this
point, no literature is available on the application of gas-permeable membrane aeration for radon
removal.
2.1.1.8 Air Sparging
Air sparging is the inj ection of pressurized air into water reservoirs or directly into well water
or an aquifer. For reservoir sparging, air is injected, under pressure, near the bottom of the reservoir,
creating air bubbles in the reservoir. Radon is transferred from the reservoir water to the air bubbles
and then carjied to the surface by the air bubbles and released to the atmosphere. For aquifers and
wells, air is injected under pressure below the water table. The air bubbles, pushing in three
dimensions through the soil column, carry radon into the vadose zone. From the vadose zone, radon
migrates to ground surface and continues its movement into the atmosphere.
Like air stripping and diffused bubble aeration, air sparging will volatilize radon. Radon
removal efficiency using air sparging varies and depends on the amount and pressure of the air
injected, reservoir or well column depth, surface area of water in direct contact with the atmosphere,
and venting conditions in the reservoir or well column. Design of an air sparging system may
require pilot tests to determine the radius of influence of the air sparging point(s) and the
effectiveness of the system. For reservoir sparging, design of the air sparging system requires
knowledge about reservoir venting, water inlet and outlet conditions, and mixing conditions. For
2-18
May 1999
-------
aquifer sparging, design of the, air sparging system requires knowledge about the hydrogeology of
the aquifer.
Some of the complications from air sparging include biological growth and precipitation of
metals, particularly in hard or iron and manganese-containing ground waters. Water sparged with
air maintains high dissolved oxygen which enhances biological growth. Biological growth might
create water quality problems including the formation of slime material and odor and color problems.
Air sparging also enhances precipitation of metals, which may cause pitting and corrosion of the
pump and motor. Moreover, pressurizing air and delivering pressurized air to water in wells,
aquifers, and reservoirs is energy demanding.
The application of air sparging to well ground water for radon removal has been limited.
Experimental trials have not been encouraging because of the problems mentioned above (Hess,
1998). A feasibility study of in-well aeration conducted by the North Perm Water Authority found
this treatment technology to have relatively low efficiency and identified several disadvantages,
including causing the water to appear milky as a result of dissolving large quantities of air into the
water (AWWA and ASCE, 1998).
2.7.2 Removal Efficiency and the Effect of Key Design Criteria
Reviewed studies on aeration technologies provide data on removal efficiencies for radon.
These data are summarized in Table 2-3. Removals for packed tower aeration ranged from 78.6 to
greater than 99 percent, with most removals reported at 90 percent or greater. For the two diffused
bubble aeration facilities, removal efficiencies were 93 and 95 percent. Removal efficiencies for
multi-staged bubble aerators ranged from 71 to 100 percent. These studies showed wide variation
in removal efficiencies for spray aerators, with removals ranging from 35 to 99 percent depending
on operating conditions. For tray aeration, the studies evaluated reported radon removals of 70 to
99 percent. POE systems using diffused bubble and multi-stage diffused bubble aerators
technologies achieved radon removals of 92 to 99.9 percent. Spray aerators installed in homes have
shown radon removals of 82 to 93 percent.
2-19
May 1999
-------
•a
1
H
2
<2
|
Q
§
•0
I
Comments
•M ^»
||
DS 5
"O )M
S CU
es -3
gfi
System Design, Si
Other Design Pars
O> g
cu c
i- es
3 b<
0 *
on ft.
c ^
1 g
Q.
u "3
9 >
I"3
O
V
5
C/3
C
e
<
1
f2
•s
Cl
K
C.
*-
O CO
= o =
.2 «" § o S
- Tested on water drawn from 1 well
- Similar removals at all A:W ratios;
- Packing height had greatest effect on
removals;
- Strong correlation between concentral
CO, and radon in raw and treated water
CO2 could be a surrogate for testing rad
removal;
- Minimal maintenance; several possibi
retrofitting options; lack of redundancy
service stoppage; aesthetics concerns
0 t*. -
i_ K ° ~
"• 02 i fLZ
O ••— " "O cfl CN
2 § c .2 ~"
cV- 00 in tt- ts 77
^ "5 (Q £ ^ » "~
m 5 u- D. * ' T^ o
ON -2 o ofl <, m (^i
«J CO °° 2
«2ic3i> 12a3eO~
VO "^ ^ ^ ^ •" 'rN fll
S 1 "- 2 ^ "« « o "2 o
O-^Q «^ g SlVO'"*''
co S* ^5 c "C Q^ ^ O ? Si
*S ,S2 4_* ^* ^" w ^ £ co
'S'O^c'p^"— i)"§"f?
ilflliillii
eu jz com u Ei: cnHifcis
o
— T3
u >>
O ra
> g
o
o
7
§
— "
O
CO
5 Os"
C oo
^ ON
< c.
2
_
•a
1 1 systems evaluated
Compact system, but aesthetic concerns
tall towers (some >25 ft)
£
ON
°!~
ON
.E
1 i
II
5= aJ
Ic
t. 00 •=
> o
s ° «*>
2 o —
•0 ON" 0
4> Q "*"*
Ja o\ CM'
0. 1 1
^
7.
7.
O o?
S ON
Q C.
E c
"to .£
Design recommendations for full-scale :
based on results of pilot test;
Packing bed depth is key factor in deter
removal efficiency
s=
O
ON
acked tower
450 gpm flow
14 ft high, 5 ft diameter
0.5 blower horsepower
CU 1 1 1
^t
7.
o
o
—
0 o~
§ ON
_J ^,
E c
CD .=
« .£
>•» E
Design recommendations for full-scale i
based on results of pilot test;
Packing bed depth is key factor in deter
removal efficiency
v^
o
ON
acked tower
350 gpm flow
14 ft high, 5 ft diameter
0.33 blower horsepower
0. I I |
^
7.
0
o
— "
8 o1
S ON
g ON
E C?
w c
£> 1 .
Design recommendations for full-scale i
based on results of pilot test;
Packing bed depth is key factor in deter
removal efficiency i
ee
O
Is
xO v= C.
\f\ ^ (4^
ON ' ON O
acked tower
450 gpm flow
21 ft high, 5 ft diameter
0.75 blower horsepower
CU I | I
~
'Sou
_j _o 2^
"ob -g g
2. CO Q>
8 2 f
0
o
rn
O o"
V °^
~J o
E/5
System designed and installed for VOC:
Avg daily consumption 87,000 gpd
Gas chlorination disinfection
! '
*O
V
S? i J1
» 1 C
ON ^^ fi.
acked tower
tower 33+ ft tall
supply capacity 450 gpm
counter current flow
a. I I I
•
-------
(Continued)
«
D£
_O
"o
C
1
H
c
L.
b.
"«
O
^_^
c*
e
.2
"S
S
i
«
i
_w
R
H
Comments
1 1
II
"§ o
« "S
.£2 CB
cc Js
(? C"
.2f B
oj •••
O % ^
1 « 1
iti
15 g 8
Q- 1 1
z
/— »
5
OI
eo o"
> -^
C*4
VO "
•sr" C-
c
o ^_^
ll
Avg daily consumption 14,000 gpd
Chlorine hypochlorite disinfection
Os **"-•
JL d
|O
g E g"
£,
v> o
t"*- fj^
>> •»-
U ^
fc- eo c
III
!li
0. I 1
X
o
O
O
Os"
i
"O £?
(/i rn
T3 ^
_O Os
W
IM
o
CL) (3
*- .0
•S tlJ
'5. 'c
Avg daily consumption 46,000 gpd
Liquid chlorine disinfection
3 bag filters installed to remove preci
iron (result of radon retrofit)
Water passes through ion exchange u
tower to reduce iron and hardness
o
o
^ l~
5^ 3 -J
!
60 £
10 O
oo c
>% «-
J CO o
1 II
CU 1 1
z
o
0
•— '
§
"8 <^
•o ^
Q O^
Avg daily consumption 82,000 gpd
Gas chlorination disinfection
Major retrofit to existing system
V
•*_»
c
2? 3 j"
Q.
00
0 S
""> ij O
— j: C
>>.2P -
•3 .8 S
"~ B u! *-
1 t« 1
•g 3 ro O
,.„ « m o
a. i i i
Z
o
r^
pr\
2
^^
r^"
C
o
"s> g^
"o os
S^ 3 J
so 'o O
co -o a.
n. ^
o
Os rj~
>\ •>-
"u §
i- co t
0 Q. 3
US
111
£ 1 1
z
o
o
tN
of
C
o
•a ^
O Os
•n ^,
Avg daily consumption 240,000gpd
Chlorination
Retrofit for C02 removal
m
oo
II
c
ID s-^.
^|u
1
o
^r
& ^
"o rr
u. W ^^
^— * 3
•§ §• §
CU 1 1
7.
m
^
™«*
oo
o
01^
—
o _.
•o ^
tn JT>
Jl
-a .
0. I
SO u
o M
o' S
°. o
Avg daily consumption 300,000-400
Chlorination
Major retrofit for radon removal & si
new high lift station & clearwell
oo
Os
II
CL
00
O
v^
^J
& ^
'3 [7-
fc- CO ^"^
fli Q_ •*-•
1 « g
ill
0- 1 1 .
Z
O
OS
t""^
^M
m
r~
vi^
—
c
o
"S f>
•a Os
S —
2-21
May 1999
-------
(Continued)
8
•g»
"o
B
JS
CJ
H
c
o
^C
2
CJ
dg
C5
CM
-£
O
e>
i
"es
i
£
,
r*5
<4
&>
^5
«
H
Comments
•s >
> e
O CJ
E '«
o £
K fjj
•a £
C O
O
1|
BO s
°5> B0
fi |
>> —
co O
O* £
5^ ^
"^ a>
s s
s £
O €5
cc fri
C _.
o 'j
^5 •^
*5 o
CM ~^
•W sS*
c ZT
O QJ
15
0
3
O
^
5?
\/i
*^*!
s
i
e
^o
«z
1
15
ea
"s
tn
S
O
CO
5 « o S3
O > -T-t ^.
Tested on water drawn from 1 well;
Strong correlation between concentrati
CO2 and radon in the raw and treated \
C02 could be a surrogate for testing ra<
removal;
minimal maintenance;
redundancy possible with standby blov
" $? " -4-*
» *^ («^ ^^ > go
'l^ ft J^ >M ^ >O Q C
"^ "^ O 4_*
^3 m ^eo i: c
E "1 - §1 £ •§*!
^/i O r*» . -^ K3 t- t.
^°» Q *T*- ^""" ^ O "O — .
_o o — «" .c ^ ta t^
^ f s « 1 i 1 1
CO co cc c-sj *^3 .£ u co
""S^vd i"S2 = «
« CO gj .„-=• ODE *>
5lS"o Q. 2 "» co"g S
o
CO .^
CO QJ
en N
U .£>
O a
^^ A
o
o
7
o
0
—
00 ^_^
^ ON
^» CO
•£• ON ,
< c.
8 systems evaluated
Less compact than PTA
c£
+
Os
°\
t—
c
*— CO C
to S o
S & 1
CO O 4>
S s= -a
-§ D..C
If ^l t^
to "— o
J 2 o
^^
7-
<
"Z-
If
Q O
o
•o
CO
CO
C
J
O
o.
o
0
.CO
'c
=s
1
o"
T
-
&o o
- i
O _u C3
1 1 i „ S
c/5'ln 1 1 1
^f
•z.
o
0
o
CN"
3: 0s
o 2
_j
At A:W=8, achieve effluent of 100 pC
(instead of 200) up to 230 gpm
cN
O
i
3
O
— tT"
00 0
^ 2
Stripper Aeration Model
stage diffused bubble ae
- 3 1 to 325 gpm flow
- 1 module
-A:Wof5.7
^f
•z.
o
o
0
CN"
5 ON
0 =*
_J ^,
o
»rt-
2 §
_ o
ci. CN
1 s,
§ i «
s J 1
-" 3
00 -0 =
a. o g.
_u E eo
00 -= >V1
C g CN
•s; o ^
« ig 2
7. !§•
^ -5
... (U -J
S 1 o 0
U- 00 U. Q.
^
O
ON
,
5
1
o
"^ /— *
00 O
a- 2
Stripper Aeration Mode
stage diffused bubble ae
- 326 to 650 gpm flow
- 2 modules
-A:Wof5.7
^£
z
o
o
o,
CN
> ON
o 2
— i O
Avg daily consumption 100,000 gpd
Sodium hypochlorite for disinfection
0^
ON
d.
00
0
m
ro vo
'o o
i i
^f
•z.
o
o
0
o
vT
_w
r- ^> £>
00 cj 5r
'C <5 ON
^ CU d'
2-22
May 1999
-------
'•S
o
W
«
• PM
6C
^
O
C
w
.
s c
o »
£ '«
S a
« w
System Design, Size, and
Other Design Parameters
»g
§ i
•s S
w £
B _.'
o 3^
1 s
^* Q.
•^^ -i?
C ^1
CU <£
11
S
O
VD
T3 £
o..2
»tJ
O ,OJ
o ^
o .=
o" .=2
iO "O
CN t.
i|
s fe
Avg daily consul
Sodium hypochli
CO
-Capacity of 450 gpm
-A: W of 6.77
7.
0
o
£1
0
oo
'^ tt> C^
^ (E O
W)
cs
33
JU
"M
c
1
E
'•?
£
4?
jj
s
ea
i
V)
£
5
c
r"
.1
*J
C
.2
^
o
u*
•a u
U W)
1 system evaluat
Significantly Ion
MSBA
o\
Diffused bubble aeration
— 79 gpm flow rate
- 60 min detention
7
^~*
7
o S?
Q O
*a
S
o
t-^
Avg daily use 4f
Chlorination
Retrofit
0
o
o
II
s
CJ /•—*
||G
ca
Diffused aeration
- supply capacity 60 gpm
- 3 clearwells, 2 w/diffusers supplied t
single blower
7.
c,
o
c"
(N
0
s s
= C-
g;
™
£
V
SB
ty>
___,
C3
^
o
E
•§
•- c _
H ° >
*Jii
00 ^ CO §
00 O J3 "
1
.ll
S E
. SN
'«" £ g» "E £
— g _, 0 ^
S O CN Tt ca
r, ^> — r- >
CN fc- « i-4 o
E S -S o E
o c n ** S
tfa O O —
^ en • • C
c — -2 <^ o
2 i o o 2 ^.
•o c E *• c co
i- O u o O >
5J C 1- .2^0
§ i1! ^11
C ^3 ^. ^^ ^ os
1 ll <1l
^ i"l|ili
«£ E If-S 1
1/1 IM O t*— "? ^ tti
m o o o > -c £:
§3 g J .. s I ' a-
~i£:Ofc'-2*^3u-J>cLto ^aB^-'^
^S C^ ci t,Q ^* ^-* ^*-^ f-s 7^5 -^^i ^^ ?^ r** ^^ t**-
illiiilllilliyiii
'S >»
iirl id! ill,
i nil
C , 4>SS 4>£'1l>*"'O.i_t "Q !>
* a3 T I? ^5 "~* *^-' P tJ IJ
11 !|| i! fli fil
= °-c3-5-= *-=_^sSr3o-s:
H i Q. S S 13-S^wf-o?
o
•s ^
CO fli
en N
U ^
> i
0 O
o o
(N "3-
0 C3
O 0
r» r-
— c<1
b
? 00
^ o\
< ^
2-23
May 1999
-------
^^^
•§
S
J
"£
c
E/
a
*K
j|
"o
a
•O
8
H
0
•-S
g
o
<
jraoval Efficiency Data for
w
•
r<»
IS
cs
42
P"
Conimei
IS 0*
0 .0
G .£5
o> £
•o S2
System Design, Size, an
Other Design Paramete
o> e
> ei
o e
w B
i- «
s u
0 65
GO a,
n ^
a ^ZT
a) ~
|3
C
CO J2
3 *™
1 1
« S.
2 e
c
— o
S, ">
In1 8
vo J
^
°f
CO
Slal tray aeration
- 365 to 450 gpm flow rate
-3.3 to 10 min detention
<
*JP
MO?
5 ON
Q C-
j*j
•s;
m
E
"Q
JC en
E J
y A
> ^
^n -a
To «
> ra
E o
fll •"
« s
•5 o
"° "S
S eO
C -C
<5 "°
3 O
TS °
'S «
S •£ vo
^J ^f ^
< < <
© ©_ ©
oq ^ oj ^.' o\ vo
ON C*M (N ^*— ™" *4—
OG O ^ O ON O
en "Q
« 3 S
-3- en 3
—_ u "g
3 t- .e
CT* J2 in
en ^ .
u. C GO *^
*"* O\ W ^ ^
s . ~ « * "tr
8 « C C 1 .£
>» — O J> C 0
e = 2 .1 S S"
•03 J2 « ® « |
T3 .S 5 . "^ .,-. O
8|^|-SSg3
^2 "S "So .S -.. — t>T to
— 1 IE 1 0 1 | T3
<
tN
+ 1
»
(N
^ ON
g ON
a o
(-
a>
"S> ^s
J^ C/l
S 03
? ^
J2 "^
^ "ti
11
51
1 §
1 1
"o to
C X)
« *H ^"
111 1
VO
ON
II
C
O i*^N
S? t§ ^
ON !E'o.
'5
-------
1
S
*p
e
0
U
en
cu
"ob
_O
"o
S
JS
u
S
c
o
'•£
R
&M
moval Efficiency Data for
CU
tf
S
3
H
en
g
S
s
a
•— • 5**
cs u
O c3
S '«,
tf M
^3 in
5*
a"!
E Is
j?> B
o 's
| u
j3 CU
cJ?O
^ i
u b
U (S3
3 !•
0 «
00 &•
Q
li S»
OS W
"£ ^iT
cu (i>
CZ *u
""
w
b
00
c
'EL
E
3
Q.
;x,
"5.
D.
«
-o • fe "g •
00 CO T3
gl-i*
S-J li
g .S 00 fe
3 .22 > S
^^ " °
'<3 c c^ -2-
*o 'n o S
> J £ n-
Ov
fe
S u
•§>!
3 CO
£ «
"S "Jn
00 3
E .S o
n. K o
— ao co ;"
's ^
•o o
= C-
i
>^
C w
O >v
I1
<]=. 3
.£ w
&o ^
^5 •*— ^^
° .2 3
^ O 00
^> J5 O
A C Cfc^
•O 3 O
g?^ t
> O CO
•< C/D Cu
o
•—
1'.
C
o> ^^
3 *-J
i i u
D
5 *•*•
f 1
Shallow tray aeration1
- supply capacity 25 gpm
- single treatment unit
- air filtered before passing thro
and is vented directly to outside
use
Z
^_^
OO ®
!n S
S
"° fn1
o c^
i
•^
c M
0 >,
11
^- 3
C tn
•o !S fe
sl|
•*„ .2 3
_>» jz g
CO £ c+_
•0 _3 0
> O CO
< oo cu
oo
o
II
11
c
_ 3 J
^E u
S ^
i3
"ee S
^ *"
3 rt
Shallow tray aeration1
- supply capacity 39 gpm
- single treatment unit
- air filtered before passing thro
and is vented directly to outside
use
Z
o1
o
s*0/
li
CN Ci
c
O ^
•S S!
E C-
o.
C 3"
.2 ">
tj O
lil I
OD ^cn T3
o *5 §
ill
c^ jr cj
| 11
"•"* (^
« E «•-
.2 °
< en i£
o
o
tN
^L
c
(D >T^
_ 3 -J
^ GZ ^
^^ r \
CJ
to
n. o.
oo a.
§ !
CN" «
CJ 0 >
en *rs "^
^cM g
>> C^. 4J
*CO *^ CfcM
"O « o
> o |g
o
o
II
g /-N
111
4> C^
"§ > ^ "c
•^ 0 C CO
S p CO
•- 1 » s.
•O e c
Shallow tray aeration1
- supply capacity 25 gpm
- single treatment unit
- system is 8'wide * 15' long an
a water softener (for manganese
shallow tray aerator, polyethylei
used as a clearwell, high lift pur
hydropneumatic tanks
Z
0
oo_
ocf
c
o
•g $3
•O £>
Q ^N
2-25
May 1999
-------
; (Continued)
0
0£
P^^
O
e
"3
£
C
•*•*
>
^ J™
o fiJ
£: u
U TjT«
«!—
Cu
System Design, Size, and
Other Design Parameters
CXS2
eJ C
3 2
o s
e _
||
^j B«
n ^
c ST*
c —3
»•«
9
•s
•2 J£> "3
o cx"3
•n * 3 i
Q..S M £
ao.w h S
Avg daily use 720,000
Sodium hypochlorite d
Part of new groundwat
Pumping & separate cl
o
vo
*«•
11
c
- § Q
OS *- _-* f^ ;
W
re ,
U
35
oo
c X
0 u
Tj JiJ >
a. u o
Avg daily use 145,000)
Gas Chlorination disinf
Towers installed to rerr
10
c
III
Slatted tray towers
- 2 redundant towers w/ total supply
capacity of 800 gpm
- counter current flow
-A:Wof50:l
- 8'xg'x 10' towers with redwood
horizontal slats (not packed media)
<;
Z
0
o
Os
§
"O ^*
in £T>
T3 Cs
O Os
u p
•° -o
c >»
•alt
Avg daily use 40,000 g
Liquid chlorine disinfe
Slat tower first put in f(
removal
eo
o\
CS
JL
c
SO ^ D.
c re
in •—
Slatted tray tower
- supply capacity 435 gpm
- tower is specially built unit w/wood
horizontal slats instead of packed med
*£
Z
0
o
Os
O ^
™O ON
0 £
e
o
•*•*
£
<
V
a
u
es
U
o o
E «g
re o «
"S "So
*^~ yj ^|™*
£ 2 s
22 5J -s
.c .0 re
Sre f^Q
t> ^_/
S en re
» -2 «
01 -o 5
8. 1 §
§ E £
O W ^
o "H. L-
ui -J3 0
Q. -2 -=
cs 5 .2?
cs 2 .n
^_
12 Is « E" «'P
xp (^1 ^ g^ jf •
eNoS t^O60^ —
O^OPn fg ^ O J ^,
lifilflfi
Multi-staged diffused bubble POE
aeration system (Stripper model)
- 3 stages (compartments) in 1 modul
^
Z
o
o
o*
"P
0
o
o
CS
« 0?
.: oo
>. Os
o ^
J 73
2-26
May 1999
-------
(Continued)
en
"S
c
JS
SJ
H
e
^o
"ea
s-
cS
a
•*••
a
Q
-~
"w
s
IS
"w
S
i
cU
CS
H
Comments
CQ ^
g §
S.'J
« fc
System Design, Size, and
Other Design Parameters
a c
u c
g 2
O 8*
cn a.
O j
8> U
s ^
s >•
e — 1
4)
t.
0
Effluent < 200 pCi/L even when air flow rate
restricted due to iron precipitation
accumulated on diffusers
For very high concentrations, monitoring may
be needed to see if higher A:W (more dilution)
or treatment of off gas is needed
Iron oxidation occurs readily - leads to
precipitation and release or gradual decrease
in air flow rate, so iron treatment probably
required
More expensive O&M than GAC, extra pump
needed for repressurization, blowers possibly
needed for ventilation (noise issue)
A
m
in -n ro
cs
IN
t-^
¥i
CN
S
•5
fc
c
c
•2
/-~\
jg
S
^
CO
a
<*?
&.
"5 ^
3
O
O
13
O
U
O
r«1
O
c
o
c.
•a
1
o
(D
o
T3 «
(O SO
i
-
2 .>
?"S
o ^
o o
•3 «
II
1!
II
2-21
May 1999
-------
2.1.2.1 Packed Tower Aeration
Packed tower aeration can achieve very high removals of radon ranging from 90 percent to
higher than 99.9 percent. The design parameters that affect the removal of radon include packing
height, A:W ratio, packing type, and loading rate. Systems installing packed tower aeration should
also consider issues such as pretreatment, additional disinfection requirements, and pump retrofitting
(discussed in Sections 2.1.3 and 2.1.4).
Effect of Packing Height—Packing height is the most critical design parameter for radon
removal (Dixon et al., 1991; Cummins, 1988). Dixon et al. (1991) suggest a minimum packing
height of 10 feet.
Effect of A: W Ratio—The removal of radon is not very sensitive to A:W ratio as long as the
ratio is sufficiently high. Typical A: W ratios for packed towers in drinking water treatment plants
range from 30:1 to 100:1 (AWWA and ASCE, 1998). Cummins (1988) noted radon removals drop
rapidly for A:W ratios lower than 2:1. Khmer et al. (1988) observed that radon removal efficiencies
were similar at A:W ratios of 5:1,10:1, and 20:1, and were only slightly lower for an A: W ratio of
2:1, so increasing the A: W ratio beyond the range of 2:1 to 5:1 impacted removals very little. Kinner
et al. (1988) also noted that using an A: W ratio of 1:1 provided a significantly lower removal. Dixon
et al. (1991) showed that radon removal is not sensitive to A:W ratio. Dixon et al. (1991) reported
radon removals greater than 93 percent for an A:W ratio of 3:1 and a packing height of 10 feet.
Based on Dixon et al. (1991), an A:W ratio of 5:1 should be sufficient for obtaining high radon
removals. More recent research on radon removal efficiencies shows that radon removal is sensitive
to the A:W ratio over a wider range of ratios than reported above. Spencer and Brown (1997)
showed that radon removals tend to level off at an A:W ratio of about 10:1, and that increasing the
A:W above 10:1 has minimal effect on radon removal. According to Spencer and Brown (1997),
the theoretical A:W ratio necessary to remove more than 90 percent of the radon from water is about
5:1, while the practical A:W ratio is 6.5:1 for 90-percent removal. An A:W ratio of 19:1 should be
sufficient to remove nearly 100 percent of the radon, as shown in Figures 2-1 and 2-2.
2-28
May 1999
-------
% Radon Actual
% Radon Theoretical
fcass
0 i 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Actual Air:Water Ratio From Lowiy Engineering, 1993
Figure 2-1. Practical Radon Removal with Increasing Air:Water Ratio for PTA
(Source: Spencer and Brown, 1997)
Ideal % Radon
Ideal % Carbon Dioxide
Ideal A:W Ratio
Figure 2-2. Theoretical Radon Removal with Increasing Air:Water Ratio for PTA
(Source: Spencer and Brown, 1997)
2-29 May 1999
-------
Effect of Packing Type—Kinner et al. (1988) observed that radon removals with saddle
packing were slightly lower than removals achieved with pall rings; however, radon removals were
above 90 percent with either packing type for a similar packing height and A:W ratio.
Effect of Loading Rate—Dixon et al. (1991) reported high removals of radon at loading
rates of 50 gpm/sq ft. However, in order to prevent potential flooding, a loading rate of 25-30
gpm/sq ft may be a practical limitation.
Other Considerations—Other considerations such as pretreatmenl: for iron and manganese
removal, additional disinfection requirements, and pump retrofitting are relevant to ground water
systems that currently do not have any treatment in place. These factors are discussed in Sections
2.1.3 and 2.1.4. In addition, if the treatment facilities are located indoors, dehumidification and
ventilation needs must be addressed.
2.1.2.2 Diffused Bubble Aeration
Diffused bubble aeration (DBA) can achieve very high removals of radon ranging from 71
to >99 percent, with removals often greater than 90 percent. The design parameters that have been
studied for their effect on the removal of radon for this technology include the A:W ratio and flow
rate. Systems installing DBA should also give consideration to pretreatment, disinfection, and pump
retrofitting requirements.
The removal of radon by MSBA does not vary significantly with A:W ratio and flow rate
(Dixon et al., 1991). A slight increase in radon removal occurred by increasing A:W ratio or
decreasing the flow rate. According to Dixon et al. (1991), designs of MSBA permit a maximum
flow of 800 gpm for radon removals greater than 95 percent, and 1,800 gpm for removals of less
than 85 percent. Industry brochures claim treatment capacity of more than 1,000 gprn (Lowry
Engineering, 1989).
2-30
May 1999
-------
2.1.2.3 Spray Aeration
Kinner et al. conducted pilot tests using spray aeration and reported the following removal
efficiencies (U.S. EPA, 1988), presented in Table 2-4.
Table 2-4. Radon Removal for Spray Aeration Pilot Tests
Detention Time (hr)
9
12
Percent Removal of Radon
Decay
7
9
Total
63-73
62-65
Dixon et al. (1991) noted that variations in A:W ratio had a negligible effect on radon
removal efficiency. Dixon et al. achieved 77-percent radon, removal rates using a baffled steel tank
with a flow of 70 gpm, an A:W ratio of 6:1, and a detention time of 20 minutes. For the same
system with a flow of 50 gpm, Dixon et al. achieved radon removals of 83-91 percent for A:W ratios'
ranging from 3:1 to 17:1 (removal was 88 percent at A:W ratio of 6:1). Further investigation is
needed to assess how removal efficiency is affected by design parameters, including water drop size,
spray height, and ventilation.
2.1.2.4 Tray Aeration
Radon removals across tray aerators at three different plant sites are reported in a 1987 report
by the American Water Works Service Company, Inc. Each of the plants use tray aerators for the
oxidation of iron and manganese, with one site using a slat tray cascade aerator. The radon removal
efficiencies of the aerators were between 77 and 91 percent. The site with the cascade aerator
achieved 80- to 81 -percent radon removal (U.S. EPA, 1988).
Other studies have shown radon removals of 86 to 94 percent for slat tray aerators with flow
rates of 365-450 gpm and detention times ranging from 3.3-10 minutes (Drago, 1998). Brown
(1995) reported radon removals of 89.8 to 93.8 percent for two pilot test systems. One pilot system
contained an inducted draft aerator that was tested at A: W ratios ranging from 3:1 to 6:1; the radon
removals were similar for all three A:W ratios. The second system was a forced draft aerator with
an A:W ratio of 3:1 and a hydraulic loading rate of 17.5 gpm/sf (Brown, 1995).
2-31
May 1999
-------
For shallow tray aerators, the key design parameters are residence time and A:W ratio.
Radon removals of 90 percent and higher are expected using a residence time of 30 to 60 seconds
and an A:W ratio between 5:1 and 15:1. (Hodsdon, 1993 draft)
2.1.2.5 Point of Entry Devices
Studies and pilot tests of POE systems have focused on diffused bubble and spray aeration
devices, since packed towers have generally been considered impractical for home use because of
their size and cost. Radon removals from 95 percent to more than 99 percent are reported for
diffused bubble aeration devices installed at the point of entry to homes (Lowry et al., 1984; Kinner
et al., 1990; Kinner, et al., 1993). Kinner et al. (1993) note that the diffused bubble and bubble plate
aerator POE units tested had high A:W ratios (which is common since POE units are generally
overdesigned) and therefore the units should handle changes in influent radon activity and the water
flow rate without a significant increase in effluent radon activity. For spray aeration systems
installed in homes and tested, radon removals have ranged from 82 to 93 percent (Rost, 1981).
2.1.2.6 Comparison of Technologies
There is little difference between PTA and DBA in terms of radon removal. Both PTA and
DBA achieve high removals of radon and are available commercially. Although shallow tray
aerators can achieve radon removals of greater than 90 percent, removals are generally lower than
those obtained from PTA and DBA. The maintenance requirements for both PTA and DBA are low.
While DBA is favored for aesthetic reasons, PTA is favored for large flows for both practical and
economic considerations. MSB A is also limited in treating larger flows. Dixon et al. (1991) and
industry brochures show an upper limit capacity of 800 to 1,000 gpm for currently available DBA
units based on practical considerations. Shallow tray aerators also offer aesthetic advantages since
they are compact. The units are generally modular, so they can generally be used with systems of
all sizes since multiple units can be used together to increase capacity.
The costs of PTA and MSB A are generally comparable for flows below 1 mgd, while the
process costs of MSBA are higher than PTA for larger flows (based on Cummins, 1992; Lowry,
2-32
May 1999
-------
1990). Construction, engineering, operations and maintenance, and other indirect costs play a major
role in determining the feasibility of a treatment technology. Building costs for compact aerators,
like MSB A, should be much lower than for PTA. Other considerations such as pretreatment for iron
and manganese, disinfection, pump retrofitting, and obtaining air permits are site-specific and may
be required for either technology. For tray aerators, copper sulfate or chlorine may need to be added
to control the growth of slime and algae on the trays.
2.1.3 Pretreatment
Some-radon removal systems may require pretreatment, particularly treatment for iron and
manganese, to reduce operational problems associated with aeration.
2.1.3.1 Iron and Manganese
Iron (Fe) and manganese (Mn) in influent water can precipitate when a water supply is
aerated. Precipitation can foul packing in aeration units, thus decreasing the efficiency of these
processes.
Existing ground water systems that will be required, to reduce radon may not need additional
treatment for iron and manganese since water systems normally treat their water to reduce iron and
manganese levels below their secondary MCLs of 0.3 mg/L and 0.05 mg/L, respectively. This is
confirmed by the results of an analysis of NIRS data (WMA, 1992) which correlated the occurrence
of radon-222 with combined Fe and Mn levels. These results are summarized in Table 2-5.
Table 2-5. Correlation of Occurrence of Fe and Mn with Radon
Total No. of Systems with Rn-222 >300 pCi/L
347
Percentage of Systems with Combined Fe&Mn
>0.3 mg/L
14.7
>1 mg/L
3.5
> 2.5 mg/L
0.3
As detailed in the subsequent sections, suggested treatment techniques that can be applied
to avoid fouling in aeration units depend on the concentrations of iron and manganese in the influent
water to aeration units, and are presented in Table 2-6.
2-33
May 1999
-------
Table 2-6. Treatment Levels for Iron and Manganese
Combined Fe & Mn (mg/L)
<1
>1
Suggested Treatment
Addition of a Sequestrant
Oxidation/filtration or greensand filtration
Sequestration
Sequestration is a treatment method by which iron and manganese are prevented from
causing objectional turbidity and color without actually removing iron or manganese from the treated
water. Typical sequestering agents are sodium silicate and polyphosphate and other phosphate-
containing compounds. Sequestrant chemicals are normally used simultaneouly with chlorine. For
manganese-containing waters, polyphosphate is a more effective Sequestrant than sodium silicate
(Robinson et al., 1990).
The Ten State Standards (the Standards) recommend the use of polyphosphates only when
iron, manganese, or a combination of both metals does not exceed 1 mg/L (GLUMRB, 1997). The
Standards also limit the use of polyphosphates to 10 mg/L as phosphate. Because phosphates may
enhance fouling of the distribution system, the Standards require a disinfectant residual in .the
distribution network. The Standards also require that polyphosphate stocks be disinfected and
maintain about 10 ppm of free chlorine residual for solutions with pH above 2. The Standards
strongly recommend the addition of chlorine or chlorine dioxide with or before the addition of the
Sequestrant agent sodium silicate. Maintaining a residual disinfectant in the distribution network is
also strongly recommended by the Standards to avoid biological breakdown of the sequestered iron
(GLUMRB, 1997).
States may have different limits for using sequestration. For example, in New Jersey
sequestering limits are 0.6 mg/L for iron and 0.1 mg/L for manganese (Dixon, 1999). At greater
concentrations of iron and manganese, removal treatment is required.
High doses of sequestrants (>2.5 mg/L) may cause turbidity particularly in very hard waters.
Kleuh and Robinson, as cited in Malcolm Pirnie (1992), found that 1 mg/L of polyphosphate is
2-34
May 1999
-------
sufficient to sequester iron (1-2 mg/L). However, the turbidity level in hard waters (100 mg/L
calcium hardness) treated with 2.5-10 mg/L of polyphosphate was above 1 NTU. Robinson and
Reed (1990), as cited in Malcolm Pirnie (1992), reported that 1 mg/L iron can be sequestered
without excessive silicate dosage even when high levels of hardness are present.
Greensand Filtration
Greensand filtration consists of a conventional filter box using greensand instead of sand or
anthracite as the principal filtration medium (U.S. EPA. 1993 a). Manganese greensand filtration has
been successfully used for iron and manganese treatment for many years. Manganese greensand6
media is prepared by treating glauconite7, a natural zeolite, with manganous sulfate and potassium
permanganate to coat'the media with manganese oxide. This process gives the media adsorptive
characteristics, which allows for the removal of soluble materials through adsorption, as well as
filtration of insoluble materials (U.S. EPA, 1993a). The manganese oxide acts as a catalyst in the
filtration process to assist in the complete oxidation of iron and manganese. Potassium permanganate
is typically added to water ahead of greensand filtration. This serves to oxidize contaminants to
insoluble forms for subsequent filtration, provides disinfection, and restores adsorptive capacity to
the media. Greensand filtration is particularly advantageous when using potassium permanganate
to oxidize iron and manganese. If reasonable dosages of potassium permanganate are used,
greensand filtration will remove any excess potassium permanganate from water, preventing pinkish
water from entering the distribution system (U.S. EPA, 1993a).
Iron and manganese are oxidized and converted to iron and manganese oxides before being
filtered out by the greensand media. The catalytic properties of the manganese greensand assist in
removing the iron and manganese. A short reaction time before filtration is desirable to avoid large
growth of particle size which may clog the surface area of the filter media. To help prevent clogging
of the greensand media, the greensand can be capped with anthracite. This prevents microbial
growth on the greensand since organic matter is removed on the anthracite.
6 greensand: a sedimentary' deposit consisting of glauconite mingled with sand and clay
7 glauconite: iron potassium-silicate mineral
2-35
May 1999
-------
The filters can be regenerated by two techniques, continuous regeneration and intermittent
regeneration. Intermittent regeneration intermittently passes potassium permanganate through the
greensand bed which usually requires about one hour. This includes washing and rinsing of the
media. Intermittent regeneration allows for a higher flow rate and longer runs between
regenerations. Intermittent regeneration is recommended when removal of only manganese or
manganese with a small amount of iron is needed (U.S. EPA, 1993a). Continuous regeneration
involves the constant feeding of potassium permanganate solution and other oxidizing chemicals to
the raw water ahead of the filters. The filters are periodically taken offline for conventional washes
that require approximately 20 minutes. If excess potassium permanganate is fed ahead of the filters,
the filter media will pick up the excess and minimize leakage into the effluent, up to a point.
Potassium permanganate breakthrough may occur. Continuous regeneration is recommended for
water where iron removal is the main objective with or without the presence of manganese (U.S.
EPA, 1993a).
For removal of iron and manganese, greensand filtration is very sensitive to pH. When the
pH is lower than 7.1, deterioration of the bed occurs and the addition of a pH adjuster to the
incoming stream is required. When pH is between 7.5 and 9, optimum conditions occur and the
oxidation reaction with potassium permanganate is complete and rapid. The rate of reaction is very
important as this process is usually employed in installations with direct pumping where the contact
time is very short (i.e., from the well through the filters to the distribution system) (U.S. EPA,
1993a).
The entire process can be a closed process which does not require the water to be exposed
to the air. Major components of the process include filtration with greensand backwash facilities,
and potassium permanganate feed systems.
2.1.3.2 Other Factors
In addition to Fe and Mn, other factors that can affect fouling of aerators are microbial
growth, pH, pE, and the hardness of the water. A potential but generally labor-intensive alternative
to pretreatment is periodic replacement of packing, or cleaning of packing in aeration units (either
2-36
May 1999
-------
by removing the packing for cleaning and replacing it with spare packing, or by cleaning the packing
in place with acid, chlorine, or pressure washing).
For hard water with a high CO2 concentration, precipitation of calcium carbonate can occur
when aeration reduces the CO2 concentration, also resulting in an increased pH level. As stated in
NRC (1998), "In a study of the aeration units used for VOC treatment, the American Water Works
Association (1991) reported that the effect of CO2 removal, with the greater stability of CaCO3 at
the higher pH, negated the effect of the increased oxygen concentration hi water. There was no
increase in the corrosivity of water." NRC (1998) also reported that a very small water supply
system in Colorado found that removing radon through aeration eliminated the need to add lime to
prevent corrosion. In addition, a small system in New Hampshire experienced a decrease in
corrosivity and a reduction in the lead and copper measured in the drinking water as a result of
aeration. Although the effects of aeration can include decreased corrosivity of the water, it may not
eliminate the need to add corrosion inhibitors. (NRC, 1998)
2.1.4 Post Treatment
2.1.4.1 Disinfection Following Aeration
During the aeration process, atmospheric air is blown into the water supply. The air blowers
are equipped with influent screens to prevent any large particulate matter from entering the water
supply. However, airborne bacteria or viruses are usually introduced into the supply. For ground
water systems, this is likely to be the only contact with the air prior to the water reaching consumers.
The exposure of a clean ground water supply to air increases the risk of microbiological
contamination. In keeping with good engineering practice, ground water supplies that are aerated
should be disinfected, even if the ground water supply may otherwise be classified as naturally
disinfected. Based on this approach, if a ground water system currently does not disinfect and it adds
aeration for radon removal, it would also need to install disinfection.
2-37
May 1999
-------
The Ten State Standards (GLUMRB, 1997) state that the following features be provided when
water is subjected to aeration:
Disinfection application points both ahead of and after the tower to control biological
growth.
Disinfection and adequate contact time after the water has passed through the tower and
prior to the distribution system.
Methods for disinfection include:
• Chlorination
• Ultraviolet light treatment
• Ozonation
• Chloramine addition
• Chlorine dioxide addition
• Mixed oxidant addition (anodic disinfection)
These methods are described in the sections below.
The Ten State Standards (GLUMRB, 1997) do not specify minimum contact times for
i
disinfection, but specify that aerated water must receive chlorination as minimum additional
treatment. It is difficult to estimate the extent of microbiological contamination that would result
due to aeration. Insects such as chironomus fly may lay eggs in the stagnant portion of a tray aerator
(U.S. EPA, 199la). Also bacteria and viruses may be introduced into the water through the blown
air. As a conservative estimate, affected groundwater systems that would install aeration for radon
removal should provide adequate disinfection for 4-log (99.99%) inactivation of virus. Water
systems with little or no distribution network and with minimal likelihood of cross-contamination
do not need to provide residual disinfection. For other systems that need to install disinfection and
provide a residual, UV and ozone should not be used.
2-38
May 1999
-------
Chlorination
Chlorination, in gaseous, solid, and liquid-feed forms, is the most widely used disinfectant
at public water supplies (U.S. EPA, 1997). Chlorine can inactivate bacteria, giardia, and viruses.
The level of inactivation is determined by the chlorine contact time, pH, temperature, and free
chlorine residual. Chlorine feed systems can be either direct solution/dry chemical, gas to solution,
or direct gas injection. Direct solution or dry chemical feed systems use commercially-available
sodium or calcium hypochlorite to deliver the necessary chlorine dosage. Gas to solution systems
inject compressed chlorine gas from cylinders into feedwater to form a chlorine solution which is
then added to the process stream. Direct gas injection into the process stream is possible but is not
a common procedure.
The size of the clearwell depends on the inactivation required, temperature, pH, chlorine
residual, and requirements for efficient pumping. The inactivation concentration time (CT)8 value
from the Guidance Manual for Compliance With the Filtration and Disinfection Requirements for
Public Water Systems Using Surface Water Sources (U.S. EPA, 1991) for chlorine disinfection to
a level of 4-log (99.99%) inactivation of virus at a temperature of 10°C, and a pH between 6 and 9,
is 6 mg-min/L. Assuming a chlorine dose of 1.5 mg/L, the contact time for this conservative case
is 4 minutes. Assuming that the theoretical contact time is 70 percent of practical residence time
(due to inadequate mixing or short circuiting in the cleanvell) the clearwell should be sized to
provide a contact time of about 5.7 minutes. For a chlorine dose of 1 mg/L, the theoretical contact
time would be 6 minutes, so the practical contact time would need to be about 8.6 minutes.
For ground waters with low chlorine demand and a short distribution system, 3-log
inactivation of viruses may be sufficient. A chlorine dose of 1 mg/L and a contact time of 4 minutes,
at a temperature of 10 °C, and a pH range of 6 to 9, will provide the needed CT value of 4 mg-min/L.
To account for short circuiting and inadequate mixing, a practical residence time of 5.7 minutes in
8 CT is the product of the residual disinfectant concentration in mg/L © and the disinfectant contact time in
minutes (T). Disinfectant contact time is the time needed for the water being treated to flow from the point of
disinfectant application to a point before or at the first customer during peak hourly flow (U.S. EPA, 1997).
2-39
May 1999
-------
clearwell will be adequate for disinfection. On the basis of disinfection considerations alone,
systems would provide a theoretical residence time in the range of 5 to 10 minutes. The clearwell
in a PTA system provides*a theoretical residence time of about 14 minutes for very small systems
and 7 minutes for systems treating 1 mgd (U.S. EPA, 1984). Minor improvements to the PTA
clearwell, such as adding baffles, will improve mixing and disinfection conditions in the PTA
clearwell.
Ultraviolet Light Treatment
Ultraviolet light treatment in drinking water involves the direct exposure of the water stream
to ultraviolet light. Exposure to the ultraviolet light damages nucleic acids and changes their mode
of action in microorganisms, thus preventing microorganisms from propagating or remaining active.
Ultraviolet light is generated by striking an electric arc through mercury vapor. The inactivation of
microorganisms in drinking water by means of ultraviolet light is a function of the intensity of the
radiation, proper wavelength, exposure time, water quality, flow rate, type and source of the
microorganisms (natural or culture), and the distance from the light source to the targeted
microorganisms. The intensity is measured in milliwatts per square centimeter (niwVcm2) and time
is measured hi seconds(s), resulting in a dose measurement in milliwatts-seconds per square
centimeter (mWs/cm2).
At sufficient intensity and appropriate wavelength and exposure time, ultraviolet light is an
effective disinfection agent for drinking water. Since ultraviolet treatment is more suitable for clean
water sources with little suspended matter, water should be pretreated (e.g.., for iron removal) before
reaching the ultraviolet disinfection unit (U.S. EPA, 1996). The scientific literature shows that, in
general, under laboratory conditions and using distilled water, a 3-log reduction of bacteria is
achieved using an ultraviolet light dose of 30 mWs/cm2, a 4-log reduction of viruses is achieved at
a dose of 50 mWs/cm2, and a minimum 3-log reduction of bacterial spores is achieved at a dose of
60 mWs/crrr (without a safety factor). Studies conducted on ground water show that a 4-log
reduction of bacteriophage MS-2 (a surrogate test organism) is achieved at about 90 mWs/cm2
(Snicer et al. 1996 as cited in U.S. EPA, 1996). Based on their study of the practical experience of
2-40
May 1999
-------
ultraviolet disinfection in the Netherlands, Kruithof et al. (1992) recommended the use of ultraviolet
light (without use of a secondary disinfectant) for disinfecting drinking water from all sources of
water provided that two conditions are met: (1) the water has to be low in biodegradable compounds
so regrowth would not occur; and (2) the distribution network does not need any additional
protection (no biofilm growth is likely and no cross contamination is likely). For ground water
systems with little or no distribution network, UV disinfection is very feasible.
Given the inconsistent sensitivity of microorganismsto UV light and assuming a safety factor
of 1.5 for dose requirements, a dose of 90 mWs/cm2 should produce a minimum 3-log reduction of
pathogenic (and indicator) microorganisms (bacteria and viruses) that are in ground waters. In
addition, a dose of 140 mWs/cm2 should produce a minimum of 4-log reduction of microorganisms
that are in ground water sources.
Ultraviolet systems come in two types, closed and open, with closed systems more
commonly used in potable and sterile water applications and thus used in this analysis. The design
and operation of an ultraviolet system needs to consider equipment operational factors (e.g., lamp
output, fouling of unit surfaces), water quality factors (e.g., rnicrobial and chemical characteristics),
and hydraulic design elements (e.g., dispersion, flow rate). For small systems, multiple modular
units are recommended and should be easy to install and operate. To be effective, ultraviolet light
treatment must be applied after aeration. Information collected from case studies has shown that
generally, ultraviolet systems require little supervision and users are satisfied with the performance
of the equipment. For the case studies, the main factors cited for choosing ultraviolet technology
over traditional disinfection technologies include minirnum service time, low operation and
maintenance costs, and the absence of a chemical smell and taste in finished water (U.S. EPA, 1996).
Ozonation
Ozone is a powerful disinfectant and has been used primarily at large drinking water
treatment facilities, but is also applicable to small systems. Ozone is generated on-site from air or
oxygen and introduced into water using mass transfer equipment. Ozone reacts quickly with organic
2-41
May 1999
-------
and inorganic substances in the water and auto decomposes, so its residual is much less stable than
that of other disinfectants and dissipates rapidly. To prevent health risks, any excess dissolved or
entrained ozone must be destroyed or removed before the water enters the distribution system.
Every effort must be made to avoid exposure to plant personnel. This may require the use of a
secondary disinfectant, (U.S. EPA, 1993b).
Design criteria include ozone residual, competing ozone demands, and minimum contact
time. One study showed complete inactivation of bacteriophage MS-2 and Hepatitis A virus for a
pH range of 6-10, temperature of 3-10°C, and ozone residuals of 0.3-2.0 mg/L. For a short contact
time of 5 seconds, inactivations of >3.9-log to 6-log occurred. Other studies have shown inactivation
ofGiardia muris and enteroviruses of 3-log and 4-log removals, respectively, for 5 minutes contact
tune and ozone residuals of 0.5-0.6 mg/L. (U.S. EPA, 1997)
Chloramine Addition
Chloramines are formed when chlorine is added to water containing ammonia and the
ammonia then reacts with hypochlorous acid. The three chloramine species that can form are
monochloramine, dichloramine, and trichloramine (or nitrogen trichloride). The amount of each
chloramine formed depends on pH, temperature, time, and the initial chlorine to ammonia ratio.
Some of the characteristics of chloramine disinfection include a long residual effect, low production
of disinfection byproducts, the need for careful management of the ratio of chlorine to ammonia to
prevent odor and taste problems and biological instability in the water, and longer inactivation
contact times than chlorine and ozone (since chloramines are less potent). Chloramines can be used
as a primary disinfectant or as a secondary disinfectant, but their use as a primary disinfectant is
limited by the long contact times necessary for adequate disinfection. The longer inactivation time
translates into larger contact basins than those required for chlorination, chlorine dioxide, ozonation,
or UV treatment. The CT value of 4-log removal of viruses using chloramine at 10°C is
1,491 mg-min/L (U.S. EPA, 1991). Information on the use of chloramination at small systems is
limited. Theoretically, chlorination may be applicable to small systems. However, the relative
stability of chloramines and their long residual effect in distribution systems makes them effective
as a secondary disinfectant.
2-42
May 1999
-------
Chlorine Dioxide Addition
Chlorine dioxide can be used as a primary disinfectant, but is not applicable as a secondary
disinfectant because its reactivity means it is rapidly consumed and there is little or no residual in
the distribution system. Because of its instability and explosivity, chlorine dioxide cannot be
transported and must be generated at the application site, usually by chlorinating aqueous sodium
chlorite.
Since chlorine dioxide has more than 2.5 times the oxidizing capacity of chlorine, its CT
requirements for Giardia cysts are lower than the CT requirements for free chlorine. However, for
viral inactivation, chlorine dioxide has higher CT requirements than those for free chlorine. For 3-
log and 4-log inactivation of viruses at 10°C, CT values are 12.8 and 25.1 mg-min/L, respectively,
for pH between 6.0 and 9.0. CT values for the inactivation of viruses by chlorine dioxide are
affected by temperature, but are independent of changes in pH over a range of 6.0 to 9.0 (U.S. EPA,
1993b).
As with chloramination, information on the use of chlorine dioxide disinfection at small
drinking water treatment facilities is limited. This is because chlorine dioxide is an expensive
technology requiring skilled labor, more careful handling than other forms of chlorine, and has high
monitoring requirements. In addition, the production of chlorite and chlorate (disinfection
byproducts or DBFs) can be problematic. Taste and odor can develop if the chlorite reacts with free
chlorine used for residual disinfection.
Mixed Oxidant Addition (Anodic Disinfection)
For mixed-oxidant disinfection, a solution containing oxidants (mostly free chlorine, but it
may also contain ozone, chlorine dioxide, hypochlorite ion, and hypochlorous acid), is generated on-
site by sending an electric current through a continuous-flow salt solution. This multiple oxidant
solution is then added to the water for treatment. Mixed-oxidant disinfection can be more effective
than a method using one oxidant since each oxidant provides different advantages. Advantages
include broader ranges of conditions where they are effective and different residual effects,
2-43
May 1999
-------
production of potentially fewer disinfection byproducts, and combination effects (the presence of
one disinfectant can make another more effective). This method may be particularly useful at sites
where chemical supply is not reliable or where local codes do not permit the transportation of
chlorine. Some of the limitations of mixed-oxidant disinfection include the difficulty of determining
the relative ratios of each oxidant to include in the mixture, inability to significantly reduce color,
questionable reliability for removing turbidity, and the limited data on the contact time necessary
for adequate disinfection. Although it is expected that mixed-oxidant disinfection requires a shorter
contact time than other technologies, EPA recommends that chlorine CT values be used until more
data specific to mixed-oxidant disinfection are available. (Mixed oxidants have been used for
disinfection and other full scale water treatment applicants.) Results from laboratory and pilot tests
indicate that mixed-oxidant disinfection can achieve 3-log to 6-log inactivation for parasitic
microorganisms at four hours contact time and 5 ppm residual, and 3-log to 4-log inactivation for
Giardia (U.S. EPA, 1997).
2.1.4.2 Water Pump Modifications
When a ground water system installs a process that is open to the atmosphere, pumping
modifications and additions may be necessary. Existing ground water systems normally provide
minimal treatment—usually only disinfection—before pumping directly under pressure to the
distribution system. If a process open to the atmosphere, such as aeration, is installed, the affected
water system has the following options: (A) throttle existing well pumps; (B) restage existing well
pumps; and © replace existing well pumps with pumps providing a lower head. In any case, finished
water pumping will be needed to boost the pressure before distribution since most aeration
technologies need to be operated at atmospheric pressure for radon to be released to the air. Some
small water systems that choose to replace well pumps could use the old well pumps for pumping
from the clearwell to the distribution system. Therefore, processes such as aeration will need both
raw and finished water pumping, but often do not require more than one additional pump.
2.7.5 Off-Gas Emissions
2.1.5.1 Worker Radiation Exposure
Personnel in water treatment facilities that use PTA or other aeration techniques for radon
I
removal may be exposed to higher-than-background radiation levels. This is because radon is
2-44 May 1999
-------
heavier than air and can build up in areas with stagnant air or in poorly ventilated facilities that house
PTA or open DBA treatment units. Water treatment facilities should set work practices and
monitoring to attain exposure levels as low as reasonably achievable in the work place (U.S. EPA,
1994). For example, areas immediately surrounding or immediately downwind of a PTA should be
well-ventilated. Additionally, the water treatment plant buildings and areas where workers spend
their time should be adequately ventilated all year round.
2.1.5.2 Air Emissions
Off-gas is not expected to be a regulatory or engineering concern for typical systems. There
are no Federal regulations for off-gas emissions of radon from drinking water treatment plants. One
source in the reviewed literature (Martin and Myers, 1992) cited a California State emissions
standard for radon, Hut personnel at the California Department of Health Services (DOHS) said that
California has no radon emission standards for water treatment facilities (Quinton, 1998). Martin
and Myers (1992) stated that DOHS regulates the discharge of radionuclides and has a radon
discharge standard that sets a 3 pCi/L concentration limit for radon "at the boundary of the controlled
area" (CaliforniaDOHS Regulations, Title 17, Section 30269). Using modeling, Martins concluded
that compliance with a 3 pCi/L standard was not difficult to attain. For water with an influent radon
concentration of 350 pCi/L and aeration treatment at an A:W ratio of 20:1, the stack discharge is
estimated at 17.5 pCi/L. EPA's SCREEN model (which uses highly conservative default
meteorology) predicts that this discharge at 60 ft above ground would dilute radon to less than 0.01
pCi/L at ground level (Martin and Myers, 1992).
No other potential State regulations setting limits for radon off-gas emissions were identified.
The Technology Transfer Handbook: Management of Water Treatment Plant Residuals (U.S. EPA,
1996a) notes that some States have treatment requirements for radon off-gas and some States limit
gas phase emissions from stripping processes and reactivation systems using GAC, but States are
not identified and requirements are not given. In some cases, local and/or State restrictions could
pose extra engineering or permitting requirements. Radon off-gas emissions regulations were being
considered by the South Coast Air Quality Management District (SCAQMD) of California in 1991,
but their development has been hindered by the unavailability of a unit risk factor for radon
concentrations in ambient air (Balagopalan, 1998).
2-45 May 1999
-------
Although a human health risk may be created by radon emissions from aerators treating
drinking water, this risk is far less (about 2 to 4 orders of magnitude smaller) than the risks from
radon in homes due to untreated drinking water (U.S. EPA, 1994; U.S. EPA, 1993). In addition, by
considering the location of off-gas emissions sources (e.g., on a roof) when designing an aeration
system, the risks posed by human contact with these emissions can be minimized (U.S. EPA, 1993).
I !
In a field evaluation of packed tower aeration using a 1-ft diameter stainless steel packed
tower with 12 ft of plastic media packing, Kinner et al. (1990) reported off-gas emissions ranging
from 2,410 to 21,200 pCi/L, which is 4 to 5 orders of magnitude higher than the average outdoor
' ' '
level of radon (0.2 pCi/L). The study used A:W ratios varying from 20:1 to 1:1 and water flow
conditions of half and full. Kinner et al. (1990) noted that although the fate of the plume was not
i ! !
studied, it can likely be sufficiently diluted by proper venting (e.g., constructing the release point
high enough off the ground). In an associated study using a diffused bubble aeration system, Kinner
I > !
et al. (1990) found that the off-gas radon activity increased (4,167 to 18,600 pCi/L) as the A:W ratio
decreased (less dilution), similar to findings with the packed tower system.
2.1.6 Treatability/Case Studies
2.1.6.1 Packed-Tower Aeration
PTA Blairsville, Georgia
EPA-TSD conducted a field evaluation of radon removal by packed tower aeration from a
ground water supply in Blairsville, GA (Cummins, 1988). The key design parameters were as
follows:
Diameter
Tower Height
Packing Height
Packing Type
A:W Ratio Range
2 feet
24 feet
18 feet
1" plastic saddles
0.05:1 to 15:1
2-46
May 1999
-------
The influent radon concentrations ranged from 4,000 to 6,200 pCi/L. Eight different runs
were conducted with different air-to-water ratios. A profile of radon removal with packing height
was obtained by sampling at different heights on the column.
The following observations were made:
» Radon removal ranged from 99.84 percent for an A:W ratio of 15:1 to only 17 percent
for an A:W ratio of 0.05:1.
a The mass transfer coefficient decreased from 0.0225 sec'1 at an A:W ratio of 15:1, to
0.0027 sec'1 at an A:W ratio of 0.05:1.
Dixon and Lee Case Studies
Dixon and Lee (1988) performed studies for the removal of radon from ground water by
packed-tower aeration. The key design parameters for the packed tower were as follows:
Diameter
Tower Height
Packing Height
A:W Ratio
4.5 feet
16 feet
14 feet
50:1
Although this column was originally designed to remove 70 percent of the VOCs present in the
source water, the column removed greater than 95 percent of the radon present as shown in
Table 2-7.
2-47
May 1999
-------
Table 2-7. Packed-Tower Aeration
Run
1
2
3
4
Type of Sample
Raw water
Aerator effluent
Raw water
Aerator effluent
Raw water
Aerator effluent
Raw water
Aerator effluent
Mean Radon
Concentration
pCI/L
783
20
649
21
646
24 .
646
34
Reduction Percent
97
97
96
95
Kinner et al. Pilot Studies
Kinner et al. (1988) performed pilot scale studies for the removal of radon from water by
packed tower aeration. The stainless steel packed tower aerator was 18 feet in height and 1 foot in
diameter. Glitsch mini-rings and saddles, and Koch pall rings packing media were tested. The
overall packing height was 12.3 feet for both of the Glitsch media and 11.8 feet for the Koch media.
Several runs were conducted at various A:W ratios, high flow rates between 4.25 and 17 gpm, .and
low flow rates between 0.1 and 7 gpm. The packed tower removed 97 to 99 percent when mini or
pall rings media were used, and 90 to 94 percent when saddles were used. Temperature changes had
little effect on radon removal efficiencies.
Dixon et al. Case Study in Pennsylvania
Packed-tower aeration was one of three aeration technologies evaluated for the removal of
radon from a ground water supply in Pennsylvania (Dixon et al., 1991). The design and operating
parameters of the packed tower were as follows:
Diameter
Tower Height
Packing Height
Packing Type
23 inches
16 feet
10 feet
2" Polypropylene Tripacks
2-48
May 1999
-------
Loading Rates 50 and 25.5 gpm/sq ft
A:WRatios 3:1, 8:1, 12:1, and 30:1
The influent radon levels ranged from 1,700 to 2,700 pCi/L. Removal of radon ranged
between 93 and 98 percent. The following observations were made from the study:
Removals of radon were sensitive to packing height. The increase in removal efficiency
by increasing packing height from 5 to 10 feet ranged from 17 to 22 percent at a loading
rate of 50 gpm/sf, and from 14 to 16 percent at a loading rate of 25.5 gpm/sf.
Removals of radon were insensitive to the A:W ratios tested in the study. The removal
efficiencies radon ranged from 93 to 98 percent for A: W ratios ranging from 3:1 to 30:1,
with a packing height of 10 feet.
The removals were marginally improved (1- to 4-percent increase in removal) by
decreasing the loading rate from 50 gpm/sf to 25.5 gpm/sf.
A strong correlation was noted between the concentrations of carbon dioxide and radon
in raw and treated water. Carbon dioxide was suggested as a potential surrogate
parameter for assessing radon removal efficiency.
Hodsdon Case Studies
Data collected by A.E. Hodsdon (Hodsdon, 1993 draft) as part of a unpublished survey for
the American Water Works Association (AWWA) showed removal efficiencies ranging from 78.6
percent to more than 99 percent for packed tower aerators, with three of the facilities reporting 99
percent or greater radon removals. The survey included eight operating packed tower aerators at
drinking water treatment facilities of various sizes (supply capacities ranging from 9-450 gpm) in
Wisconsin (3 sites), Pennsylvania and New Hampshire (2 sites each), and Colorado (1 site). Influent
and effluent radon concentrations varied from 1,208 to 9,000 pCi/L and <15-300 pCi/L, respectively.
The draft survey report did not provide information on ainwater ratios or most tower heights (the
two tower heights noted were both 33 feet). Five packed towers had counter current flow and three
had concurrent flow, though this did not appear to be a significant factor. All of the facilities
provided disinfection but did not install it as a result of radon treatment. Two facilities have ion
exchange units before the tower to reduce iron and hardness and had this pretreatment in place prior
2-49
May 1999
-------
to installing radon removal equipment, but indicated that they installed post-tower bag filters to
remove precipitated iron as a result of adding radon treatment.
i ' i
Valley County Water District
i
Secondary Effects of Packed Tower Aeration
The Valley County Water District(VCWD) of San Gabriel Valley, California, provides water
to the city of Baldwin Park and a portion of Irwindale about 30 miles east of Los Angeles. The water
is drawn from ground water sources at 10 wells. A monitoring program revealed the presence of
trichloroethylene (TCE) at levels greater than 1 mg/L. Since these levels exceeded action levels for
TCE that were established by the State and would lead to the shutdown of wells, VCWD began
evaluating the potential use of a packed tower aeration system and the secondary effects of aeration
(Umphres and Van Wagner, 1986). The study covered several phases (pilot study; design,
construction, and monitoring of the full-scale packed tower aeration facility) and examined the
secondary effects of mineral scaling, corrosivity, microbiological quality, equipment noise, air
pollution, and water particulates. Observations are reported below for each of these secondary
effects evaluated.
Mineral Scaling—Calcium carbonate scaling occurred in the lower part of the tower and in
pump casings downstream of the PTA. A dose of 1 mg/1 of hexametaphosphate added to the aerator
influent prevented further scaling.
Corrosivity—Copper corrosion rates were not influenced by aeration, while mild steel
corrosion rates decreased slightly.
Microbiological Quality—Without chlorination of the PTA influent, standard plate counts
of the effluent were higher than the influent. This suggests that microorganisms were growing in
the oxygen-rich tower or were scrubbed from the influent air.
Equipment Noise—Instantaneous noise levels were measured with a Columbia Research
Laboratory Model SPL-103 Sound Pressure Level meter. "A scale" readings taken indicated that
2-50 May 1999
-------
the blower contributed little to the base noise level for the area either during working hours or on the
weekend. Noise levels of the centrifugal blower were well below OSHA standards and dropped to
background levels 100 feet from the aerator.
Air Pollution—Computer modeling of the air emissions from the aerator prior to the detailed
design indicated that ground level concentrations of organics as a result of aeration would be
insignificant. Air samples were analyzed upwind and downwind of the aerator to check the findings
of the model. At a detection level of 1 ppb (about 5.8 ug/L for TCE), TCE was not detected in any
of the samples. Since 1 ppb (which was the standard detection level for that analyses) is
considerably higher than the 21.8 ppt TCE predicted by the model, model estimates could not be
fully verified.
The potential impact of ambient air concentrations of volatile organics on the pilot tower
performance was examined. As the concentration of the volatile organics in the influent air
increased, the driving force for transfer from water to air diminished. In the extreme case where the
concentration in air is high enough and concentration in water is low enough, the aerator would
function as an air scrubber by transferring the contaminant back into the water.
Water Particulates—Air stripping removal efficiency was increased by improving the
distribution of water onto the packing and by adding packing to compensate for packing that settled
as a result of start-up. Accumulated residue at the blower inlet indicated that dust and particulates
were entrained into the PTA. Particle counts and turbidities did not increase significantly as a result
of packed tower aeration.
Overall Conclusions of the VCWD Study—The impact of packed tower aeration on
particulates, microbiological contaminants, and scaling is site dependent. Aeration facilities should
be provided with the capability to feed chlorine to the influent and/or effluent to reduce the impact
of microbiological growth. The water quality of a potential aerator site should be reviewed with
respect to scaling. As a minimum, the design of the facility should provide for future installation
of a chemical feed system to prevent scaling.
2-51 May 1999
-------
2.1.6.2 Diffused-Bubble Aeration
Lowry et al Pilot Studies
Lowry et al. (1984) performed pilot scale shallow depth diffused aeration studies in a 120-gal
(454-L) vessel. The aerator consisted of a series of fine bubble ceramic stones that achieved a
practical aeration pattern. The process had a 60-minute aeration period and an aeration rate of 50
scfh (1,400 L/h), which provided an A:W ratio of 3:1. The unit was operated for 3 days with an
average influent radon concentration of 76,000 pCi/L. The average effluent concentration was
4,900 pCi/L corresponding to a 93.6-percent removal.
Deep Tank DBA Belstone, England
A deep-tank diffused aeration system was installed in Belstone, England (Rafferty, 1983) to
remove radon from a water supply with a capacity of 2.5 mgd. The water supply contained 10,000
pCi/L of radon. Designed and constructed in the early 1960s, the treatment facility consisted of an
aeration tank that was divided into two parallel compartments Each compartment was 70 feet long
by 10 feet wide and was equipped with a weir at each end to provide a 4-foot depth of water. The
tank was equipped with 2,800 diffusers, each designed to diffuse air at a rate not to exceed 0.8 cubic
feed per minute (cfrn). Air was supplied by two of four 30-hp blowers, each capable of providing
1,125 cfin of air. At the design water flow rate of 2.5 mgd and air flow rate of 2,250 cfrn, the plant
operated with a 24-minute detention time and an A: W ratio of 8:1. This treatment resulted in a long
term radon removal rate of 97 percent.
Shallow Depth DBA Kinner et al. Case Study
Kinner et al. (1988) installed a shallow-depth diffused-bubble aeration system that consisted
of three 270-gallon polyethylene tanks connected in series. Water flowed through each tank at a
flow rate of approximately 9,960 gpd with an influent radon concentration of approximately 77,500
pCi/L. Air was provided through the spiral tube diffusers that contained holes of .0015 in. diameter
and were placed 14 hi. above the bottom of the tanks. The system was tested at A:W ratios ranging
i
from 2:1 to 15:1 and two water flow rates (12 gpm and 27-33 gpm). At A:W ratios of 5:1 and
greater (at both flow rates), the diffused bubble system obtained 91 - to 99-percent radon removal and
effluent levels ranged between 700 and 6,542 pCi/L.
2-52
May 1999
-------
MSBA Dixon et al. Case Study in Pennsylvania
Multi-stage diffused-bubble aeration was one of three aeration technologies evaluated for the
removal of radon from a ground water supply in Pennsylvania (Dixon et al.. 1991). The relevant
design and operating parameters were as follows:
No. of Stages
Flow Rates
A:W Ratio
50 gpm, 100 gpm
16.6:1,11.1:1, 5.5:1 for 100 gpm
33.2:1,22.2:1,11.1:1 for 50 gpm
The influent radon concentration ranged from 1700 to 2700 pCi/L. The removals ranged
from 97 to 100 percent, except for one run conducted with A:W ratio of 5.5:1 which achieved 86-
percent removal. Neither flow rate nor A:W ratios (at or above 11.1:1) had a significant effect on
removal efficiency.
2.1.6.3 Spray Aeration
Jet Aeration Case Study, Pennsylvania
Spray jet aeration was one of three aeration technologies evaluated for the removal of radon
from a ground water supply in Pennsylvania (Dixon et al., 1991). Well water was pumped through
a spray jet unit into a baffled steel tank with a detention time of 20 minutes. A series of tests were
conducted to study the effect of air-to-water ratio, number of passes, remote mounting of the unit
(attaching a 50-ft hose to the discharge end of the unit), and flow rate on the removal of radon. The
following observations were made:
The variation in the air-to-water ratio had only a marginal effect on radon removal
efficiency. Radon removal of 87 percent was achieved at an A:W of 3:1, and different
runs at A:W ratios of 6:1 to 17:1 achieved radon removals in the range of 83 to 91
percent.
2-53
May 1999
-------
The removal of radon improved with increase in number of passes. Radon removal was
68-74 percent for a single pass and increased to 99 percent for four passes.
Removal efficiency decreased significantly due to back pressure when the spray jet was
remote mounted.
2.1.6.4 Slat or Cascade Tray Aeration
Shallow Tray Air Stripper, New Boston, New Hampshire
North East Environmental Products, Inc. installed a pilot air stripper with a 24x16 ft
footprint at the New Boston Air Force Station, New Hampshire (Alexant, 1995). The design flow
through rate for the air stripper was 80 gpm. The system included a 600 cfm air blower and four
aeration trays. The unit was expected to achieve 99.9% radon removal; for a projected maximum
influent radon concentration of 76,000 pCi/L, treated water projected concentrations would be 300
pCi/L or less. The A:W ratio was 56:1. Analyses done during trial operation of the pilot system
showed removals greater than 99% (influent radon of 93,593 pCi/L and effluent values of 144,138,
and<100pCi/L).
Wooden Slat Tray Aeration
Smith et al. (1961) evaluated the performance of a wooden slat tray aeration system for radon
removal, although the aeration system had not initially been installed for radon removal. The height
of the aerator and the water and air flow rates were not reported. The average influent radon
concentration was 6,780 pCi/L. The slat tray aerator achieved an average removal of 71 percent,
with the treated water having a radon concentration of 1,950 pCi/L.
Cascade Tray Aeration
Dixon and Lee (1988) studied the effect of cascade tray aeration on radon removal at three
sites. Details of the setup were not provided. The cascade tray aerator removed greater than 75
percent of the radon as shown in Table 2-8.
2-54
May 1999
-------
Table 2-8. Cascade Tray Aerator
Site
1
1
2
3
3
Run
1
1
2
i
2
Type of Sample
West aerator influent
West aerator effluent
East aerator influent
East aerator effluent
Raw water
Effluent
Raw water
Effluent
Influent
Effluent
Influent
Effluent
Mean Radon
Concentration
pCi/L
327
50
342
37
465
108
521
46
269
54
260
50
Percent
Removal
89
77
91
80
81
2.1.6.5 Point of Entry (POE) Devices
Several prototype home diffused-bubble aeration and spray aeration treatment devices have
been field tested. These studies are described below.
Kinner et al POE Diffused-Bubble and Bubble-Plate Case Study in Derry, NH
Kinner et al. (1993) evaluated the performance of diffused-bubble and bubble-plate aerators
in POE applications. The study investigated radon removal efficiencies, potential problems (e.g.,
emissions and equipment malfunctions), and economic issues. The aeration devices were installed
inside the pump house at an abandoned groundwater well in Derry, N.H. Radon activity in the
influent ranged from 22,837 pCi/L to 54,765 pCi/L, with an average of 35,620 + 6,727 pCi/L. The
diffused-bubble device contained three compartments (each 24-cm long x 40-cm wide x 40-cm high)
in series. Each compartment had an internal diffuser, with all diffusers fed from one header. The
diffusers had variably spaced holes of 0.51-mm diameter, producing relatively small bubbles to
provide a larger surface area for gas transfer. A 38-cfm capacity blower was used for the air feed
and a water flow rate of 2.3 gpm was used to provide an A:W ratio of approximately 119:1. The
2-55
May 1999
-------
system consistently achieved radon removals higher than 99 percent (effluent radon < 200 pCi/L),
even during periods of restricted air flow due to accumulated iron precipitation on the diffusers.
The bubble-plate unit was contained in a molded plastic casing (60-cm long x 38-cm wide
x 23-cm high) with a 10-cm diameter PVC vent pipe. A 0.95-cm spiral diffuser fed water into a 7.6-
cm wide polyethylene channel containing 4.8-mm diameter holes along the bottom, spaced 1.9 cm
apart. Air was fed in through these holes, causing the influent water to rise up to 17 cm. At the end
of the polyethylene channel, the treated water flowed over a weir and into a holding tank. With an
air flow rate of 125 cfm and a water flow rate of 6.0 gpm, the unit had an A: W ratio of 156:1. The
bubble-plate aerator generally achieved radon removals of more than 99 percent, but did have
diminished effluent quality on a few occasions, particularly when there were problems with clogging
of the blower's air intake filter.
Kinner et al. noted that pretreatment for iron would probably be needed since iron oxidation
'
occurred readily in both aeration units, causing iron precipitate to accumulate in the diffusers
(decreasing air flow rates) or to be released in pulses when the units were started. The study found
that the plume from the off-gas emissions is diluted fairly rapidly, but venting should be set above
the roof line to keep radon gas from entering the home. Treatment of off-gas emissions could be
necessary hi States where the emissions are regulated.
Lowry et al. POE Diffused Aeration Field Study
Lowry et al. (1984) tested a home diffused aeration unit during a field study. The basic
components of the system were an aeration tank, a fine bubble diffuser, a liquid level pump control,
and a timer-controlled air supply. A shallow well pump was provided to repressurize the treated
water after aeration at atmospheric pressure. The diffuser was a composite of eight one-inch
spherical ceramic porous diffusers made of fused crystalline alumina grains which were arranged
uniformly within a 12-inch diameter area. The blower supplied an air flow rate of 30 scf/h to
I
provide an A:W ratio of 3.4:1. The system reduced the influent concentration of radon from 50,000
pCi/L to an average 2,500 pCi/L in the treated water.
2-56
May 1999
-------
Kinner et al. POE Shallow Diffused-Bubble Pilot Study
Kinner et al. (1990a) conducted pilot scale studies for the removal of radon by shallow-depth
diffused-bubble aeration applied at the point of entry to a home water system. Influent radon levels
ranged between 22,900 and 54,800 pCi/L with an average of 35,700 pCi/L. The unit consisted of
three aeration tanks (46 cm long, 40 cm wide, 24 cm deep). Each tank was fed by an interval
diffuser from a common header. A 56-cfin capacity blower was used for the air feed to provide an
A: W ratio of 119:1. The system achieved consistent radon removals higher than 99 percent.
Rost Spray Aeration POE Case Studies
Rost (1981) evaluated the performance of a spray aeration system built and tested hi
Hallowell, Maine by the Department of Human Resources. The unit was installed in a private home
during the spring of 1981 and tested for a period of 4 months. The unit consisted of two atomizing
spray heads and a 50 gallon receiving tank. The system provided 12 minutes of recirculation through
the spray aerator prior to use. Influent radon concentrations ranged from 44,000 to 63,000 pCi/L.
The process produced water with an effluent radon concentration ranging between 2,460 and 7,600
pCi/L, for an average removal of 93 percent.
As part of the same study (Rost, 1981), a separate spray aeration unit was installed in a single
family house and operated for 1 month. This unit consisted of a stainless steel tank with two
separate compartments of equal size, each with a capacity of approximately 100 gallons. The liquid
volume in the first and second compartments was 30 and 60 gallons, respectively, to allow for air
space. The raw water was introduced into the first compartment as a horizontal spray which
impacted onto the vertical wall separating the compartments. A half horsepower pump, with its
suction near the bottom of the first compartment, sprayed the water through a similar nozzle
arrangement into the second compartment. The water was then pumped into the pressure tank for
use in the home's water system as needed. The results of the study are summarized below:
2-57
May 1999
-------
Average Influent Radon Concentration 22,380 pCi/L
Effluent Concentration After 1st Spray 5,460 pCi/L
Effluent Concentration After 2nd Spray 2,290 pCi/L
Average Overall Removal 82 percent
After 20 days of operation, the unit was drained and thoroughly vented by blowing air
i
through the system. The radon removal then increased from 82 to 87 percent for a period of 1 week.
It was hypothesized that since radon is heavier than air, it was lying above the water surface in the
I
two chambers and radon would re-enter the aerated water again as the spray passed through the layer
of radon. A suction tree was substituted for the nozzle arrangement in Tank 2 for 1 day, with less
effective results. The chamber removal efficiencies dropped first to 67 percent then to 47 percent
and finally to 14 percent.
2.2 LOW-TECHNOLOGY AERATION METHODS
2.2.1 Process Description
Several low-technology aeration alternatives are available. These technologiesprovide some
transfer between air'and water. They include free-fall aeration, spray aeration and bubble aeration
(both using simple devices), Venturi aeration, and mechanical surface aeration. These low-
technology methods, however, will only provide low removals (relative to the aeration technologies
discussed in Section 2.1). Therefore, for waters with high levels of radon, achieving the removals
necessary to meet requirements might not be feasible with these methods.
Although several low technology radon treatment processes have achieved some degree of
removal the following barriers might limit the implementation of any of these technologies in full-
scale treatment:
• Existing storage tanks may not provide adequate headroom and ventilation to prevent the
accumulation of a gaseous layer of radon, which can re-enter the aerated water.
2-58
May 1999
-------
» Existing pumps designed to discharge into an atmospheric clearwell would be inadequate
to pump against the additional head imposed by distribution manifold piping and a spray
nozzle system.
• Modifications to storage facilities may interrupt service.
« The capacity of existing storage tanks will probably be insufficient to provide the
necessary detention time to reduce radon concentrations to acceptable levels if high
influent levels are present.
• As stated earlier, these systems will probably not achieve removals that are sufficient to
attain desired water qualities, if influent levels are extremely high.
• Rechlorination may be necessary after storage to provide a residual in the distribution
system.
As with other aeration systems, off-gas emissions, iron and manganese precipitation, and
corrosion control issues can be concerns and need to be addressed before implementing low
technology aeration treatment systems.
2.2.1.1 Free-Fall Aeration
Free-fall aeration involves the flow of contaminated water over a weir or similar structure
to provide a free-fall effect into a storage tank. This enhances the transfer of contaminants into the
atmosphere due to the increased surface area exposed to the atmosphere and additional turbulence.
One type of free-fall aeration is cascade aeration, where water flows down over a set of steps
or baffles that increase the time the water is exposed to air and the area-volume ratio. The simplest
structures are basic sets of concrete steps, where the greater the number of steps the longer the
exposure time. Baffles produce turbulence and can be used to increase the area-volume ratio.
Limitations associated with these structures include the need to house them with adequate ventilation
in cold climates, and the occurrence of slime and algae buildup and corrosion. (AWWA and ASCE,
1998)
2-59
May 1999
-------
2.2.1.2 Low Technology Spray Aeration
For low technology spray aeration, a simple spray nozzle is installed above an atmospheric
storage tank or basin. The influent is then discharged into the tank or basin from the sprayer.
2.2.1.3 Low Technology Bubble Aeration
Low technology bubble aeration involves delivering air bubbles into a storage tank or basin
using a simple device consisting of a hose or pipe with holes and a blower.
2.2.1.4 Venturi Aeration
For venturi aeration, a venturi laboratory device is installed above an atmospheric storage
tank or basin to discharge the influent into the tank or basin.
2.2.1.5 Mechanical Surface Aeration
Mechanical surface aeration involves using a mechanicalmixer to agitate the surface of water
in a basin. The agitation brings more air into contact with the water to enable increased radon
transfer from the drinking water to the air. One advantage of mechanical surface aeration is that it
can often be retrofit to existing basins. There are several disadvantages to this technology, including
the need for large basins, long residence times, and high energy inputs. (NRC, 1997)
2.2.2 Removal Efficiency
Since removal rates can vary considerably for low-technology methods, it may be necessary
to pilot test the system to determine actual removal rates and avoid site-specific problems. For small
systems with waters requiring 50-percent radon removal rates, low-technology spray aeration may
be the best choice.
Removal efficiency data reported for various low-technology aeration techniques are
presented in Table 2-9.
2-60
May 1999
-------
2.2.3 Treatability/Case Studies
Kinner et al. (1990) conducted pilot-scale studies of the effects of several low technology
techniques that involved modifications to a simulated atmospheric storage tank. Most of the
techniques involved simple aeration mechanisms. The techniques studied were:
Water entry at the bottom of the tank
Table 2-9. Low-Technology Aeration Removal Efficiencies Observed
Technique
Mechanical surface aeration
Free-fall into a tank
Free-fall and simple bubble aeration
Simple spray aeration with free-fall
Simple bubble aeration
Percent Removals Observed
83-92%
50-70%
86-96%
60-70%
80-95%
Source
Drago(1998)
Kinner etal. (1990)
Kinner etal.( 1987)
Kinner etal. (1990)
Kinner etal. (1990)
• Water entry at the bottom of the tank with minimal bubble aeration
• Water entry via free fall from 2 ft above the tank water level
• Water entry via free fall from 2 ft above the tank water level with minimal bubble
aeration
• Water entry via free fall from garden spray nozzle 2 ft above the tank
• Water entry via free fall from laboratory venturi apparatus 2 ft above the tank.
For each of these techniques, the radon removal was measured for hydraulic detention times of 8 to
23 hours. The results are shown in Table 2-10.
Following the pilot-scale study, Kinner et al. (1990) installed a minimal aeration system to
an existing storage tank at a housing development. The system consisted of 2 in. PVC pipes with
1/8 in. diameter holes and a blower injecting 20 cfm of air. The detention time for the system was
5.3 hours. Radon removal efficiencies ranged from 80 to 88 percent.
2-61
May 1999
-------
Table 2-10. Radon Removals for Low Technology Techniques
Technique
Bottom entry
Bottom entry w/aeration
Free fall
Free fall w/aeration
Free fall from sprayer
Free fall from venturi
Radon removal % for 8 hr
detention time (flow=0.9 gpm) -
~10<»
83
50
86
60-70
(3)
Radon removal % for 23 hr
detention time (flow=0.24 gpm)
~ 34(I>*
.95
70
96
no data'2'
(3)
(1) Slightly above removal expected from decay and volatilization without any treatment.
(2) Nozzle unable to operate at low flow rate.
(3) Water flow rates too low to obtain good venturi action.
Drago (1998) reported removals of 83 to 92 percent for a mechanical surface aeration system.
The system had flow rates between 5,450 and 7,600 gpm and detention times of 288 to 408 minutes.
I
Drago noted that mechanical surface aeration was only appropriate with a large existing clearwell.
Kinner et al. (1987) assessed the performance of free-fall aeration for radon removal. The
I
researchers observed removals of 50 and 70 percent for detention times of 12 and 30 hours,
respectively when water was allowed to free fall into a storage tank.
•
j
Dixon and Lee (1988) observed a decrease of approximately 18 percent in the levels of radon
in water that was stored within a water supply distribution system. This decrease was attributed
largely to volatilization of the gas caused by pumping and agitation of the water and by ventilation
within the storage vessel. Decay of the radon accounted for only approximately 34 percent of the
noted decrease in the radon concentration.
2.3 GRANULAR ACTIVATED CARBON (GAC)
In drinking water treatment the use of GAC in the United States has been limited primarily
to applications for the control of synthetic organic chemicals and taste and odor compounds.
However, since the detection of radon in drinking water supplies, a number of research and
2-62
May 1999
-------
pilot-scale studies have been undertaken to evaluate the effectiveness of GAC for controlling radon.
Based upon the results of this research and pilot-scale work, GAC appears to be effective in
removing radon from water.
2.3.1 Process Description
R.adon is removed from water by adsorption using granular activated carbon (GAC). The
adsorption process occurs when the radon molecules diffuse through the water to the surface of the
GAC. Radon sorbs at the interface between the water and the carbon. Therefore, a high surface area
is an important factor in the adsorption process. Although the outer surface of the carbon provides
some available area for adsorption, the majority of the surface area is provided in the pores within
the carbon particles.
Adsorption systems usually operate hi a.downflow mode where the contaminated water is
introduced at the top of the carbon bed and flows through the bed to the bottom. As the water moves
down through the bed, the radon is adsorbed to the carbon until all the available interfacial area is
taken up. The radon moves with the water through the bed until there is available area for adsorption
to take place. Contaminant removals are a function of the available interfacial area between water
and carbon, and are a function of time.
The design of a GAC system needs to account for competitive adsorption from natural
organic matter in the water that may compete with radon for adsorption sites and thus increase the
amount of carbon needed to sufficiently remove radon (NRC, 1997). Adsorption can be limited by
suspended solids accumulation in the carbon columns or beds, which can cause hydraulic short-
circuiting (bypassing portions of the bed) and also by coating (blinding) the outer surface of the
carbon. Adsorption of large molecules can also block the adsorption of other materials. The
adsorption of oxides and carbonates, such as Fe, Mn, and Pb, can reduce the adsorption capacity of
GAC for radon. However, the actual surface density of adsorbed decay daughter products does not
2-63
May 1999
-------
greatly affect the capacity for radon adsorption since the decay products tend to be at much lower
concentrations than Fe and Mn hi groundwater.
A minimum contact time is required to reduce the contaminant level to any particular
concentration. This required residence time increases as the column or bed becomes partially
exhausted (saturated with contaminants). The empty bed contact time (EBCT) is a nominal contact
time that can be defined as the nominal volume of the contact vessel, divided by the design flow.
The actual bed contact time is affected by other adsorbable materials in the water that can reduce the
adsorption capacity for radon. At high radon levels, the bed/column size required for effective
removals may become so large that GAG is impractical. The long EBCT required for the removal
of radon from drinking water sources by GAC may not be feasible particularly when radon levels
. are far higher than the MCL levels.
GAC has a finite adsorption capacity that is determined by the type of GAC and the
characteristics of the target contaminant. If VOCs are present during the GAC adsorption process,
the VOC concentration in the water leaving the GAC will be similar to the influent VOC
concentration when all available interfacial area has been exhausted. When the contaminant begins
to appear in the effluent, breakthrough has occurred. Once the breakthrough concentration reaches
a set level, the carbon must be regenerated or replaced. However, in the radon adsorption process
(where VOCs are not present), an adsorbed radon atom decays, reducing the interfacial concentration
of radon and restoring some adsorption capacity to allow new radon atoms to become adsorbed.
Thus, the effective life of GAC is extended through the in-situ decay of the adsorbed radon (radon
decays to ultimately become lead; 6.48 mg of radon are equivalent to 1 Curie and 1 Curie is 1012
picocuries).
i
Some of the major design considerations for GAC systems are:
• Adsorption/Decay Steady State
2-64
May 1999
-------
• Empty Bed Contact Time (EBCT)
• Contactor Configuration.
These major design factors are discussed below.
Additional design considerations include:
• Type of GAC - The adsorption capacity of GAC is partly determined by the type of GAG
(as discussed in the adsorption section below), so this may influence system design.
However, the selection of a particular type of GAC also often depends on the sources of
GAC that are available nearby since major costs can be incurred for transporting GAC.
Lowry and Lowry (1987) reported that the type of GAC used "has significant bearing
upon the performance achieved." Rutherford, as cited in Hess et al. (1998), found that
coconut charcoal works best for absorbing radon. NRC (1998) also noted that coconut-
based GAC has been found to be the best for radon sorption. Since coconut-based GAC
has a larger percentage of micropores than other types of GAC, it is thought that
micropores may be the most effective for sorbing small molecules and atoms like radon
(NRC, 1998).
• Temperature - Rutherford, as cited in Hess et al. (1998), also found that the cooler the
GAC, the better the absorption of radon.
AdsorptiowVecay Steady State—Carbon usage rates dictate the rate at which carbon will
be exhausted and how often it needs to be replaced (AWWA and ASCE, 1998). In the classic
application of granular activated carbon adsorption to the removal of organic compounds, adsorption
isotherms have been found to be useful screening tools not only for determining preliminary carbon
usage rates, but also for evaluating the effectiveness of different types of GAC. The Freundlich
isotherm model provides a method of empirically evaluating the adsorption characteristics of GAC.
The Freundlich equation for radon is:
X/M = KrC
l/n
where:
2-65
May 1999
-------
X = radon adsorbed in pCi
M = mass of carbon in mg
C = equilibrium radon concentration in pCi/L
1/n, Kr = Freundlich isotherm constants.
A log-log plot of the experimental data generally yields a straight line whose slope is 1/n and
intercept is K,. The slope or 1/n provides an indication of adsorption intensity and the intercept or
KT provides an indication of adsorption capacity (Lowry and Brandow, 1985). Table 2-11 lists the
radon Freundlich constants for a number of different carbons. As indicated on this table, the
adsorption capacity can vary greatly among different types of carbon.
Table 2-11. Freundlich Isotherm Data and Relative Ranking for Six Activated Carbons
and Radon at 10°C
Carbon type
A
B
C
D
E
F
Mesh Size
12x40
12x30
12x40
8x30
8x30
8x30
1/n
1.02
0.99
0.91
1.28
1.27
0.82
K,
pCi Radon
Adsorbed/ing Carbon
1.5600 x lO'3
1.2300 x 10-3
2.0600 x 10°
0.0044 x IQ-3
0.0046 x lO'3
4.6500 x lO'3
Source: Lowry and Brandlow (1985)
However, while adsorption isotherms may give an indication of the potential of a given
carbon for radon removal, they do not yield sufficient data to develop design criteria for GAC
treatment systems. This is due to the effect of decay and the adsorption/decay steady state
relationship that dictates the performance of a carbon column/bed in continuous service (Lowry and
Brandow, 1985).
Breakthrough is a key factor in the design of a GAC system and, as noted earlier, occurs
when the effluent concentration of a contaminant after GAC treatment exceeds a maximum
2-66
May 1999
-------
acceptable criteria. When breakthrough occurs depends on the design of the carbon bed/column
(e.g., type and depth of carbon) and the quality of the influent. Breakthrough curves are used to
define the relationship between the physical and chemical factors of the GAC system (e.g., flow rate,
bed/column size, carbon exhaustion rate), the number of beds/columns and their arrangement (i.e.,
in series or parallel), and treatment plant effluent requirements (AWWA and ASCE, 1998). For
radon, the removal relationship shows an initial period where the adsorption is maximum and decay
is minimum. A typical breakthrough curve for radon shows that the radon concentration in the
treated water begins to increase gradually, and then over time it levels off ata steady state value.
A steady state is established within the carbon bed when the adsorption rate equals the rate of decay.
Knowing the half life of radon (3.82 days), the decay constant of radon can be calculated as
follows:
M(t) = M0 e
,-A.t
where:
M = mass in mg at time t
t = time in days
M0 = initial mass in mg or concentration in mg/L or pCi/L
A, = decay constant in days'1.
Knowing the half-life of radon (3.82 days), solving the above equation can be as follows:
M (3.82) = !/2 MO, since 3.82 days is the half-life
In 0.5 = -3.82A
= 0.18/day.
Therefore, M (t) = M0e
,-0.18t
2-67
May 1999
-------
Researchers have found that the removal constant Kr varies by the type of GAC and by the
source water quality. The different system constants reported by various studies are presented in
Table 2-12. Early research showed a radon removal constant ranging between 1.25/hr and 3.2/hr
in laboratory controlled studies and a radon removal constant ranging between 1.34/hr to 1.89/hr in
field studies. As can be seen from Table 2-13, the Kr constant ranges from 1.35/hr to 5.8/hr. In
more recent research, the K, changed significantly between two different sites (and source waters)
(AWWARF, et al, 1997). Iron, manganese, and TOC hi source waters affect radon adsorption by
GAC. One study reported a decrease in the rate constant from 4.57/hr to 2.76/hr when the source
I
water iron concentration was increased from 0.09 to 3.9 mg/L (AWWARF, et al. 1997).
Pretreatment to sequester/remove iron and manganese, or other pretreatment such as pH adj ustment,
can improve radon adsorption and extend carbon life. (Pretreatment is discussed further in Section
2.3.3.)
As the system rate constant (Kr) increases, the EBCT decreases. Since the system contact
time is site specific, a treatability pilot study is required to determine the rate constant for a particular
i
GAC type and source water. For small systems that cannot afford a treatability pilot study, a
conservative Kr should be used.
Table 2-12. GAC Kr Constant by Carbon Type(1)
Carbon Type
A
B
C
D
E
F
G
H
K«(l/hr)W
Co = 50,600 Laboratory*3*
1.25
1.76
3.20
2.09
2.28
2.61<5>
1.24
1.97
Co = 700,000 Field(4)
1.34
-
-
1.58
1.89
-
-
-
Notes:
(1) Unpublished data. EPA Grant No. R-81-829-01-0 Office of Research and Development
(2) KB(K at steady state) calculated from Empty Bed Contact Time (EBCT)
(3) Continuous flow
(4) Full-scale household systems
(5) Data limited.
2-68
May 1999
-------
Table 2-13. GAC Kr Constant by Carbon Type as Reported From Different Sources
Rate Constant K in 1/hr or hr'1
4.57-5.15
4.76-5.31
2.69-3.64
3.17-4.52
1.81-3.67
2.43(2.99-2.28)
5.8
3.3
1.35
2.09
1.53
3.02
1.48
2.98 without cation exchange
pretreatment
3.29 with cation exchange
pretreatment
GAC Type
Column 1 Barneby & Sutcliffe PE 12x30
Column 2 Barneby & Sutcliffe PE 12x30
Column 3 Barneby & Sutcliffe PE 12x30
Column 4 American Norit HD 3000
Columns CalgonFSQO
American Norit HD 3000
BC1002
BC1002
American Norit Peat (8x20)
ICI Americas* HydroDarco 4000 (12x40)
Calgon F-400 (12x40)
Barneby Cheney 299 or 1002
American Norit HydroDarco 4000
BC1002
Source
AWWARF, 1997
New Jersey site
AWWARF , 1997
Amherst site, EPA, 1990
Mont Vernon site, EPA, 1990
Lowry and Lowry, 1987
Lowry and Lowry, 1987
Lowry and Lowry, 1987
Lowry and Lowry, 1987
Lowry etal., 1987
Kinner, et al., 1993
* Now American Norit
Empty Bed Contact Time—The required empty bed contact time can be calculated, given
a radon influent level, desired radon effluent level, and a carbon steady state system constant. The
system is modeled by the following equation:
In
c = c e
t o
or t =
K
where:
C0 = influent radon level(pCi/L)
Ct = desired effluent level (pCi/L)
t = EBCT (hours)
Kss = steady state system constant (1/hr).
The empty bed contact time is a function of contactor volume and flow rate.
2-69
May 1999
-------
Calculations by AWWARF (1997) indicate that the EBCT can range from 7 minutes at a
of 4.5/hr, an influent radon concentration of 500 pCi/L, an effluent radon concentration of 300
pCi/L, and a removal efficiency of 40 percent to 293 minutes at a K^ of 1.5/hr, an influent radon
concentration of 300,000 pCi/L, an effluent radon concentration of 200 pCi/L, and a removal
efficiency of 99.9 percent. As with the system rate constants discussed above, EBCT can be
assessed through treatability pilot studies.
Contactor Configuration — The three basic configurations for contactor operation are
downflow fixed bed, upflow fixed (packed) or expanded (moving) bed, and pulsed bed. These
configurations can have single adsorbers, or multiple adsorbers operated in series or parallel
(AWWA and ASCE, 1998). Single-stage operation is applicable to radon removal in ground waters
because of the relatively low carbon usage rates. A system of multiple contactors in series has the
advantage of being able to capture first-stage breakthrough. Two or more fixed beds operated in
parallel typically are used when a high flow rate would require a vessel diameter too large to be
economical or feasible.
Downflow fixed bed contactors offer the simplest and most common contactor configuration
for radon removal in ground water. The contactors can be operated either under pressure or by
«
gravity. Pressure contactors may be more applicable to ground water systems because of the nature
of these systems. The use of gravity contactors for most ground water systems would involve
repumping the treated water to the distribution system. On the other hand, pressure contactors might
be used without repumping, thus reducing both capital and operating costs. Downflow contractors
are used when the unit is also being used to filter out suspended solids (AWWA and ASCE, 1998).
The suspended solids are removed periodically by backwashing.
When suspended solids concentrations are high, solids accumulation and resulting head
losses may be high if downflow contactors are used. In these situations, upflow beds used with
subsequent filtration processes to remove suspended solids may be preferable. Upflow beds can also
be used in waters with low suspended solids concentrations since the bed is not needed as a filter
(AWWA and ASCE, 1998). Upflow beds have also been applied to situations where very long
2-70
May 1999
-------
contact times (greater than 120 minutes) are required. For pulsed bed adsorbers, the bed is
occasionally pulsed to release some of the exhausted carbon from the bottom, while some fresh
replacement carbon is added to the top of the bed (AWWA and ASCE, 1998).
Point of entry GAC units typically consist of a metal or fiberglass pressure vessel containing
gravel and the activated carbon. Adequate volume is provided above the carbon bed for expansion
during backwashing. Capacities are usually between 1.0 cubic feet to 3.0 cubic feet (28 to 85 L) of
carbon. (Lowry et al., 1987)
2.3.2 Removal Efficiency
Several studies have documented the removal efficiency of GAC for radon. Typical removal
efficiencies in the range of 80 to 99 percent have been recorded. These studies have involved
household unit or Point-of-Entry (POE) devices, pilot studies, and full scale plants. Flows have
ranged from 125 gpd to approximately 10,000 gpd with average influent radon levels between
approximately 2,500 pCi/L to approximately 750,000 pCi/L.
However, to remove 99 percent of the influent water radon, a long EBCT is needed. Dyksen,
Hiltebrand and Guena (1986), as reported in Dixon and Lee (1987), concluded that EBCT for the
removal of 99 percent of radon is 130 minutes. Experiments conducted by Dixon and Lee indicated
a 35-percent reduction in radon concentrations using GAC filters with EBCT of 10.5 minutes.
Hiltebrand,DyksenandRaman (1987) conducted pilot studies that showed approximately 70 percent
reduction of radon levels using a GAC filter with an EBCT of 30 minutes. Their research also
showed that pressure GAC filters with EBCT of 2 minutes and loading rates of about 1 gpm/ft2 may
reduce radon concentrations up to 10 percent of influent concentration. Table 2-14 presents a
summary of radon removal efficiencies reported by various sources.
2.3.3 Pretreatment
GAC systems may require some kind of pretreatment to prevent clogging of the carbon bed
and to minimize the organic loading on the carbon. Clogging of the bed could be caused by
suspended solids in the raw water and/or by precipitation of iron and manganese on the carbon. The
2-71
May 1999
-------
U
<:
u
i
o
•o
cs
•*
F-"
I
O
i
i
c
u
I i
* t;
11
to.
£
J
*l
£.
1
h.
c
§
1
S.
£
u
u
1
^
U« Jj
§|
"3
g
u
l|g
H •§ a
5~
Il*£
: 5 J S
,
I
•ill
11!
e: t-
55
« i
t>
1
u .u
S i
|
1
|
j
a
fill
•^ S~* *o — *
r** r**
^*£J
"*- ~
2 2
1.
00
3 O S*
i^ s?§
V gw"
i.*|i.=
r £, so" S JJL_
"sif =•
fill
* _ -
111
o e s
8- 2 "2
•§ - 8
if!
nil
8^.P|
1 1 1 I
2 r* §"= at
§0 H o •;
cr «> w c
r* < S <2 £
1
oo
^J*
Z
Z
1
g
>o
_
1
|
00
~ M
^ ea
|i
3 f^
°-
1
•*
"3
E C^
"a
3
CD
2
c:
•o
- - e
11 I
.&S „ § _
= a '= H §
c ^ ~* ~"
1 g. .1
£ a °
r-i
o
3
O
a
i.
D
NO
_
o — .e —
I s * s
S K °f S
A f2 f
0,' C.^ B.
^3- o
— *—, °
rt t/j o^
^^ H i
§o o
£ 2
7 S
s|
™ ™
I1!
ca
1
2
o
!i
I g
1 Full-scale labor
5 dilTerenl bed
o
OO
§
C4
I
Wt
0
O
7? -^
K o
|-"
O
3
1
2
<
O
— "3
VI
1
00
g'
•^
9
3
>
—
||
! °*
[I
ta
"5
10 household u;
o
-
«J*
z
&
T3
|8
S Tr
M
g
ll
fN
5
l\
i
01
O
3"
r»
ii
sO, OO"
3
1
)
1
.S S
S °
2 contactors in
Iron levels wen
(3
•*.!-_ u
-0 = <
0 _' "g
*y *c3 w-i
" 0 °°.
= — 0
r^
o
<^.>
^
2 2
s
"
•I -1
= £
•v o*
0 O
£ 0
k oo
3 —
I ^ 1
J *O O C^
^\ *F • •
Sigi
?"- -s
' " "*"
J . ^ ,
E | z
IP|
s a"8!
- = 1 "3 8
- '» .s ^
— r*l 2 — = "H
1 5 If *i
S | '5 "2 « 2 S
u u 8 ' i 1 !
S i -s 4 1 3 *
« 8 s s <3 .2 a
o o o o o
o o" — — —
it n ii ii ii
— r^ m TT «n
= e = c e
i||||
•§ J •§ "I I
0 U 0 0 0
-_••= 1
s 1 1 1 1 ti
"** ^> *^ O "O ^
II = f •= g g
w J: 2 2 'S o.
'* ?' a |
f-{ o , __ , g
"^ i A "= g 'S 2
5 ,7 | * - §
1 |a=a§
o
111
— s
o
!n Jn § § 2
1 " II II II
1 1 i I 1
^ 3 J = 3
^ w Cj U O
2 ffl
2 r^
*^ S •< K
— t: w^ o —
g I* ll
i !*• j*
,
*•
> 2
\ _
• P£
"(3 «
» r«- r£
2 s a
• ^3, —
; <
.Si
S
3 contactors in :
II
c "3
fN O
0 ,0
H
S o
o
2
1
n
9
<
5.
so
o
(N
TT
0
D
9
2
Oo
"vi 5"
3
« o
j- r^* "°
;=-!
.s
_e
1
••= «s
si
4= "3
>• 2
"5
4J
H o
So
-»rf
^S
2
1
m
g
yr
<
i
O
^~
J
o
(*?
* s.
< 1
oo
1
>.
.0
ea
03
CQ
§
rr
^
&
S
2
£
c
o
E?
3
O
O
ti.
m
CN Q
0v O
1 S
^ X
>,!•
^ u
| |
1 f
" rN
ON O
CM —
U U
8
1
Is-.,
- 1
c ^
||
li
§1
a? s
232
SJ= ii
< Z <
* < Z
2-72
May 1999
-------
former is typical of surface water systems, while iron and manganese in the soluble form may be
encountered in ground water systems. Clogging may also be caused by biological growths when the
carbon bed life is long. The influent water may also contain competing organics. Competitive
adsorption and the influence of high concentrations of background organics have been found to
affect the removal of organics but their impact on radon removal is unknown. (McCreary and
Snoeyink, 1980).
AWWARF (1997) found that iron (and possibly TOC) decreased the K,s of GAC from 4.5
h'1 to 3.2 fr1. NRC (1998) noted that the pattern and rate of accumulation of uranium, radium, and
Pb-210 can vary greatly when iron is present. Cornwell et aL, as cited in NRC (1998), reported that
high levels or uranium, radium, and Pb-210 occurred with iron-rich backwash residuals from GAC
units. In another study, iron, manganese, and turbidity at the levels shown in Table 2-15 were not
found to affect radon removal (U.S. EPA, 1990).
Table 2-15. Turbidity, Iron, and Manganese Levels
Parameter
Turbidity, NTU
Iron, mg/L
Manganese, mg/L
Mont Vernon
average influent level
0.17 -Phase I
0.05 - Phase II
0.06 - Phase I
0.06 - Phase II
0.02 - Phase I
0.02 - Phase II
Amherst
average influent level
1.75 -Phase I
2.99 - Phase II
0.50 - Phase I
0.78 - Phase II
0.09 - Phase I
0.10 -Phase II
Source: EPA, 1990
Fouling of the GAC by oxidized metals, organics, particulates, and/or microorganisms can
be controlled by pretreatment and/or frequent backwashing. Periodic backwashing of the GAC units
can remove some suspended solids which might otherwise clog the bed. Disinfection with chlorine
prior to GAC adsorption creates chlorine by-products during the reduction of chlorine on GAC.
Since some of the chlorine by-products are adsorbed by carbon, design of GAC systems needs to
account for this competitive effect. Filtration ahead of the GAC system is a common solution to
prevent clogging of the bed. For POE units a small sediment filter can be installed immediately
2-73
Mav 1999
-------
upstream of the GAG unit (Lowry, et al., 1987). Pretreatment can be provided to reduce the organic
loading on the carbon, thereby decreasing the carbon usage rate. The need for pretreatment should,
however, be justified on the basis of cost and performance. Examples of processes that may be used
for pretreatment include conventional treatment and ozonation.
2.3.4 Post Treatment
The dynamic behavior of bacterial populations on GAG has been the subject of several
studies. (Klotz et al., 1976, Cairo et al., 1979, McElhaney and McKeon, 1978, and Parson et al.,
1980). While the results of these studies have not presented a consistent picture of the dynamics of
bacterial growth on GAG and in effluent from GAG contactors, all found the average number of
bacteria in the effluent from GAG systems to be significantly higher than influent levels. More
recent GAG radon removal studies measured the bacterial population in the effluent and found that
at both the Mont Vernon and Amherst sites, bacterial populations in the effluent were significantly
greater than those hi the influent. At Mont Vemon the effluent level was as high as 362,000
CFU/100 ml while at Amherst the effluent ranged from 20,000 to 40,000 CFU/100 ml (EPA, 1990).
No incidents of waterborne disease outbreaks have been linked to GAG systems; however, these
studies indicate a need to ensure adequate disinfection of GAG contactor effluents prior to
distribution of the water and a need for careful monitoring of the disinfected water. If chlorination
is used to disinfect, the formation of disinfection byproducts (DBFs) is unlikely to be a concern until
the GAG is saturated since the natural organic matter (NOM) needed for DBF formation will sorb
to the GAG (NRG, 1998). If the GAG becomes saturated, DBF formation after chlorination could
be a concern.
In addition, well pumps may need to be modified or replaced to address the additional head
loss hi the pressurized treatment system (Kennedy/Jenks Consultants, 1991).
2.3.5 Operational Considerations
Gamma radiation exposure from process units and waste disposal issues related to the
accumulation of radioactive lead-210 on the media are two concerns associated with using GAG for
radon removal. The decay of radon within the GAC bed results in the growth of radon progeny.
2-74
May 1999
-------
Beta and gamma emissions are associated with daughters Bi-214 and Pb-214 which have short half
lives. The radon adsorbed on the GAC decays into its radioactive daughter products, resulting in a
buildup of radioactive lead-210. Pb-210 accumulates on the bed because of its long half-life (over
20 years), and its tendency to be adsorbed on the media.
2.3.5.1 Gamma Emissions
The accumulation of radon (and other radionuclides such as uranium and radium) on the
activated carbon poses a potential health risk to full-time operators of the water treatment system,
maintenance personnel, and handlers of the spent carbon. Radioactive air emissions could also
necessitate an analysis of radiation exposure to the nearby community (persons living or working
in the immediate vicinity or downwind).
According to research conducted on gamma emissions from a GAC, the level of radiation
surrounding the bed depends on the influent radon level, radon effluent level, and distance from the
bed. The exposure rate is significantly lower a few feet from the GAC bed compared to the
maximum exposure rate found at the surface of the GAC vessel. Field monitoring studies indicate
that gamma exposure rate dissipates with distance, and shielding with lead or water decreases
exposure rates further (Lowry, 1988). Lowry (1988) reported that encasing a small POE GAC unit
(1.7 to 3.0 cubic feet) in a tank of water can virtually eliminate the problem of gamma exposure
surrounding the GAC bed.
The levels of gamma emissions recorded from GAC contactors by several field monitoring
studies are listed in Table 2-16.
2-75
May 1999
-------
Table 2-16. Gamma Emissions from GAC Contactors
Gamma emissions . •'_,_
500 urem/hr
500 urcm/hr
4-5 mrem/hr
Max exposure at surface of 2.5 cu ft bed
4.6-16 mrem/hr at surface of unit #1
1.1-1.8 mrem/hr at bottom of unit #2
40-100 mrem/hr for workers in direct
contact with GAC bed during coring for
1-1.5 hours
Influent Radon
Level (pCi/L)
14,830-17,110
150-7,500
53,600-74,900
191,113
Site
Atkinson, New Hampshire
Colorado
Full-scale laboratory study
Mont Vernon, NH
Source
AWWARF (1997)
AWWARF (1997)
Lowryetal. (1987)
Kinneretal.(1988)
Lowry and Brandow (1985) showed that radiation at the surface of the household contactor
vessel decreased to less than 1.0 mrem/hr at 3 feet from the tank surface.
Some studies using field observations and models have investigated the exposure due to
gamma radiation from GAC beds. Based on data collected during a study of GAC household POE
units, Lowry (1988) developed a relationship to predict the maximum gamma exposure rate based
on the radon level in the raw water. Equations that model the relationship between acceptable
i
distance and raw water radon level were developed. The maximum gamma exposure rate that can
be expected can be predicted by the relationship that 1.0 mrem/hr is produced for each 10,360 pCi/L
of radon in raw water and applies for GAC units that have reached a steady-state operation. (Lowry,
1988)
Analyses and conclusions/recommendations by researchers are listed below:
A GAC unit located 3 feet from a living area would require no shielding up to an
influent radon level of 21,000 pCi/L (Lowry, 1988).
There is no increased gamma exposure if raw water is less than 5,000 pCi/L and
GAC is adequately shielded.
2-76
May 1999
-------
• A GAC unit in a cellar needs no special precautions up to an influent radon level of
30,000 pCi/L.
Based on their analysis, researchers recommended an upper level of approximately 5,000
pCi/L for influent radon concentration to avoid hazard due to gamma radiation for a domestic water
supply. The analysis took into consideration the standard promulgated by EPA which limits
residential exposure to gamma radiation to 20 uR/h above the background exposure rate, or
approximately 170 mrem/year; an exposure period of 8 hours per day; and a safe distance of a few
feet from the surface of the GAC unit. It was estimated that the exposure from a unit that removes
95 percent of the influent radon concentration of 5,000 pCi/L at a flow rate of 300 gpd was less than
0.058 mR/hr (equivalent to 170 mR/yr) at a distance of approximately two feet from the GAC unit
(Rydell, 1989).
Based on research studies, AWWARF (1997) verified the following equation (developed by
Lowry and Brandlow, 1985) to provide an estimate of the gamma emissions for any given radon
influent concentration.
T =-
max
Rn
17.8
pCi/L
urem/hr
where:
Rn
= maximum gamma emission (urem/hr)
= influent radioactivity (pCi/L).
A computer program called CARBDOSE models radon removal by a domestic style GAC
filter (Rydell et al., 1989). The program can estimate the exposure dose from a GAC unit at a
specified distance from the unit. It also models the accumulation of Pb-210 on the media as a
function of years of operation, and estimates the Pb-210 activity per gram of wet carbon. Martins
(1992) developed graphs that depict the exposure rate versus the radon influent concentrations for
various system configurations and sizes.
2-77
May 1999
-------
The gamma radiation exposure rate to operators in the work area depends on several factors
including radon concentration, distance between the work area and the GAG units, time spent in the
work area, and the extent of shielding. The exposure rate can be reduced by shielding with materials
such as lead or water, and engineering controls that maximize the distance between the beds and
workers, and minimize the tune spent by workers near the GAG units. Possible measures to reduce
I
radon exposure by workers include:
• Automate the treatment system
• Install remote instrumentation and a remote control center
i i
• Implement a radiation protection program that establishes operator exposure time
management and measures radiation exposure (time logs, dosimetric badges, periodic
Geiger-Mueller counter tests)
1 ! ' l
• Add vessel shielding, such as lead liners or water jackets around the GAG system
• Provide physical barriers such as caging around the contactors at one meter distances to
prevent casual contact and therefore limit unnecessary body exposure
1 i i
• Change site configuration to maximize distances between vessels or parallel trains
• Add a storage silo for offline storage of spent carbon (option for waste treatment)
• Improve ventilation/dispersion (Martins, 1992).
2.3.5.2 Spent GAG Disposal
A GAG bed used for radon removal can last for many years with little decrease in efficiency,
assuming no limiting water quality conditions exist. Radionuclides (Pb-210 from the decay of radon
and from the source water, and uranium and radium from the source water) accumulate at the GAG
media. The presence of iron increases the ability of radium and uranium to adsorb to the media since
iron is reactive with these compounds (AWWARF, 1997). Depending on the contaminants and the
extent of accumulation, the disposal of the spent carbon containing Pb-210 and/or other
contaminants could pose problems.
2-78
May 1999
-------
Currently there are no Federal regulations governing the disposal of radioactive wastes
generated by water treatment facilities. The US Nuclear Regulatory Commission's (USNRC)
definition of low level radioactive wastes is not based on the level, of radioactivity but on how the
material was generated. That agency does not regulate naturally occurring radioactive material
(NORM). EPA's Suggested Guidelines for Disposal for Drinking Water Treatment Wastes
Containing Radioactivity (U.S. EPA, 1994) and Management of Water Treatment Plant Residuals
(U.S. EPA, 1996a) review disposal options and recommend disposal criteria. EPA's guidelines are
summarized in Table 2-17.
Table 2-17. EPA Guidelines for Disposal of Radioactive Water Treatment Plant Residuals
Radionuclide level (pCi/g)
.. Radium
<3
3 to 50
50 to 2,000
>2,000
Uranium 238
<30
30 to 75
75 to 750
>750
Recommended Disposal
Landfill - waste should be dewatered and mixed with other wastes.
Covered landfill to prevent release of radon to air.
Dewatered and disposed in RCRA hazardous waste facility.
Handle on a case-by-case basis.
RCRA hazardous waste facility.
Low level radioactive waste facility.
As permitted by State regulations.
Low level radioactive waste facility.
Source: "Assessment of GAC Adsorption for Radon Removal," (AWWARF, 1997).
AWWARF (1997) notes that most States deal with the disposal of radioactive water
treatment plant residuals on a case-by-case basis. The States have no specific regulations or
guidelines for these radioactive residuals but would apply existing solid waste or hazardous waste
disposal requirements. New Hampshire requires that wastes containing radium at 0.444 pCi/g or
uranium 238 at greater than 58.4 pCi/g be disposed of in a low level radioactive waste facility.
Illinois licenses water treatment facilities and landfills receiving radium bearing sludges as radiation
installations. Sludges containing less than 5 pCi/g total radiation can be disposed of in a permitted
landfill. Sludges with radioactivity levels between 5 and 50 pCi/g can also be disposed of in a
permitted landfill but under more stringent requirements. Sludges with radioactivity levels greater
2-79
May 1999
-------
than 50 pCi/g are handled on a case-by-case basis by the Illinois EPA and the Illinois Department
of Nuclear Safety. (AWWARF, 1997)
(
AWWARF (1 997) provides a method for predicting the accumulation of Pb-2 1 0 on the GAC
using the following simplified equation:
(-2.57 xlO"1) C K
^ ' r s
Pb-210 radioactivity/g of carbon after year 1 =
I C
--
where:
Cr = radon removal (pCi/L)
Ka = rate constant for a given GAC (1/hr)
d = density of GAC (lb/ft3)
C0 = influent radon concentration (pCi/L)
C, = effluent radon concentration (pCi/L).
As indicated hi the AWWARF report (1997), this equation can be used as a general
indication of Pb-210 buildup. The AWWARF (1997) report shows a set of Pb-210 buildup curves
for different K^ values. The above constant of 2.57 x 10"' is believed to be the result of
multiplication of a unit conversion factor of 546.92 and a curve slope value of 0.47 x 10'3 for a K^
!
value of 4.5/hr.
Hess, et al. (1999) studied the effects of washing GAC in acid reagents and bases to
determine whether they could be used to regenerate carbon. The GAC filters were used to remove
radon from well waters with up to 100,000 pCi/L. Carbon used as radon filters was washed with
either hydrochloric acid, nitric acid, acetic acid, EDTA, sodium hydroxide, potassium hydroxide,
or distilled water. This series of washes was found to remove many of the radionuclides that had
built up in the GAC as a result of the decay of radon that had adsorbed during use as a filter to
2-80
May 1999
-------
remove radon from drinking water. Pb-210 was reduced by factors ranging up to 65%, but in some
cases also increased by as much as a factor of 2 due to a reduction in the mass and density of the
sample. Thus, washing GAC in certain acids and bases may reduce the levels of Pb-210 and other
radionuclides in spent GAC to levels not requiring special waste disposal. The researchers did not
indicate whether spent acids and bases would require special waste disposal arrangements. (Hess,
etal., 1998)
2.3.6 Case Studies
Full-scale studies to date using GAC for removing radon have involved central treatment
systems and point-of-entry (POE) treatment devices. The majority have been on POE.
ML Vernon, New Hampshire—A GAC system for radon removal was installed in Mt.
Vernon, New Hampshire by Dr. Jerry Lowry. The water supply consisted of two wells which
supplied approximately 6,500 gallons per day to 40 mobile homes. The concentration of radon in
the water averaged 155,000 pCi/L (Lowry, et. al., 1984)
The treatment system consisted of two GAC contactors connected in series. The contactors
contained a total of 48 cubic feet of carbon which provided an empty bed contact time for each unit
of 80 minutes. After achieving steady state operations in 10 to 15 days, the levels of radon in the
treated water ranged from 4,000 to 15,000 pCi/L. This corresponds to a removal efficiency of 90.3
to 97.4 percent. Following 2 months of operation, major leaks developed in the distribution system
that resulted hi the wells being pumped dry several times. Following repairs, the levels of radon in
the treated water rose to 22,000 pCi/L which corresponds to an 85.8 percent removal efficiency. The
reasons for the decreased removal efficiency were unknown. However, one possibility was that
sediments containing radium may have been deposited in the contactors.
Deny, New Hampshire—Kinner et al. (1990) conducted a pilot scale study at Deny, NH
using two different GAC systems applied at the point of entry (POE). The first system consisted of
a sediment filter with pleated paper for pretreatment and a GAC unit with 0.047m3-of Bameby
Cheney 1002 coconut-based carbon. The second system consisted of a sediment filter with pleated
2-81
May 1999
-------
paper, an ion exchange unit with 0.042m3 of strong cationic resin and a GAG unit with-0.0465m3 of
1002 carbon. Pretreatment and/or backwashing was applied to prevent fouling. The first system
(with cation exchange pretreatment) achieved removals ranging between 85 and 99.7 percent, while
the second system (without cation exchange pretreatment) achieved removals ranging between 79
and 99.7 percent. The systems were operated at a flow rate of 270 gpd at six 18-min intervals and
a 30-min interval each day.
Mont Vernon, New Hampshire—Kinner et al. (1988), performed a full-scale study using
a GAG system designed by Lowry Engineers. The system consisted of two filters operating in
series. The first filter (30" diameter) contained 20 ft3 of Barneby Cheney 1002 coconut-based
carbon. The second filter (36" diameter) contained 27 ft3 of carbon. The system achieved removals
I I
ranging between 74 and 88 percent. This decreased efficiency was attributed to clogging possibly
i i I
from iron, manganese, bacteria or organics. The system was operated at an average flow of about
40,700 gpd with influent radon levels of about 191,000 pCi/L. Backwash was provided at 10 gpm.
Spent carbon was shipped to an approved low level radioactive waste landfill in the western United
States.
Leed, Maine—Lowry et al. (1990) performed a full scale GAG adsorption study in Leed,
Maine. Influent radon levels were 1,124,000 pCi/L and the average flow was 886 L/day retained
I i
over 22 months. A standard GAG setup was used for the study. During adsorption, 99.5 percent
radon removals were achieved while all progeny, Po-218, Pb-214, Bi-214 and Pb-210, were retained
on the bed. Desorption occurred only when Rn-free (< 1 pCi/L) water with pH < 3.0 was applied
to the bed. One hundred percent of the progeny was retained in the bed at pH > 3.0. It was
concluded that optimal elution occurs when pH is between 2.0 and 3.0.
Hodsdon (1993) conducted a field survey of radon treatment facilities using GAG. The
facilities surveyed included very small to medium sized water systems. A summary of the findings
are presented hi Table 2-18.
2-82
May 1999
-------
Point-of-EntryUnits—Lowry et al. (1989) summarized data obtained from various full-scale
point of entry (POE) GAC studies conducted between 1984 and 1989. The total number of selected
units was 121 distributed in several states as shown hi Table 2-19. Typical GAC systems were used
for the studies. These systems included pump/hydropneumatic pressure systems, sediment filters,
manual control valves, bottom injection flow inlets, and pressurized GAC treatment units with 6
inches of support gravel and 36 inches of GAC depth. However, different sizes of GAC units were
used as shown in Table 2-20.
2-83
May 1999
-------
Cv
S,
o
•a
en
•S3
C
S
ss
s
03
CO
CQ P-, £J
o to
^-^ ~^ 3
•a O
< s §
t» ^__ CO
1 1 1 1
S S « ts
O cs I I
u
.
.S"
j=
en
T3
C
.H
*j™^
£
o
0
o
o"
1
1
«-"
u
e
<*.
o
t;
•o S.
Avg daily use 2,000 gpd
- max capacity 43,200 gp
Radon removal system is
groundwater supply
T}*
£:
u
4-4
ts
«t"i =
3 S- O
C g- _ 0
2 S g rs
O 1 oo 1
U
S
T3~
"u
T3
03
U
if
^"
—
^_^ gj
"2 3
so u
o °°
Avg daily use 3,000-4,00
Sodium hypochlorite
Pretreatment - ion exchan
0
VO
M
II
•0*
C
u ^
ill
1
C CO
0 U
-O c «*•
c5 I. °
O ao "^^ O
« .t; CN
ranular Activ
supply capaci
contactor w/
0 I T
0
O
o
U
0
o
m"
LlJ
a
<2
I
~^
CQ
C/l
73 -S
Avg daily use 280,000 gp
Chlorine
GAC contractors initially
removal
•J
^ * .0 G
0 g g Q.
, 1 I O
1 . o
— fc to oo
.5 >» 1
il .£ A* CN
o
(N
o
OO CQ
C — . [|j|1
^ oo -c
u >, w
w .ts £
•s a 2
4? S-s
^ 0 0 JB
L. •*. ca u
JS ^ c *
= g. § U
C £• O 03
2 i
^
00
> cs
ei
1 ^i.
CN" O
2-84
May 1999
-------
Table 2-19. Categorization of POE GAC Units by State
Categorization of POE GAC Units by State
State
Maine'
New Hampshire
New Jersey
Kentucky
Pennsylvania
Massachusetts
Number
61
20
12
1
6
5
State
Colorado
Rhode Island
Connecticut
New York
North Carolina
Vermont
Number
3
3
6
1
1
1
Table 2-20. Relative Use of Different Sized GAC Units
GAC Model
Not designated
GAC10
GAC 17
GAC30
GAC (cu ft)
2.0/3.0
1.0
1.7
3.0
Vessel Size
12" x 48"
10" x 35"
10" x 54"
14" x 47"
Number Installed
12/3
15
72
16
Note: Bameby Cheney Type 299 coconut-based carbon was used in Models 10,17, and 30. Calgon F-400 was used in 11 of the
12 (2.0 cu ft) not designated units. Norit was used in 1 of the not designated 2.0 cu ft units and in the 3.9 cu ft not designated units.
From the 121 selected installations, 64 percent acliieved removals greater than 95 percent,
30 percent achieved removals between 80 and 90 percent, and 6 percent experienced premature
failure that is believed to be water quality related. Based on these results, it was concluded that in
order to meet maximum contaminant levels (MCLs), the POE GAC application should be limited
to well supplies with Rn <. 5,000 pCi/L.
2-85
May 1999
-------
In addition, a more specific evaluation of selected full-scale units was conducted. Eleven
locations were selected from the 121 based on the following criteria:
1. All installations had a long period of service.
2. One or more had several performance checks.
3. One or more had particularly water quality problem other than Rn.
4. One or more showed a progressive premature failure.
The removals ranged between 90 and 99 percent except for one installation where the
removal ranged between 30 and 60 percent.
2-86
May 1999
-------
Best Management Practices
3.0 INTRODUCTION TO BEST MANAGEMENT PRACTICES
Best management practices (BMPs) are combinations of activities or modifications to
treatment processes that under certain conditions will prevent the use of water sources high in radon
levels and achieve acceptable water quality.
3.1 DESCRIPTION OF PRACTICES
3.1.1 Geologic Controls (Siting)/Alternate Sources
Development of alternative water sources located within a reasonable distance of a
community which do not exceed the MCL for radon may provide a satisfactory solution to a
community water quality problem. Although surface water sources may be free of radon, they may
require treatment, such as clarification and disinfection, which may be as expensive as removing
radon from the original source water. However, the radon contaminant in the original source may
be peculiar to a localized geological formation and developing wells of different depths and at
different locations should not be precluded.
Many communities may have existing facilities that can be utilized in the development of
an alternative source. For example, systems with filtration or chlorination processes may adapt those
processes to treatment of a new water source. Different combinations of new and existing processes
relative to the size of the system and quality of water source impact significantly on cost and
complexity.
3.1.2 Regionalization
A feasible option, especially for small water systems that are out of compliance with the
radon MCL requirements, is to join with other small or Large systems that do comply with MCL
requirements to form a regional water supply system. A schematic of such a system is shown as
3-1
May 1999
-------
Figure 3-1. This differs from alternate source development in that the basic water source has
already been developed by the host community.
Generally it is the smaller communities or systems that may encounter radon concentrations
in excess of the MCL because these systems tend to use ground water sources. Ground water
. j
sources are used by 91 percent of the systems serving under 500 people and 74 percent of the
systems serving 501-3,300 people (NRC, 1997). Small systems also tend to have more operational
reliability problems than do the larger systems (U.S. EPA, 1987). The cost of treating poor quality
ground water and the technical difficulties that many small systems face make regionalization an
appealing idea. With centralized treatment, treatment costs (capital and O&M) are frequently much
lower than decentralized treatment. Regionalization can also provide other cost savings through
shared maintenance service and/or central billing, which show economies of scale.
1 j
The cost of regionalization is independent of the particular contaminant being considered
since regionalization involves supplanting the contaminated water supply with water from a
(presumably larger) host community. A community may arrange for more than one host community
to supply it with water. However, if multiple transmission pipelines are involved, this may increase
i
costs. Since regionalization is one community arranging with a neighboring community to supply
it with potable water, the distance between communities has shown to be the most sensitive factor
affecting regionalization costs in this analysis. Piping treated water may prove prohibitive in
sparsely populated regions. Regionalization requires a feasibility and economic study, preferably
one that includes value engineering. A regionalization plan should include alternative solutions with
cost comparisons such as pipe laying costs above ground versus below ground (where feasible), the
use of one large pumping station versus several smaller pumping stations, and the use of multiple
elevated reservoirs versus one large ground reservoir.
3-2
May 1999
-------
Figure 3-1. Regionalization
-------
Griffin (1989) noted the advantages of regionalization as follows:
• Cheaper treated water due to economics of scale
• Dependability of supply
• Longer life expectancy of pipelines than water treatment plants
i i
• Long term economic benefits because of better quality, steady water supply, and less
spending on water treatment.
Griffin (1989) also lists problems associated with regionalization:
• Requires the cooperation of individuals and groups with different mentalities, needs, and
economic capacities
• May mean the abandonment of existing treatment facilities
• Challenges hi places where water rights are an issue and water is scarce.
3.7.5 Extended Atmospheric Storage
Extended storage allows sufficient time for some radon reduction to occur through decay and
losses to the atmosphere. During extended storage, the radon-contaminated water can be exposed
to the atmosphere so that the contaminant may be naturally transferred to the atmosphere. Storage
may be provided by reservoirs or water tanks, though water tanks would be easier to monitor for
percent removals. Removals can be increased in the tanks by providing ventilation through the tanks
to minimize the buildup of radon, which decreases removal efficiencies by hindering the natural
diffusion process. The dimensions of the tanks also affect removal efficiencies, though more studies
will be required to determine the removal efficiencies and costs of shallower tanks with large surface
areas for greater contact with the atmosphere versus deeper tanks with the same storage capacity.
Advantages of storage can be summarized as follows:
5-4
May 1999
kill;, 1,91
-------
• Provides radon treatment with and without ventilation or aeration because of natural
decay
« Provides continuous water supply when key treatment units are out of service and during
. emergencies
• May provide savings on energy consumption by treating and storing water during off-
peak energy rates.
Disadvantages associated with storage are included below in Section 3.1.5.
3.1.4 Blending
Many systems rely on ground water and surface water sources. In many cases, ground water
is used during peak seasons or during dry years. For system with access to surface water, blending
might be a viable option for reducing radon levels in drinking water, particularly for small systems.
Since radon concentrations in surface water are typically very low, blending surface water with
ground water will decrease, often significantly, the concentration of radon in the water supply. In
addition, blending a low radon ground water source with the source high in radon is an option.
3.1.5 Limitations of BMPs
Although the above best management practices (BMPs) achieve some degree of removal, the
following barriers limit their implementation in full-scale 'treatment:
« Existing storage tanks may not provide adequate headroom and ventilation to prevent the
accumulation of a gaseous layer of radon, which can re-enter the aerated water. Storage
must be carefully managed to avoid water stagnation. Deterioration of water quality can
be dramatic otherwise, especially in the summer.
• Modifications to storage facilities may interrupt service.
• These systems will probably not achieve removals that are sufficient to attain desired
water qualities, if influent levels are extremely high.
• Removals will not be consistent, depending on seasonal usage patterns and radon
occurrence levels.
3-5
May 1999
-------
The capacity of existing storage tanks will probably be insufficient to provide the
necessary detention time to reduce radon concentrations to acceptable levels if high
influent levels are present.
The use of other sources, such as blending a surface water source with an existing ground
water source, may reduce radon levels but may introduce other contaminants that require
the addition of other treatment. Additional monitoring may be necessary to track source
water quality, thus increasing costs.
Additional transmission pipelines or pumps may be needed (e.g., storage systems may
operate at atmospheric pressure, requiring either elevated tanks or repumping).
Water quality may be less consistent (e.g., if storage times fluctuate greatly based on
demand or if the quality of a blended source is not as reliable).
Disinfection may be needed (e.g., if water is exposed to the atmosphere or blended with
a surface water source).
3.2 REMOVAL EFFICIENCY
Limited data are available on removal efficiencies for the BMPs described above. These data
are shown in Table 3-1.
Table 3-1. Removal Efficiencies for BMPs
Best Management Practice
Atmospheric Storage
Blending (6.34 MG of well water with 18.34 MG surface water)
Effect on Radon Level
7-13% removal - 9-hr detention time
33-36% removal - 30-hr detention time
79% reduction
3.3 DESIGN CRITERIA
The half-life of radon is 3.82 days. Therefore, to achieve a minimum of 50 percent removal
of radon, a 4-day storage capacity is needed (if radon removal occurs by decay alone). While this
might seem to be cumbersome for large systems with available sources of ground water, large and
small systems with intermittent flows and fluctuating demands may find storage to be the best
technology.
3-6
May 1999
-------
3.4 TREAT ABILITY/CASE STUDIES
Storage: Kinner et al. (1987) performed a laboratory study at the University of New
Hampshire to monitor radon removal from a still pool of water to determine the effect of storage.
A plastic storage tank with a capacity of 30 gallons wasfilled with water containing radon to a depth
of 27 inches. The radon concentration was monitored for 5 to 6 days during four test runs. The
results of this bench study are presented in Table 3-2.
Table 3-2. Bench Studies
Detention Time (hr)
9
30
Percent Decay™
7
20
- Percent Removal
7-13
33-36
(1) The values for percent removal due to decay were calculated assuming first order decay and a radon half-life of 3.82 days.
The results of the study were used to develop design criteria for storage tanks. Although it
is difficult to scale-up the removals cited above, the study determined low level radon removal may
be attainable through the installation of storage tanks.
Blending: Based on volumetric dilution, Dixon arid Lee (1987) blended 6.34 MG of well
water with an average radon concentration of 1079 pCi/L and 18.34 MG of surface water with no
radon. The average radon concentration in the blended wetter was 226 pCi/L.
3-7
May 1999
-------
-------
Treatment Cost Analysis for Radon Removal Technologies
4.0 INTRODUCTION
This chapter presents capital and operation and maintenance (O&M) costs for treating radon
in drinking water. Costs are presented for aeration technologies, granular activated carbon (GAC),
pre-engineered centralized treatment devices, and regionalization. The cost equations and point
estimates of costs presented in this chapter are developed to be used in the regulatory impact analysis
process to estimate regulatory compliance costs for the prospective Radon Rule. Section 4.1
describes the cost estimating approach. Section 4.2 presents capital and O&M costs (in the form of
summary cost tables) for aeration and GAC. Section 4.3 presents costs for inter-connecting small
public water supplies (regionalization) as an alternative to installing treatment devices. Section 4.4
presents equations for estimating capital and O&M costs for pre-engineered centralized treatment
devices for systems with very small flows (less than 10,000 gpd). Section 4.5 presents a
comparative analysis of the aeration and GAC costs using case studies and other cost-estimating
models. Section 4.6 presents the best-fit equations for aeration and GAC developed in Sections 4.2
through 4.4 in a summary form.
4.1 DESCRIPTION OF COST ESTIMATING APPROACH
For aeration and GAC, point estimates of capital and O&M costs for a series of plant sizes
ranging from 10,000 gallons per day (gpd) to 100 million gallons per day (mgd)9 were developed
using cost estimating software. Costs were also estimated for aeration technologies for small plant
sizes (0.1 to 2 mgd) using current cost estimating literature and according to generally recommended
engineering practices. Since Radon's compliance costs could be borne primarily by small systems,
the alternative costs using this approach provide another resource for analysts for estimating
regulatory compliance costs. In addition to the basic technology costs for aeration and GAC, costs
9 System sizes and flow ranges used in this costing exercise apply primarily to community water systems
(CWS).
4-1 May 1999
-------
were also developed for indirect cost elements using various cost-estimating tools. Indirect elements
il,
include costs for structures, land, permitting, disinfection, and pretreatment.
I
Cost estimates for treatment units for capacities less than 10,000 gpd were obtained from the
I
published literature on point-of-entry (POE) systems.10 For centralized treatment of radon using
POE devices, capital and O&M costs were estimated using the report entitled "Cost Evaluation of
Small System Compliance Options: Point-of-Use and Point-of-Entry Treatment Units. EPA. Draft.
April 20, 1998." Costs for regionalization were based on best professional judgment and using
published unit cost data.
The following sections provide more detailed descriptions of the cost approach, cost
estimating models, model inputs (design parameters and cost factors) and assumptions used for
estimating capital and O&M costs.
;l
1 1
4,1.1 Description of the Approach Used for Estimating and Validating PTA and GA C Costs
Specifically, the following approach is used for estimating PTA, DBA and GAC costs for
plant sizes greater than 10,000 gpd:
• Available design and cost estimating models for PTA and GAC were modified and
adapted for estimating radon treatment costs.
i
• Point estimates of capital and O&M costs were generated for 11 plant sizes ranging from
10,000 gpd to 100 mgd. The plant sizes were selected to provide representation across
small and large water treatment plants.
• Indirect capital costs for permitting, structures, piloting, land, pretreatment, and post-
treatment were also developed.
• In addition to model-based costs for PTA, costs were estimated for small systems using
an alternative approach.
10 POE technologies can be used by very small community water systems (C WS) or transient and non-transient
non-community water supply systems (TNCWS and NTNCWS).
4-2
May 1999
-------
• Best-fit curves were fitted to the data points and cost equations (flow versus capital and
O&M costs) were generated.
• Costs were compared to costs developed using other models and to case studies.
4.1.2 Description of the Cost Estimating Models For PTA and GAC
Two models, the PTA-COST and the GAC-COST, were used to estimate centralized capital
and O&M costs for PTA and GAC, respectively11. Both models were developed by EPA but were
modified by S AIC for estimating Radon treatment costs. Costs from other models, historically used
by EPA for generating unit treatment costs, were also used but only for comparison with PTA-COST
and GAC-COST estimates.
4.1.2.1 The PTA-COST Model
The PTA-COST Model was developed by EPA in the early 1980s for generating cost-
optimized designs for counter-current packed tower aeration systems for removing volatile organic
compounds (VOCs) from water. The PTA-COST Model was designed to generate process designs
and costs for small to large PTA configurations. The PTA-COST Model has been modified several
times over the past decade and was used most recently in 1992 for estimating radon treatment costs.
The PTA-COST Model (the 1992 version), written in an old version of the Hewlett Packard
Basic language, was translated to a spreadsheet-based program and its design routines were modified
to produce realistic PTA designs. These modifications include constraining the program to base
costs on realistic tower dimensions (the 1992 version generated unrealistic tower dimensions) and
simplifying the way the model checks for flooding conditions.
Significant changes were not made to the cost estimating routines. Data are not available to
reliably change the equations in the model that describe the base capital and O&M costs. Therefore,
11 Note that in this report, the PTA costs are assumed to also represent costs for technologies that are similar
• in performance and costs to PTA. These technologies include multi-staged bubble aeration (MSBA) and shallow tray
aeration (STA). Costs for Diffused Bubble Aeration (DBA) were estimated separately.
4.3 May 1999
-------
the PTA-COST Model's estimated costs are validated by comparing them with costs from case
studies and outputs from other models. While changes were not made to the underlying cost-
estimating routines in the PTA-COST Model, cost indices and the labor rate were changed to update
model costs to December 1997. The indices and labor rates used are discussed later in this section.
i
The design generated by the PTA-COST Model (1998 version) is for a system with a steel
tower and internals, plastic packing, an air blower, influent pumping, clearwell (whose size can be
adjusted based on detention time), piping, instrumentation, and electrical and air duct. The PTA-
COST Model design does not consider a building (for housing pumps and other support equipment)
or a chemical wash system. Costs for additional items are estimated separately, (see Section 4.2).
I '
O&M costs generated by the PTA-COST Model include labor (mainly for taking readings),
administrative costs, and power for pumps and the blowers. Appendix A-0 presents a conceptualized
diagram of the assumed configuration.
4.1.2.2 The GAG-COST Model
EPA developed the original GAC-COST Model written in FORTRAN computer language
in the late 1980s. EPA's application of this model is documented in a report entitled "EPA's
Drinking Water and Groundwater Remediation Cost Evaluation: Granular Activated Carbon." The
original GAC-COST Model was updated, converted to a spreadsheet, and additional automation
added to simplify use. The current model performs GAG design calculations (e.g., it calculates
empty bed contact time, contactor volume, and quantity of GAC) based on user supplied design
criteria (e.g., design and average flow rate, percent radon removal, number of contactors, and
loading). It then estimates capital and O&M costs for these designs.
j
For small systems, the model assumes that systems will use preengineered (package)
treatment units. For larger systems, the model automatically selects the lowest cost of either a steel
pressure contactor design or a concrete gravity contactor design. For these systems, capital costs
include construction costs, backwash pumping, and the initial GAC load. Further details of the
4-4
May 1999
-------
elements included in construction costs are not provided in the original model documentation.
However, given the similarity in capital cost found between the GAC-COST Model and other
models (see Section 4.2), the GAC-COST Model is believed to include the same basic capital cost
elements associated with "standard" GAC units. O&M costs in the model include labor, power,
maintenance, process and GAC transport , GAC replacement, and spent carbon transport and
disposal.
Construction costs generated by the model can be updated using cost indices. Cost factors
(e.g., engineering, construction), used for estimating total capital costs, can also be adjusted by the
user. O&M costs can be updated using the Producer Price Index, Bureau of Labor Statistics (BLS)
labor rates, and water and energy unit costs.
4.1.2.3 Other EPA Models
EPA has traditionally used three models for estimating drinking water treatment technology
costs. They are the Very Small Systems Model (VSS), the WATER Model, and the Water/
Wastewater (WAV) Cost Model. These models are capable of estimating capital and O&M costs for
aeration and GAC as well as numerous other technologies. Use of a particular model is predicated
on the size of the treatment system being costed.
These models are used only for making comparisons with the PTA-COST and GAC-COST
Model outputs. Although they offer breadth in terms of the number technologies they cover, they
have several limitations. First, the models cannot generate designs. Designs have to be generated
separately by hand or by using another model. Second, the input design parameters for these models
have to be within the original design constraints on which these models' cost equations are based.
Thus, costs cannot be linked directly to a removal efficiency. Finally, these models are several years
older than the unmodified PTA-COST Model and the GAC-COST Model. Despite these limitations,
these models, provided they are used with care, are still good indicators of generalized technology
costs and are generally sufficient for regulatory-type cost analysis or as a comparative tool.
4-5
May 1999
-------
The main features of the VSS, Water, and W/W Costs Models are described below.
The VSS Model
The VSS is a spreadsheet containing tables for 18 water- treatment technologies. The
spreadsheet was developed by SAIC using cost equations presented in an EPA Manual entitled
"Very Small Systems Best Available Technology Cost Document, September 1993." Each table in
the VSS spreadsheet corresponds to a treatment technology. Users input a public water system's
design and operating flow, an escalation factor (e.g., average CCI increase), and the treatment plant's
design and average flows . Capital and O&M costs are generated by the spreadsheet using the
underlying cost equations. The VSS Model is applicable for treatment plants with design flows from
10,000 gpd to about 270,000 gpd.
The WATER Model
The WATER Model is also a spreadsheet program based on cost estimating equations
presented in the EPA Report entitled "Estimation of Small System Water Treatment Costs 1984."
The WATER Model can estimate capital and O&M costs based on user supplied design criteria and
cost indices for 45 unit processes. Costs can be generated for a single process or a combination of
processes (a treatment train). The costs in the WATER Model are applicable for plant sizes from
15,000 gpd to 1 mgd.
Construction cost data in the WATER Model are based on unit equipment cost data supplied
by manufacturers, cost data from actual plant construction, unit takeoffs from actual and conceptual
designs, and published data. Operation and maintenance requirements are based on operating data
at existing plants, BPJ, and information from equipment manufacturers.
j
|
Construction costs generated by the WATER Model can be updated using cost indices.
Engineering cost factors, used for estimating total capital costs, can also be adjusted by the user.
O&M costs can be updated using labor indices, labor rates, and chemical costs.
4-6
May 1999
-------
The W/W Costs Model
The WAV Costs Model is a computerized DOS-based model and is applicable to large plants
(1 to 200 mgd) but is also capable of estimating costs for package plants for some technologies (e.g.,
GAG). Version 2.0 of the WAV Costs Model was developed in 1994 by CulpAVesner/Culp; an
engineering consulting firm, based on various information sources including an EPA Report titled
"Estimating Water Treatment Costs. Cost Curves Applicable to 1 to 200 mgd Treatment Plants,
Volume 2, August 1979." This report provides conceptual designs and cost curves for 99 unit
processes.
The WAV Costs Model allows the user to select from among 149 individual unit processes.
About 90 unit processes are solely applicable to drinking water treatment. Similar to the WATER
Model, the W/W Costs Model can be used to estimate costs for a single unit process or a
combination of unit processes. Key inputs include construction cost indices (e.g., Engineering News
Record indices), engineering cost factors, labor rates, and chemical costs.
4.1.3 Case Studies
Costs reported in case studies from several different sources and costs from other models
were used to validate the costs generated by the PTA-Model and GAC Model. These case studies
encompass projects for small to medium water treatment plants that have implemented aeration to
treat radon and other VOCs. A brief description of each of the sources and general information vis-
a-vis each case study is presented below.
A.E. Hodsdon Engineers Report: This document is an unpublished report entitled "Field
Verification of Radon Treatment Costs for Very Very Small to Medium Sized Water Systems"
prepared by A.E. Hodsdon Consulting Engineers for the A WWA. The study collected costs incurred
by utilities to install various types of radon treatment systems including PTA, STA, DBA, and GAC.
Data was collected via a questionnaire which requested a breakdown of costs. Only the raw data
from this document was used for comparative analysis.
4-7
May 1999
-------
Operating Experiences at VOC Treatment Facilities: This report was prepared by
Malcolm Pimie, Inc. and presents process performance and costs for air stripping and GAC at
facilities treating volatile organic compounds (VOCs). Since the same types of technologies that
remove VOCs are applicable for Radon, the costs presented in this report were used.
I i
Evaluation of Full-Scale Treatment Technologies at Small Drinking Water Systems,
Summary of Available Cost and Performance Data: This document is a recently developed
(1998) reference tool that contains cost and performance data for treatment technologies applicable
to small drinking water systems.
MSB A Information from Lowry Engineers: Information was also obtained from
I
memoranda from Lowry Engineers on the performance and costs of multi-staged bubble aeration
(MSBA). Lowry Engineers have extensive experience in implementing aeration technologies for
I
Radon removal.
Wright-Pierce Engineering Case Studies: Cost data for installed multi-stage bubble
aeration was provided via personal communication to EPA by Wright-Pierce Engineers.
I
Assessment of GAC Adsorption for Radon Removal: This report was prepared by the
American Water Works Association Research Foundation (AWWARF) and is a research effort
aimed at determining radon removal efficiency of different types of GAC. Cost data are also
provided for hypothetical treatment plants.
American Waterworks Service Company: This report entitled "Pennsylvania-American
Water Company, Eastern Region, Comprehensive Planning Study, 1990" details the costs of
upgrading a ground water treatment facility in Frackville, Pennsylvania. Costs are included for
various upgrades including a PTA system for VOC removal.
4-8
May 1999
-------
AWWARF: This report "Critical Assessment of Radon Removal Systems "for Drinking
Water Supplies, 1998" The report critically compares old EPA cost estimates with AWWA cost
estimates and provides cost estimates from case studies.
4.1.4 Costing ofPTA and DBA Units Using Direct Engineering Costing Methods
Since small systems may be impacted the most by the Radon regulation, a separate cost
estimate for PTA was developed for small plants using best professional judgment and the cost
estimating literature such as R.S. Means publications. Also, this cost estimate is based on using
alternative materials of construction to those dictated by existing configurations. Costs were also
estimated for DBA. Diffused bubble aeration is a technology that can be retrofitted into existing
water treatment plants. It offers another alternative to installing a full-fledged PTA or similar unit
and can provide a facility with a low-cost yet effective Radon control technology.
4.1.5 Assumptions for Engineering Cost Factors and Other Costing Inputs
In November of 1997, EPA convened the Technology Design Panel (TDP) at an EPA-
sponsored workshop, in Denver. The TDP was comprised of experts from the drinking water field.
They provided recommendations on a host of issues vis-a-visEPA's regulatory cost analysis process.
Among the issues discussed was the need to bring better consistency in assumptions for estimating
costs. Therefore, the assumptions for cost-related items in this report are based largely on the
discussions among these experts which are embodied in two documents. These documents are:
"Discussion Summary: EPA Technology Design Workshop." November 6 to 7, 1997 Denver,
Colorado and the "Technology Design Information Package." November 1997.12 Each of the cost
elements is discussed below.
Engineering Cost Factors
12 In many cases, the TDP did not provide specific recommendations. They provided ranges of expected values
for various cost items. In cases where ranges were provided or in areas where they disagreed, the costing assumptions
were based on an interpretation of the TDP's intent.
4-9
May 1999
-------
Based on an EPA Manual entitled "Innovative and Alternative Technology Assessment
Manual," total capital costs are comprised of construction and other non-construction costs.
Construction costs include the installed cost of components, miscellaneous structures, piping,
instrumentation and site preparation. Non-construction costs include engineering design fees,
overhead and profit, and contingencies. Non-construction costs are estimated based on assumed
percentages. These percentages are known as engineering cost factors. For example, engineering
design fees, which are based on the complexity and scale of a project, range from 4 to 15 percent of
construction costs.
In the past, engineering cost factors in EPA's Technology and Cost (T and C) documents
were based on best professional judgment and the literature. To promote better consistency in the
application of engineering cost factors, the TDP made some recommendations (see Table 4-1).
As Table 4-1 shows, the TDP left
some leeway for interpretation of their
guidelines and this is appropriate because the
complexity of the technology under
consideration and other factors will permit
analysts some flexibility when developing
treatment technology costs for T&C
documents. At the same time, the lumping of
many capital cost elements into three categories
allows analysts to prepare cost estimates more
quickly. The approach by the TDP is suited for
regulatory cost analysis and where applicable
their guidelines were integrated into the PTA
and GAC Cost Models and applied to the VSS,
WATER, and W/W Costs Model as well.
Small Systems
Installed Process Equipment
Engineering Design
Construction [Indirect]
Total Capital Costs
Large Systems
Process Equipment
Engineering Design
Construction [Indirect]
Total Capital Costs
40%
20%
40%
100%
22-29%
25%
25-50%
100%
Source: Discussion Summary: EPA Technology Design
Workshop. November 6 and 7, 1997. EPA.
Note: Construction costs include miscellaneous indirect
items like sitework, profit, and contingencies.
Table 4-1. Summary of Percentages Recommended
by the TDP
4-10
May 1999
-------
However, a review of cost estimating literature and case studies indicate that the percentages
recommended by the TDP are conservative. For example engineering design fees are reported in
cost estimating literature to vary from 4 to 15 percent of total construction costs (Humphreys, 1993),
with the higher percent being more applicable for complex chemical processing plants or small
projects. Economies-of-scale results in indirect construction costs like engineering to become a
relatively small component of the total capital costs.
An evaluation of case studies on PTA places mean engineering and indirect construction
costs at 17 percent and 17 percent of total capital costs, respectively [Figure 4-1]. The percentages
in Figure 4-1 are based on data for small drinking water plants with flows ranging from 9 to 1,400
gallons per minute (gpm). The 17 percent for engineering is high based on best professional
judgment and the published literature but still below the TDP-recommendations (Table 4-1).
However, for very small plants, engineering design can represent a large portion of total capital
costs.
4-11
May 1999
-------
; 8
i :g
i *•
CQ
2
•I
GO
«\
^
•S
•i
o
fi.
93
u
««
o
c
o
3
4>
C-
s
61)
4-12
May 1999
-------
The case study data showed that engineering comprised only 8 percent of the capital costs
for the 1,400 gpm plant, which is in line with expectations.
Figure 4-1 shows construction costs only account for 17 percent of total capital costs. This
is much lower than the percentages recommended by the TDP, however, five of the 11 case studies
had construction costs ranging from 3 to 10 percent of total capital costs. These may be low because
some of the case studies were retrofits using innovative installation techniques rather than full-
fledged installations. Without these data points, construction costs account for 25 percent of the total
capital costs. This is hi line with the TDP recommendations for small systems but at the lower range
of the values for large systems.
Only four case studies were available for GAC, but the data shows that process equipment
costs and installation account for two thirds of the total capital costs (Figure 4-2).
The implication of the TDP assumptions of process equipment costs can be seen by the
following example.
Case 1 (Using TDP-recommended percentages)
Model-generated process equipment costs = $100 or 40 percent of total capital costs
Engineering costs = $50 or 20 percent of total capital costs (imputed by setting process = 40 percent of total capital
costs)
Construction costs = $100 or 40 percent of total capital costs (imputed by setting process = 40 percent of total
capital costs)
Total Capital Costs = $100 (process) + $50 (engineering) + $100 (construction) = $250
Case 2 (Using Case Study data where process equipment is a greater percentage of total capital costs)
Process Equipment Costs = $100 or 66 percent of total capital costs
Engineering costs = $25.75 or 17 percent of total capital costs
Construction costs = $25.75 or 17 percent of total
Total Capital Costs = $100 (process) + $25.75 (engineering) + $25.75 (construction) = $151.52
4-13
May 1999
-------
.1
•3
•w
00
-------
The difference between the case study data and setting model-based estimates of process
costs at 40 percent of total capital costs are 65 percent. This suggests that using the TDP-based
approach, which assumes that equipment costs account for no more than 50 percent of the total
capital costs may result in conservative estimates.
Because of the relatively high degree of conservatism introduced into the PTA and GAC
Model Cost estimates if the TDP recommendations were adopted, costs estimates for alternative
designs for PTA and DBA were based on applying percentages for individual elements based on the
literature and best professional judgment.13 This "two-pronged" approach for generating costs was
designed to acknowledge the recommendations of the TDP while providing an alternative source for
comparative analysis of the results. Table 4-2 lists the percentages for engineering and construction
costs used for PTA and GAC Models. The percentages used for the direct engineering approach are
presented in Table 4-7.
Table 4-2. Percentages Used in the PTA-COST and GAC-COST Models1
Cost
Component
Installed Process Components2
Engineering
Indirect Construction Costs3
Total Capital Costs
Treatment Plant Size2
Small
PTA
50%
15%2
35.0%
100.0%
GAC
40.0%
20.0%
40.0%
100.0%
Medium >
PTA
40.0%
22.5%
37.5%
100.0%
GAC
35.0%
25.0%
40.0%
100.0%
Large
PTA
30.0%
30.0%
40.0%
100.0%
GAC
30.0%
30.0%
40.0%
100.0%
1 Based on TDP guidelines.
2 The TDP did not define small and large systems by flow. For this report small plants are 1 mgd; medium 1 to
10 mgd; and large plants are greater than 10 mgd.
3 Includes miscellaneous indirect items like sitework, interest during construction, administrative and legal costs, and
contingencies.
13 Note that the TDP also provided recommendations that allows flexibility in making assumptions for
engineering cost factors. Their recommended ranges for individual elements of indirect costs agree well with values
reported in the literature and best professional judgment.
4-15
May 1999
-------
Labor Rates
Another issue discussed during the TDP workshop hi Denver were operator labor rates used
to estimate O&M costs. Old T&C documents show that analysts assumed labor rates for water
treatment professionals to be about $ 15 dollars per hour. The TDP suggested that loaded labor rates
should be used and that labor rates should account for system size.' The TDP Panel recommended
loaded labor rates ranging from $28 to $75, with the higher labor rates for highly trained water
i
treatment professionals. They recommended checking labor rates from the Bureau of Labor
Statistics (BLS) or from State operator surveys.
An evaluation of the labor rates (unloaded) from the BLS " 1998-1999 Occupational Outlook
Handbook- Water and Wastewater Treatment Plant Operators" reported unloaded weekly labor rates
ranging from $335 to $1,034 with a mean of $668 which translates to $16.70 for a 40 hour
workweek. Based on the BLS data, the maximum value of $1,034 translates to $25.85 per hour.
Based on this data, the following loaded labor rates were selected.
Small Systems (< 1 ragd) - $28
Medium (1 to 10 mgd) - $40
Large (>10 mgd) - $52
Note: Above rates reflect a loading of 70-100 percent
of base salary for fringe and other benefits
The above rates incorporate the TDP's recommendations for using loaded rates and
acknowledge that labor rates increase as a function of system size.
i I
i I
Redundancies
Prudent design practices generally allow for redundancies. These may include an extra pump
to allow for uninterrupted service. However, drinking water treatment plants typically have storage,
which would allow for some down-time of equipment for repair. In the TDP workshop, the experts
agreed that redundancy is not a major concern for small systems, particularly those with storage
1 i i • i
capacity. Most members of the TDP agreed that it was prudent to include redundancy for pumps and
4-16
May 1999
-------
chemical feeders. Accordingly, an extra pump and blower was assumed to be required for PTA and
DBA.
Permits
In almost all cases, new construction will require the need for permits. Permits are required
for construction and for discharge to air, water, or land. The TDP members suggested that permits
be included as part of design costs or a value of 3 percent of total construction costs be assumed with
a minimum floor of $2,500. Note that the TDP has already recommended relatively high
percentages for design and including permitting costs at 3 percent of the construction costs may
overinflate permitting relative to other costs. Permitting was not broken out separately for the
available case studies that reported differentiated cost data (about 20 studies). Only one out of the
20 case studies reported permitting as a separate line item. The permitting costs for this facility were
about seven percent of the total capital costs. Figure 4-3 shows a breakdown of the mean percent
for permitting and other "extra" line items. It is inferred from these case studies that permitting is
typically not a cost driver and may already be rolled into items like engineering design. Thus,
assuming an additional 3 percent of capital costs for permitting adds a level of conservatism that may
not be necessary, particularly when the assumed engineering cost factors are already conservative.
Since percentages for engineering are already believed to be conservative, permitting is not included
as part of the basic technology installation costs but the final decision to use the 3 percent suggested
by TDP or to assume it as part of the design is better made during the RIA process. Accordingly,
permitting costs are presented in this document as a separate line item.
Land
Treatment plants typically have land available to install a new unit process. Neither PTA,
nor GAG are particularly land-intensive technologies but some systems may need to purchase land
to install these systems. The TDP members discussed the fact that land could be an important item
driving costs, particularly for scenarios where treatment plants are located in dense urban areas such
as Los Angeles county. However, the TDP members did not provide specifics on valuing land but
generally agreed that regulatory analysis costs should acknowledge that treatment plants may have
4-17
May 1999
-------
en
"O
s
,!= 00
-O
i— •
en
^O
CS
I
3
cu
e
TT
I
en
4-18
May 1999
-------
to purchase land and that the value of this land would vary based on whether it was located in a rural,
suburban, or an urban area.
For the purposes of this document, the TDP's recommendations for land were addressed by
assigning land values based on system size. The following land values were used.
Small Systems (< 1 mgd)
Medium (1 to 10 mgd)
- $1,000 per acre
- $10,000 per acre
The cost per acre for small systems is based on the assumption that small systems will
typically be located in rural areas or hi areas where land is not expensive. Larger systems would be
located hi small towns or metropolitan areas where land prices are higher. Note that of the 20 case
studies with disaggregated costs for aeration, only two reported land acquisition. Land accounted
for about 15 percent and 3 percent of the total capital costs for these two case studies. Across all
case studies, land accounts for about one percent of total costs. This suggests that it may not be a
significant cost driver.
Pre-and Post-Treatment Costs
As noted previously, the installation of PTA or GAC could trigger the need for other
technologies. Disinfection of previously undisinfected groundwater water or post-treatment may be
necessary because the water passing through the packing or the GAC system could become
microbiologically contaminated. Also, PTA or GAC systems can be fouled if the raw water contains
too much iron and/or manganese. In this case, the water would need to be pre-treated prior to the
PTA or GAC unit.
Costs for disinfectionare included in this document for estimating disinfection costs for those
systems requiring disinfection. Costs for disinfection are based on the document "Evaluation of
Central Treatment Options As Small System Treatment Technologies." Note that about half of the
ground water PWS universe already disinfects. Thus, a limited subset of potentially impacted
facilities would need to install disinfection.
4-19
May 1999
-------
While some utilities are expected to require pre-treatment to improve the raw water quality
for PTA and GAC, the number is expected to be small. First, it is presumed that water supplies with
;3,000
Totals
Diss. Fe (mg/L)
ND
0.67%
2.17%
7.55%
18.89%
6.42%
2.10%
37.80%
<0.3
0.36%
1.72%
10.20%
22.61%
9.05%
3.82%
47.76%
0.3-1.5
0.21%
0.53%
2.67%
3.08%
0.74%
0.31%
7.54%
1.5-2.5
0.02%
0.12%
1.34%
0.57%
0.10%
0.02%
2.17%
>2.5
0.31%
0.48%
1.74%
1.31%
0.62%
0.26%
4.72%
Totals
1.57%
5.02%
23.50%
46.46%
16.93%
6.51%
100.00%
Radon
(pCi/L)
ND
>100
100-300
300-1,000
1,000-3,000
> 3,000
Totals
Diss. Mn (gm/L)
ND
0.69%
2.67%
8.00%
21.99%
6.45%
1.43%
41.23%
< 0.02
0.26%
0.84%
5.97%
11.84%
5.90%
3.39%
28.20%
0.02-0.05
0.05%
0.36%
2.20%
3.17%
1.24%
0.53%
7.55%
>.05Q
0.57%
1.15%
7.33%
9.48%
3.34%
1.17%
23.04%
Totals
1.57%
5.02%
23.50%
46.48%
16.93%
6.52%
100.00%
Source: EPA, 1999.
4-20
May 1999
-------
Housing
Housing can also be required to enclose the entire process or components of the process.
PTA systems do not require housing except for components like pumps, blowers, and control
instrumentation but an entire G AC unit can be enclosed in a building (although many GAC units are
also found outdoors). For the purposes of this document, housin'g requirements for PTA and GAC
are based on best professional judgment. A review of engineering literature reports that building
costs can account for up to 20 percent of the process equipment costs (AWWA/ASCE, 1990). For
the 20 case studies, eight facilities reported requiring structures. These eight facilities' costs for
structures were about 19 percent of the total costs, which corroborates the literature. The mean
across all case studies was 10 percent (Figure 4-3).
For PTA, it is assumed that building costs account for 10 percent of the process equipment
costs because housing requirements need only address pumps, blowers, and instrumentation. For
GAC, it is assumed that housing costs are 20 percent of the process equipment costs.
4.1.6 Summary of Design and Cost Assumptions Used for Each Model
The key design inputs for the PTA-COST Model are the design and average plant flows,
Henry's law co-efficient for Radon, the desired removal efficiency, air-to-water ratios, cost indices,
arid a labor rate. Other variables are "internal" and can only be manually altered. These variables
include size ratios for the clearwell, physical constants, and cost/design relationships developed by
the original authors of the program. These design factors were also used for sizing PTA units costed
using the direct engineering costing approach.
The basic input parameters for GAC design are the design and average flows of the system
and the rate constant (K^). The higher the K^ value, the more the contaminant adsorbs to the carbon.
Higher K^ values result in lower empty bed contact times (EBCT), the time the water needs to be
in contact with the carbon to achieve the desired performance goal (AWWARF, 1997). This
translates into smaller process units and lower costs. In a study conducted by AWWARF, K^ values
used for conceptual design scenarios for treating radon ranged from 1.5 to 4.5. Costs in this report
are based on a K^. of 3.0, the mid-point of the AWWARF scenario values.
4-21 May 1999
-------
The key design and cost input variables for PTA and GAC are presented in Tables 4-4 and
I
4-5 for the models used. Table 4-6 presents a summary of input cost factors and cost indices used
in the models.
4. L 7 Summary of Assumptions Used for Alternative PTA
'i
As noted earlier, PTA costs were estimated using an alternative approach, which entailed
considering different materials of construction and a potentially less expensive configuration. This
approach was used only for a series of small plants, ranging in size from 0.05 to 2.9 mgd. This
i i I
method was used because radon compliance costs are expected to be incurred by small systems and
having an alternative to the model-based costs for PTA would provide analysts with more flexibility
during the RIA.
For this approach, PTA designs were based on standard design equations available in
engineering literature. Table 4-7 presents the design and cost estimating assumptions used. The key
ii |
difference from the PTA Cost Model's conceptual configuration and the direct engineered design
is the fact that the tower rests on a slab and that the clearwell is at grade. These differences are
:i ' :
expected to result hi a more cost-effective design (excavation and concrete work are minimized) and
i
may be more realistic for small systems dealing strictly with a Radon problem. Appendix A-0
presents a conceptualized diagram of the design.
' ' i
i
i
4.1.8 Summary of Assumptions Used for Alternative DBA
• , ;• | ' i
Diffused bubble aeration has also been demonstrated to be effective for Radon removal. Cost
1
estimates for DBA were developed by selecting and costing components that are common with PTA
• : | • i
(primarily the blower). Specifically, costs for the tower and installation, internals, packing material
were "backed out" of the cost estimates for the direct-engineered PTA treatment unit. The same
j
number of plants as the direct engineered PTA were used. The air blower was sized using the same
equations and assumptions as the direct engineered PTA using a water height of 3 feet. This "back-
of-the-envelope" cost estimate for DBA yields conservative estimates of capital and O&M cost
; j : i
estimates. The design assumptions for this configuration are presented in Table 4-8.
4-22
May 1999
-------
Table 4-4. Summary of Design Inputs for Estimating PTA Costs
Model
PTA-COST
VSS*
Water*
W/W Cost*
" "• V ;! "'. "' ;:';•'. .; . . .., : Design Inputs ' '.' '" ' "".' .' ' ".".." '• .' "•• •• •'.'•"•'
Flow: Range of flows to cover plant sizes from 0.01 to 100 mgd
Loading Rate: 30 gpm/ft2 (flooding considerations)
Air-to- Water Ratio: 15 (This can vary from 2 to 50 based on flooding checks)
Tower Dimensions:
Column diameter: Between 0.5 and 10 feet
Packing height: Between 1 and 20 times the diameter but less than 40 feet
Removal Efficiencies: 80 and 99 percent.
Clearwell: Generated by the model based on detention time.
(5- and 10-minute detention times were considered.)
Flow: Range of flows to cover plant sizes from 0.015 to 0.270 mgd (the model's flow
domain)
Costs based on designs generated by PTA-COST Model
Flow: Range of flows to cover plant sizes from 0.015 to 1 mgd (the model's flow domain)
Costs based on design generated by PTA-COST Model
Flow: Range of flows to cover plant sizes from 1 to 100 mgd
Costs based on design generated by PTA-COST Model
* These models cannot perform process design. The PTA-COST Model was used to generate the process design for use in
these models. Outputs from these models are used for comparison with PTA-COST Model results only.
Table 4-5. Summary Design Inputs for Estimating GAC Costs
Model
GAC-COST
VSS*1
Water*1
W/W Cost*1
Design Inputs
Flow: Range of flows from 0.01 to 100 mgd
K^: 3.0 (Mean of the values reported in AWWARF's Radon Report)1
Removal Efficiency: 50, 80, and 99 percent.
GAC replacement frequency: 555 days1
Density of carbon: 26 Ibs/ft3
Transportation and disposal of GAC as a non-radioactive waste costs are included (see
Appendix A).
Flow: Range of flows to cover plant sizes from 0.015 to 0.270 mgd (the model's flow
domain)
Costs based on designs generated by GAC-COST Model
Flow: Range of flows to cover plant sizes from 0.015 to 1 mgd (the model's flow domain
Costs based on design generated by GAC-COST Model
Flow:
Costs based on design generated by GAC-COST Model
1 Assessment of GAC Adsorption for Radon Removal. Revised Final. November, 1997. AWWARF. Denver,
replacement frequency is based on length of study in the AWWARF Report.
* These models cannot perform process design. The GAC-COST Model was used to generate process designs.
from these models are used for comparison with GAC-COST Model results only.
CO. GAC
Outputs
4-23
May 1999
-------
Table 4-6. Cost Indices and Other Factors for Models
Cost Index/Parameter
Engineering and Design (percent of total capital costs) '
Construction (percent of total capital costs) '
Land Cost (S/Acre)2
Electricity Cost (S/kwh) 3
Labor (S/hr)4
Diesel Fuel (S/gal)3
Natural Gas (S/ft2)3
Building Energy (kwh/frVyr) 5
ENR Skilled Labor
(1967 base)
(19 13 base)
ENR Building Costs
(1967 base)
(1913 base)
ENR Construction Cost Index
(1983 base)
Average CCI Increase
PPI for Finished Goods
BLS Commodity Code No. 1 14
BLS Commodity Code No. 132
BLS Commodity Code No. 1017
BLS Commodity Code No. 1 149
BLS Commodity Code No. 1 17
Index/Factor Value
See Section 4. 1.5
See Section 4. 1.5
$ 1,000 for rural (< 1 mgd)
$10,000 for urban (> 1 mgd)
0.090
$28 per hour for systems < 1 mgd)
$40 per hour for systems between 1 and 10 mgd
$52 per hour for systems > 10 mgd
0.979
0.0085
19.5
531 (for WATER Model)
5,294 (for WAV Costs Model)
499.1 (for WATER Model)
3,370 (for WAV Costs Model)
542 (for PTA-Cost Model)
3.2 (for WAV Costs Model)
367.9 (for WATER Model)
441.9 (for WATER Model)
447.9 (for WATER Model)
405.6 (for WATER Model)
5 17.8 (for WATER Model)
281.9 (for WATER Model)
1 Engineering and construction cost percentages are based on evaluation of constructed projects and from recommendations from experts
(seeSection4.U).
1 Land costs are based on "Technology Design Conference Information Package. USEPA. Nov. 1997" and best professional judgment.
' Bureau of Labor Statistics.
' Labor costs are based on range of labor values presented in TDP Report. Rates are loaded values.
3 WATER Model Document.
4-24
May 1999
-------
Table 4-7. Design and Cost Assumptions for Alternative PTA Configuration
Process Item or Parameter
Plant Sizes
Performance
Tower Height
Tower Diameter
Tower Construction
Pumps
Clearwell
Blower
Piping and valving
Packing
Air-to- Water Ratio
Overall mass transfer coefficient
O&M labor
Maintenance
Packing reconditioning
Other Factors
Value
Five plants sizes: 0.1 to 2.2 rngd design flows.
80 and 99 percent removal
Restrained to 20 ft by varying diameter within a 1 to 5 foot range to
maintain low profile. Tower height is three feet greater than packing
height.
Between 1 and 5 feet.
Fiberglass Unit with steel internals mounted on 10' x 15' 8" concrete slab.
Dual centrifugal pumps at the inlet. Designed to deliver water to top of the
tower. 20 percent allowance for suction head and another 20 percent for
frictional losses. Pumps and blowers are housed in sheds.
At grade ready-made tank
Standard air blower
Pipe diameter based on 6 ft/s delivery. Three butterfly valves to regulate
and distribute flow (see conceptual diagram) •
Plastic packing
15-25 to keep tower height at 20 feet
0.015 per second (see Appendix for other physical constants used)
15 minutes per shift to take readings. Two shifts for plants under 1 mgd,
three shifts for plants greater than 1 mgd. Labor rate of $28 per hour for
plants under 1 mgd; $40 per hour for plants greater than 1 mgd.
Two days per year
Once every year
Engineering Design — 15 Percent
Other Indirect Costs - 12 Percent (Profit); 15 Percent (Contingencies)
4.2 CAPITAL AND O&M COSTS AND EQUATIONS
Point estimates of capital and O&M costs for aeration and GAC were first generated based
on the approaches) described in Section 4.1. Cost curves were then generated using the trendline
function in Microsoft Excel, which generates best-fit equations for the flow and cost relationship.
Polynomial equations were used to provide the best possible fit.
4-25
May 1999
-------
Table 4-8. Design and Cost Assumptions for Direct Engineered DBA
Process Item or Parameter
Plant Sizes
Performance
Water Height
Tank size
Tank
Pumps
Blower
Piping and valving
O&M labor
Maintenance
Other Factors
'- - • • • • Value .-• ' '.-'..- -. .. .' ... . '
Plants sizes: 0.05 to 2.9 mgd design flows.
80 and 99 percent removal
Restrained to 3 ft.
Based on 5 minutes retention time. At grade ready-made tank
Fiberglass Unit with steel internals mounted on 10' * 15' 8" concrete slab.
Dual centrifugal pumps at the outlet. Designed to boost water delivery to
the next treatment step.. With 20 percent allowance for suction head and
another 20 percent for frictional losses. Pumps and blowers are housed in
sheds.
Standard air blower
Pipe diameter based on 6 ft/s delivery. Three butterfly valves to regulate
and distribute flow
15 minutes per shift to take readings. Two shifts for plants under 1 mgd,
three shifts for plants greater than 1 mgd. Labor rates of $28 per hour for
plants under 1 mgd and S40 per hour for plants greater than 1 mgd.
Two days per year
Engineering Design - 15 Percent
Other Indirect Costs - 12 Percent (Profit); 15 Percent (Contingencies)
4.2.1 Capital and O&MCosts for PTA and DBA
Table 4-9a presents a summary of capital and O&M costs from the PTA-COST Model for
all of the scenarios considered for PTA. Table 4-9b presents costs for other indirect items
I
potentially associated with the base aeration technology. Table 4-9c presents capital and O&M
costs for direct-engineered PTA. Table 4-9d presents capital and O&M costs for DBA. Equations
'i
based on the costs shown in Tables 4-9a through 4-9d are presented at the end of this chapter.
Appendices A-l and A-2 present more detailed breakdowns of estimated costs, cost curves
and resulting best fit equations.
4-26
May 1999
-------
-
2
I
H
•a
^
u
OS
p-l
O
u
«
"3
•*•
'c.
~
IS
a
., ...
8
u
s
•S
t-
on
8
£
*M
».
<
Q
!U
S ^
i ^
* 0
O 00
•JJ; •-•
is
s£
O ON
02 —
•5 ^
3 C-
1*
2 o
v> *>
» x-s
"S O
•j£
•£ o\
m °v.
a.
a f
S —
> fa
t>
•** ^"s
s r"
B >-'
ig
O 00
u
•4— /— \
3 i—
B — '
i^
*= ON
o o\
A3
•*•* — T
j*
2 o
in «°
fl? • i
** CM
3 O
£ s?
"5 o1-
2 S
«n ^
OH
153
ea
/—\
•a
s
s— '
i
o
tf)
a
ea
%
a
S3
8
ca
ea
^?n
S£
oo
to
m
o
t-
^~
00
m
m
^
•^~
fN
O
0
o
f-
rr
m
fS
m
rr
ON
0\
CN
TT
OO
VO
0
o
CN
•0
0
>n
CO
VO
r~
o
o
wo
t—
o\
CN
CO
CN
O
I~-
ON
VO
CO
CN
00
g
o
TT
CN
O
ON
VO
O
0
<0
CO
oo
CN
CO
oo
CN
VO
os
o
CO
r~-
r-
_
o
ON
"3-
r—
CN
ON
t^
rf
CO
VO
CN
C~
CN
VO
m
•*
CO
ON
O
0
o
VO
wo
CO
o
"3-
CN
•q-
00
CN
CN
U1
00
CN
ON
VO
r~
r-
co
o
oo
CO
•*
r~-
TT
OO
VI
VO
o
CO
•*r
r-
r~
w>
0
TT
CN
0
OO
VO
r-~
U">
CN
oo
00
r-
ON
VO
P»
O
0
2;
o
•5T
CO
o
oo
wo
r~
o
t--
n-
T
O
CO
n
CO
CN
CO
TT
Tf
W)
o
0
CO
CO
CN
oo
ON
r-
cs
TT
•*"
w^
ON
w^
wo
ON
VO
ON
00
00
r-
wo
0_
o
00
r-
o
S
CO
ON
NO
00
r-
oo
VO
VO
oo
0
ON
TT
CO
t^
0
o
0
—
o
VO
o
VO
ir>
CO
CN"
o
TT
in
?
CO
CO
CO
VO
CO
VO
0,
CN"
CO
r-
CN
w^
•5T.
co"
0
o
CO
CN
•ro
CN
O
•3-
•fl-
OO
wo
Wl
oo
C--
VO
CN
vo
oo
CO
CN
TT
r-
o
CN
CN
VO
VO
§
o
T
TT
o
ON
0
VO
CO
oo_
VO*
s
Wl
CO
v>
0
oo
o
oo
00
wo"
s
CO
ON
ON
o"
0
oo
r-
co
VO
•*
VO
W!
t~
w^
ON
>n
VO
oo
o\
CN
CO
i?
W">
oo
rr
VO
CO
•*r
oo
o
VO
w^
s
VO
0
ON
o_
oo"
s
TT
CN
WO
co^
co"
r-
t—
'O
cK
ON_
vo"
r-
o
CO
VO
m
CN_
CN"
o
o
o
o
^J*
c
ID
s
s
s
a.
4— '
c
u
1
I
a.
-o o
c >
so w
c 2>
1 8
si
« o
ll
'o u
£ 18
=5 -^
C 1>
'Z &
= 1
.s •« .1
0 0 C
c | g
^«sl
v> £ £
" « •&
8 g~
w 'o 3?
III
s -i °
O u ^N
Z oi C-
4-27
May 1999
-------
o
1
CJ
3§
"S
•*-
.2
O
s
-5
Is
•*••
S
o
1* .
i
t3
u
•2
42
en
ja
.2
•ft
f2
».
* ****** ° ^" °* ^
£|joUC{U£!C
AI*S§Q!ES
£ J,§ 8 2? S S g
Jsliggi
^go^^S^^
g_ g - | R s
%ss§ss
« r* »o — • »n —
e _r r*r *-r «r v— *
i 2
| ; | g S S |
§ e - a a" a a
3 € •**• OO O — 00
" « s — - S" 5 I?
il
§| « _r ^j ft T?
tn
c
g S IA 1« H K> f>
fit
•|: sss s s
§ ,i a a s s K
« O O O *rt <*i
C O O O O M-»
« u*t *n v-t v\ oo
• - % S X X S
5 2
d ; 8 s s § ^
n s = a a a a a
; *
i ; s s s 8 s
« C *i M *T -T m
1 1
(2 ; S S S g g
g S ;;" g rj rj rJ
£-
. J 3 S r^ — S
1 | R a 2 S g
§! : = S 1 1 1
§ s a u 5 = s
b i/)
• ^ ' o
[ ^ -™* — <^ r- — —
1 * •«*
vo o r- vo — r-
Vt fc^ trt fc^ V
V> CN W» VQ OO w
s 5 ™ S s s
3 § S S 2 S
so o r-j n w> r-
S\ oo c\ — r- <^
5 S S 3 S B
ao_ r\ ~-^ rn vq. ««_
o» oo \o r-T -rr o"
**** *^ ~ gj g
i S s S i 1
w " « a a
— m rN c*t n —
£•» Ct a
S ~ S g a. r~ «
o S S § ? S 5!
—
® S o § G S §
*^ 5? S
S R S S S S S
"» i -r" ] - O «
"-BBS!
3 o r*- — r«j c* °
— *o m oo ea oa t-
*' w « a a i
> -v ^r rM m —
N r» »n oo oo r^ •»
•^ *o r- vo o r- •
3" p^" o vT r^" ' — '
*^ ** ^ S £* i
i f*i VO O «n r^. _
0^ C» O — \O CS* §
*** *^* S U |
- — jg t; go m g
S 3 § S § Jr
5 § o § o H E
^ ** *^ o >. g
5 *« —
^ ill
4-28
May 1999
-------
Table 4-9c. Capital and O&M Costs for Direct-Engineered PTA
Flows (mgd)
Design
0.0432
0.072
0.1296
0.36
0.72
1.08
1.44
2.88
Average
0.0216
0.036 ".'"
0.0648
0.18
0.36
0.54
0.72
1.44
• " .' :': "'" •''.' ./'Capital Costs' '•• >: • •" - ' ' ."•'- •
80% Removal
$27,300
~" $28,600 "
$28,700
$36,600
$50,400
$57,800
$59,900
$82,000
99% Removal
$29,000
" $30,700 '
$31,000
$41,800
$57,200
$69,000
$70,100
$99,400
1 O&M Costs
80% Removal
$8,700
$8,700
$8,800
$9,900
$10,600
$12,200
$12,200
$12,800
99% Removal
$8,700
' " $8,800 ' "
$8,900
$10,600
$11,900
$15,700
$15,700
$17,500
Table 4-9d. Capital and O&M Costs for DBA
Design Flow (mgd)
0.0432
0.072
0.1296
0.36
0.72
1.08
1.44
2.88
Average Flow (mgd)
0.0216
0.036
0.0648
0.18
0.36
0.54
0.72
1.44
Capital Cost
$19,100
$20,300
$20,000
$22,300
$34,700
$37,800
$40,900
$56,900
O&M Cost
$8,600
$8,600
$8,600
$9,200
$9,200
$10,300
$10,300
$10,300
Note: Costs are the same for 80 and 99 percent removals.
4.2.2 Capital and O&M Costs for GAC
Table 4-10a presents a summary of Capital and O&M costs for GAC for all of the
performance scenarios considered. Table 4-10b presents costs for other indirect items potentially
associated with the base GAC costs in Table 4-10a. Appendix A-3 present the breakdown of
estimated costs and charts.
4-29
May 1999
-------
QJ
<
O
I
42
S
U
S
<*
O
J
u
C3
U
•*T
&
I
42
3
u
s
fiX
o
42
3
u
S
c.
a
j
.si
'o
c
"3
(••
If
o
E
«
'5
"3
td
Remove
«>
si
<
a &
ff? £
Nd
O\
ON
eN
O
00
X*
o
V)
n?
ON.
ON
O
oo
0
§
*•*'
an
,£•,
w
^f
en
g
en
60
f"«fc
ON
CS
VO
S
;?:
en
6O
i
r-
oo
t-
o
o
cs
6O
8
0
o
o
o
ON
0
cn
-
ON
CS
m
io
r-
o
ON
^
60
VO
t—
S
VO
6O
VO
ON
CS
cn
60
en
en
cs
en
ON
CS
ON
vo
O
O
o
3
0
cs
cs
o
cs
&O
ON
r-
r—
60
f*^
ON
vo
60
ON
~
ON
6O
CS
OO
ON
VO
OO
vo
CO
—
6O
•n
3
o
S
0
VO
cn
en
O
VO
ON
60
*n
cn
00
60
O
00
00
0
_~
6O
vo
ON
VO
O
"3-
•O
OO
cs
o
o
„
0
^"
cn
r-
en
VO
CS
r-
cs
VO
00
*n
en
CS
6O
CN
CN
in
0
CN
&%
cs
oo
cs"
6O
O
NO
cn
^S
6O
Tt
0
•n
o
f^
^3*
(^
CO
00
r-
cs
cs
ON
VO
vo
CS
cn
VO
60
en
ON
en
NO
_M
^M
y^
cn
O
o
-------
a
£
$
H3
u
o
Gfl
ce
es
• MM
**
«
-«
O
p-
g
•3
c
en
O
+1
s
s*
£1
*
1
Ov
r-
^
OO
r*.
c^
r^"
vd"
•
ON
t^
$569,282
0.01745
5
0
m
ON
S
fl
r--
r*-
ON
V>
m
ON
ON"
S
o
8
V*
r-
ON
r^
o"
f*l
^>
fS
r*-
r-
00
58
(N
ON
f-
a
°
vO
ON
(A
in
vO
m
m"
(N
(A
m
O
vO
ON"
&
m
S
vO
m
w»
$1,208,840
vO
ON
5-
^
VO
64
OO
fN
O
O
o
vO
oo
O,
m"
64
in
m
r-
m
64
S
OO
n
m
vO
in
r*i
&4
m
^l
i^-
m
64
S
r~
O
r*l
M
CTv
I
CN
(N
<£
64
t*»
TT
(*^
r^-"
s
vO
-.
r^
n
^
fT
r*
vo"
64
$205,781,817
m
m
fn
^;
i
•5T
*
$352,936
1
m
r»-
64
t*.
*n
-«T
S
64
S
t-^
r*i
vO^
tr"
64
~<°
g"
OO
r*-"
64
ON
OO
i"
a"
$7,316.351
$243,878.375
r*\
SL
oo
oo"
o
&4
vO
in
o
o
es
4-31
May 1999
-------
4.3 INTERCONNECTION (REGIONALIZATION) COSTS
1
An alternative option for a utility to achieve compliance is through regionalization, which
entails linking up with other water systems that can cost-effectively provide water that meets the
regulatory requirement. It is well known from many studies (Castillo et al., 1997) that small water
systems have historically (pre-regulation periods included) had problems delivering a safe and"
affordable level of water service to its customers.
Interconnection and consolidation of services is one way to achieve economies of scale for
j
compliance. Associated costs may include studies, legal fees, and interconnection costs (i.e., costs
for pipes and appurtenances).
/
Cost data for estimating interconnection costs were not readily available. A "back-of-the-
envelope" estimate was assumed by taking the cost of installed cast iron pipe at $44 per linear foot
(an average cost for several pipe diameters) from the R.S. Means Plumbing Cost Data (1998) and
applying 20 percent for fittings, excavation, and other expenses resulting in a cost of $53 per linear
foot or $279,840 per mile. This cost, even if halved, indicates that interconnection costs could be
as high as treatment using aeration.
4.4 CENTRALIZED TREATMENT FOR SYSTEMS WITH LESS JTHAN 10,000 GPD
Point-of-entry (POE) costs were taken directly from the document entitled Cost Evaluation
of Small System Compliance Options: Point-of-Use and Point-of-Entry Treatment Units. Draft.
April 20, 1998.
The equations for POE treatment of Radon to 300 pCi/1 are as follows:
GAC: y = 1
Aeration: y = 12.24x-°-°~
•where y — Sper thousand gallons; x = number of households
Note that unlike previous equations, the independent variable is x or the number of
households served.
4-32
May 1999
-------
4.5 COMPARISON OF PTA CAPITAL COSTS WITH CASE STUDIES
Figure 4-4 presents the PTA Cost Model and the direct engineered 99-percent capital cost
curves and the case study data described previously in Section 4.1,14 The case study data costs, from
different years, were escalated to 1997 dollars assuming an annual inflation rate of three percent.
A power fit for the case study data was also generated using the trendline function in Microsoft
Excel.
I
O
o
D.
Rt
O
I
100000000
10000000
1000000
100000
10000
Best fit line - PTA-Cost Mode
99%Removal(1)
AWWARF 1998
cost estimates
Power Fit Line for all
case studies
1000
0.001
0.01
0.1
10
100
1000
Design Flow, MOD
(1) Only for the base technology: Does not include costs for
.structures, land, additional permitting, pre-and post-treatment.
~S Case Studies STA and .
A Case Studies, "Full Scale Report"
,_£ Direct Engineered - 99%
Xj Case Studies PTA Rn
_*L__Pow er (AWWARF 1998 cost estimates)
epor
.AWWARF 1993 Estimates
AWWARF 1998 cost estimates
.Power (AWWARF 1993 Estimates)
Figure 4-4. Comparison of PTA-COST Model Capital Costs to Case Studies
The case study data in Figure 4-4 demonstrate scatter across the range of plant sizes. The
scatter stems from several reasons. For example, some of the case studies reported in the A.E.
14 On a log scale, the 80-percent cost curves are almost identical to the 99-percent cost curves. To minimize
clutter, the 80 percent cost curves are not shown in Figure 4-4.
4-33
May 1999
-------
Hodsdon Report were implemented using volunteer labor. These projects, at the lower-end of the
j
flow scale, are below the PTA Cost Model's and the direct engineering capital cost curve in
Figure 4-4. This is logical because if standard construction methods were used, the cost of these
projects would probably be closer to the PTA. However, data from other case studies, which were
constructed using conventional means agree well with the PTA Cost Model's estimates.
' • I
i • ' !
Scatter also stems from the fact that some projects include some components that are not
necessarily associated with the base technology. These might include relocation of existing pumps
': ii ' . i
or other equipment, re-configuration of the piping, and installation of ancillary items not directly
related to the specific technology but implemented as part of the overall project. Other reasons for
•ii | • i
the scatter include the variability in labor rates (costs can vary by geographical region), the year the
: . • ,1
project was implemented, and soil conditions.
i
Thus, one can conclude that utilities across the United States will encounter different costs
when implementing a given technology for their system and that a single utility's experience is not
necessarily a good indicator of costs that other utility's will incur in complying with the same
regulation. Thus, if modeled costs lie within the boundaries of a range of case studies, then the
modeled costs can be taken as a reasonable indicator of potential compliance costs for the purposes
of an RIA, which attempts to model costs at the'national level.
I ''
As Figure 4-4 shows, the PTA Cost Model's costs are remarkably close to a power fit
generated from the case study data for the larger systems. Note also that the PTA Cost Model Cost
I
curve is just for the basic PTA Cost technology. Separate cost curves are presented in the
j i
appendices for instances when utilities will require disinfection, buildings, land, and permitting.
Including these costs would push the PTA technology cost curve towards the higher end of the case
study scatter.
In Figure 4-4, the direct engineering cost curve has a flatter slope than the PTA Cost Model
curve. The costs generated based on the direct engineering approach are also less than the PTA Cost
4-34
May 1999
-------
Model's curve. The lower costs developed using the direct engineering approach are the result of
using a fiberglass tower and an above grade clearwell. But the capital costs using the direct
engineering approach are still greater than the costs reported in some case studies because certain
fixed costs result hi a flat curve at the lower end of the flow range.
In addition, capital cost estimates (escalated to 1997 dollars) presented by AWWARF, 1995,
for six plants in its report "Estimating the Cost of Compliance With Drinking Water Standards: A
User's Guide." The report noted that the costs for the six plants addresses issues related to very
small systems and the AWWARF capital costs for aeration were shown to be two to three tunes
greater than EPA's capital cost estimates for PTA prepared at that time. In addition, the report also
concluded that EPA's capital costs tended to fall in the lower ranges of actual constructed projects
and that the EPA costs tended to produce "minimum" cost requirements. A 1998 AWWARF report
"Critical Assessment of Radon Removal Systems for Drinking Water Supplies" also notes that actual
bid costs were well above EPA's capital cost curve for PTA.
Figure 4-4 shows that the capital cost curve from the revised version of the PTA Cost Model
is a better fit to the case study data, some of which are the same ones used in the AWWARF reports
noted above. As can be seen from the AWWAF's 1998 reports (Figure 8.5) , the "best fif'is
well-below the "PTA Cost" for small systems.
What is reassuring is that EPA's traditional models, also appear to agree with independent
model costs (Figure 4-5) and with the case studies. This provides further support to the PTA-Cost
Model as a good tool for estimating conceptual-level cost estimates for treating Radon.
4.5.1 Comparison ofO&M Costs With All Case Studies
Figure 4-6 presents a comparison of the case study O&M data with the PTA-COST
Model and the direct engineered data. The PTA-COST Model's estimates for O&M are greater than
those reported in the case studies for plants less than 1 mgd. This trend is reversed as the plant size
increases but the PTA-COST Model is still within the lower end of the scatter. Also, note that the
O&M costs display more scatter than the capital costs. Estimates of O&M costs are greater using
4-35
May 1999
-------
a
i
3
4
tinnnnnnnn
I
l
i
j
i
1
$10 0"" "in ' 3C_
\ S
i S * *
. «innnn '
•
i
-------
-------
8
fefl
o
3
8
•fi
o
a
0
tf
c
o
'S
€3
(3
h
(S
w
.1
•»••
§
cr
W
•fc*
w
o
U
<
•o
^
«
CA
g
v> U
•j 1 J
c
o
C A
.2 S3
nj1
-
««
o
c5
«—•
3
¥
O
i_
J3
J3
3
03
ts
<
u>
eg
3
§
0
B II II II II
o
< O
CQ <
Q O
?.*!
»»
'. '"
Vr'*'
.MW"V«
7?c
Ss
<-e
:l
{2
;C
>
jf
•*•»
«
er
W
•fa*
I
5
••*s.
'<3-
»
.: '..;.
- '.-•»
-.-"••"
g
»•»
en
6'
1
5
y
$-.
-*M, *'
"f
CO *"""J
vo o
S S
^M l«l-«
+ +
x x
OS ^ _.®
OS OO O
oo ?? vo "^
in c — oo
+ os1 + E
CN 0. CN VO
X ? X V
§ £ § £
+ jp os V
UJ V CN —
T| Tl
X X
vo r^
o oo
+ •<*•
tU -"S;
C- o
11 II
vo
oo
os"
T Os
i^!
x *;,
CO SO
°V. ~a 4-
vo" ao T3
S E 3g
If ?I
co — ; x ro
A§ 1°
X ^
jo II
+ >,
VO
II
Si
1 (U
V~l ID
- C
|M
1 §
u ..
£ 'K'
S.-2
S S
<* "O
fc
>n oo
S 5
— OS
+ + •
X X
n u
— r .S
O m
||
(U ^
t- C
S.-2
i rs
§|
<• -a
fc
vo •<»•
wi os
o o
m vo
I— 4 ~~
_1_ J_
+06xA2 + 59164x
=x<=0.091mgd
841xA2+16608x
\
u
VO
II
-a
JL fe
O Q>
~ c
•5"'^
Ej C
I §
c ..
(U /— \
o c
«3 .2
^ *^
OS U
«T"5
f~* "3
'i
Os cvj
^^ SO
— o
~ CN
+ +
i596xA2 + 49550x
=x<=0.091tngd
071xA2+10453x
r^ o
II II
VO P-
^
il
c ..
1) ,— V
B c
n. '5
i C
o ^>
< "S
E~* "3
1
z
o
vo
oo
CN
r
X '
CN
i| |t
CN • P 0
< CN < 0
7j || ^ j^i'
r-- v "^ II
CO X S3 V
co o' >>
"x
^
CN
_O
C
4)
•o
3
C
1
*-_-*
=
C8
ID
U
4-38
May 1999
-------
rp
^ ^
?ON
22
r- o
II II
^ §
JN S
NO
+ ^U
.. ' '
f^ X
rn ^
l! 11
U V « —
«2 S3
? S
M —•
^^ II
II "
>» ^
f!
§-a>
I1!
4J "•
c o
•*-* 1^
c *^
S o
||
S'
10 ON
§ §
X X
<$ OO
*O "Q O
5 |> S^
10 C — SO
1! il
? if S v
rfl * / ^j ^^
_! i^"\ ^^
1 § ' 0
<^ C5 <^
X X
? NO
ffl •«
r~ o
II II
ON
O)
+
X »
u
NO
II
2l
— • OO
Cw p"
If
£1
II
o\ «>
< "S
fcl
i
ON cs
Si 2
— o
— CN
X X
O en
^* SO ^^
ON ^ f^ "^3
^ C »-^ SO
o^ V ?^ x
10 ?5 O V
NO V ^ S
ON rn ™ o
1 § « 0
JP O <
X x
VO ON
? rn
BJ n
t- o
II II
^ 00
— r-
+ +
x x
pS m
m o
fN ON
If 2-|
+ — + J
1 5 |T
I§ J2
X X
s 9
+ rj
lil °°
NO 0
II II
•o
2 |
1 §
S £
te .2
?• c
o «
OO q>
^*^ "O
< V
fcl
E
Z
o
NO
oo
CM
~
>~
X
(N
JT1 13 CN
ON 00 (N 13
00 p O OO
+ °°. s
r^ V !Q T
m x ^* v
;_! j^ °° x
i °. " r>i
*
X
O
en
II
's?
1
u
•o
u
3
1
O
cd
-------
(U
3
O
f
"o
c
"3
o ° °^
1 0 I M
X X
CO >
0
>
o
_r
o
If
c g
" ^s
U ^3
0 3
00 C
"M'E
33
3
O
T"
<
t--
o
ON
X
ON
It 2^
if ?l
§? Sv*
§ * s^
10 II 2
en 0' >»
X
in
N
II
>^
_-
§
|J
4-> C
C U
" S
= o
« ~~
3
3
*"^
T
z
oo
X
VO
vo"
ts
•n
^ -a 1.
O 00 in 7«
0 E CN I 2P
oo iy-, vq E
~^~ o in 2
X II | ||
(S V \f
r- x -^i- x
2 X CN JQ
n °. o! o
>»® «
0
1
o
o
o
1
II
•— r
o
II
C S3
« S
11
cr\ 3
oo E
C i
'OT
3
0
Z
-i
Z
n-
•n
X
ON
r-
m
_i_
-: -o P>
-------
I
o
o
a
o
t3
S3
09
O
•-C
05
er
o
—
3
s
s
02
<*°
*
-
. s.
• 13'
a*
W
1
;S
'o'
*,
>
,,
-
Capital Cost Equation
j?
-***
^
."
*«.
E
«
Cft
> O-
i
"o
a
•JS
"
en
CN
C--
en
,
±£
ON
-—> °^
X Tg o\
"2 oo o
OO p CN f
ON CN + g?
Il If
II ll
N 1 + 2
II d ?
^ X
en
r~
en
o
II
°°. CN
ON en
ut O
•* —
X "^
CN X
cA3 + 6828 13xA2 + 43796!
0.01<=x<=0.75mgd
A3- 1451.8xA2 + 341048
0.75 f^
r- > >t
•4-1
E
2
2
a.
CN
t--
en
_i_
^
~ »
,2 -a .ov
;2 oo o
OO If rvj "O
oo E ,00
ON CN + p
Ox ?* *3
B-? 1?
| JT g N^
r~- S3 '"" oo
CN § + 0
II o 0
>» X
en
f—
en
9
u
CN
in
en
en
+
x
oo
oo -o "">
!! Il
w ^* ^~" f^J
II o ^ °
en
9
ii
>,
C
u
E
2
CO
Q
cu
VO
^f ^«j
t <**
S °°
4-
+•
ci5 s
en fXj
oo e J2 £
1 °- +" eo
?2 T ^7
*=» v § ^
<^ p ""
o o
°- fh
o
II "
•* o
^^ en
CN NO
"*" "*"
p*^
fl ct|
X ^ ^ CN
"^ X CN V
?i ?I
^ en ' y(
^ <*
O O
O i
O CJ
o r*
1 II
II p^
1
11
2 |
I'l
&• tj
OO **""
^^ Q
^
ON
O
CN
oo
+ 0
v °°
2 5
^ £ °. §
X, 5*-, mo
en II • T
«- V CN A
"""" ^ btf
XV oo V
«n ™ !C
1 _. o ^-
en q <*">
CN II
§ >>
o
o
II
p-
CN
+ 0
X ^T
OO O
O oo
:| if
= | XV
O V ON o
+ en — t--
S ?
§ *
t"|-
11
>-,
0
1 "i
s '!>
8?
Q- o
^ •—
^D
fc
t-;
NO
O
00
1. « i. 1 1 1
0 eo r*: oo Si?? 2o
§3 28 +g S§
oo v + »
o
oo
r-
ON ^
en en +
-— • NO •«<
NO 0 CN
Vt CN t^j
+ E x g - 1
X Q. iS Q. +80
§ |> S3 |> ? |
+ v- 7V g 7
g | £ $ | %
T o ?
U o ^
>> I" o
>> ci
cs
II
^.^
S |
CJ C
fe "5i
i S
0
^""^ en
^^ ?J
^^ E
4-41
May 1999
-------
ued)
o
&
3
1
o
e
U
_4>
K
O
V
C
O
•a
i,
.2
e of Cost Equatio
C3
E
S
s
m
&M Cost Equation
Capital Cost Equation
I
G>
,
"S
E
v
°
o
E
-------
REFERENCES
Alexant, J.E. (1995). "Installation of additional 80 gpm water treatment capacity pilot project, Phase
II Mode 2, New Boston Air Force Station, New Hampshire." Environmental Science and
Engineering Inc., Amherst, New Hampshire.
American Water Works Association and American Society of Civil Engineers (A WWA and ASCE)
(1998). Water Treatment Plant Design, Third Edition. McGraw-Hill, New York.
Atoulikian, R.G., C.A. Close, M.A. Lucca, B.E. Juergens, and P.E. Malmrose (1995). "Radon
Removal - Designing to Meet Upcoming Regulations,"Proceedings oftheAWWA Annual
Water Quality Conference, pp. 587-610.
American Water Works Association Research Foundation (AWWARF), (1997). "Assessment of
GAC Adsorption for Radon Removal." revised final prepared by Cornwell, D.A., N. Kinner
and N. McTique.
American Water Works Association ResearchFoundation(AWWARF),(l 995). Estimating the Cost
of Compliance with Drinking Water Standards: A User's Guide
American Water Works Association Research Foundation (AWWARF), (1998). Critical Assessment
of Radon Removal Systems for Drinking Water Supplies. An AWWA and AWWARF sponcered
study conducted by Kennedy-Jenkins consultants.
Balagopalan, M. (1998). Telephone conversation between Mohan Balagopalan, Manager,
SCAQMD, San Bernardino, California and Diane Loy, SAIC, McLean Virginia. May 22.
Brown, D., PE. (1995). "Aeration Reduces Radon in Groundwater Supply." Public Works, May.
Cairo, P.R., J. McElhaney, and I.H. Suffet (1979). "Pilot Plant Testing of Activated Carbon
Adsorption Systems." Journal AWWA, Vol 71, pp. (500-673.
Castren, O. (1977). "The Contribution of Bored Wells to Respirating Radon Daughter Exposure in
Finland." IRPA, Paris.
Clark, R.M. and J.Q. Adams (1991). EPA's Drinking Water and Groundwater Remediation Cost
Evaluation: Granular Activated Carbon. Lewis Publishers, Chelsea, Michigan.
Crawford-Brown, D. J. (1990). "Analysis of the Health Risk from Ingested Radon." Chapter 2 from
Radon, Radium, and Uranium in Drinking Water, Cothern, C.R. and P.A. Rebers, editors.
Cummins, M.D. (1992). "Packed Tower Aeration Cost Estimates for Radon Removal,"
Communication from M.D. Cummins of EPA-TSD to Marc Parrotta of EPA.
R-l
May 1999
-------
Cummins, M.D. (1988). "Removal of Radon from Contaminated Ground Water by Packed Column
Air Stripping." U.S. EPA, April.
C WC Engineering Software (1994). W/WCosts and Design Criteria Guidelines. User Manual for
Computer Software for Estimating Water and Wastewater Treatment Costs. Version 2.0.
1 ' | !
Deb, A.K. (1992). Contribution ofWaterborne Radon to Home Air Quality. Prepared for the
American Water Works Association Research Foundation (AWWARF).
Dixon, K.L., et al. (1991). "Evaluating Aeration Technology for Radon Removal." Journal A WWA,
8:4:141,
Dixon, K.L. and R.G. Lee (1988). "Occurrence of Radon in Well Supplies." Journal AWWA, pp.
65-70.
Dixon, K.L. and R.G. Lee (1987). "Radon Survey of the American Water Works System." Radon
in Ground Water. Barbara Graves Editor. Lewis Publishers, Inc. pp. 311—346.
•I
Drago, J.A. (1998). Critical Assessment of Aeration Technologies for Radon Removal From Water
'Supplies. Sponsored by AWWA Research Foundation. Denver, Colordao.
Dyksen, I.E., D.J. Hiltebrand, and R. Guena (1986). "Treatment Facilities and Costs for the
Removal of Radon From Ground Water Supplies," in Proceedings of the Environmental
Engineering Specialty Conference, published by the American Society of Civil Engineers.
pp.510-52i:
I
Field, R.W., et al. (1995). "Radium-bearing Pipe Scale Deposits: Implications for National
Waterborne Radon Sampling Methods." American Journal of Public Health, Vol. 85, No. 4,
April.
: j i
Goldstein, N. (1997). "The State of Garbage in America." Biocycle, April, pp. 60-67.
i
Great Lakes Upper Mississippi River Board (GLUMRB) (1997). Recommended Standards for
Water Works.
I •
Griffin, W.E. (1989). "Meeting Water Quality Objectives in Periods of Financial Constraint."
Drinking Water Treatment Small Systems Alternatives. Toft, Tobin and Sharp, pp. 46-52.
Pergamon Press.
|
Hess, A.F. (1998). Personal communication between Dr. Alan Hess and Dr. Faysal Bekdash, SAIC,
June 2.
i
Hess, C.T., J. Michel, T.R. Horton, H.M. Prichard, and W.A. Coniglio (1985). "The Occurrence of
Radioactivity in Public Water Supplies in the United States," Health Physics 48: 553-553^
R-2
May 1999
-------
Hess, A.F., J.E. Dyksen, and H.J. Dunn (1983). "Control Strategy - Aeration Treatment
Technique." Occurrence and Removal of Volatile Organic Chemicals From Drinking Water,
American Water Works Association Research Foundation, Denver, Colorado.
HDR Engineering (1998). On-going EPA cost-estimating project for GAC.
Hiltebrand, D.J., I.E. Dyksen, and K. Raman (1987). "Radon in Water Supply Wells Treatment
Facility Requirements and Costs." Radon in Ground Water. Barbara Graves Editor, pp.
521-534. Lewis Publishers, Inc.
Hodsdon, A.E. (1993). Field Verification of Radon Treatment Costs for Very Very Small to Medium
Sized Water Systems. Unpublished draft; information used was predominantly from the
survey data in Appendix 3. Prepared for the American Water Works Association (AWWA).
Humphreys, K.K., editor (1984). Project and Cost Engineer's Handbook, Second Edition. Marcel
Dekker, Inc., New York, New York.
Humphreys, K.K. andL.M. English, editors (1993). Project and Cost Engineer's Handbook, Third
Edition. Marcel Dekker, Inc., New York, New York.
Kennedy/Jenks Consultants (1991). Analysis of Costs for Radon Removal from Drinking Water
Systems. American Water Works Association. Pre-publication copy.
Kinner, et al. (1993). "Using POE Techniques to Remove Radon." Journal AWWA, Vol 85, No. 6,
pp. 75-86, June.
Kinner, N.E., J.P. Malley, J.A. Clement, and K.R. Fox (1993). "Using POE Techniques to Remove
Radon." Journal AWWA, June.
Kinner, N.E., P.A. Quern, G.S. Schell, C.E. Lessard, and J.A. Clement (1990). "Treatment
Technology for Removing Radon from Small Community Water Supplies." Chapter 4 from
Radon, Radium, and Uranium in Drinking Water, Cothern, C.R. and P.A. Rebers, editors.
Kinner, N.E., J.P. Malley, et al. (1990a). "Radon Removal Using Point-of-Entry Water Treatment
Techniques." EPA-RREL, June.
Kinner, N.E., J.P. Malley, et al. (1989). "Radon Removal Techniques for Small Community Public
Water Supplies." Prepared under Cooperative Agreement CR-812602-01-0 with U.S. EPA
Office of Research and Development.
Kinner, N.E., G. Schell, et al. (1988). "Radon Removal from Drinking Water Using Granular
Activated Carbon, Packed Tower Aeration and Diffused Bubble Aeration Techniques."
Presented at Rn and Rn Reduction Technology (October 17-21), Denver, Colorado.
R-3
May 1999
-------
Kinner, N.E., C.E. Lessard, and G.S. Schel (1987). "Low-cost/Low Technology Aeration-related
Treatment Techniques for the Removal of Radon from Drinking Water." EP'A, WERL.
Klotz, M., P. Werner, and R. Schweisfurt (1976) "Investigations Concerning the Microbiology of
Activated Carbon Filters." Translations of Reports on Special Problems of Water
Technology, H. Sontheimer, ed., EPA Report, Vol 9, EPA 600/9/76-030.
,,i i 'i
Kruithof, J.C., R.C. van der Leer, and W.A.M. Hijnen (1992). "Practical Experiences with UV
Disinfection in the Netherlands." Aqua, Vol. 41, No. 2, pp. 88-94.
' j !
Land and Water Resources Center (1983). "Radon in Water and Air-Health Risks and Control
Measures." University of Maine at Orono, February.
Lenzo, F., PE. (1990). "Radon - Removing It From Ground-water." Public Works, November.
Longtin, J.P. (1988). "Occurrence of Radon, Radium and Uranium in Ground Water." Journal
A WWA, Vol. 80, No. 7, pp. 84-93.
Lowry Engineering (1993). Multi-Staged Diffused Bubble Aeration System Design Guide for
Drinking Water Applications.
Lowry Engineering, Inc. (1989). Design and Application Guide for the STRIPPER* Multi- Staged
Aeration Products.
" ; • j i :,
Lowry, J.D. (1990). Personal Communication from Dr. J.D. Lowry of LEI to Joe Drago of Kennedy
Jenks Chilton Engineers, San Francisco, CA, dated September 27.
Lowry, J.D., S.B. Lowry, et al. (1989). "Rn Removal by Field POE GAC Systems: Design,
Performance and Cost-Draft." Lowry Engineering, Inc., January.
Lowry, J.D. (1988). "Radon Progeny Accumulation in Field GAC Units." Final Report for
Research Contract No.88001, Department of Human Resources, Bureau of Health
Engineering, Augusta, Maine, March.
I I
Lowry, J.D., et al. (1988). "New Developments and Considerations for Radon Removal from Water
Supplies." Symposium on Radon and Radon Reduction Technology, Denver, Colorado,
October 17-21.
Lowry, J.D. and S.B. Lowry (1987) "Modelling Point-of-Entry Radon Removal by GAC." Journal
A WWA, Vol. 79, No. 10, pp. 85-88, October.
Lowry, J.D, et al. (1987). "Point-of-Entry Removal of Radon from Drinking Water." Journal
AWWA, Vol. 79, No. 4, pp. 162-169, April.
R-4
May 1999
-------
Lowry, J.D. and J.C. Brandow (1985). "Removal of Radon from Water Supplies." Journal of
Environmental Engineering, ASCE, 111:511-527.
Lowry, J.D., W.F. Brutsaert et al. (1984). "GAG Adsorption and Diffused Aeration for the Removal
' of Radon from Water Supplies." Annual Conference AWWA, p. 315, Dallas, Texas.
Malcolm Pimie (1992). "Pre-treatment for Iron and Manganese in Air Strippers." Memorandum
from Arun Gera, Malcolm Pirnie to Marc Parrotta, EPA OGWDW, September 25.
Malcolm Pirnie (1992a). "Iron and Manganese Pre-Treatment; Additional Disinfection after
Aeration." Memorandum from Arun Gera, Malcolm Pirnie to Marc Parrotta, EPA
OGWDW, January 29.
Martins, K.L. (1992). "Practical Guide to Determine the Impact of Radon and Other Radionuclides
on Water Treatment Processes." Wat. Sci. Tech., Vol.26, No. 5-6, pp. 1255-1264.
Martins, K. L., and A.G. Myers (1992?). "Controlling Radionuclides in Water Treatment Plants;
Removal, Disposal, and Exposure Minimization."
McCreary, J.J. and V.L. Snoeyink (1980). "Characteristics and Activated Carbon Adsorption of
Several Humic Substances." Water Research, Vol. 14, p. 151.
McElhaney, J. and W.R. McKeon (1978). "Enumeration and Identification of Bacteria in Granular
Activated Carbon Columns." Proceedings of the AWWA Water Quality Technology
Conference, Miami, Florida.
Michel, J. (1990). "Relationship of Radium and Radon with Geological Formations." Chapter 7
from Radon, Radium, and Uranium in Drinking Water, Cothern, C.R. and P.A. Rebers,
editors.
Mills, W.A. (1990). "Risk Assessment and Control Management of Radon in Drinking Water."
Chapter 3 from Radon, Radium, and Uranium in Drinking Water, Cothern, C.R. and P.A.
Rebers, editors.
National Research Council (NRC) (1997). Safe Water From Every Tap: Improving Water Service
to Small Communities. Committee on Small Water Supply Systems, Water Science and
Technology Board, Commission on Geosciences, Environment, and Resources. National
Academy Press, Washington, D.C.
National Sanitation Foundation (NSF) (1988). Internet site at http://www.nsf.org/programs/
dwtu.html accessed on July 27. Also, personal communication between Lisa Quiggle, NSF,
and Bill Labiosa, U.S. EPA, June.
R-5
May 1999
-------
Parrotta, M.J. (1991). "Radioactivity in Water Treatment Wastes: A USEPA Perspective." Journal
AWWA, Vol 83, No. 4, pp. 134-140, April.
i
Parson, F., P.R. Wood, and J. DeMarco (1980). "Bacteria Associated with Granular Activated
Carbon Columns." Proceedings of the AWWA Water Quality Technology Conference,
Louisville, Kentucky.
. I
Partridge, I.E., T.R. Horton, and E.L. Sensintaffar (1979). "A Study of Radon-222 Released from
Water During Typical Household Activities." Report for Office of Radiation Programs
Technical Note ORP/EERF-79-1.
Pennsylvania-American Water Company, Eastern Region, Comprehensive Planning Study. July,
1989.
j
I
Perry, R.H. and D.W. Green (1984). Chemical Engineer's Handbook. 6th Edition. McGraw-Hill
Book Co., New York, New York.
I ' J
i|
j
Quinton, D. (1998). Telephone conversation between David Quinton, Manager of the Radon Grant,
Environmental ManagementBranch, CalifomiaDepartment of Health Services, Sacramento,
California and Diane Loy, SAIC, McLean, Virginia. May 22.
Raffety, J.C. (1963). "Some Notes on a Plan for the Removal of Radon by Aeration." J. Instn. Wat.
Engrs., 17, 23-35.
Robinson, B., G. Reed, D. Christodos, and V. Chidambaviah (1990). Sequestering Methods of Iron
and Manganese for Treatment, pp. WT53-WT54, AWWARF.
.. |
Rost, K.L. (1981). "Report on Spray Aeration for Division of Health Engineering." Maine
Department of Human Services.
R.S. Means Company (1998). Plumbing Cost Data.
|
Sasser, M. and I.E. Watson, Jr. (1978). "An Evaluation of the Radon Concentration in North
Carolina Ground Water Supplies." Health Physics, Vol. 34, June, pp. 667-671.
Smith, B.M., Gruane, W.N., Higgins, F.B., Jr., and Terrill, J.G., Jr. (1961). '^Natural Radioactivity
in Ground Water Supplies hi Maine and New Hampshire." Journal A WWA, January.
: ' •: I'
Spencer, C.M., and W.E. Brown (1997). "pH Monitoring to Determine Aeration Effectiveness for
Carbon Dioxide and Radon Removal." Proceedings of the Winter AWWA Conference.
' :'
Treybal,R.E. x(1980). Mass Transfer Operations, Third Edition. McGraw-Hill, New York, New
York.
R-6
May 1999
-------
Umphres, M.D., and J.H. VanWagner (1986). "An Evaluation of the Secondary Effects of Air
Stripping." Prepared under Cooperative Agreement CR-809974 with U.S. EPA Office of
Research and Development.
Department of the Air Force (USAF) (1995). ASDC: A Microcomputer-Based Program for Air
Stripper Design and Costing. Prepared by Carnegie Mellon University. CE Report No.
92-204, USAF Report No. ESL-TR-91-40. October.
U.S. Environmental Protection Agency (U.S. EPA) (1998). Cost Evaluation of Small System
Compliance Options: Point-of-Use andPoint-of-Kntry Treatment Units. Prepared by the
Cadmus Group, Inc. April 20.
U.S. Environmental Protection Agency (U.S. EPA) (1998). Packed Tower Aeration (PTA) Cost
Model Operation Manual. Prepared by SAIC for the EPA. September.
U.S. Environmental Protection Agency (U.S. EPA) (1997). Technology Design Conference
Information Package. Prepared by SAIC for the EPA. November.
U.S. Environmental Protection Agency (U.S. EPA) (1997). Small System Compliance Technology
List for the Surface Water Treatment Rule. Office of Water, Washington, D.C. EPA815-R-
97-002.
U.S. Environmental Protection Agency (U.S. EPA) (1996). Ultraviolet Light Disinfection
Technology in Drinking Water Application - An Overview.
U.S. Environmental Protection Agency (U.S. EPA) (1996a). Technology Transfer Handbook:
Management of Water Treatment Plant Residuals. Office of Research and Development,
Washington, D.C. EPA/625/R-95/008.
U.S. Environmental Protection Agency (U.S. EPA) (1994) "Suggested Guidelines for the Disposal
of Drinking Water Treatment Wastes Containing Radioactivity." Draft.
U.S. Environmental Protection Agency (U.S. EPA) (1994a). Report to the United States Congress
on Radon in Drinking Water -Multimedia Risk and Cost Assessment of Radon. Office of
Water. EPA 81 l-R-94-001.
U.S. Environmental Protection Agency (U.S. EPA) (1993). Draft Response to Major Comments on
Treatment, Costs, and Occurrence ofRadionuclides. May.
U.S. Environmental Protection Agency (U.S. EPA) (1993a). Very Small Systems Best Available
Technology Cost Document. Prepared by Malcolm Pirnie. September.
R-7
May 1999
-------
U.S. Environmental Protection Agency (U.S. EPA) (1993 b). Guidance Manual for Compliance -with
the Disinfection Requirements for Public Water Systems Using Ground Water Sources.
Draft of October 1993. Prepared by Malcolm Pimie.
U.S. Environmental Protection Agency (U.S. EPA) (1992). Technologies and Costs for the Removal
ofRadionuclidesfrom Potable Water Supplies. Prepared by Malcolm Pirnie, Inc.
|.
U.S. Environmental Protection Agency (U.S. EPA) (1991). Guidance Manual for Compliance With
the Filtration and Disinfection Requirements for Public Water Systems Using Surface Water
Sources.
\ • j i!
U.S. Environmental Protection Agency (U.S. EPA) (1991a). Manual of Small Public Water Supply
Systems. EPA 570/9-91-003.
•, „ .'; i • i
U.S. Environmental Protection Agency (U.S. EPA) (1990). "Suggested Guidelines forthe Disposal
of Drinking Water Treatment Wastes Containing Naturally Occurring Radionuclides."
EPA/570/F-90/018.
1 i
U.S. Environmental Protection Agency (U.S. EPA) (1988). Aeration Alternatives for Radon
Reduction, Addendum to Technologies and Costs for the Removal of Radon from Potable
Water Supplies. Prepared by Malcolm Pirnie for EPA.
i
1 !
U.S. Environmental Protection Agency (U.S. EPA) (1987). Technologies and Costs for the Removal
of Radon from Potable Water Supplies. Fourth Draft. Prepared by Malcolm-Pirnie for the
EPA.
U.S. Environmental Protection Agency (U.S. EPA) (1985). Estimating Sludge Management
EPA 625/6-85-010. October.
U.S. Environmental Protection Agency (U.S. EPA) (1984). Estimation of Small System Water
Treatment Costs. Prepared by Culup/Wesner/Culp Consulting Engineers for U.S. EPA.
Municipal Environmental Lab. Cincinnati, Ohio.
U.S. Environmental Protection Agency (U.S. EPA) (1979). Estimating Water Treatment Costs;
Volume 2 Cost Curves Applicable to 1 to 200 mgd Treatment Plants. EPA
600/2-79-162b. August.
Valentine, R.L., et al. (1992). "Radon and Radium from Distribution System and Filter Media
Deposits," Proceedings of the AWWA Water Technology Conference, Toronto, Canada,
November 15-19.
Wade Miller Associates (WMA) (1992). Communication from Jeff Mosher, WMA to Arun Gera,
Malcolm Pirnie. January 30.
R-8
May 1999
-------
Zumdahl, S.S. (1989). Chemistry. D.C. Heath and Company, Lexington, Massachusetts.
R-9
May 1999
-------
-------
Appendix A-0
Conceptualized Diagram for PTA Configurations Assumed in the PTA-Cost Model
and the Direct Engineered Approach
-------
-------
Figure A-0-1
Influent Pipe
Pump Blower
Clearwell
Liquid Distributor
304 SS Shell
1 inch Plastic Saddle Packing
Liquid Redistributor
I
Effluent Pipe
Pump
A.0.1
-------
Figure A-0-2
Influent Pipe
| Pump Blower
Liquid Distributor
304 SS Shell
1 inch Plastic Saddle Packing
Liquid Redistributor
Effluent
Pipe
Clean/veil
A.0.2
-------
Appendix A-l
Detailed Breakdown of Estimated Costs for PTA and Cost Curves
Based on PTA-Cost Model
-------
' .iiiiiiL Jiiiiti
-------
1!
11
Si-
X
Q
UJ
Q.
C.
i§
i 2
. '5.
So
sSsassSaSIs
11
oo(.-
3 (N 00 OS I
tO 5 «3_ '
22S
O
i ~
51
*
si
55
— J
II
I*
I-I
I
= _
iiss
[2§i
>**«»«
sss
— S3
•ssss-suas:
) U
13
15
£ 2 S S S "
«H «. « » o S "
» S. S S j;. S. r>
« S
u <
il
a I
e c
IF
1!
.si
•i ^^
s
8SSS.S. S.|. S.».».».
°si
' •» S s 3. S
1!
II
!
_^,
3s
II
II
= sr
» s
u <
•sf
i!
I
loss^sss-sas
SSB.SI.IS-.-.-.-.
iM~~n»!S°>S-oo
S88ft*!-SiS«i».».
SSSSSftSS||s
S^SiSJ
° o. o = J
:aaaaa
!I.Sg.S|§lll
•==«*»S55
""•••.^^wSSjjjj
-------
!;
.11
t
'"3
is
I
s
II
H
II
ll
SsS|SS§|||i
»5S5S5ii§|i
aSiSMlB|i§
Sllliii'Sssg
*»**«»*»«»*»g4jj^;
•»'-
sgSspsSjIS
SSsaSgSsSSi
*********»v» — JJj
lls;is«S|5sis
lilsllssisisi
-------
Key for Tables A.l.l-A.1.4
Item
Design Flow
Average Flow
Initial Process
Pump and Blower
Effluent pumping
Clearwell volume
Clearwell depth
Clearwell area
Revised process
Engineering
Construction
Total Aeration Capital
Aeration O&M
Clearwell
Permitting
Assumption/Rationale
The selected design flow for cost analysis (1 1 plant sizes).
The average flow for the selected design flow (1 1 plant sizes).
Process equipment cost calculated by PTA-Cost Model. Includes steel support
column shells, internals, 2" plastic packing, blower, pump and motor, clearwell, ..
pipng, instrumentation, air duct, and electrical.
The cost of a redundant pump and blower to allow continued operation in case of
failure at a single point. Estimated from PTA Cost Model output as [SPump +
$Blower]/Number of columns.
When clearwell is added, a pump is necessary to pump from clearwell to
distribution system. Pump costs estimated by regression of pump cost versus
plant size from RS Means data.
Estimated by the PTA Cost Model
Estimated by the PTA Cost Model
Clearwell volume divided by clearwell depth (used for estimating land
requirement).
The sum of initial process, pump and blower, and effluent pumping costs. This
estimate accounts for redundancies.
Small systems (<1 mgd): 30 percent of Revised Process (15 percent of total
capital costs) based on" average of case study data for small systems.
Large systems (>10 mgd): 40 percent of revised process (20 percent of total
capital costs) based on TOP recommendations.
Medium systems (1 to JO mgd): 35 percent of process (17.5 percent of total
capital costs). Average of small and large system percentages.
Small systems (<1 mgd): 70 percent of process costs (35 percent of total capital
costs) based on average of case study data for small systems.
Large Systems (>10 mgd): 100 percent of process (40 percent of total capital
costs) based on TDP recommendations.
Medium systems (1 to 10 mgd): 85 percent of process (37.5 percent of total
capital costs). Average value of small and large system percentages.
The sum of Revised Process, Engineering, and Construction.
O&M costs (including labor, administration, and power) for aeration estimated
by the PTA Cost Model.
The cost of a clearwell is included in the PTA Cost Model output (i.e., initial
process). However, the cost of the clearwell may need to be "backed off' for RIA
scenarios.
The greater of 3 percent of Total Aeration Capital or $2,500 based on TDP
recommendations.
A. 1.3
-------
Key for Tables A.l.l-A.1.4 (Continued)
Item
Land (sq. ft.)
Land(S)
Housing
Pre-treatment Capital
Post-treatment Capital
Pre-treatment O&M
Post-treatment O&M
Assumption/Rationale
Land area is based on process footprint (clearwell area with allowance for
access) plus a sixty foot buffer zone between tower and adjacent land
The utility can purchase land for $ 1,000 per acre in rural areas (< 1 mgd), or $
10,000 per acre in urban areas (> 1 mgd).
20% of Revised Process Cost.
Generalized capital cost equations for iron and manganese removal
Generalized capital costs for disinfection following aeration
Pre-treatment O&M costs based on generalized O&M equations
O&M costs based on generalized O&M equations
A. 1.4
-------
CO
CO
+
X
= 1
o en
1 2
o- °°
ai +•
3
f-
crf
CO
+
X
'I-
CN
CO V
? x
LU V
co j-
•
S E
co —'
11
o o:
I »
O. (A
CO CD
O Q
II
g
G)
'<0
c
O
•o
(0
a:
ro
If
en a>
en "o
"- cu
o -•=
^ 3
O !£.
c.
TO
O
£
3
G)
($) JSOD
|B;OI
-------
CO
8
X
e o>
I §
3 m
•S +
UJ W
8
O
en
o
u i
>> i
CM
TT
"x % o
^C
O (I)
O 3>
s 2
il
u H
>. X
0.
ra
I
I
a
O
15
8 !£.
O
I
3?
u.
CD
<
S
O
O
O
o"
o o
o
o
WSO
-------
I
+
g o
I 8
I"
? +
CO CN
O X
°?-g>
111 C
co £
" IO
™x <=>
8 x
1!
•I p
>. o"
93243X + 6564
1 .2483X3 - 364.69X
o
o
2
C3
0
o
0.
o
o
o
o"
o
0
8
o
-------
to
CO
en
cq
^
5
•»—
+
X
pg
I
+
"x
§
s
v
?
O (Q
O O)
— (0
«3 >
O <
II II
>. x
5
s
8
q
o
oq
<
o
q I
^- E
8
o
p
o
o
q
O
($) jsoo W3O
-------
to
ff
O.
0)
*w
O
•o
O
"5
Oi
O >
£
'o.
re
O
3
O
o
o
§
o
S
o
pi
S££
II
O Q
II II
>. X
o
p
CJ
o
o>
•a
a
s I
•^ u.
e
01
o
o
o
s
p
o
-------
CD
O
CO
I • '•
s
. 05
•5 10
iS"^
1 °
o +
O
°
to O
5 «
? if
UJ v
I*- CO
X
CO
O
CD
CO
n CD
X If?
in II
OJ V
II p
>. d
TJ
D)
§ "sT
CO O
«j «=
O 0)
O D>
S 2
i §
O <
II II
>. X
i
la
o
•u
a
O
"to
o
D; «
£ I
o
o
<9
T^
|
ra
IT
§
8
O
o —
o -a
p en
I
0>
O)
o §.
o <
o
o
o
o
o
o
a
o
o
p
o
-------
5?
in
to
CO
r- CM
I §
I
•<•
O
04
O +
O LU
I
13)
.!=
jn
i=
o
13
ffi
K.
"e"
§
8 ^
o
5
o.
(0
o
I
o
ii q
>. o
11 CO
1 I
O U.
1 i
Q. CO
(0 V
O Q
ll II
>. x
o
o
O
o
o
o
8
o"
o
o
o
o"
o
p
o
(S) isoo IBlidBO
-------
CM
-------
O CM
It
D S2 CM
M V < H
* x $ x
2 j- g cd
S3?*1
I
I
_c
o
o
to
O"
Ul
i
•(•
X
CM
m o> w 2*
2 ^ « o « £
* $ ? ? HI
x
S
(0 ,D>
JJ »»
CO
^
0)
o
o>
•o
c
(8
o
00
CO
jj>
O
42
w
o
o
Is
'E
(0
o
1
D)
co
•a
ra
-------
CM
o'
O)
+
CO
in
CM
m
CM
.z »*/ *
« Si? £
•g
fe
co
CN
-(•
•1 B 5 3 ^
01 , O
p
o
II
O. Q
II II
>< x
o
II
£
JS?
u.
=0
O)
1
U.
_0)
(A
o
o
-------
r-
LO
S
c
.0
o>
Q §
X «
.2 CM
IS
UJ ^.
T— co co_
d
u
v .
if »
V
in
m
3
tr
UJ
O
o
d
to
x
m o
N. to o
l> " 7
^ + x
-^S
o CM r-
s s
Q CM
Ol
d
b §>
<"•(•<-
-2 x ^
.IS-?"
CO CO
g ... .-. P 'if oo
o
o
S
a
u
X
n
g
I
o
•D
(0
vg
I;
en
U
O)
a>
D.
2
3
Ol
o
o
•o
O)
I
tZ
CD
'w
($) sjsoo
-------
1
CO
M ^^ —^
CO £3 £^
sP CO . »
I » S Sf
?! S «N " o «
e w +0-53
O +• V CM TT o O
= ~x jf 4f U O C
g- ~^1 S v E g,
IU •qf q co CM co -5
o CM
—' 0>
II Q
>N II
x
»
§
O
a>
at
CO
b
I
I
0£
to
O
O
CM
2
I
u.
(O
-s <
T3
O)
8
($)
puen
-------
o
.1 5
«> •*•
e?»
o •:
o o
o,u.
'« c5>
CO X
"
c
•s.S
"55 CO
o
583
17 o o
x g"t
V •— c
-i
l!s££
•C§- I
IT5
II
>.
CO
§
o
c
o
•o
co
.8
(0
o
o
O)
'35
o
I
O)
o
o
•o
O)
1
u.
c
O)
o
CD
o
o"
o
o
o
o"
o
($) SJSOQ BuisnoH
-------
c
o
"c
=i
c
E
m
.0
Ta
3
cr
UJ
"in
o
O
(0
8
CO
X
CM
a>
CO
II
>»
in
h-
C5
0.01<=x<=
216.6X + 4866
in
N
CM
(O
s
1
CO
X
CNJ
§
(O
d
+
x
CO
o
0
•^—
0.75
05 o
"7 ii
+ v
Iv"
-85
*~ ci
II
>,
• X
CO
CT
LU
75
o
O
CO
d
as
CO
o
q
9
ii
1
i
o:
o
05
*•*
W
O
O
I
CO
•o
O)
E
I
LL
O)
($) s;soo BUISHOH
-------
CO
cri
in
o
CO
to
en
o ^
_g
to
o
ii
UJ T-
*8^ 5S;
CO V
.
(0 CO
O «
X
CO
x
CO
CO
CM
O5
in
II
Q)
*-*
n
£
fi
w
o
O
To
•**
'a.
a
O
V)
O)
•a
u>
I
in
a>
a
) S;SOQ
-------
CO
CM
o
CN
u.
0)
O)
($) SJSOQ
-------
It
>.
i
£
o
o
(O
o
CL
O
u
TJ
O)
o
£L
n
O
e
J3J
tn
I
O)
iZ
o
o
o
o"
o
o
o
o
o
o"
o
o
o
o_
T~
to-
o
o
($)
-------
S
O
O
CM
CO
o
cci it
•,<=
CO
CM
X
o>
o>
cri
T3
—. O)
^ CD
«-+(? o|
^ x <-> "-
ic^S g,
. oo °a co
t 2 O o>
x
CO
CD
II
o
0.
o
eg
O
oo
8
I
•<- -a
O)
LL.
0)
O)
5 <
O
O
O
8
O
O
o
o
o
o_
($) SISOQ W8O
-------
Appendix A-2
Raw Design and Cost-Estimating Data and Cost Curves
for Direct Engineered PTA
-------
I .'"ifpi',"1 ' 'I1'1 'NT '!' ' lilll!!'"' I'1" '"'
''I'!"1!?!1'!!!'!'!" •'! "''"I!!!'!'"IT "•» ! f|i'"• I'!" II
« 'jy'ii
-------
to
8
00
o
g
o
00
g
u
75
'5.
re
O
o.
to
I
0)
D)
2
dJ
c
O)
s
03 § «>- °l ^ ^
eo" co" oo" ° T~ ^ ^
N-_ r^ oo en <".
co" eo" oo" of 2
ooo
ooo
o
o
o
o
o
o
o
o
<«•«»««•<«•€«•««•««•««•
ooo
ooo
o
o
o
o
o
o
o
o
m
a
a>
o
in
1
5
(0
o
O
00
o
•o
n
5
'a.
n
O
n
E
3
(O
e> i
co" oo" oo" oi" o" 2 2 2"
0]
o"
§88
CO
-------
_
re
§
£
s?
o>
o
1
re
>P
£
CO
CD
re
CD
a
w
§
a
•a
c
re
.
re
O
re
E
3
(0
*•*
CO
Q
o
&S
O
4-1
W
o
O
i
Q.
re
O
"§•
Q.
2
3
0
u.
CD
O)
Q)
Q.
2
§T
iZ
c
'35
0)
Q
o
0
g
CO
o"
£
o
o
0
8
co"
in
*>»
o
o
o
o
o
o
CM
o
o
8
CO
o"
£
0
o
0
o
CO
o"
&3*
8
LD
o
0
o
o
o
o
o
CO
o"
£
8
0
o
oo
&^
in
CO
0
tn
o
0
o
of
o
0
o
0
CO
&^
o
in
CM
0
o
in
o
0
o
0
CM
co"
«•
o
o
o
o
CO
CM"
CM
m
CM
0
in
CM
8
o
o
CO
oo"
-------
s
X
in o S Q.
»-• o 2 9 « -2
o-
UJ £
o..S>
* *H tU H*
°S"S
S
9
u
1
oo
"O
Q>
*O)
o
£
Q
kt
O
o
CL
ra
O
£
3
O>
CO
csi
Q.
1
c
O)
o
§
o
g
o"
($) sjsoo
-------
CD
CO
+
X
CM
CO
O)
CM
CO
CO
X
CO
a
CO
w-
.o
to
o-
tu
8
o
oo in
~x ¥
OO V
h- x
CM II
CM V
o £
? 8
ii O
•8
O
§
\h
CO
ii
„
X
o
o
Q.
O)
1
0
o>
&
I
g
o
o"
Q
O
o
o
o
o
o
($) sjsoo
-------
^ + •„- o
o- x v co v •= "-
uf <° * • x B c
••* ^~ tl ». w -~ o)
I ° o £S ™f
."x
in
o
p
o
II
5
O
o>
g
OJ
0)
UJ
«rf
o
0)
o
"3
±i
a.
ca
O
*
05
iZ
o
o
o
o
in
CNJ
a
D)
8 I
_
10
0)
Q
o
o
o
8
($)
-------
O)
o
cv
o
i
1 5
to
j. tS O
C3 ^T "—
x o J2 3
5 {} O E
«' ff s §>
^Si
•^ 10 oo ^
o
II
o
p
o
II
re
I
5?
s
1
0
-------
O
oo
CO
CO
in
c
.2
111 +
O
O
CM
0>
CO
CD
O
.
CO
o
o
II
m o
*- o
~+?
X »
CO V
*<3* ?5
O i.
o ^
o to
CM
II
Q..
co
jn
to
o
£
55
o>
o>
•o
co
.
s
O
0)
"03
UJ
4rf
o
2
Q
XL
ID
O
O
I
a,
co
O
•9
CN
D)
O
O
O
o
CN
O)
8 1
" U.
0)
o
o
o
o
CO
o
o
o
o"
o
o
o
o"
o
o
o
o"
CO
o
o
o
o"
CN
o
o
o
o"
($)
-------
.9
Ha
U)
o
O
in
o -fr
o ii
f *
II v
>> in
(O
in
o
oo
»
« H1 +
+ x °o
~ V en
o
§
O «
II CM
II V
in
o>
o
o
J2
§
o
oo
1
0)
-------
Appendix A-3
Detailed Breakdown of Estimated Costs for GAC and Cost Curves
Based on GAC-Cost Model
-------
-------
X
O
UJ
a
OL
s«
!*
15
P
1!
•?
I!
TiS
£ J
^ •
• s
II
ff
11
gT-oa
O**t
-
*» *•
., -9 3. O W> O. — CO. -» tO -»
sssssss
i— ^.Moor
^i--
W i~" ^ of pi « to
_JoqqpSqqq5
w » « « r-, CM. «n «» co •*
^-''— *
3«— oo S (O to _
'"" sss'S
«>r-.^i--mcM~»-nintor-
5SSSSSS8
ss's
a s s x a s g s s s s
SSuuss2SsS^
**"*** r- r-'
8SSSSS53
£2 2 IS X *""•
*™<**™"'-—
("••_ •» •» *rt «D *» CO C»
wo'-ripciS
*~- '"*• "- - *~- """•
-
5
^tnSSSSSriNOco
". •». •- <*. ~. •*- •>. T ". S. •».
* s 5 s S S *• 31- 5 3
S" S S 9 9 S S S S S S
n- • "• to-
£
I
it
is
It
Toul
Capital
EHIuonl Revl
Pumping Proc
n
(mgd)
•- T
88S53HSS3
*» V»*» C U
sa
•»•» s 5 s •• 5
"5
sssssisssss
SSSS2SS2SS3!
a s u a s s °- ™- 2 3!
^ oi CM rt V (-" *-IV(i*:J«lOCs.O(MT-
ssgassssss|
S S 1 3 3 I 3 S 5 S
8 S S S 5" = S S 3" 5
•* ** **ss
"*s*sass
i s s s s s i 15 i S
•ir-Srtor-r^r^PiV*-
to w -T *o_ in
cs" o" 0" <•> o" <0" vn" tnn(Meoo>**>n
S y S K S r Sf S3 = S S
5 »
-------
Key for Tables A.3.1-A.3.3
Item
Design Flow
Average Flow
Initial Process
Effluent pumping
Revised process
Engineering
Construction
Total Capital
GAC O&M
Clearwell
Permitting
EBCT
Land (sq ft)
Land ($)
Contactor Volume
Housing
Pre-treatment Capital
Post-treatment Capital
Pre-treatment O&M
Post-treatment O&M
Total GAC O&M
Assumption/Rationale
The selected design flow for cost analysis (1 1 plant sizes).
The average flow for the selected design flow (1 1 plant sizes).
Process equipment cost calculated by GAC-COST Model.
When clearwell is added, a pump is necessary to pump to distribution_system.
Pump costs estimated by regression of pump cost versus pump size from RS
Means Plumbing Cost Data.
The sum of initial process and effluent pumping costs
Small systems (<1 mgd): 50 percent of Revised Process cost (20 percent of total
capital costs) as per TDP recommendations.
Medium systems (1 to 10 mgd): 71 percent of Revised Process cost (25 percent of
total capital costs) based on TDP recommendations.
Large systems (>10 mgd): 100 percent of Revised Process cost (30 percent of total
capital costs) as per TDP recommendations.
Small systems (<1 mgd): 100 percent of Revised Process cost (40 percent of total
capital costs) as per TDP recommendations.
Medium systems (1 to 10 mgd): 1 14 percent of Revised Process cost (40 percent of
total capital costs) as per TDP recommendations.
Large systems (>10 mgd): 133 percent of Revised Process cost (40 percent of total
capital costs) as per TDP recommendations.
The sum of Revised Process, Engineering, and Construction.
Output from GAC-COST model (includes transportation and disposal costs).
The cost of a clearwell estimated using the PTA-COST Model.
The greater of 3 percent of Total Aeration Capital or $2,500 based on TDP guides.
Empty bed contact time as output from the GAC-COST Model using a mean Kss
of 3.0 from a range of values reported by AWWARF 1997..
Land area is based on process footprint plus a sixty foot buffer zone between plant
and adjacent land. Calculated as (SQRT(1 .5*flow*EBCT/depth/Tt + 60))2 *TC.
The utility can purchase land for $1,000 per acre in rural areas (<1 mgd), or
$10,000 per acre in urban areas (>1 mgd).
Calculated as flow * EBCT.
20% of Revised Process costs.
Generalized capital cost equations for iron and manganese removal were used.
Generalized capital cost for disinfection following GAC.
Pre-treatment O&M costs based on generalized O&M equations.
O&M costs based on generalized O&M equations.
The sum of Pre-treatment O&M, Post-treatment O&M, and GAC O&M.
A.3.2
.liil. iiti, :•& ik
., in ,; 1.-,
i1, llliii,iiiilf'lil,iiiiil
iiteliii.i,.:-..:;? I 111 in. ln;/^.,; -fc;..i .,'
-------
CO
CO
CO
en
CM
f
LU
CO
LU
« I if
O |]j V
° TO
<•> o
I.
m OJ
^ E
+ o
^
h- v
o> m
•r- r--
o
CO
II
>.
ef
o o
O u.
1 i
o. 'in
co o>
O a
it II
>. X
o
•o
n
g
i
IO
£
1
o
'5.
n
U
eo
3
D)
CO
CO
I
($) iSOQ |BJ|dBO IBJ01
-------
s
m
*T
^~
+
8
Ii
I!
3
o
i
CM
8
'&*
CO
CT>
O
II
>>
Sx
CM
CO
o>
in
It
•o
O)
s,
ii ii
>. X
O
O
eo
i
8
g,
I
O
o
o
o"
o
o
o
o
o
s
O
($) jsoo WSO
-------
?
Ill
git
•2 in to
3+ •*•-
« o
— ox
SAO
? 5
"x°
8
X
CD
?T3
LLI 0>
CM £
V
O) X
CM V
1
co
o o
O u.
Q. V)
(0 CD
O Q
It II
>» X
II
>.
o
<
o
O
T3
CO
o:
1
iq
to
•o
D)
CO
o
(0
CO
a
O
Q.
IB
O
CO
I
p
o
o
o"
o
o
o
8"
o
o
a
§
o
o"
o
o
($) JSOO
|BiOJ.
-------
CD
CO
CO
•»•
£
3= +• CO
5 ~ d
It tl
>> x
o
o
o
O
<
O
i
ra
o:
I*.
o
IS
g
o
CD
CO
o
o
C3
g
o"
o
g
o
o
o
o"
o
o
T—
*»
{$) ?soo W5O
-------
i
CN
in
3
«o -t- •••:
o-'-x °
UI x
csi v
T— ia
o> r-
"x
CO
O U.
IS
Q. tn
m »
LU
w
o
o
£
'a.
re
O
eo
<
£
O)
o
o
o
o"
o
o
o"
p
o
o
o
o
o
o
($) isoo lejideo
-------
CO
co
CO
CO
O -Q
g en
™ cz
c T =
g ••- CM
ro * "?
= M« 9
cr .?* II
.iff £ v
1 § v
O co g
I
§
O V
V X
<=> CN
T- °°
i O
•a
o>
V) ^
s "•
O
-------
Appendix A-4
Investigation of Possible Off-Gas Emissions Regulations
-------
-------
Appendix A-4
Investigation of Possible Off-Gas Emissions Regulations
Purpose and Methodology
EPA is required by Congress to set a maximum contaminaat level (MCL) for radon in drinking
water. In order to meet an MCL requirement, many drinking water systems (particularly those
using ground water sources) will need to install treatment facilities to remove radon from
drinking water. Since air stripping is generally highly effective for removing radon from water,
many drinking water systems may install aeration facilities to comply with the radon rule. As
part of their operation, aerators generate off-gas emissions, which may result hi air permitting
requirements in the States.
To assess potential procedures (e.g., permit applications, off-gas risk modeling) and costs that
could be associated with radon off-gas from aeration facilities, U.S. EPA gathered information
from agencies responsible for air permitting. U.S. EPA contacted representatives from nine air
districts in California via telephone to determine the likely response of their district to
promulgation of a radon rule with an associated radon MCL requirement. The representatives
responded to the following questions:
• What is the likely response of your permitting board to water systems installing aeration
treatment to comply with the radon rule?
• What are the likely permitting procedures and costs for water systems installing aeration
for radon? Who would be responsible, the permitting board or the water system, for
carrying out each procedure and paying the costs?
• Will large water systems and small water systems follow different procedures, or are
procedures uniform regardless of water system size (e.g., off-gas volume)? How do
permitting costs change with the applicant's system size?
• Will water systems be required to perform off-gas risk modeling as part of the permitting
procedure or will they be required to do other environmental impact analyses?
• Would there be annual renewal procedures (e.g., reapplication, compliance monitoring)
and costs? Who would be responsible for carrying our the procedures and bearing the
costs?
• Is ongoing monitoring likely to be required?
Where possible, representatives provided estimates of time and cost that could be incurred by
water systems and the districts as a result of the potential district response to the radon rule.
A-4-1
-------
Results
The likely response to a radon rule is similar across the California air districts contacted. Most
districts indicated they are likely to follow the lead of the State. Following the State's lead
means that if the State includes radon on its Toxic Air Contaminants List and establishes unit
risk factors and exposure levels for radon, air districts will probably regulate drinking water
system aeration facilities through permits. Permitting procedures are similar across air. districts
and generally do not vary for facilities of different sizes. However, permitting costs and who
bears those costs can vary significantly from air district to air district. Some portion of the costs
are likely to vary based on facility size or emissions level.
;" ••: • , ! I •
A summary table of the information collected is provided below, followed by a more detailed
description of each air district's likely response to the radon rule and its associated permitting
procedures.
A few of the air district representatives raised the issue of the lack of technologies that water
systems could use to treat radon off-gas. As noted in the National Research Council's 1998
report titled Risk Assessment of Radon in Drinking Water1, treatment of radon off-gas emissions
from aerators could prove to be problematic since technologies tested for this purpose have
shown limited effectiveness. One air district representative suggested the possibility of
collecting the radon off-gas emissions and allowing the radon to decay in a controlled space.
Risk Assessment of Radon in Drinking Water. 1998. National Research Council
Assessment of Exposure to Radon in Drinking Water. National Academy Press, Washington
A-4-2
., Committee on the Risk
.DC.
-------
ill
60 2 fa
I 2 -
s-s
-•8
= 3 C
•5 j= o
•= « -o
1 g.8
U 3 c3
?- 8 e
« oo
111
III ills i
QO ~ a fe^-^ 8 E
-- -
«
T3
U
;-^*j«ttf'l»s
a
ys
n
O
T3
t> -^
W W
OT •!-•
•° >,
.— —
Js 'i
°^1
ill
ill
* E
s
•§
5
81
§,
00
"E
•-
s
~ p
§
O
1
O
00
o
(D
js
§•
o
o
s
= u
.2 o, a.
g o s s
•-
o
«>
§
0
c
'
P -a
cu aj
S § -1 OH
So o- -
_
>%ffu
M U
=3 <
>»
—
-------
~ o
o
S S
3
O
J2
i
3
O
•S
CO
3
"5,
o
'
* ** t01"™*1
1
CO
O
CJ
JS
submit single application
W)
(D)
operation (
-evaluation
ved, issue permit (D)
ction of aerator (W)
ion (D) & testing (W o
C
to
^
g, hea
odel
C- 60
U. C
2 '^
2 c *v ^~
co '5 C U O J O
T3 _3 .3 0 .S o ~
g "3 2 S § S •§
tllll^
2 -c
a -
in
C C4_ QA
Q
O,
g ^
CO >
o
&Q
S u
o ^
I &
i-l C
co
A-4-4
-------
i
I
IS
li
-"
W)
ly modeling,
t (D)
"8
y co
00
1
y te
ke
(D) & fac
(D)
& D)
atio
w/l
ss
ss
io
sue
n
ed
W
lic
w
ACP app
ce revie
ion, risk as
nce inspecti
ved, PO issu
O renewal
^ 3 _O O .2 O ~
sk
a
u
£J£>
O^
a, =
S*
£•3
ue ACP (D) & construct (\
ssued (D) & source test do
ns limit & issue PO (D)
is
io
l
S I""! ««
~ w co
•O °- '=
S £>s
2 S a
00
1
renewa equired (D) likely afte
review of testing protocol (D)
ric
2
5
c
o
o s
PH g
u. — CO
a a> o
S « o
I ^ ^—
3J
E
K = .=•
§• g 5
? ? g^
00 •<
c
I
U CH
P <
C
I
IQ
c O
u Cu
A-4-5
-------
Call to South Coast AOMD - Mohan Balagopalan (909-396-2704) - 2/23/99
. !l .|»l I, ' I
Currently, radionuclides (including radon) are on the Toxic Air Contaminant Identification List
developed by the California Air Resources Board. The List contains substances identified as
toxic air contaminants by the Board. Substances on the List are split into categories, and
depending on what category a substance is in, the substance may or may not have potency
numbers developed by California's Office of Environmental Hazard Health Assessment
(OEHHA) and approved by the State's Scientific Review Panel (SRP). These numbers include
unit risk factors and reference exposure levels. At the present time, OEHHA has not published
values for radon unit risk factor and reference exposure level. Radon is currently in Category 4A
of the List, and is proposed to be evaluated for entrance to Category 3. Substances in Category 3
are considered to be nominated for development of the above values.
• ' I !| j
Air quality districts generally follow the lead of OEHHA, meaning that if OEHHA publishes a
unit risk factor and reference exposure level for radon, air districts would likely evaluate whether
radon should be considered in their air permit programs. If OEHHA does not publish the
numbers, air districts are not likely to take action because they need unit risk factors and
reference exposure levels to establish permit limits. The process would likely be as follows:
1) U.S. EPA publishes an MCL for radon
2) OEHHA decides whether or not to evaluate radon and develop a unit risk factor and
reference exposure level (development of values could take as long as 2 years or more)
3) A- If OEHHA publishes values - then some air districts would take action
For SCAQMD, the first step would be amending Rule 1401 to add radon to the
list of compounds to be evaluated for permits. This would require a rulemaking
process, which would probably take around 6 months. If SCAQMD thinks there
are many sources in its District, they would go through this process. Whether a
facility would be required to add controls for radon off-gas from aerators would
be a case-by-case determination. Not all air districts would go through the process
of looking at health risk assessments and considering controls for aeration towers.
3) B- If OEHHA does not publish values - then air districts would probably not take any action.
SCAQMD is not sure what facilities would do to treat radon off-gas emissions since methods are
not readily apparent. One possible approach they might consider is looking at the endpoint of
what level facilities would need to meet for the emissions and back calculate to a certain flow
rate/volume that a facility would have to stay under (based on the radon level in its source water
and assuming all radon is stripped through aeration) in order to meet the emissions limit.
i
I
Permitting procedures are the same for all water systems, regardless of size: apply for permit to
construct (fee of approximately $2,000 paid by system); the District evaluates application and
decides whether or not to issue a permit; a permit is issued and the system constructs the aerator;
the District conducts an inspection and the system may or may not have to perform testing; a
public notice is issued if required by risk level and proximity of schools (cost of $400-500); the
District issues a permit to operate (no fee); the system must pay an annual renewal fee of about
$600-650 (no monitoring or inspection likely). Facilities with less than $500,000 gross income
and less than 10 employees receive a 50% reduction in the initial permit application fee, but
A-4-6
-------
water systems in SCAQMD are unlikely to be small enough to qualify for this reduction. In
determining what controls to require, the District does not look at cost and distinctions are not
made based on system size. Water systems are likely to need to do a risk assessment and perform
modeling as part of their permit application. Depending on which agency in the water system's
municipality takes the lead on the permitting process, the municipality may also produce an
environmental impact report (EIR), which could be costly.
Call to Bay Area AOMD - Kenneth Lim (415-749-4710) - 2/23/99
How the Bay Area AQMD (BAAQMD) would address the issue of radon off-gas emissions from
aerators at drinking water plants would depend on the concentration limit and the concentration
of radon in the water. BAAQMD has not had experience with facilities in the Bay Area that
remove radon from water. If the District were to consider issuing a permit to a water system for
radon off-gas, the District would perform simple, screening level modeling to compare off-gas
concentrations to acceptable exposure levels. The modeling would assume that the aerator is
100% efficient (i.e., all radon is stripped from the water by the aerator), and would take into
account factors such as local meteorology and the distance to the nearest residents. The analysis
would provide a risk assessment for screening purposes. If the screening analysis shows that the
risk is significant, the burden would then shift to the applicant and the applicant would need to
do more detailed modeling and may need to do a more comprehensive risk assessment
(requirements would depend on the magnitude of the emission loads involved).
B AAQMD has permitting categories that are based on facility type and the contaminants and
emissions produced. Permitting fees vary from category to category. A drinking water system
being issued a permit for radon off-gas would likely fall under the miscellaneous category since
there is currently no category for this type of operation. If permitting this type of facility became
a major issue (e.g., affected a significant number of facilities), the District might create a new
category. There are a range of permit fees within each category, depending on throughput (e.g.,
amount of emissions, fuel use, solvent use), so different size water systems could incur different
fees. The permitting process would be the same regardless of system size. Permitting fees are
automatically doubled if a toxic contaminant is involved. Permitting fees can range from
$50,000 for a large industrial facility to a couple hundred dollars for a little aerator at a corner gas
station.
Call to San Joaquin Vallev APCD (SJVUAPCD) - Steve Bonaker (209-230-6000) - 2/25/99
The impact on SJVUAPCD of a radon rule would largely be unknown since radon is not a
criteria air contaminant. If the state developed unit risk levels, S JVUAPCD would likely issue
permits to drinking water systems that install aeration treatment for radon. If that occurs and a
drinking water system decided to install an aerator to remove radon, the following process' would
likely occur:
1) Water system submits an application to the District for an authority to construct (AC) permit
(application fee is $60).
A-4-7
-------
2) The District evaluates the application. The parts of the evaluation may vary depending on how
radon is classified (air toxic, criteria contaminant, public health risk), but would likely include
modeling, a determination of whether the source is a major, health risk analysis (acute; chronic,
cancer), whether offsets are triggered, etc. The District would perform this analysis. The
analysis would probably cost the District around $1,000 (minimum for a typical evaluation is
around $480 for 8 hours of an engineer's time at about $60/hr), so the application fee does not
cover the cost of the evaluation.
3) If the District approves the construction, the AC permit is issued and the water system would
construct the aeration facility. The AC permit contains all operating conditions that will apply to
the constructed facility.
4) After construction is complete, the District determines if the facility meets the requirements of
the AC permit. To do this, the District inspects the facility and the water system may need to
perform source testing and would need to monitor influent and effluent. If requirements are met,
the facility is allowed to continue operating and the AC permit is converted to a permit to
operate.
5) To continue operating, the facility must pay an annual permitting fee, which is based on the
type of facility (aerators would likely fall under the electric motor category due to the power
supply to the aerator). Permitting fees vary by category and by size (horsepower for electric
motor category) within category (fee range for electric motor category is $74-882/yr). The
facility would need to do some ongoing monitoring of the water (influent and effluent).
If a water system had an existing aeration facility, the District would only need to issue the
Operating permit.
Call to Sacramento Metro AOMD - Bruce Nixon (916-386-6623) - 2/25/99
At the state level, California has its own large toxics board within its Air Resources Board
(ARB). The ARE develops air toxics control measures (ATCMs) that the districts use in their
permitting process. If a Federal pollutant control level (e.g., a NESHAPS) is promulgated, ARB
may directly adopt it, but instead usually performs its own evaluation and develops its own
values. Under CA law, a NESHAPS is automatically adopted as an ATCM. Regardless of
whether a pollutant control level is established by a NESHAPS or an ATCM, the Sacramento
Metro AQMD (SMAQMD) would use the value in its permitting process. The District has six
months to adopt a regulation after an ATCM is promulgated; if the District does not, an ATCM
automatically becomes enforceable by the District under CA law.
If radon becomes regulated, a facility that wants to construct an aerator for radon removal would
need to submit an application to SMAQMD for a construction permit. For most toxics, the
District performs short-term dispersion modeling using a screening model. If the District
approves the permit, the water system would then construct/install the aeration facility and would
need to obtain a permit to operate the facility. To issue the operating permit, the District would
need some verification that the facility is in compliance either through the facility performing
A-4-8
-------
testing or the District conducting an inspection. The permit would be valid for one year, and
would be renewable each year after verification that the facility is in compliance (through
facility's own testing or District inspection). Most permits do not require regular/ongoing
testing. Permitting procedures are the same regardless of facility size. Permitting costs are
generally associated with the air pollution control equipment (# BTUs/yr or # of motors/unit).
Initial permitting costs for most types of facilities range from $880-3,760, with an annual permit
renewal cost of half the initial amount.
Call to Ventura County APCD - Kerby Zozula (805-645-1421) - 2/25/99
Air districts do not develop individual toxics rules; the State takes that action. The CA Air
Resources Board (ARE) adopts MACTS for the entire State. Most MACTS do not apply to
drinking water systems and most systems do not have permits with the Ventura County APCD
(VCAPCD) since drinking water plants generally do not have engines, boilers, or similar units
that trigger the need for air permits. So VCAPCD may or may not issue permits to aeration
facilities at drinking water plants, depending on actions taken by the State.
VCAPCD uses a two-step permitting process. The first step involves an authority to construct
(AC) permit, and the second step involves a permit to operate (PO). A facility applies for an AC
permit and pays an application filing fee of $400 and an equipment/processing fee of around
$1,000. VCAPCD then conducts an evaluation to determine if the facility should be allowed to
construct the facility (e.g., if aeration is an acceptable treatment). For point sources that would
discharge contaminants with established unit risk values, the District performs a health risk
assessment that involves air modeling (estimated cost of $1,000 for about 20 hours of labor at
$60/hr). If the proposed facility is approved, the District issues an AC permit and the applicant
constructs the facility.
Once the facility is constructed, VCAPCD issues a temporary PO to allow the facility to operate
and perform a source test ($400 filing fee is charged to the facility, source test probably costs the
facility about $600). The source test is used to develop an emissions limit. The PO is issued for
one year, and can be renewed each year for an annual emissions fee that would probably be about
$500/yr for a drinking water aerator. The facility would probably be required to perform testing
once per year; VCAPCD would evaluate the protocol for the test and oversee the testing (fee
would probably be around $600/yr, unclear whether this is in the same as the annual emissions
fee or in addition to it). Since a drinking water aerator would not fall under any of VCAPCD's
current equipment types for permitting, the District would probably determine permitting fees
based on time and materials costs, so the fees could vary from system to system. VCAPCD
might need to add a staff person if there are many facilities to address as a result of radon limits
(cost could be as much as $70,000/yr).
Call to San Diego County APCD - Mike Lake (619-694-3313) - 2/25/99
If the aeration process could cause an air pollution problem, the San Diego County APCD
(SDCAPCD) could require facilities to obtain permits, regardless of what action is taken on the
A-4-9
-------
State level If the State is not involved and OEHHA does not establish risk factors and exposure
levels, SDCAPCD would pursue development of such numbers if the District determined there
seemed to be a significant risk. A single district is unlikely to develop public health'exposure
values on its own since it probably does not have the expertise. They would either join with
other districts to develop values, or more likely use any EPA values for acceptable exposure
levels for hazardous air pollutants or encourage EPA to develop exposure numbers. The District
has a general nuisance rule that can be used to regulate a source thought to be a risk, even if the
State has not taken action. For toxic ah- contaminants, OEHHA generally establishes risk factors
and exposure levels which the District then uses.
SDCAPCD does not typically require permits for municipal drinking water systems, so there
would be a cost to the District to set up a program. If the District determined after review that
emissions for aerators at drinking water systems needed to be controlled, there would a cost and
burden associated with developing permits and conducting ongoing inspections. Initially, the
District would assess what contaminants would be involved (since there may be contaminants in
the water other than radon that would also be removed through the air stripping process) and
whether they were at levels that need control. The District might ask facilities intending to use
aeration to conduct testing to determine what contaminants are present in the water to be aerated
ind provide me data to the District for its analysis. An evaluation could take 3-6 months. If the
District is conducting a general evaluation to assess whether to permit such facilities, the
Evaluation would cost around $15-20,000. If the District conducts facility-specific evaluations
because of differences in facilities, the cost would be around $5,0007facility. If public health
impacts are identified by EPA or OEHHA, SDCAPCD would need to do site-specific studies
Based on factors such as distance to the nearest residents. These studies would involve the
facility's collection of emissions data and submission of data to the District, and then the
District's evaluation of the data (District's evaluation would cost about $2,pOO/facility, with the
cost charged to trie facility). SDCAPCD is a full cost recovery facility, so all costs incurred by
the District need to be recovered through charges to the facilities involved.
If the District determines that aeration facilities at drinking water systems need to be under
fermit, the water system would submit a single application for both construction of the facility
and its operation The District performs an on-paper evaluation of the application and, if
approved, issues a construction permit. The water system then constructs the aerator.
SPCAPCD would then inspect the aerator and (District or water system) would conduct onsite
testing.
Call to San Luis Obisoo County APCD - David Dixon (805-781-5912) - 3/1/99
;:, .: , • / • ' ': ' I • " •
:. j.
San Luis Oblspo County APCD (SLOCAPCD) has never regulated radiation sources, since they
have fallen under the purveyance of the Federal or State government. However, SLOCAPCD
does regulate any facility that emits an air contaminant. So the District would likely issue a
permit to a water system that intends to install and operate an aerator to remove radon. The
District would hope that risk values would be developed that it could use for risk assessment
purposes. Any project that comes under permit in the District has to meet criteria of no excess
cancer risk greater than one in a million or no excess non-cancer hazard index of more than one.
A-4-10
-------
The permitting process begins when the water system applies for an authority to construct permit
(ACP). The application should include a risk assessment and modeling. A screening model is
sufficient, but if it shows significant impacts then more detailed modeling must be done. The
District then reviews the application and decides whether to issue an ACP. The initial
application fee is $100 for non-governmental facilities, with no charge for government facilities.
All water system would be charged for application review at a rate of $46 per hour. Once the
aerator is constructed, the District would inspect it (approximately 2 hours of labor charged at .
$46 per hour) and may require the water system to conduct testing to verify compliance with its
ACP. If the facility is approved, a permit to operate is issued by the District. The fee is likely to
be about $340 (facility would probably be in the miscellaneous category), plus a permit
processing fee for about 2 hours of labor at $46 per hour. The permit would need to be renewed
every year at a fee of about $340 (no difference for facility size). Annual or ongoing testing may
be required.
Call to Tehama County APCD - Gary Bovee (530-527-3717) - 2/26/99
To determine whether to permit aeration facilities at drinking water plants, Tehama County
APCD (TCAPCD) would need to assess whether the emissions would be significant enough to
warrant permits. TCAPCD does not currently issue permit to either drinking water plants or
sewage treatment plants (the county is fairly rural). If the State developed risk factors and
exposure levels for radon, TCAPCD would probably issue permits to aeration facilities used to
remove radon from drinking water.
The permitting process would entail:
1) Water system applies for an authority to construct (AC) permit (base application fee of $100).
2) TCAPCD conducts a new source review that would likely include modeling, projecting the
amount of emissions that would be released, and conducting a risk assessment that looks at the
risks to surrounding neighborhoods. The water system would be charged on a per hour basis for
the time it takes TCAPCD to process the application (including the source review) at a rate of
$63/hr (total cost would probably run between $500 and $1,000 for the AC permit process.
3) If the project is approved, the AC permit is issued and the water system constructs the aeration
facility. The District performs a compliance walk-through inspection to check if the facility was
built in accordance with the AC permit. The water system would probably be required to
conduct some testing to determine if the facility meets the emissions limits it said it would.
Testing could cost the water system as much as a couple thousand dollars.
4) If facility is approved to operate, the District issues a permit to operate (PO) that is valid for
one year. The water system would be assessed a PO fee that would also need to be paid each
year for PO renewal. A drinking water system aeration facility would fall under the
miscellaneous category for permits and the fee assessed would be based in the amount of
emissions ($135 for up to 5 tons/yr, with an additional $26/ton for amounts over 5 tons). The
District does not generally require annual testing in its POs, but the District's Air Pollution
A-4-11
-------
Control Officer (APCO) has the authority to require testing at any time (e.g., when a problem is
suspected).
Call to Monterey Bay Unified APCD - Fred Thoits T831-647-94ID - 3/20/99
Monterey Bay Unified APCD (MBUAPCD) has not traditionally treated radon as a point source,
so the District has not issued permits for radon. If EPA promulgates a rule for radon,
MBUAPCD would assess what action it needed to take, if any, based on the risk presented.
MBUAPCD does not develop its own risk values or exposure levels, but relies on the State
(OEHHA) to develop values. The District hopes that if EPA regulates radon that it would also
issue some heaJti effects data for determining exposure and risk levels. If there is potential for
public health exposure and there are risks of concern (but not at unacceptable levels), the District
would require permits for aeration facilities and would rely on EPA/OEHHA to develop risk and
exposure values.
The permitting process would be the same regardless of facility size. The water system would
apply for an authority to construct permit (ACP) and should include a risk assessment with the
application. Water system aerators would probably fall under the miscellaneous category, so the
ACP fee would likely be $674 ($103 for filing fee, $103 for toxics fee, and $468 for processing).
If the applicant did not perform an adequate risk assessment, the District would conduct one
during its application review process and would charge the water system on an hourly basis for
the work. If the aerator is approved, the District would issue an ACP and the water system would
Construct the aeration facility. After construction, the District would inspect the facility to
determine compliance with ACP conditions and may require the water system to perform testing.
If the facility is in compliance, a permit to operate (PO) is issued. There is no fee for the PO
Unless the inspection finds problems and more work is needed (work would be charged to the
yvater system on an hourly rate basis). The PO would need to be renewed on an annual basis,
with the fee.Jjasejj on the amount of emissions. No ongoing testing would likely be required
unless the facility was operating near the threshold for health/cancer risks. The facility would
heed to report the value for a surrogate emissions factor, such as gallons of water treated, that is
more easily measured than actual off-gas. The surrogate factor would be developed when the
ACP is issued and would be a means of estimating emissions using a multiplier to relate the
surrogate to emissions.
A-4-12
------- |