EPA-600/8-77-015
November 1977
NITRATE REMOVAL FROM WATER SUPPLIES BY ION EXCHANGE
Executive Summary
by
Dennis A. Clifford
and
Walter J. Weber, Jr.
University of Michigan
Ann Arbor, Michigan 48105
Grant No. R-803898
Project Officer
Thomas J. Sorg
Water Supply Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, U. S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U. S. Environmental Protection Agency nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
ii
-------
EPA-600/J8-f77-015
Novembei- 1J977
NITRATE REMOVAL FROM WATER BY ION EXCHANGE
Executive Summary
, /NO-
Np3
Municipal Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports ol the Office of Research and Development. U S Environmental
Protection Agency. have been grouped into nine series These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology Elimination of rraditior.al grouping was consciously
planned to foster technology transfer and a maximum interface in related fields
The nine senes are
1 Environmental Health Effects Research
2 Environmental Protection Techr 1 ology
3 Ecological Research
4 Environmental Monitoring
5 Socioeconomic Environmental Studies
6 Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8 “Special Reports
9 Miscellaneous Reports
This report has been assigned to the “SPECIAL REPORTS series This series is
reserved for reports targeted to meet the technical information needs of specific
user groups The series includes proolem-oriented reports, research application
reports, and executive summary documents Examples include state-of-the-art
analyses. technology assessments design manuals, user manuals and reports
on the results of major research and development efforts
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161
-------
FOREWORD
The Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health
and welfare of the American people. Noxious air, foul water, and spoiled
land are tragic testimony to the deterioration of our natural environment.
The complexity of that environment and the interplay between its components
require a concentrated and integrated attack on the problem.
Research and development is that necessary fIrst step in problem solu-
tion and it involves defining the problem, measuring its impact, and
searching for solutions. The Municipal Environmental Research Laboratory
develops new arid improved technology and systems or the prevention, treat-
ment, and management of wastewater and solid and hazardous waste pollutant
discharges from municipal and community sources, for the preservation and
treatment of public drinking water supplies, and to minimize the adverse
economic, social, health, and aesthetic effects of pollution. This publi-
cation is one of the products of that research, a most vital comunications
link between the researcher and the user community.
Serious and occasionally fatal poisonings in infants have occurred
following the ingestion of water containing high concentrations of nitrate.
This report presents the results of an investigation on the removal of
nitrate from water supplies by two-bed (strong-acid, weak-base) ion-exchange
treatment systems and by single-bed (chloride form) ion-exchange systems.
Detailed information is given on nitrate selectivity, rates, and capacities
for nitrate and competing ions, and regeneration requirements for various
commercially available anion—exchange resins. Also, an economic comparison
is made between the single—bed and the two-bed ion—exchange systems.
Director Francis 1. Mayo
Municipal Environmental Research Laboratory
•111
-------
ABSTRACT
Anion exchange using synthetic organic resins is a proven and practi-
cable technology for the removal of nitrate from water supplies. However,
disposal of the spent regenerant brine solution containing nitrate is a
potential problem. Two processes were examined in detail in this report--
single-bed strong-base anion exchange with NaC1 regeneration and two-bed
strong-acid, weak—base ion exchange with HC1 and NH 4 OH regeneration. Both
systems must be operated to nitrate breakthrough to minimize regeneration
costs. The two-bed process is one and one-half to two times as expensive
to build and operate as is the single—bed process, but produces softened
low-TDS, low-nitrate water, and has a readily disposable, spent regenerant
with fertilizer value. Important design considerations were found to
include the nitrate and sulfate concentrations In the raw water, the
service flow rate, the resin bed depth, and the nitrate/chloride selectivity
of the resin. The sulfate, nitrate, chloride, and bicarbonate selectivities
and multicomponent column behavior of the anion resins available from U.S.
manufacturers were examined and are reported in detail. An important
peripheral finding was that significant quantities of non-volatile organics
were leached from “clean” resins into the treated water.
This report was submitted in fulfillment of Grant No. R-803898 by the
University of Michigan under the sponsorship of the U.S. Environmental
Protection Agency. Work was completed as of December 31, 1976. A more
complete report will be published at a later date.
iv
-------
CONTENTS
• . . 111
• . . iv
• . . vi
• . . vii
• . .vlll
1
2
5
• . . 6
• . . 10
• . . 20
• . . 32
• . . 36
• . . 38
Foreword
Abstract
Figures
Tables
Abbreviations and Symbols
1. Introduction
2. Conclusions
3. Recomendations
4. Ion-Exchange Processes Studied . •
5. Experimental Procedures
6. Discussion of Results
7. Design Considerations for Water Supply Applications
References
Glossary
V
-------
F IGURES
Number Page
1 Conventional single-bed, ion-exchange process
2 Proposed two-bed, ion-exchange process
3 Chromatographic enrichment of ground water anions
in an exhausted anion exchanger
4 Resin phase concentration profile
5 Sulfate—nitrate and chloride-nitrate isotherms
for Duolite ES-368
6 Sulfate-nitrate and nitrate-chloride isotherms
for Niiberlite IR-45
7 Sulfate—nitrate and chloride-nitrate isotherms
for Dowex WGR
8 Experimental column set-up
9 Typical two-bed effluent concentration profile . .
10 Single-bed effluent concentration profile
11 Effect of sulfate selectivity on column efficiency 26
• • . • 11
• . • . 12
• • . 16
• . . 17
• • 18
• • • 19
• • • 22
• • • 23
vi
-------
TABLES
Number Page
1 Anion Resin Characteristics . . . . 14
2 Composition of Simulated Ground Water . . 15
3 Effect of Nitrate Concentration of Feedwater . . . 25
4 Effect of Bed Depth and Service Flow Rate . . . . . . 27
5 Column Performance Characterics 29
6 Economics and Regenerant Wastewater Comparisons Between
the Single-Bed and Two-Bed Processes 30
7 Calculated Chemical Regenerant Costs 31
8 Organics Leached by Agitating Conditioned Anion Resins . . . 32
9 Organics Leached from Conditioned Ion-Exchange Resins
Upon Standing 33
10 Ranking of Resins for Use in Nitrate Removal Service . . . . 35
vii
-------
LIST OF ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS (See glossary for definitions of terms)
BV - bed volumes
CaCO 3 - Calcium Carbonate, 50 mg/meq
GEL - gel or microporous polymer
HCHO - fcrrr.aldehyde
ID - inside diameter
Meq - milliequivalents, 10 gm equivalents
MR - macroporous polymer
N0 3 -N - nitrate measured as nitrogen, mg/i
NVOC - non-volatile organic carbon, mg/i
POLY - polyamlne functionality
Q-l - quaternary amine type 1 functionality
Q-2 - quaternary amine type 2 functionality
Quat — quaternary amine functionality
STY-DVB - styrene-divinylbenzene polymer matrix
TDS — total dissolved solids, mg/i
tert - tertiary amine functionality
TOC — total organic carbon, mg/l
Typ - typical
VOC - volatile organic carbon, mg/i
viii
-------
SYMBOLS (see glossary for definitions of terms)
N,Cl - the nitrate/chloride separation factor, dimensionless
aCl - the nitrate/chloride separation factor, dimensionless
aSN - the sulfate/nitrate separation factor, dimensionless
- the sulfate/nitrate separation factor, dimensionless
C 0 — initial concentration of ion in water, meq/l
C 1 - total concentration of ions in water, meq/l
Cl — chloride when used as a subscript
EM - maximum possible chemical efficiency, dimensionless
- overall chemical efficiency, dimensionless
ER - regeneration efficiency, dimensionless
N - nitrate when used as a subscript
N - normality, qm equivalents/i
Q - ion-exchange capacity of resin, meq/ml
S - sulfate when used as a subscript
t - superficial detention time, minutes
I - throughput, eq. of solution/eq. of exchanger
- bed volumes of effluent to nitrate breakthrough
x 1 - equivalent fraction of ion i in liquid, dimensionless
- equivalent fraction of ion i on resin, dimensionless
— average equivalent fraction of nitrate on the resin,
dimensionless, equals EM
7. - average equivalent fraction of ion I on the resin,
1 dimensionless
ix
-------
SECTION 1
INTRODUCTION
The maximum contaminant level (MCL) allowable for nitrate in community
and non-community water supplies has been set at 10 mg/l as N in the Interim
Primary Drinking Water Standards £19]. These standards were developed as a
result of the Safe Drinking Water Act of 1974 and became effective on
June 24, 1977. This nitrate level is equivalent to the long-standing,
recommended limit established by the U.S. Pdblic Health Service for the
prevention of methemoglobinemia in infants; furthermore, the National
Academy of Sciences [ 17] reported, in June of 1977 that, on the basis of
available scientific information, there was no real justification for
relaxing the MCL for nitrate. Public and private water supplies in nearly
all of the fifty states, especially CA, NY, and TX, and in many foreign
countries have been found to be polluted with nitrates in amounts regularly
exceeding this 10 mg/i limit. Nitrate removal by ion exchange with syn-
thetic, organic, anion—exchange resins is the treatment method which appears
to offer the most readily available, proven technology at a cost which is
lower than the alternatives -- reverse osmosis and electrodialysis. However,
disposal of the spent nitrate-containing, regenerant-brine solution is an
unsolved problem and, previous to the time of this research, there was a
lack of technical information in the literature regarding the selectivity
of the various anion exchange resins for nitrate with respect to the impor-
tant ground water anions: chloride, sulfate, and bicarbonate. Neither was
there sufficient, useful information available for the prediction of multi-
component effluent concentration profiles from ion-exchange columns economi-
cally operated by chromatographically eluting the ions not intended to be
removed.
The research described here was undertaken to provide the missing data
and to propose hypotheses concerning the prediction and control of anion
exchange selectivity in general. A further objective was to provide a
means of describing the muiticomponent chromatographic column behavior of
anion-exchange resins, especially weak-base resins, in nitrate removal
service. A final objective was to perform technical and economic evalua-
tions comparing a conventional, single-bed, strong-base, nitrate removal
process to a two-bed, strong-acid, weak-base, nitrate removal process which
would produce a spent ammonium nitrate regenerant amendable to disposal as
a fertilizer.
1
-------
SECTION 2
CONCLUSIONS
The preferences which comercially available anion exchangers exhibit
for the coninon ground water anions varies drastically, especially in dilute
multicomponent solutions with ions of differing valences. Naverthaless,
these drastic differences do not translate simply into good or bad column
performance in nitrate removal applications. Sulfate is always much pre-
ferred over nitrate and, below 3000 ppm TDS, the selectivity sequence for
ll comercially available resins is sulfate > nitrate > chloride >> bicar-
bonate. Strong and weak-base resins used for nitrate removal should be
chosen on the basis of high nitrate/chloride selectivity not on the basis of
low sulfate/nitrate selectivity. Surprisingly, high sulfate/nitrate
selectivity actually improves the economics of nitrate removal in column
operations.
The desirable anion resin characteristics for both single-bed and two-
bed nitrate removal service are high nitrate/chloride selectivity, moderate
to high sulfate/nitrate selectivity, moderate to high capacity, and macro-
porosity compared to microporosity. All thirty-two of the anion resins
tested could be ranked in preference order for nitrate removal service but
the differences among the reconinended resins were only slight.
Ion-exchange system operating costs for single-bed and two-bed systems
can often be reduced greatly by operating the anion bed to breakthrough of
the ion intended to be removed (nitrate) rather than by operating to the
ion-exchange capacity of the bed.
Kinetic considerations are important when eluting the ions n t
intended for removal. High service flow rates (> 2.5 gal/mm. ft ) and
shallow beds N 2 ft.) can result in premature breakthrough of the ion
intended to be removed and in a resin loading favoring the faster ions not
intended for removal.
Although predictably more costly to build and operate, feasible though
complicated systems can be designed for nitrate removal applications which
yield useful land—disposable regenerants.
The advantages (+) and disadvantages (-) of the single-bed and two-bed
processes for nitrate removal are as follows:
2
-------
Single-bed, strong-base anion with NaC1 regeneration
(+) Simple, no balancing of beds and regenerants
(+) Low cost regeneration, typically 9 /lOOO gal. water supplied
(+) Lower capital cost compared to the two-bed process
(-) Very difficult and costly to dispose of regenerants in noncoastal
locations where natural evaporation is impossible
(-) Iron must be removed to prevent resin fouling
(-) Continuous nitrate analysis required for process control
(-) Relatively high total operating costs, roughly 28 /lOOO gal. water
supplied if no cost Is assumed for regen rant disposal, and
37 /lOOO gal. water supplied assuming trucking and dilution
disposal.
Two-bed, strong-acid, weak-base, HC1 & NH regenerants
(+) TDS reduction and partial softening in addition to nitrate removal
(+) No problem with iron fouling. Precipitated iron is removed from
the cation bed during each regeneration cycle.
(+) Regenerant wastewaters expected to be easy to dispose of by land
application as fertilizer
(-) Complex system: bed sizes and regenerants must be balanced
(-) Degasifier for CO 2 removal required
(-) Continuous pH and nitrate na1ysis required for process control
(-) High regenerant costs, typ a1ly 28t/l000 gal. water supplied
(-) Higher capital costs comparea to the single—bed process
(-) Nigh total operating costs, roughly 55 /lOOO gal. water supplied
with regenerants given away as fertilizer
The usual method of downflow exhaustion with downflow regeneration
typically results in the need for 300 + I excess of regenerants for both
strong acid and strong-base resins. The need for excess regenerants becomes
greater whenever ions hig l,y preferred by the ion exchanger (e.g., SOA for
an anion exchanger and Ca for a cation exchanger) are involved. Although
not specifically investigated in this research, countercurrent (upflow)
regeneration can minimize the inefficient regeneration problem with its high
costs for regenerant chemicals and their ultimate disposal.
3
-------
All of the resins tested leached significant amounts of non-volatile
organic matter into the treated water even after the resins were cleaned and
conditioned. The total organic carbon (TOG) concentrations ranged from 3 to
90 mg/i for 1% resin solutions tumbled for 16 hours at 13 rpm and 25°C.
Non-volatile organic carbon (NVOC) concentrations ranged from 61 to 3870
mg/l for the supernatant waters from 50% resin solutions left standing in
distilled water for 6 months.
4
-------
SECTION 3
RECOMMENDATIONS
The degree to which various cation and anion resins yield leachable
organic compounds from breakdown of their polymer structures should be
quantified and the compounds identified. These unwanted hydrocarbons are
likaly to be chlorinated during the traditional, water supply disinfection
practices with the subsequent formation of possible carcinogenic compounds.
The problem has some degree of urgency as the legal provisions of the 1974
Safe Drinking Water Act will necessitate the more widespread use of syn-
tt’etic organic ion exchangers as the best available water treatment tech-
nology for removal of trace ionic contaminants, viz.: toxic metals,
fluoride, and nitrate.
Pilot plant studies of the single-bed and two—bed ion-exchange systems
described here should be undertaken to verify the cost estimates, assess
the relative design, operation and control complexities, and evaluate the
alternative means of regenerant disposal especially land application as
a fertilizer.
Mathematical models for the prediction of nitrate breakthrough profiles
should be developed based on the data presented here in conjunction with
multicomponent chromatography theory.
Instrument research should be undertaken with the objective of develop-
ing a simple on—stream nitrate analyzer for monitoring nitrate in water
supplies and in ion—exchange column effluents.
Polymer research should be undertaken with the objective of designing
ion-exchange polymers which will be selective for monovalent-nitrate ions
over divalent-sulfate ions based on the finding that the distance of nitro-
gen functional group separation is the most significant factor in univalent!
polyvalent ion separations.
The anticipated, beneficial, catalytic effect of carbonic acid and
bicarbonate ions on the nitrate removal efficiency of anion exchangers in
multicomponent ion-exchange service should be investigated. Results of
such an investigation should resolve the question of whether to place the
system degasifier upstream or downstream of the anion exchanger.
5
-------
SECTION 4
ION-EXCHANGE PROCESSES STUDIED
THE SINGLE BED PROCESS (Figure 1)
The.”usual” [ 1, 8, 9, & 13] and conceivably the most economical means
of nitrate removal from dilute solutions (<3C00 mg/i IDS) using ion—exchange
is a single-bed (fixed or moving) strong-base anion exchanger in the chlo-
ride form as described in Figure 1. With a typical ground water having the
analysis shown, at least 25% of the feedwater may be safely bypassed and
still provide a water of acceptable nitrate concentration (5--b mg/i) right
up untii nitrate breakthrough. Offsetting the innerent economical advan-
tages of the process are the following serious disadvantages.
(1) Resin selectivity for nitrate was originally though to be a
potentially serious problem because sulfate was expected to be preferred
with a selectivity ratio of more than 2/b over nitrate.
(2) When present, ferrous iron oxidizes, precipitates, and seriously
fouls the resin.
(3) Regeneration and brine disposal are the major economic and environ-
mental problems yet to be solved even with low-cost NaCl regeneration.
THE TWO-BED STRONG-ACID WEAK-BASE PPOCESS (Figure 2)
Because this two—bed process see’ried to have certain advantages with
respect to regeneration efficiency, iron removal, regenerant disposal as
fertilizer and nitrate selectivity, “e little-studied weak-base resin
portion of the two-bed system shown in Figure 2 was examined in detail.
The thermodynamic and kinetic data from the weak-base resin studies were
compared and contrasted to the data obtained from strong-base resins of the
type used in the single-bed process shown in Figure 1.
Evans [ 7] reported on a similar two-bed nitrate removal process, but
with HC1 and lime as the regenerants. His data indicated that, even after
the cation bed was exhausted, and sodium was being eluted, the system con-
tinued to provide softening and nitrate removal thereby delivering greater
than stoichiometric efficiency due to the weak-base anion resin’s apparent
selectivity for nitrate over all the other ions present including sulfate.
Actually, in the total concentration range experienced with water supplies
(100 to 5000 mg/i TDS), nitrate is thermodyn mical1y not the most preferred
6
-------
ion. The often reported sequence for both strong and weak-base resins is
(1, 5, 6 10 & 16]
sulfate > nitrate > chloride > bicarbonate
In this concentration range, both cation and anion exchangers of the syn-
thetic organic resin type much prefer divalent over monovalent ions (although
carbonate is an exception, it being lesspreferred than nitrate [ 16)). This
preference for multivalent ions at low concentrations is referred to as
“electroselectivity” [ 10); it diminishes with increasing solution concentra-
tion, and in the cases of sulfate/nitrate and sulfate/chloride exchange,
reverses to favor the monovalent ions at TDS concentrations above approxi-
mately 5000 mg/i as CaCO 3 [ 1, 3 & 15].
7
-------
Raw Water Influent
Flow :Q
Nitrate-N 20mg/I
TDS =380 ma/I
Hardness :225mg/I
NaHCO 3
Ca (NO 3 ) 2
Mg SO 4
Ca Cl2
Fe SO 4
Bypass L ___
I
Spent Regenerant
NaCI —NaNO 3 Brine
(Disposal Problem)
— Ion Exchange Column
Effluent
Ca Cl 2
MgC I 2
NaCI
Fe Cl 2
— Blended Product Water
Nitrate -N :5-10 mg/I
TDS 296-380mg/I
Hardness 225 mg/I
Chloride :53-195mg/I
Figure Conventional Sing!c Bed Ion Exchange Process
Regene rant
1 NaCI
I (low cost)
Strong
Base
Anion
Exchanger
Chloride
Form
0.25 Q
0.75Q
Raw
Water
8
-------
—Raw Water (Typical)
Flow=Q NH 4 OH
Nitrate N= 20 ppm Regenerant
TDS =380 ppm
Hardness =225 ppm
NaHCO 3
Ca (NO 3 ) 2 HNO Regénerant I
Mq So 4 I I Alternatively,] I
CaCI 2 LHcl or H 2 S0 4 I
_______________________________________ I
FeS O 4 ___
—Ii_J_,
Strong Weak
Acid Base
Cation Anion
Exchanger Exchanger
H
Form Free
Base
_____ Form
— - Cation Effluent — —
I H 2 C0 3
HNO 3
I H 2 S0 4
HCI
1’
L_ ___ ___
Spent Acid Spent -Ion-Exchange
I Amonia Column Flow:.75Q
Bypass Raw Water I
Flow=.25Q
y Blended
Combined Regenerants Product Water
NH 4 NO 3 Solution Nitrate - N = 5 -IOppm
(Fertilizer) lOS = 95-380 ppm
Hardness :56-225
ppm
Figure 2 Proposed Two-Bed, ron-Exchange Process
9
-------
SECTION 5
EXPERIMENTAL PROCEDURES
THEORY
With experimentally determined separation factors, multicomponent
chromatography theory [ 11, 2, 14 & 18] can be used with some simplifying
ssLImptions to predict the resin-pha!e concentration profiles of the anion
exchangers used in the single—bed and two-bed processes. The ions removed
by the resin tend to become separated and concentrated into enriched zones
contairdng predominantly one species as depicted in Figure 3. A hypotheti-
cal, semiquantitative profile based on the selectivity sequence given above
showing the idealized distribution of sulfate, nitrate, chloride, and bicar-
bonate is given in Figure 4 for the case in which the feed water contains the
same equivalent concentrations of the four anions shown. The most preferred
species, sulfate, is removed preferentially and is concentrated near the
inlet to the column. Bicarbonate, the least preferred species, is removed
nearer to the exit end of the column and is the first ion other than the
exchanging ion to show up in the effluent. Nitrate and chloride being of
intermediate affinity are concentrated in zones between these two extremes.
Quantification of the breakthrough curves for all the ions may be approxi-
mated if the selectivity coefficients (K’s) or separation factors (a’s) are
known for the ions of interest [ 12, 14]. These constants are defined below
for an example sulfate/nitrate ion-exchange reaction on a typical anion
resin:
2 RNO 3 + SO R 2 S0 4 • 2NO
124 3
K = Sulfate/Nitrate -_____________
S,N Selectivity Coefficient rsO=1 1 RNO
I 4 J L
= Sulfate/Nitrate = 3 ’s /YN
aS,N Separation Factor XS/ XN
CO xN
KS,N = cLSN
10
-------
inlet
Zone I Sulfate
rich
Zone 2 Nitrate-
rich
ANION
Zone 3 Chloride - EXCHANGER
rich
Zone 4 Bicarbonate-
rich
outlet
FIGURE 3
Chromatographic Enrichment of Ground
Water Anions in an Exhausted Anion
Exchanger
11
-------
1.0 — _____
YcI3 Y 54 ,’
SO 4 N,2 NO 3
ZONE I ZONE 2 ZONE 3 ZONE4
(Sulfate) (Nitrate) (Chloride) (HCO )
NO 3 CI 1 2 CI
CI,I Cl
0.0 DISTANCE INTO BED
meq Exchanger/meq Soin. 1.0
FIGURE 4
HYPOTHETICAL RESIN PHASE CONCENTRATION PROFILE
= eq. fraction of chloride in zone i
y , 1 =eq. fraction of sulfate in zone I
N,I =eq. fraction of nitrate in zone I
N,2 :eq. fraction of nitrate in zone 2
B,4 =eq. fraction of bicarbonate in zone 4
12
-------
where: R ion-exchange site on resin
R 2 S 0 4 = sulfate in the resin phase
SO = sulfate in the liquid phase (water)
[ ) = denotes concentration in eq/l
= equivalent fraction sulfate on the resin
XS = equivalent fraction sulfate in the water
C 0 = Total conc. of liquid phase in eq/i
Q = Total capacity of the resin in eq/l
Separation factors and selectivity coefficient can be detennined from
(isothermal) plots of resin phase concentration vs. water phase concentra-
tion of the ions being exchanged -- binary ion-exchange Isotherms. One
important objective of the research was to develop sulfate/nitrate and
chloride/nitrate isotherms for all the commercially available strong and
weak-base anion exchangers manufactured in the U.S. Column two of Table 1
contains a listing of all the resins evaluated during the research study.
SELECTIVITY STUDIES
The experimental sulfate/nitrate isotherms were constructed by equili-
brating (for 16 hrs.) weighed samples of resins in the nitrate form with
0.005 N H SO 4 (250 mg/l as CaCO ) solutions and measuring the redistribution
of ions, t.e., ion exchange, which took place. At least five data points
were obtained for the construction of each isotherm with each of the 32
resins. Nitrate/chloride isotherms were constructed for 20 of the resins by
equilibrating weighted samples of resins in the nitrate form with 0.005 N
HC1 (250 mg/l as CaCO .. ). Some examples of sulfate/nitrate and nitrate/
chloride isotherms ar given in Figures 5-7. Bicarbonate/nitrate selec-
tivity studies were also conducted on 12 of the resins utilizing a similar
technique in a closed system to prevent the escape of CO 2 .
iON EXCHANGE COLUMN STUDIES
After a preliminary evaluation of the binary equilibrium data, resins
were selected for laboratory scale (1” ID by 24” deep) column experiments in
which the chrornatographic elution of ions was to be observed and hopefully
quantified with the help of the experimentally determined separation factcrs.
A simulated ground water (See Table 2) was used in most of these studies.
Ten of the eleven multicomponent column runs were made with the two-bed
system described in Figure 2 as simulated by the experimental flow system
shown in Figure 8. One comparison run was made with the single-bed system
described in Figure 1 by simply bypassing the cation column shown in Figure 8.
Complete experimental details, results, and analysis for both the selec-
tivity and column studies can be found in Reference 4.
13
-------
TABlE 1: ANION RESIN CHARACTERISTICS
UN Resin
Number
Manufacturer’s
Designation
Matrix
Functionality
Porosity
meq/ml
Advertised
Capacity
meq/ml
Measured
HC1 Capacity
pKa
Average
S
N
Average
N
Cl
15
AMBERLITE IRA-400
STY-DVB
Q-1
MICRO
1.40
1.53
>13
1.89
-
17
AMBERLITE IRA-900
STY-DVB
Q-1
MACRO
1.00
1.10
>13
1.71
3.41
21
DOWEX SBR
STY-DVB
Q-1
MICRO
1.40
1.66
>13
1.89
2.90
27
IONAC ASB-1
STY-DVB
Q-1
MICRO
1.40
1.39
>13
1.87
-
32
IONAC AFP-100
STY-DVB
Q-1
MACRO
1.20
1.07
>13
1.76
2.97
16
AMBERLITE IRA-400
STY-DVB
Q-1
ISO
1.25
1.16
>13
3.09
3.11
19
DOWEX SBR-P
STY-DVB
Q-1
ISO
1.20
1.02
>13
2.96
-
22
DOWEX 11
STY-DVB
Q-1
ISO
1.20
1.17
>13
3.37
-
24
DUOLITE A-1O1-D
STY-DVB
Q-1
ISO
1.30
1.32
>13
2.59
-
28
IONAC A-641
STY-DVB
Q-1
FM
1.16
1.21
>13
3.30
3.30
30
IONAC ASB-1P
STY-DVB
Q-1
ISO
1.35
1.13
>13
2.59
-
14
AMBERLITE IRA-910
STY-DVB
Q-2
MACRO
1.00
1.31
>13
3.26
2.85
18
AMBERLITE IRA-410
STY-DVB
Q-2
MICRO
1.35
-
>13
2.40
-
20
DOWEX SAR
STY-DVB
Q-2
MICRO
1.40
1.50
>13
3.04
-
23
DUOLITE A-102-D
STY-DVB
Q—2
MICRO
1.40
1.48
>13
3.26
-
29
IONAC ASB-2
Sfl-DVB
Q-2
MICRO
1.52
1.33
>13
3.04
3.64
1
AMBERLITE IRA-93
STY-DVB
TERTIARY
MACRO
1.25
0.98
7.7
3.75
4.86
5
DOWEX MWA-1
STY-DVB
TERTIARY
MACRO
1.10
1.15
7.6
2.67
4.43
8
DUOLITE ES-368
STY-DVB
TERTIARY
MACRO
1.30
1.43
7.8
2.83
3.87
12
IONAC AFP-329
STY-DVB
TERTIARY
MACRO
1.25
1.26
8.5
3.07
4.14
3
AMBERLITE IP—45
STY-DVB
POLY
MICRO
1.90
1.76
7.9
12.7
3.89
2
AMBERLITE IRA-68
ACRYLIC-AMINE
TERTIARY
MICRO
1.60
1.42
11.1
23.4
1.89
10
DUOLITE ES-374
ACRYLIC-AMINE
POLY*
MACRO
3.0
2.59
9.9
94.0
3.85
6
DUOLITE A-7
PHEN OL-HCHO-PA
POLY**
MACRO
2.4
1.67
7.7
108
3.35
9
DUOLITE ES-561
PHENOL-HCHO-PA
POLY
MACRO
2.0
1.22
6.8
109
2.65
11
IONAC A-26O
ALIPHATIC-AMINE
POLY
MICRO
1.8
1.81
10.6
54.0
2.25
4
7
13
DOWEX WGR
DUOLITE A-340
IONAC A-3O5
EPOXY-AMINE
EPOXY-AMINE
EPOXY-AMINE
POLY
POLY
POLY+
MICRO
MICRO
MICRO
1.0
2.6
3.5
1.53
2.54
1.51
7.9
8.7
137
82.9
108
1.99
1.70
-
= Polyamine not including quaternary amine
= Quarternary Amine - Type 1
= Quarternary Amine - Type 2
= Isoporosity or ‘lmproved Porosity”
= Fixed Macropore (MANUFACTURER’S TERMINOLOGY)
= Advertised as tertiary amine but titrates as polyamine
= Advertised as secondary amine but titrates as polyamine
= Polyamine Including quaternary amine
POLY
Q- 1
Q- 2
ISO
FM
POLY*
P0LY**
POLY+
-------
TABLE 2: COMPOSITION OF.SIMULATED GROUND WATER
Ion X 1 meq/1 mg/i
Cations
Ca .546 3.0 60
Mg .273 1.5 18
Na .181 1.0 23
++
Fe nil nil 1
Total Cations 1.00 5.5 102
Anions
S0 4 .273 1.5 72
NO 3 - .273 1.5 93
Cl .273 1.5 53
HC0 3 . 181 1.0 61
Total Anions 1.00 5.5 279
Total Concentration of ions, C 1 = 0.0055 N
Hardness = 225 mg/i as CaCO 3
Total dissolved solids, lOS = 381 mg/i
Nitrate as nitrogen, N0 3 -N = 21 mg/i
Equivalent fraction of nitrate, XN = 0.27
15
-------
RESIN NUMBER 8
DUOLITE ES 368. rIACROPOHOUS RESIN
STYRENE—OVB MATRIX
TERTIARY—AM I NE FUNCTIONALITY
TOTAL CAPACITY=l.3 MEQ/ML
FIGUHE 5
25° C. BINRI9Y IONI—EXCHRNGE
ISOTFIEI9H
0
0
.-.
rJ,
z
a-
z
L J
z
0
-a
C-)
0
I-
I-
C - )
I-
-J
>
L U
0.20 O.t 0 0.60 0.80
‘( . EQUIVALENT FRFICT ION SQ IN LIQUID PHASE
‘c,. EQUIVALENT FRACTION CL IN LIQUID PHASE
1.00
16
-------
0
0
I I I I I
LU
L 1 JCI)
(flcl
=cLo
—U,
‘ (‘L i i
LAJ
D
— —
U’-’
ac
LLLL.
I-
jLU
—i--J
>>
tJJ. e SULFATE—NITRATE
a
o CHLORIOE-NITRATE
0
o £
I I I I
bOO 0.20 0.t40 0.60 0.80 1.00
EQUIVALENT FRACTION SO IN LIQUID PHASE
X , EQUIVALENT FRACTION CL IN LIQUID PHASE
RESIN NUMBER 3
AMBERLITE IR LLSP MICROPOROUS GEL
STIRENE—OVB MATRIX
POLYAMINE FUNCTIONALITY
TOTAL CAPRCITY=1.9 MEQ/ML
FIGURE 6
25° C, BIN RRT ION—EXCHRNGE ISOTHERM
17
-------
0.20 0.40
EQUIVRLENT FRACTION
EQUIVALENT FRACTION
0.60 (‘.80
SOq IN LIQUID PHASE
CL IN LIQUID PHASE
RESIN NUMBER q
DOWEX WGR, MICROPOROUS GEL
EPOXT—AMINE MATRIX
POLYAMINE FUNCT IONALITY
TOTAL CRPACITY=1.O MEQ/ML
25° C, BINRH’(
FIGUHE 7
I ON—EXCHANGE
0
0
U
C’,
cc
Q o
z.
.— .0
U,
Li
c i
z
0
0
0
C i
c i
L& ’
i_0
z
U
.1•
cc
Do
c’J
LU.
0
U
U,
cc
=
a-
z
U,
LU
ci
0
-J
L)
0
Ci
cc
ci
I-
z
LU
-J
cc
>
I-
LU
—4-_-_-
0
SULFATE—NI TRATE
0
0
c 00
CHLORIDE—NITRATE
.10
so1I&.
1.00
ISOTHE19M
18
-------
Two
Plexiglas
Columns
2.54 cm 1.D.
I .52m. long
Resin Depth
61 cm (Typ.)
Feed water
Pump
0-450 mI/mm
NH 4 OH Pump
0-20 mI/mm
Automatic Sampler
(24:500 ml Bottles)
Strip Chart Recorder
N. 0. Valve
N.C. Valve
Figure 8 Experimental Column Set-up _
0
To Waste
40/
NH 4 OH
1.14 N.
-------
SECTION 6
DISCUSSION OF RESULTS
RESIN SELECTIVITIES
AU nineteen strong-base and thirteen weak-base re3ins coimi ercial1y
available from American manufacturers were tested for sulfate, nitrate,
chloride, and bicarbonate selectivities. These selectivities were then
related to the following resin properties: matrix, functionality, porosity,
capac 1 ty, basicity, and type. A sigrificant finding was that resin matrix
is the most influential factor in the determination of divalent/monovalent
selectivities. It is hypothesized that if many pairs of closely spaced
ion-exchange sites are present in the resin matrix then the resin is very
divalent ion selective. High divalent selectivity can be designed into a
resin first by choosing the proper resin polymer structure (non-polystyrene >
polystyrene) and second by selecting the proper functionality (primary >
secondary > tertiary > quaternary amine). See Reference 4 for details.
Sulfate was always preferred over nitrate by all the strong and weak-
base resins tested which exhibited an extremely wide range of selectivities:
= i.71--3.37 for strong—base resins and 2.67—-137 for weak-base resins.
(14O e: The larger is cx the greater is the resin preference for sulfate
over nitrate.) It is eX ’êcted that the sulfate preference observed will hold
true for any resin tested with feed dters having total dissolved solids
concentrations up to at least 0.06 N 3000 mg/i as CaCO . ). Example isotherms
for moderately (Figure 5) strongly ‘ ‘ire 6) and very strongly (Figure 7)
sulfate-selective resins are given ‘ ‘ comparison purposes. The uppermost,
convex, curve in each figure depicts ‘e sulfate/nitrate equilibrium rela-
tionship for that resin. The degree )f curvature exhibited by the isotherm
indicates the resin’s preference for u1fate over nitrate. Compare Fig-
ures 5 and 7: when x = 0.10, y = C 41 for resin No. 8 (Figure 5) while
y = 0.98 for resin Nd. 4 when x 0.10. In other words, when sulfate
r presents 10% of the negative i8nic charges in the equilibrium aqueous
phase, 41% of the resin phase exchange sites are occupied by sulfate in the
moderately sulfate selective resin No. 8 (Figure 5) while 98% of the exchange
sites are occupied by sulfate in the very strongly sulfate selective resin
No. 4 (Figure 7). In both examples, the sites not occupied by sulfate are
occupied by nitrate, as these isotherms were derived from binary, sulfate-
nitrate mixtures.
Nitrate was always preferred over chloride by all the anion resins
tested although the range of preferences was relatively narrow:
2.85--3.64 for strong-base resins and 1.70--4.Bó for weak-base res’fñ
I-
-------
As expected, , was independent of total solution concentration.
The chloride/nitrat’ ’c otherms, the lower curves in Figures 5-—7, demon-
strate a weak nitrate preference (Figure 7) and moderate nitrate preferences
(Figures 5 and 6).
Bicarbonate and carbonic acid are not si nificant1y taken up by ion-
exchange resins in binary equilibrium with N0 and HN0 . . The expected
selectively sequence has been verified as sulfate > ni&ate chloride
bicarbonate. Table 1 summarizes the selectivity characteristics of all the
anion resins studied during the course of this research. When using the
selectivities, as measured by the separation factors listed in Table 1, note
that the sulfate/nitrate selectivities are strictly valid only at a total
ionic concentration of 0.005 N (250 ppm as CaC0 ) while the nitrate/chloride
selectivities are independent of total ionic coflcentration.
MULTICOMPONENT EFFLIJENT PROFIL.ES
The curves depicting the effluent concentration vs. bed volumes of
treated effluent for all eleven run; ara tjpified by Figure 9, a two-bed
run, and Figure 10, the single—bed run. Consider Figure 9 (Run 11) as
typical of the general effluent behavior of the four anions of interest and
note that, as predicted from multicomponent chromatography theory, there are
four plateaus each corresponding to a maximum concentration of one of the
anions, and that these plateaus are separated by rather abruEt transition
zones. The first compongnt to appear is always H ,C0 or HC0. followed by
C1 , NO , and finally S0 , the most preferred spe ie . 0bse? ve also that,
as expe ted, all species except for the most preferred species (sulfate)
appear at some time in the effluent in concentrations from 130--350% of
their concentrations in the feed water. An abrupt increase in the concen-
tration of one component in the effluent is always accompanied by a corre-
spondingly abrupt concentration decrease in a second component once the
has been eluted. The single-bed effluent profile (Run 7, Fig. 10)
d’fff rs from the two-bed profile (Figure 9) in that the strong-base anion
bed was presaturated with Cl , the exchanging ion, which was always present
in the effluent; nevertheless, the same general effluent behavior of the
anions prevails.
Upon examination of the chloride and nitrate breakthrough curves from
the two-bed run in Figure 9 it is noted that by not terminating the run on
conductivity breakthrough (Cl breakthrouyh) the capacity of the bed in
terms of bed volumes of effluent treated is increased from 320 to 480 bed
volumes (V ). This is a 50% efficiency increase and is due to elution of
species 1e s preferred than nitrate. In Figures 9 and 10, C ’s indicate the
influent concentrations in meq/l of each of the anions; the omp1ete test
water composition is detailed in Table 2. Nitrate breakthrough has been
defined as 0.48 meq/l of nitrate or 6.7 mg/l of nitrate-N. At that time
the composition of the blended water will have reached the maximum permis-
sible concentration of 10 mg/l.
21
-------
8
Max :5.13
RMBERLITE Iii Lj5
STY-DVB, POLYAtI I NE
TWO—BED SYSTEM. NEUTEIRL
COLUMN DIA. — I INCH (2.5’& CH )
BED DEPTH • 2 1k INCHES (61.0 CM.)
F10N RATE — 2.111k GRL./PIIt1.FT.’ (SIjIlIN.)
+
AVE. SO/NO 3 SEPARATION FACTOR — 12.7
AVE M1/CL SEPARATION FACTOR - 3.89
£ NO
C 0 :1.5
so
8
30000
00 ‘500.00 eoo.oo
BED VOLUMES OF EFFLUENT
700.00
FIGURE 9
11, EFFLUENT CONCENTRATION PROFILE
1000.00
RUN NO.
-------
C 0 1.5
lIVE. SO /NO 3 SEPARATION FACTOR — 1.76
AVE. N(J /CL SEPARATION FACTOR — 2.97
A- NO
w- s
IONAC RFP—100
STY—OVB, QUAT. (1) I9ESIN
SINGLE BED. NEUTRAL ELUTION
COLUMN DIR. — 1 INCH (2.514
BED DEPTH — 214 INCHES (61.0 CM.)
RATE — 2.41* GF&./PUN.FT. 3 (3.1 PUN.)
FLON
S
HC0
C 0 10
100
200
300
BED VOLUMES OF
‘400
EFFLUENT
500
FIGURE 10
7, EFFLUENT CONCENTRATION
600
RUN NO.
PROFILE
-------
MEASURES OF EFFICIENCY
- Maximum possible chemical efficiency (EM) has been defined simply as
y the average equivalent fraction of nitrate on the resin at the end of the
rUn. Since N’ the equivalent fraction of nitrate on the resin, varies with
distance into the bed, the weighted average value (y ) must be used to
represent the ratio of nitrate removed t o all ions r moved. In the ideally
efficient process this would of course approach 1.0 which would only be
possible if nitrate were much preferred over all other anions, and it was
not in these experiments. -
- meq NO on resin at end of run
M — ‘N — Total meg of ions on resin at end of run
meq NO in - meq NO out
EM = Initial meq of all ions on resin + meg of all ions in
-meq of all ions out
Overall chemical efficiency (En) is the product of the maximum possible
chemical efficiency (EM) and the ob erved regeneration efficiency (ER).
E 0 = EMER
ER = Regeneration Efficiency meqtot 1capacityof
meq N0 renioved
E 0 = Overall Chemical Efficiency = meq anion regenerant applied
HOW SELECTIVITY EFFECTS COLUMN PERFORMANCE
Nitrate/Chloride selectivity (cx ) is the most important selectivity
in determining the relative amount o ’ trate on the resin at nitrate break-
through, i.e., in determining the maximum possible chemical efficiency
(EM = N • This is both good and bad: good because all the resins were
ni’erate selective with respect to chloride, bad because little variation
existed in the values of (cx ) among the thirty-two resins tested (cx 1 =
1.85 — 4.33) and no real si ñ icant effects on selectivity seem possi 1 by
further manipulating those variables which were found to be most influential
in the determination of a , i.e., matrix and relative degree of cross-
linking as indicated by Reasons for the importance of N Cl are
discussed below.
Sulfate/nitrate selectivity a 5 is nearly irrelevant in determining
the average equivalent fraction of hTtrate on the resin at the end of a run
(yM). Surprisingly, slight increases in YM are possible as a result of
in’dreasing rather than decreasing the sulf te selectivity-—cx N The
explanation proposed for this is that (1) all the sulfate will be removed
24
-------
from the feedwater regardless of its actual selectivity because it is the
most preferred species and (2) high sulfate selectivity promotes a short
sulfate-rich zone near the column entrance in which almost no nitrate is
removed thereby leaving essentially all of that species to compete with the
lesser preferred chloride in the second equilibrium zone of the column which
is where nearly all of the nitrate is concentrated; see Figures 3 & 4. The
inevitability of sulfate removal in the first enriched zone of the bed and
the major nitrate--chloride competition.in the second enriched zone of the
bed are the reasons why the nitrate/chloride selectivity is the most impor-
tant one to be considered. If either a strong or weak-base anion column is
operated to nitrate breakthrough, that breakthrough will occur when the
second enriched zone, containing primarily nitrate and chloride, reaches the
outlet of the bed. The higher the nitrate/chloride selectivity is (measured
here by aN the higher will be the ratio of nitrate to chloride in the
exhausted ä, and the greater will be N’ the maximum possible chemical
efficiency.
Regardless of the explanation given, the effect of the selectivity of
the most preferred species, sulfate, is predictably slight when the objec-
tive is to remove nitrate, invariably a less-preferred species. That is
graphically demonstrated by the results from Runs 5 & 6 in Figure 11 where
the sulfate/nitrate selectivity varies from 2.83 to 94 with no effect on the
maximum possible chemical efficiency or on the throughput--a normalized
measure of the equivalents of ions fed to the bed per equivalent of bed
capacity.
THE EFFECT OF FEEDWATER NITRATE CONCENTRATION
The most important influence on the average amount of nitrate on the
exhausted resin is, predictably, xN the equivalent fraction of nitrate
in the feed water; when it’s low, proce s efficiency will be correspondingly
low because the exhausted resin will comprise mostly sulfate and chloride--
species not intended 3 to be removed. In these studies the influence of
x at 2.5 gal/mm ft , aN ri = 3.9 and x = 0.3 is shown in Table 3. Rela-
t9’ve efficiency has been 1P1 luded in that table to illustrate that PH is
not simply a multiple of XN.
TABLE 3. EFFECT OF NITRATE CONCENTRATION OF FEEDWATER
Equivalent Fraction
Nitrate Feed Water
of
xN
Average Equivalent Fraction
of Nitrate on Spent Resin
Relative
Efficiency
.20
.27
.34
.40
1.70
1.48
EFFECTS OF BED DEPTH, DETENTION TIME, AND SULFATE CONCENTRATION ON EFFICIENCY
In addition to and xN the interrelated variables--service flow
rate, bed depth and s i ficial detention time (T), are quite significant.
25
-------
a -94, YN°• 41
A. .&
O COLUMN RUN 5
£ = COLUMN RUN 6
I
-J
I
‘a
E
8
END OF RUNS
0.50
1.00 1.50 200 2.50
T = THROUGHPUT EQUIVALENTS SOLUTION / EQUIVALENTS EXCHANGER
FIGURE 11
COLUMN EFFLUENT PROFILES (NITRATE)
OF SULFATE SELECTIVITY ON COLUMN EFFICIENCY
3.00
3.50
EFFECT
-------
Short detention times (t < 3.0 mm.) 3 shallow beds (depth < 60 cm) and high
service flow rates C> 2.5 gal/mm ft ) reduce y by causing relatively more
chloride, apparently the kinetically favored an9’on, to be on the resin at
nitrate breakthrough. Indications are that the detrimental effect is
greater with microporous compared to macroporous resins and with high
capacity compared to low capacity resins. These observations are derived
from the data in Table 4 which compares the nitrate efficiencies (y ’s)
of three different resins in shallow beds (r = 1.5 mm) and deep beds (r =
3.0 mm).
TABLE 4. EFFECT OF BED DEPTH AND SERVICE FLOW RATE
Resin
Resin
Porosity
Capacity
meq/ml
Service
Flow Rate 3
(gal/mm ft
)
De
t
tention
Time
(mm)
Resin
Depth
(cm)
‘N
WGR
Micro
1.5
4.88
1.5
30.5
.36
ES-374
Macro
2.6
4.88
1.5
30.5
.36
ES-374
Macro
2.6
2.44
3.1
61.0
.41
ES-368
Macro
1.4
4.88
1.5
30.5
.39
ES-368
Macro
1.4
2.44
3.1
61.0
.41
Although x , the equivalent fraction of sulfate in the feed water) was
not a variable ‘In the column experiments, it will greatly influence y
because all the sulfate fed to the column will be on it at nitrate break-
through. When xS is high, the efficiency, N’ will be low.
THE EFFECT OF REGENERATION STOICHIOMETRY ON EFFICIENCY
Because we are advocating the operation of the anion bed to a point
considerably beyond exhaustion, the cation bed must not contain exchangeable
hydrogen ions after the anion bed is exhausted. If it contains these excess
hydrogen ions, they will be added to the cation bed effluent water but will
not be removed or neutralized in the anion bed. On the other hand, if the
cation bed doesn’t contain a sufficient number of exchangeable hydrogen
ions, it will breakthrough first and the potential capacity of the anion
bed will not be realized to the detriment of process efficiency. For these
reasons the capacities of the beds must be matched as closely as possible.
In actual practice this will prove to be somewhat difficult because of the
variable nature of the ion—exchange capacities of the beds. The ion-
exchange capacity of the cation bed is a function of the cationic composition
of the feedwater and the regeneration level, the latter being a measure of
the quantity of regenerant applied compared to the stoichiometric amount
theoretically required. The weak-base anion bed capacity is predominantly
a function of the anionic composition of tha feedwater and is not much
influenced by the regeneration level because of the resin’s great affinity
for the regenerant hydroxide ions, i.e., its strong preference for the
free-base form.
27
-------
Regeneration level influences both the overall chemical efficiency
(En) and the maximum possible chemical efficiency (EM or . M). It has been
d&fermined here that a downf low regeneration level o’f 240-300% of the
theoretical HC1 required should be applied to the cation bed if calcium and
magnesium are the primary- cations on the resin; see simulated ground water
composition in Table 2. Regeneration levels lower than 300% caused premature
cation breakthrough, increasing pH, and reduced anion bed capacity with
smaller values of y,. at breakthrough; see Runs 9 and 10, Table 5. In order
to study the elutio i behavior of the anions under true ion-exchange conditions
in the absence of a significant concentration of hydroxide ions, the capacity
of the cation bed used in Runs 1-6 was deliberately made much larger than
the capacity of the weak-base anion bed. By observing the minimum and
final pH’s of these runs in Table 5, the acidic nature of the system effluent
during elution to nitrate breakthrough is evident for the case of cation
bed capacity greater than anion bed capacity. In actual practice, the beds
will have to be sized based on the resins chosen and the ground water
composition. The efficiency of the regeneration operation should remain
constant as long as the composition of the feedwater doesn’t change drastically
and the resins do not become fouled with organics, iron, or silica.
For the single-bed strong-base anion process regenerated with NaCl it
is expected that a regeneration level of 300% or greater will be required
for efficient regeneration. This is based on published rather than experi-
mentally determined information [ 1].
The overall chemical efficiency (En) for waters similar to the test
water in Table 2 can be expected to be about 13.3% for both the single-bed
and two-bed processes. This estimate is based on operation to nitrate break-
through with the observed average equivalent fraction of nitrate on the
resin at the end of the runs (y = 0.40) utilizing a feedwater containing the
same equivalent concentrations df nitrate, chloride and -sulfate with an
irrelevant amount of bicarbonate which undergoes no net removal in either
process. Higher equivalent fractions (xM’s) of nitrate in the feedwater will
increase E 0 and lower equivalent fractioHs will reduce it.
COST COMPARISONS FOR THE SINGLE-BED Al 1D NO-BED PROCESSES
A comparative economic evaluation of the processes reveals that the
two-bed process with HC1 and NH 0H as regenerants has chemical plus disposal
costs which are approximately 50% higher than the single-bed process if the
following assumptions are made: an overall chemical efficiency of 13.3%,
25% bypass water, a feedwater with the composition of the test water in
Table 2, NaCl-NaNO 3 brine disposal by trucking 8 miles before discharging
into a stream, and no disposal costs for the high-nitrogen content waste-
waters from the two-bed process which are given away for their fertilizer
value. Considering only the chemical costs for regenerants, the two-bed
process costs three times as much to operate as the single-bed process but
yields nitrate free, partially softened water and a land disposable regenerant
with fertilizer value. See Table 6 for cost details and chemical composi-
tions of regenerants, and Table 7 below for the comparative costs of possible
regenerants to remove 1 lb. equivalent of nitrate, i.e., 14 lbs. of N or
62 lbs. of NO 3 ion.
28
-------
TABLE 5: COLUMN PERFORMANCE CHARACTERISTICS
**
Run
No.
*
Flow
gal
min.ft
Minimum
pH
Final
ph
Bed
Depth
cm
Resin Description
(Cation Regeneration Level)
S
N
N
aCl
i
HCO
3
N
Final
Column
Capacity
meq/ml
t
ye
BV
1
2.34
2.5
63.5
Duolite ES-368
STY-DVB, Tert-Amine, MR
2.83
3.87
.13
.53
.00
.34
1.65
582
2
4.88
2.5
2.5
30.5
Duolite ES-374
Polyacrylic, Polyamine, MR
94.
3.85
.26
.36
.02
.36
2.93
720
3
4.88
2.4
T4
30.5
Duolite ES-368
STY-DVB, Tert. Amine, MR
2.83
3.87
.20
.40
.01
.39
1.36
364
4
4.88
2.5
2.5
30.5
Dowex WGR
Epoxy-Amine, Polyamine, Gel
137.
1.99
.27
.37
.00
.36
1.62
391
5
2.44
2.4
2.4
61.0
Duolite ES-368
STY-DVB. Tert. Amine, MR
2.83
3.87
.16
.43
.00
.41
1.48
423
6
2.44
2.3
61.0
Ajj1Ite [ .3/4
Polyacrylic, Polyamine, MR
94.
3.85
.15
.44
.00
.41
3.12
920
7
2.44
6.1
T 4
61.0
lonac AFP-100
STY-DVB, Quat.(I)Amine, MR
1.76
2.97
.14
.43
.01
.42
1.03
295
8
2.88
2.8
5.8
61.0
Duolite ES-368 (600%)
STY-DVB, Tert. Amine, MR
2.83
3.87
.21
.40
.00
.39
1.39
375
9
2.44
4.5
6.7
61.0
Duolite ES-368 (120%)
STY-DVB, Tert. Amine, MR
2.83
3.87
.31
.34
.02
.33
0.84
190
10
2.44
4.6
6.3
61.0
Duolite ES-368 (240%)
STY-DVB, Tert. Amine, MR
2.83
3.87
.14
.44
.00
.42
1.15
334
11
2.44
4.7
5.5
61.0
Amberlite IR-45 (300%)
STY-DVB, Polyamine, Gel
12.7
3.89
.08
.45
.03
.44
1.61
480
x 7.48 =
Superficial
detention time, r, minutes
1
gal/min.ft
t Ve = Bed volumes of effluent to 0.5 meq/1 N0 3 -breakthrough (end of run)
** Final Column Capacity is greater than measured HC1 capacity because resin has higher capacity for sulfate which occupies
a significant fraction of the available sites at the end of the run.
Final pH refers to the pH of the system effluent at nitrate breakthrough.
Minimum pH was the minimum pH observed during the course of the run.
-------
Total Dissolved Solids, ppm .
Undesirable Cation (Nat), ppm
Calcium ion, ppm
Magnesium ion, ppm
Ammonium ion, ppm . . 0
Sulfate ion, ppm
Nitrate ion, ppm
Chloride ion, ppm
Nitrogen, ppm
Nitrogen Fertilizer Produced, lb N/100C gal
Nitrogen Fertilizer Produced, kg/rn 3 H 2 0 Supplied .
Per capita Fertilizer Production, lb F1/capita .year
Per capita Fertilizer Production, kg N/capita .year
TABLE 6: ECONOMIC AND
AND TWO-BED
REGENERANT WASTEWATER
PROCESSES
COMPARISONS BETWEEN THE
SINGLE-BED
Item
Single-Bed
Process
Two-Bed
Process
Regenera nt
Regenerant
Regenera nt
Regenerant
Regenera nt
Regenerant
Regenerant
Regenera nt
Chemical Costs, /l000 gal H 0 Supplied. . . . 9.2
Chemical Costs, /m H 2 0 Supplied 2.43
Disposal Costs, /l0OO gal H 2 0 Supplied . . . 9.2
Disposal Costs, /m 3 H 2 0 Supplied 2.43
plus Disposal Costs, /lOOO gal H 0 18.4
plus Disposal Costs, /m H 2 0 Supplied . . . . 4.86
Volume, % Total Water Supplied 0.78
Composition: Total Concentration, N 1.08
69,200
24,900
0 2,480
0
27.8
7 .34
Nil
Nil
27.8
7 .34
1 .73
0.650
37,400
0
7,870
8,960
27,500
2,020
H 2 0 Supplied 0
465
8,780
3,540
4,030
18,100
7,740
1.12
o .134
0 40.9
0 18.6
30
-------
TABLE 7: CALCULATED CHEMICAL REGENERANT COSTS
Regenerant
Chemical
$
3
m
.supplied**
lb_equivalent*
1000 gal supplied**
H 2 S0 4
NaC1
1.23
1.30
8.63
9.15
2.28
2.42
NH 3
HC1
1.53
2.43
10.7
17.1
2.78
4.52
NH 4 C 1
HNO 3
5.64
6.62
39.6
46.5
10.5
12.3
* 1 lb-equivalent 14 lbs of nitrogen or 62 lbs of nitrate
**Assumes regeneration level of 300%, 25% bypass and N = 0.40
Note that the cost estimates in Table 6 include only chemical and dis-
posal costs. They do not include capital costs for plant and equipment,
depreciation, labor, interest, or pumping costs. Clearly these excluded
costs are a significant portion of the total cost of supplying potable water
of acceptable nitrate concentration but these items were not the objectives
of this study and consequently no detailed calculations of them were made.
Nevertheless, one can obtain conservative “ball-park” estimates of the
total costs by doubling the 1970 amortization, power, labor, and maintenance
costs given by Holzmacher [ 13) and adding that figure to the chemical plus
disposal costs calculated here. This results in very rough estimates of
3fl/1000 gal for the single—bed process with regenerant disposal, and
55 /10O0 gal for the two-bed process with the regenerants given away, and
further estimating that amortization, power, labor, and maintenance are
150% of the single-bed process. Holzmacher’s system included 4 MGD of
installed, continuous ion-exchange capacity operated 40% of the time. Again,
as a rough estimate, we might conclude that treating a water containing
approximately 400 mg/l TDS and 21 mg/I nitrate nitrogen by ion exchange is
going to add between 37 /lOO0 gal and 55 /1O0O gal to the cost of the
water delivered to the distribution system if the cost of regenerant disposal
is considered. If we neglect this disposal cost, the single-bed process can
probably be operated in 1977 at a total added cost of 28 t/lO00 gal of water
supplied. These “ball-park” estimates are also subject to the assumptions of
25% bypass water, downflow regeneration with 300% of the theoretical require-
ment, and an exhausted resin which is 40% in the nitrate form at nitrate
breakthrough.
31
-------
SECTION 7
DESIGN CONSIDERATIONS FOR WATER SUPPLY APPLICATIONS
THE USE OF NITRIC ACID
Nitric acid is definitely not recomended as a regenerant in the two-bed
process even though it would greatly enhance the fertilizer value of the
regenerant wastew ters. Nitric Acid is too costly, 46.5 /lOOO gal water
supplied (l2.3 /m ), requires excess cation bed rinsing to reduce the residual
nitrate concentration, and allows the possibility of disastrous nitrate and
acid pollution of the water supply in the event of an operating error. Even
though HC1 is more costly than H ,SO it may be more economical where large
excesses of H SOA are required dUe to CaSOA fouling of the cation bed. Where
H SO can be t1se , however, it should be u ed because of its lower cost.
R fe to Table 7 for a cost comparison of regenerants for both cation and
anion beds.
ORGANIC EXTRACTABLES LEACHED FROM THE RESINS
Even after conditioning by extensive backwashing followed by two com-
plete exhaustion-regeneration cycles including the usual backwashes and
rinses, all the anion resins tested continued to bleed measurable amounts
of organics into the treated water. The organic extractables leached from
these conditioned resins gave rise to total organic carbon (TOC) concentra-
tions in the 3--90 ppm range in acidic aqueous solutions containing about
1.0 weight % resin tumbled at 13 rpm for 16—-20 hours; see Table 8 below.
TABLE 8: ORGANICS LEACHED BY AGITATING CONDITIONED ANION RESINS
PPM
Resin
Description
TOC
3
STY-DVB, Weak Base
35
2
Acrylic-Amine, Weak Base
5
6
Phenol-HCHO, Weak Base
26
11
Aliphatic-Amine, Weak Base
90
7
Epoxy-Amine, Weak Base
25
16
STY-DVB, Strong Base
6
1% Resin Solutions Tumbled 16 Hours at 25°C
32
-------
The leached organics are primarily non-volatile at 100°C. This is indicated
in Table 9 which is a listing of the volatile (VOC) and non-volatile carbon
(NVOC) analyses of the supernatants from conditioned resins left standing
approximately six months in distilled water at room temperature. Some of
these organic carbon concentrations are alarmingly high especially in light
of the recent interest in low levels of organics in water supplies as
precursors to carcinogens formed upon reaction with chlorine. This problem
definitely merits further investigation.
TABLE 9: ORGANICS LEACHED FROM CONDITIONED IX RESINS UPON STANDING
Resin
Description
PPM
VOC
PPM
NVOC
3
STY-DVB, Weak Base
lC
182
6
Phenol-HCHO, Weak Base
12
137
li
Aliphatic-Amine, Weak
Base
3
3870
7
Epoxy-Amine, Weak Base
47
520
16
STY-DVB, Strong Base
3
61
C—I
STY-DVB, Strong Acid
3
428
50% Resin Solutions Standing 6 Mos. at 25°C
A DEGASIFIER TO REMOVE CO 2
In the two-bed system, the first species to be eluted is carbonic acid
(H CO 3 ) at which time the pH of the treated water drops to approximately
4. . This acid gas (COo) must of course be removed before permitting the
treated water to enter the distribution system. Usually, in such a two-bed
system, a degasifier would be installed between the cation and anion beds
to remove CO , under the very acidic conaitions produced by the mineral
acids present at that point. This may rot be good practice in systems where
it is desired to operate to nitrate breakthrough. A better location may
well be following rather than preceding the weak-base anion bed. Some
beneficial kinetic effect due to the presence of H 2 C0 in column experiments
was observed during the course of the eperimental wo ’k and reported elsewhere
(I. Abrams, Diamond Shamrock Chemical Co., Personal Communication). Apparently,
in column operation, H ,CO 3 is neutralized by the weak-base resin, whereupon
the HCO 3 anions are taken up thereby sw. 11ing the resin beads in the lower
reaches of the bed where they compete with no other anion; finally the
swollen, bicarbonate-form resin takes up_the next most preferred species
(chloride) by rapidly exchanging the HCO 3 for it. Simply stated, the
bicarbonate anion is a catalyst for the removal of the more preferred
species in ion-exchange column operations. Thus, it is questionable whether
removing CO from waters before anion-exchange in nitrate removal is good
or bad desi n since the closer the approach to equilibrium, the more
chemically efficient is the operation of this process.
33
-------
RANKING THE STRONG- AND WEAK-BASE ANION RESINS
The recomended anion resins for nitrate removal service are listed in
Table 10 considering that high nitrate/chloride selectivity high capacity,
and moderate sulfate/nitrate selectivity are desirable characteristics.
Organic extractables as evidenced by the TOC of resin equilibrates were not
considered in making the rankings because of the very preliminary nature
of those measurements. However, an asterisk (*) has been used to indicate
a resin producing markedly colored water in addition to high TOC.
Although the resins are ranked in preference order, the differences
among the recomended resins are not large; they are all expected to give
nearly the same maximum possible chemical efficiency EM. Some overall pro-
cess efficiency is gained by using high capacity resins while some might
be lost with the highly sulfate selective resins should they require large
rinse volumes. Note that the rankings are not endorsements by the authors
or the EPA of any of the resins listed.
34
-------
TABLE 10: RANKING OF RESINS FOR USE IN NITRATE REMOVAL SERVICES
Recomended
STY—DVB, Polyamine, Resins
Amberlite IR-45
STY-DVB, Tertiary-amine, MR Resins
Amberlite IRA-93
Dowex MWA-1
Tonac AFP-329
Duolite ES-368
STY-DVB, Quat. (I & II) Amines, Gel,MR and Improved Porosity Resins
lonac ASB-1, AFP-l00, A-641, ASB-1P, ASB-2
Duolite, A-l0l-D, A-102-D
Dowex 11, SAR, SBR-P, SBR
Amberlite IRA-400, IRA-900, IRA-402, IRA-910, IRA-410
lonac A-550, A-540 (No DVB crosslinking)
Duolite A-104 (mixed Types I and II Amines)
Acrylic—Amine, Polyamine, MR Resins
Duolite ES-374
Phenol-HCHO, Polyamine, MR Resins
Duolite A-7
Duolite ES-561
Not Reconiiiended
Epoxy-amine, Polyamine, Gel Resins
Dowes WGR
Duolite A-340
lonac A-305
Acrylic-Amine, Tertiary Amine, Gel Resins
Amberlite IR.A-68
Aliphatic-Amine Polyamine, Gel Resins
*Ionac A-260
35
-------
REFERENCES
1. Beulow, R. W., K. L. Kropp, J. Withered, and J. M. Symons, “Nitrate
Removal by Anion Exchange Resins,” Water Supply Research Laboratory,
National Environmental Research Center, U.S. EPA, Cincinnati, Ohio,
May, 1974.
2. Bi gham, E. C., “Fertilizer Maker Stops Nitroqen,” Water and Wastes
Engineering , P.E.-4, November, 1972.
3. Boari, G., L. Liberti, C. Merli, and R. Passino, :Exchange Equilibria on
Anion Resins,” Desalination , V. 15, p. 145-166 (1974).
4. Clifford, Dennis A., “Nitrate Removal from Water Supplies by Ion Exchange:
Resin Selectivity and Multicomponent Chromatographic Column Behavior of
Sulfate, Nitrate, Chloride, and Bicarbonate,” Ph.D. Thesis, University
of Michigan, University Microfilms, Ann Arbor, Michigan, Pub. No. 77-
7893 (1976).
5. Diamond Shamrock Chemical Co., Duolite Ion-Exchange Manual , Redwood
City, CA (1969).
6. Dorfner, K., Ion Exhchangers: Properties and Applications , 3rd. Ed.,
Ann Arbor Science, Ann Arbor, Mich. (1972).
7. Evans, S., “Nitrate Removal by Ion-Exchange,” J. WPCF , V. 45, No. 4,
pp. 632-36, April, 1973.
8. Gauntlett, R. B., “Nitrate Remo a from Water by Ion—Exchange,” Water
Treatment and Examination, V. 24, p. 172 (1975).
9. Gregg, J. C., “Nitrate Removed at Water Treatment Plant,” Civil
Engineering - ASCE , p. 45, April, 1973.
10. Heifferich, Friedrich, Ion-Exchange , McGraw-Hill Book Co. Inc., New York,
N.Y. (1962).
11. Heifferich, F. G., “Multicomponent Ion Exchange in Fixed Beds,” I & EC
Fund., V. 6, No. 3, p. 362 (1967).
12. Helfferjch, F. and G. Klein, Multicomponent Chromatography: Theory of
Interference , Marcel Dekker, New York (1970).
13. Holzmacher, R. G., “Nitrate Removal from a Ground Water Supply,” Water
and Sewage Works , p. 210, July, 1971.
36
-------
14. Klein, G., D. Tondeur, 1. Vermeulen, “Multicomponent Ion Exchange in Fixed
Beds,” I & EC Fund., Vol. 6, No. 3, p. 339 (1967).
15. Kunin, R., Ion Exchange Resins , 2nd Ed., John Wiley & Sons Inc., New York
(1958).
16. Midkiff, W. S. and Weber, W. J., Jr.,’ “Operating Characteristics of
Strong-Base Anion Exchange Reactors,” Proceedings of the 25th
Purdue Industrial Waste Conference , May, 1970.
17. National Academy of Sciences, “Sutmiary Report: Drinking Water and
Health,” National Academy of Sciences, Washington, D. C., 1977.
18. Tondeur, D. and G. Klein, “Multicomponent Ion-Exchange in Fixed Beds,”
I & EC Fund., V. 6, No. 3, p. 351, August, 1967.
19. U. S. EPA, “Interim Primary Drinking Water Standards,” Federal Register ,
December 24, 1975.
37
-------
GLOSSARY
“as CaC0 ”: Normality (N) can be converted to calcium carbonate equivalents.
TheI’e are 50 mg of CaC0 per millequivalent. Any 0.005 N solution
contains 5 milliequival nts/l or the equivalent of 250 mg/l of CaCO 3 .
bed: The ion—exchange resin contained in a column. Water to be treated by
ion—exchange is passed downward through the column.
breakthrough: The appearance of a sharp increase in the concentration of an
ion in the effluent from the bed.
capacity: The total number of ion-exchange sites available per unit volume
of resin measured in equivaIents/l or milliequivalents/mi. Resins were
equilibrated with 0.005 N acids (HC1, H ,S0A or HNO. ) for the experi-
mental capacity determinations. This w s done to gimulate the expected
capacities in typical groundwater applications.
chromatographic elution: Continued application of the feed water to an
exhausted ion-exchange bed so as to “elute” or sequentially drive off
those less-preferred feed water anions previously removed during the
exhaustion cycle. In this operation, the ions being driven off the
resin are separated into zones in which the aqueous concentration of the
primary ion in a given zone exceeds the concentration of that ion in the
feed water.
downflow regeneration: Cocurrent regeneration, i.e., the regenerant solution
is passed down through the bed in the same direction as the feed water
was passed through the bed.
effluent profile: A plot of the effluent concentration of an ion or ions vs.
the volume of effluent water from the bed.
elution: The displacement of non-preferred ions previously removed from the
feed water by continued application of the feed water or an “eluting
solution” containing an ion or ions more preferred by the ion exchanger.
equivalent: One gm equivalent (6.023 x ]Q23) of ionic charges in the aqueous
phase or that number of fixed charges in the resin phase.
equivalent fraction: That fraction of the total negative or positive charges
present which is due to a given ion. If x( = 0.27, then 27% of the
negative ionic charges in a given volume of water are due to sulfate
ions.
38
-------
exhaustion: The step in an ion-exchange cycle in which the undesirable ions
are removed from the water being treated. The resin bed is said to be
“exhausted” when the ions originally on the resin have been essentially
completely exchanged for feed water ions.
functionality: A description of the nature of the amine groups attached to
the resin matrix which give an anion resin its ion exchange properties,
e.g., quaternary amine functionality’.
ion-exchange: A physicochemical process in which ions in the water being
treated replace and are exchanged for ions in a solid phase (the resin).
In the single—bed process, nitrate, the pollutant ion, is placed on the
resin phase in exchange for an innocuous ion such as chloride.
isoporous resins: Resins having slightly greater uniform porosity than
typical microporous resins.
isotherm: A constant temperature plot of resin phase concentration of an ion
vs. the water phase concentration of that ion. In a binary isotherm,
e.g., sulfate/nitrate, the resin phase exchange sites not occupied by
sulfate are occupied by nitrate. Similarly, the significant anions in
the water which are not sulfate are nitrate.
macroporous resins (also referred to as macroreticular resins): Very porous
resins whose beads comprise aggregates of gel resins with large internal
voids having diameters up to several hundred angstroms (A). This
porosity is non—uniform with areas of very high crosslinking. Macro-
porous resin beads are opaque.
matrix: The polymer backbone of a synthetic organic ion-exchange resin.
microporous resins (also referred to as gel resins): resins wjth porosity of
atomic dimensions, i.e., having “pores” which are 10-20 A in diameter.
Gel resins are relatively uniform in porosity and the beads are trans-
parent.
milliequivalent: (Abbreviated meq.) 1/1000 of an equivalent. An 0.005 N
solution contains 0.005 equivalents/i or 5 meq/1.
porosity: A measure of the degree of openness of the polymer matrix which is
related to the nature and degree of crosslinking.
regeneration: The displacement from the exhausted ion-exchange resin of the
undesirable ions removed from the water during the exhaustion cycle.
Performed by passing through the bed, a relatively concentrated (1 N)
solution of the ion desired on the resin.
regeneration level: A measure of the inefficiency of regeneration expressed
here in %. The level indicates the amount of regenerant which must
actually be applied compared to the amount theoretically required. For
downf low regeneration a level of 300% is typically required; that means
a 200% excess of regenerant must be applied.
39
-------
selectivity: A measure of the relative affinity for one ion over another
exhibited by the resin. In this report selectivity (relative affinity)
is measured by the separation factor, a. This a should not be confused
with the selectivity coefficient , K.
selectivity sequence: A listing of ions as preferred by the ion exchanger
ordered from most preferred to least preferred.
separation factor (binary): The ratio of the distribution of ions between
the water phase and the resin phase. a N is the ratio of the dis-
tribution of sulfate ions between phase ”fo the distribution of nitrate
ions between phases. If > 1, the resin prefers sulfate over nitrate.
service flow rate: The rate of application of feedwater to the resin bed.
Because the exchanger capacity is relat d to the volume of resin, the
rate is usually specified as gal/mm ft or volume of feed water per
volume of resin per unit time. With proper units this is reciprocal
superficial detention time. Recommended exhaustion rates are 1-5 gall
mm ft corresponding to detention times of from 7.48 to 1.50 minutes.
softening: In ion exchange, a process by which polyvalent cations, e.g.,
calcium, magnesium, and iron are exchanged for a monovalent cation such
as hydrogen or sodium.
spent regenerant: A wastewater containing the excess regenerant ions and the
undesirable ions removed from the exhausted resin. Its volume will be
determined by the volume of rinses included as “spent regenerants.”
strong-base resin: An anion exchange resin containing fixed positively
charged quaternary amine functional groups which prefer all common
anions over hydroxide ions. Simply, a resin which tends to readily give
up hydroxide ions in exchange for nearly any other anion. The capacity
of strong-base resins to exchange ions does not depend on the presence
of excess hydrogen ions (acidity) to form the positively charged
exchange sites as is the case with weak-base resins. Thus, they may be
used as ion exchangers in acid, ;‘eutral, and basic solutions.
superficial detention time (T): The tune a particle of feed water spends in
the empty resin bed assuming plug flow. It is calculated as the empty
bed volume divided by the feed flow rate.
upflow regeneration: Countercurrent regeneration, i.e., the regenerant
solution is passed up through the bed in a direction opposite to that
taken by the feedwater. Countercurrent regeneration is reportedly more
efficient than cocurrent regeneration because the most preferred ions
are not driven through the entire bed.
weak-base resir : An ion-exchange resin comprising primary, secondary, or
tertiary amine functional groups or a mixture of those groups which
acquire positive charges when excess hydrogen ions (acidity) are present.
These charged sites can exchange anions if the feed solution remains
acidic. Thus, these resins are said tc “adsorb” acids. In neutral to
40
-------
basic solutions they have no charged sites and consequently no signifi-
cant anion exchange capacity. They are readily regenerated with weak
bases or even neutral water solutions.
41
-------
TECHNICAL REPORT DATA
(Please read Insnvcnons on the reverse before corn plenng)
1 REPORT NO.
EPA—600/8—7 7—015 12
3. RECIPIENTS ACCESSION NO.
4. TITLE AND SUBTITLE
NITRATE REMOVAL FROM WATER SUPPLIES BY ION EXCHANGE
Executive Summary
5 REPORT DATE
November 1977 (Issuing Date)
6.PERFORMINGORGANIZATIONC ODE
7. AUTHOR(S)
Dennis A. Clifford*
Walter J. Weber, Jr.
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
University of Michigan
Ann Arbor, Michigan 48105
10. PROGRAM ELEMENT NO.
1CC614
11.CONTRACT/GRANTNO.
Grant No. R—803898
12. SPONSORING AGENCY NAME AND ADDRESS
Hunicipal Environmental Research Laboratory——Cm., OH
Office of Research and Development
US Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Fin 1 — Executive Stimniary
14 SPONSORING AGENCY CODE
EPA/600/14
15 SUPPLEMENTARY NOTES
*Presently with University of Houston, Houston, Texas 77004
Project Officer: Thomas J. Sorg (513—684—7228)
16 ABSTRACT
Anion exchange using synthetic organic resins is a proven and practical
technology for the remâval of nitrate from water supplies. However, disposal of the
spent regenerarit brime solution containing nitrate is a potential problem. Two
processes were examined in detail in this report——single—bed strong—base anion
exchange with NaC1 regeneration and two—bed strong—acid, weak—base ion exchange with
HC1 and NH 4 OH regeneration. Both systems must be operated to nitrate breakthrough to
minimize regeneration costs. The two—bed process Is one and one—half to two times
as expensive to build and operate as is the single—bed process, but produces
softened low—TDS, low—nitrate water, and has a readily disposable, spent regenerant
with fertilizer value. Important design considerations were found to include the
nitrate and sulfate concentrations in the raw water, the service flow rate, the resin
bed depth, and the nitrate/chloride selectivity of the resin. The sulfate, nitrate,
chloride, and bicarbonate selectivities and inulticoniponent column behavior of the
anion resins available from U.S. manufacturers were examined and are reported In
detail. An important peripheral finding was that significant quantities of non-
volatile organics were leached from “clean” resins into the treated water.
17. KEY WORDS AND DOCUMENT ANALYSIS
a DESCRIPTORS
b IDENTIFIERS/OPEN ENDED TERMS
C. COSAr FleldlGroup
Water treatment——Ion exchanging, Water sup-
ply, Ion exchanging, Ion exchange resins,
Demineralizing, Nitrate deposits——inorganic
nitrates, Sulfates, Chlorides, Cost esti-
mates, Experimental data
Nitrate removal, Ion ex—
change——two—bed process
13B
18. DISTRIBUTION STATEMENT
.
Release to Public
19 SECURITY CLASS (This Report)
Unclassified
21 NO OF PAGES
52
20 SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Fo,m 2220—1 (R.v. 4—77)
42 * U S GOVERNMENT PRINTING OFFICE 1978-757-140/6632 Region lb 5-Il
------- |