EPA/600/A-97/063
Pilot-Scale Evaluation of the Iron-Enhanced Dechlorination Technology for
              Remediation of Contaminated Groundwater
                           Chien T. Chen
                 U.S. Environmental Protection Agency
                 Urban Watershed Management Branch
                       Edison, NJ 08837-3679
                            Presented at
              World Environmental Congress (WORLD '96)
                           Cincinnati, OH
                        October 26-29,1996

-------
  Pilot-Scale Evaluation of the Iron-Enhanced Dechlorination Technology for
                   Remediation of Contaminated Groundwater

                                     Chien T. Chen
                          U.S. Environmental Protection Agency
                          Urban Watershed Management Branch
                           2890 Woodbridge Avenue, MS-104
                                 Edison, NJ 0883 7-3679

                                       Abstract

       The iron-enhanced dechlorination technology was evaluated under the U.S.
Environmental Protection Agency (EPA) Superfund Innovative Technology Evaluation (SITE)
program at a contaminated site in New Jersey... This paper describes the ability of this
technology to destroy chlorinated volatile organic compounds (VOCs) in contaminated
groundwater.  Specifically, this paper discusses performance and economic data from a
demonstration of the technology.
       The technology involves oxidation of iron and reductive dechlorination of chlorinated
VOCs in aqueous media. During reductive dehalogenation, VOCs are converted to hydrocarbons
and inorganic chlorides. The process can be used for either in situ or ex situ groundwater
treatment.
       This process was demonstrated using a pilot-scale, aboveground treatment reactor
containing the  reactive iron medium.  A flow rate of about 0.5 gallons per minute was maintained
throughout a 13-week  demonstration period; about 60,800 gallons of groundwater were treated.
During the demonstration of the aboveground reactor, water samples were collected at influent,
intermediate, and -effluent sampling locations and analyzed for VOCs and inorganic parameters
(e.g., concentrations of dissoved metals, chloride, sulfate,total inorganic carbons; pH; Eh).
VOCs present  in influent groundwater or generated as degradation by-products were considered
critical analytes for the SITE demonstration.
       The analytical results indicated that influent groundwater was contaminated with
(1) trichloroethene (TCE) at concentrations ranging from 54 to 590 micrograms per liter (,ug/L);
(2) tetrachloroethene (PCE)  at concentrations ranging from 4,100 to 13,000 //g/L; and, (3)
eis-l,2-dichloroethene  (cDCE) at concentrations ranging from 35 to 1,600 jwg/L. Vinyl chloride
(VC) was not detected in the influent groundwater.  Analytical results for the effluent samples
indicated that the iron-enhanced dechlorination process signficantly reduced the total
concentrations of chlorinated VOCs in water treated, and consistently achieved the demonstration
effluent target level of  1 //g/L for TCE and PCE.  During the last  two weeks of the demonstration
the process did not consistently achieve the effluent target levels of 2 jUg/L for VC and 5 ^g/L for
cDCE. Although some cDCE was present in the influent groundwater, most of the cDCE and VC
appears to have formed through the degradation of PCE and TCE. The incomplete dechlorination
of cDCE and VC in the latter portion of the SITE demonstration may have been caused by PCE
persisting to greater depths within the reactor than anticipated. These factors may have resulted
from a gradual reduction in the iron's  reactive surface area through formation of precipitates.

-------
       Based on information obtained from the SITE demonstration, groundwater remediation
 costs for an aboveground reactor using the iron-enhanced dechlorination process are estimated to
 be about $91 per 1,000 gallons treated.  This cost was estimated based on data from a pilot-scale
 system operating at a flow rate of 0.5 gpm extrapolated to a 30-years operational period.

                                       Introduction

       Common technologies (e.g., air sparging/soil vapor extraction with carbon adsorption) for
 remediating groundwater contaminants with chlorinated solvents, involve the transfer of
 contaminants from water to another medium.  As regulatory requirements for the treatment of the
 contaminated medium become more stringent and more expensive to comply with, the iron
 enhanced technology may offer a major advantage over other treatment technologies because it
 destroys hazardous substances rather then transferring them to another medium.
       The iron-enhanced dechlorination technology involves oxidation of iron and reductive
 dechloenation of chlorinated VOCs in aqueous media.  Although aluminum, copper, brass,
 standard steel, and zinc have also been shown to promote reductive dechlorination of VOCs
 (O'Hannesin and Gillham,  1992), metallic iron has been chosen for use in large-scale
 applications of the  technology.  Metallic iron is readily available, inexpensive, and induces rapid
 dechlorination of organic compounds.  The technology induces conditions that cause substitution
 of chlorine atoms by hydrogen atoms. Equations (I) through (5) may describe the reactions that
 take place in the presence of water, zero^valent iron (Fe°), and a chlorinated hydrocarbon (Rcl)
 (Gillham and O'Hannesin,  1994):
                                 2Fe°-->2Fe2* + 4e-                             (1)
                                3H20 ->3H+ + 30H                             (2)
                            2H+ + 2e--->H2(g)                                   (3)
                        RCI + H* + 2 e --> RH + CI                               (4)
                  2 Fe°  + 3 H20 + RCI -> 2 Fe2+ + 3 OH + H2(g) + RH + CI         (5)

The overall reaction that takes place (Equation 5) results in the formation of Fe2+,  hydroxyl ions
(OH"), hydrogen gas [H2(g)], nonchlorinated hydrocarbons (R-H), and chloride ions (CI").
       This technology can be installed and operated in either an aboveground reactor or in situ
as a permeable treatment wall (EPA, 1995). An aboveground, pilot-scale reactor was used for the
Superfund Innovative Technology Evaluation (SITE) demonstration at a contaminated site in
New Jersey. The pilot-scale reactor was designed to evaluate the technology's suitability for full-
scale  remediation at this site, and to gather data regarding fiill-scale system design and operating
parameters.
       The primary purpose of the SITE Program is to advance the development and
demonstration, and thereby establish the commercial availability, of innovative treatment
technologies applicable to Superfund and other hazardous waste sites.  The SITE Program was
established by the U.S. Environmental Protection Agency (EPA) Office of Solid Waste and
Emergency Response (OSWER) and Office of Research and Development (ORD) in response to
the Superfund Amendments and Reauthorization Act of 1986 (SARA), which recognized the
need for an alternative or innovative treatment technology research and demonstration program.

-------
       This process has shown very effective  for the dechlorination of various VOCs (e.g.,
chlorinated methanes, ethanes, and ethenes) (Gillham and others, 1993) over a wide range of
concentrations in laboratory-scale tests. However, its effectiveness, possible problems and cost
for a large scale treatment has not been evaluated.  This paper describes the results of a pilot-
scale.demonstration under natural conditions.  These results can be used for control, modification,
and optimization of design and operating parameters for either the above-reactor or in situ
permeable wall.

                                       Experimental

       Groundwater at the demonstration site was treated in a reactor (see Figure 1).  The
reactor was a 9-foot-high, 8-foot-diameter fiberglass-reinforced  plastic tank containing a 5.5-foot-
thick layer of reactive granular iron. The reactive iron rested on top of a layer of coarse silica
sand, referred  to as "well sand," placed in the bottom of the reactor.  The well sand in the bottom
of the reactor prevented granular iron from washing out into the effluent pipe.  Pea gravel or well
sand can also be placed on top of the reactive iron to act as a prefilter, but was not used during
the SITE demonstration as the reactor feed line was equipped with an in-line 5-micron prefilter.
Eliminating the upper filter layer also allowed observation of and direct access to the top of the
iron.  The top of the reactor was equipped with a passive gas vent to prevent accumulation of
excess pressure, and a manhole with a sightglass to allow observation of the reactive iron surface
and access to the vessel interior.
       The influent groundwater fed to the reactor was collected from the shallow,
unconsolidated zone and the underlying, fractured bedrock aquifer (PRC, 1994).  Two trenches
passively collected contaminated groundwater from the shallow zone. The trenches drained to a
common sump; water was pumped directly from the sump to the feed line for the reactor.  Two
pumping wells extracted groundwater from the bedrock aquifer.  Water from these wells flowed
to a common pipe, and then directly into the reactor feed pipe. The influent groundwater
contained TCE at concentrations ranging from 54 to 590 /t/g/L, PCE at concentrations ranging
from 4,100 to  13,000 //g/L, and cDCE at concentrations ranging from less than 25 to 1,200 //g/L.
       The influent groundwater passed through an air eliminator, a 5-micron water filter (to
remove suspended solids, which may inhibit flow through the reactive iron medium), and then
entered the reactor. Water was pumped into the reactor at a sufficient rate to maintain a 2-foot-
deep layer of water ponded above the iron at all times, to prevent rust from forming on the iron
surface and to  minimize variations in volumetric flow through the reactor.  The water then flowed
through the reactive iron layer by gravity.  The treated water flowed to a perforated pipe in the
well sand, and  then out through an effluent pipe. The volumetric flow rate, flow velocity and
residence time were controlled by the thickness, porosity, and permeability of the iron layer and
the configuration of the effluent piping. A flow rate of about 0.5 gpm was maintained throughout
the SITE demonstration period. The estimated residence time of 27.5 hours equates to a vertical
flow velocity of about 4.8 feet per day, based on an assumed iron porosity of about 40 percent
(Vogans  and others, 1995).
       The operating parameters (system design and flow rate) were determined with laboratory
column tests using the groundwater from the site.  The SITE program evaluated the treatment
reactor's  effectiveness over a period of 13 weeks. During this period, three types of data were

-------
 collected: 1) analytical data for water samples collected from the reactor; 2) analytical data for
 samples of other media (reactive iron and air); and 3) operating data and observations, such as
 cumulative volume treated, flow rate, and electrical consumption, .
       The total amount of iron placed in the reactor was 42,920 pounds.  The total volume of
 reactor space filled by the iron was 277 cubic feet; therefore, the iron's bulk density after settling
 was 155 lbs/ft3'.  During the demonstration: the flow rate (through the reactor), cumulative
 volume treated, and electrical power consumption were recorded; and groundwater samples at the
 reactor's influent (II), control (Rl), intermediate (R2 through R5), and effluent (El) sampling
 locations (Figure 1).  Sampling locations II and El were taps on the reactor's influent and
 effluent lines, respectively.  The other locations (Rl through R5) consisted of slotted stainless
 steel tubes that extended to the reactor's interior.  The tubes were capped when not in use; to
 obtain samples, the tubes were uncapped, and water flowed out into the sample containers..
 Sampling locations R2 through R-5 were spaced at various depths through the layer of reactive
 iron, to evaluate changes as water passed through the reactive medium.
       The influent and effluent water were collected weekly during the demonstration to
 determine and monitor the concentrations of various chlorinated hydrocarbons.  These include
 tetrachloro ethene (PCE, trichloro ethene (TCE), 1,1-dichloro ethene (1,1-DCE), eis-l,2-diehloro
 ethene (cDCE), and vinyl chloride (VC). During weeks 1, 5, 9, and  13 of the demonstration,
 water samples from locations Rl,  R2,  R3, R4, and R5, as well as the influent and effluent
 locations were collected. In addition to the aforementioned chlorinated hydrocarbons, the
 dissolved metals, chloride,  sulfate, and total inorganic carbons were also analyzed for these
 collected water samples. Field measurements, of dissolved oxygen (DO), temperature, specific
 conductance, pH, and Eh. Were conducted on all the water samples collected. In addition to
 water, air gas samples from the headspace of the reactor interior during weeks 5, 9, and 13 were
 also collected. The concentrations of PCE and TCE in the gaseous samples were analyzed.
       The first sampling event (week 1) was performed after about three pore volumes of
 groundwater had passed through the  treatment reactor.  One pore volume equals the volume of
 pore space of the reactive iron medium and is estimated by the developer as about 40 to 45
 percent of the total volume of the reactive iron medium. Based on the volume of iron in the
 reactor, the pore space was about 110 cubic feet, indicating that the pore space of the iron
 probably held approximately 827 gallons of water at any given time during the demonstration.

                                  Results and Discussion

       Table 1 presents  a summary of the concentrations of chlorinated VOCs detected in
 samples collected at the influent and effluent sampling locations during weeks 1 through 13.
Figures 2 through 5 summarizes concentrations of these VOCs detected at all (influent, control,
intermediate, and effluent)  sampling locations during weeks 1, 5, 9, and  13.
       The relatively high PCE concentrations in the influent groundwater necessitated dilution of
some samples to bring the PCE concentrations within the quantifiable range. In diluted samples
the detection limits were adjusted for the dilution factor.  For this reason, it is possible that
relatively small amounts  of cDCE, VC, and possibly other VOCs, were present in the diluted
samples and were not detected. No effluent samples required dilution; therefore, the detection
limits achieved for these  samples were all lower than the applicable effluent target levels.

-------
Applicable effluent target levels for all VOCs detected are summarized in Table 1. Compliance
with the target levels was evaluated by comparing the effluent VOC concentrations with the most
stringent effluent target levels.
       The analytical results shown in Table 1 indicate that the TCE and PCE were detected in
the influent during all weeks of testing; however, the influent concentrations of PCE and TCE
were variable.  Influent TCE concentrations ranged from 54 to 590 Mg/L, and influent PCE
concentrations ranged from 4,100 to 13,000 jUg/L. The types and concentrations of VOCs in the
influent may have varied due to the effects of mixing of groundwater from the two different zones
(the shallow zone and the bedrock aquifer), which typically contain different concentrations of the
various contaminants.
       Based on comparison of influent and effluent samples, the iron-enhanced dechlorination
process significantly reduced the total chlorinated VOC concentrations in groundwater treated by
the reactor.  Concentrations of PCE and TCE in the effluent were consistently below the
detection limit of 0.9 ^g/L during all weeks of testing, and thus were also below the applicable
target effluent level of 1 ,ug/L for both compounds. As shown in Figures 2 and 3, PCE and TCE
concentrations at intermediate  sampling locations generally increased over the demonstration
period, but were reduced to below detectable levels before exiting the reactor in all weeks of
testing.
       cDCE was not detected in influent groundwater during weeks 1 to 5, or during weeks 7
and 9.  cDCE was detected in the influent groundwater in weeks 6, 8, 10, 11, 12, and 13.  The
detection limit for VOCs (including cDCE)  in the influent groundwater samples was 25 ug/L for
all weeks except weeks  1 and 7; for these 2 weeks the influent detection limits were 250 ug/L and
50 ug/L, respectively. The detection limit of 1.0  /ug/L for cDCE was achieved for effluent
samples during all weeks.  cDCE was not detected in the effluent during the first 8 weeks of the
demonstration, but was detected in the effluent during weeks 9 through 13. The technology
acheived the NJDEP site-specific discharge limit of 5 /ug/L for cDCE for all weeks except week
13. Although cDCE was detected in the influent  groundwater during some weeks, during weeks
1, 5, 9, and 13 the highest cDCE concentrations were detected at the intermediate locations,
indicating that cDCE was also introduced as a by-product of the dechlorination of PCE and TCE
(see Figure 5).  Generally, the concentrations of cDCE in the effluent groundwater increased
consistently from 1.3 /ug/L during week 9 to 37 /ug/L during week 13 (see Figures 4 and 5).

-------
Table 1
Chlorinated VOC Concentrations at Influent and Effluent Sampling Locations
VOC
PCE
VC
TCE
cDCE
Weekl
I
5,900
<250
110a
<250
E
<0.9
<1
<0.9
<1
Week 2
I
9,800
<25
69
<25
E
<0.9
<1
<0.9
<1
Week 3
I
9,700
<25
130
<25
E
<0.9
<1
<0.9
<1
Week 4
I
9,800
<50
120
<50
E
<0.9
<1
<0.9
<1
WeekS
I
13,000
<50
110
<50
E
<0.9
<1
<0.9
<1
Week 6
I
7,200
<25
350
1,200
E
<0.9
<1
<0.9
<1
Week 7
I
6,900
<50
54
<50
E
<0,9
<1
<0,9
<1
Target Effluent
Levels
Federa
1MCL
5
2
5
70
NJDEP
Discharg
e Limit
1
5
1
5
VOC
PCE
VC
TCE
cDCE
Week 8
I
5,900
<25
79
65
E
<0.9
<1
<0.9
<1
Week 9
1
8,900
<25
54
<25
E
<0.9
<1
<0.9
1.3
Week 10
I
7,300
<25
110
67
E
<0.9
1.4
<0.9
2.4
Week 1 1
1
4,100
<25
590
1,600
E
<0.9
1.2
<1
2.8
Week 12
I
7,100
<25
99
35
E
<0,9
2.8
<0.9
2.3
Week 13
I
7,900
<25
180
330
E
<0.9
8.4
<0.9
37
Target Effluent Levels
Federal
MCL
5
2
5
70
NJDEP
Discharge
Limit
1
5
1
2
Notes:
        All analytical results are presented in /wg/L.
        I = Influent groundwater
        1 Value estimated without sample reanalysis
                               < = Less than the detection limit shown.
                               E = Effluent groundwater

-------
       VC was not detected in the influent groundwater during the SITE demonstration.
Because the detection limits were adjusted for dilutions, it is possible that VC was present in the
influent samples at low concentrations and was not detected.  However, past groundwater
monitoring data do indicate that VC is typically present in site groundwater at significant
concentrations. VC was detected in the effluent samples collected during weeks 10, 11, 12 and
13. The effluent concentrations of VC during these weeks increased from 1.2 Mg/L during week
10 to 8.4 Aig/L during week 13 (see Figure 5). VC concentrations in the effluent exceeded the
applicable maximum contaminant level (MCL) of 2 //g/L during weeks 12 and 13, but were
relatively low (2.8 /ug/L duirng week 12 and and  8.4 /ug/L during week 13).  VC is a common by-
product of PCE, TCE, and cDCE dechlorination.  Because VC was not detected in the influent
groundwater it was probably formed during the reductive dechlorination of PCE, TCE, and
cDCE.
       Based on the demonstration results, the following reaction mechansium is suggested
(Chen, 1995):

                          Fe -> Fe2* + 2e                                       (6)
                        H2O->H + + OH                                       (7)
        Cl 2C = CC12 + H+ + 2e -> CICH=CCI2 + CI                                (8)
       CICH = CC12 + H+ + 2e -> CICH=CHCI + CI                               (9)
      CICH = CHCI + H+ + 2e -> CH2 = CHC1 + CI                               (10)
        CH2 = CHCI + H+ + 2e -> CH2 = CH2 + CI                                 (11)

       During the early part of the SITE demonstration, the iron was still very reactive and was
able to rapidly reduce all by-products (TCE, cDCE, and VC) generated as PCE degraded.
However, as the demonstration progressed, the reactivity of the iron decreased and the produced
TCE could not be  immediately reduced, leading to increases in TCE concentrations and
incomplete dechlorination  of cDCE and VC. For this reason, during this demonstration the
dechlorination of mulit-chlorinated VOCs appeared to be continuous and sequential, rather than
occuring in one precipitous step.
       Table 2  presents the pH values measured at the influent  and effluent sampling locations
during all weeks of testing. Table 3 shows the pH values measured at all locations during weeks
1, 5, 9, and 13,  Generally, the  pH increased progressively as groundwater moved through the
reactor during all weeks except week 13. Equations 1 through 4 may explain the increase in pH.
In these reactions,  H+ is consumed  so that the pH rises and significant amounts of OH' ions
appear.
       Sulfate concentrations were measured to evaluate, in  part, the potential for sulfate-
reducing bacterial growth and precipitation of metal sulfates.  Table 2 shows that, except during
week 1, the concentration of sulfate did not change significantly during or after treatment. During
week 1, the influent sulfate concentration was 27.3 mg/L, and the effluent sulfate concentration
was less than 5 mg/L.  However, even during week 1 the decrease in sulfate concentration did
not progress consistently through the reactor.

-------
ForTable 2   pH at all locations
Sampling Location
Rl Distance(inches)
Weekl
WeekS
Week 9
Week 13
11
-
7.52
7.52
7.66
8.90
Rl
0
7.30
7.43
7.77
8.63
R2
12
7.36
8.75
7.78
8.74
R3
24
7.89
8.48
7.92
8.81
R4
42
8.14
10.72
9.17
Not
measured
R5
60
8.32
10.96
9.04
8.73
El
66
8.72
10.69
9.36
8.91
Note: Rl Distance = Distance through reactive iron.

Table 3       pH at Influent and Effluent Sampling Locations
Week
1
2
3
4
5
6
7
8
9
10
11
12
13
pHatll:
7.52
7.67
7.90
8.42
7.52
7.58
8.11
7.73
7.66
7.57
7.17
7.66
8.90
pHatEl:
8.72
8.63
8.72
10.82
10.69
9.57
9.63
9.69
9.36
9.37
9.76
9.60
8.91
these reasons the sulfate data provide no evidence of metal-sulfate precipitation or bacterial
growth.
       Chloride concentrations were determined because they may correlate with dechlorination
of VOCs.  However, because the background chloride concentrations are relatively high
compared to the VOC concentrations, no significant trends in chloride concentrations were noted
during treatment as a result of VOC dechlorination.
       TIC concentrations generally decreased from concentrations measured at the influent and
control sampling locations as the groundwater moved through the reactor (see Table 4). This
decrease in TIC concentration may be caused by the precipitation of metal carbonate compounds.

-------
 Precipitation of calcium carbonate (CaC03) (as well as iron carbonates) may be attributed to the
 removal of C032". The OH" produced from the dissolution of water as described in equation 2
 may react with bicarbon ate ions (HC03") in the groundwater to produce carbonate ions
 Table 4
Weekl
Summary of Inorganic Analyte Data

Analyfe:
Chloride
Sulfate
Aluminum
Barium
Calcium
Iron
Magnesium
Manganese
Potassium
Sodium
Zinc
TIC

T!
85,8
27.3
U
0.311
76.1
0.114
25.1
0.012
U
11.8
0.043
40
y~*
Rl
73.7
33.3
U
0.258
63.1
U
19.2
0.037
9.36
15.7
0.043
36

R2
75.9
30.4
U
0.171
48.6
U
21.5
0.812
2.58
11.8
0.041
18

II 111 iTH1lll|»
R3
76.4
46
U
0,230
34.8
U
15.7
0.246
4.17
12.6
0.031
4.0
1" T 4
&
R4
78.7
8.12
U
0.192
23.1
U
2.55
0.278
6.12
17.5
0.024
1.0

R5
73.2
U
U
0.125
19.4
U
1.32
0.099
6,92
21.6
0.023
1,0

El
73.7
U
2.16
U
33.4
1.05
10.9
U
16.6
29.1
0.025
1.0
WeekS
Concentration at Sampling Location:
Analyte
Chloride
Sulfate
Aluminum
Barium
Calcium
Iron
Magnesium
Manganese
Potassium
Sodium
Zinc
TIC
11
77.7
34.1
U
0.299
72.1
U
22.8
U
1.05
9.28
U
38
Rl
78.1
34.6
U
0.297
71.7
U
22.5
U
U
9.70
0.012
39
R2
81.1
34.2
U
0.038
34.3
0,091
21.8
0.311
U
9.27
U
18
R3
84.7
34.4
U
0.034
28
U
21.9
1.40
1.11
9.06
U
12
R4
82.9
35.6
U
0.102
24.3
U
19.2
0.270
1.36
9,11
0.012
4.8
R5
82.8
34.7
U
0.136
24
U
19.1
0.025
1.26
9.49
U
3.2
El
82.9
33.8
U
0.148
25.5
U
17.9
0.135
1.45
9.43
U
3.7
Notes:   All concentrations in milligrams per liter (nig/L)
       U = Analyte not detected    TIC = Total inorganic carbon

-------
 Table 4 Summary of In organic Analyle Data (Continued)
 Week 9
Concentration at Sampling Location:
Analyte
Chloride
Sulfate
Aluminum
Arsenic
Barium
Calcium
Iron
Magnesium
Manganese
Potassium
Sodium
Zinc
TIC
11
70.9
30,2
U
0.0211
0.298
68.8
U
22.6
U
U
9.71
U
18
Rl
69.9
30.5
U
0.0169
0.255
66.3
U
21.1
0.0449
1.74
13.1
U
18
R2
74.4
31.2
U
0.0117
0.0356
37.0
0.557
20.8
0.166
1.86
13.6
U
14
R3
76.6
28.8
U
0.0135
0.0347
28.7
U
18.4
0.470
2.34
14.2
U
11
R4
75.7
25.2
U
U
0.0313
24.6
U
15.7
0.883
2.95
15.9
U
6.5
R5
72.9
23,6
U
0.0121
0.0653
23.1
U
13.8
0.222
3.14
16.0
U
4.7
El
76.2
26.8
U
0.0160
0.0639
21.1
U
15.2
0.0781
3.04
16.0
U
5.3
Week 13
Concentration at Sampling Location:
Analyte
Chloride
Sulfate
Aluminum
Barium
Calcium
Iron
Magnesium
Manganese
Potassium
Sodium
Zinc
TIC
11
83.3
32.0
U
0.327
77.6
U
24.8
0.0260
U
10.5
U
40
Rl
78.7
31.5
U
0.322
74.8
U
23.7
0.0165
1.24
10.3
U
37
R2
81.1
32.7
U
0.0715
54.2
2.11
23.7
0.142
U
10.2
U
31
R3
81.9
30.2
U
0.0306
39.3
.228
23.6
0.284
U
10.9
U
19
R4
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
R5
83.7
31.7
U
0.0454
25.8
U
20.7
0.213
1.29
10.3
U
5.3
El
83.5
31.9
U
0.0259
26.8
U
23.1
0.649
3.27
10.5
U
8.3
NOTES: All results in milligrams per liter (mg/L)
        U = Analyte not detected; associated value is quantitation limit
        N/A ~ Sample not collected - sampling location R4 inaccessible during week 13
TIC = Total inorganic carbon
                                                   10

-------
       The concentrations of dissolved calcium, magnesium, and barium generally decreased as
water moved through the reactor (see Table 2).  During weeks 1, 5, and 9, the decrease in
concentrations of these metals coincided with an increase in measured pH values, suggesting that
geochemical conditions were conducive to decreased solubility and increase precipitation of some
metal compounds  The decreasing concentrations of barium, calcium, and magnesium as water
moved through the reactor are probably indicative of metal compounds such as calcium carbonate
precipitating from the water.
       Th cost for using the iron-enhanced dechlorination technology, in an aboveground reactor,
to treat contaminated groundwater is estimated.  The cost estimates are based on a reactor
designed to treat the types and concentrations of chlorinated VOCs observed at the
demonstration site, and were based on data compiled during the SITE demonstration. Costs for
this technology are highly variable and are dependent on the types and concentrations of the
contaminants present, regulatory cleanup requirements, and other site-specific factors.  Estimates
for total cost and cost per gallon of water treated are also heavily influenced by assumptions
regarding the duration of the treatment program and the cumulative volume treated.  The purpose
of the pilot-scale system was to determine the optimal design and operating parameters for a full-
scale system; differences between the capabilities of the pilot-scale and full-scale systems could
significantly affect costs. For these reasons, costs for full-scale systems designed for optimal
performance at other sites may vary significantly from estimates presented herein.
       Due to the numerous factors that potentially affect the cost of using this technology,
various assumptions  were necessary to prepare the economic analysis.  Some of the most
significant assumptions were (1) the aboveground reactor is identical to the pilot-scale reactor
used at the demonstration site (2) the reactor will treat water contaminated with PCE, TCE,
cDCE, and VC at concentrations observed during the demonstration (3) the reactor operates at
0.5 gpm, as demonstrated. Also, the cost evaluation is based on data obtained during the SITE
demonstration, extrapolated  to a 30-year operational period. The 30-year timeframe was selected
for consistency with cost evaluations of other innovative technologies evaluated by the EPA SITE
Program,  and because it facilitates comparison to typical costs associated with conventional, long-
term remedial options.
        Costs have been distributed among 12 categories applicable to  typical cleanup  activities
at Superfund  and RCRA sites (Evans 1990).  Costs  are presented in July 1995 dollars and are
rounded to the nearest 100 dollars. These factors include : (1) treatability study, (2) system
design, (3) site preparation, (4) permitting and regulation, (5) Mobilization and startup, (6) capital
equipments, (7) demobilization, (8) supplies, (9) utilities, (10) treatment and disposal of effluent
and residual waste, (11) analytical  services and (12) equipment maintenance.  After operating for
30 years, the total costs of the groundwater remediation scenario presented in this analysis are
$714,700. Costs were not adjusted for inflation. A total of nearly 7.9 million gallons of
groundwater would be treated over this time period. Based on these criteria, the total cost per
1,000.gallons treated is $91,  or roughly 9.1 cents per gallon.

                                        Conclusion

       The SITE demonstration of the iron-enhanced dechlorination technology produced the
following  key findings:
                                            11

-------
              *      The iron-enchanced dechlorination process significantly reduced the total
                     concentrations of chlorinated VOCs present in the water treated. The
                     effluent water met the target concentration of 1 fj,g/L for TCE and PCE,
                     during each of the 13 weeks of testing.  No cDCE or VC was detected in
                     the effluent during weeks 1 through 8; however, low concentrations of
                     cDCE and VC were detected in the effluent in the latter part of the
                     demonstration.  The effluent groundwater met the target levels of 2//g/L
                     for VC during weeks 1 through 11, and 5 j/g/L for cDCE during weeks 1
                     through 12.  VC concentrations during weeks 12 and 13 (2.8 Atg/L and 8.4
                     /ug/L, respectively) and cDCE concentrations duirng week 13 (37^g/L)
                     exceeded the target levels.
              •      The PCE removal efficiencies were consistently greater than 99.9 percent
                     during each week of testing. Results from weeks 1, 5, 9, and 13 indicate
                     that PCE concentrations  increased  at the intermediate sampling locations,
                     suggesting that PCE was persisting to increasingly greater depths as the
                     demonstration progressed,
              •      The concentrations of chloride and sulfate did not change significantly as
                     water moved through the reactor.  The groundwater TIC concentration
                     decreased as a function of reactor depth.
              •      The concentrations of dissolved metals such as calcium, magnesium, and
                     barium changed as groundwater moved through the reactor,  apparently as
                     a result of pH increase induced by the process.  Generally, the decrease in
                     concentrations of calcium, magnesium, and barium coincided with an
                     increase in pH.  Iron concentrations were higher at intermediate sampling
                     locations than iron concentrations in the influent and effluent samples, also
                     possibly due to the effects of pH on solubility of iron compounds.
              *      The main operating problem observed during the demonstration was the
                     deposition precipitated.  A hard, crust-like layer formed in the upper few
                     inches of the reactive iron, which according to the developer was primarily
                     carbonate compounds that had been produced during the dechlorination
                     process. The crust was periodically manually broken up during the
                     demonstration to maintain unrestricted flow through the reactor.
              «      Precipitates may also have affected  the reactor's performance by blocking
                     the iron surfaces available for reaction, thereby reducing the reactivity of
                     the upper portion of the iron.  This nay have contributed to  the  increasing
                     persistence of PCE over the demonstration period and the incomplete
                     dechlorination of cDCE and VC in  the latter part of the demonstration.
       Costs for using the iron-enhanced dechlorination process are highly dependent on site-
specific factors, and highly variable.  The cost estimate assumed that groundwater was
contaminated with the same types and concentrations of chlorinated  VOCs present in
groundwater at this demonstration site, and assumed that design and operating parameters for the
treatment system were the same as for the pilot-scale system at this site. Based on these
assumptions, the total costs directly related to the metal-enhanced dechlorination process are
estimated to be $91 per 1,000 gallons  of groundwater treated.  Due  to potential differences
between the capabilities of pilot-scale systems and full-scale systems designed for optimal

                                            12

-------
performance, and varying site-specific factors, costs per gallon treated could be significantly less
for a fiill-scale application of the metal-enhanced dechlorination process at other sites.

                                     References

Chen, Chien T. 1995a. "Iron Enhanced Dechlorination of Chlorinated Hydrocarbons", 21st
       Annual Research Syposium Abstract Proceedings, April, pp 74-78

Chen, Chien T. 1995b, Excerpts from Presentation Titled "Iron Reactive Wall," Innovative Site
        Remediation Workshop, Sturbridge, Massachusetts, Sept. 13-14.

EnviroMetal Technologies, Inc. Report on Performance of Test Reactor,  former SQL Printed
        Circuits Facility, Wayne, New Jersey.  September 28.

Gillham, Robert W., and others.  1993.  "Metal Enhanced Abiotic Degradation of Halogenated
       Aliphatics: Laboratory Tests and Field Trials."  Paper Presented at the 1993 HazMat
       Central Conference.  Chicago, Illinois.  March 9-11.

Gillham, Robert W., and Stephanie F. O'Hannesin.  1994. "Enhanced Degradation of
       Halogenated Aliphatics by Zero-Valent Iron." Ground Water.  Vol. 32, No. 6, pp. 958 -
       967.

O'Hannesin, Stephanie F., and Robert W. Gillham.  1992. "A Permeable Reaction Wall for In
       Situ Degradation of Halogenated Organic Compounds." Paper Presented at the 1992,
       45th Canadian Geotechnical Society Conference.  Toronto, Ontario, Canada.  October.

PRC Environmental Management, Inc. (PRC). 1994. EnviroMetal Technologies, Inc. "Metal
       Enhanced Abiotic Degradation Technology Demonstration Final Quality Assurance
       Project Plan," Submitted to EPA ORD, Cincinnati,  Ohio. October.

EPA. 1995.  Metal-Enhanced Dechlorination Technology, Demonstration Bulletin,
       EPA/540/MR-95/5100, May.

Vogan, John L., and others.  1995. "Site-Specific Degradation of VOCs in Groundwater Using
       Zero-Valent Iron." Preprint Extended Abstract.  Presented Before the Division of
       Environmental Chemistry.  American Chemical Society (ACS). Anaheim, California.

Yamane, C.L., and others. 1995.  "Installation of a Subsurface Groundwater Treatment
       Wall Composed of Granular Zero-Valent Iron." Preprint Extend Abstract Presented
       before the Division of Environmental Chemistry,  American Chemical Society. Anaheim,
       California.
                                          13

-------
                                        figure  1
                         Schematic of Aboregromid Reactor Design
             Row

     Water   Meter     Air
Manhole
                        Gas Vent
9'
                                                                     Sealing Flange
                                                                        Top
                                                                                Outflow
                                                                          Efluent
                                                                           Line
     - Sampling Location

-------
                                      Figure 2
              PCE Concentration vs. Distance through Reactive Iron
^ s
•J^

2
§
(j
1
o
+^
2
§
0
U
UJ
U
1UVJVJVJ
9000
j
8000
7000 E

6000 <
5000

4000
3000
2000

1000
/•>

-
^
— \
\
*\'\. 	
\ \ ••..^.
\\\\

- \ \ \
: \ ^\\

\ ^N\ '^•••.
- \ ^\ 	
X .\^_ " 	 -,.-
Week
~

•* WeekS
a Week 9

-D. Week 13







|
0 10 20 30 40 50 60 7(
                         Distance Through Reactive Iron (inches)

Notes: 1) Sampling location Rl used as distance x = 0 inches. 2) Sampling location R4 (42 inches)
     not sampled during week 13. 3) All non-detect values assumed to be 0 for plotting purposes.
                                   Figure 3
           TCE Concentration vs. Distance through Reactive Iron
JVJU
0
^^
g 400

eb
s
1 300
c
0

2 200

c
0>
u
c <
U 100,
U
n <

«
~ f \
* *
f %
^ ^ %
^ %
« »
A* »
/
' \
/ */
r / x
• Week 1

A WeekS

B Week 9

0 Week 13

/ / . »
y * *

"" / m/^ \^^^^ \

+ **J ^^^^1 *
^ ** ^^ \ ^^^. *
~ rf* * J ^^. *
>V \ >s
(/ \ \\
/ » \ \
V-
      0
10        20        30        40         50
       Distance Through Reactive Iron (inches)
60
 Notes: 1) Sampling location Rl used as distance \ = 0 inches. 2) Sampling location R4 (42 inches)
      not sampled during week 13. 3) All non-detect values assumed to be 0 for plotting purposes.
70

-------
                                             Figure 4
                     cDCE Concentration vs. Distance through Reactive Iron
n (micrograms/li
>ncentratic
U
u
n
JUU
400
300
200
i nn
lUU <

— * %
"* iX" ^^ *
* j ^W *
* f * X
— * * *
* .f ^v %
*&f *
(•** ^s^^^^t fc ^^
~^~ %%x"* 	 \.
*£_ 	 1 • 1 	 El 	 -^ 	 i • «, I. ^| 	 L,

+ Weekl
-A- WeekS
3 Week 9
«&• Week 13

..»-:l'l'.




      u
                         10
 20         30         40         50
Distance Through Reactive Iron (inches)
60
70
      Notes: Sampling location Rl used as distance x *• 0 inches.
      Sampling location R4 (42 inches) not sampled during week 13.
                                                Figure 5
                               Effluent Concentrations of VC and cDCE
Week of Testing
aVC
BcDCE
9
<1
1.3
10
1.4
2.4
11
1.2
2.8
12
2.8
2.3
13
8.4
37.0
Note: Only concentrations greater than applicable detection limits are plotted.

-------
                                       TECHNICAL REPORT DATA
                                  (Please read instructions on the reverse before completinal
1.RETORT NO.
  EPA/600/A-97/063
3. RECU,
4.TITLE AND SUBTITLE

    Pilot-Scale Evaluation of the Iron-Enhanced Dechlorination
    Technology for Remediation of Contaminated Groundwater
5. REPORT DATE
                   10/15/96
          6. PERFORMING ORGANIZATION CODE
 7. AUTHOR(S)
  C.T. Chen, US EPA, UWMB, Edison, NJ 08837
 8.. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
    PRC
10. PROGRAM ELEMENT NO.
     TD1Y1A
                                                              11. CONTRACT/GRANT NO.
 12. SPONSORING AGENCY NAME AND ADDRESS
   National Risk Management Research Laboratory
   Office of Research and Development
   U.S. Environmental Protection Agency
   Cincinnati, Ohio 45268
 13. TYPE OF REPORT AND PERIOD COVERED
   Peer Reviewed Conference Paper - Proceedings
14. SPONSORING AGENCY CODE
     EPA 600/14
 15. SUPPLEMENTARY NOTES
    Project Officer: Chien T. Chen (908) 906-689?
 16. ABSTRACT
         The iron-enhanced dechlorination technology was evaluated under the U.S. Environmental Protection Agency
(EPA) Superfund Innovative Technology Evaluation (SITE) program at a contaminated site in New Jersey...This process was
demonstrated using a pilot-scale, aboveground treatment reactor containing the reactive iron medium. A flow rate of about
0.5 gallons per minute was maintained throughout a 13-week demonstration period; about 60,800 gallons of groundwater were
treated.
        Analytical results for the effluent samples indicated that the iron-enhanced dechlorination process signficantly
reduced the total concentrations of chlorinated VOCs in water treated, and consistently achieved the demonstration effluent
target level of 1 ptg/L for TCE and PCE. During the last two weeks of the demonstration the process did not consistently
achieve the effluent target levels of 2 ^4g/L for VC and 5 ^g/L for eDCE.   Most of the cDCE and VC appears to have formed
through the degradation of PCE and TCE. The incomplete dcchlorination of cDCE and VC in the latter portion of the SITE
demonstration may have resulted from a gradual reduction in the iron's reactive surface area through formation of precipitates.
       Based on information obtained from the SITE demonstration, groundwater remediation costs for an aboveground
reactor using the iron-enhanced  dechlorination process are estimated to be about $91 per 1,000 gallons treated. This cost was
estimated based on data from a pilot-scale system operating  at a flow rale of 0.5 gpm extrapolated to a 30-ycars operational
period.
17.
a. DESCRIPTORS
ion enhanced dichloroethene
tetra-chloroethene
tri-ehlorocthene
cis- 1 ,2-diehloroethene
vinyl chloride
18. DISTRIBUTION STATEMENT
Release to Public
KEY WORDS AND DOCUMENT ANALYSIS
b. IDENTIFIERS/OPEN ENDED TERMS
19. SECURITY CLASS (This Report)
Unclassified
20 . SECURITY CLASS (This page)
Unclassified

e. COSATI Field/Group
21. NO. OF PAGES
22, PRICE

-------