EPA/600/R-02/094
September, 2002
Demonstration of the HiPOx Advanced
Oxidation Technology for the Treatment of
MTBE-Contaminated Groundwater
Final Report
By
Thomas F. Speth
Water Supply and Water Resources Division
National Risk Management Research Laboratory
Cincinnati, Ohio 45268
Greg Swanson
TetraTech EM Inc.
San Diego, CA 92101
National Risk Management Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
Disclaimer
The information in this document has been funded by the U.S. Environmental Protection
Agency under Contract No. 68-C-00-181 to Tetra Tech EM Inc. It has been subject to the
Agency's peer and administrative reviews and has been approved for publication as an EPA
document. The results described herein should not be interpreted as USEPA policy or guidance.
Mention of trade name or commercial products does not constitute endorsement or
recommendation for use by the US Government.
11
-------
Foreword
The U.S. Environmental Protection Agency (EPA) is charged by Congress with
protecting the Nation's land, air, and water resources. Under a mandate of national
environmental laws, the Agency strives to formulate and implement actions leading to a
compatible balance between human activities and the ability of natural systems to support and
nurture life. To meet this mandate, EPA's research program is providing data and technical
support for solving environmental problems today and building a science knowledge base
necessary to manage our ecological resources wisely, understand how pollutants affect our
health, and prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's center
for investigation of technological and management approaches for preventing and reducing risks
from pollution that threaten human health and the environment. The focus of the Laboratory's
research program is on methods and their cost-effectiveness for prevention and control of
pollution to air, land, water, and subsurface resources; protection of water quality in public water
systems; remediation of contaminated sites, sediments and ground water; prevention and control
of indoor air pollution; and restoration of ecosystems. NRMRL collaborates with both public
and private sector partners to foster technologies that reduce the cost of compliance and to
anticipate emerging problems. NRMRL's research provides solutions to environmental problems
by: developing and promoting technologies that protect and improve the environment; advancing
scientific and engineering information to support regulatory and policy decisions; and providing
the technical support and information transfer to ensure implementation of environmental
regulations and strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term
research plan. It is published and made available by EPA's Office of Research and Development
to assist the user community and to link researchers with their clients.
E. Timothy Oppelt, Director
National Risk Management Research Laboratory
in
-------
Abstract
The HiPOx technology is an advanced oxidation process that incorporates high-precision
delivery of ozone and hydrogen peroxide to chemically destroy organic contaminants with the
promise of minimizing bromate formation. A MTBE-contaminated groundwater from the
Ventura County Naval Base in Port Hueneme, CA was used to evaluate this technology. Due to
extremely high concentrations of bromide in the feed water (1.3 mg/L) and the desire to limit
bromate formation, an experimental system was operated with 630 ozone injector ports in series.
In all trials, the HiPOx system reduced MTBE from 748 |ig/L to below its regulatory limit of 5
l-ig/L; however, bromate was not maintained below its regulatory limit of 10 |ig/L. The oxidative
intermediate tert-butyl alcohol (TEA) was below its regulatory effluent limit of 12 |ig/L in two
of the three trials. Both MTBE and bromate were under their regulatory limits at intermediate
sampling ports that corresponded to 330, 470, and 540 injector ports for the three runs.
However, TEA was above its regulatory limit at these locations for all three runs. To control
TEA, more injection ports were required. However, as shown above, additional injection ports
increased the bromate concentration above its regulatory limit. Therefore, the experimental
HiPOx system was not fully successful with this atypical water at the chosen oxidant doses.
A model calculation is presented that uses many simplifying assumptions to show that
this HiPOx system may have been fully successful at this location under the chosen oxidant
doses if the influent bromide concentration was 0.56 mg/L, or less. Since a bromide
concentration of 0.56 mg/L is still extremely high for a drinking water source, the HiPOx system
appears to hold promise for destroying MTBE and its oxidative byproduct TEA while
controlling bromate formation, even in waters that have high bromide concentrations. However,
before application to other sites, pilot testing will be needed due to the uncertainty in
performance resulting from source-water quality differences.
Appendix A contain the manufacturer's supporting data from other sites and data
collected by the manufacturer during the demonstration runs described herein.
IV
-------
Table of Contents
Foreward Hi
Abstract iv
Tables vi
Figures vii
Acknowledgment viii
1 Background 1
1.1 Technology description 1
1.2 Process chemistry 2
2 Methods and Materials 4
3 Results and Discussion 5
3.1 General water quality 6
3.2 MTBE and byproducts 8
3.3 Bromide/bromate modeling 16
4 Conclusions 18
5 References 19
6 Appendix A 20
-------
Tables
1-1 Total Ozone and Peroxide Doses for the Three Runs 5
3-1 Summary of Analytical Results for Metals and General Chemistry Parameters... 7
3-2 MTBE Removal Efficiency 8
3-3 Oxidation Intermediates and By-products 9
3-4 Summary of Analytical Results for Organic Parameters 10
3-5 Hypothetical Influent Bromide Concentration That Would Result in 10 mg/L
of Bromate in the Effluent Water for 35 Cycles 17
6-1 Comparison of Results on Run #3 21
6-2 Bromate Control at other Locations with HiPOx Technology 23
6-3 TEA Destruction and Acetone Reduction at other Location with
HiPOx Technology (and Bioreactor Technology) 24
VI
-------
Figures
1-1 HiPOx Technology Process Flow Diagram 3
3-1 Destruction / Formation Curves for HiPOx Study at Port Hueneme (Test 1) 11
3-2 Destruction / Formation Curves for HiPOx Study at Port Hueneme (Test 2) 12
3-3 Destruction / Formation Curves for HiPOx Study at Port Hueneme (Test 3) 13
3-4 Combined Destruction / Formation Curves with Standard Deviations for HiPOx
Study at Port Hueneme (Tests 1-3) 15
6-1 Bromate Control Chemistry 20
vn
-------
Acknowledgments
The authors would like to acknowledge Albert Venosa who was Work Assignment
Manager for this demonstration, Dr. Fran Kremer (USEPA) who oversaw the entire EPA / Port
Hueneme project, Dr. Carl Enfield (USEPA) who was responsible for installing the wells that
provided water to this project, Katherine Baylor (USEPA) who helped run the system, and Sam
Hayes (USEPA) and Michael Elovitz (USEPA) who provided inhouse reviews of this research
brief. The authors would also like to acknowledge Terry Applebury, Reid Bowman, and Doug
Gustafson of Applied Process Technologies, Inc. (APT) for their help in determining the optimal
operating conditions, for the set up of the unit before testing, and for supplying Appendix A.
Finally, the authors would like to acknowledge Emmet Black (TetraTech EM Inc.) who operated
and sampled the HiPOx unit.
Due to lack of funds at the time of this technology's scheduled testing, a CRADA
agreement between the USEPA and APT was created where APT supplied the USEPA with the
funding necessary to cover the analytical costs of this project (approximately $7K). The USEPA
used the funds through its contractor (TetraTech) who ran the study, collected the samples,
shipped the samples, paid for the analyses, and summarized the results. At no time did APT
influence the results or the subsequent discussion. Please see Appendix A for APT's data and
comments regarding this project.
Vlll
-------
1 Background
The U.S. Environmental Protection Agency (EPA), National Risk Management Research
Laboratory (NRMRL) and the U.S. Navy entered into a memorandum of understanding to
conduct a multi-year program involving demonstration and evaluation of innovative technologies
for treatment of methyl fert-butyl ether (MTBE) in groundwater. Technology vendors were
identified through an open solicitation that requested proposals for processes to treat MTBE.
Vendors participating in the program were selected based on internal and external peer reviews.
One of the vendors selected for the demonstration program was Applied Process
Technology, Inc. (APT), the developer of an advanced chemical oxidation technology called
HiPOx. The site selected for the demonstration was the Ventura County Naval Base in Port
Hueneme, CA, where a plume of MTBE-contaminated groundwater was present. The HiPOx
technology was demonstrated as an ex situ application at a location known as the Wellhead
Protection Zone, which is toward the toe of the MTBE plume. At this location, MTBE
concentrations in the groundwater were attenuated to less than 1000 micrograms per liter (i-ig/L),
and other gasoline components were not present. The purpose of the demonstration at this
location was to show the capabilities of the technology for remediating MTBE-contaminated
groundwater for potential reuse as a drinking water supply.
Besides MTBE, the primary contaminants of interest for the demonstration included tert-
butyl alcohol (TEA), acetone, and bromate. TEA and acetone are intermediates in the oxidative
degradation of MTBE, while bromate results from the oxidation of bromide during ozone
treatment. Treatment goals for the demonstration were established for these parameters based on
Maximum Contaminant Levels (MCL) and other regulatory standards for drinking water in the
State of California. California currently regulates MTBE at 5 |ig/L (Enforceable Secondary
MCL), TEA at 12 |ig/L (Action Level), and bromate at 10 |ig/L (EPA Stage 2 Disinfection By-
product Rule). There were no drinking water standards for acetone at the time of testing, and
thus no treatment goal was set for this parameter.
1.1 Technology description
The HiPOx technology developed by APT is similar to other advanced oxidation
technologies that use ozone (O3) and hydrogen peroxide (H2O2) to destroy organic compounds in
contaminated waters. However, the vendor claims that the high-precision delivery of the
oxidants and the use of multiple oxidant injection ports enhances process efficiency. In
"traditional" applications of advanced oxidation technology, 2 to 3 percent ozone by weight is
injected at a single point through a diffuser and is allowed to bubble up at atmospheric pressure
through the contactor. Volatile organic compounds (VOCs) and other contaminants are
destroyed over contact times as long as 20 minutes. The HiPOx technology enhances the mass
transfer of ozone into the water by using higher ozone injection concentrations (8 to 10 percent
by weight), higher operating pressures, and in-line mixers to promote efficient mixing. The
vendor claims that the multiple reaction zones utilized in the HiPOx technology further enhances
the process efficiency by keeping the localized ozone concentration low which in theory
minimizes bromate formation.
-------
A general process flow diagram for the HiPOx process is shown in Figure 1-1 . In the
HiPOx process, liquid hydrogen peroxide is injected into the influent stream. Ozone, which is
generated from liquid oxygen on site, is injected through multiple ports along a serpentine
reactor. The reactor is maintained at a pressure of about 35 to 45 pounds per square inch gauge
(psig) to provide efficient mass transfer of ozone into solution. Following each ozone injection
port, the dosed fluid immediately flows through an in-line mixer to ensure that the ozone is
mixed into solution, and then through a brief reaction zone. Thus, the reactor consists of a series
of injection/mixing/reaction modules in series, the purpose of which is to maintain low
instantaneous O3:H2O2 mole ratios, as stated earlier.
1.2 Process chemistry
In aqueous solution, ozone dissociates and reacts with hydrogen peroxide to produce
hydroxyl radicals (•OH). Hydroxyl radicals are powerful oxidizing agents and work alone or in
concert with its precursors to oxidize organic contaminants. Given enough time, complete
mineralization to carbon dioxide and water can be achieved. The following equation presents
the overall balanced equation for hydroxyl radical formation from ozone and hydrogen peroxide:
2 O3 + H2O2 -> 2 .OH + 3 O2 Eq. 1
The following equation represents the oxidation of organic compounds:
Organic Compounds + »OH — * Intermediates + »OH — * CO2 + H2O Eq. 2
The oxidative degradation of MTBE involves a series of chemical reactions. The initial
oxidation product is tert-buty\ formate (TBF), a short-lived intermediate that is converted to a
longer-lived intermediate, TEA. TEA is subsequently converted to acetone prior to conversion
to carbon dioxide and water. Thus, the overall reaction sequence can be summarized as follows:
MTBE — > TBF — > TEA — > Acetone — > CO2 + H2O Eq. 3
In addition to TEA and acetone, bromate may be formed by the oxidation of bromide by
ozone and hydroxyl radicals. This is a drawback for ozone-based advanced oxidation processes.
Therefore, a key question for the use of ozonation systems in drinking water applications is
whether bromate formation can be controlled. The HiPOx vendor claims that formation of
bromate is minimized by the HiPOx system's use of multiple ozone injection ports and in-line
mixers that keep local ozone concentrations low. Appendix A contains information from APT
on this technology, APT's data from this study, and supporting data from other studies/locations.
-------
R:\G1093\66140\ FLOW.dwg 03/05/2001 SANDOVK DN
> COOLING WATER )
LEGEND
^^— GROUNDWATER FLOW LINE
LIQUID LINE
VAPOR LINE
r^—i PUMP
—{X3-
CKl
-HX1--
I
I I
VALVE
TEMPERATURE SENSOR/TRANSMITTER
FLOW SENSOR/TRANSMITTER
PRESSURE SENSOR/TRANSMITTER
CONDENSATE
Figure 1
HiPOx Technology Process
Flow Diagram
-------
2 Methods and Materials
The APT HiPOx Pilot Test Unit (PTU) was used in this demonstration. The PTU was
housed in a mobile, box van approximately 14 feet in length. The PTU was designed as a
simplified version of the HiPOx process for use in pilot studies. The PTU was configured with
18 reactor modules, which are sufficient to destroy most contaminants. For this demonstration,
however, a larger number of modules was desired to demonstrate the ability of the technology to
completely destroy MTBE and other oxidative intermediates while not producing bromate in this
high bromide water. To address this need, the PTU was operated in a recirculating mode, in
which a batch of contaminated groundwater was recycled through the PTU multiple times to
simulate a reactor with a larger number of reactor modules. As stated above, a reactor module
consists of an ozone injection port followed by a static mixing zone.
The PTU was set up at the Wellhead Protection Zone early in November 2001. The
influent process water was drawn from a manifold that connected nine wells installed in the
MTBE plume at the Wellhead Protection Zone. The optimal ozone and peroxide doses were
determined from numerous trial studies conducted by the manufacturer. These data are not
presented in this document. In preparation for the demonstration, a 1900-liter storage tank was
filled to near capacity to facilitate the introduction of process water into the PTU.
Approximately 300 liters were drawn into an intermediate 380-liter storage tank for initial
hydrogen peroxide dosing.
The process demonstration occurred on November 15, 2001 and consisted of three runs at
the same operating conditions. To initiate a run, the process water in the 380-liter intermediate
storage tank was dosed with 60.8 mg/L of hydrogen peroxide, and the PTU reactor was filled
with 70 liters of process water (the volume of the reactor plus the 7.6 liter recirculating tank).
The system pump was then initiated and the process water was pumped continuously around the
PTU reactor, with the reactor effluent being recycled back to the recirculating tank at a flow rate
of 13 liters/min. The 7.6 liter recirculating tank was located at the beginning of the system. The
380-liter tank was not in the recycle loop. Since each run consisted of 35 cycles through the
PTU, and the PTU incorporated 18 reactor modules, each run simulated a HiPOx treatment
system with 630 reactor modules. Each process run took approximately 3 hours to complete all
35 cycles.
The ozone to hydrogen peroxide mole ratio was monitored and controlled using
adjustments to the ozone's mass flow indicator and the hydrogen peroxide's metering pump.
Ozone was dosed into the injection port of each reactor module continuously during each run
(3.4 mg O3 / L per cycle). The initial hydrogen peroxide dose was supplemented with a low dose
(1.72 mg H2O2 /L per cycle) that was injected continuously into the effluent end of the PTU
reactor. The total hydrogen peroxide and ozone doses for each run are listed in Table 1-1.
-------
Table 1-1. Total Ozone and Peroxide Doses for the Three Runs.
H2O2 dose (mg/L)
Ozone Dose (mg/L)
Runl
121
119
Run 2
121
119
Run3
121
119
Average
121
119
Samples of the process water were collected at five separate times/locations during each
run. Initial (influent) samples were collected from the 380-liter intermediate storage tank before
any chemical oxidants were added; treated process water samples were collected after 12, 24, 30,
and 35 cycles (at the final effluent sample port). Prior to sample collection, the sampling port
was purged to ensure that any stagnant water was flushed from the port. Separate samples were
collected for analysis of VOCs, bromate, and general chemical parameters. Sample containers,
preservatives, and other sampling procedures were in accordance with standard reference
methods.
Ozone analysis of treated water samples was performed in the field with a Hach® ozone
field kit to monitor ozone concentrations for process monitoring and control purposes. All other
samples were shipped to a certified laboratory for chemical analysis. Samples were analyzed for
VOCs in accordance with EPA Method 5030 (purge and trap) and EPA Method 8260 (capillary
column gas chromatography/mass spectrometry). These EPA methods were modified to include
MTBE and TEA as target analytes after method evaluation studies confirmed that these analytes
were satisfactorily recovered and that quantitation limits below the treatment goals could be
routinely achieved. Samples were also analyzed for bromate by ion chromatography in
accordance with EPA Method 317. Metals and other general chemical parameters were analyzed
in accordance with the appropriate EPA reference methods for water samples.
Off-gases from the reactor were passed through a condenser to separate water, and then
through an ozone destruction unit, which contained a chemical reducing agent to destroy any
residual ozone prior to venting to the atmosphere. Gases vented from the enclosure surrounding
the ozone generation unit were also passed through this unit to ensure that fugitive emissions of
ozone were not allowed into the process/control room. The treatment system was equipped with
influent and effluent sampling ports, as well as flow measurement and process
monitoring/control equipment, to assure that process operations were effectively monitored and
controlled.
-------
3 Results and Discussion
This section presents the analytical results for the process water samples collected from
the PTU, and compares these results to the demonstration treatment objectives. As stated above,
the primary objective for the demonstration was the reduction of MTBE to below 5 |ig/L in the
final effluent. The second objective for the demonstration was the reduction of TEA, acetone,
and bromate to below their respective regulatory limits.
-------
To ensure that this study would produce valid data that were suitable for their intended
use, a data quality review was conducted using results of both field quality control (QC) samples
and laboratory QC samples. For the VOC analyses, results for matrix spike/matrix spike
duplicate samples indicated that recoveries of detected VOCs were consistently within the
acceptance range of 75 percent to 125 percent, and that the relative percent differences were less
than the acceptance criterion of 25 percent. Trip blanks did not reveal contamination with any of
the detected VOCs, although method blanks revealed occasional MTBE contamination below the
laboratory quantitation limit. QC sample results for other analytical parameters were in
compliance with method acceptance criteria. Thus, the only significant qualification of the data
was the estimated nature of some MTBE results that were below the laboratory quantitation
limit, which was set at the lowest standard concentration (1 i-ig/L), or that were above the
calibration range of the instrument, as noted in the data tables below.
3.1 General water quality
The groundwater used in this study was relatively hard with high alkalinity as shown in
Table 3-1. Most importantly, the bromide levels were very high (1.3 mg/L). This allowed for an
evaluation of this technology under very challenging conditions for bromate formation.
Turbidity levels were high, likely due to inadequate well development and/or iron precipitation.
Sodium levels were also elevated as compared to a typical drinking water. The influent
dissolved organic carbon (DOC) and total organic carbon (TOC) results were confusing in that
the DOC values were greater than the TOC values in all cases. Also, the influent DOC
concentrations went down from Run 1 through Run 3, whereas the influent TOC concentrations
and influent UV absorbances did not. The synthetic organics: benzene, ethylbenzene, toluene, or
total xylene were all below their detection limits of 10, 10, 10, and 30 |ig/L, respectively. The
water did contain cis 1,2 dichloroethene at concentrations estimated between 4 and 6 |ig/L. This
suggests that either cis 1,2 dichloroethene was following the MTBE plume closely or that the
wells were contaminated from another source.
The effluent results for metals and general chemistry parameters are also listed in Table
3-1. Iron and manganese appear to be reduced in the effluent, likely due to precipitation and
settling induced by the oxidative environment. Alkalinity and possibly calcium dropped slightly,
but the other inorganics were not affected by treatment. This includes bromide, which is known
to be the precursor to bromate; however, the small amount of conversion of bromide to bromate
would not have impacted the quantitation of the high levels of bromide. Ozone was not found in
the effluent samples due to excess hydrogen peroxide.
Both DOC and TOC were somewhat reduced in the effluent as compared to the influent
indicating that some organic carbon was destroyed by the advanced oxidation treatment.
However, consistently higher levels of DOC than TOC in both the influent and effluent could not
be explained, nor could the drop in DOC influent values from Run 1 through Run 3 without a
parallel drop in either effluent DOC, influent UV absorbance, influent TOC, or effluent TOC.
-------
Table 3-1. Summary of Analytical Results for Metals and General Chemistry Parameters.
Run
1
2
3
Analytical Parameter
Calcium
Iron
Magnesium
Manganese
Potassium
Sodium
Alkalinity
Bromide
DOC
Ozone
pH
TOC
Turbidity (NTU)
UV@254nm(l/cm)
Calcium
Iron
Magnesium
Manganese
Potassium
Sodium
Alkalinity
Bromide
DOC
Ozone
pH
TOC
Turbidity (NTU)
UV@254nm(l/cm)
Calcium
Iron
Magnesium
Manganese
Potassium
Sodium
Alkalinity
Bromide
DOC
Ozone
pH
TOC
Turbidity (NTU)
UV@254nm(l/cm)
Influent Concentration
(mg/L)*
340
7.4
130
2
8.1
270
446
1.3
8.3
0.0
7.10
3
46.2
0.061
360
7.7
140
2.1
8.4
280
458
1.2
5
0.0
7.09
3
48.6
0.059
340
7.5
130
2
8
270
452
1.3
3.6
0.0
7.17
3
24.0
0.061
Final Effluent
Concentration (mg/L)*
320
5.2
130
1.5
8.1
270
406
1.3
3.1
0.1
7.52
2.5
37.4
0.027
320
5.3
130
1.4
7.9
270
406
1.3
2.7
0.0
7.97
2.2
23.0
0.011
330
5.7
130
1.5
7.9
270
408
1.3
3.7
0.0
8.04
2.2
42.6
0.011
* unless noted otherwise
-------
As part of this study, the influent and effluent waters were chlorinated at Uniform
Formation Conditions (Summers et al., 1994). A chlorine dose of 2.7 mg/L was delivered to
each of the waters. The reported final chlorine residual ranged between 0.90 and 1.25 mg/L
implying that the organic constituents in the water reacted with the chlorine. However, the
amounts of trihalomethane, haloacetic acid, and total organic halide formed were extremely low.
The field notes indicate that the influent sample was mistakenly taken after hydrogen peroxide
addition. The effluent sample also contained hydrogen peroxide, as expected, albeit at
concentrations higher than first envisioned. Unfortunately, the high levels of hydrogen peroxide
quickly reacted with the spiked chlorine (Connick, 1947; Held et al., 1978) preventing the
reaction between chlorine and the natural organic material. The hydrogen peroxide also likely
interfered with the DPD method used to measure chlorine (Sengupta et al., 1978; Bader et al.,
1988). Therefore, the reported chlorine residual after 24 hours of between 0.90 and 1.25 mg/L
was likely a false positive resulting from the presence of large amounts of hydrogen peroxide. In
conclusion, the UFC results tell us nothing about the effect of HiPOx treatment on subsequent
disinfection byproduct formation. In any event, it is unclear as to the worth of such information
for such an atypically high-bromide water as that used in this study.
3.2 MTBE and byproducts
Table 3-2 shows the influent and final effluent (35 cycles) concentrations of MTBE
during each of the three replicate runs. The percent removal of MTBE for each run is also listed.
As shown in this table, the HiPOx system was able to reduce MTBE from an average of 748
l-ig/L in the influent to less than 1 |ig/L, which was the laboratory quantitation limit. The average
MTBE removal efficiency was greater than 99.87 percent, which reflects the near total
destruction of MTBE by the HiPOx system. The treatment goal of 5 |ig/L MTBE in the final
effluent was easily achieved for all three runs.
Table 3-2: MTBE Removal Efficiency.
MTBE (jig/L)
Removal Efficiency
Run 1
Start / End
744 / 0.3*
99.96 %
Run 2
Start / End
751 / <1
>99.87 %
Run 3
Start / End
749 /< 1
>99.87%
Average
748 /99.87%
* Estimated value below laboratory quantitation limit.
Table 3-3 shows the concentrations of intermediates and by-products generated by the
HiPOx treatment. The TEA concentration in the final effluent was below the regulatory goal of
12 |J.g/L in Runs 2 and 3, but not in Run 1. One possible explanation may be that the higher
DOC levels in Run 1 may have scavenged the hydroxyl radicals, preventing them from
destroying the TEA. However, as stated above, the DOC and TOC results are somewhat
questionable. The concentration of acetone was high in the effluent (average =135 |ig/L) for all
three runs. However, due to the lack of a drinking water regulation for acetone, no conclusion
could be made as to the acceptability of these levels. Bromate was clearly generated as a by-
product of the chemical oxidation treatment with effluent concentrations above the regulatory
goal of 10 |ag/L in all three runs.
-------
Table 3-3: Oxidation Intermediates and By-products.
TBAOig/L)
Acetone (jig/L)
Bromate (jig/L)
Runl
Start / End
<40 / 29
<40/154
<1/12.1
Run 2
Start / End
<40 /8.1
<40/135
<2/19.9
Run3
Start / End
<40 75.9
<40/117
<2/27.0*
Average
<40/14.3
<40/135
<2/19.7
* Estimated value above instrument calibration range.
Table 3-4 shows all the sample results for MTBE, TEA, acetone, and bromate for the
three runs. The results are also shown in Figures 3-1, 3-2, and 3-3 for Runs 1, 2, and 3,
respectively. Data that were estimated, as noted in Table 3-4, are presented in Figures 3-1
through 3-3. However, if a data point was reported as less than a certain value in Table 3-4, it
was not plotted in the figures. The figures show that most of the MTBE was destroyed in the
first 24 reactor cycles (432 injectors). TEA, acetone, and bromate exhibit different relationships
with the number of reactor cycles, depending on the chemistry involved in their formation.
TEA, which is an intermediate in the degradation of MTBE, initially increased in concentration
(at 12 cycles), but then decreased over the remaining reactor cycles. Acetone concentrations
increased rapidly through 24 cycles, in conjunction with MTBE destruction, but then appeared to
stabilize. These trends are consistent with the reaction sequence for MTBE degradation
suggested previously, wherein TEA is an initial intermediate, and acetone is the next-to-fmal
product of MTBE degradation. Bromate concentrations increased steadily with the number of
reactor cycles since its formation is related to oxidant dose, bromide concentration, and time, and
is independent of MTBE degradation.
-------
Table 3-4. Summary of Analytical Results for Organic Parameters.
Analytical
Parameter
MTBE
TEA
Acetone
Bromate
Run
1
2
3
Average
Std Dev.
1
2
3
Average.
Std Dev.
1
2
3
Average
Std Dev.
1
2
3
Average
Std Dev.
Influent
Concentration
fog/L)
756 & 731
766 & 736
758 & 740
748
14.0
11*&<40
<40 & <40
<40 & <40
NA
NA
<40 & <40
<40 & <40
<40 & <40
NA
NA
<1&<2
<2 & 2.6
<2&<2
NA
NA
After
12 Cycles
fog/L)
31
15
10
19
11
126
96
99
107
16
101
117
79
99
19
4.3
1.9
6.7
4.3
2.4
After
24 Cycles
fog/L)
1.7
0.3*
3.9*
2.0*
1.8*
67
33
21
40.3
23.9
156
150
103
136
29
8.0
6.9
14
9.6
3.8
After
30 Cycles
fog/L)
0.6*
<1
3.8*
<1.8*
1.7*
48
20
<40
<36
14
164
151
81
132
45
10.1
16.7
18.4
15.1
4.4
After
35 Cycles
(Hg/L)
0.3*
<1
<1
NA
NA
29
8.1
5.9
14
13
154
135
117
135
18
12.1
19.9
27.0*
19.7*
7.5*
* Estimated concentration (outside calibration range or below quantitation limit)
NA = Not applicable
10
-------
1000 -3
100 --
d
o
• I—I
s-a
cd
'S
o
u
10 -
1 -
0.1
0
• MTBE
T TEA
• Bromate
T
10 20 30
Cycles (18 injectors each)
40
Figure 3-1. Destruction / Formation Curves for HiPOx Study
at Port Hueneme (Test 1).
ll
-------
1000
d
o
• I—I
s-a
cd
0)
O
ti
o
u
100 -
10 -
i -
0.1
0 10 20 30
Cycles (18 injectors each)
Figure 3-2. Destruction / Formation Curves for HiPOx Study
at Port Hueneme (Test 2).
40
12
-------
1000
d
o
• I—I
s-a
cd
0)
O
ti
o
u
100 -
10 -
0
TBA@12 [ig/I
Bromate@10 jig/L
MTBE
@5ng/L
T
MTBE 3
TEA
Bromate-
10 20 30
Cycles (18 injectors each)
40
Figure 3-3. Destruction / Formation Curves for HiPOx Study
at Port Hueneme (Test 3).
13
-------
For the first run where the effluent TEA was above the regulatory limit, one of the two
influent samples had a estimated TEA concentration of 11 i-ig/L. Also, for the second run, one of
the two influent samples had a bromate concentration of 2.6 |ig/L. These data raise questions as
to whether all of the TEA or bromate in the system were formed from the oxidation process, or
whether they were present in the influent water. TEA could have come from the natural
breakdown of MTBE in the subsurface; however, it is more difficult to explain why bromate was
present in the anoxic groundwater.
The location (number of cycles) where bromate reached 10 |ig/L was estimated to be 30,
26, and 18 cycles (Figures 3-1, 3-2, and 3-3, respectively), with an average of 25 cycles (450
injectors). At these points, MTBE was below its regulatory limit of 5 |ig/L, but TEA was above
its regulatory limit of 12 |ig/L for all three runs, and acetone was present at concentrations above
100 |ag/L. Therefore, at no point was the system fully successful in treating this atypical water
that represented a challenging scenario for bromate formation.
Figure 3-4 shows the three runs averaged together with the standard deviations plotted as
error bars. The results show the same pattern as that discussed above. The MTBE was removed
quickly, reaching the regulatory limit of 5 |ig/L after 18 cycles (450 injectors). The average
TEA concentration remained above its regulatory limit of 12 |ig/L throughout the reactor,
although the error bars on the final effluent samples are large, reaching below 12 |ig/L. The
bromate concentration reached its regulatory limit of 10 |ig/L after 25 cycles (450 injectors).
Also, the acetone concentrations remained relatively high.
14
-------
10000 -
a
o
0)
CJ
=
o
U
1000 i
100 -
0
10
20
30
Cycles (18 injectors each)
40
Figure 3-4. Combined Destruction/Formation Curves with Standard
Deviations for HiPOx Study at Port Hueneme (Tests 1-3).
15
-------
3.3 Bromide/bromate modeling
Although bromate was not successfully controlled in this water, it is possible to calculate
the bromate formation for this water at hypothetically lower influent bromide levels. That is,
assuming that all other water parameters and operational parameters remain the same, a
calculation can be made to determine the influent bromide concentration that would result in an
effluent bromate concentration of 10 |ig/L. The formation of bromate is, in its simplest form, a
reaction of bromide and hydroxyl radical.
Br' + »OH —> BrCV Eq. 4
It can be further assumed that in this oxidative system there is no reverse reaction. Therefore,
the formation of bromate can be expressed as
d[BrO3-]/dt = k [Br ] [»OH ] Eq. 5
Because this system continually replenished ozone by using 630 injector ports, it is safe
to say that if the system was operated in a similar dosing situation (Table 1-1), the concentration
of hydroxyl radicals would be relatively stable over the reactor, and would be very similar to
other hypothetical Port Hueneme groundwaters (where only the bromide concentration would
change), regardless of the chosen influent bromide concentration. Therefore, bromide would be
the only parameter that impacts the formation of bromate from one hypothetical Port Hueneme
water to another under these operating conditions. By also assuming that the amount of bromate
formed is insignificant when compared to the concentration of bromide, the bromide
concentration remains constant over the reactor. Therefore, the rate of bromate formation is
constant, and the integration results in a linear relationship. This linear relationship (shown
below) can be utilized to calculate the influent bromide concentration that would result in an
effluent bromate concentration of 10 |ig/L in this water. This is done in Table 3-5 for each run.
An implied assumption is that the hypothetical influent bromide concentration is still high
enough not to be influenced by losses to the formation of bromate, as assumed in the original
development. Since the hypothetical influent bromide concentrations range from 0.48 mg/L to
1.07 mg/L, and the final bromate concentration is set at 10 |ig/L, this appears to be a safe
assumption.
Hypothetical influent = (1.3 mg/L bromide)*(10 [ig/L bromate in effluent) Eq. 6
bromide (mg/L) (Experimental effluent bromate cone., i-ig/L)
16
-------
Table 3-5. Hypothetical Influent Bromide Concentration That Would Result in 10 mg/L of
Bromate in the Effluent Water for 35 cycles.
Run / Test Number
1
2
3
Average of Runs 2 & 3
Effluent Bromate
Concentration from
Experiment
(M-g/L)*
12.1
19.9
27.0E
23.4
Hypothetical Influent Bromide
Concentration that would Result in
10 ng/L of Bromate in the Effluent
(mg/L)
1.07#
0.63
0.48
0.56
* Influent bromide concentration =1.3 mg/L for all runs.
# However, effluent contained TEA above its regulatory limit.
E = Estimated, outside of calibration range.
To determine an influent bromide concentration that will result in the achievement of all
regulatory goals, it is not appropriate to use the data from the first run because the effluent TEA
concentration was above its regulatory limit. Apparently, the system in Run 1 needed higher
hydroxyl radical doses to further destroy the TEA. However, this would also have resulted in
higher ozone concentrations, and hence, higher bromate formation. Because of this, the
following discussion will only concern Runs 2 and 3.
The calculations for Runs 2 and 3 show that to achieve a bromate concentration of 10
l-ig/L, the influent bromide concentration would have to be 0.63 and 0.48 mg/L, respectively.
The average of these two is 0.56 mg/L. Therefore, if all the operational and water quality
parameters other than influent bromide remained the same, this HiPOx system could have treated
this water if it contained less than 0.56 mg/L of bromide in the influent, without violating any of
the water quality regulations, including bromate. Because 0.56 mg/L of bromide is extremely
high for a typical source water, these calculations suggest that the HiPOx system, as operated
here, is able to control bromate formation in the presence of high bromide concentrations, as
theorized. It should be noted that this modeling effort made a number of simplifying
assumptions that may not be entirely correct for other waters or operating conditions. Any future
application of this technology would have to be piloted with the source water in question.
17
-------
4 Conclusions
The results of this study were a function of the uniqueness of the feed water which
represented a very challenging scenario for bromate formation/control. The high influent
bromide concentrations (1.3 mg/L) necessitated operating under unusual conditions where ozone
was injected at 630 separate ozone injector ports in series. The HiPOx process achieved greater
than a 99.87 percent reduction in MTBE concentration and easily met the treatment goal of
reducing the concentration of MTBE to below 5 |ig/L. However, significant concentrations of
MTBE degradation intermediates and oxidation by-products were present in the final effluent.
TEA was produced early during the chemical oxidation process. Its concentration was
diminished by further oxidation, reaching below its regulatory limit of 12 |ig/L in two of the
three runs. Acetone was generated and a sizable percentage was left unoxidized in the final
effluent (>100 i-ig/L). Bromate concentrations increased with the number of reactor cycles, and
the final effluent concentrations exceeded the drinking water standard of 10 |ig/L for all three
runs. Bromate formation was controlled up to 330 injection ports in all three runs. At this point,
MTBE was reduced to below its target concentration of 5 |ig/L, but TEA was above its
regulatory target of 12 |ig/L. Therefore, the HiPOx technology, as operated, effectively
destroyed MTBE, but the effluent water quality did not conform to drinking water standards due
to by-product formation.
Using a number of simplifying assumptions, it was calculated that if all the operational
and water quality parameters other than influent bromide remained the same, this HiPOx system
would have been able to meet all regulatory limits, including bromate, if the influent water at this
site contained less than 0.56 mg/L of bromide. Because 0.56 mg/L of bromide is extremely high
for a typical source water, these calculations suggest that the HiPOx system is able to control
bromate formation in the presence of high bromide concentrations, as theorized. It is quite
possible that the results of this study will allow for modifications to the HiPOx experimental unit
that would improve performance to the point of controlling MTBE, its oxidation byproducts,
and bromate at this location. More research is warranted. Because of the importance of water
quality on final effluent quality, any future application of this technology would have to be
piloted with the source water in question.
18
-------
5 References
Bader, H, Sturzenegger, V, and Hoigne, J (1988) "Photometric Method for the
Determination of Low Concentrations of Hydrogen Peroxide by the Peroxidase Catalyzed
Oxidation of N,N-Diethyl-p-phenylenediamine (DPD)," Water Research, 22:9:1109-
1115.
Connick, RE (1947) "The Interaction of Hydrogen Peroxide and Hypochlorous Acid in
Acidic Solutions Containing Chloride Ion," Journal ACS, 69:6:1509-1514.
Held, AM, Halko, DJ, and Hurst, JK (1978) "Mechanisms of Chlorine Oxidation of
Hydrogen Peroxide," Journal ACS, 100:18:5732-5740.
Sengupta, C, Helz, GR, Getz, JW, Higgins, P, Peterson, JC, Sigleo, AC, Sugam, R (1978)
"Survey of Chlorine Analytical Methods Suitable for the Power Industry," Electric Power
Research Institute, EPRI-EA-929.
Summers, RS, Hooper, SM, Shukairy HM, Solarik, G, and Owen, DM (1994) "Assessing
DBP Yield under Uniform Conditions," Journal AWWA, 88:6:80.
19
-------
6 Appendix A:
HiPOx Advanced Oxidation Treatment at Port Hueneme and Other Sites
(As Supplied by: Applied Process Technology, Inc.)
This Appendix explains how the HiPOx technology controls the formation of bromate in
bromide-containing waters, comments on the results obtained during the EPA demonstration at
Port Hueneme, and provides additional examples of sites where HiPOx has destroyed VOC's
and controlled the formation of bromate to below the regulatory limits for drinking water. This
Appendix also provides examples of real-world applications in which HiPOx has destroyed TEA
to below detectable levels and provides data pertaining to the reduction of acetone, an oxidation
by-product that forms during the destruction of MTBE and TEA.
How HiPOx Controls Bromate Formation
The bromate control process employed in the HiPOx technology is based on the
chemistry reported by Urs von Gunten1 (see Figure 6-1). HiPOx is designed to minimize the
amount of molecular ozone in solution by increasing the number of ozone injectors the system
utilizes this has been patented by APT. By minimizing the amount of ozone in solution, less
ozone is available to oxidize hypobromite (HOBr/OBr") before hydrogen peroxide reduces
hypobromite back to bromide. The reduction of hypobromite by hydrogen peroxide (k= 1 X 107
M"1 sec"1) is orders of magnitude faster than the oxidation of hypobromite to bromate by ozone
(k=100M4sec-1).
Figure 6-1. Bromate Control Chemistry
Br0
1 U. von Gunten and Y. Oliveras, Envir. Sci. and Tech., 32, 63 (1998); U. von Gunten, Y. Oliveras, Wat. Res., 31,
900 (1997); W.R. Haag, and J. Hoigne, Envir. Sci. and Tech., 17, 261(1983); U. von Gunten, J. Hoigne and A.
Bruchet, Water Supply, 13, 45 (1995)
20
-------
Bromate Control at Port Hueneme, CA
During screening tests conducted at Port Hueneme, the HiPOx system utilized 216, 432
and finally 504 ozone injectors to treat the site's water. Analytical data from these test runs
indicated that effluent bromate concentrations were 9.7, 7.0 and 5.2 |ig/L, respectively.
Therefore, in order to control bromate to less than 5 |ig/L during the actual demonstration, it was
determined that 632 ozone injectors would be required. During the screening tests conducted by
APT, all bromate analyses were performed by Clinical Laboratory of San Bernardino (Clinical
Lab) and VOC analyses were performed by Alpha Analytical, Inc.
During the actual demonstration, VOC destruction and bromate formation were measured
by laboratories selected by the EPA. However during Run #3, a split sample was analyzed by
both EPA-selected and APT-selected laboratories for comparison. Analytical data from all three
runs, including the split taken during Run #3, are shown in Table 6-1 below.
Table 6-1. Comparison of Results on Run #3
Component
MTBE
TEA
Acetone
Bromide
Bromate
Concentrations (i-ig/L)
EPA Labs
Run#l
Influent | Effluent
744 0.3
11 29
<40 154
1,300 1,300
<1 12.1
Run #2
Influent | Effluent
751 <1
<40 8.1
<40 135
1,200 1,300
<2 19.9
Run #3
Influent | Effluent
740 <1
<40 5.9
<40 117
1,300 1,300
<2 27
APT Lab
Run #3l <2
Influent | Effluent
750 0.57
11 7.6
<20 140
1,400 1,390
<5 <5
1 VOC's Analysis conducted by Alpha Analytical, Inc., 255 Glendale Ave., Suite 21, Sparks, Nevada. On Run #3,
sample splits submitted by Applied Process Technology.
2 Bromides and Bromates Analysis conducted by Clinical Laboratory of San Bernardino Inc., 21881 Barton Rd.,
Grand Terrace, CA 92313. On Run #3 sample splits submitted by Applied Process Technology.
As shown in Table 6-1, VOC data reported for Run #3 by the EPA laboratories were very
similar to those reported by Alpha Analytical. The bromide concentrations reported by both the
EPA labs and Clinical Lab were also in very good agreement. However, the effluent bromate
values reported by the EPA labs and Clinical Lab differed. The EPA labs reported influent and
effluent bromide concentrations of 1,300 |ig/L and an effluent bromate concentration of 27.0
l-ig/L. Clinical Lab reported an effluent bromate concentration below the detection limit of 5
l-ig/L, which would indicate the HiPOx system successfully controlled the formation of bromate.
If the bromate values reported by the EPA labs are correct, then the HiPOx system would require
additional injectors in order to control the formation of bromate.
21
-------
Bromate Control at other Sites
Bromate control is a characteristic unique to the HiPOx technology. APT has conducted
several trials demonstrating control of bromate formation and has placed a commercial HiPOx
unit that treats contaminated ground water to produce drinking water. The locations of these
sites, the contaminant destruction achieved, and the bromate results are listed in Table 6-2.
The site trials in Table 6-2 were conducted between 1998 and 2002 using a variety of
source water. The system that operated at the LADWP Headworks site ran continuously for four
weeks at 1,000 gallons per minute. Samples were taken every 6 hours to demonstrate
reproducibility and scalability of the HiPOx technology.
The South Lake Tahoe Commercial HiPOx system is operating continuously, producing
drink water for the South Lake Tahoe Public Utility District at 800 gallons per minute.
TEA Destruction and Acetone Reduction
Drinking water applications sometimes require complete destruction of TEA and acetone
depending on the state regulations. Complete destruction of these constituents was not required
during the Port Hueneme demonstration but has been achieved at other HiPOx installations.
The HiPOx technology has destroyed TEA concentrations to below detectable levels at a
number of sites. Data from two of these sites are provided in Table 6-3. Both sites are
commercial installations.
Acetone forms during destruction of MTBE and TEA by the HiPOx system, but it can be
destroyed as oxidation continues. At sites where acetone formation during destruction of other
contaminants is significant, then it becomes more cost effective to reduce acetone using a
bioreactor on the effluent of the HiPOx system. This method, HiPOx followed by a bioreactor,
has been used to control acetone at several sites treated by HiPOx including the two sites
detailed in Table 6-3. Acetone formation and biodegradation data from these sites, both post-
HiPOx and post-bioreactor, are also provided in Table 6-3.
22
-------
Table 6-2. Bromate Control at other Locations with HiPOx Technology
Location (Month/Year)
Southern California (8/98)
Southern California (8/98)
Southern California (1 1/98)
Southern California (1 1/98)
Headworks (4/99)
Headworks (4/99)
South Lake Tahoe (4/00)
South Lake Tahoe (4/00)
New York (1/01)
New York (1/01)
New York (1/01)
New York (1/01)
New York (1/01)
New York (1/01)
South Lake Tahoe (7/02)6
South Lake Tahoe (7/02)6
Number
Injectors
18
18
162
162
18
18
12
12
18
18
18
18
18
18
16
16
Sample
Location
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
MTBE
54
0.86
9801
0.58
NA
NA
11
<0.2
6.6
<0.5
284
<0.5
905
<0.5
0.2
<0.2
TEA
<100
<100
<100
40
NA
NA
<10
<10
<20
<20
<20
<20
<20
<20
<10
<10
Contam
Acetone
<10
<10
<10
120
NA
NA
<10
<10
<5
<5
<5
<5
<5
7.8
<10
<10
inate C
PCE
NA
NA
NA
NA
50
2.63
NA
NA
10
<0.5
11
<0.5
11
<0.5
NA
NA
"oncenti
TCE
NA
NA
NA
NA
100
<0.5
NA
NA
2.6
<0.5
2.5
<0.5
2.6
<0.5
NA
NA
•ation (ppb)
Bromide
321
318
10332
1025
211
210
34
35
120
NA
100
NA
100
NA
10
9
Bromate
<5
5.4
<5
7.0
<5
<5
<1
<3
<5
<5
NA
<5
NA
<5
<1
<3
1 The concentration of 980 ppb MTBE was achieved by spiking the Raw Southern California Water which contained 54 ppb MTBE.
2 The concentration of 1033 ppb bromide was achieved by spiking the Raw Southern California Water which contained 321 ppb bromide with additional sodium bromide.
3 The requirement was to achieve an effluent concentration of <3.0 ppb.
4 The concentration of 28 ppb MTBE was achieved by spiking the New York Raw Water which contained 6.6 ppb MTBE with additional MTBE.
5 The concentration of 90 ppb MTBE was achieved by spiking the New York Raw Water which contained 6.6 ppb MTBE with additional MTBE.
6 Operating at 800 gpm.
23
-------
Table 6-3. TEA Destruction and Acetone Reduction at other Locations with HiPOx Technology
(and Bioreactor Technology).
Location (Month/Year)
Tustin, CA (March 2002)
Tustin, CA (March 2002)
Tustin, CA (March 2002)
Turlock, CA (May 2002)
Turlock, CA (May 2002)
Turlock, CA (May 2002)
Contaminant Concentration (ppb)
Sample Location
Influent
HiPOx Effluent
Bioreactor Effluent
Influent
HiPOx Effluent
Bioreactor Effluent
MTBE
17,000
<2
<2
3,400
<0.5
<0.5
TEA
200
<4
<4
220
<5
<5
Acetone
110
680
8.2*
<200
110
<10
Trip blank acetone: 5.7ppb.
24
------- |