(DWRD0014)
NITRATE REMOVAL FROM DRINKING WATER
by
Thomas J. Sorg
A Paper Presented
at
EPA Seminar on Nitrates in Groundwater
October 3-4, 1979
Kansas City, Missouri
DRINKING WATER RESEARCH DIVISION
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
U.S. ENVIRONMENTAL PROTECTION AGENCY
26 W. ST. CLAIR STREET
CINCINNATI, OHIO 45268
MARCH 1980

-------
NITRATE REMOVAL FROM DRINKING WATER
INTRODUCTION
The two principal sources of nitrate contamination of water sources
are fertilizers and septic tank wastes. Thus, the most vulnerable water
supplies to nitrate contamination are ground waters in agricultural areas
and in areas having large numbers of septic tanks.
ALTERNATIVES TO TREATMENT
Solutions to the nitrate contamination problem in ground water are
several with treatment being just one of the potential remedies.
Good engineering practice requires that all alternatives to treatment be
explored for their feasibility and cost effectiveness in comparison to
treatment. Because most small communities have limited resourses, treat-
ment will very likely be selected as a last resort approach to the problem.
Three alternatives to treatment are:
(1)	development of another water supply;
(2)	blending of two or more water supplies; and
(3)	connecting to approved water supply.
Each alternative has advantages and disadvantages depending on the
conditions and circumstances in each community (Fig. 1).
Obviously, the most simple approach is the development of a new water
supply, such as drilling a new well. This solution assumes that a nitrate
free water source exists in the area. With advanced drilling techniques,
water can be drawn from specific strata or the contaminated source can be
sealed off. The major disadvantage is the potential for the water quality
to change with pumping and time.

-------
2
FIGURE 1. ALTERNATIVES TO TREATMENT
ALTERNATIVE
ADVANTAGES
DISADVANTAGES
Development of a new
water supply.
Less expensive
Short time implementation.
Modifications to distri-
bution system.
Water quality may change.
Blending of two or
more water supplies.
Connect to an approved
water supply.
Less expensive.
Short time implementation.
Le s s expe ns ive.
Short time implementation.
Few modifications.
Extensive modifications
may be required for
blending.
No control over water
supply.
Dependent on another
utility.

-------
3
Another approach is to blend water from different sources to achieve
a nitrate concentration in the blended water below the Maximum Contaminant
Level (MCL) of 10 mg/L (as N). This solution assumes that a good source
of water exists for blending with the high nitrate water. Modifications
to the distribution system will be required to provide mixing; modifica-
tions that could be very expensive.
A third solution is to obtain water from a neighboring community.
This solution results in the community giving up control over its
drinking water and becoming dependent on another community or utility.
Some communities may not be willing to relinquish this control and thus
would be willing to pay the higher cost of developing a new supply or
providing treatment. Although each alternative seems to be a straight-
forward simple approach to the problem, each solution has certain dis-
advantages from either an economic, engineering, or political viewpoint.
In the final analysis, each solution may be more expensive than treatment
or unacceptable.
TREATMENT METHODS
Several treatment techniques have been studied for the removal of
nitrate from drinking water: (1) chemical reduction*-^ (2) biological
denitrification;^-^ (3) ion exchange;^-® and (4) reverse osmosis^"-'-®
(Fig. 2). Of these methods only the latter two, ion exchange and reverse
osmosis, are considered feasible and practical for full-scale treatment
of drinking water. Electrodialysis is very similar to reverse osmosis
and is also effective for nitrate removal.^ Neither conventional
coagulation nor lime softening are effective removal methods because
nitrate is very soluble in water.

-------
4
FIGURE 2. TREATMENT METHODS FOR NITRATE REMOVAL
TREATMENT METHOD
Chemical coagulation
Lime Softening
Chemical reduction
Biological denitrification
Ion exchange
Reverse osmosis
Electrodialys is
EFFECTIVENESS
Not effective
Not effective
Has potential, but not practical
Has potential, but not accepted
Effective
Effective, but costly
Effective, but costly

-------
5
Several chemicals have been Investigated for reducing nitrate to
nitrogen gas. Of the various chemicals studied, only ferrous iron has
been determined to be economically attractive. The process requires a
catalyst, copper, for denitrification however and the process must
take place in an alkaline solution. Other drawbacks to this process
are that only about 70 percent of the nitrate is reduced and large
amounts of ferrous iron are required, eight moles of iron per mole of
nitrate.
Biological denitrification has been studied for the purification
of wastewater and drinking water. The process requires the use of deni-
trifying organisms in a filter bed to reduce the nitrate to nitrogen
gas. The process also requires an organic energy source for the
bacteria, such as methanol or ethanol, to be added to the water because
most groundwaters are low in organic content. In general, the water
utility.industry has not accepted this process because of several reasons:
(1) the need to add an organic material to water that is generally free
of organics; (2) the process requires careful control; (3) the need to
develop a large bacterial population in water that is generally free of
organisms; and (4) the system will be out of service if the biological
mass is lost. As long as other alternatives exist for nitrate removal,
this process will probably be given very little serious consideration
for drinking water treatment, particularly in small communities.
Currently, ion exchange treatment is the only treatment method being
used on a full scale level to remove nitrate from drinking water. In
1974, a full scale plant was constructed by the Garden City Park Water
District, Nassua County, New York to treat ground water containing 20-30
mg/L of nitrate-nitrogen. The plant was designed to treat 0.08 m^/s
(1200 gpm) and to lower the nitrate-nitrogen level to less than 2 mg/L.

-------
6
The Garden City Park facility is a different ion-exchange process that
employs a continuously regenerated ion-exchange process in a closed loop.
This process is used in industrial applications but uncommon in drinking
water treatment. The major advantage of this system is that the resin material
is continuously regenerated; large slugs of brine solution for disposal are thus
avoided. Greater efficiency of the media for nitrate capacity is also projected.
As of September, 1979, the system was operating successfully, although
not on a full time basis. Because most small communities cannot afford
such an elaborate system, the more common fixed bed system would have
greater application. The Garden City Park system, however, has demon-
strated that ion exchange technology will work for nitrate removal. The
economics, removal efficiency, chemical usage, and so forth will vary with
the quality of water and, therefore, the operation of the Garden City
Park plant cannot be universally projected to other locations. For
example, the most significant factor of the water is the sulfate concentra-
tion because up to about a total dissolved solid (TDS) concentration of
3000 mg/L, sulfate is preferred over nitrate. High sulfate water, therefore,
is more costly to treat than low sulfate water because the sulfate competes
with nitrate and thus lowers the nitrate removal efficiency.
The Garden City Park ground water is low in sulfate (30-40 mg/L) and
therefore the ion exchange system is relatively efficient for nitrate removal.
To evalute the influence of higher sulfate water, the Drinking Water Research
Division, USEPA funded a research project in McFarland, California that has a
ground water containing near 20 mg/L of nitrate-nitrogen and about 300 mg/L
of sulfate. The project is investigating the effectiveness and efficiency
of ion exchange and reverse osmosis treatment. The project is about 50
percent completed (Jan. 1980) and is scheduled for completion by
December, 1980.

-------
7
Two other treatment methods, reverse osmosis and electrodialysis,
will remove nitrate, but these methods are used primarily for treating
high TDS and salt water. The nitrate removal efficiency by reverse
osmosis varies with the type of membrane; a wide rejection range of
60-95 present is reported.Neither method would probably be selected
solely for nitrate removal because of their high cost. If the water has
several other contaminants or is high in TDS, the techniques may be
practical and economical, however.
EQUIPMENT AVAILABILITY
Ion exchange, reverse osmosis, and electrodialysis equipment are
readily available from a variety of equipment manufacturers. Because
the efficiency of ion exchange treatment for nitrate removal is dependent
on the water quality, pilot-plant studies on the specific water are
recommended for developing design and operational data. Once these data
are established, most water treatment equipment manufacturers can easily
specify the size of anion exchange beds and other ancillary equipment.
Reverse osmosis and electrodialysis equipment is also readily
available from a large number of manufacturers and pilot studies are not
necessary as long as the water quality is known.
TREATMENT COSTS
Of the three most effective and practical methods, ion exchange,
reverse osmosis and electrodialysis, the former, ion exchange, is the
most economical for low TDS water. Because the equipment for ion
exchange nitrate removal is similar to ion exchange softening equipment,

-------
8
and even activated alumina fluoride removal equipment, the construction
costs for all three processes should be comparable. A recent EPA publi-
cation12 on the costs of water treatment process, developed by Culp,
Wesner, Culp, Consulting Engineers, Santa Ana, California, copfirms this
as shown in Table 1 and Fig. 3. In assending order of total construction
costs up to about 1 mgd, the costs of four comparable processes are (1)
activated alumina for fluoride removal; (2) ion exchange softening; (3)
ion exchange nitrate removal and (4) reverse osmosis. The costs for the
first three processes differ by only a small amount, while reverse osmosis
costs are about 2-3 times more than the other methods.
The principal difference in the construction costs for ion exchange
softening and nitrate removal is the media expense. An example to
illustrate this difference is shown in Table 2 that provides a breakdown
in the construction costs of 1136 m^/d (0.3 mgd) plant. All of the cost
categories shown are approximately equal, except for one, the ion exchange
media. The anion resin for nitrate removal cost about three times more
than that cation resin for the softening process.
Operating cost data for the various processes are compared in
Table 3 and Fig. A. These data are total costs for maintenance materials,
energy and labor and exclude the cost of regeneration and brine disposal.
These data indicate that the operation costs for ion exchange nitrate
removal are about 5-30 percent higher than ion exchange softening and
reverse osmosis treatment is 2-5 times more costly than ion exchange.

-------
9
TABLE 1. CONSTRUCTION COSTS - WATER TREATMENT SYSTEMS12
CONSTRUCTION COSTS - $
Plant
Capacity
gpd
Ion Exchange
Softening
Activated Alumina
Fluoride Removal
Ion Exchange
Nitrate Removal
Reverse
Osmosis
50,000
100,000
250,000
500,000
750,000
1,000,000
60,500
68,000
85,000
106,000
131,000
150,000
50,000
58,000
72,000
92,000
110,000
130,000
65,000
74,000
98,000
129,000
170,000
221,000
78,000
131,000
245,000
430,000
625,000
776,000

-------
10
Fig. 3, Construction Costs - Ion Exchange Softening, Activated Alumina
Fluoride Removal, Ion Exchange Nitrate Removal and Reverse Osmosis
10', 000,000
olon Exchange: Softening
~ Activated Alumina Flouride Removal
Olon Exchange: Nitrate Removal
^Reverse Osmosis
CO
4J
CO
O
V
c
o
•H
¦U
y
3
M
U
(0
(3
O
o
1,000,000
100,000
10,000
1,000
1,000
a/
-i—I III I B]	l-l I 11 Ittj	1 I I 111 ll|
10,000	100,000
Plant Capacity (gpd)
1,000,000

-------
11
TABLE 2. CONSTURCTION COST - ION EXCHANGE PLANTS FOR
SOFTENING AND NITRATE REMOVAL12
CONSTRUCTION COST - $
Cost Ion Exchange Ion Exchange
Category Softening Nitrate Removal
	280,000 gpd	 270,000 gpd
Site work
640
110
Manufactured Equipment


Equipment
16,000
16,500
Media
6,790
21,860
Concrete
1,400
490
Steel
2,170
680
Labor
7,430
5,990
Pipe and valves
12,600
12,440
Electrical & instrumentation
21,600
21,460
Housing
8,900
8,900
Subtotal
77,270
88,430
Miscellaneous


& Contingency
11,590
13,260
TOTAL
88,860
101,690

-------
12
TABLE 3. OPERATION AND MAINTENANCE COST - WATER TREATMENT SYSTEMS12
YEARLY O/M COST - $ and (1/1000 gal)
Plant
Capacity
(gpd)
Ion Exchange
Softening
Activated Alumina
Fluoride Removal
Ion Exchange
Nitrate Removal
Reverse
Osmosis
50,000
11,500
13,000
12,400
21,300

(63)
(71)
(68)
(117)
100,000
12,200
15,000
15,000
33,500

(33)
(41)
(41)
(92)
250,000
15,500
16,000
20,000
63,000

(17)
(18)
(22)
(69)
500,000
19,800
18,000
26,900
104,000

(11)
(10)
(15)
(57)
750,000
23,000
20,800
.34,800
142,000

(8)
(8)
(13)
(52)
,000,000
25,300
23,700
46,000
180,000

(7)
(6)
(13)
(49)
*Costs Include - Maintenance Material, Energy, and Labor
Costs Exclude - Regeneration, Brine Disposal

-------
13
Operation and Maintenance Cost-Ion Exchange: Softening,
Activated Alumina Fluoride Removal, Ion Exchange: Nitrate
12
Removal, and Keverse Osmosis
Olon Exchange: Softening
oActivated Alumina Fluoride Removal
Olon Exchange: Nitrate Removal
/JReverse Osmosis
NOTE: 0/M Costs Excludes
Regeneration and Brine
Disposal Costs
250,000
500,000	750,000
Plant Capacity - gpd
1,000,000

-------
14
Regeneration and brine disposal costs will increase the costs
and are dependent on the efficiency of each process. For ion exchange,
regeneration costs are a function of the quality of the water and in
particular the total amount of sulfate, nitrate, and other anions in
the water. Because sulfate is more preferred than nitrate, the regenera-
tion costs for nitrate ion exchange are primarily dependent on the amount
of sulfate in the raw water.
To evaluate the effect of sulfate on the regneration costs for ion
exchange nitrate removal, cost data were developed assuming: (1) 100
percent efficiency in removal of sulfate and nitrate; (2) a capacity of
1.2 meq/ml of resin; (3) 240 kg of salt per cubic meter of resin
(15 lbs/cu. ft.) for regeneration; and (4) 1.5^/lb of salt. These
data given in Fig. 5 show that the sulfate concentration can increase
regeneration cost substantially; the costs will double as the sulfate
concentration increases from 50 to 250 mg/L. Increases in salt usage
and salt costs will also increase these costs. The data from Fig. 5
also show that the regeneration costs are a major portion of the operat-
ing costs that can amount to 40-50 percent of the total costs. Including
regeneration costs in the total operating costs for ion exchange treatment
also narrows the gap between this treatment luechod and reverse osmosis.
SUMMARY
Currently, ion exchange treatment is the most economically and
practical method for nitrate removal from groundwater. The construction
costs for ion exchange nitrate removal are slightly higher than those of
an ion exchange softening plant because of the higher cost of the anion

-------
17
16
15
14
13
12
11
10
9
8
7
6
5
4
15
Fig. 5. Regeneration Costs - Nitrate Removal by Ion Exchange
Treatment
0^
Resin Capacity - 1.2 meq/L
Salt Usage - 15 lbs/ff-* Resin
Salt Cost - 1.5c/lb
I	f-	1	!	1-
50	100	150	200	250
Sulfate Concentration of Raw Water - mg/L
300

-------
16
resin. Excluding regeneration and brine disposal, operating costs are
about 20-40 percent higher for nitrate removal than for softening. The
regeneration costs are a very significant portion of the total cost,
30-50 percent. Regeneration costs are directly dependent on the total
amount of nitrate and sulfate in the water.
Reverse osmosis and electrodialysis are effective, but more costly
methods for nitrate removal. The methods become more competative with
ion exchange when the sulfate and TDS concentrations are high. These
methods are also advantageous when the drinking water has a multiple
contaminant problem that cannot be solved"by anion ion exchange treatment
alone.

-------
17
REFERENCES
1.	Gunderloy, F. C. et al. Development of a Giemi©a|.Denitrification Process.
Project No. 17010 EEX. Wtr. Pol. Control Res. Ser*" EPA (Oct. 1970). .
2.	Gunderloy, F. C. et al. Dilute Solution Reactions of the Nitrate Ion as
Applied to Water Reclamation. Rpt. No. TWRC-1, FWPCA, Cincinnati, Ohio
(Oct. 1968)•
3.	O'Brien, W. J. Chemical Removal of Nitrate frcan Potable Water Supplies.
Contribution No. 23. Kansas Water Resources Res. Instit., Manhattan,
Kan. NTIS PB-193023 (Jun. 1968).
4.	Jeris JV S. et al. Biological Fluidized-Bed Treatment for BOD and
Nitrogen Removal. Jour. WPCF, 49:5:816 (May, 1977).
5.	St Amant, P. P* & McCarty, P. L. Treatment of High Nitrate Waters.
Jour. AWWA, 61:12:659 (Dec. 1969).
6.	Gregg, J« D* Nitrate Removal at Water Treatment Plant. Civ. Engrg.,
43:4:45 (Apr. 1973).
7.	Buelow, R. W. et al. Nitrate Removal by Anion Exchange Resins. Jour.
AWWA, £7:9:528 (Sept. 1975).
8.	Clifford, D. A. & Weber, W. J. Nitrate Removal from Water Supplies by
Ion Exchange, Executive Summary. EPA 600/8-77-015, Cincinnati,
Ohio (1977)
9_ Sword, B. R. Desalination of Irrigated Return Waters. Am. Geophys.
Union Nat. Fall Meet., Hydrology Section, San Francisco, Calif.
(Dec. 1969).
10.	Hindin E.; Dunstan, G. H.; & Bennett, R. J. Water Reclamation by
Reverse Osmosis. Bull. 310. Tech. Ext. Service., Washington State Univ.,
Pullman, Wash. (Aug. 1968).
11.	Tfckenaka, H. H. j Chen, C. L»j & Miele, R. P« Demineralization of
Wastewater by Electrodialysis. Ofce. Res./Devel. EPA, Cincinnati, Ohio
(Oct. 1975).
12.	Gumerman, R. C., Culp, R. L. and Hansen, S. P. Estimating Costs for .Water
Treatment as a Function of Size and Treatment Efficiency. EPAr600/2-
78-182, Cincinnati, Ohio (1979).
» U^aOVSMffifTPmnDieOffiCC: ittao-657-146/5*27
LIBRARY
US EPA Region 4
AFC/MI) FL Tower
61 Forsyth St. S.W.
Atlanta, GA 30303-3104

-------