United States
Environmental Protection
Agency
EPA/540/SR-93/520
March 1995
&EPA
SUPERFUND INNOVATIVE
TECHNOLOGY EVALUATION
Technology Demonstration
Summary
Magnum Water Technology
CAV-OX® Cavitation Oxidation
Process
As part of the Superfund Innovative
Technology Evaluation (SITE) Program,
the U.S. Environmental Protection
Agency (EPA) demonstrated the Mag-
num Water Technology (Magnum) CAV-
OX® cavitation oxidation process at
Edwards Air Force Base (Edwards) Site
16 in California in March 1993. The CAV-
OX18 technology treated about 8,500 gal
of groundwater contaminated primarily
with trichloroethene (TCE) and benzene.
Demonstration results showed that
the CAV-OX® process was capable of
achieving average removal efficiencies
from contaminated groundwater of over
99.9% for both TCE and benzene. Efflu-
ent from one process configuration met
the 95% confidence level for State of
California drinking water action levels
and federal maximum contaminant lev-
els (MCL) for TCE and benzene.
This technology may be applied at
Superfund and other hazardous waste
sites where groundwater or other aque-
ous wastes are contaminated with or-
ganic compounds. Economic data
indicate that the costs of remediating
groundwater using the CAV-OX® pro-
cess range from $13 to $31 per 1,000
gal. These amounts include capital and
operation and maintenance costs. Ac-
tual site remediation costs will, how-
ever, depend on the CAV-OX® process
configuration and waste characteristics.
This Summary was developed by
EPA's Risk Reduction Engineering
Laboratory In Cincinnati, OH, to an-
nounce key findings of the SITE Pro-
gram demonstration that is fully
documented in separate reports (see
ordering information at back).
Introduction
The SITE Program was established in
1986 to accelerate the development, dem-
onstration, and use of new, innovative
technologies that offer permanent cleanup
solutions for hazardous wastes. One com-
ponent of the SITE Program is the Dem-
onstration Program, which develops
reliable engineering, performance, and cost
data for innovative treatment technologies.
Data developed for the SITE Demonstra-
tion Program enables potential users to
evaluate each technology's applicability to
a specific waste site.
The SITE demonstration of the CAV-
OX® technology was conducted at
Edwards Site 16 over a 4-wk period.
Edwards is a 310,000-acre U.S. Air Force
-------
base about 75 mi northeast of Los Ange-
les. Edwards Site 16 occupies about 12
acres in an area that includes taxiways,
hangars, office buildings, outside storage
buildings, and open fields. The site con-
tamination is attributed to a military-grade
jet fuel (JP-4) release from a leaking fuel
transfer line that runs along the western
boundary of Site 16. The fuel leak, esti-
mated to be between 250,000 and 300,000
gal, began about September 1983.
In November 1984, the Lahontan Re-
gion of the State of California Regional
Water Quality Control Board issued Clean-
up and Abatement Order No. 84-10 for
Site 16. This order was followed by reme-
dial actions including pipeline replacement
and monitoring well installation. Further in-
vestigation included a feasibility study, re-
medial action, and remedial design to clean
up the fuel release. The resulting remedial
action combined a recovery well system,
a recovery fuel storage system, and a
fuel-water separator system. This com-
bined system collected fuel for later re-
use. Potentially contaminated water was
then routed to the base sanitary sewer
system. Dissolved TCE was, however, later
detected in the discharge water. Because
water containing TCE could not be dis-
charged to the sanitary sewer system, the
recovery well system ceased operation in
December 1987. A site assessment was
completed in June 1989 under the De-
partment of Defense Installation Restora-
tion Program. A pilot plant study completed
for the site proposed a treatment system
that included an ultraviolet-oxidation unit
and air stripping to treat contaminated
groundwater from Site 16 recovery wells.
In February 1992, Magnum responded
to EPA's annual solicitation for proposals
to participate in the SITE Program. At that
time, Magnum discussed the possibility of
demonstrating the CAV-OX® technology
to treat contaminated groundwater at
Edwards. EPA subsequently accepted the
CAV-OX® technology into the SITE Dem-
onstration Program. The CAV-OX® tech-
nology demonstration had the following
primary objectives:
• Determine TCE and benzene, tolu-
ene, ethylbenzene, and xylene (BTEX)
removal efficiencies in the treatment
process under different operating con-
ditions
• Determine whether TCE and BTEX
concentrations in treated groundwa-
ter meet the 95% confidence level for
applicable discharge limits to the sani-
tary sewer
• Compare TCE and BTEX removal ef-
ficiencies among three treatment pro-
cess configurations
Secondary objectives for the demon-
stration were:
• Collect information needed to estimate
treatment costs, including those for
process chemicals and utility require-
ments
• Assess the presence of degradation
by-products in the treated water
• Collect characterization data for both
influent and effluent streams
Technology Description
Magnum developed the CAV-OX® tech-
nology to destroy organic contaminants
dissolved in water. The technology uses
hydrodynamic cavitation, ultraviolet (UV)
radiation, and hydrogen peroxide to oxi-
dize organic compounds present in water
at parts per million (ppm) levels or less.
The technology uses a cavitation cham-
ber to generate hydrodynamic cavitation,
which occurs when a liquid undergoes a
dynamic pressure reduction while under
constant temperature. The pressure reduc-
tion causes bubbles to explosively develop,
grow, and then collapse. Acoustic cavita-
tion decomposes water into extremely re-
active hydrogen atoms and hydroxyl
radicals. Based on the assumption that
hydrodynamic cavitation produces similar
effects to those of acoustic cavitation, Mag-
num combines hydrodynamic cavitation
with UV radiation and hydrogen peroxide.
This combination generates powerful oxi-
dants that react with organic contaminants.
End products of oxidation include water,
carbon dioxide, halides (for example, chlo-
ride), and in some cases, organic acids.
Each portable, skid-mounted CAV-OX®
technology configuration consists of the
following main components: cavitation
pump, cavitation chamber, UV reactor, and
control panel unit. Contaminated water is
pumped through the cavitation chamber,
and then enters either the UV reactor or is
recycled through the cavitation chamber.
The selected hydraulic retention time (or
flow rate) determines the recycle rate. In
permanent applications of the technology,
hydrogen peroxide is usually injected di-
rectly into the water downstream from the
cavitation chamber.
For the demonstration, groundwater
from Edwards Site 16 monitoring wells
was first pumped into an equalization tank.
From this tank, the water was pumped to
an influent holding tank, where hydrogen
peroxide was added. The water was then
pumped through a flow indicator, past an
influent sampling port, through the cavita-
tion chamber, and then to a UV reactor.
Three configurations of the CAV-OX®
technology were demonstrated. The CAV-
OX® I low-energy process, operating at
360 watts, contained one UV reactor with
six 60-watt UV lamps. The CAV-OX® II
process consisted of two UV reactors with
one UV lamp each, and operated at 5 or
10 kilowatts (kW). The low-energy and
high-energy process configurations oper-
ated simultaneously. Figure 1 illustrates
the main and ancillary components of the
CAV-OX® process configurations.
Site Preparation
The demonstration required about 2,000
ft2 of ground surface for the CAV-OX®
technology and support equipment and
facilities. Support equipment for the dem-
onstration included a 7,500-gal bladder
tank for untreated groundwater, a 21,000-
gal storage tank for treated groundwater,
a solid waste dumpster for disposal of
nonhazardous wastes, 55-gal drums for
storage of decontamination rinse water, a
forklift for unloading and loading support
equipment, three submersible positive dis-
placement pumps with gasoline powered
generators, sampling and analytical equip-
ment, and health and safety equipment.
Utilities required for the demonstration
included water, electricity, and telephone
service. Water was required for equipment
and personnel decontamination, field and
mobile laboratory use, and drinking. Elec-
tricity was required to power the CAV-OX®
process, a field trailer, and mobile labora-
tory equipment. Two cellular telephones
located in the office and laboratory trailer
were used to schedule deliveries and to
order equipment, parts, reagents, and
other chemical supplies.
Technology Demonstration
The technology was demonstrated at
Edwards Site 16 over a 4-wk period in
March 1993. During the demonstration, the
CAV-OX® process treated about 8,500 gal
of groundwater contaminated with volatile
organic compounds (VOC). Groundwater
was pumped from three wells into a 7,500-
gal bladder tank to minimize variability in
influent characteristics. Treated groundwa-
ter was stored in a 21,000-gal steel tank
before discharge.
Principal groundwater contaminants in-
cluded TCE and benzene, which were
present at concentrations of up to 2,100
and 500 micrograms per liter (ng/L), re-
2
-------
High-energy
UV reactor
Figure 1. The CAV-OX® technology as demonstrated.
spectively. Other VOCs, such as carbon
tetrachloride, chloroform, 1,1-dichloro-
ethane, 1,2-dichloroethane, 1,1-dichloro-
ethene, trans-1,2-dichloroethene,
tetrachloroethene, 1,1,1-trichloroethane,
and 1,1,2-trichloroethane, were present in
groundwater at concentrations less than
300 ng/L.
During the demonstration, samples were
collected at three locations. Samples of
untreated groundwater were collected from
the feed line exiting the influent holding
tank. Samples of treated groundwater were
collected from the effluent lines of both
the CAV-OX® I and CAV-OX® II configura-
tions. Influent and effluent samples were
analyzed for the following parameters:
TCE, BTEX, VOCs, hydrogen peroxide,
total organic carbon (TOC), total carbon
(TC), purgeable organic carbon (POC),
metals, alkalinity, hardness, total recover-
able petroleum hydrocarbons (TRPH),
semivolatile organic compounds (SVOC),
and turbidity. Measurements were made
of pH, temperature, flow rate, and specific
conductance. Bioassay tests were also
conducted to determine the toxicity of in-
fluent and effluent to freshwater organ-
isms. Samples were also analyzed to verify
hydrogen peroxide concentrations in the
influent holding tank.
The principal operating parameters for
the Magnum CAV-OX® process include
hydrogen peroxide concentration, UV lamp
output, and flow rate. These parameters
were varied to observe treatment process
performance under different operating con-
ditions. Initial operating conditions for the
demonstration were based on the profes-
sional judgment and experience of Mag-
num personnel, and on groundwater
characterizations performed by Edwards
in 1990 and 1991. Preferred operating con-
ditions were determined as conditions that
1) provided the greatest reduction of VOC
concentrations, and 2) reduced VOC con-
centrations below target levels.
Operating information, such as power
consumption rates, was recorded during
the demonstration to estimate costs. Mag-
num provided operating and maintenance
cost information for the technology.
Demonstration Results
Demonstration results were based on
extensive laboratory analyses under rigor-
ous quality control procedures, as well as
observations made by SITE personnel dur-
ing the CAV-OX® process demonstration.
The following sections summarize the pri-
mary and secondary objective results of
the CAV-OX® demonstration.
Summary of Primary Objective
Results
Throughout the demonstration, only TCE
and benzene were detected in the influent
at concentrations well above target levels.
Toluene, ethylbenzene, and xylenes were
present in the influent at relatively low
concentrations when compared with TCE
and benzene. Therefore, this summary of
the CAV-OX® technology's effectiveness
pertains only to TCE and benzene.
Table 1 compares TCE and benzene
removals among the three treatment pro-
cess configurations. The major differences
among the three configurations are UV
lamp output and flow rate. As expected,
contaminant removal efficiency increased
with higher UV lamp output and with lower
flow rates. The choice of a particular pro-
cess configuration depends, however, on
individual site-specific characteristics.
Under the preferred conditions — a flow
rate of 0.6 gal per minute (gpm) and a
hydrogen peroxide concentration of 23.4
milligrams per liter (mg/L) — effluent from
the CAV-OX® I process configuration met
State of California drinking water action
levels and federal MCLs for benzene and
TCE at the 95% confidence level. If a
greater flow rate is required, the CAV-OX®
3
-------
Table 1. Contaminant Removal
Run Influent Hydrogen
Peroxide
TCE Benzene Concentration Flow
(fig/L) (tig/L) (mg/L) (gpm)
CAV-OX* I
CAV-OX* II
Effluent
Flow (gpm)
Effluent
TCE
(ng/L)
Benzene
(ng/p
TCE
(»g/L)
Benzene
(m/L)
5-kW
10-kW
5-kW
10-kW
5-kW
10-kW
Low Hydrogen Peroxide Concentration
1 1,980 263
0
0.6
1,570
315
2.0
2.0
275
31.0
189
99.1
2 2,000 265
0.4
1.5
1,500
300
1.6
1.6
169
17.3
150
86.0
3 1,780 243
0.4
1.5
1,790
301
3.9
4.0
632
270
234
174
Medium Hydrogen Peroxide Concentration
4 1,930 285
4.9
1.5
657
36.9
4.0
3.9
278
41.5
35.6
4.44
5 1,560 333
4.9
1.5
617
65.9
3.9
3.4
287
96.3
44.1
18.0
6 1600 355
5.9
0.5
60.6
238
1.5
1.4
9.56
21.1
3.18
4.26
7 1,980 503
5.9
0.7
270
19.7
1.9
2.0
54.7
39.6
41.3
6.41
8 1,850 283
6.0
0.7
268
9.90
1.9
1.9
38.9
20.8
4.18
2.97
9 1,800 433
6.1
1.5
789
67.3
4.0
4.0
262
26.3
29.4
3.21
High Hydrogen Peroxide Concentration
10 1,680 252
23.4
0.6
2.71
0.18
2.0
2.1
8.46
7.84
1.89
2.05
11 1,480 240
33.1
0.5
4.07
0.38
1.6
1.5
8.80
18.2
2.07
4.52
12 1,530 250
48.3
0.6
12.1
0.12
1.4
1.4
5.61
7.97
1.25
1.14
Vendor Selected Conditions
13' 2,000 500
1.8
NA *
NA
NA
1.9
1.9
427
57.9
328
20
14*" 1,680 410
9.1
NA
NA
NA
NA
7.9
NA
257
NA
39.1
15*"* 2,000 468
NA
NA
NA
NA
NA
0.8
NA
185«
NA
43.4«
Notes: " CAV-OX* I not operated.
f NA - not applicable.
CA V-OX" I and 5-kW CA V-OX* II not operated.
Influent pH adjusted.
CA V-OX8II high-energy process operated at 2.5 kW.
Sample collected after cavitation prior to the UV reactor.
Values for TCE and benzene represent the 95% confidence limit.
t
s
ss
II 5-kW or 10-kW process may be more
suitable. Average effluent contaminant con-
centrations from the CAV-OX® II process
configurations, operating under the pre-
ferred conditions, met State of California
drinking water action levels and federal
drinking water MCLs for TCE and ben-
zene. However, because of data variabil-
ity, many of these effluents did not achieve
the 95% confidence level for TCE and
benzene. Since the demonstration evalu-
ated numerous operating conditions, not
all of them produced effluent that met
these discharge limits.
Figure 2 compares removal efficiencies
of the three CAV-OX® process configura-
tions for TCE. Figure 3 compares removal
efficiencies of the three CAV-OX® process
configurations for benzene.
Summary of Secondary
Objective Results
The technology demonstration was also
designed to evaluate analytical results of
several other parameters and collect in-
formation to determine the costs of apply-
ing the technology. Results of demonstra-
tion procedures designed to satisfy sec-
ondary objectives are summarized below.
Process Chemical and Utility
Requirements
A secondary objective of the demon-
stration was to estimate treatment costs
by determining process chemical and util-
ity requirements. Hydrogen peroxide is the
only process chemical used in the CAV-
OX® process. During the various demon-
stration runs, the process used between 0
and 48.3 mg of hydrogen peroxide per L
of influent water. During the demonstra-
tion, hydrogen peroxide was added to the
influent holding tank in an attempt to
achieve optimal concentration of hydro-
gen peroxide.
The demonstration results show that re-
moval efficiency increased with the addi-
tion of hydrogen peroxide. Magnum
reports, however, that this increase in effi-
ciency is limited, and that efficiency be-
gins to decrease at some point specific to
the influent water characteristics and op-
erating parameters used. Table 1 shows
the effects of varying hydrogen peroxide
concentrations on contaminant removal
efficiency.
Runs 13 and 14 were conducted under
conditions chosen by Magnum. For these
runs, Magnum chose not to operate the
CAV-OX® I low-energy process. Run 13
repeated the conditions of Run 11 without
hydrogen peroxide added. The results of
these runs show that although some con-
taminant removal does occur without hy-
drogen peroxide, this additive is vital to
achieving high removal efficiencies. Run
14 was conducted with a hydrogen perox-
ide concentration of 9.1 mg/L and a flow
rate of 7.9 gpm. Results of this run showed
that increased hydrogen peroxide concen-
trations did not compensate for the in-
creased process flow rate (and consequent
decreased retention time). However, be-
cause the influent's pH was lowered, it is
difficult to draw definitive conclusions
based on this run.
Electricity and water are required for
the CAV-OX® process. Electricity demand
4
-------
&
§
G
€
ft
100 -¦
90 --
80 --
70 --
60 --
50 --
40 --
30 +
20
1 0
0
3 4 5 6
10 11 12 13 14 15
¦1
CAV-OX* 1
26.6
30.0
12.7
71.4
61.7
96.4
87.1
87.8
60.6
99.9
99.9
99.7
NA 1
NA
NA
r~i
5-kW CAV-OX1* II
88.1
94.1
65.8
87.8
85.1
99.6
97.8
98.4
86.2
99.7
99.6
99.8
80.1
NA
0*
~
W-kW CAV-OX* II
98.6
99.2
86.1
98.1
97.1
99.4
99.2
99.3
98.9
99.7
99.2
99.7
97.6
91.6
96.3
Notes: ' CA V-OX" II high-energy process operated at 2.5 kW.
* not applicable.
* Sample collected alter cavitation chamber prior to UV reactor.
Figure 2. Trichloroethene removal efficiency comparison for three operating conditions.
$
S
y
ig
Q>
5
0
1
-------
was 2.2 kW for the CAV-OX® I process,
6.4 kW for the 5-kW CAV-OX® II process,
and 13.0 kW for the 10-kW CAV-OX® II
process. During the demonstration, water
was required in the following approximate
amounts: 10 gallons per day (gpd) of po-
table water for equipment and personal
decontamination; 5 gpd of distilled, deion-
ized water for field laboratory use; and 5
to 10 gpd for drinking water.
Degradation By-Products
Gas chromatography/mass spectrometry
(GC/MS) analysis of influent and effluent
samples for VOCs and SVOCs indicated
that no new target compounds or tenta-
tively identified compounds (TIC) were
formed during the treatment. TICs are com-
pounds whose retention times on GC col-
umns indicate, but do not confirm, the
presence of a given compound. Several
unknown TICs were identified in both the
influent and effluent samples. Acetone and
2-butanone concentrations were generally
higher in the effluent than in the influent.
Additional Groundwater
Characterization
Groundwater characterization was also
performed during the demonstration. This
demonstration included bioassay testing
to determine the toxicity of process influ-
ents and effluents.
Bioassay tests were performed to evalu-
ate the acute toxicity of influent and efflu-
ent from Runs 13, 14, and 15. Two
freshwater test organisms, the fathead min-
now (Pimephales promelas) and the wa-
ter flea (Ceriodaphnia dubia), were used
in the bioassay tests. Toxicity was mea-
sured as the concentration at which 50%
of the organisms died (LCM) and was ex-
pressed as the percent of effluent or influ-
ent in the test water. One influent and one
effluent sample were collected and ana-
lyzed in each run.
Bioassay analyses showed that (1) in-
fluent was generally toxic to both the
fathead minnow and the water flea; (2)
treated effluent from runs without hydro-
gen peroxide was nontoxic to the fathead
minnow, but moderately toxic to the water
flea; (3) treated effluent from runs with
hydrogen peroxide added was toxic to both
the fathead minnow and the water flea.
Comparison of effluent toxicity with re-
sidual hydrogen peroxide concentration in
the effluent indicates that effluent toxicity
may be partially due to hydrogen peroxide
residual rather than treatment by-products.
These results represent conditions during
the demonstration only and are site- and
waste-specific. Additional studies are
needed to draw definitive conclusions on
the effluent toxicity.
No significant changes in pH, alkalinity,
hardness, TOC, TC, POC, or specific con-
ductance were observed during treatment.
Also, the CAV-OX® process did not re-
move iron or manganese from the influent
groundwater. Influent TRPH concentrations
ranged from 0.64 to 1.15 mg/L. Effluent
TRPH concentrations ranged from below
detection limits to 0.78 mg/L.
Turbidity readings for the influent
samples ranged from 2.97 to 3.58 nephelo-
metric turbidity units (ntu). Turbidity read-
ings for the effluent samples ranged from
2.07 to 3.38 ntus.
Water temperature increased at an av-
erage rate of about 0.26°F/min of UV ex-
posure in the CAV-OX® I process. In the
5-kW CAV-OX® II process, the water tem-
perature increased at an average rate of
about 2.36°F/min of UV exposure; and in
the 10-kW CAV-OX® II process, the water
temperature increased at an average rate
of about 4.29°F/min of UV exposure. Be-
cause the treatment equipment is exposed
to the surrounding environment, the tem-
perature increase may vary with the ambi-
ent temperature or other atmospheric
conditions.
Estimated Treatment Costs
Using information from the SITE dem-
onstration, an economic analysis exam-
ined 12 separate cost categories for the
CAV-OX® process applied to a hypotheti-
cal, 5-yr groundwater remediation project
at a Superfund site. This analysis exam-
ined costs for the CAV-OX® I low-energy
process and the CAV-OX® II high-energy
process using flow rates of 10 and 25
gpm. Costs were determined in October
1993 dollars.
For the CAV-OX® I low-energy process,
capital costs are estimated to be $314,500
for a 10-gpm process and $342,500 for
the 25-gpm process. Annual operation and
maintenance (O&M) costs are estimated
to be $71,000 for the 10-gpm process and
$78,000 for the 25-gpm process. Costs to
treat 1,000 gal of contaminated, ground-
water are estimated to be $30 for the 10-
gpm process and $13 for the 25-gpm
process.
For the CAV-OX® II high-energy pro-
cess, capital costs for both the 10- and
25-gpm processes are the same as those
for the CAV-OX® I low-energy process.
Annual O&M costs are estimated to be
$75,000 for the 10-gpm process and
$86,000 for the 25-gpm process. Costs to
treat 1,000 gal of contaminated ground-
water are estimated to be $31 for the 10-
gpm process and $14 for the 25-gpm
process.
Conclusions
The following preferred operating con-
ditions were determined for the CAV-OX®
I low-energy process: (1) an influent hy-
drogen peroxide concentration of 23.4 mg/
L and (2) a flow rate of 0.6 gpm. Under
these conditions, TCE and benzene con-
centrations in the effluent met State of
California drinking water action levels and
federal MCLs (target levels) for TCE and
benzene (5 jag/L and 1 (ig/L, respectively)
at the 95% confidence level. The average
removal efficiencies for TCE and benzene
were both greater than 99.9%.
No preferred operating conditions were
identified for the CAV-OX® II high-energy
process. However, certain conditions did
reduce the concentrations of either TCE
or benzene, but not both, below target
levels. Based on the results of the demon-
stration, the optimal conditions (as deter-
mined by percent reductions) appear to
be: (1) an influent hydrogen peroxide level
of 48.3 mg/L and (2) a flow rate of 1.4
gpm. Under these conditions, the average
TCE and benzene concentrations in the
effluent were reduced below target levels;
however, due to data variability, these con-
stituents were not reduced below target
levels at the 95% confidence level. The
average removal efficiencies for TCE and
benzene were about 99.7% and 99.8%,
respectively.
Bioassay tests were inconclusive as to
whether the toxicity of effluent was caused
by hydrogen peroxide or by CAV-OX® pro-
cess treatment residuals. Because hydro-
gen peroxide appears to be toxic to certain
aquatic organisms at levels greater than
10 mg/L, effluent hydrogen peroxide con-
centrations should be kept well below 10
mg/L. Also, additional bioassay tests
should be performed on the process efflu-
ent to determine whether the effluent can
be safely discharged to aquatic bodies.
6
------- |