v>EPA
United States
Environmental Protection
Agency
EPA/540/SR-93/501
August 1993
SUPERFUND INNOVATIVE
TECHNOLOGY EVALUATION
Technology Demonstration
Summary
Peroxidation Systems, Inc.
perox-pure™ Chemical
Oxidation Technology*
As part of the Superfund Innovative
Technology Evaluation (SITE) program, the
U.S. Environmental Protection Agency
(EPA) demonstrated the Peroxidation Sys-
tems, Inc. (PSI), perox-pure™ chemical oxi-
dation treatment system. The SITE dem-
onstration was conducted at Lawrence
Livermore National Laboratory (LLNL) Site
300 in Tracy, CA. Over a 3-wk period in
September 1992, about 40,000 gal of
groundwater contaminated with trichloroe-
thene (TCE), tetrachloroethene (PCE), and
other volatile organic compounds (VOC)
was treated in the perox-pure™ system.
The SITE demonstration results showed
that the perox-pure™ system removed TCE
and PCE from contaminated groundwater
at the LLNL site to concentrations below
detection limits. The perox-pure™ system
achieved TCE and PCE removal efficien-
cies greater than 99.7% and 97.1%, re-
spectively. For other VOCs, the system
achieved removal efficiencies of 81.8%,
98.3%, and 93.1% for 1,1,1-trichloroethane
(TCA); 1,1-dichloroethane (DCA); and chlo-
roform; respectively. The treatment sys-
tem effluent met California drinking water
action levels and federal drinking water
maximum contaminant levels (MCL) for
TCE, PCE, TCA, DCA, and chloroform at
the 95% confidence level.
Potential sites for applying this technol-
ogy include Superfund and other hazard-
ous waste sites where groundwater or
other liquid wastes are contaminated with
organic compounds. Economic data indi-
cate that groundwater remediation costs
could range from about $7 to $11/1,000
gal depending on contaminated ground-
water characteristics. Of these costs,
perox-pure™ system direct treatment costs
could range from about $3 to $5/1,000 gal.
This demonstration summary was de-
veloped by EPA's Risk Reduction Engi-
neering Laboratory in Cincinnati, OH, to
announce key findings of the SITE pro-
gram demonstration that is fully docu-
mented in two separate reports (see or-
dering information at back).
Introduction
The SITE program was established in 1986
to accelerate the development, demonstra-
tion, and use of new, innovative technologies
that offer permanent cleanup solutions for
hazardous wastes. One component of the
SITE program is the demonstration program,
which develops reliable engineering, perfor-
Mention of trade names or commercial products does
not constitute endorsement or recommendation for
:• Printed on Recycled Paper
-------
mance, and cost data for innovative treat-
ment technologies. Data devebped for the
SITE demonstration program enable poten-
tial users to evaluate each technology's appli-
cability for a specific waste site.
The SITE demonstration of the perox-pure™
technology was conducted at LLNL Site 300
in Tracy, CA, over a 3-wk period in Septem-
ber 1992. The technology demonstration
had the following primary objectives:
• Determine the ability of the perox-pure™
system to remove VOCs from ground-
water at the LLNL site under different
operating conditions.
• Determine whether treated groundwater
could meet applicable discharge require-
ments at the 95% confidence level.
• Gather information necessary to estimate
treatment costs, including process chemi-
cal dosages and utility requirements.
The secondary objective of the technology
demonstration was to obtain information on
the presence and types of byproducts formed
during treatment.
Technology Description
The perox-pure™ chemical oxidation treat-
ment technology was developed by PSI to
destroy dissolved organic contaminants in
water. The technology uses ultraviolet (UV)
radiation and hydrogen peroxide to oxidize
organic compounds present in water at parts
per million (ppm) levels or less. This treat-
ment technology produces no air emissions
and generates no sludge or spent media that
require further processing, handling, or dis-
posal. Ideally, end products include water,
carbon dioxide, halides (for example, chlo-
ride), and in some cases, organic acids. The
technology uses medium-pressure, mercury-
vapor lamps to generate UV radiation. The
principal oxidants in the system, hydroxyl radi-
cals, are produced by direct photolysis of
hydrogen peroxide at UV wavelengths.
The perox-pure™ chemical oxidation treat-
ment system (Model SSB-30) used for the
SITE technology demonstration was as-
sembled from the following portable, skid-
mounted components: an oxidation unit, a
hydrogen peroxide feed module, an acid feed
module, a base feed module, a UV lamp
drive, and a control panel. The oxidation unit
has six reactors in series, with one 5-kilowatt
(kW) UV lamp in each reactor, and a total
volume of 15 gal. The UV lamp is mounted
inside a UV-transmissive quartz tube in the
center of each reactor so that water flows
around the quartz tube.
A schematic flow diagram of the perox-
pure™ chemical oxidation system is shown in
Figure 1. Contaminated water enters the oxi-
dation unit through a section of pipe contain-
ing a temperature gauge, a flow meter, an
influent sample port, and hydrogen peroxide
and acid injection points. Contaminated water
is dosed with hydrogen peroxide before the
water enters the first reactor; however, a split-
ter can be used to add hydrogen peroxide
before any of the six reactors within the oxi-
dation unit. In some applications, acid is added
to bwer the influent pH and shift the carbonic
acid-bicarbonate-carbonate equilibrium to car-
bonic acid. This equilibrium is important be-
cause carbonate and bicarbonate ions will
scavenge hydroxyl radicals. After chemical
injections, the contaminated water flows
through a static mixer and enters the oxida-
tion unit. Water then flows through the six UV
reactors. Treated water exits the oxidation
unit through a pipe equipped with a tempera-
ture gauge, effluent sample ports, and a base
injection point. Base may be added to the
treated water to adjust the pH to meet dis-
charge requirements.
Circular wipers mounted on the quartz tubes
housing the UV lamps are used periodically
to remove any solids that may have accumu-
lated on the tubes. Solids may accumulate as
a result of metals oxidatbn (such as iron and
manganese), water hardness, or solids pre-
cipitation. Accumulated solids could eventu-
ally coat the tubes, thus reducing treatment
efficiency.
Site Preparation
About 10,000 ft2 of relatively flat ground
surface was used for the perox-pure™ chemi-
cal oxidation system, support equipment and
facilities, and a parking area. A temporary
encbsure covering about one-fourth of the
demonstration area was erected to provide
shelter for the perox-pure™ system during
inclement weather. Some of the laboratory
analyses were conducted onsrte in a field
trailer, which also served as an office for field
personnel and provided shelter and storage
for small equipment and supplies.
Support equipment for the perox-pure™
system demonstration included a cartridge
filtration system to remove suspended solids
from groundwater, storage tanks for untreated
and treated groundwater, an acid feed mod-
ule for untreated groundwater, a base feed
module for treated groundwater, a spiking
solution feed system, a static mixer, two 55-
gal drums for collecting equipment washdown
and decontamination rinsewater, a dumpster,
aforklift, pumps, sampling equipment, health-
and safety-related gear, and a van.
Technology Testing
During the 3-wk demonstration period,
about 40,000 gal of groundwater contami-
nated with VOCs was treated. The principal
groundwater contaminants were TCE and
PCE, which were present at concentratbns
of about 1,000 and 100 micrograms per liter
(u.g/L), respectively. Influent groundwater was
extracted from Wells W-7-O and W-875-08 at
6 gal/min (gpm) and 2 gpm, respectively. The
groundwater was mixed inline using a static
mixer and then pumped into a 7,500-gal blad-
der tank to minimize any variability in influent
characteristbs. The bladder tank also mini-
mized the bss of VOCs to volatilizatbn, and it
provided a fbw rate higher than the ground-
water yield for several runs. Cartridge filters
were used to remove suspended solids greater
than 3u,m from the groundwater before it
entered the bladder tank. Treated groundwa-
ter was stored in two 20,000-gal steel tanks
before being discharged.
During the demonstration, a total of 14
runs were performed to test the technology's
ability to meet the objectives. Table 1 shows
the operating conditbns for each run, includ-
ing the influent pH, the hydrogen peroxide
dose in milligrams per liter (mg/L), and the
fbw rate in gpm. Operating conditbns were
varied during Runs 1 through 8 to determine
the preferred operating conditbns. The influ-
ent was spiked for Runs 9 through 14, and
the results of Runs 4 and 9 were used to
determine the effect of spiking. Runs 10
through 12 were reproducibility runs, and Runs
13 and 14 evaluated quartz tube cleaning.
The principal operating parameters for the
perox-pure™ system, including influent pH,
hydrogen peroxide dose, and flow rate, were
varied during Runs 1 through 6 to observe
treatment system performance under differ-
ent conditions. PSI used quick turnaround
analytbal data from Runs 1 through 6 and its
professional experience to determine the
system's preferred operating conditions, —
conditions under which the concentrations of
effluent VOCs would be reduced below target
levels during spiked groundwater runs.
Influent groundwater for Runs 9 through 14
was spiked with about 200 |j.g/L each of TCA,
DCA, and chloroform. These compounds were
chosen because they are difficult to oxidize
and because they were not present in the
groundwater at high concentrations.
Runs 10 through 12 involved reproducibil-
ity tests. These runs were designed to evalu-
ate the reproducibility of treatment system
performance at the operating conditbns of
Run 3, which were determined by PSI to be
the preferred conditions from Runs 1 through
6.
During Runs 13 and 14, the effectiveness
of quartz tube wipers was evaluated by per-
forming two runs using scaled and clean
quartz tubes.
During the demonstratbn, samples were
collected at the folbwing locations shown in
Figure 1: Reactor 1 influent and Reactors 1,
2, 3, and 6 effluent, as needed. Samples
were analyzed for the following parameters:
VOCs, semivolatile organic compounds
(SVOC), total organic halides (TOX), adsorb-
-------
Groundwater from
W-875-08 and W-7-O
> To Disposal
UVLamp
Reactor
Spiking
Solution
Figure 1. perox-pure™ chemical oxidation treatment system sampling locations.
able organic halides (AOX), total organic car-
bon (TOC), total carbon (TC), and purgeable
organic carbon (POC). In addition, samples
of Reactor 1 influent and Reactor 6 effluent
were collected and analyzed for acute toxicrty
to freshwater organisms. The hydrogen per-
oxide, acid, and base solutions were also
sampled and analyzed to verify concentra-
tions.
Demonstration Results
SITE demonstration results are based on
extensive laboratory analyses under rigorous
quality control procedures and the observa-
tions of the SITE team during system opera-
tion.
VOC Removal Under Different
Operating Conditions
Operating conditions were varied (see Table
1) to determine the ability of the perox-pure™
system to remove VOCs from groundwater at
the LLNL site. Table 2 presents VOC con-
centrations in the influent to Reactor 1 and in
the effluent from Reactors 1, 3, and 6 during
Runs 1 through 9 under steady-state condi-
tions. VOC removal efficiencies were based
on VOC concentrations in the influent to Re-
actor 1 and effluent from the specified reac-
tor.
VOC Removal as a Function of
Influent pH
The influent pH for Runs 1, 2, and 3 was
8.0, 6.5, and 5.0, respectively, whereas the
fbw rate and hydrogen peroxide level were
the same for all three runs. In all three runs,
effluent TCE and PCE concentrations were
well below the target level of 5 p.g/L Based
on TCE and PCE results, the perox-pure™
system performed best in Run 1, when the
influent pH was 8 (the unadjusted pH of
groundwater). In Run 1, the Reactor 1 efflu-
ent had lower levels of TCE and PCE than in
Runs 2 and 3, and it had the same levels of
Table 1. Operating Conditions for perox-pure™ System
Run
Number
Hydrogen Peroxide at
Influent Influent to Reactor 1,
pH mg/L
Raw Groundwater Runs
1 8.0
2 6.5
3 5.0
4 5.0
5 5.0
6 5.0
7 5.0
8
5.0
Spiked Groundwater Runs
9 5.0
10 5.0
11 5.0
12 5.0
13 5.0
14 5.0
Hydrogen Peroxide at
Influent to Reactors 2 to 6, Flow Rate,
mg/L gpm
40
40
40
70
30
240
240
60
70
40
40
40
40
40
25
25
25
50
15
Hydrogen peroxide was added
at influent to Reactor 1 only
50
25
25
25
25
25
10
10
10
10
10
10
40
40
10
10
10
10
10
10
-------
TCE and PCE as the Reactor 6 effluent in
Runs 2 and 3. The Reactor 6 effluent TCA
concentration, however, was lowest in Run 3
at 1.4 ng/L. Because TCA is difficult to oxi-
dize, PSI selected Run 3 as the preferred
operating condition, with an influent pH of 5.0.
VOC Removal as a Function of
Hydrogen Peroxide Level
Hydrogen peroxide levels at influent to Re-
actor 1 for Runs 3, 4, and 5 were 40 mg/L, 70
mg/L, and 30 mg/L, respectively. At Reactors
2 through 6, the hydrogen peroxide levels for
Runs 3, 4, and 5 were 25 mg/L, 50 mg/L,
and 15 mg/L, respectively. The influent pH
and flow rate were the same for these three
inns. Although the Reactor 6 effluent TCE
and PCE concentrations were the same in all
three runs, the data show that Reactor 1
effluent TCE and PCE concentrations were
lowest in runs with the highest and lowest
hydrogen peroxide level (Runs 4 and 5, re-
spectively), and Reactor 1 effluent TCE and
PCE concentrations were highest at the inter-
mediate hydrogen peroxide level (Run 3).
The Reactor 6 effluent TCA concentrations in
Runs 3, 4, and 5 showed no correlation to
hydrogen peroxide level. These data cannot
be explained, and no definite trend can be
identified based on TCE, PCE, and TCA data
in Runs 3, 4, and 5.
Runs 7 and 8 had hydrogen peroxide lev-
els at Reactor 1 of 240 mg/L and 60 mg/L,
respectively, and flow rate and influent pH
were the same for both runs. A comparison
of TCE and PCE levels shows that both TCE
and PCE concentrations in Reactor 1 effluent
were higher in Run 7 than in Run 8. Effluent
TCA levels at Reactors 1, 3, and 6 were
about the same in both runs. Higher Reactor
1 effluent TCE level in Run 7 may be attrib-
uted to higher influent TCE level in that run.
Reactor 1 effluent TCE levels correspond to
99.5% and 99.9% TCE removal in Runs 7
and 8. Similarly, Reactor 1 effluent PCE lev-
els correspond to 92.9% and 99.2% PCE
removal in Runs 7 and 8. These data seem
to indicate that higher doses of hydrogen
peroxide may have scavenged hydroxyl radi-
cals or excess hydrogen peroxide reduced
UV transmrttance through water, which re-
sulted in lower removal efficiencies for Run 7
than those for Run 8.
VOC Removal as a Function of the
Method Used to Add Hydrogen
Peroxide
Runs 4 and 6 were performed at the same
flow rate and influent pH. In these runs, the
same total amount of hydrogen peroxide was
added to the contaminated groundwater. In
Run 4, however, hydrogen peroxide was
added at multiple points in the system using
the splitter, whereas in Run 6, all hydrogen
peroxide was added at the influent to the
system. Based on a comparison of TCE and
PCE levels in Runs 4 and 6, the effect of
adding hydrogen peroxide at multiple points
in the perox-pure™ system cannot be evalu-
ated, because in both runs, effluent TCE and
PCE levels were below the detection limit of
1.0 (ig/L. However, TCA levels in Reactors 1,
3, and 6 were less in Run 4 than in Run 6.
Based on TCA data, adding hydrogen perox-
ide at multiple points in the perox-pure™ sys-
tem appears to enhance the system's perfor-
mance.
VOC Removal as a Function of Flow
Rate
The fbw rates for Runs 6 and 7 were 10
gpm and 40 gpm, respectively, and hydrogen
peroxide level and influent pH were the same
for both runs. A comparison of TCE, PCE,
and TCA levels in Runs 6 and 7 shows that
the effluent concentrations of these three
VOCs were higher in Run 7 than in Run 6.
These observations are consistent with the
operating conditions, because contaminated
groundwater had a much longer UV expo-
Table 2. VOC Concentrations for Runs 1 through 9
Run
1
2
3
4
5
6
7
8
9
Contaminant
TCE
PCE
TCA
TCE
PCE
TCA
TCE
PCE
TCA
TCE
PCE
TCA
TCE
PCE
TCA
TCE
PCE
TCA
TCE
PCE
TCA
TCE
PCE
TCA
TCE
PCE
TCA
DCA
Chloroform
Influent, iig/L
Reactor 1
1300
150
17?
1300
100
13+
1100
130
9+
980
110
&
910
100
7*
990
120
8t
1100
85
6t
890
71
6t
690
63
110
160
150
Reactor 1
0.5*
0.5*
11.5
9.6
3.4
8.9
1.2
1.6
6.8
0.5*
0.5*
4.4
0.5*
0.5*
5.3
0.5*
0.5*
7.0
5.3
6.0
5.7
0.5*
0.6*
5.3
3.2*
3.2*
84
23
89
Effluent, \ig/L
Reactor 3
0.5*
0.5*
9.6
0.5*
0.5*
5.9*
0.5*
0.5*
3.7
0.5*
0.5*
2.5
0.5*
0.5*
3.7
0.5*
0.5*
3.7
0.5*
0.5*
3.9
0.5*
0.5*
4.3
2.4*
2.4*
47
3.0*
37
Reactor 6
0.5*
0.5*
6.7
0.5*
0.5*
3.1
0.5*
0.5*
1.4
0.5*
0.5*
1.8
0.5*
0.5*
2.1
0.5*
0.5*
3.0
0.5*
0.5*
3.9
0.5*
0.5*
4.0
2.4*
2.4*
7.8*
2.9*
14
* More than one of the four replicate samples had analyte concentrations at nondetectable levels.
For these replicate samples, one-half the detection limit was used as the estimated concentra-
tion. If more than one replicable sample had concentrations at nondetectable levels, 0.5, 0.4, 0.6,
and 0.4 times the detection limit were used as estimated concentrations for the first, second,
third, and fourth replicate samples, respectively.
t The reported concentration is from analysis of one sample by the gas chromatography/mass
spectroscopy (GC/MS) method. All other reported concentrations are the mean concentrations of
four replicate samples analyzed by the GC method.
-------
sure time in Run 6 than in Run 7. UV expo-
sure times were 1.5 and 0.4 minutes in Runs
6 and 7, respectively.
VOC Removal as a Function of
Influent Groundwater Spiking
A comparison of the perox-pure™ system's
performance in treating spiked groundwater
(Run 9) and unspiked groundwater (Run 4)
shows that TCE and PCE levels in treated
groundwater were higher in spiked ground-
water than in unspiked groundwater. These
data suggest that spiking compounds (TCA,
DCA, and chloroform) affected the perox-
pure™ system's performance in removing TCE
and PCE, perhaps because they present an
additional oxidant demand. TCE and PCE
concentrations presented in Table 2, how-
ever, are estimated concentrations. The con-
centrations in Run 9 were estimated higher
than in Run 4 for two reasons: (1) the detec-
tion limit for TCE and PCE in Run 9 was 5
Hg/L and in Run 4 was 1 jag/L, and (2) TCE
and PCE were present at nondetectable lev-
els in both runs. Therefore, the estimated
data are inconclusive with regard to the effect
of spiking compounds on the removal of TCE
and PCE.
VOC Removal as a Function of
Quartz Tube Cleaning
Table 3 presents VOC concentrations in
Runs 12,13, and 14, which were conducted
to evaluate quartz tube cleaning. In Run 12,
quartz tubes from the previous demonstration
runs were used. In Run 13, scaled quartz
tubes were used. The tubes had been ex-
posed to an environment that encouraged
scaling, but they had not been maintained
with cleaners or wipers. In Run 14, quartz
tubes that had been maintained by cleaners
or wipers were used.
A comparison of removal efficiencies for
TCE after Reactors 1 and 2 shows that TCE
removal efficiencies were about the same in
all runs. PCE removal efficiencies were about
3% to 4% less in Run 13 than in Runs 12 or
14. In general, removal efficiencies for TCA,
DCA, and chloroform were less in Run 13
than in Run 14; this indicates that periodic
cleaning of quartz tubes by wipers is required
to maintain the perox-pure™ system's perfor-
mance. Without such cleaning, the removal
efficiencies will likely decrease in an aqueous
environment that would cause scaling of quartz
tubes. For example, after Reactor 2, chloro-
form removal efficiency in Run 13 was 53.6%,
compared with 61.3% removal efficiency in
Run 14. Because the quartz tubes used in
Run 12 had little coating, removal efficiencies
in Run 12 were expected to be higher than in
Run 13. The demonstration did not confirm
this for all VOCs, however. For example, Run
12 TCA removal efficiencies were less than
Run 13 TCA removal efficiencies; this incon-
sistency cannot be explained.
Reproducibility of Treatment
System Performance
Based on the results from Runs 1 through
6, PSI selected Run 3 operating conditions
as the preferred operating conditions for spiked
groundwater. As a result, Runs 10 through 14
were performed at Run 3 conditions.
VOC removal efficiencies in reproducibility
runs (Runs 10, 11, and 12) are plotted in
Figure 2. Figure 2 shows that for TCE and
PCE, which are relatively easy to oxidize,
most of the removal occurred in Reactor 1,
leaving only trace quantities of TCE and PCE
to be removed in the rest of the perox-pure™
system. For TCA, DCA, and chloroform, how-
ever, which are difficult to oxidize, consider-
able removal occurred beyond Reactor 1.
During the three reproducibility runs, average
removal efficiencies for TCE, PCE, TCA, DCA,
and chloroform after Reactor 1 were 99.5%,
95.9%, 17.4%, 67.0%, and 41.3%, respec-
tively. After Reactor 6, average overall re-
moval efficiencies for TCA, DCA, and chloro-
form increased to 81.8%, 98.3%, and 93.1%,
respectively. In general, overall removal effi-
ciencies of the perox-pure™ system were
reproducible for all VOCs. For certain com-
pounds, removal efficiencies after Reactor 1
were quite variable (for example, chloroform
removal efficiencies ranged from 27.4% to
56.3%).
Compliance with Applicable
Discharge Requirements
Figure 3 compares the 95% upper confi-
dence limits (UCL) of effluent VOC concen-
trations with target levels in reproducibility
runs. For this project, the target level for a
given VOC was set at the most stringent limit
in cases where the VOC has multiple regula-
tory limits. For all VOCs but chloroform, the
most stringent limit is the California drinking
water action level. For chloroform, the most
stringent limit is the MCL specified in the Safe
Drinking Water Act. Figure 3 shows that perox-
pure™ system effluent met the target levels at
the 95% confidence level in all three repro-
ducibility runs; this indicates that the system
performance was reproducible.
Byproducts Formed During
Treatment
GC/MS analysis of influent and effluent
samples for VOCs indicated that new target
compounds or tentatively identified com-
pounds (TIC) were not formed during the
treatment. GC/MS analysis of influent and
effluent samples for SVOCs showed that tar-
get SVOCs were not present at detectable
levels. Several unknown TICs were, how-
ever, present in both the influent and effluent
samples.
During Runs 10, 11, and 12, bioassay
tests were performed to evaluate the acute
toxicity of influent to and effluent from the
perox-pure™ systems. Two freshwater test
organisms, a water flea (ceriodaphnia dubia)
and a fathead minnow (pimephates promelas),
were used in the bbassay tests. Toxicity was
measured as the lethal concentration at which
50% of the organisms died (LCJ, and was
expressed as the percent of effluent (or influ-
ent) in the test water. One influent and one
effluent sample were tested in each run. One
control sample was also tested to evaluate
the toxicity associated with hydrogen perox-
ide residual present in the effluent. The con-
trol sample had about 10.5 mg/L of hydrogen
peroxide (average effluent residual in Runs
10, 11, and 12), and had characteristics (al-
kalinity, hardness, and pH) similar to those of
effluent in Runs 10, 11, and 12.
In general, the influent was not found to be
acutely toxic to either test organism. The
effluent was found to be acutely toxic to both
test organisms. The influent LC50 values for
both organisms indicated that in the undiluted
influent sample more than 50% of the organ-
isms survived. LCW values for the water flea,
however, were estimated to be 35%, 13%,
and 26% of effluent in Runs 10, 11, and 12,
respectively; and LCSO values for the fathead
minnow were estimated to be 65% and 71%
of effluent in Runs 10 and 11, respectively. In
Run 12, more than 50% of the fathead min-
nows survived in the undiluted effluent. The
LC,, value for the water flea was estimated to
be 17.7% in the control sample, indicating
that the sample contained hydrogen peroxide
at a concentration that was acutely toxic to
water fleas; however, more than 50% of the
fathead minnows survived in the undiluted
control sample indicating hydrogen peroxide
was not acutely toxic to fathead minnows at
the concentration of 10.5 mg/L. This observa-
tion however, is not entirely consistent with
observations made by the Department of En-
vironmental Protection, State of Connecticut
(CDEP). The CDEP Water Toxics Section of
Water Management Division reports LCM value
of 18.2 mg/L of hydrogen peroxide with 95%
confidence limits of 10 mg/L and 25 mg/L.
Comparison of the LCSO value of the control
sample with LC!0 values of effluent samples
for water fleas indicates the toxicity associ-
ated with the effluent samples is probably
due to hydrogen peroxide residual in the ef-
fluent. No conclusion can, however, be drawn
on the effluent toxicity to fathead minnows
because the control sample toxicity results
from the SITE demonstration data are not
entirely consistent with the data collected by
CDEP.
-------
o
I
100
80
60
40
20
Run 10
100
80
I 60
§ 40
I
20
0
100
_ 80
2
o
I 60
QC
40
20
TCE
PCE
TCA
Run 11
TCE
PCE
TCA
Run 12
DCA
Chloroform
DCA Chloroform
TCE
PCE
TCA
DCA Chloroform
G
Reactor 1
Reactor 3 mi$m Reactor 6
Figure 2. VOC removal efficiencies in reproducibility runs (influent pH=5.0; hydrogen peroxide level
at Reactor 1= 40 mg/L; hydrogren peroxide level at Reactors 2 through 6 = 25 mg/L; flow
rate = 10 gpm.
Miscellaneous Parameters
The technology demonstration also evalu-
ated analytical results of several parameters
other than VOCs, including TOX, AOX, TC,
TOG, and POC.
TOX and AOX were analyzed upon re-
quest by the German Federal Ministry of
Research and Technology, under a U.S.-
German bilateral technology transfer program.
Average Reactor 1 influent TOX and AOX
levels were 800 u,g/L and 730 ng/L, respec-
tively. The perox-pure™ system achieved TOX
removal efficiencies that ranged from 93% to
99% and AOX removal efficiencies that ranged
from 95% to 99%.
Runs 10,11, and 12 Reactor 1 influent and
Reactor 6 effluent samples were analyzed for
TC, TOC, and POC. Average TC concentra-
tions in the influent and effluent were 75 mg/L
and 55 mg/L, respectively. The decrease in
TC concentration in the perox-pure™ system
may be due to the bss of dissolved carbon
dioxide that occurred as a result of the turbu-
lent movement of contaminated groundwater
in the perox-pure™ system. TOC decreased
about 38% during treatment, which corre-
sponds to the amount of organic carbon that
was converted to inorganic carbon (carbon
dioxide) during treatment. Organic carbon may
have originated from the VOCs or from some
other compounds present in groundwater. Ef-
fluent POC concentration was about 0.02
mg/L, which is bebw the reporting limit of
0.035 mg/L. POC concentration data show
that the average POC removal efficiency was
about 93%. Assuming that the majority of
organic carbon associated with VOCs could
be measured as POC, these data show that
about 93% of the volatile organic carbon was
converted to either carbon dioxide or
nonpurgeable organic carbon.
Estimated Treatment Costs
With the use of information obtained from
the SITE demonstration, an economic analy-
sis examined 12 separate cost categories for
perox-pure™ systems treating about 260 mil-
lion gal of contaminated groundwater at a
Superfund site. This analysis examined two
cases based on groundwater characteristics.
In Case 1, the groundwater was assumed to
have five contaminants, of which two are
easy to oxidize (TCE and PCE) and the
remaining three are difficult to oxidize (chloro-
form, DCA, and TCA). In Case 2, the ground-
water was assumed to have only two con-
taminants that are easy to oxidize (TCE and
PCE). For each case, costs for three different
fbw rates (10, 50, and 100 gpm) were esti-
mated. Costs for the 50-gpm flow rate sce-
nario for each case are summarized below.
For Case 1, capital costs are estimated to
be about $906,000 of which the perox-pure™
system direct capital cost is $185,000. An-
nual operation and maintenance (O&M) costs
are estimated to be about $188,000 of which
perox-pure™ system direct O&M costs are
$125,000. Groundwater remediation costs to
treat 1,000 gal of contaminated water are
estimated to be about $11 of which perox-
pure™ system direct treatment costs are $5.
For Case 2, capital costs are estimated to
be about $776,000 of which the perox-pure™
system direct capital cost is $55,000. Annual
O&M costs are estimated to be about
-------
1,000
^j 100
5
10
TL=200
TL=100
32
34
Target Level
_ _ _TL=5_ _ _
2.9 3.2 3.1 2.9 3.2 3.1
18
33
15
3.7
13
TCE
PCE
TCA
DCA
Chloroform
Run 10
11
Run
Figure 3. Comparison of 95% UCLs for effluent VOC concentrations with target levels in reproducibility runs (influent pH = 5.0; hydrogen peroxide level
at Reactor 1 =40 mg/L; hydrogen peroxide level at Reactors 2 through 6 =25 mg/L; flow rate = 10 gpm.
$111,000 of which perox-pure™ system di-
rect O&M costs are $61,000. Groundwater
remediation costs to treat 1,000 gal of con-
taminated water are estimated to be about $7
of which perox-pure™ system direct treat-
ment costs are $3.
Conclusions
The folbwing conclusions about the PSI
perox-pure™ technology are based on the
results of the SITE demonstration:
For the spiked groundwater, PSI deter-
mined the following preferred operating con-
ditions: (1) influent hydrogen peroxide level of
40 mg/L, (2) hydrogen peroxide level of 25
mg/L in the influent to Reactors 2 through 6,
(3) influent pH of 5.0, and (4) flow rate of 10
gpm. At these conditions, the effluent TCE,
PCE, and DCA levels were generally below
detection limit (5 fig/L) and TCA levels ranged
from 15 to 30 ng/L. The average removal
efficiencies for TCE, PCE, chloroform, DCA,
and TCA were about 99.7%, 97.1%, 93.1%,
98.3%, and 81.8%, respectively.
For the unspiked groundwater, the effluent
TCE and PCE levels were generally bebw
the detection limit (1 jig/L), with correspond-
ing removal efficiencies of about 99.9% and
99.7%. The effluent TCA levels ranged from
1.4 to 6.7 |ig/L with removal efficiencies rang-
ing from 35% to 84%.
The perox-pure™ system effluent met Cali-
fornia drinking water action levels and federal
drinking water MCLs for TCE, PCE, chloro-
form, DCA, and TCA at the 95% confidence
level.
The quartz tube wipers were effective in
keeping the tubes clean and appeared to
reduce the adverse effect scaling has on
contaminant removal efficiencies.
TOX removal efficiencies ranged from 93%
to 99%. AOX removal efficiencies ranged
from 95% to 99%.
For spiked groundwater, during reproduc-
ibility runs, the system achieved average re-
moval efficiencies of 38% and greater than
93% for TOC and POC, respectively.
The temperature of groundwater increased
at a rate of 12°F/min of UV exposure in the
perox-pure™ system. Since the oxidation unit
is exposed to the surrounding environment,
the temperature increase may vary depend-
ing upon the ambient temperature or other
atmospheric conditions.
Table 3. Percent Removal Efficiencies for VOCs in Quartz Tube Cleaner Runs*
TCE
PCE
TCA
DCA
Run
12
13
14
Reactor
1
99.7
99.4
99.4
Reactor
2
99.7
99.4
99.4
Reactor
1
96.7
92.1
95.5
Reactor
2
96.7
92.1
95.5
Chloroform
Reactor
1
13.9
42.0
48.5
Reactor
2
46.9
57.5
63.5
Reactor
1
44.2
82.4
84.2
Reactor
2
96.5
97.0
96.9
Reactor
1
56.3
25.4
37.2
Reactor
2
76.7
53.6
61.3
Influent pH = 5.0 ; hydrogen peroxide level at Reactor 1 = 140 mg/L; hydrogen peroxide level at Reactors 2 through 6 = 25 mg/L; and flow rate = 10
gpm
•kv.S. GOVERNMENT PRINTING OFFICE: t993 - 750-071/8003*
-------
The EPA Project Manager, Norma Lewis, is with the Risk Reduction Engineer-
ing Laboratory, Cincinnati, OH 45268 (see below)
The complete report, entitled "Technology Evaluation Report: SITE Program
Demonstration of the Peroxidation Systems, Inc., perox-pure™ Chemical
Oxidation Technology," (Order No. PB93-213528AS; Cost: $27.00,
subject to change) discusses the results of the SITE demonstration and
will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA22161
Telephone: 703-487-4650
A related report, entitled "Applications Analysis Report: SITE Program Demon-
stration of the Peroxidation Systems, Inc., perox-pure™ Chemical Oxidation
Technology," discusses the applications of the demonstrated technology.
The EPA Project Manager can be contacted at:
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268
Official Business
Penalty for Private Use
$300
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
EPA/540/SR-93/501
------- |