United States
Environmental Protection
Agency
Office of Research and
Development
Washington DC 20460
EPA/540/R-97/503
May 1997
Matrix Photocatalytic, Inc
Photocatalytic
Technology
Innovative Technology
Evaluation Report
SUPERFUND INNOVATIVE
TECHNOLOGY EVALUATION
-------
EPA/540/R-97/503
May 1997
Matrix Photocatalytic,
Inc. Photocatalytic
Oxidation Technology
Innovative Technology
Evaluation Report
National Risk Management Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
-------
Notice
This document was prepared for the U.S. Environmental Protection Agency's
(EPA) Superfund Innovative Technology program under Contract No. 68-C5-0037.
This document was subjected to EPA's peer and administrative reviews and was
approved for publication as an EPA document. Mention of trade names or
commercial products does not constitute an endorsement or recommendation for
use.
-------
Foreword
The U.S. Environmental Protection Agency is charged by Congress with protecting
the Nation's land, air, and water resources. Under a mandate of national environ-
mental 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 is the Agency's center for
investigation of technological and management approaches for reducing risks
from threats to human health and the environment. The focus of the Laboratory's
research program is on methods for the prevention and control of pollution to air,
land, water and subsurface resources; protection of water quality in public water
systems; remediation of contaminated sites and groundwater; and prevention and
control of indoor air pollution. The goal of this research effort is to catalyze
development and implementation of innovative, cost-effective environmental tech-
nologies; develop scientific and engineering information needed by EPA to support
regulatory and policy decisions; and provide technical support and information
transfer to ensure effective implementation of environmental regulations and
strategies.
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
-------
Abstract
This report evaluates a photocatalytic oxidation technology's ability to destroy volatile organic
compounds (VOC) and other contaminants present in liquid wastes. Specifically, this report
discusses performance and economic data from a Superfund Innovative Technology Evalua-
tion (SITE) demonstration and one case study of the technology.
The photocatalytic oxidation technology was developed by Matrix Photocatalytic, Inc. (Matrix).
This technology involves exposing titanium dioxide (TiOJ particles to ultraviolet (UV) light
having a predominant wavelength of 254 nanometers, me TiO2 is activated by UV light to
produce highly oxidizing hydroxyl radicals. Matrix also uses hydrogen peroxide and ozone to
enhance the treatment system's performance. Target organic compounds are either mineral-
ized or broken down into low molecular weight organic compounds, primarily by hydroxyl
radicals.
The Matrix technology was demonstrated over a 2-week period in August and September 1995
at the K-25 Site of the U.S. Department of Energy Oak Ridge Reservation in Oak Ridge,
Tennessee. The Matrix system used for the SITE demonstration is housed in an 8- by 20-foot
mobile trailer and is rated for minimum and maximum flow rates of 1 and 2.4 gallons per
minute, respectively. During the demonstration, the Matrix system treated about 2,800 gallons
of K-25 Site groundwater contaminated with more than 30 VOCs. The principal groundwater
contaminants were 1,1-diehloroethane (DCA) and 1,1,1-trichloroethane (TCA), which were
present in K-25 Site groundwater at concentrations up to about 840 and 980 micrograms per
liter (jig/L), respectively. The groundwater also contained low concentrations of total xylenes;
toluene; cis-1,2-dichIoroethene (DCE); and 1,1-DCE at concentrations up to about 200, 85,
100, and 165 gg/L, respectively. Although groundwater alkalinity ranged from 270 to 295
milligrams per liter as calcium carbonate, groundwater did not require pH adjustment prior to
treatment by the Matrix system.
During the technology demonstration, groundwater was spiked with trichloroethene (TCE),
tetrachloroethene (PCE), and benzene. After spiking, the concentrations of these spiking
compounds ranged from about 125 to 1,120 jig/L in Matrix system influent. PCE, TCE, and
benzene were selected as spiking compounds because they are present in groundwater at
many Superfund sites but are not present in K-25 Site groundwater at significant concentra-
tions.
Seven test runs were performed during the demonstration using the spiked groundwater to
evaluate Matrix system performance under different operating conditions. In general, high
percent removals (PR) of up to 99.9 percent were observed for benzene; toluene; xylenes;
TCE; PCE; cis-1,2-DCE; and 1,1-DCE. However, low PRs were observed for 1,1-DCA and
1,1,1-TCA (the highest PRs for 1,1-DCA and 1,1,1-TCA were 40 and 21, respectively). System
effluent met the Safe Drinking Water Act maximum contaminant levels (MCL) for benzene; cis-
1,2-DCE; and 1,1-DCE. However, the effluent did not meet the MCLs for PCE; TCE; 1,1-DCA;
and 1,1,1-TCA. VOC PRs were generally reproducible for most VOCs when the Matrix system
was operated under identical conditions. Treatment by the Matrix system did not reduce the
groundwater toxicity to fathead minnows and water fleas. Purgeable organic carbon and total
organic halide removals of up to 92 and 50 percent, respectively, suggest that some VOCs
were mineralized. However, the formation of aldehydes and haloacetic acids indicated that not
all VOCs were completely mineralized.
Potential sites for applying this technology include Superfund and other hazardous waste sites
where groundwater or other liquid wastes are contaminated with organic compounds. Eco-
nomic data indicate that groundwater remediation costs for the Matrix system used for the SITE
demonstration would be about $65 per 1,000 gallons treated. Of these costs, Matrix system
direct costs would be about $28 per 1,000 gallons treated.
iv
-------
Contents
Page
Notice ii
Foreword , Hi
Abstract .....iv
Figures...... viii
Tables ix
Acronyms, Abbreviations, and Symbols x
Conversion Factors xiii
Acknowledgments xiv
Executive Summary 1
1 Introduction 5
1.1 Brief Description of SITE Program and Reports 5
1.1.1 Purpose, History, and Goafs of the SITE Program 5
1.1.2 Documentation of SITE Demonstration Results.... 6
1.2 Purpose and Organization of the ITER 6
1.3 Background Information on Matrix Technology under the SITE Program 7
1.4 Technology Description 7
1.4.1 Process Chemistry 7
1.4.2 Matrix Treatment System 9
1.4.3 Innovative Features of the Technology 11
1.5 Applicable Wastes 11
1.6 Key Contacts 11
2 Technology Effectiveness and Applications Analysis 13
2.1 Overview of Matrix Technology SITE Demonstration 13
2.1.1 Project Objectives 15
2.1.2 Demonstration Approach 15
2.1.3 Sampling and Analytical Procedures 17
2.2 SITE Demonstration Results 17
2.2.1 Critical VOC PRs under Different Operating Conditions 18
2.2.2 Compliance with Applicable Target Effluent Levels 23
2.2.3 Effect of Treatment on Groundwater Toxicity 24
2.2.4 Reproducibility of Treatment System Performance 25
2.2.5 Treatment By-Products and Additional Parameters 26
2.2.6 Operating Problems 28
-------
Contents (Continued)
Section Page
2.3 Additional Performance Data 28
2.4 Factors Affecting Performance 28
2.4.1 Influent Characteristics 28
2.4.2 Operating Parameters 29
2.4.3 Maintenance Requirements 29
2.5 Site Characteristics and Support Requirements 30
2.5.1 Site Access, Area, and Preparation Requirements 30
2.5.2 Climate 30
2.5.3 Utility and Supply Requirements 30
2.5.4 Required Support Systems 30
2.5.5 Personnel Requirements 31
2.6 Material Handling Requirements 31
2.7 Technology Limitations 31
2.8 Potential Regulatory Requirements 31
2.8.1 Comprehensive Environmental Response, Compensation, and Liability
Act 32
2.8.2 Resource Conservation and Recovery Act 33
2.8.3 Clean Water Act 34
2.8.4 Safe Drinking Water Act 34
2.8.5 Clean Air Act 34
2.8.6 Toxic Substances Control Act 35
2.8.7 Atomic Energy Act and Resource Conservation and Recovery Act 35
2.8.8 Occupational Safety and Health Administration Requirements 35
2.9 State and Community Acceptance 35
3 Economic Analysis 36
3.1 Introduction 36
3.2 Issues and Assumptions 36
3.2.1 Site-Specific Factors 36
3.2.2 Equipment and Operating Parameters 38
3.2.3 Financial Calculations 39
3.2.4 Base-Case Scenario Premises and Assumptions 39
VI
-------
Contents (Continued)
Section Page
3.3 Cost Categories 39
3.3.1 Site Preparation Costs 40
3.3.2 Permitting and Regulatory Costs , 40
3.3.3 Mobilization and Startup Costs , ..„.....,„., 41
3.3.4 Equipment Costs 41
3.3.5 Labor Costs 41
3.3.6 Supplies Costs 42
3.3.7 Utilities Costs , 43
3.3.8 Effluent Treatment and Disposal Costs 43
3.3.9 Residual Waste Shipping and Handling Costs 43
3.3.10 Analytical Services Costs 43
3.3.11 Equipment Maintenance Costs , 44
3.3.12 Site Demobilization Costs 44
3.4 Conclusions of Economic Analysis 44
4 Technology Status 45
5 References 46
Appendix
A Vendor's Claims for the Technology 48
B Case Study 50
VI!
-------
Figures
Figure Page
1-1 Simplified TiO2 Photocatalytic Mechanism 8
1-2 Single Photocatalytic Cell Profile of Matrix Treatment System 10
1-3 Flow Configuration in a Matrix Wafer 10
1-4 Flow Configuration in the Matrix System 12
2-1 Matrix Technology Demonstration Layout 14
2-2 PRs at Various Path Lengths for Critical Aromatic VOCs 19
2-3 PRs at Various Path Lengths For Critical Unsaturated VOCs 19
2-4 PRs at Various Path Lengths for Critical Saturated VOCs 20
2-5 PRs at Equivalent CTs for Critical Aromatic VOCs 20
2-6 PRs at Equivalent CTs for Critical Unsaturated VOCs 21
2-7 PRs at Equivalent CTs for Critical Saturated VOCs 21
3-1 Distribution of One-Time and Annual O&M Costs for a Groundwater
Remediation Project 45
3-2 Distribution of One-Time and Annual O&M Matrix Treatment System
Direct Costs 49
3-3 Distribution of Matrix Treatment System Direct Costs per 1,000
Gallons Treated 50
viii
-------
Tables
Table Page
ES-1 Superfund Feasibility Evaluation Criteria for the Matrix Technology ,. 4
1-1 Correlation Between Superfund Feasibility Evaluation Criteria and
ITER Sections 6
1-2 Comparison of Technologies for Treating VOCs in Water 12
2-1 Target Effluent Levels for Critical VOCs 15
2-2 Demonstration Approach and Relationship of Runs to Project Objectives 16
2-3 Critical VOC Concentrations in Matrix System Influent 18
2-4 PRs for Critical VOCs in Run 1 (No Oxidants) and Run 2 (Oxidants) 22
2-5 Effluent Compliance with Applicable Target Effluent Levels 23
2-6 Acute Toxicity Data 24
2-7 Reproducibiiity Run VOC PRs 25
2-8 Mean, LCL, and UCL Values for Critical VOC PRs at PL 48 in
Reproducibiiity Runs 26
2-9 Haloacetic Acid and Aldehyde Concentrations , 27
2-10 TIC, TOC, POC, and TOX Concentrations , 27
2-11 Summary of Applicable Regulations 32
3-1 Costs Associated with the Matrix Technology 37
3-2 Matrix Treatment System Direct Costs 46
3-3 Costs Associated with the Matrix Technology for Projects Lasting 30 Years 47
IX
-------
Acronyms, Abbreviations, and Symbols
ACL
AEA
AOP
APHA
ARAR
ATL
CAA
CaCO3
CDEP
CERCLA
CFR
C02
CT
CWA
DCA
DCE
DOE
DS1TMS
ECHOS
EPA
gpm
H20
HA
IEA
ITER
kW
kWh
LCL
LDR
Matrix
MCL
Means
mg/L
Alternate concentration limit
Atomic Energy Act
Advanced oxidation process
American Public Health Association
Applicable or relevant and appropriate requirement
Aquatic Testing Laboratories
Clean Air Act
Calcium carbonate
Connecticut Department of Environmental Protection
Comprehensive Environmental Response,
Compensation, and Liability Act of 1980
Code of Federal Regulations
Carbon dioxide
Contact time
Clean Water Act
Dichloroethane
Dichloroethene
U.S. Department of Energy
Direct sampling, ion-trap mass spectrometer
Environmental Cost Handling Options and Solutions
U.S. Environmental Protection Agency
Gallons per minute
Water
Hydrogen peroxide
Irreversible electron acceptor
Innovative technology evaluation report
Kilowatt
Killowatt-hour
Lethal concentration at which 50 percent of the test
organisms die
Lower confidence limit
Land Disposal Restriction
Matrix Photocatalytic, Inc.
Maximum contaminant level
R.S. Means
Milligram per liter
-------
Aeroynms, Abbreviations, and Symbols (Continued)
MS/MSD
NAAQS
nm
NPDES
NRMRL
NSPS
O&M
02
o--
OH-
ORD
ORR
OSHA
OSWER
PAH
PCB
PCE
PFS
POC
PPE
ppm
PR
PRO
PRQL
PVC
QA/QC
QAPP
Quanterra
RCRA
RPD
RSD
SARA
SDWA
SITE
Matrix spike and matrix spike duplicate
National Ambient Air Quality Standard
Nanometer
National Pollutant Discharge Elimination System
National Risk Management Research Laboratory
New Source Performance Standard
Operation and maintenance
Oxygen
Superoxide ion
Ozone
Hydroxide ion
Hydroxyl radical
Office of Research and Development
Oak Ridge Reservation
Occupational Safety and Health Administration
Office of Solid Waste and Emergency Response
Polynuclear aromatic hydrocarbon
Polychlorinated biphenyl
Tetrachloroethene
Precipitation/flocculation/sedimentation
Purgeable organic carbon
Personal protection equipment
Part per million
Percent removal
PRO Environmental Management, Inc.
Project-required quantitation limit
Polyvinyl chloride
Quality assurance and quality control
Quality assurance project plan
Quanterra Environmental Services, Inc.
Resource Conservation and Recovery Act of 1976
Relative percent difference
Relative standard deviation
Superfund Amendments and Reauthorlzation Act of 1986
Safe Drinking Water Act
Superfund Innovative Technology Evaluation
xl
-------
Acronyms, Abbreviations, and Symbols (Continued)
TCA Trichloroethane
TOE Trichloroethene
TER Technology evaluation report
TIC Total Inorganic carbon
TiO2 Titanium dioxide
TOC Total organic carbon
TOX Total organic halides
TSCA Toxic Substances Control Act
TSS Total suspended solids
TU, Acute toxicity unit
UCL Upper confidence limit
UV Ultraviolet
UV-A Ultraviolet A
UV-C Ultraviolet C
VOC Volatile organic compound
WQS Water quality standard
u.g/L Microgram per liter
> Greater than
< Less than
XII
-------
Conversion Factors
To Convert From
la
Multiply Bv
Length:
Area:
Volume:
Mass:
Energy:
Power:
Temperature:
inch
foot
mile
square foot
acre
gallon
cubic foot
pound
kilowatt-hour
horsepower
(Fahrenheit - 32)
centimeter
meter
kilometer
square meter
square meter
liter
cubic meter
kilogram
megajoule
kilowatt
Celsius
2.54
0.305
1.61
0.0929
4,047
3.78
0.0283
0.454
3.60
0.746
0.556
XIII
-------
Acknowledgments
This report was prepared under the direction and coordination of Mr. Richard
Eilers, U.S. Environmental Protection Agency (EPA) Superfund Innovative
Technology Evaluation (SITE) program project manager of the National Risk
Management Research Laboratory (NRMRL) in Cincinnati, Ohio. Contributors
and reviewers for this report were Mr. Gordon Evans, Dr. John Ireland, Ms. Ann
Kern, and Ms. Norma Lewis of EPA NRMRL, Cincinnati, Ohio; Mr. Robert
Henderson of Matrix Photocatalytic, Inc., London, Ontario; Ms. Elizabeth
Fiedler of Lockheed Martin Energy Systems, Inc., Oak Ridge, Tennessee; and
Ms. Elizabeth Phillips of the U.S. Department of Energy, Oak Ridge, Tennes-
see.
This report was prepared for EPA's SITE program by Dr. Kirankumar Topudurti,
Ms. Mary Wojciechowski, Ms. Sandy Anagnostopoulos, and Mr. Jeffrey Swano
of PRC Environmental Management, Inc. (PRC). Mr. Ted Tharp and Ms. Kim
Taibot of PRC developed and managed analytical data spreadsheets. Special
acknowledgment is given to Dr. Harry Ellis, Mr. Stanley Labunski,. Ms. Shelley
Fu, Mr. Gary Sampson, and Ms. Jeanne Kowalski of PRC for their technical,
quality control, editorial, graphic, and production assistance, respectively,
during the preparation of this report.
XIV
-------
Executive Summary
The photocatalytic oxidation technology developed by
Matrix Photocatalytic, Inc. (Matrix), can destroy organic
compounds in liquid wastes. This technology was
demonstrated under the U.S. Environmental Protection
Agency's Superfund Innovative Technology Evaluation
(SITE) program. The technology demonstration was
conducted overa2-week period in August and September
1995 at the K-25 Site of the U.S. Department of Energy
(DOE) Oak Ridge Reservation in Oak Ridge, Tennessee.
The purpose of this innovative technology evaluation
report (ITER) is to present information that will assist
Superfund decision-makers in evaluating the Matrix
photocatalytic oxidation technology for application to a
particular hazardous waste site cleanup. The report
provides an introduction to the SITE program and Matrix
technology (Section 1), analyzes the technology's
effectiveness and applications (Section 2), analyzes the
economics of usi ng the Matrix system to treat groundwater
contaminated with volatile organic compounds (VOC)
(Section 3), summarizes the technology's status (Section
4), and presents a list of references used to prepare the
ITER (Section5). Vendor'sclaimsfortheMatrixtechnology
are presented in Appendix A, and a case study of the
technology application performed in Canada is
summarized in Appendix B.
This executive summary briefly describes the Matrix
technology and system, provides an overview of the
SITE demonstration of the technology, summarizes the
SITE demonstration results, discusses the economics of
using the Matrix systemtotreatgroundwatercontaminated
with VOCs, and discusses the Superfund feasibility
evaluation criteria for the Matrix technology.
Technology and System Description
The Matrix technology involves the exposure of titanium
dioxide (TiOa) particles to ultraviolet (UV) light having a
predominant wavelength of 254 nanometers (nm). The
TiO2 is activated by UV light to produce highly oxidizing
hydroxyl radicals. Matrix also uses hydrogen peroxide
(H2O2) and ozone (Os) to enhance the treatment system's
performance. Target organic compounds are either
mineralized or broken down into low molecular weight
compounds, primarily by hydroxyl radicals.
The basic component of a Matrix system is a photocatalytic
reactor cell. Each cell measures 5.75 feet in length and
has a 1.75-inch outside diameter. A 75-watt, 254-nm UV
light source is located coaxially within a 5.4-foot long
quartz sleeve. The quartz sleeve is surrounded by multiple
layers of fiberglass mesh bonded with the anatase form
of TiO2.
The Matrix system used for the SITE demonstration is
housed in an 8- by 20-foot mobile trailer and is rated for
minimum and maximum flow rates of 1 and 2.4 gallons
per minute (gpm), respectively. The system consists of
two units positioned side by side in the trailer. Each unit
consists of 12 wafers, and each wafer consists of six
photocatalytic reactor cells joined by manifolds. Matrix
placed a block in each wafer so that contaminated
groundwater flowed in parallel mode into three reactor
cells at a time. Each set of three cells along the path
where contaminated groundwater flows is defined as a
path length. During the demonstration, H2O2 was injected
at path lengths 1, 9, 17, 25, 33, and 41 and O3 was
injected at path length 17. The Matrix system does not
have any vents and does not generate air emissions.
The Matrix technology is applicable for treatment of
VOCs, semivoiatile organic compounds, polynuclear
aromatic hydrocarbons, and pesticides in liquid wastes,
including groundwater, wastewater, landfill leachate, and
drinking water.
Overview of the Matrix Technology SITE
Demonstration
The Matrix technology was demonstrated over a 2-week
period in August and September 1995 at the K-25 Site of
DOE's Oak Ridge Reservation in Oak Ridge, Tennessee.
During the demonstration, the Matrix system treated
about 2,800 gallons of K-25 Site groundwater
contaminated with VOCs. The principal groundwater
contaminants were 1,1-dichIoroethane (DCA) and 1,1,1-
triehloroethane (TCA), which were present at
concentrations up to about 840 and 980 micrograms per
liter (p.g/L), respectively. The groundwater also contained
low concentrations of total xylenes; toluene; cis-1,2-
dichloroethene (DCE); and 1,1-DCE at concentrations
up to about 200, 85, 100, and 165 jig/L, respectively.
Although groundwater alkalinity ranged from 270 to 295
milligrams per liter (mg/L) as calcium carbonate,
groundwater did not require pH adjustment prior to
treatment by the Matrix system.
1
-------
During the technology demonstration, groundwater was
spiked with trichloroethene (TCE), tetrachloroethene
(PCE), and benzene. After spiking, the concentrations of
these spiking compounds ranged from about 125 to
1,120 pg/L in Matrix system influent. PCE, TCE, and
benzene were selected as spiking compounds because
they are present in groundwater at many Superfund sites
but are not present in K-25 Site groundwater at significant
concentrations.
For the SITE technology demonstration, 1,1-DCA; 1,1,1-
TCA; total xylenes; cis-1,2-DCE; and the spiking
compounds were considered critical VOCs. The VOCs
1,1 -DCE and toluene were not considered critical because
during the planning stages of the demonstration, available
data did not indicate that 1,1 -DCE or toluene was present
at significant concentrations in K-25 Site groundwater.
However, during the SITE demonstration, 1,1 ,-DCE and
toluene were found to be present at significant
concentrations.
The primary objectives of the technology demonstration
were as follows:
Determine percent removals (PR) for critical VOCs
in groundwater achieved by the Matrix treatment
system under different operating conditions (by vary-
ing flow rate, number of path lengths, and O3 and
H2O2 doses)
» Determine whether the Matrix treatment system
effluent meets maximum contaminant levels (MCL)
promulgated under the Safe Drinking Water Act
(SDWA) for the critical VOCs at a significance level
of 0.05
• Evaluate the change in acute toxicity of groundwa-
ter (measured as the lethal concentration [expressed
as percent sample] at which 50% of test organisms
die after treatment by the Matrix system at a signifi-
cance level of 0.05
Evaluate the reproducibility of the Matrix treatment
system performance in terms of PRs for critical
VOCs and its ability to meet applicable target efflu-
ent levels for the critical VOCs
Estimate costs for the Matrix system to treat ground-
water contaminated with VOCs
Tnesecondary objectives of the technology demonstration
were as follows:
« Document the concentrations of potential treatment
by-products in groundwater (for example, haloacetic
acids and aldehydes) formed by the Matrix treat-
ment system
Determine PRs for noncritical VOCs in groundwa-
ter achieved by the Matrix system under different
operating conditions (by varying flow rate, number
of path lengths, and O3 and H2O2 doses)
Document observed operating problems and their
resolutions
During the demonstration, seven test runs were conducted
using spiked groundwater under different system
operating conditions to evaluate the Matrix system in
accordance with the project objectives. The operating
parameters varied include influent flow rate, path length,
and Os and h^Oa doses. The demonstration also included
three test runs (Runs 5, 6, and 7) to evaluate the
reproducibility of the system's performance at the
technology developer's preferred operating conditions.
Because K-25 Site groundwater contained high
concentrations of iron and manganese (16 and 9,9 mg/L,
respectively), the groundwater was pretreated using an
ion-exchange system to prevent fouling of the
photocatalytic reactor cells. The pretreatment system
also included a 3-micron cartridge filter system to remove
solids so that the ion-exchange columns would not clog.
During the demonstration, groundwater samples were
collected at Matrix system influent, intermediate, and
effluent sampling locations. These samples were analyzed
for VOCs, acute toxicity, aldehydes, haloacetic acids,
total inorganic carbon, total organic carbon, purgeable
organic carbon (POC), total organic halides (TOX),
alkalinity, metals, pH, turbidity, total suspended solids,
temperature, Oa (treated groundwater only), and HaOa
(treated groundwater only). Process chemical
concentrations (O3 and HgGa), system flow rates, and
electrical energy consumption were also measured to set
system operating conditions and gather information
needed to estimate treatment costs.
SITE Demonstration Results
Key findings of the Matrix technology are listed below:
In general, high PRs (up to 99.9%) were observed
for both aromatic VOCs (benzene, toluene, and
total xylenes) and unsaturated VOCs (PCE, TCE,
1,1-DCE, and cis-1,2-DCE). However, the PRs for
saturated VOCs were low (the highest PRs for 1,1-
DCA and 1,1,1 -TCA were 40 and 21, respectively).
• The PRs for all VOCs increased with increasing
number of path lengths and oxidant doses. At
equivalent contact times, changing the flow rate did
not appear to impact the treatment system perfor-
mance for all aromatic VOCs and most unsaturated
VOCs (except 1,1-DCE). Changing the flow rat©
appeared to impact the system performance for
saturated VOCs.
The effluent met the SDWA MCLs for benzene; cis-
1,2-DCE; and 1,1-DCE at a significance level of
0.05. However, the effluent did not meet the MCLs
for PCE; TCE; 1,1-DCA; and 1,1,1-TCA at a signifi-
cance level of 0.05. The influent concentrations for
toluene and total xylenes were below the MCLs.
-------
• In tests performed to evaluate the effluent's acute
toxicity to water fleas and fathead minnows, more
than 50% of the organisms died. Treatment by the
Matrix system did not reduce the groundwater toxic-
ity for the test organisms at a significance level of
0.05.
• In general, the PRs were reproducible for aromatic
and unsaturated VOCs when the Matrix system was
operated under identical conditions. However, the
PRs were not reproducible for saturated VOCs. The
Matrix system's performance was generally repro-
ducible in (1) meeting the target effluent levels for
benzene; cis-1,2-DCE; and 1,1-DCE and (2) not
meeting the target effluent levels for PCE; TCE; 1,1-
DCA;and 1,1,1-TCA.
• POC and TOX results indicated that some VOCs
were mineralized in the Matrix treatment system.
However, formation of aldehydes (formaldehyde,
acetaldehyde, propanal, butanal, glyoxal, and me-
thyl glyoxal), haloacetic acids (mono- and
dichloroacetic acids), and several tentatively identi-
fied compounds indicated that not all VOCs were
completely mineralized.
• Several problems were experienced during the SITE
demonstration. Some of these problems involved (1)
the system's inability to maintain a steady flow at the
anticipated minimum flow rate of 0.5 gpm, (2) the
system's inability to inject O3 at multiple path lengths,
and (3) frequent breakage of the quartz tubes. Al-
though these problems resulted in significant down-
time, Matrix resolved these problems and the SITE
demonstration was completed on schedule.
Economics
Based on information obtained from the SITE
demonstration, Matrix, and other sources, an economic
analysis was performed to examine 12 separate cost
categories under hypothetical cases in which three Matrix
systems were assumed to treat about 28.4 million gallons
of contaminated groundwater at a Superfund site.
Groundwater characteristics and Matrix system
performance were assumed to be similar to that observed
during the SITE demonstration. The costs of using 2-,
12-, and 24-gpm Matrix systems were estimated. The
estimated cost under the base-case scenario of an 11-
kilowatt (kW) system operating at 2-gpm flow rate for 30
years with an annual downtime of 10% is summarized
below.
The total direct costs related to procuring and operating
the Matrix system are estimated to be $28.53 per 1,000
gallons treated. Of these costs, the three largest cost
categories are supplies, utilities, and equipment
maintenance costs. Specifically, supplies, utilities, and
equipment maintenance costs represent 47, 25, and
10% of the total direct costs, respectively.
The estimated cost for a 65-kW system operating for 5
years at 12 gpm with an annual downtime of 10% to
remediate 28.4 million gallons of contaminated
groundwater is $42.96 per 1,000 gallons treated. The
estimated cost for a 130-kW system operating for 2.5
years at 24 gpm with an annual downtime of 10% to
remediate 28.4 million gallons of contaminated
groundwater is $50.76 per 1,000 gallons treated.
Superfund Feasibility Evaluation Criteria
for the Matrix Technology
Table ES-1 briefly discusses the Superfund feasibility
evaluation criteria for the Matrix technology to assist
Superfund decision-makers considering the technology
for remediation of contaminated groundwater at
hazardous waste sites.
-------
Tablo ES-1. Superfund Feasibility Evaluation Criteria for the Matrix Technology
Criterion Discussion
Overall Protection of
Human Health and the
Environment
Compliance with
Applicable or Relevant
and Appropriate
Requirements (ARAR)
Long-Term
Effectiveness and
Permanence
Reduction of Toxfcily,
Mobility, or Volume
Through Treatment
Short-Term
Effectiveness
Implementability
Coat
State Acceptance
Community Acceptance
The Matrix technology is expected to protect human health by significantly lowering the concentrations
of aromatic and unsaturated VOCs in treated water. The technology's ability to treat water contaminated
with saturated VOCs is questionable.
Overall reduction of human health risk should be evaluated on a site-specific basis because of the
potential for formation of harmful treatment by-products such as aldehydes and haloacetic acids.
The technology protects the environment by curtailing migration of contaminated groundwater.
Protection of the environment at and beyond the point of treated water discharge should be evaluated
based on uses of the receiving water body, concentrations of residual contaminants and treatment by-
products, and the dilution factor of the receiving water body.
The technology has the potential to comply with existing federal, state, and local ARARs (for example,
MCLs) for several organic contaminants (for example, aromatic VOCs and unsaturated VOCs).
However, the technology's ability to meet federal, state, and local ARARs for some organic
contaminants (for example, saturated VOCs) is questionable.
The technology's ability to meet any future chemical-specific ARARs for by-products should be
considered because of the potential for formation of aldehydes and haloacetic acids during treatment,
The technology's ability to meet any state or local toxicity-related requirements such as bioassay tests
should be considered because of the potential for treatment by-product formation.
Human health risk can be reduced to acceptable levels by treating groundwater to a 10-6 cancer risk
level. The time needed to achieve cleanup goals depends primarily on contaminated aquifer
characteristics.
The technology can effectively control groundwater contaminant migration because it is operated in a
pump-and-treat mode.
The treatment achieved is permanent because photocatalytic oxidation is a destruction technology.
Periodic review of treatment system performance is needed because application of the technology to
contaminated groundwater at hazardous waste sites is fairly new.
Although contaminants are destroyed by the technology, the reduction in overall toxicity should be
determined on a site-specific basis because of the potential for formation of by-products (for
example, aldehydes, and haloacetic acids).
The technology reduces the volume and mobility of contaminated groundwater because it is operated in
a pump-and-treat mode.
During the pump-and-treat operation, aquifer drawdown may impact vegetation in the treatment zone.
The technology can be implemented using a mobile, transportable, or permanent treatment system.
State and local permits must be obtained to operate the Matrix system. A National Pollutant Discharge
Elimination System permit is usually required to implement the technology,
Treatment costs vary significantly depending on the size of the treatment system used, contaminant
characteristics and levels, cleanup goals, the volume of contaminated water to be treated, and the length
of treatment. For the K-25 Site groundwater cleanup operation, the treatment cost is expected to be $28
to $51 per 1,000 gallons of contaminated water treated.
This criterion is generally addressed in the record of decision. State acceptance of the technology will
likely depend on the concentrations of residual organic contaminants and treatment by-products in treated
water and the toxicity of treated water.
This criterion is generally addressed in the record of decision after community responses are received
during the public comment period. Because communities are not expected to be exposed to harmful
levels of noise or fugitive emissions, the level of community acceptance of the technology is
expected to be high.
-------
Section 1
Introduction
This section briefly describes the Superfund Innovative
Technology Evaluation (SITE) program and SITE reports;
states the purpose and organization of this innovative
technology evaluation report (ITER); provides background
information on the development of the Matrix
Photocatalytic, Inc. (Matrix), photocatalytic oxidation
technology underthe SITE program; describes the Matrix
photocatalytic oxidation technology; identifies wastes to
which this technology may be applied; and provides a list
of key contacts that can supply information about the
technology and demonstration site.
1.1 Brief Description of SITE Program
and Reports
This section provides information about the purpose,
history, and goals of the SITE program and about reports
that document SITE demonstration results.
1.1.1 Purpose, History, and Goals of the
SITE Program
The primary purpose of the SITE program is to advance
the development and demonstration, and thereby
establish the commercial availability, of innovative
treatmenttechnologies applicable to Superfund and other
hazardous waste sites. The SITE program was
established by the U.S. Environmental Protection Agency
(EPA) Office of Solid Waste and Emergency Response
(OSWER) and Office of Research and Development
(ORD) in response to the Superfund Amendments and
Reauthorization Act of 1986 (SARA), which recognized
the need for an alternative or innovative treatment
technology research and demonstration program. The
SITE program is administered by ORD's National Risk
Management Research Laboratory (NRMRL). The overall
goal of the SITE program is to carry out a program of
research, evaluation, testing, development, and
demonstration of alternative or innovative treatment
technologies that may be used in response actions to
achieve long-term protection of human health and welfare
and the environment.
The SITE program consists of fourcomponent programs:
(1)the Emerging Technology program, (2) the
Demonstration program, (3) the Monitoring and
Measurement Technologies program, and (4) the
Technology Transfer program. This ITER was prepared
under the SITE Demonstration program. The objective of
the Demonstration program is to provide reliable
performance and cost data on innovative technologies
so that potential users can assess a given technology's
suitability for specific site cleanups. To produce useful
and reliable data, demonstrations are conducted at actual
hazardous waste sites or under conditions that closely
simulate actual waste site conditions.
Data collected during the demonstration are used to
assess the performance of the technology, the potential
need for pretreatment and post-treatment processing of
the treated waste, the types of wastes and media that can
be treated by the technology, potential treatment system
operating problems, and approximate capital and
operating costs. Demonstration data can also provide
insight into a technology's long-term operation and
maintenance (O&M) costs and long-term application
risks.
Under each SITE demonstration, a technology's
performance in treating an individual waste at a particular
site is evaluated. Successful demonstration of a
technology at one site does not ensure its success at
other sites. Data obtained from the demonstration may
require extrapolation to estimate a range of operating
conditions over which the technology performs
satisfactorily. Also, any extrapolation of demonstration
data should be based on other information about the
technology, such as case study information.
Implementation of the SITE program is a significant,
ongoing effort involving ORD, OSWER, various EPA
regions, and private business concerns, including
technology developers and parties responsible for site
remediation. The technology selection process and the
Demonstration program together provide a means to
perform objective and carefully controlled testing of field-
ready technologies. Innovative technologies chosen for
a SITE demonstration must be pilot- or full-scale
applications and must offer some advantage over existing
technologies. Mobile technologies are of particular
interest. Each year, the SITE program sponsors about 10
technology demonstrations.
-------
1.1.2 Documentation of SITE
Demonstration Results
The results of each SITE demonstration are usually
reported in four documents: the demonstration bulletin,
technology capsule, technology evaluation report (TER),
and ITER.
The demonstration bulletin provides a two-page
description of the technology and project history,
notification that the demonstration was completed, and
highlights of the demonstration results. The technology
capsule provides a brief description of the project and an
overview of the demonstration results and conclusions.
Because of budget restrictions, the demonstration bulletin
and technology capsule may not be prepared for the
Matrix technology demonstration.
The purpose of the TER is to consolidate all information
and records acquired during the demonstration. It contains
both a narrative portion and tables and graphs
summarizing data. The narrative portion discusses
predemonstration, demonstration, and postdemonstratipn
activities, any deviations from the demonstration quality
assurance project plan (QAPP) during these activities,
and the impact of such deviations, if applicable. The TER
data tables and graphs summarize test results in terms of
whether project objectives and applicable or relevant and
appropriate requirements (ARAR) were met. The tables
also summarize quality assurance and quality control
(QA/QC) data and data quality objectives. The TER is not
formally published by EPA. Instead, a copy is retained by
the EPA project manager as a reference for responding
to public inquiries and for recordkeeping purposes. The
purpose and organization of the ITER are discussed in
Section 1.2.
In addition to the four documents, a videotape is also
prepared that displays and discusses the technology,
demonstration site, equipment used, tests conducted,
results obtained, and key contacts for information about
the technology. The videotape is typically about
15 minutes long. Again, because of budget restrictions,
avideotape may not be prepared forthe Matrix technology
demonstration.
1.2 Purpose and Organization of the
ITER
Information presented in the ITER is intended to assist
Superfund decision-makers in evaluating specific
technologies for a particular cleanup situation. Such
evaluations typically involve the nine remedial technology
feasibility evaluation criteria, which are listed in Table 1-
1 along with the sections of the ITER where information
related to each criterion is discussed. The ITER represents
a critical step in the development and commercialization
of a treatment technology. The report discusses the
effectiveness and applicability of the technology and
analyzes costs associated with its application. The
technology's effectiveness is evaluated based on data
collected during the SITE demonstration and from other
case studies. The applicability of the technology is
discussed in terms of waste and site characteristics that
could affect technology performance, material handling
requirements, technology limitations, and other factors.
This ITER consists of five sections, including this
introduction. These sections and their contents are
summarized below.
Section 1, Introduction, presents a brief description
of the SITE program and reports, the purpose and
organization of the ITER, background information
about the Matrix technology under the SITE pro-
gram, a technology description, applicable wastes
that can be treated, and key contacts for informa-
tion about the Matrix technology and demonstration
site.
Section 2, Technology Effectiveness and Applica-
tions Analysis, presents an overview of the Matrix
technology SITE demonstration, SITE demonstra-
tion results, additional performance data for the
Matrix system, factors affecting the Matrix system
performance, site characteristics and support re-
quirements, material handling requirements, tech-
nology limitations, potential regulatory requirements,
and state and community acceptance.
Table 1-1. Correlation Between Superfund Feasibility Evaluation Criteria and ITER Sections
Evaluation Criterion8 ITER Section
Overall protection of human health and the
environment
Compliance with ARARs
Long-term effectiveness and permanence
Reduction of toxicity, mobility, or volume through
treatment
Short-term effectiveness
Impiementability
Cost
State acceptance
Community acceptance
2.2.1 through 2.2.5
2.2.2 and 2.2.4
1.4 and 2.2.4
2.2.1 and 2.2.3 through 2.2.5
2.2.1 through 2.2.4
1.4, 2.2, and 2.4
3.0
1.4, 2.2.1 through 2.2.6, and 2.9
1.4, 2.2.1 through 2.2.6, and 2.9
•Source: EPA 1988b
-------
Section 3, Economic Analysis, presents issues and
assumptions, cost categories, and conclusions of
the economic analysis.
Section 4, Technology Status, discusses the devel-
opmental status of the Matrix technology.
Section 5, References, lists references used to
prepare this ITER.
In addition to these sections, this ITER has two
appendixes: Appendix A, Vendor's Claims for the
Technology; and Appendix B, Case Study.
1.3 Background Information on Matrix
Technology under the SITE
Program
The Matrix technology was accepted into the SITE
Emerging Technology program in May 1991. The
Emerging Technology program promotes technology
development by providing funds to developers with bench-
or pilot-scale innovative technology systems to support
continuing research, in 1994, the Matrix technology was
accepted into the SITE Demonstration program. The
technology was demonstrated over a 2-week period in
August and September 1995 at the K-25 Site of the U.S.
Department of Energy (DOE) Oak Ridge Reservation
(ORR) in Oak Ridge, Tennessee.
1.4 Technology Description
This section describes the Matrix technology process
chemistry, the treatment system, and innovative features
of the technology.
1.4.1 Process Chemistry
The Matrix photocataiytic oxidation process is an
advanced oxidation process (AOP) developed by Matrix
to remediate wastewater and groundwater contaminated
with organic pollutants including solvents, pesticides,
polynucleararomatichydrocarbons (PAH), and petroleum
hydrocarbons at ambient temperatures. In general, AOPs
involve hydroxyl radicals (OH.) as oxidants. The OH, can
be generated in aqueous solutions by use of any one of
the following:
Ultraviolet (UV) light and hydrogen peroxide (H2O2)
UV light and ozone (O3)
• 03andH202
UV light and a semiconductor photocatalyst
High-voltage electron beam
Photocataiytic AOPs use UV light in the presence of
oxidants (O3 and H2Os) or with semiconductors to produce
OH-. The Matrix system utilizes UV light, asemiconductor,
and oxidants to generate OH-.
Semiconductors are solids that have electrical
conductivities between those of conductors and those of
insulators. Semiconductors are characterized by two
separate energy bands: a low-energy valence band and
a high-energy conduction band. Each band consists of a
spectrum of energy levels in which electrons can reside.
The separation between energy levels within each energy
band is small, essentially forming acontinuousspectrum.
The energy separation between the valence and
conduction bands is called the "band gap" and consists
of energy levels in which electrons cannot reside.
Light, a source of photons, can be used to excite an
electron from the valence band into the conduction band.
When an electron in the valence band absorbs a photon,
the absorption of the photon increases the energy of the
electron and enables the electron to move into one of the
unoccupied energy levels of the conduction band.
However, because the energy levels of the valence band
are lowerthan those of the conduction band, electrons in
the conduction band will eventually move .back into the
valence band, leaving the conduction band empty. As
this occurs, energy corresponding to the difference in
energy between the bands is released as photons or
heat. Because photons can be used to excite a
semiconductor's electrons and enable easy conduction,
semiconductors are said to exhibit photoconductivity.
Semiconductors that have been studied for commercial
photocataiytic processes include titanium dioxide (TiOa),
strontium titanium trioxide, and zinc oxide. Because of
TiO2's high level of photoconductivity, ready availability,
low toxicity, and low cost, TiOa is generally preferred for
use in commercial AOP applications. TIO2 has three
crystalline forms: rutile, anatase, and brookite. Studies
indicate that the anatase form provides the highest OH-
formation rates (Tanaka and others 1993).
exhibits photoconductivity when illuminated by
photons having an energy level that exceeds the TiO2
band gap energy level of 3.2 electron volts. For TiOa, the
photon energy required to overcome the band gap energy
and excite an electron from the valence to the conduction
band can be provided by UV radiation having a wavelength
between 200 and 385 nanometers (nm). When an electron
in the valence band is excited into the conduction band,
a vacancy or hole is left in the valence band. These holes
have the effect of a positive charge. The combination of
the electron in the conduction band and the hole in the
valence band is referred to as an electron-hole pair.
Because the electron is in an unstable, excited state, the
electron-hole pairs within asemiconductortend to reverse
to a stage where the electron-hole pair no longer exists;
however, the band gap inhibits this reversal long enough
to allow excited electrons and holes near the surface of
the semiconductortoparticipate in reactions at thesurface
of the semiconductor.
Because of the relatively low costs and hazards associated
with UV-A lamps with a predominant wavelength of 350
nm, they are frequently used in TiO2 photocataiytic
oxidation applications. However, according to one study,
photocataiytic reactors that use UV lamps with a
predominant wavelength of 254 nm (UV-C) are more
effective in promoting organic compound destruction
-------
than reactors using UV-A lamps with a predominant
wavelength of 350 nm {Matthews and McEvoy 1992).
One possible reason for the improved performance of
254-nm light is that 254-nm light is strongly absorbed by
TIOz; therefore, the penetration distance of photons is
relatively short, allowing electron-hole pairs to form closer
totheTIOasurface.where contaminant destruction occurs.
Also, many organic molecules are excited by 254-nm
light and as a result may be destroyed solely by 254-nm
HOW.
A simplified TiO2 photocataiytic mechanism is
summarized !n Figure 1 -1. This mechanism is still being
researched, and published research indicates that the
primary photocataiytic mechanism is believed to proceed
as follows (AI-Ekabi and others 1993);
TiO
hv
where
hv
hvu
light energy (photon)
electron in the conduction band
hole in the valence band
At the TiOa surface, the holes either react with water
molecules (HaO) or hydroxide ions (OH~) from water
dissociation to form OH* as follows:
where
h4 +H2O-OH- + H*(i-2)
h^ + OH" - OH»(i-3)
=> proton
An additional reaction may occur where the electron in
the conduction band reacts with dissolved oxygen (Og) in
water to form superoxide ions (Qa"") as follows;
"(1-4)
These CV can then react with
OH-, OH", and Oz as follows:
2O2-
2OH» + 2OH" + O2
provide additional
(1-5)
The OH" then can react with the hole In the valence band
in accordance with Equation 1-3 to form additional OH-.
One practical problem of semiconductorphotoconductivity
is the electron-hole reversal process. The overall result of
this reversal is the generation of photons or heat instead
of OH-, This process significantly decreases the
photocataiytic activity of a semiconductor {AI-Ekabi and
others 1993). One possible method of increasing the
photocataiytic activity of a semiconductor is to add
irreversible electron acceptors (IEA) to the aqueous
Electron-Hole
Pair Reversal
Photon
T1O2 Particle in Water
Figure 1-1. Simplified TiO, Photocataiytic Mechanism.
-------
matrix to be treated. Once lEAs accept an electron in the
conduction band or react with O%~, the lEAs dissociate
and provide additional routes for OH. generation. H2O2 is
an IEA and can illustrate the role lEAs may play in AOPs.
When the IEA H202 accepts an electron in the conduction
band, it dissociates in accordance with the following
reaction:
H2O2
d-6)
H2O2 not only inhibits the electron-hole reversal process
and prolongs the lifetime of the photogenerated hole, it
also generates additional OH-.
O3 is also used as an IEA and may undergo the following
reaction:
2O
2e
CB
02+202.
(1-7)
The O2 and O2r can then react with electrons in the
conduction band and H2O in accordance with Equations
1-4 and 1-5, respectively, to form additional OH-.
Organic compounds can be destroyed by a variety of
reactions with OH-. These reactions include addition,
hydrogen abstraction, electron transfer, and radical-
radical combination.
If sufficient OH- are not generated to completely oxidize
contaminants to carbon dioxide (CO2) and H2O, stable
intermediates may be formed. The types of intermediates
formed depend on the initial levels and types of
contaminants. Studies of TiO2 photocatalytictechnologies
have analyzed potential stable intermediates resulting
from treatment of chlorinated organics. These studies
show- that the photoeatalytic degradation of 1,1,1-
trichioroethane (TCA) yields the stable intermediate
monochloroacetic acid. The photoeatalytic degradation
of trichloroethene (TCE) and tetrachioroethene (PCE)
yields the stable intermediates dichloroaeetic acid and
trichloroacetic acid, respectively. Organic compounds
with double bonds between carbon atoms yield aldehydes
as stable by-products (Glaze and others 1980).
Photocatalytiedegradation of aromatic compounds yields
the stable intermediates acetic acid and formic acid
(Pichat and others 1993). Of these, the haloacetic acids
and aldehydes are considered toxic by-products.
Some compounds commonly present in water may react
with the reactive species formed by the Matrix treatment
system, thereby exerting an additional demand for reactive
species on the system. These compounds are called
scavengers and can potentially impact system
performance. A scavenger is defined as any compound
in water otherthan the target contaminants that consumes
reactive species such as OH-. Carbonate and bicarbonate
ions are examples of OH- scavengers present in most
natural waters and wastewaters. Alkalinity is therefore an
important operating parameter. If alkalinity is high, influent
pH or alkalinity adjustment may be required to shift the
carbonate-bicarbonate equilibrium from carbonate (a
scavenger) to carbonic acid (not a scavenger). Other
potential OH- scavengers include sulfide; nitrite; cyanide
ions; and oxidizable, nontarget or"background"organies.
1.4.2 Matrix Treatment System
The Matrix treatment system used for the demonstration
contains 144 photoeatalytic reactor cells. Each cell
measures 5.75 foot long and has a 1.75-inch outside
diameter. A 75-watt, 254-nm UV light source is located
coaxially within a 5.4-foot long quartz sleeve. The quartz
sleeve is surrounded by seven or eightlayers of fiberglass
mesh bonded with the anatase form of TIO2 and is
enclosed in a stainless steel jacket (see Figure 1-2).
Each cell is rated fora maximum flow rate of approximately
0.8 gallon per minute (gpm).
The Matrix treatment system used for the demonstration
consists of two units positioned side by side in a mobile
trailer. Each unit consists of 12 wafers, and each wafer
consists of six photocataiytic reactor cells joined by
manifolds. Matrix placed a block in each wafer so that
contaminated groundwater flowed in parallel mode into
three reactor cells at a time. The flow configuration in a
wafer is shown in Figure 1-3, The overall maximum flow
rate for this configuration is 2.4 gpm. For the
demonstration, each set of three cells along the path
where the contaminated groundwater flows is defined as
a path length. Therefore, each wafer has two path lengths.
Each unit has 24 path lengths, resulting in a total of 48
path lengths for the two units. The flow configuration for
the Matrix treatment system is shown in Figure 1-4. The
system has a pressure pump after every four wafers to
ensure constant flow in the system. Beginning with the
first wafer, contaminated groundwater enters the first
path length (first set of three reactor cells of Unit 1) and
then the second path length (second set of three reactor
cells of Unit 1), completing treatment in the first wafer.
Contaminated groundwater then flows to the second
wafer and enters the third path length (first set of three
reactor cells of Unit 2) and then the fourth path length
(second set of three reactor cells of Unit 2), and so on,
until it passes through all 24 wafers (48 path lengths).
For the demonstration, the Matrix system was housed in
an 8- by 20-foot secure trailer. The trailer was equipped
with fans to keep the unit cool and contained an area of
approximately 25 square feet for the system technician.
The trailer also housed the Os-generating system.
Because the quantity of Oa to be added to the Matrix
system was small, Matrix used bottles of dry oxygen to
generate 03. This system produced Oa concentrations of
3,000 to 3,700 parts per million (ppm) by volume. Oa was
introduced into contaminated water using a venturi
injector. The contaminated water that flowed past the
injector was mixed with very small bubbles of Oa from the
injector at the injection point. According to Matrix
representatives, sufficient turbulent mixing would occur
at the injection point so that a uniform concentration of Oa
would be distributed to each of the three cells associated
with one path length. The system did not contain vents
from which O3 or volatile organic compounds (VOC)
could be vented.
-------
NOT TO SCALE
Mao
.ilolci
-------
In addition, the trailer housed a container of 50% H2O2
stock solution. Dilutee1 HpQ? stock solution was injected
into the Matrix system. For the SITE demonstration.
H?O? injection ports were placed at path lengths 1,9,17,
25, 33, and 41. Matrix also placed Os inj&clion ports at
path lengths 1 and 17. However, during the demonstration,
O3 was injected only at path length 17.
1.4.3 Innovative Features of the
Technology
Common methods for treating groundwater contaminated
with solvents and other organic compounds include air
stripping, steam stripping, carbon adsorption, biological
treatment, and chemical oxidation. As regulatory
requirements for treatment residuals and by-products
become more stringent and more expensive to comply
with, technologies involving free radical chemistry offer a
major advantage over other treatment techniques Free
radical chemistry technologies destroy contaminants
rather than transfer them to another medium such as
activated carbon or ambient air. Also, technologies
invol/ing free radical chemistry offer faster reaction tales
than some other technologies such as biological treatment
processes (Topudurti and others 1993).
The Matrix technology generates powerful oxidizing free
radicals (OH-) through the combined use of (1) 254-nm
UV-C light. (2) a semiconductor photocaialyst, and (3)
fEAs (HjO? and O3), Conventional technologies that
oxidize organics by UV light, h^Q?, orOs are much more
selective than OH radicals or have kinetic limitations
restricting their applicability to a narrow range of
contaminants (Topudurti and others 1993). As a result of
these limitations, such technologies have not been cost-
competitive treatment options compared to AOPs such
as Matrix photocatalytic oxidation. Also, as discussed in
Section 1.4,1, the use of UV-C light with a wave length of
254 nm has been shown to enhance photocatalytic
reactor performance.
The Matrix technology does not generate residues or
sludges that require further processing, handling, or
disposal. However, routine maintenance of the Matrix
system may include the disposal of UV lamps that contain
mercury and used TiOj-bonded fiberglass n ->h.
The Matrix technology either completely oxidizes target
organic compounds to CO?, H?O, and halide ions or
breaks them down into low molecular-weight compounds,
According io studies of react ion mechanisms for aromatic
and double- and single-bond chlorinated aliphatic
compounds, incomplete oxidation can result in the
formation of low molecular-weight aldehydes and organic
acids (Glaze and others 1930). Table 1-2 compares
several treatment options for water contaminated with
VOCs, including the Matrix technology.
1.5 Applicable Wastes
Based on SITE demonstration results and results from
the case study, the Matrix technology can be used to treat
organics in liquid wastes, including groundwater,
wastewater, landfill leachate, and drinking water.
1.6 Key Contacts
Additional information about the Matrix technology, the
SITE program, and the K-25 Site can be obtained from
ihe following sources:
1. The Matrix Technology
Mr. Bob Henderson
Matrix Photocatalytic, fnc.
22 Pegter Street
London, Ontario N5Z 2B&
Canada
Telephone No.: (519) 660-8669
2, The SITE Program
Mr. Richard Filers
U.S. Environmental Protection Agency
Office of Research and Development
National Risk Management Research Laboratory
26 West Martin Luther King Drive
Cincinnati, Ohio 45268
Telephone No.: {5t3} 569-7809
3. The K-25 Site
Ms Elizabeth Phillips
U.S. Department of Energy
3 Main Street
Oak Ridge, Tennessee 37830
Telephone No.: (423)241-6172
11
-------
Wafer J
(typical) 1
C
c
c
c
c
Path " '
Length
Number
(typical)
Flguro 1-4. Flow C
48©©©
41 WWW
40©©©
33WWW
32©©©
25®®®
24©©©
17®®®
16©©©
9W®W
8©©©
*"v *"» **»
Unitl
onfiguration in the Matrix
WWW47
0©©42
WWW39
©©©34
WW®31
©0026
WWW23
©0©18
WWW 15
0©©10
WWW 7
©0© 2
System.
4600©
43WWW
380©©
35®®®
30©©©
27WWW
®®®45
©0©44
WWW37
©©©36
l& *&' '^' 29
©©©28
22©0© WWW 21
19t'x)®®
1400©
1 1 W IX; t^X;
6©0©
3 WWW
©©©20
WWW13
©0012
WWW 5
©0© 4
Unit 2
>
Legend
i*x) Water flow in
0 Water flow out
Table 1-2. Comparison of Technologies for Treating VOCs in Water
Technotogy Advantage Disadvantage
Air stripping
Steam stripping
Air stripping with carbon
adsorption of vapors
Air stripping with carbon
adsorption of vapors and
spent carbon regeneration
Carbon adsorption
Biological treatment
Chemical oxidation
Effective for high VOC concentrations,
mechanically simple; relatively inexpensive
Effective for all VOC concentrations
Effective for high VOC concentrations
Effective for high VOC concentrations;
no carbon disposal costs; product can
be reclaimed
Low air emissions and effective for
high VOC concentrations
Low air emissions and relatively
inexpensive
No air emissions; no secondary waste;
VOCs destroyed
Matrix photocatalytto oxidation No secondary wastes; multiple
mechanisms for powerful oxidant (OH-)
production to destroy VOCs
Inefficient for low VOC concentrations; VOCs
discharged to air
VOCs discharged to air; hjgh energy consumption
Inefficient for low VOC concentrations; requires
disposal or regeneration of spent carbon
Inefficient for low VOC concentrations;
high energy consumption
Inefficient for low VOC concentrations: requires
disposal or regeneration of spent carbon;
relatively expensive
Inefficient for high VOC concentrations; slow rates of
removal; sludge treatment and disposal required
Not cost-effective for high VOC concentrations:
may be restricted to narrow range of contaminants
Difficult to oxidize VOCs with single bonds between r
carbon atoms; incomplete oxidation produces toxic
intermediates including aldehydes and haloacetic
acids; relatively expensive
12
-------
Section 2
Technology Effectiveness and Applications Analysis
This section addresses the effectiveness and applicability
of the Matrix technology for treating water contaminated
with VOCs. Vendor claims regarding the effectiveness
and applicability of the Matrix technology are included In
Appendix A. Because the SITE demonstration provided
extensive data on the Matrix treatment system, this
evaluation of the technology's effectiveness and potential
applicability to contaminated sites is based mainly on the
demonstration results presented in this section. However,
demonstration results are supplemented by data from
other applications of the Matrix technology, including a
case study conducted by Atomic Energy Canada
Laboratories on the Matrix system. This section also
summarizes the additional performance data. The case
study is discussed In detail in Appendix B.
This section also provides an overview of the SITE
demonstration and discusses the following topics in
relation to the applicability of the Matrix technology:
additional Matrix technology performance data, factors
affecting technology performance, site characteristics
and support requirements, material handling
requirements, technology limitations, potential regulatory
requirements, and state and community acceptance.
2.1 Overview of Matrix Technology
SITE Demonstration
The Matrix technology demonstration was conducted at
the K-25 Site of DOE's ORR in Oak Ridge, Tennessee,
during a 2-week period in August and September 1995.
During the demonstration, about 2,800 gallons of VOC-
contaminated groundwater from the SW-31 spring was
treated. The principal groundwater contaminants were
1,1-dichloroethane (DCA) and 1,1,1-TCA, which were
present at concentrations of about 655 to 840 and 675 to
980 micrograms per liter (p,g/L), respectively. The
groundwater also contained low levels of total xylenes;
toluene; cis-1,2-dichloroethene (DCE); and 1,1-DCE at
concentrations of 55 to 203,44 to 85,78 to 98, and 123
to 163 jig/L, respectively. In addition, a spiking solution
containing PCE, TCE, and benzene was injected into the
groundwater at the influent line to the Matrix system. The
resulting PCE, TCE, and benzene concentrations in
influent groundwater ranged from 125 to 205; 225 to 613;
and 400 to 1,123 |ig/L, respectively. PCE, TCE, and
benzene were selected as spiking compounds because
they are present in groundwater at many Superfund sites
but are not present in groundwater from the SW31 spring
at significant concentrations. Influent critical VOC
concentrations are presented in Table 2-3 of Section 2.2.
Forthe SITE technology demonstration, 1,1 -DCA; 1,1,1-
TCA; total xylenes; cis-1,2-DCE; and the spiking
compounds were considered critical VOCs. The VOCs
1,1 -DCE and toluene were not considered critical because
during the planning stages of the demonstration, available
data did not indicate that 1,1 -DCE ortoluene was present
at significant levels in SW-31 groundwater. The VOCs
1,1-DCE and toluene were found to be present at
significant levels only during the SITE demonstration.
The SITE demonstration consisted of seven test runs,
Runs 1 through 7, Each run consisted of a predetermined
set of operating conditions. These conditions are
discussed in detail in Section 2.1.2.
Groundwater used forthe technology demonstration had
a high alkalinity of 270 to 295 milligrams per liter (mg/L)
as calcium carbonate (CaCOs) and a pH of about 6.5 to
7.2 standard units. Groundwater samples collected by
DOE in November 1994 contained high concentrations
of iron and manganese at about 16 and 9.9 mg/L,
respectively. Metals present in their reduced states, such
as ferrous and manganous ions, can be oxidized to less
soluble forms that can precipitate and foul the
photocatalytic reactor cells. To prevent fouling of the
photocatalytic reactor cells during the demonstration, an
ion-exchange pretreatment system was used to remove
iron and manganese in the groundwaterwithoutaffecting
the concentrations of VOCs. The pretreatment system
also removed solids using a 3-micron cartridge filter
system consisting of two cartridge filter units arranged in
parallel prior to iron and manganese removal so that the
ion-exchange columns would not clog.
Figure 2-1 shows the layout of the Matrix technology
demonstration area. The 200-gailon spiking solution
container; static in-line mixer; 2,000-gallon bladder tank;
Matrix treatment system trailer; and an approximately
1,800-gallon treated groundwater accumulation tank were
located in a 45- by 16-foot portable, secondary
containment system. The ion-exchange pretreatment
system was located in a separate portable, secondary
containment system.
13
-------
Pump
Pretreatment
System
Influent
Sampling
Port
>-*
Pretreatment
System
Effluent
Ion-Exchange SystemSamP|ln9
3-Mcron
Cartridge Filter _
(two units in parallel)
Contamlnatsd
Groyndwater
Lepnd
• Liquid sampling port
*• Gaseous monitoring port
PL Path length
Secondary
-Containment
System
•QZ>
Static In-Line
Mixer
1,800-Galion
Treated
Groundwater
Tank
H202
Feed Solution'
Sampling Port,
PL PL PL
41 33 25
H202
Feed Solution
PL PL
9 j Influent
Sampling
Port
(PLO)
PL PL PL
48 36 24
03 Monitoring —
Port
(PL 17)
PL
12
1-Micron
Cartridge
Filter
03 Feed Line to PL 17
200-Gallon
Spiking
Solution
Container
X^lMQ-GailonN
\BladderTank/
Matrix Treatment System (in trailer)
NOT TO SCALE
Figure 2-1. Malrjx Technology Demonstration Layout.
From the ion-exchange system, pretreated groundwater
flowed to a static In-line mixer, where the spiking solution
containing PCE.TCE, and benzene was injected into the
groundwater before it entered the static mixer. Spiked
groundwater then flowed to a 2,000-galIon bladder tank,
where It was stored for a short period before treatment in
the Matrix system.
From the 2,000-gallon bladder tank, groundwater flowed
through a 1-micron cartridgefilterto remove solids before
groundwater entered the first photocatalytic reactor cell.
Figure 2-1 also shows oxidant (63 and l-feOa) injection
and sampling locations. Matrix used bottles of dry Os to
generate Oa and a commercially available, 50% HaOa
solution as the hfeC^ stock solution. Matrix planned to
feed Oa into the pretreated groundwater using a venturi
Injector at path lengths 1 and 17 during Runs 2,5,6, and
7. However, because of difficulties encountered by Matrix
with injecting Og and measuring Osflow rates, 0.4 mg/L
of Oa was injected during Run 2 only at path length 17.
However, because the Oa flow meter malfunctioned after
Run 2 was completed, Matrix did not inject Os during
Runs 5,6, and 7 (PRC 1996a). The H2O2 feed solution
was prepared by diluting the stock solution with distilled
water to obtain desired total H2O2 doses of 70, 26, 21,
and 19 mg/L during Runs 2, 5, 6, and 7, respectively.
H2C>2 was fed Into the Matrix system using peristaltic
pumps at path lengths 1, 9,17, 25, 33, and 41.
The Matrix system was equipped with sampling ports
immediately after path lengths 0,12,24,36, and 48. For
this demonstration, samples collected after path lengths
12, 24, 36 and 48 were considered system effluent
samples representing varying degrees of treatment, and
samples collected after path length 0 (before path length
1} were considered system influent samples (see Figure
2-1).
The following sections describe the project objectives for
the Matrix technology demonstration, the demonstration
approach followed to meet project objectives, and
sampling and analytical procedures.
14
-------
2.1.1 Project Objectives
Project objectives were developed based on EPA's
understanding of the Matrix technology and system,
SITE Demonstration program goals, and input from the
technology developer. The Matrix technology
demonstration had both primary and secondary
objectives. Primary objectives were considered critical
for the technology evaluation. Secondary objectives
involved collection of additional data that were useful, but
not critical, to the technology evaluation. The technology
demonstration objectives listed below are numbered and
are designated by the letters "P" for primary and "S" for
secondary.
The primary objectives of the technology demonstration
were as follows:
P1
Determine percent removals (PR) for critical VOCs
in groundwater achieved by the Matrix treatment
system under different operating conditions (by vary-
ing flow rate, number of path lengths, and O3 and
H2O2 doses)
P2
Determine whether the Matrix treatment system
effluent meets maximum contaminant levels (MCL)
promulgated under the Safe Drinking Water Act
(SDWA) for the critical VOCs listed in Table 2-1 at a
significance level of 0.05
P3 Evaluate the change in acute toxicity of groundwa-
ter (measured as the lethal concentration [expressed
as percent sample] at which 50% of test organisms
die [LC50]) after treatment by the Matrix system at a
significance level of 0.05
P4 Evaluate the reproducibility of the Matrix treatment
system performance in terms of PRs for critical
VOCs and its ability to meet applicable target efflu-
ent levels for the critical VOCs listed in Table 2-1
P5 Estimate costs for the Matrix system to treat ground-
water contaminated with VOCs
The secondary objectives of the technology demonstration
were as follows:
S1 Document the concentrations of potential treatment
by-products in groundwater (for example, haloacetic
acids and aldehydes) formed by the Matrix treat-
ment system
S2 Determine PRs for noncritical VOCs in groundwa-
ter achieved by the Matrix system under different
operating conditions (by varying flow rate, number
of path lengths, and O3 and H2O2 doses)
S3 Document observed operating problems and their
resolutions
2.1.2 Demonstration Approach
Seven test runs were performed during the SITE
demonstration to evaluate the performance of the Matrix
system. The demonstration approach is summarized in
Table 2-2. Table 2-2 also shows the relationship between
each test run and the primary and secondary project
objectives.
During the week before the demonstration, Matrix
performed predemonstration test runs. These runs
consisted of experiments to determine the initial operating
conditions for the SITE demonstration. During these
experiments, Matrix treated spiked'groundwater with
characteristics similar to the groundwater used for the
actual demonstration. A field-transportable, direct
sampling, ion-trap mass spectrometer (DSITMS) operated
by DOE's contractor, Lockheed Martin Energy Systems,
Inc., was used to measure VOC concentrations in influent
and effluent samples.
Table 2-1. Target Effluent Levels for Critical VOCs
Critical VOC
Target Effluent Level (ng/L)°
Benzene
Xylenes (Total)
cis-1,2-DCE
PCE
TCE
1,1-DCA
1,1,1-TCA
Aromatic VOCs
Unsaturated VOCs
Saturated VOCs
5
10,000"
70
5
5
5
200
" Target effluent levels are MCLs; however, because no federal MCL exists for 1,1-DCA, the MCL for 1,2-DCA was used as the target effluent
level for 1,1-DCA for the demonstration.
b The MCL for total xylenes exceeds the concentration of total xylenes present in demonstration groundwater. This VOC is, however, considered
a critical VOC for this demonstration because total xylenes are present in SW-31 groundwater at high enough concentrations (about 200 \ig/
L) to allow the reporting of meaningful PRs.
15
-------
Tabta 2-2, Demonstration Approach and Relationship of Runs to Project Objectives
Run
No.
1
2
3
4
Flow
Rate
(gpm)
1.0
1.0
2.0
1.5
No. of
Path
Lengths
48
48
48
48
Sampling
Locations
(Path
Length No.)
0, 12, 24 36,
and 48
0, 12,24,
38, and 48
0, 24, and 48
0 and 36
H202
Dose
(mg/L)
0
See note"
0
0
03
Dose
(mg/L)
0
0.4*
0
0
P1 p?
P1.P2,
P1.P2,
P1.P2,
Project
Objective
P3, P5.S2,
P3, P5, S2,
P3, P5, S2,
P3, P5.S2,
and S3
and S3
and S3
and S3
5
6
7
2.0
2.0
2.0
48
48
48
0,12,24,
36, and 48
0,12,24,
36. and 48
0, 12, 24,
36, and 48
See note *
See note •
See note •
0
0
0
P1.P2, P3, P4.P5, S1.S2,
and S3
P1, P2, P3, P4, P5, S1.S2,
and S3
P1, P2.P3.P4, P5, SI, S2,
and S3
" The HzOj dosa at path lengths 1,9,17,25,33, and 41, respectively, was as follows:
Run 2: 14,13,18,13,9, and 3 mg/L
Run 5: 7,4,4,4,4, and 3 mg/L
Run 6: 6,3,3,3,3, and 3, mg/L
Run 7: 5,3,3,3,3, and 2 mg/L
* O, wos added at path length 17 only.
Run 1 was preceded by a start up run. The startup run
was conducted at the end of the predemonstration runs.
Conditions for the startup run were identical to those for
Run 1, including the use of spiked groundwater. The
purpose of the startup run was to identify and resolve any
problems arising from sampling and field analysis
protocols. Only field analyses using the DSITMS and
Held measurements (such as pH and groundwater
temperature) were performed during the startup run. No
groundwater samples were sent for off-site analysis
during the startup run.
The technology demonstration began with Run 1, which
was performed at a flow rate of 1.0 gpm. No H2Oa or Og
was added during Run 1. The influent flowed through all
48 path lengths, and groundwater samples were collected
at path lengths 0,12,24,36, and 48.
During Run2,atotal r-feOa dose of 70 mg/L and a total 03
dose of 0.4 mg/L were added to the system. T he flow rate,
number of path lengths, and sampling locations were
Identical to those under Run 1. Results from Run 2 were
compared to those from Run 1 to evaluate whether HsC^
and Og Improve system performance. Samples collected
during Runs 1 and 2 were sent off site to Quanterra
Environmental Services, Inc. (Quanterra), and Aquatic
Testing Laboratories (ATL) for quick-turnaround VOC
and acute toxicity analyses, respectively.
During Runs 3 and 4, flow rates of 2 and 1.5 gpm were
used, respectively. Groundwater flowed through all 48
path lengths during both runs. Neither H2O2 nor 03 was
added during either of these runs. Samples were collected
at path lengths 0,24, and 48 during Run 3, and at path
lengths 0 and 36 during Run 4. The results of Runs 1,3,
and 4 were compared to evaluate the effect of flow rate,
if any, on system performance by comparing PRs attained
at equivalent contact times (CT). CT is defined as the
amount of time that water is in contact with the
photocatalytlc reactor cells. PRs at a flow rate of 1 gpm
after path length 12 and at a flow rate of 2 gpm after path
length 24 were compared. CTs under both flow rates and
path lengths are considered equivalent (approximately 6
minutes) based on an approximate wafer volume of 1
gallon. Similarly, PRs at a flow rate of 1 gpm after path
length 24, at a flow rate of 1.5 gpm after path length 36,
and at a flow rate of 2 gpm after path length 48 were
compared. CTs under all three flow rates and path
lengths are considered equivalent (approximately
12 minutes) based on an approximate wafer volume of 1
gallon.
Runs 5, 6, and 7 were reproducibility runs performed
under the same operating conditions. Matrix selected
preferred operating conditions for all three runs, such as
a flow rate of 2 gpm and an average total H2O2 dose of
about 22 mg/L (see Table 2-2). As explained in Section
16
-------
2.1, Os was not added during these runs. The selection
of preferred operating conditions was based on the
results of the quick-turnaround VOC and acute toxicity
analysis results from Runs 1 and 2. Samples were
collected at path lengths 0,12,24,36, and 48 during all
three of these runs.
2,1.3 Sampling and Analytical
Procedures
During the demonstration, samples of Matrix system
influent and effluent, pretreatment system influent and
effluent, and HaO2 feed solution were collected.
Demonstration runs lasted from approximately 2.5 to 6.5
hours, depending on the number of samples collected
during a given run. Each run was divided into four
sampling events. The system was allowed to operate for
about 1 hour at the beginning of each run so that steady
state conditions could be reached before sampling was
first conducted. Thus, to reach steady state conditions for
the 1 -, 1.5-, and 2-gpm flow rates, 60,90, and 120 gallons
of water were flushed through the system, respectively.
This approach allowed flushing more than one volume of
water through the Matrix system (24 gallons). To ensure
that representative samples were collected, sample lines
were purged for a few minutes before each sampling
event.
Matrix system influent and effluent samples werecollected
during all runs for VOC, acute toxicity, alkalinity, pH,
temperature, total suspended solids (TSS), and turbidity
analyses. Pretreatment system influent and effluent
samples for on-site iron and manganese analyses were
also collected throughout the demonstration when the
ion-exchange system operated. During the reproducibility
runs, Matrix system influent and effluent samples were
collected for aldehyde, haloacetic acid, iron and
manganese (off-site laboratory), purgeable organic
carbon (POC), total inorganic carbon (TIC), total organic
carbon (TOC), and total organic halides (TOX) analyses.
H2O2 feed solution samples and Matrix system effluent
samples for ^Oz analysis were collected during Run 2
and the reproducibility runs. Matrix system effluent
samples for O3 analysis were collected during Run 2.
Samples for VOC analysis were collected during each of
the four sampling events per run so that average VOC
concentrations during each run could be calculated from
four replicate data points. Samples for acute toxicity
analysis were collected once during the same sampling
event in each run. Samples for all other analyses were
collected during two of the four sampling events.
Preservatives were added to all samples sent off site for
analysis as necessary. Samples for onsite analysis of
h^Oa, iron and manganese, Os, pH, and temperature
were not preserved either because they were analyzed
immediately after collection or because the analytical
methods for these parameters do not require sample
preservation. Quenching agents to neutralize residual
oxidants were added to samples for aldehyde, haloacetic
acid, POC, TIC, TOC, TOX, and VOC analyses. AH
samples were analyzed using EPA-approved methods
such as those presented in Test Methods for Evaluating
Solid Waste (E PA 1994b), Methods for Chemical Analysis
of Water and Wastes (EPA 1983), or other standard or
published methods (APHA and others 1992; Boltz and
Howell 1979).
Measurements of the Matrix system influent flow rate
were recorded at the beginning of each run and once
every hour while the run was in progress. Electrical
energy measurements were recorded at the beginning
and end of each run. The H2O2 feed solution influent flow
rate was measured and recorded at the beginning of
Runs 2, 5, 6, and 7, and then once every hour during
these runs. Osfeed gas influent flow rate and concentration
were measured and recorded at the beginning of Run 2
and once every hour during this run.
In all cases, EPA-approved sampling, analytical, and
QA/QC procedures were followed to obtain reliable data.
These procedures are described in the QAPP written
specif icallyforthe Matrix technology demonstration (PRC
1995) and are summarized in the TER, which is available
from the EPA project manager (see Section 1.6).
2.2 SITE Demonstration Results
This section summarizes results from the Matrix
technology SITE demonstration for both critical and
noncritical parameters and discusses the effectiveness
of the Matrix technology in treating groundwater
contaminated with VOCs. Table 2-3 presents the average
critical VOC concentrations in the influent to the Matrix
system for each test run.
Performance data collected during the demonstration
are presented in this section in tabular and graphic form.
In most cases, reported data are based on average
values derived from replicate sampling event results. For
influent samples with analyte concentrations at
nondetectable levels, half the detection limit was used as
the estimated concentration to calculate the average
concentration unless all replicate sampling results were
at nondetectable levels. If all influent replicate samples
contained nondetectable levels of any analyte, the
detection limit was used to calculate the average
concentration for that analyte. The average is then
reported as a "<" (less than) value, and the PR was not
calculated. Because effluent samples were analyzed at
two dilutions, lower dilution results were used to calculate
average concentrations, except when the analyte
concentrations in the effluent samples exceeded the
calibration range. In this case, results for the higher
dilution were used to calculate the average concentrations.
For effluent samples with analyte concentrations at
nondetectable levels, half the detection limit of the lower
dilution was used as the estimated concentration to
calculate the average concentration unless all replicate
effluent sample results were at nondetectable levels. If all
effluent replicate samples contained nondetectable levels
of any analyte, the detection limit of the lower dilution was
used to calculate the average concentration for that
analyte and the average analyte concentration was
reported as a "<" value. However, the PR is reported as
17
-------
Table 2-3. Critical VOC Concentrations in Matrix System Influent
Critical VOC
Run 1
ftig/L)
Run 2
(ng/L)
Run 3
(tig/L)
Run 4
(H9/L)
Run 5
(KI/L)
Run6
Oig/L)
Run 7
(H9/L)
Aromatic VOCs
Benzona
Xylenes {Total)
c!s-l,2-DCE
PCE
TCE
1,1-DCA
1,1,1-TCA
1,123
148
78
190
613
655
675
930
168
92
183
550
763
908
995
203
1,025
55
Unsaturated VOCs
98
205
570
87
133
510
Saturated VQCs
840
980
793
845
655
118
90
125
335
818
885
428
117
84
130
238
685
733
400
158
90
153
225
820
878
a V (greater than) value and the 95% upper confidence
limit (UCL) was not calculated.
The remainder of this section is organized to correspond
to the project objectives presented in Section 2.1.1.
Specifically, Sections 2.2.1 through 2.2.4 address primary
objectives except for objective P5 (estimation of costs),
which Is discussed in Section 3. Parts of Sections 2.2.1,
2,2.2, and 2.2.4 also address secondary objective S2
(determination of noncritical VOC PRs). Sections 2.2.5
and 2.2.6 address secondary objectives S1 and S3,
respectively.
2,2.1 Critical VOC PFts under Different
Operating Conditions
During the Matrix technology demonstration, VOC PRs
were measured at different path lengths, flow rates, and
oxidant doses. The VOC PRs observed when each of
these operating conditions was varied are discussed
below. The PR at a given path length was calculated
using the average influent VOC concentration as a
baseline.
Path Length
Varying the number of path lengths used for treatment
changes the CT of the treatment system, thus varying the
timeavailablefor VOC destruction. Groundwater samples
were collected at path lengths 0 (influent), 12,24,36, and
48 during Run 1 and at path lengths 0,24, and 48 during
Run 3. The system was operated at flow rates of 1 and 2
gpm during Runs 1 and 3, respectively, and no oxidant
was added to the system during these runs. Figures 2-2,
2-3, and 2-4 summarize PRs for critical aromatic,
unsaturated, and saturated VOCs at different path lengths
during Runs 1 and 3. As shown in these figures, critical
VOC PRs increased with increasing path length.
During Run1 among the critical aromatic VOCs, the
largest increase in PR (42 percentage points) was
observed for benzene (53% at path length 12 to 95% at
path length 48). During Run 3, the largest increase in PR
(34 percentage points) was again observed for benzene
(53% at path length 24 to 87% at path length 48).
During Run 1 among the critical unsaturated VOCs, the
largest increase in PR (59 percentage points) was
observed for TCE (27% at path length 12 to 86% at path
length 48). During Run 3, the largest increase in PR (37
percentage points) was again observed for TCE
(32 percent at path length 24 to 69% at path length 48).
During Run 1 among the critical saturated VOCs, the
largest increase in PR (22 percentage points) was
observed for 1,1-DCA (-6% at path length 12 to 16% at
path length 48). During Run 3, the largest increase in PR
(12 percentage points) was again observed for 1,1-DCA
(-2% at path length 24 to 10% at path length 48). Negative
PRs were observed for 1,1-DCA during Run 1 at path
length 12, during Run 3 at path length 24, and also during
Run 6 at path lengths 12 and 24. For 1,1,1-TCA, negative
PRs were observed during Run 1 at path lengths 12,24,
and 36; during Run 3 at path length 24; and during Run
6 at path lengths 12,24, and 48; and during Run 7 at path
lengths 12 and 24. Because photocatalytic oxidation
literature does not suggest the formation of saturated
VOCs during treatment, the PRs for saturated VOCs
were examined more closely. When a negative PR was
observed at a particular effluent sampling location, the
95% confidence interval for the effluent concentration
was compared with the 95% confidence interval for the
influent sample concentration. In ail cases, the influent
and effluent intervals overlapped, indicating that no
statistically significant difference exists between influent
and effluent concentrations. For example, for 1,1,1-TCA
during Run 1 at path length 12, the influent and effluent
95% confidence intervals are 601 to 749 jig/L and 607 to
969 ng/L, respectively. Because these intervals overlap,
the conclusion that the observed minus 17 PR resulted
from the formation of 1,1,1-TCA cannot be statistically
supported. Therefore, the conclusion that negative PRs
result from the formation of saturated VOCs cannot be
statistically supported.
18
-------
Run1
(1 gpm, No Oxidants)
95 96
Path Length 12 Path Length 24 Path Length 36 Path Length 48
1 Benzene •Xylenes (total)
Noncritical VOCs detected in Matrix system influent at
concentrations exceeding their project-required
quantitation limit (PRQL) of 50 pg/L during Runs 1 and 3
include 1,1 -DCE and toluene. During Runs 1 and 3,1,1 -
DCE was detected at concentrations of 123 and 158 jig/
L and toluene was detected at concentrations of 67 and
85 pg/L, respectively. During Run 1, PRs for 1,1-DCE
increased from 50% at path length 12 to 94% at path
length 48, and PRs for toluene increased from 59% at
path length 12 to 95% at path length 48. During Run 3,
PRs for 1,1-DCE increased from 40% at path length 24
to 79% at path length 48, and PRs for toluene increased
from 55% at path length 24 to 88% at path length 48.
Flow Rate
During the Matrix technology demonstration, the effect of
flow rate on system performance was determined by
comparing VOC PRs at equivalent CTs using different
flow rates and path lengths. VOC PRs were compared at
a CT of approximately 6 minutes, which was achieved at
1 gpm at path length 12, and at 2 gpm at path length 24.
VOC PRs were also compared at a CT of approximately
Run3
(2 gpm, No Oxidants)
Run 1
(1 gpm, No Oxidants)
Path Length 24
Path Length 48
Path Length 12 Path Length 24 Path Length 36 Path Length 48
\m cis-1,2-DCE D PCE • TCE I
1 Benzene IXylenes (total)
Figure 2-2. PRs at Various Path Lengths for Critical Aromatic
VOCs.
The figures also show that PRs for critical aromatic and
unsaturated VOCs are higher than those for critical
saturated VOCs. This difference is probably due to the
presence of double bonds between the carbon atoms in
the unsaturated VOCs and aromatic bonds between the
carbon atoms in the aromatic VOCs. In general, VOCs
with multiple bonds between carbon atoms are more
amenable to oxidation by OH- than single bonds between
carbon atoms because the electrons of these multiple
bonds can react with the OH- to form relatively stable
intermediates that survive long enough to rearrange or
react further. Similar intermediates from saturated
compounds with single bonds between the carbon atoms
have such short lifetimes that they will generally revert to
their original form.
Run3
(2 gpm, No Oxidants)
100
Path Length 24
Path Length 48
|B Cis-1,2-DCE D PCE •TCE I
Figure 2-3. PRs at Various Path Lengths for Critical Unsaturated
VOCs.
19
-------
Run 1
(1 gpm, No Oxidants)
Path Length 12 Path Length 24 Path Length 36 Path Length 48
I O1.1-DCA B1.1, '
6 percentage points or less at equivalent CTs, For critical
saturated VOCs, the greatest variation in PR (20
percentage points) was observed for 1,1,1-TCA.
Therefore, changing the flow rate at equivalent CTs
appeared to impact system performance on 1,1,1-TCA,
indicating that mass transfer limitations might exist for
this compound. The variation in PRs for 1,1 -DCA (4 to 10
percentage points) was not as large as that for 1,1,1-
TCA.
Because diffusion coefficients in water for the VOCs
addressed in this report are not available in published
literature (Perry and Chilton 1973), diffusion coefficients
in water were estimated for the critical VOCs using a
published method (Lyman and others 1990). The method
estimates diffusion coefficients using the viscosity of
water and the molar volume of each compound of interest.
The estimated diffusion coefficients showed no correlation
Run3
(2 gprn, No Oxidants)
30
20
10
'-10
-20
-30
10
-2
Path Length 24
Path Length 48
|ai.1-DCA
Figure 2-4. PRs at Various Path Lengths for Critical Saturated
VOCs.
12 minutes, which was achieved at 1 gpm at path length
24; at 1.5 gpm at path length 36; and at 2 gpm at path
length 48. If flow rate were not to have an effect on
system performance at equivalent CTs, PRs should be
comparable; otherwise, PRs may differ because of mass
transfer limitations at low flow rates (Turchi and Ollis
1988). According to Matthews (1988), a relationship may
exist between photocataly tic oxidation reaction rates and
diffusion coefficients for individual organic compounds,
but it is unlikely that differences in reaction rates result
from differences in diffusion coefficients only,
Photocatalytic oxidation rates are also believed to depend
on reactor geometry (Turchi and Ollis 1988) as well as
contaminant concentration, reactivity, and adsorption on
the photocatalytic surface.
Figures 2-5 through 2-7 present critical VOC PRs for
critical aromatic, unsaturated, and saturated VOCs,
respectively, at equivalent CTs of 6 and 12 minutes.
These figures show that at equivalent CTs, changing the
flow rate did not significantly affect system performance
on critical aromatic and unsaturated VOCs. For critical
aromatic and unsaturated VOCs, PRs varied by
Equivalent CT of 6 Minutes
1 gpm
Path Length 12
(Run 1)
2 gpm
Path Length 24
(Run 3)
Q Benzene • Xylenes (total)
Equivalent CT of 12 Minutes
1 gpm
Path Length 24
(Run1)
1.5 gpm
Path Length 36
(Run 4)
2 gpm
Path Length 48
(Run 3)
H Benzene • Xylenes (total)
Figure 2-5. PRs at Equivalent CTs for Critical Aromatic VOCs.
20
-------
Equivalent CT of 6 Minutes
1 gpm
Path Length 12
(Run 1)
2 gpm
Path Length 24
(Run 3)
varied by as much as 14%. Although this variation is not
as high as that observed for 1,1-TCA, it is possible that
mass transfer limitations might exist for this compound
also. For toluene, PRs at equivalent CTs varied by 3 to 4
percentage points. Therefore, at equivalent CTs, changing
theflow rate did notappearto impact system performance
on toluene.
Oxidant Dose
Table 2-4 presents critical VOC PRs for Runs 1 and 2,
which were both conducted at a flow rate of 1 gpm. No
oxidants were added during Run 1. During Run 2, the
system received a total HaOa dose of 70 mg/L added at
six injection points distributed along the system and a
total 63 dose of 0.4 mg/L added at path length 17 only. As
shown in Table 2-4, critical VOC PRs were greater during
®eis-1,2-DGE D PCE »TCE
Equivalent CT of 12 Minutes
Equivalent CT of 6 Minutes
1 gpm
Path Length 24
(Run1)
1.5 gpm
Path Length 36
(Run 4)
2 gpm
Path Length 48
(Run 3)
Hcis-1,2-DCE DPCE • TCE
1 gpm
Path Length 12
(Run 1)
2 gpm
Path Length 24
(Run 3)
BS1.1-DCA
• 1,1
,1-TCA'
Figure 2-8. PRs at Equivalent CTs for Critical Unsaturated VOCs,
to the critical VOC PRs. Among the other parameters that
could impact the reaction rate, reactor geometry,
contaminant reactivity, and contaminant concentration
do not change when flow rate is changed. In addition, little
information is available on VOC adsorption on the TiOa
bonded mesh as a function of flow rate. Therefore, the
apparent flow rate effect observed for 1,1,1 -TCA cannot
be explained at this time.
The only noncritical VOC detected in the Matrix system
influent at concentrations exceeding its PRQL during
Runs 1, 3, and 4 is 1,1-DCE. Toluene was detected at
concentrations above the PRQL in Runs 1 and 3. During
Runs 1,3, and 4,1,1 -DC E was detected at concentrations
of 123, 158, and 160jag/L, respectively. Toluene was
detected at concentrations of 67 and 85 jjg/L in Runs 1
and 3, respectively. For 1,1-DCE, PRs at equivalent CTs
Equivalent CT of 12 Minutes
1 gpm
Path Length 24
(Run 1)
1,5 gpm
Path Length 36
(Run 4)
2 gpm
Path Length 48
(Run 3)
1,1-DCA • 1,1,1-TCA
Plgure 2-7, PRs at Equivalent CTs for Critical Saturated VOCs.
21
-------
Run 2 than during Run 1 and the greatest increases in
PRs generally occurred at path length 12, except for 1,1 -
DCA where the greatest increase in PRs occurred at path
length 48. Among the critical aromatic VOCs, benzene
exhibited the greatest PR increase of 41 percentage
points (from 53% during Run 1 to 94% during Run 2).
Among the critical unsaturated VOCs, TCE exhibited the
greatest PR increase of 52 percentage points (from 27%
during Run 1 to 79% during Run 2). Among the critical
saturated VOCs, 1,1,1-TCA exhibited the greatest PR
Increase of 17 percentage points (from -17% during Run
1 to 0% during Run 2). Run 2 PRs were greater than Run
1 PRs at all other path lengths. However, the increase in
PRs from Run 1 to Run 2 was not as large at these other
path lengths. In addition, the negative PRs for critical
saturated VOCs observed during Run 1 were not observed
during Run 2.
Also as shown In Table 2-4, the PRs at the path length 48
In Run 1 (no oxidant added) are approximately the same
as the PRs at path length 24 in Run 2 (oxidant added),
suggesting that the same performance can be achieved
with a significant reduction in capital costs and processing
time through the addition of oxidants. Based on the
example cited above, a reduction in capital costs and
processing time of about 50% could be achieved when
treating watersimilar to that usedduring the demonstration
by adding 70 mg/L of h^Oa and 0.4 mg/L of 03. However,
because only a small quantity of 03 was added, the
improved system performance appears to be primarily
due to the addition of
Noncritical VOCs detected at concentrations exceeding
their PRQL during Runs 1 and 2 include 1,1 -DCE and
toluene. During Runs 1 and 2, 1 ,1-DCE was detected at
concentrations of 123 and 140 jig/L and toluene was
detected at concentrations of 67 and 76 ng/L, respectively.
PRs for these VOCs were also greater in Run 2 than in
Run 1 , and the greatest increase in PR occurred at path
length 12. For 1,1 -DCE, the Run 2 PRs exceeded the
Run 1 PRs by 36 percentage points. For toluene, the Run
2 PRs exceeded the Run 1 PRs by 35 percentage points.
Tabls 2-4. PRs for Critical VOCs in Run 1 (No Oxidants) and Run 2 (Oxidants}
Critical VOC Run 1 PR
Run 2 PR
Path Length 12
Benzene
XySenes (Total)
cis-1,2-DCE
PCE
TCE
1,1-DCA
1,1,1-TCA
Path Length 24
Benzene
Xytenes (Total)
cis-1,2-DCE
PCE
TCE
1,1-DCA
1,1,1-TCA
Path Length 36
Benzene
Xylenes (Total)
tis-1.2-DCE
PCE
TCE
1,1-DCA
1,1,1-TCA
Path Length 48
Benzene
XyJenes (Total)
Cis-1,2-DCE
PCE
TCE
1,1-DCA
1,1,1-TCA
53
62
49
30
27
-6
-17
89
92
83
67
70
0
-11
92
94
87
75
76
6
-3
95
96
93
86
86
16
4
94
95
86
58
79
7
0
99
97
98
84
95
17
4
100°
>98
>96
95
99
29
13
100*
98
>96
96
99
40
14
*Tne PR Is actually 99.8 but dge to rounding is presented as 100
22
-------
Toluene PRs at path lengths 36 and 48 during Runs 1 and
2 could not be compared because the PRs for Run 2 were
calculated to be ">" values (during Run 1, the PRs at path
lengths 36 and 48 were 93 and 95%, but during Run 2, the
PRs at both path lengths were ">" 93%).
2.2.2 Compliance with Applicable Target
Effluent Levels
Applicable target effluent levels are presented in Table 2-
1. Compliance with these target effluent levels was
evaluated by comparing the 95% UCL of effluent VOC
concentrations with the target effluent levels. In some
cases, the 95% UCL could not be calculated because
VOC concentrations were below detectable levels.
However, in all such cases, the detection limit was below
the target effluent level.
Table 2-5 presents target effluent levels, critical VOC
95% UCLs, and the average concentration at path length
48 for Runs2,5,6, and 7, which displayed the best overall
performance in terms of VOC PRs. The table shows that
the Matrix treatment system achieved target effluent
levels at path length 48 for cis-1,2-DCE during Runs 2,5,
6, and 7 and for benzene during Runs 2, 6, and 7. The
95% UCL for benzene at path length 48 in Run 5 was 1
jig/L above the target effluent level, but the average
benzene concentration did not exceed the target level.
The Matrix treatment system also achieved target effluent
levels for cis-1,2-DCE at path lengths 12, 24, and 36
during Runs 2,5,6, and 7 and for benzene at path length
36 during Run 2. For VOCs that are relatively easy to treat
such as PCE and TCE, target effluent levels were not
achieved. This failure may be due to the fact that these
compounds had influent concentrations that were
significantly higher than the target effluent levels. Target
effluent levels also were not achieved for any saturated
VOCs. This failure may be due to the fact that these
compounds are difficult to oxidize. In addition, saturated
VOCs had relatively high influent concentrations (see
Table 2-3). The only noncritlcal VOC detected at a
Table 2-5. Effluent Compliance with Applicable Target Effluent Levels
Critical VOC
Run 2
Benzene
Xylenes (Total)
cis-1 ,2-DCE
PCE
TCE
1,1 -DCA
1,1,1-TCA
RunS
Benzene
Xylenes (Total)
cis-1 ,2-DCE
PCE
TCE
1,1 -DCA
1,1,1-TCA
Run 6
Benzene
Xylenes (Total)
cis-1 ,2-DCE
PCE
TCE
1,1 -DCA
1,1,1-TCA
Run 7
Benzene
Xyienes (Total)
cis-1 ,2-DCE
PCE
TCE
1,1 -DCA
1,1,1-TCA
Target
Effluent Level
(WJ/L)
5
10,000°
70
5
5
5
200
5
10,000"
70
5
5
5
200
5
10,000"
70
5
5
5
200
5
10,000"
70
5
5
5
200
Path Length
12
72
14
15
83
140
755
950
229
36
42
88
184
885
989
161
46
38
116
154
794
880
147
57
46
125
165
842
918
95% UCL
Path Length
24
15
7
2
32
32
692
969
52
9
16
52
86
692
824
39
10
16
68
71
760
844.
33
13
22
80
67
838
984
fufl/LI
Path Length
36
2
NCb
NC
11
5
630
910
17
4
7
34
49
650
763
13
3
9
50
42
678
763
11
2
9
48
38
785
883
Path Length
48
3
5
NC
10
g
469
872
6
2
4
22
25
675
824
4
NC
3
31
21
705
785
3
NC
3
31
18
680
863
Avg Concentration
at Path Length 48
(W/L)
2
3
<4
8
6
458
783
5
2
3
20
23
623
698
3
<4
3
26
19
665
745
2
<4
3
27
16
665
798
* Influent concentrations for total xylenes were below target effluent levels (see Table 2-3).
b NC = Not calculated because analyte concentrations were nondetectable; however, detection limit is below target effluent level
23
-------
concentration exceeding its MCL of 7.0 pg/L in Matrix
system influent was 1,1-DCE during Runs 2,5,6, and 7.
The MCL for 1,1-DCE was achieved at path length 48
during Runs 2,5,6, and 7. The MCL for this compound
was also achieved at path lengths 24 and 36 during Run
2.
2.2.3 Effect of Treatment on Groundwater
Toxicity
Bioassay tests were performed during each demonstration
run to evaluate the change in acute toxicity of the
groundwater after treatment by the Matrix system. For
each run, one influent sample and one effluent sample
from path lengths 12,24,36, and 48 were tested. Two
common freshwater test organisms, the water flea
(Ceriodaphnia dubia) and the fathead minnow
(Pimepbales promeias), were used in the bioassay tests.
Toxicity data are presented in Table 2-6 as LCso values
and as acute toxicity units (TUa). The LCso is the sample
concentration, expressed as percent sample, at which
50% of the test organisms die. Toxicity is expected to
decrease for groundwater after treatment by the Matrix
system. As the toxicity of the treated groundwater
decreases, the LCso value increases but the
corresponding TUa value decreases. If the LCso value
was less than 100%, TUa values were calculated using
the following equation {State of California 1990):
TUa
LC50
X100
(2-1)
If the LCso value was greater than 100%, TUa values
were calculated using the following equation (State of
California 1990):
TUa= log f100-SI
1.7
(2-2)
where
S = percentage of organism survival in undiluted sample
Although project objectives specify that LCso values will
be used to analyze the effect of the Matrix system
treatment on groundwater toxicity, TUa values were used
instead because LCso values were greater than 100% at
several path lengths. By using TUa values, a more
comprehensive analysis of groundwatertoxicity is possible
because specific TUa values could be calculated for path
lengths where the LCso values exceeded 100 percent.
Using nonparametric statistics to evaluate the change in
groundwatertoxicity resulting from treatment In the Matrix
system shows that toxicity increased or decreased
depending on the test organism evaluated. Specifically,
the change in toxicity was evaluated by comparing the
change in toxicity between influent and path length 48
effluent over six runs regardless of operating conditions.
Run 4 was not used for this evaluation because path
length 48 effluent was not collected during this run.
Nonparametric statistics (Wilcoxon signed ranks test)
show that for C, dubia, the probability that groundwater
Table 2-6. Acute Toxicity Data
Run
Ceriodaphnia dubia
Influent Effluent
PL'O PL 12 PL 24 PL 36
LCM (Percent)
PL 48
Plmephales promeias
JnllUSDt Effluent
PLO PL 12 PL 24 PL 36
PL 48
1
2
3
4
5
6
7
36.3
41.2
84.6
61,7
79,4
61,8
50.7
61.0 70.7
68.8 36.5
—" 93.7
76.5 70,7
72.0 66,0
68.2 >100
76.0
37.9
61.3
73.9
71.3
>100
>100
76.0
85.2
65.4
99.9
>100
64.2
66.4
>100
70.7
93.7
>100
>100
70.7
88.8
79.4
>100
81.7
73.8
73.8
89.1
—
73.8
>100
89.1
72.0
76.5
—
53.6
82.0
>100
82.0
>100
85.2
82.0
—
87.7
93.7
79.4
TU.
Ceriodaphnia dubia
Run
1
2
3
4
5
6
7
*PL-
"— .
Influent
PLO
2.75
2.43
1.18
1.62
1,26
1,62
1.97
Path length
Not measured
Effluent
PL 12
1.64
1.4S
—
—
1.31
1.39
1.47
PL 24
1.41
2.74
1.07
—
1.41
1.52
0.82
PL 36
1.32
2.64
. —
1.63
1.35
1.40
0.82
PL 48
0.91
1.32
1.17
—
1.53
1.00
0.59
Influent
PLO
1.56
1.51
0.76
1.41
1.07
0.82
0.59
Pimsphales promeias
Effluent
PL 12
1.41
1.13
—
—
1.26
0.59
1.22
PL 24
1,36
1.36
1.12
—
1.36
0.82
1.12
PL 36
1.39
1.31
—
1.87
1.22
0.76
1.22
PL 48
0,41
1.17
1.22
—
1.14
1.07
1.26
24
-------
toxicity decreased as a result of treatment is greater than
95%. For P. promelas, the probability that groundwater
toxicity increased as a result of treatment is greaterthan
65%,
Evaluation of the demonstration's toxicity data using
nonparametrlcstatistics provides general information on
the change in groundwater toxicity after treatment in the
Matrix system for the test organisms. To provide a more
quantitative analysis, the change in groundwater toxicity
resulting from treatment in the Matrix system was also
statistically evaluated using data from the reproducibility
runs, which were conducted under the same treatment
conditions. Specifically, the mean difference calculated
over the three reproducibility runs (Runs 5, 6, and 7)
between influent and effluent TUa values was compared
to zero using a two-tailed, paired Student's t-test. The null
hypothesis is that the mean difference between influent
and effluent TUa values equal zero at a 0.05 significance
level. The critical t value at this significance level with two
degrees of freedom is 4.303. For path length 48 of Runs
5, 6, and 7, the calculated t values for C. dubia and P.
prome/aswere-1.22 and 1.86, respectively. These results
indicate that treatment in the Matrix system did not
statistically change groundwater toxicity for the test
organisms. Calculated t values for effluent at path lengths
12, 24, and 36 were also below 4.303; therefore, the
Matrix system did not significantly change groundwater
toxicity at these path lengths.
During the reproducibility runs, the TUa values at path
length 48 ranged from 0.59 to 1.53 for C. dubia and from
1.07 to 1.26 for P. promelas. Corresponding LC50 values
ranged from greater than 100 to 65.4% for C. dubia and
from 93.7 to 79.4% for P. promelas. Because of the large
variability in the C. dubia TUa values for runs conducted
under the same conditions, the VOC and by-product data
were reviewed to determine if higher VOC or by-product
concentrations corresponded to increased toxicity;
however, no correlation was observed.
During Runs 2,5,6, and 7, hkC^ was added to the system
during treatment. Residual h^Oa concentrations were
generally less than the detection limit of 1 mg/L Based on
literature data, these concentrations are considered low
enough to not have contributed to the overall toxicity of
the treated groundwater. Literature data indicate that the
LCsoforHgOaforC. dirf?/aisabout2mg/L In addition, the
Connecticut Department of Environmental Protection
reports an LCgo value of 18.2 mg/L of HaOg with 95 percent
confidence limits of 10 and 25 mg/L for P. promelas
(CDEP1993).
2.2.4 Reproducibility of Treatment
System Performance
Critical VOC PRs observed in the reproducibility runs
(Runs 5, 6, and 7) are shown in Table 2-7. Table 2-7
shows that the PRs for critical aromatic and unsaturated
VOCs were generally reproducible within 5 percentage
points at all path lengths except forTCE and PCE at path
length 12, where a difference of up to 11 percentage
points was observed. However, PRs for saturated VOCs
varied by as much as 23 percentage points indicating that
the PRs for saturated VOCs were not reproducible.
These observations were more closely examined to
determine whether the apparent PR variation is real and,
if real, whether it is due to the inherent irreproducibility of
the process or is an artifact of sampling and analysis
procedures.
Because influent and effluent samples collected during a
given sampling event are not paired samples, the mean
and confidence limits for PR for a given VOC in each
reproducibility run were estimated from the PRs for each
VOC generated by performing Latin Hypercube
simulation. This simulation technique was selected
because It is more accurate than the commonly used
Monte Carlo simulation technique (Crystal Ball® 1996).
The mean and standard deviation data for influent (path
length 0) and effluent (path length 48) samples were
used as Inputs for the simulation. Table 2-8 shows the
mean, 95% lower confidence limit (LCL), and 95% UCL
values for critical VOC PRs in the reproducibiiity runs
after 1,000 simulation trials. Table 2-8 shows that except
for TCE in Runs 5 and 7 and for benzene in Runs 5 and
6, the confidence intervals did not overlap, indicating that
the PRs were statistically different at the 95% confidence
level. The same conclusion was drawn by comparing the
means using Tukey's method (Kleinbaum and others
1987).
The sampling and analysis error associated with PR
determination was estimated using Gauss's taw of
propogation of errors (Gellert and others 1989). This
approach involves using the MS/MSD RPD values for
influent and effluent samples to estimate the error
associated with PR determination. The estimated error
Table 2-7. Reproducibility Run VOC PRs
Critical
VOC
Benzene
Xylenes (Total)
cis-1,2-DCE
PCE
TCE
1,1 -OCA
1,1,1-TCA
PL12
69
72
57
36
49
5
0
Run 5
PL24 PL36
92
93
84
61
76
16
8
98
98
93
78
87
22
18
PL48
99
99
96
84
93
24
21
PL12
70
70
57
25
45
-9
-13
aun_6
PL24 PL36
92
93
82
66
73
**Q
-9
97
98
91
68
84
4
0
PL48
99
97
96
80
92
3
-2
PL12
69
72
53
27
38
4
-1
Run 7
PL24 PL36
92
94
80
56
72
1
-2
98
99
92
72
86
10
5
PUS
99
>97
97
82
93
19
9
25
-------
Tablo 2-8. Mean, LCL, and UCL Values for Critical VOC PRs at PL 48 in Reproduclbility Runs
Critical VOC
Bonzena
Xylones (Total)
cl»-1,S-DCE
PCE
TGE
1,1-OCA
1,1,1-TCA
Mean
99.19
98.51
96,32
84.24
93.00
22,56
1953
RunS
LCL
99.18
98.48
96.28
84.12
92.95
21.81
18.09
UCL
99.20
98.54
96.37
84.35
93.06
23.31
20.37
Mean
99.20
98.12
96.41
79.30
91.70
2.76
-1.95
RunS
LCL
99.19
98.00
96.40
78.92
91.60
2.36
-2.38
UCL
99.22
98.20
96.42
79.67
91.97
3.17
-1.52
Mean
99.44
98.67
97.20
82.39
93.09
18.77
8.75
Run?
LCL
99.43
98.65
97.16
82.21
93.03
18.54
8.22
UCL
99.44
98.69
97.24
82.57
93.15
19.00
9.29
ranged from 2 to 6%. Specifically, the error was 2% for
TCE and total xylenes; 3% for eis-1,2-DCE, -PCE, and
benzene; 5% for 1,1-DCA; and 6% for 1,1,1-TCA.
Because the mean PRs for unsaturated and aromatic
VOCs were within the sampling and analysis error except
for the PR of PCE in Runs 5 and 6, the PRs for these
VOCs are considered reproducible. However, because
the mean PRs for saturated VOCs were generally not
within the sampling and analysis error the PRs for the
saturated VOCs are not considered reproducible. The
high variability of PRs for the saturated VOCs is probably
due to variability associated with the Matrix treatment
process rather than variability associated with sampling
and analysis procedures.
As stated in Section 2.2.2, the target effluent levels for
cis-1,2-DCE and 1,1-DCE were consistently achieved
during Runs 5, 6, and 7. The target effluent levels for
PCE,' TCE; 1,1-DCA; and 1,1,1-TCA were not achieved
during the reproducibility runs.
2.2.5 Treatment By-Products and
Additional Parameters
During Runs 5, 6, and 7, samples were collected for
analysis for several additional parameters. These
additional parameters include aldehydes, haloacetic
acids, TOC, POC, TIC, TOX, tentatively identified
compounds as part of VOC analysis, alkalinity, pH,
temperature, TSS, and turbidity. The analytical results
for the additional parameters are discussed below.
Research studies have shown that incomplete oxidation
of chlorinated compounds during phptocatalytic oxidation
processes can result in the formation of low molecular
weight aldehydes and organic acids (see Section 1.4.1).
Sample analytical results for haloacetic acids and
aldehydes aresummarized in Table 2-9.Twochloroacetic
acids, mono- and dfchloroaceticacids, andsix aldehydes,
formaldehyde, acetaidehyde, propanal, butanal, glyoxal,
and methyl glyoxal, formed during treatment. Four
haloacetic acids, bromochloroacetic, dibromoacetic,
monobromoacetic, and trichloroacetic acids, were
anaiyzedforbutwere not detected. The aldehyde pentanal
was analyzed for and detected in effluent samples at
concentrations of less than or equal to 3 jig/L. Of the six
aldehydes formed during treatment, formaldehyde and
acetaidehyde showed increasing concentrations with
increasing path length numbers. Formaldehyde and
acetaidehyde have one and two carbon atoms,
respectively, which would explain why these aldehydes
appear to form more readily during treatment. The other
aldehydes have progressively increasing numbers of
carbon atoms or oxygen substitutions, making them
more complicated and increasingly difficult to form. The
increase in concentration of these by-products seems to
indicate that the by-products form as VOC oxidation
progresses through the Matrix system. In addition, the
by-products formed appear relatively stable or not easily
oxidized by the Matrix system because concentrations of
the by-products either increase with increasing path
length number or remain relatively unchanged.
The TIC, TOC, and POC concentrations in influent and
effluent samples collected during Runs 5, 6, and 7 are
presented in Table 2-10. As a result of oxidation, the
overall TOC and POC concentrations decreased, but the
TIC concentrations increased. TOC concentrations
decreased between influent and path length 48 effluent
by 21 and 28% in Runs 6 and 7, respectively, but
increased by 3% in Run 5. Data review does not yield an
explanation for this relatively insignificant increase in
TOG concentration for Run 5. POC was removed by 73,
greater than 88, and greater than 92% during treatment
between influent and path length 48 effluent for Runs 5,
6, and 7, respectively. Assuming that most of the organic
carbon associated with VOCs could be measured as
POC, the data show that the majority of volatile organic
carbon was converted to either nonpurgeable organic
carbon or bicarbonate ion. The TIC concentrations
increased between influent and path length 48 effluent by
10 and 9% in Runs 5 and 7, respectively, but decreased
by 1 % in Run 6. Data review does not yield an explanation
for this relatively insignificant decrease in TIC
concentration for Run 6.
Also shown in Table 2-10 are TOX concentrations. The
Matrix system achieved TOX reductions of 23,27, and
50% between influent and path length 48 effluent for
Runs 5, 6, and 7, respectively. Chloride concentrations
were not measured during the demonstration because
the amount of chloride ions that could have formed if all
VOCs were mineralized would not be easily differentiated
from the average background chloride concentration of
35 mg/L in K-25 Site groundwater.
26
-------
Table 2-9. Haloacetic Acid and Aldehyde Concentrations
Influent
Concentration5
Parameter*
Effluent Concerrtration°4ng/L.)
Path Length Path Length Path Length Path Length
12 24 36 48
RipJS
Dichloroacetic add
Monoehloroaeette acid
Formaldehyde
Acetaldehyde
Propanal
Butanal
Glyoxal
yethyl glyoxal
RUDJ6
Dichloroacetic acid
Monoehloroaeette acid
Formaldehyde
Acetaldehyde
Propanal
Butanat
Glyoxal
Methyl glyoxal
1
1
8
2
1
1
4,3
3
1
1
8
2
1
1
4.3
3
5.5
4
60
23
8
12
43
22
8
5
91
29
9
14
44
31
9
7
135
37
11
14
57
34
12
8
154
46
14
17
50
39
11
11
181
43
11
13
52
39
14
12
209
50
14
15
45
42
11
12
217
46
9.5
10
41
38
16
14
249
52
12
10
41
37
Run 7
Dichloroacetic acid
Monochloroacetic acid
Formaldehyde
Acetaldehyde
Propanal
Butanai
Qlyoxal
Methyl glyoxal
1
1
• 8
2
1
1
4.3
3
10
5
81
28
10
13
38
26
15
8
162
44
12
16
45
33
17
11
206
47
12
15
41
34
18
14
246
61
10
11
37
36
The following parameters were analyzed for but were not detected at all path lengths for ail runs: bromochloroacetie acid, dibromoacetic acid,
rnonobromoacetic acid, and trichloroacetic acid. Pentanai was detected in effluent samples but at concentrations than or equal to 3 fig/
L.
Influent concentration was measured during two sampling events of Runs 5,6, and 7, and mean values are reported in this table.
Effluent concentration was measured during two sampling events of Runs 5,6, and 7, and mean values are reported in this table.
Table 2-10, TIC, TOO, PQC, and TOX Concentrations
Parameter
RunS
TIC
TOG
POC
TOX
Run 6
TIC
TOG
POC
TOX
Run?
TIC
TOG
POC
TOX
Influent
Concentration
(mg/L)
71.3
6.1
0.95
1.050
74.8
7
0.87
1.100
66.8
6.1
1.2
1.150
Path Length
12
85.3
6.8
0.6
1.150
70.6
7.7
0.56
1.100
68.7
5.3
0.56
0.860
Effluent,, Concentration fmg^L>
Path Length Path Length Path Length
24 36 48
75.8
7.2
0.43
0.805
78.2
7.1
0.20
0.760
69.2
4.7
0.39
0.790
74.9
6.8
0.32
0.725
81.1
6.5
<0.10
0.725
71.8
4.5
0.36
0.700
78.6
6.3
0.26
0.810
73.8
5.8
<0.10
0.800
72.9
4.4
<0.10
0.575
27
-------
Tentatively Identified compounds were also measured
during the demonstration. The maximum number of
tentatively identified compounds in influent samples is
12. This number increased to 31 for path length 12 but
decreased to 19 for path length 48. Tentatively identified
compounds detected in influent or effluent samples include
ethanes, ethenes, sulfides, methyl benzenes, unknown
hydrocarbons, unknown alkanesandchloroflourocarbons
such as 1,1 »2-trichloro-1,2,2-triflouroethane (Preen 113).
Chemicals such as Freon 113 are commonly used as
solvents and are highly resistant to oxidation by the
Matrix system. Estimated concentrations of tentatively
identified compounds detected in effluent samples that
were not detected in influent samples ranged from about
5to13Qng/L
Other additional parameters measured during the
demonstration include alkalinity, pH, temperature, TSS,
and turbidity. Each of these parameters, with the exception
of temperature, remained relatively constant throughout
the demonstration. Thetemperature of samples increased
about 5 to 10 degrees Celsius (°C) as groundwater
progressed through the Matrix system. This increase is
attributable to the heating of groundwater during treatment
as the groundwater contacts activated TiOa particles and
the quartz tubes surrounding the UV lamps.
2.2,6 Operating Problems
The Matrix system's operation was observed throughout
the technology demonstration to record problems and
their resolutions. Some of the problems were directly
related to thesystem's operation, but others were specific
to demonstration activities. These problems and their
resolutions are described below.
Prior to the demonstration, Matrix anticipated injecting
Og into the system at two path lengths. However, the
Matrix Oa injection system was unable to inject Qa at
more than one point into the system; therefore, 0$ was
injected Into the system only at path length 17 during Run
2.
Run 1 was preceded by a startup run performed under
conditions Identical to those for Run 1, including the use
of spiked groundwater. During the startup run, the field
sampling team noted that groundwaterflow atthe sampling
portswas considerably lessened orstopped if all sampling
ports were open simultaneously for several minutes. To
lessen this effect. Matrix adjusted the system pressure
pumps and sampling was conducted using a phased
approach, which minimized the time when all effluent
sampling ports were open simultaneously.
During the demonstration, Matrix anticipated conducting
Runs 1 and 2 at a flow rate of 0.5 gpm, which was the
anticipated minimum system flow rate. Because Matrix
was not able to keep a steady flow rate of 0.5 gpm, the
minimum flow rate at which the demonstration was
conducted was changed to 1 gpm.
An additional operating problem encountered during the
demonstration was that on three occasions, the system
was shut down to replace several cracked quartz tubes.
Matrix believes that the breakage resulted from improper
leveling of the Matrix system trailer, which piaced stress
on the quartz tubes. Because the quartz tubes are
susceptible to damage if stressed, proper care must be
taken in transporting and setting up the Matrix system.
2.3 Additional Performance Data
This section summarizes performance data for the Matrix
technology obtained from sources other than the SITE
demonstration. Significant results were obtained from
one study conducted by the Atomic Energy Canada
Laboratories using low-level nuclear laboratory waste at
the Chalk River Laboratories !n Canada. Additional details
about the study are presented in Appendix B.
During the study, the Matrix system and a UV/Qg-oxida-
tion/carbon reactor system were comparatively tested to
determine the preferred treatment option for a liquid, low-
level nuclear waste stream from the Chalk River Labora-
tories. The Matrix system was more efficient at treating
the waste stream than the U V/O$-oxidation/carbon reac-
tor system. Tests were conducted using the following
oxidants as part of the Matrix system: HaCte, compressed
air, and 02- Results show that Oa was the most effective
oxidant and that about 80% of the organic carbon in the
waste stream was converted to COg. Results also indi-
cate that the Matrix system reduced concentrations of
phenol, naphthalene, methylnaphthalenes, biphenyls,
toluene, xylenes, and ethylbenzene in the waste stream
between 50 and 99%.
2.4 Factors Affecting Performance
Several factors influence the effectiveness of the Matrix
technology. These factors can be grouped into three
categories: (1) influent characteristics, (2) operating
parameters, and (3) maintenance requirements. These
categories are discussed below.
2,4.1 Influent Characteristics
The Matrix technology is applicable for the treatment of
organic contaminants in water. Under a given set of
operating conditions, PRs depend on the chemical
structure of the contaminants. PRs are high for organic
contaminants with double bonds between carbon atoms,
such as cis-1,2-DCE; PCE; and TCE, and compounds
with aromatic bonds between the carbon atoms, such as
benzene, ethylbenzene, toluene, and xylene, because
these compounds are easy to oxidize. Organic
contaminants without double or aromatic bonds between
carbon atoms, such as 1,1,1-TCA andi 1,1-DCA, are not
easily oxidized and are thus more difficult to remove.
The Matrix system can operate in a batch recycle mode
If the Influent cannot meet treatment goals in one pass
through the Matrix system and if the water is provided by
a source that allows controlled, intermittent feeding to the
system. Operation in batch recycle mode allows multiple
exposures of highly contaminated water to the TiOa
surface. Although this approach may enhance destruction
of VOCs and toxic by-products and eventually mineralize
the organics after multiple passes through the system,
28
-------
treatment costs will significantly increase. In addition, if
the influent is provided by a continuous source such as a
groundwater extraction system, operating in the batch
recycle mode may not be feasible unless influent flow
rates are low.
OH- scavengers such as carbonate and bicarbonate ions
may impact system performance. Alkalinity is therefore
an important influent parameter. If the alkalinity of the
influent water is high, adjustment may be required to shift
the carbonate-bicarbonate equilibrium from carbonate (a
scavenger) to carbonic acid (not a scavenger). Other
potential scavengers include sulfide; nitrite; cyanide ions;
and oxidizable, nontarget or "background" organics.
Other influent characteristics of concern include high
levels of oxidizable metals, such as iron and manganese,
in their reduced form, TSS, and oil and grease. Metal
precipitates, oil and grease, and other suspended solids
may deposit on the quartz sleeve or fiberglass mesh in
each cell. Consequently, UV light transmission to the
semiconductor surface would be reduced and the
semiconductor would become IKS active, causing
low contaminant removals. In addition, as the fiberglass
mesh becomes increasingly clogged, a significant
pressure drop may occur, resulting in operational
problems. Proper pretreatment techniques should be
used to prevent these problems.
2.4.2 Operating Parameters
Operating parameters are parameters that can be varied
during the treatment process to -achieve desired
contaminant removals and treatment goals. Principal
factors affecting Matrix system performance include path
length, flow rate, and oxidant dose. These operating
parameters are discussed below.
Changing the number of path lengths used during
treatment is one way to change the CT of the system.
When the number of path lengths is increased, the CT is
increased, increasing the opportunity for VOC destruction.
During the SITE demonstration, improved VOC PRs
were observed when the number of path lengths
increased. However, concentrations of toxic by-products,
such as aldehydes and haloacetic acids, generally
increased with increasing path length. Research
conducted by Matrix prior to the SITE demonstration
indicates that by-products such as formaldehyde may be
destroyed with increasing path length. A study conducted
by Matrix in August 1993 shows about a 50% decrease
in formaldehyde after about 30 minutes of CT (PRC
1996c).
Flow rate through the treatment system also determines
the CT. In general, decreasing the flow rate (Increasing
the CT) improves treatment system performance.
However, according to one study, mass transfer limitations
may exist at low flow rates and therefore impair the
treatment system's performance (Turchi and Ollis 1988).
According to Matrix, the system used during the
demonstration has a minimum operating flow rate of 1
gpm. Below this rate, a steady flow through the system
cannot be maintained.
Oxidants such as HaOa and Oa inhibit the electron-hole
reversal process and consequently provide more time for
the photogenerated hole to form OH-. In addition, HaOa
and Oa generate OH- upon reacting with a
photogenerated, excited electron (see Section 1 .4.1 ). In
general, oxidant dose depends on the contaminated
water chemistry, contaminant oxidation rates, and
treatment unit configuration. During the demonstration,
Matrix injected H2©2 at path lengths 1 , 9, 1 7, 25, 33, and
41 during Runs 2, 5, 6, and 7. Total ^Og doses ranged
from about 20 to 70 mg/L during these runs. In addition,
Matrix injected trace levels (0.4 mg/L) of Os at path length
17 during Run 2, However, because the Oa flow meter
malfunctioned after Run 2 was completed, O$ was not
added during subsequent runs (PRC 1996a).
Although oxidants such as Os and HaOa have been
shown to generally improve system performance, their
doses should be carefully controlled because high levels
of oxidants (for example, HsPs) can act as OH-
scavengers, which would impair system performance.
Also, as mentioned in Section 2.2.3, residual oxidants in
Matrix system effluent are known to be toxic to aquatic
life.
2.4,3 Maintenance Requirements
The maintenance requirements for the Matrix system
summarized in this section are based on direct
observations and discussions with Matrix representatives
during and after the SITE demonstration. This section
addresses maintenance requirements only for
components specific to the Matrix technology and not
genera! maintenance requirements for support
components. Regular maintenance by trained personnel
is essential for successful operation of the Matrix system.
The key system component requiring regular maintenance
is the Matrix photocatalytie reactor cell, which consists of
a low-intensity UV lamp, a quartz sleeve, and a TiOa-
bonded fiberglass mesh. Each of these components and
their maintenance requirements are discussed below.
During the demonstration, Matrix used germicidal, low
Os, 254-nm mercury vapor lamps. Decreasing the use
cycle or increasing the frequency at which a UV lamp is
turned on and off can lead to earty lamp failure. Also,
plating of mercury to the interior lamp walls, a process
called"blackening,"andsolarization of the lamp enclosure
material through regular use will reduce a lamp's
transmissibility. These factors cause steady deterioration
in lamp output at the effective wavelength and may
reduce output at the end of a lamp's life by 40 to 60%. This
reduction in lamp output requires more frequent
replacement of the UV lamps. The UV lamps need to be
replaced once every year. When the UV lamps are
replaced, the spent lamps should be analyzed to determine
if they should be disposed of as a hazardous waste
because of their mercury content
According to Matrix, the quartz sleeve surrounding the
UV lamp will break if the trailer housing a mobile Matrix
system is not leveled properly. Placing the Matrix system
on a non-level surface will create stresses in the system
29
-------
that can crack the quartz tubes, and even small
movements by sampling or operating personnel in a non-
level trailer housing the Matrix system can widen these
cracks, resulting in water leakage. During startup, the
Integrity of the quartz sleeves should be ensured by
performing standard leak checks.
Destruction of contaminants is believed to occur on the
surface of the TiOa-bonded fiberglass mesh. The open
pore configuration of the mesh creates turbulent mixing,
which improves mass transfer in the mesh. For these
reasons, the mesh should be kept free of solids and oil
and grease that could clog the mesh and reduce treatment
efficiency. Proper pretreatment of Matrix system Influent
can prevent clogging of the mesh. If the mesh becomes
clogged, Matrix recommends in s/fa/rinsing with a solution
of clean water and 1 % HgOa for 30 minutes to remove the
solids. According to Matrix, the frequency of mesh
replacement depends on the contaminants treated.
However, Matrix generally recommends that the mesh
be replaced once every 2 years. When the mesh is
replaced, it should be analyzed to determine if it should
be disposed of as a hazardous or nonhazardous waste.
!n addition, the Matrix cell end assembly utilizes a highly
chemical-resistant O-ring to seal the cell annulus.
According to Matrix, the O-rlng requires no scheduled
replacement. However, if leaks are observed in a cell or
group of cells, the CD-rings should be checked for wear
and tear.
2.5 Site Characteristics and Support
Requirements
In addition to influent characteristics, operating
parameters, and maintenance requirements, site
characteristics and support requirements affect the
operation of the Matrix technology. These requirements
should be considered before selecting the Matrix
technology for remediation at a specific site. Site
characteristics and support requirements addressed in
this section include site access, area, and preparation
requirements; climate; utility and supply requirements;
required support systems; and personnel requirements.
Information related to support requirements is based on
information collected for the mobile system used during
the SITE demonstration.
2,5.1 Site Access, Area, and Preparation
Requirements
The site must be accessible for a truck with an 8- by 20-
foot trailer weighing about 7,000 pounds. An of 8 by
20 feet must be available for the trailer that houses the
Matrix system, and additional space must be available to
allow personnel to move freely around the outside of the
trailer. The area containing the Matrix trailer should be
relatively level and paved or covered with compacted soil
or gravel to prevent the trailer from sinking into soft
ground. The trailer will house the Os generating system
and an H20a solution container. Injection ports for Os and
HaOa can be installed at one or more path lengths
throughout the Matrix system. Space outside the trailer is
required for influent and effluent holding tanks if holding
tanks are required as part of the treatment scheme. An
additional area may be required for an office, laboratory
building, or trailer. During the demonstration, an area of
about 40 by 50 feet was used for the Matrix trailer; a
2,000-galIon equalization tank; a pretreatment ion
exchange system for metals removal; an approximately
1,800-gallon effluent holding tank; a laboratory and office
trailer; an outdoor staging area; and miscellaneous
equipment.
2.5,2 Climate
All components of the Matrix system used during the
SITE demonstration were housed inside the trailer, which
provides protection from rain and snow. The trailer was
equipped with exhaust fans. If below-freezing
temperatures are expected for a long period, influent and
effluent storage tanks and associated plumbing outside
the trailer should be insulated or kept in a heated shelter.
2,5.3 Utility and Supply Requirements
The Matrix system can be operated using a 220-volt,
single-phase electrical service. Additional electrical
service may be needed for groundwater extraction well
pumps, office and laboratory buildings, and on-site office
and laboratory equipment, as applicable, in addition,
Matrix can supply process chemicals such as
Oa, as well as spare parts that include UV lamps,
mesh, and quartz tubes. Also, complex laboratory
services, such as VOC and acute toxicity analyses, that
cannot usiJally be performed in an on-site field laboratory
require an off-site analytical laboratory to support an
ongoing monitoring program.
2.5.4 Required Support Systems
In general, pretreatment requirements for contaminated
water entering the Matrix system may include removal of
suspended solids, oil and grease, and metal ions. The
influent may also require pH adjustment to reduce
carbonate and bicarbonate levels. These pretreatment
requirements, as well as effluent disposal, are discussed
below.
To prevent problems with suspended solids accumulation
in the Matrix system, depending on the partieulate
concentration, cartridge filters, sand filters, or settling
tanks may be used to remove suspended solids. Solids
removed from the influent should be dewatered,
containerized, and analyzed to determine whether they
should be disposed of as hazardous or nonhazardous
waste.
According to Matrix, water containing oil and grease
requires pretreatment to separate and remove the oil and
grease to a concentration below 150 mg/L, If such water
is not treated, the oil and grease may deposit on the
phofocatalytic reactor cell and reduce UV light
transmission, which would cause low contaminant
removals. Separated oil and grease should be
containerized and analyzed to determine proper disposal
as hazardous or nonhazardous waste.
30
-------
To prevent fouling of the Matrix system cells, high levels
of metal ions that may be present in influent should be
removed. These ions could form a precipitate on the
quartz sleeve or fiberglass mesh. For the SITE
demonstration, an ion-exchange system with sodium-
based resin was used to remove iron and manganese to
a combined total concentration of 5 mg/L, as requested
by Matrix. Since completion of the SITE demonstration,
Matrix has stated that the iron and manganese
concentration in the influent should be at or below a
combined total concentration of 1 mg/L. An ion-exchange
system was selected for iron and manganese removal
during the SITE demonstration because the SITE activities
lasted only about 4 weeks. However, for a longer term
groundwater cleanup project, metal precipitation maybe
a more cost effective method for removing metal ions
from the influent. If metal precipitation is selected, the
sludge generated should be dewatered, containerized,
and analyzed to determine whether it should be disposed
of as a hazardous or nonhazardous waste.
If the influent contains carbonate and bicarbonate ions at
high levels, pH adjustment may be required. Carbonate
and bicarbonate ions act as OH- scavengers andtherefore
reduce treatment efficiency. The only material handling
requirement associated with pH adjustment is the handling
of chemicals such as acids for pretreatment and bases
for post-treatment (if required for meeting discharge
limits). Adjustment of pH should not create any additional
waste streams requiring disposal.
Effluent can be disposed of either on or off site. Examples
of on-site disposal options for treated water include
groundwater recharge or temporary on-site storage for
sanitary reuse. Examples of off-site disposal options
include discharge intosurface water bodies, storm sewers,
and sanitary sewers. Bioassay tests may be required in
addition to routine chemical and physical analyses to
determine proper treated water disposal.
2.5.5 Personnel Requirements
Personnel requirements forthe Matrix system are minimal.
Generally, one trained operator is required to conduct a
daily system check. The operator should be capable of
performing the following: (1) starting up the system,
(2) operating the influent and in-line pressure pumps,
(3) administering oxidant doses, (4) monitoring operating
parameters including flow rate, and (5) collecting samples
for off-site analyses.
Before operating the Matrix system at a hazardous waste
site, the operator should have completed the training
requirements under the Occupational Safety and Health
Act (OSHA) outlined in 29 Code of Federal Regulations
(CFR) Part 1910.20, which discusses hazardous waste
operations and emergency response. Finally, the operator
should participate in a medical monitoring program as
specified under OSHA.
2.6 Material Handling Requirements
The Matrix system does not generate treatment residuals,
such as sludge, that require handling except for residuals
generated during the maintenance activities discussed in
Section 2.4.3. The Matrix system and its components
produce no air emissions that require special controls.
Pretreatment requirements for contaminated water are
discussed in Section 2.5.4.
2.7 Technology Limitations
Technology limitations identified during the demonstration
are related to flow rates, by-product formation, and
influent characteristics. The Matrix system is limited by
the maximum flow rate at which a single photocatalvtic
reactor cell and unit can be operated. According to
Matrix, each cell is rated for a maximum flow rate of
approximately 0.8 gpm. During the demonstration, Matrix
placed a block in each wafer so that groundwater flowed
in parallel mode into three cells at a time. The overall
maximum flow rate for this configuration was 2.4 gpm.
Treatment at a higher flow rate would require operating
additional cells or units in parallel, which would increase
space requirements and costs to achieve the same
degree of contaminant removal. In addition, flow rates
below 1 gpm could not be steadily maintained in the
Matrix system used during the demonstration.
Based on research studies performed by Matrix and
SITE demonstration results, toxic by-products can form
when VOCs are not completely oxidized in the treatment
system. To decrease by-product formation, the Matrix
system may need to be operated at higher oxidant doses
or CTs (more path lengths) than necessary to meet
treatment goals fortarget contaminants, or contaminated
groundwater may need to be passed through the system
more than once. These approaches would also increase
treatment costs.
Influent characteristics of concern include high levels of
oxidizable metals, such as iron and manganese, in their
reduced form, TSS, and oil and grease. These influent
characteristics require pretreatment to ensure the proper
functioning of the Matrix treatment system. Metal
precipitates, oil and grease, and suspended solids may
deposit on the quartz sleeve or fiberglass mesh in each
cell. Consequently, UV light transmission to the TiOs
semiconductor surface would be reduced and the TiO2
semiconductor would become less active, causing low
contaminant removals. In addition, as the fiberglass
mesh becomes increasingly clogged, a significant
pressure drop may occur, resulting in operational
problems. If the mesh becomes clogged, the in situ
rinsing techniques described in Section 2.4.3 should be
employed. Matrix recommends replacing the fiberglass
mesh once every 2 years and the UV lamps once every
year.
2.8 Potential Regulatory Requirements
This section discusses regulatory requirements relevant
to use of the Matrix technology at Superfund and Resource
Conservation and Recovery Act (RCRA) corrective action
sites. Regulations applicable to implementation of this
technology depend on site-specific remediation logistics
and the type of contaminated liquid being treated;
therefore, this section presents a general overview of the
31
-------
types of federal regulations that may apply under various
conditions. State requirements should also be considered
but because these requirements vary from state to state,
they are not discussed in detail in this section. Table 2-11
summarizes the regulations discussed below. These
regulations include the Comprehensive Environmental
Response, Compensation, and Liability Act (CERCLA),
RCRA, the Clean Water Act (CWA), SDWA, Clean Air
Act (CAA), Toxic Substances Control Act (TSCA), Atomic
Energy Act (AEA) and RCRA for mixed wastes, and
OSHA.
Depending on the characteristics of the water to be
treated, pretreatment or post-treatment may be required
forsuccessful operation of the Matrix system. For example,
solids may need to be filtered out of the water before
treatment. The Matrix treatment system used during the
demonstration included a 1 -micron filter. In addition, a 3-
micron cartridge filter was used in the pretreatment
system to remove solids during the SITE demonstration.
Dissolved metals that could precipitate during treatment
may also need to be removed before treatment. An ion-
exchange pretreatment system was used to remove iron
and manganese from influent during the SITE
demonstration. In addition, if the contaminated water
exhibits high alkalinity, alkalinity adjustment may be
required so that the VOC PRs are not reduced. Each
pretreatment or post-treatment process may involve
additional regulatory requirements that would need to be
determined in advance. This section focuses on
regulations applicable to the Matrix system only.
2.8.1 Comprehensive Environmental
Response, Compensation, and
Liability Act
CERCLA as amended by SARA authorizes the federal
government to respond to releases of hazardous
substances, pollutants, or contaminants that may present
Table 2-11. Summary of Applicable Regulations
Acl/AulhQfily Applicability Application to Matrix Treatment System
Citation
CERCLA
RCRA
CWA
SDWA
CAA
TSCA
Superfund sites
Superfund and
RCRA sites
Discharge to
surface water bodies
Water discharge, water
reinjection, and sole-
source aquifer and
wellhead protection.
Air emissions from
stationary and mobile
sources
Polychlorinated
biphenyl (PCB)
contamination
AEA and RCRA Mixed wastes
OSHA All remedial actions
Requirements
This program authorizes and regulates the cleanup of
environmental contamination. It applies to all CERCLA
site cleanups and requires that othe environmentalr
laws be considered as appropriateto protect human
health and the environment.
RCRA defines and regulates the treatment, storage, and
disposal of hazardous wastes. RCRA also regulates
corrective action at generator andtreatment.storage,
or disposal facilities.
National Pollutant Discharge Elimination System(NPDES)
requirements of the CWA apply to both Superfund and
RCRA sites where treated water isdischarged to surface
water bodies. Pretreatment standards apply to discharges
to publicly owned treatment works.
40 CFR, Part 300
40 CFR, Parts 260 through
270
40 CFR, Parts 122 through
125 and 403
MCLs and contaminant level goals should be considered
when setting water cleanup levels at RCRA corrective
action and Superfund sites. (Water cleanup levels are also
discussed in RCRA and CERCLA.) Reinjection of treated
water is subject to underground injection control program
requirements, and sole sources and protected wellhead
water sources are subject to their respective control programs.
40 CFR, Parts 141 through
149
If O3 emissions occur or hazardous air pollutants are of
concern, these standards may be applicable to ensure
that air pollution is not associated with the use of this
technology. State air program requirements should also
be considered.
If PCB-contaminated wastes are treated, TSCA
requirements should be considered to determine cleanup
standards and disposal requirements. RCRA also regulates
solid wastes containing PCBs.
AEA and RCRA requirements apply to the treatment,
storage, or disposal of mixed wastes containing both
hazardous and radioactive components. OSWER and DOE
directives provide guidance that address mixed wastes.
OSHA regulates on-site construction activities and the
health and safety of workers at hazardous waste sites.
Installation and operation of the system at Superfund
or RCRA sites must meet OSHA requirements.
40 CFR, Parts 50, 60, 61,
and 70
40 CFR, Part 761
AEA (10 CFR) and RCRA
(see above)
29 CFR, Parts 1900
through 1926
32
-------
an imminent and substantial danger to public health or
welfare (EPA 1994a), Remedial alternatives that
significantly reduce the volume, toxicity, or mobility of
hazardous materials and provide long-term protection of
human health andthe environment are preferred. Selected
remedies must also be cost effective, and Superfund site
remediation activities must comply with environmental
regulations to protect human health andthe environment
during and after remediation.
Treatment of contaminated water using the M atrix system
will generally occur on site, and effluent discharge may
occur either on or off site. CERCLA requires that on-site
actions meet all substantive state and federal ARARs.
Substantive requirements (for example, effluent
standards) pertain directly to actions or environmental
conditions. Off-site actions must comply with both
substantive and administrative ARARs. Administrative
requirements (such as permitting) facilitate
implementation of substantive requirements. Depending
on site-specific conditions, EPA allows ARARs to be
waivedforon-site actions. Six ARAR waivers are provided
for by CERCLA: (1) interim measures waiver,
(2) equivalent standard of performance waiver, (3) greater
risk to human health and the environment waiver,
(4) technical impracticability waiver, (5) inconsistent
application of state standards waiver, and (6) fund-
balancing waiver. The justification for a waiver must be
clearly demonstrated (EPA 1988b). Off-site remediations
are not eligible for ARAR waivers, and all applicable
substantive and administrative requirements must be
met.
CERCLA requires identification and consideration of
environmental laws that are ARARs applicable to site
remediation before implementation of a remedial
technology at a Superfund site. Additional regulations
pertinent to use of the Matrix system are discussed
below. No direct air emissions or residuals (such a
sludge) are generated by the Matrix treatment process.
Therefore, only regulations addressing contaminated
liquid storage, treatment, anddischarge;potentialfugitive
air emissions from Oa-generating equipment or VOC-
contaminated water storage tanks; and additional
considerations are discussed below.
2.8.2 Resource Conservation and
Recovery Act
RCRA as amended by the Hazardous and Solid Waste
Amendments of 1984 regulates management and
disposal of municipal and industrial solid wastes. EPA
and RCRA-authorized states (listed in 40 CFR, Part 272)
implement and enforce RCRA and state regulations.
Some of the RCRA requirements under 40 CFR, Part
264, generally apply to CERCLA sites that contain RCRA
hazardous wastes because remedial actions generally
involve treatment, storage, or disposal of hazardous
waste.
According to Matrix, the Matrix system can treat water
contaminated with most organic compounds, including
solvents, pesticides, PAHs, and petroleum hydrocarbons.
Contaminated water treated by the system may be
classified as a RCRA hazardous waste or may be
sufficiently similar to a RCRA hazardous waste so that
RCRA regulations are applicable. Criteria for identifying
hazardous wastes are provided in 40 CFR, Part 261.
Pertinent RCRA requirements are discussed below.
If the contaminated water to be treated is determined to
be a hazardous waste or is sufficiently similar to a
hazardous waste, RCRA requirements for hazardous
waste storage and treatment must be met. The Matrix
system may require tank storage of hazardous waste
water before treatment. Tank storage of contaminated
and treated water must meet the requirements of 40
CFR, Part 264 or 265, Subpart J.
RCRA, Parts 264 and 265, Subparts AA, BB, and CC,
address air emissions from hazardous waste treatment,
storage, or disposal facilities. Subpart AA regulations
apply to process vents associated with specific treatment
operations for wastes contaminated with organic
constituents. Because the Matrix system has no process
vents, these regulations are not ARARs. Subpart BB
regulations apply to fugitive emissions (equipment leaks)
from hazardous waste treatment, storage, or disposal
facilities that treat waste containing at least 10% by
weight of organic compounds. These regulations address
pumps, compressors, sampling of connecting systems,
open-ended valves or lines, and flanges. Subpart BB
regulations could be ARARs if fugitive emissions are
associated with the Matrix system. Although no direct air
emissions are associated with the Matrix treatment
process, any organic air emissions from storage tanks
would be subject to the RCRA organic air emission
regulations in 40 CFR, Parts 264 and 265, Subpart CC,
These regulations address airemissions from hazardous
waste treatment, storage, ordisposalfacility tanks, surface
impoundments, and containers. Subpart CC regulations
were issued in December 1994 and became effective in
July 1995 for facilities regulated under RCRA. Presently,
EPA isdeferring application of the Subpart CC standards
to waste management units used solely to treat or store
hazardous wastes generated on site from remedial
activities required under RCRAcorrectiye action, CERCLA
response, or similar state remediation authorities.
Subpart CC regulations would not immediately impact
implementation of the Matrix system. The most important
air requirements are probably associated with the CAA
and state air toxics programs (see Section 2.8.5).
Use of the Matrix system would constitute treatment as
defined by RCRA under 40 CFR, Part 260.10. Therefore,
treatment requirements may apply if the Matrix system
belongs to a treatment category classification regulated
under RCRA and if it is used to treat a RCRA listed or
characteristic waste. Treatment requirements under 40
CFR, Part 264, Subpart X, which regulate hazardous
waste storage, treatment, or disposal in miscellaneous
units, may be relevant to the Matrix system. Subpart X
requires that treatment in miscellaneous units be
protective of human health and the environment.
Treatment requirements in 40 CFR, Part 265, Subpart Q
(Chemical, Physical, and Biological Treatment), could
also apply. Subpart Q includes requirementsforautomatic
influent shutoff, waste analysis, and trial tests. RCRA
contains special standards for Ignitable or reactive
wastes, incompatible wastes, and special.categories of
waste (40 CFR, Parts 264 and 265, Subpart B). These
33
-------
standards may apply depending on the water to be
treated by the Matrix system.
The Matrix system may also be used to treat contaminated
water at RCRA-regulated facilities as part of RCRA
corrective actions. Requirements for corrective actions
at RCRA-regulated facilities are included in 40 CFR,
Part 264, Subparts F and S (these subparts generally
apply to remediation at Superfund sites). Subparts F and
S include requirements for initiating and conducting
RCRA corrective actions, remediating groundwater, and
operating temporary units associated with remediation
operations (40 CFR, Parts 260 through 299). In states
authorized to implement RCRA, more stringent state
RCRA standards must also be addressed, if applicable.
2.8.3 Clean Water Act
The CWA is designed to restore and maintain the
chemical, physical, and biological quality of navigable
surface waters by establishing federal, state, and local
discharge standards. If treated water is discharged to
surface water bodies or publicly owned treatment works,
CWA regulations apply. On-site discharges to surface
waterbodies must meet substantive NPDES requirements
but do not require an NPDES permit. A direct discharge
of CERCLA wastewater would qualify as "on-site" if the
receiving water body is in the area of contamination or
very near the site and if the discharge is necessary to
implement the response action. Off-site discharges to
surface waterbodies require an NPDES permit and must
meet NPDES permit discharge limits. Discharge to a
POTW Is considered to be an off-site activity, even if an
on-site sewer is used; therefore, compliance with
substantive and administrative requirements of the
National Pretreatment Program is required in such a
case. General pretreatment regulations are presented in
40 CFR, Part 403.
Any applicable local or state requirements, such as local
or state pretreatment requirements or water quality
standards (WQS), must also be identified and satisfied.
State WQSs are designed to protect existing and
attainable surface water uses (for example, recreation
and public water supply). WQSs include surface water
use classifications and numerical or narrative standards
(including effluent toxicity standards, chemical-specific
requirements, and bioassay requirements to demonstrate
no observable effect level from a discharge) (E PA 1988a).
These standards should be reviewed on a state- and
location-specific basis before discharge to surface water
bodies occurs. During the SITE demonstration, bioassay
tests were conducted to determine whether the treated
water was toxic to particular aquatic species. Similar
bioassay tests might be required if the Matrix system is
implemented in particular states and if treated water is
discharged to a surface water body.
2.5.4 Safe Drinking Water Act
The SDWA as amended in 1986 required EPA to establish
regulations for contaminants in drinking water to protect
human health. EPA has developed the following programs
to achieve this objective: (1) a drinking water standards
program, (2) an underground injection control program,
and (3) sole-source aquifer and wellhead protection
programs.
SDWA primary (or health-based) and secondary (or
aesthetic) MCLs generally apply as cleanup standards
for water that is or that may be used as a drinking water
source. In some cases (such as when multiple
contaminants are present), more stringent MCL goals
may be appropriate. In other cases, alternate
concentration limits (ACL) based on site-specific
conditions may apply. CERCLA and RCRA standards
and guidance should be used to establish ACLs (EPA
1987a). During the SITE demonstration, Matrix treatment
system performance was tested for compliance with
SDWA MCLs for critical VOCs.
Water discharge through injection wells is regulated by
the underground injection control program. Injection wells
are categorized as Classes I through V, depending on
their construction and uses. Reinjection of treated water
involves Class IV (reinjection) or Class V (recharge)
wells and should meet SDWA requirements for well
construction, operation, and closure activities. If the
groundwater treated is a RCRA hazardous waste, the
treated groundwater must meet RCRA federal Land
Disposal Restriction (LDR) treatment standards (40 CFR,
Part 268) before reinjection.
The sole-source aquifer and wellhead protection programs
are designed to protect specific drinking water supply
sources. If such a source is to be remediated using the
Matrix system, appropriate program officials should be
notified and any potential regulatory requirements should
be identified. State groundwater antidegradation
requirements and WQSs may also apply.
2.8.5 Clean Air Act
The CAA as amended in 1990 regulates stationary and
mobile sources of air emissions. CAA regulations are
generally implemented through combined federal, state,
and local programs. The CAA includes chemical-specific
standards for major stationary emissions sources that
are not applicable but that could be relevant and
appropriate for Matrix system use. For example, the
Matrix system would usually not be a major source as
defined by the CAA but it could leak 63, which is a criteria
pollutant under CAA's National Ambient Air Quality
Standards (NAAQS). Therefore, the Matrix system may
need to be monitored for O3| or 63 emissions may need
to be controlled in order to ensure that air quality is not
impacted. NAAQS are particularly applicable to localities
that are O3 "non-attainment" areas. The National Emission
Standards for Hazardous Air Pollutants could also be
relevant and appropriate if regulated hazardous air
pollutants are emitted and if the treatment process is
considered sufficiently similar to one regulated under
these standards. In addition, New Source Performance
Standards (NSPS) could be relevant and appropriate if
the pollutant emitted and the Matrix system are sufficiently
similar to a pollutant and source category regulated by
NSPSs. Finally, state and local air programs have been
delegated significant airquality regulatory responsibilities,
and some have developed programs to regulate toxic air
pollutants (EPA 1989). Therefore, state air programs
should be consulted regarding Matrix treatment
technology installation and use.
34
-------
2.8.6 Toxic Substances Control Act
Testing, premanufacture notification, and
recordkeeping requirements fortoxic substances are
regulated under TSCA. TSCA also includes storage
requirements for PCB-contaminated media (see
40 CFR, Part 761.65). The Matrix system may be
used to treat liquid contaminated with RGBs, and
TSCA requirements would apply to pretreatment
storage of PCB-contaminated liquid. The SDWA MCL
for PCBs is 0.05 (ig/L, and this MCL is generally the
PCBtreatmentstandardforgroundwater remediation
at Superfund and RCRAcorrective action sites. RCRA
LDRs for PCBs may also apply depending on PCB
concentrations (see 40 CFR Part 268). For example,
treatment of liquid hazardous waste containing PCB
concentrations equal to or greater than 50 ppm must
meet the treatment requirements of 40 CFR,
Part 761.70.
2.5.7 Atomic Energy Act and Resource
Conservation and Recovery Act
As defined by the AEA and RCRA, mixed waste
contains both radioactive and hazardous components.
Such waste is subject to the requirements of both the
AEA and RCRA; however, when application of both
AEA and RCRA regulations results in a situation
inconsistent with the AEA (for example, an increased
likelihood of radioactive exposure), AEA requirements
supersede RCRA requirements (EPA 1988b). Use of
the Matrix system at sites with radioactive
contamination might involve treatment or generation
of mixed waste.
OSWER in conjunction with the Nuclear Regulatory
Commission, has issued several directives to assist
in the identification, treatment, and disposal of low-
level radioactive mixed waste. Various OSWER
directives include guidance on defining, identifying,
and disposing of commercial, mixed, low-level
radioactive and hazardous wastes (EPA 1987b). If
the Matrix system is used to treat low-level mixed
waste, these directives should be considered. If high-
level mixed waste or transuranic mixed waste is
treated, internal DOE orders should be considered
when developing a protective remedy (DOE 1988).
The SDWA and CWA also contain standards for
maximum allowable radioactivity levels in water
supplies.
2.8.8 Occupational Safety and Health
Administration Requirements
OSHA regulations in 29 CFR, Parts 1900 through 1926, are
designed to protect worker health and safety. Both Superfund
and RCRA corrective actions must meet OSHA requirements,
particularly Part 1910.120, "Hazardous Waste Operations
and Emergency Response." Part 1926, "Safety and Health
Regulations for Construction," applies to any on-site
construction activities. For example, electric utility hookups
for the Matrix system must comply with Part 1926, Subpart
K, "Electrical." Product chemicals such as HgOa used with
the Matrix system must be managed in accordance with
OSHA requirements (for example, Part 1926, Subpart D,
"Occupational Health and Environmental Controls," and
Subpart H, "Materials Handling, Storage, and Disposal").
More stringent state or local requirements must also be met,
if applicable. In addition, health and safety plans for site
remediations should address chemicals of concern and
include monitoring practices to ensure that worker health
and safety are maintained.
2.9 State and Community Acceptance
Because few applications of the Matrix technology have
been attempted beyond the bench or pilot scale, limited
information is available to assess state and community
acceptance of the technology. This section therefore
discusses state and community acceptance of the Matrix
technology with regard to the SITE demonstration.
Before the demonstration, the primary concerns of project
participants involved the ability of the Matrix system to meet
effluent target levels and the formation of treatment by-
products. These concerns were addressed by performing
calculations to show that no environmental impact was
anticipated from Matrix system effluent. At other sites, state
acceptance of the technology may involve consideration of
performance data from applications such as the SITE
demonstration and results from on-site, pilot-scale studies
using the actual wastes to be treated during later, full-scale
remediation.
During the SITE demonstration, about 100 people from
ORR, the Tennessee Department of Health, several
environmental consulting firms, and the local community
attended a Visitors' Day to observe demonstration activities
and ask questions about the technology. The visitors
expressed no concerns regarding operation of the Matrix
system.
35
-------
Section 3
Economic Analysis
This economic analysis presents cost estimates for using
the Matrix technology to treat groundwater contaminated
with VOCs. Cost data were compiled during the SITE
demonstration at DOE's K-25 Site at ORR and from
information obtained from Matrix, independent vendors
(Grundfos 1996; PRC1996b), and cost estimating guides
(Echos 1996a and 1996b; Means 1995 and 1996). Costs
are organized in 12 categories applicable to typical
cleanup activities at Superfund and RCRA sites (Evans
1990). Costs are presented in July 1996 dollars and are
considered to be order-of-magnitude estimates with an
expected accuracy range of 50% above and 30% below
actual costs.
This section provides an introduction to the economic
analysis (Section 3.1), summarizes the major issues
involved and assumptions made to conduct this analysis
(Section 3.2), discusses categories of costs associated
with using the Matrix technology to treat groundwater
contaminated with VOCs (Section 3.3), and presents
conclusions derived from the economic analysis
(Section 3.4).
3.1 Introduction
Matrix designs its treatment system to meet site-specific
goals. As conditions warrant, Matrix adds individual
modular units as described in Section 2 to meet flow rate
or effluent requirements. The Matrix system used during
the SITE demonstration consisted of two units treating
groundwater at flow rates ranging from 1 to 2 gpm. Based
on the preliminary dataavailableduringthedemonstration,
Matrix selected preferred operating conditions to perform
reproducibility runs. These operating conditions include
a flow rate of 2 gpm and a total average H2O2 dose of
about 22 mg/L
Information collected from the SITE demonstration forms
the basis of this economic analysis. Thus, the analysis
focuses on costs involved with operating a system similar
tothatusedduringtheSlTEdemonstrationatthepreferred
conditions. A hypothetical groundwater remediation
project Is used as a framework to present these costs and
forms a base-case scenario for this analysis. The base
case consists of treating VOC-contaminated groundwater
ataflow rate of 2 gpm forSO years at a rural site. The base
case is described in detail in Section 3.2. Additional cost
estimates are provided for larger systems operating at
12 and 24 gpm for 5 and 2.5 years, respectively. Table 3-
1 presents a breakdown of costs into the 12 cost categories
under each flow rate evaluated in this analysis. The table
also sums up the one-time costs and total annual O&M
costs under the 12 cost categories. Total costs for the
hypothetical groundwater remediation project and the
net present value of the project are presented at the end
of the table. Finally, the costs to treat 1,000 gallons of
water are presented as calculated from the net present
value figures.
3.2 Issues and Assumptions
This section summarizes major issues and assumptions
made in relation to site-specific factors, equipment and
operating parameters, and financial calculations used in
this economic analysis of the Matrix technology. These
issues and assumptions are discussed in Sections 3.2.1
through 3.2.3. Issues are related to variable conditions
that may affect costs from one site to another. Assumptions
are summarized in the bulleted list after the discussion of
the issues and are related to the base-case scenario
analysis. Certain assumptions were made to account for
variable site and waste parameters. Other assumptions
were made to simplify cost estimation. Section 3.2.4
discusses additional premises and assumptions related
to the base-case scenario.
In general, Matrix system operating issues and
assumptions are based on information provided by Matrix
and observations made during the SITE demonstration.
Other issues and assumptions are based primarily on the
operating parameters and results observed during Runs 5,
6, and 7 (the reproducibility runs performed atthe preferred
operating conditions) of the demonstration.
3.2.1 Site-Specific Factors
Site-specific factors can affect the costs of using the
Matrix treatment system. These factors can be divided
into the following two categories: waste-related factors
and site features. Waste-related factors affecting costs
include waste volume, contaminant types and
concentrations, treatment goals, and regulatory
requirements. Waste volume affects total project costs
because larger volumes take longer to remediate. The
contaminant types and concentrations in the groundwater
and the treatment goals for the site determine (1) the
appropriate Matrix treatment system size (number of
units), which affects capital equipment costs; (2) the flow
36
-------
Table 3-1. Costs Associated with the Matrix Technology
2-gpm System
12-apm System
24-opm System
Cost Categoryb Detail
Site Preparation0
Administrative
Treatment area
preparation:
Shelter building 136,300
construction
Piping installation 4,400
Effluent discharge 5,500
line installation
Electrical service 7,900
extension
Treatability study
and system design
Permitting and
Regulatory0
Mobilization and Startup0
Equipment and
personnel
mobilization
Assembly and
shakedown
Equipment0
Matrix treatment
system
Cartridge filter
Ion-exchange
system
Labor"
Supplies'*
«A
Distilled water
TiO2 mesh
UV lamp assembly
Disposal drums
Filters
ion-exchange resins
Disposable PPE
Sampling supplies
Propane gas service
Utilities"
Matrix treatment system
Effluent Treatment and
Disposal"1
Residual Waste
Shipping and Handling"
Analytical Services'1
Equipment Maintenance11
Matrix treatment system
Site Demobilizatiorf'*
Total One-Time Costs
Total Annual O&M Costs
Groundwater
Remediation:
Total costs '-g-h
Net present value1
Costs per 1,000 Gallons'
Itemized
18,800
154,100
3,000
3,000
3,800
60,000
4,000
24,000
100
1,500
5,400
7,200
100
600
18,000
600
400
800
7,800
3,000
400
H
Total
$ 175,900
5,000
6,800
88,000
1 0,000
34,700
7,800
0
3,000
3,600
3,000
100
$ 275,800
62,200
4,670,000
1,838,500
67.74
Detail Itemized
18,800
282,400
264,600
4,400
5,500
7,900
3,000
6,000
7,600
360,000
24,000
57,500
400
8,900
32,400
43,200
500
3,600
108,000
600
400
1,600
46,800
18,000
(16,400)
Total Detail
$ 304,200
521,200
4,400
5,500
7,900
5,000
13,600
441,500
10,000
199,600
46,800
0
18,000
3,800
' 18,000
(12,800)
$751,500
296,000
2,764,900
2,237,300
78.78
Itemized
18,800
539,000
3,000
9,000
11,400
720,000
48,000
81,000
800
17,800
64,800
86,400
900
7,200
216,000
600
400
3,200
93,700
36,000
(32,500)
Total
$560,800
5,000
20,400
849,000
10,000
398,100
93,700
0
36,000
3,600
36,000
(28,800)
$1,406,400
577,400
3,556,800
3,083,200
108.56
All costs are in July 1996 dollars and are rounded to the nearest $100.
Cost categories appearing in bold Italic font are Matrix treatment system direct costs.
One-time costs
Annual O&M costs
The values presented in the itemized columns represent the cost in future values. The total columns show the appropriate 1996 current
collars. Values in parentheses represent a credit value.
One-time and annual O&M costs combined
Future value of O&M costs using annual inflation rate of 5%
The analysis assumes that a total of 28.4 million gallons of water will be treated. With a 10% downtime.this treatment will take the 2-gpm
system 30 years, the 12-gpm system 5 years, and the 24-gptn system 2.5 years to complete.
Annual discount rate of 7.5%
Net present value
37
-------
rate at which treatment goals can be met; and (3) periodic
sampling requirements, which affect analytical costs.
Regulatory requirements affect permitting costs and
effluent monitoring costs, which depend on site location
and the type of disposal selected for the treated effluent.
Site features affecting site preparation and mobilization
and startup costs include groundwater recharge rates,
groundwater chemistry, site accessibility, availability of
utilities, and the geographic site location. Groundwater
recharge rates affect the time required for cleanup and
the size of the Matrix system needed. The presence of
metals such as iron and manganese in groundwater can
decrease Matrix technology effectiveness and increase
equipment and O&M costs by requiring pretreatment.
Site-specific assumptions underthe base-case scenario
Include the following:
• Contaminated water is located in an aquifer no
more than 100 feet below ground surface.
The contaminants and their average concentra-
tions are those observed during the reproducibility
runs and include 1,1-DCA at 770 ng/L; cis-1,2-DCE
at 90 ng/L; 1,1,1-TCA at 830 ng/L; benzene at
490 ttg/L; PCE at 140 ng/L; TCE at 270 jig/L; and
total xylenes at 130 |ig/L. The groundwater does
not contain radioactive contaminants.
• The groundwater contains manganese at 10 mg/L
and iron at 16 mg/L; therefore, the groundwater
requires pretreatment so that the total concentra-
tion of iron and manganese in the influent to the
Matrix system is less than 1 mg/L, These conditions
applied to groundwater during the SITE demonstra-
tion.
Suspended solids in the groundwater require re-
moval before treatment.
* Groundwater does not require pH adjustment be-
fore treatment.
» The site is located in a rural area of the Midwestern
United States. This region has prolonged winter
months of cold temperatures and hot summer
months.
* Utilities and other infrastructure features (for ex-
ample, access roads to the site) exist within 500
feet of the treatment system locale.
* Treated groundwater is discharged to a surface
water body that exists near the site.
* Four on-site groundwater extraction wells provide
the flow rates discussed in this economic analysis.
The treatment system will be located 200 feet from
the wells.
* The groundwater remediation project involves a
total of 28.4 million gallons of water needing treat-
ment. This groundwater volume corresponds to the
total volume treated by a two-unit system operating
at 2 gpm for 30 years with a 10% annual downtime.
3.2.2 Equipment and Operating
Parameters
The Matrixtreatmentsystemcan be used totreat aqueous
waste streams such as groundwater or process
wastewatercontaminated with VOCs and other organics.
This analysis provides costs for treating groundwater
contaminated with VOCs only. Matrix will provide the
appropriate treatment system configuration based on
site-specific conditions, of which groundwater recharge
rates and contaminant types are the primary
considerations. The Matrix system is modular in design,
allowing for unit setup as needed either in series or in
parallel to treat groundwater.
This analysis focuses on the costs associated with an 11 -
kilowatt (kW) system (two 5.5-kW, 12-wafer, 72-ceil
units) operating at the preferred conditions demonstrated
at the K-25 Site. Further details on the demonstration
system are discussed in Section 1.4.2. This Matrix system
can treat contaminated groundwater at a rate of 2 gpm.
The system can operate on a continuous flow cycle 24
hours per day, 7 days per week. Based on these operating
parameters, the system can treat nearly 1,051,200 gallons
per year. Allowing for a 10% annual downtime for
maintenance activities, the annual treatment volume is
946,100 gallons. Because most groundwater remediation
projects are long-term projects, this analysis assumes
that remediation will take about 30 years to complete.
Based on this period of time, the total volume of water
needing treatment is assumed to be 28.4 million gallons.
Because it is difficultto determine both the actual duration
of a project and the volume of groundwater requiring
treatment, these figures have been assumed to perform
this economic analysis.
This analysis provides additional comparisons of larger
Matrix systems operating as 12 and 24 gpm (see Table
3-1). For these comparisons, site and groundwater
characteristics are assumed to be the same as those
outlined in Section 3.2.1. The 12-gpm case consists of a
12-unit Matrix system operating for 5 years. The 24-gpm
case consists of a24-unit system operating for 2.5 years.
Based on information provided by Matrix, the costs of
equipment and related system supply and consumable
requirement rates are linear. For example, H^Oz costs for
the 24-gpm system are expected to be approximately 12
times those of the 2-gpm system.
Groundwater may require pretreatment depending on its
characteristics. High levels of suspended solids and
metals in groundwater can adversely affect Matrix
treatment system performance and therefore require
removal. Two methods are effective in removing solids
and metals from groundwater: (1) a combined cartridge
filter and ion-exchange system and (2) a precipitation/
floeculation/sedimentation (PFS) system. PFS systems
typically have higher capital and O&M costs than combined
cartridge filter and ion-exchange systems because PFS
systems have higher residuals management costs
38
-------
(McArdie and others 1988). This analysis therefore
estimates the costs of using the combined cartridge filter
and ion-exchange system only.
In certain applications, it may be effective to use Oa as an
oxtdant to generate additional OH. An Os generator
would need to be purchased or leased in addition to the
Matrix treatment system because it is not considered part
of the standard equipment Also, either Oz or purified air
is needed to generate the Os. The need for an Os
generator is determined by Matrix after the treatability
study. Because of the variability in the need for and size
of an Oa generator and because the Matrix system did not
use Os during the reproducibility runs during the SITE
demonstration, this analysis does not include costs for
this equipment.
Equipment and operating parameter assumptions under
the base-case scenario are listed below.
» The Matrix treatment system consists of two stan-
dard units in series capable of drawing a total of
11 kW of electrical power. However, during the
demonstration, the average power consumption was
about 9 kW. According to Matrix, underpowering
the system will adversely impact the system perfor-
mance.
Groundwater is pretreated using cartridge filters to
remove suspended solids and an ionexchange sys-
tem to reduce iron and manganese concentrations
in the groundwater to less than 1 mg/L
The system operates at 2 gprn, 24 hours per day
with a downtime of 10%.
» The treatment system operates automatically with-
out requiring the constant attention of an operator
and shuts down in the event of system malfunction.
A 900-square foot, fixed facility is needed to house
the Matrix treatment system and all pretreatment
equipment.
Matrix mobilizes the system to the site, assembles
it, and conducts initial shake down activities.
* The Matrix system generates no air emissions.
3.2.3 Financial Calculations
Most groundwater remediation projects are long-term in
nature. For this reason, the total costs for completing the
groundwater remediation projects presented in this
analysis are calculated for a 30-year period. In Table 3-
1, total costs for this analysis are presented as future
values, and costs per 1,000 gallons treated are presented
as net present values. This analysis assumes a 5%
annual inflation rate to estimate future values. The future
values are then calculated as net present values using a
discount rate of 7.5%, which is the current yield on a 30-
year Treasury bond. Using a higher discount rate makes
the initial costs weigh more heavily in the calculation, and
using a lower discount rate makes future operating costs
weigh more heavily. Because demobilization costs are
incurred at the end of the project, the appropriate future
values of these costs are presented in the total columns
for this cost category.
This analysis assumes that the Matrix system has a
positive salvage value of 5% of the treatment system's
equipment cost. The salvage value is assumed to arise
from the of certain Matrix system components such
as stainless piping, pumps, tubes, and UV lamp
assemblies. The proprietary nature of the Matrix system
legally precludes an ownerfrom selling the equipment as
a complete treatment system. Salvage value is typically
a constant value applied over the life of the capital
equipment and deducted from the purchase price to
determine an annual equipment depreciation expense
for annual income tax purposes. Because financial
accounting practices differ, this analysis assumes that
the salvage value wi II be a cash receipt to the owner at the
end of the project and therefore does not calculate
equipment depreciation. Also, because the salvage value
is not considered for financial accounting purposes, the
future value of the cash receipt is assumed to be realized
by the owner at the end of the project.
3.2,4 Base-Case Scenario Premises and
Assumptions
A hypothetical groundwater remediation project has been
developed based on the issues and assumptions
described above for the purposes of formulating a base-
case scenario from which cost estimates can be derived.
Additional premises and assumptions used forthis base-
scenario include the following:
All costs are rounded to the nearest $100.
Contaminated groundwater is treated to achieve
the PRs observed during SITE demonstration re-
producibility Runs 5, 6, and 7. During the demon-
stration, the effluent did not meet all MCLs usually
required to be met at Superfund sites. For this
reason, the costs presented in this analysis may
need to be adjusted based on site-specific goals.
» The Matrix system is mobilized 500 to the
remediation site from London, Ontario, Canada.
Customs clearing expenses are paid for by Matrix.
Operating and sampling labor costs are incurred by
the client. The client also performs and pays for
routine maintenance and modification activities.
Initial operator training is provided by Matrix at no
extra cost
3.3 Cost Categories
Cost data associated with the Matrix technology are
grouped into the following cost categories: (1) site
preparation, (2) permitting and regulatory, (3) mobilization
and startup, (4) equipment, (5) labor, (6) supplies,
(7) utilities, (8) effiuenttreatmentanddisposal,{9) residual
waste shipping and handling, (10) analytical services,
39
-------
(11) equipment maintenance, and (12) site
demobilization. The basis of each cost category is the 2-
gpm treatment system demonstrated at the K-25 Site.
Additional analysis is provided for the 12- and 24-gpm
systems. Table 3-1 presents cost breakdowns under the
12 cost categories for each treatment system.
3.3.1 Site Preparation Costs
Site preparation costs Include administrative, treatment
area preparation, treatability study, and system design
costs. No site clearing or grubbing or soil stabilization is
assumed to be needed, and no postconstructipn
restoration activities are included. However, minimal site
grading is assumed to be needed for placing the Matrix
system on a level surface.
Site preparation administrative costs include project work
plan development, legal searches, access right
determinations, and other site planning and design
activities. These activities are (1) assumed to require
about 250 labor hours at $75 per hour to complete;
(2) estimated to cost $18,800; and (3) assumed to be the
same for all three treatment systems.
Treatment area preparation involves constructing a sh elter
building, installing piping from the extraction wells to the
shelter building, installing piping from the shelter building
tothe nearest surface water body, andextending electrical
service to the treatment site location. These activities
need to be conducted before the Matrix system is mobilized
tothe site. A permanent, 900-square foot shelter building
on a bermed concreteslab with a sump is required forthe
Matrix system and pretreatment equipment specified for
the two-unit system under the base case. Grading and
construction costs are estimated to be $140 per square
foot. Purchase and installation costs for a propane gas
furnace ($1,200); fan coil air conditioning unit ($700);
ductwork ($400); and 4,000-gallon propane fuel tank
$8,000) are estimated to be $10,300. In the cases of the
12- and 24-unit Matrix systems, the units can be stacked
to reduce square-footage costs. Thus, shelter buildings
are needed that measure 1,800 and 3,600 square feet,
respectively. The furnace, air conditioning unit, and
ductwork costs are assumed to increase in proportion to
bulWing size, and the propane tank costs are assumed to
remain constant. The total shelter building construction
costs are estimated to be $136,300 forthe 2-gpm system;
$264,600 forthe 12-gpm system; and $521,200 forthe
24-gpm system.
This analysis assumes that four groundwater extraction
wells exist on site and that they are located 200 feet from
the shelter building. No pumps are required to maintain
the flow rate because the Matrix system includes this
equipment. Two-inch diameter polyvinyl chloride (PVC)
piping and installation costs are assumed to be about
$5.50 per linear foot, including trenching and burial. The
total piping installation costs are $4,400 and the same for
all three treatment systems.
This analysis assumes that treated groundwater will be
discharged to the nearest surface water body using
Matrix system pumps. This surface water body is assumed
to be located 1 .OOOfeet from the shelter building. A 2-inch
diameter PVC pipe can be installed in a trench for about
$5.50 per linearfoot. The total effluent discharge line cost
is $5,500 and the same for all three treatment systems.
This analysis assumes that electrical service needs to be
extended from existing power lines on an existing utility
right-of-way to the shelter building for a distance of 500
feet. This extension will require atransformer, panelboards
with circuits, one utility pole, electric line, and an electricity
meter. The total cost of this extension is estimated to be
$3,900. An additional electrical hookup charge of $4,000
is needed to activate power. If the site already has electric
lines, this hookup charge would be the only electricity-
related startup cost. The total electrical cost for this
analysis is $7,900 and is the same for all three treatment
systems.
The total treatment area preparation costs are estimated
to be $154,100 for the 2-gpm system; $282,400 for the
12-gpm system; and $539,000 for the 24-gpm system.
A treatability study will be conducted by Matrix in order to
determine the appropriate Matrix treatment system
configuration. The cost of the treatability study varies
from $1,500 to $5,000 based on waste characteristics
and Matrix client needs. This analysis assumes a
treatability study cost of $3,000, including labor and
equipment costs, for all three treatment systems. System
design activities are also required to determine the Matrix
system configuration that will achieve treatment goals.
These costs are included in the costs of the capital
equipment and are therefore not included in this cost
category.
Total site preparation costs are estimated to be $175,900
for the 2-gpm system; $304,200 for the 12-gpm system;
and $560,800 for the 24-gpm system.
3.3.2 Permitting and Regulatory Costs
Permitting and regulatory costs depend on whether
treatment is performed at a Superfund or a RCRA
corrective action site and on how treated effluent and any
solid wastes generated are disposed of. Superfund site
remedial actions must be consistent with ARARs that
include environmental laws; ordinances; regulations; and
statutes, including federal, state, and local standards and
criteria. Remediation at RCRA corrective action sites
requires additional monitoring and recordkeeping, which
can increase base regulatory costs by 5%. In general,
ARARs must be determined on a site-specific basis.
Most permits that may be required for the Matrix system
are based on local regulatory agency requirements and
treatment goals for a particular site. Discharge to a
surface water body requires an NPDES permit. The cost
of this permit is based on regulatory agency requ irements
and treatment goals for a particular site. The NPDES
permit is estimated to cost $5,000, including fees and
preparation costs, and is the same for all three treatment
systems.
40
-------
3.3.3 Mobilization and Startup Costs
Mobilization and startup costs include the costs of
transporting the Matrix system to the site, mobilizing
Matrix personnel to the site, assembling the Matrix system,
and performing the initial shakedown of the treatment
system. Matrix will assemble and shake down the Matrix
system. Matrix personnel are trained in hazardous waste
site health and safety procedures, so health and safety
training costs are not included as a direct startup cost.
Initial operator training is needed to ensure safe,
economical, and efficient operation of the system. Matrix
provides initial operator training to its customers at no
additional cost.
Equipment mobilization costs are site-specific and vary
depending on the location of the site in relation to London,
Ontario, Canada. For this analysis, the Matrix equipment
is assumed to be transported 500 miles to allow
mobilization of the system in the Midwestern United
States. A two-person Matrix crew will transport smaller
systems to the site in a Matrix semitrailer truck. Larger
systems may require retaining a cartage company's
services. For the 12- and 24-gpm systems, combined
equipment and mobilization costs are assumed to be two
and three times, respectively, those of the2-gpm system.
Total combined equipment and personnel mobilization
costs are $3,000 for the 2-gpm system; $6,000 for the 12-
gpm system; and $9,000 for the 24-gpm system. Sites
located over 500 miles from London, Ontario, Canada,
may require the services of acartage company totransport
the Matrix system equipment and air transport for Matrix
personnel.
Matrix personnel will perform assembly and shakedown
activities. Assembly and shakedown costs vary depending
on the size of the Matrix system. For this analysis, a two-
person crew is assumed to work 5 8-hour days to unload,
assemble, and hook up the system and perform the initial
shakedown at an estimated cost of $45 per hour per
person. A forklift will be rented for 1 day at a cost of $180
to unload the system from the trailer. Assembly and
shakedown activities are expected to take two times
longer for the 12-unit system and three times longer for
the 24-unit system than the 2-unit system. Total assembly
and shakedown costs are $3,800 for the 2-gpm system;
$7,600 for the 12-gpm system; and $11,400 for the 24-
gpm system.
Total mobilization and startup costs are estimated to be
$6,800 for the 2-gpm system; $13,600 for the 12-gpm
system; and $20,400 for the 24-gpm system.
3.3.4 Equipment Costs
Equipment costs include the costs of purchasing the
Matrix treatment system, a cartridge filtration system for
solids removal before ion exchange, and an ion-exchange
system for metals removal.
Matrix will configure and provide the appropriate number
of treatment units based on site-specific conditions. Each
unit consists of 72 UV lamps along with quartz tubes
wrapped with TiOa-bonded fiberglass mesh. The Matrix
system includes a 1 -micron filter at the influent end and
an appropriately sized h^Oa feed tank.
The Matrix system can be leased or purchased. Matrix
estimates that the capital equipment costs for the 2-, 12-,
and 24-gpm systems are $60,000; $360,000; and
$720,000, respectively. These costs show thatthe Matrix
system capital costs are in linear proportion to unit size.
Matrix estimates that leasing costs are $3,000 per month
for the 2-unit system; $1 6,000 per month for the 1 2-unit
system; and $30,000 per month for the 24-unit system.
Based on these costs, it is less expensive to purchase the
systems presented in this analysis.
Filtration will be required to remove any suspended
solids from the groundwater prior to metals pretreatment.
This analysis assumes that the 2-gpm system has two 3-
micron cartridge filter units placed in parallel and located
upstream of the metals pretreatment system. Each
cartridge filter unit costs about $2,000, for a total filter
system cost of $4,000. Matrix includes a 1 -micron influent
filter with its system at no additional cost. This analysis
assumes that the number of cartridge filters increases
linearly with the flow rate. Thus, the 1 2-gpm system will
require 1 2 filter units for a total cost of $24,000; and the
24-gpm system will require 24 filter units for a total cost
of $48,000.
This analysis also assumes that a metals pretreatment
system is needed to remove iron and manganese from
the groundwater. It is further assumed that this removal
will be accomplished using an ion-exchange system
located immediately downstream of the cartridge filters
described above. During the SITE demonstration, iron
was present in groundwater at a concentration of 1 6 mg/
L and manganese was present at about 1 0 mg/L, which
are fairly high concentrations for Matrix system influent.
Matrix requires that the combined iron and manganese
concentrations in the influent be less than 1 mg/L. This
analysis assumes that the ion-exchange system is
installed and maintained in the shelter building. Based on
costs incurred during the SITE demonstration, the initial
cost of the ion-exchange system is estimated to be
$24,000 for the 2-gpm system. The costs for larger ion-
exchange units tend to increase by 41 %for every doubling
of the flow rate (McArdle and others 1 988). Thus, the ion
exchange system cost for the 1 2-gpm system is estimated
to be $57,500 and $81 ,000 for the 24-gpm system. No
other capital equipment is needed to complete the
groundwater remediation project.
Total equipment costs are estimated to be $88,000 for
the 2-gpm system; $44 1,500 for the 12-gpm system; and
$849,000 for the 24-gpm system.
3.3.5 Labor Costs
Once the system is functioning, it is assumed to operate
continuously at the designed flow rate except during
routine maintenance, which the operator is assumed to
conduct. The operator, trained by Matrix, also performs
routine equipment monitoring and sampling activities.
Matrix estimates that under normal operating conditions,
41
-------
an operator is required to monitor the system about three
times per week, regardless of system size.
This analysis assumes that the equipment monitoring
and confirmatory sampling work is conducted by a full-
time employee of the site owner, who will be considered
the Matrix system primary operator, Further, it is assumed
that a second person also employed by the site owner will
be trained to act as a backup to the primary operator.
Based on observations made at the SITE demonstration,
it is assumed that operation of the system requires about
one-quarter of the primary operator's time, Assuming
that the primary operator earns $40,000 per year, the
total direct annual labor costs are estimated to be $1 0,000.
The primary operator is not expected to spend a
significantly greater amount of time monitoring larger
treatment systems. The increased time performing
maintenance activities is assumed to be minimal between
the three treatment systems, and any differences are
accounted for In the assumptions outlined in Section
3,3.11, As a result, labor costs are expected to be the
same for all three treatment systems,
3,3.6 Supplies Costs
The supplies considered in this analysis can be grouped
Into two categories; direct supplies and indirect supplies.
The former Includes supplies directly associated with the
operation of the Matrix system. The latter includes supplies
associated with completing a grourtdwater remediation
project. For this analysts, direct supplies include H2O2,
distilled water, TiOa mesh, UV lamp assembly, and
disposal drums. Indirect supplies include filters, ion-
exchange resins, disposable personal protective
equipment (PPE) for health and safety Level D, sampling
and field analytical supplies, and propane gas service.
HaOa is commercially available as a solution of 30 to 50%
by weight. It can be purchased in bulk, delivered to the
site when needed, and stored in an appropriately sized
tank that is part of the complete treatment system. I^Oa
has a shelf life of about 1 year and a density of about
1 0 pounds per gallon. A 50% solution is estimated to cost
S0.20 per pound, including delivery (treatment scenarios
requiring larger amounts of HaOa will have a much lower
unit cost because feedstock can be purchased in bulk
quantity). Based on observations made at the SITE
demonstration, HaC^ was consumed at a rate of 173
pounds per year. The 1 2-gpm and 24-gpm systems are
assumed to consume 1,035 and 2,070 pounds of HaOa
peryear, respectively. Annual h^Os costs are about $1 00
for the 2-gpm system; $400 for the 1 2-gpm system; and
for the 24-gpm system.
Distilled water is needed to dilute the HaGa solution for
preparing the HaOt feed. Distilled water is assumed to be
purchased In bulk and stored in an appropriately sized
tank that is part of the complete treatment system.
Distilled water Is estimated to cost about $0.75 per
gallon. Based on observations made at the SITE
demonstration, distilled water is consumed at a rate of
2,000 gallons peryear. The 1 2-gpm and 24-gpm systems
are assumed to consume 1 1 ,900 and 23,800 gallons of
distilled water per year, respectively. Annual distilled
water costs are $1,500 for the 2-gpm system; $8,900 for
the 12~gpm system; and $17,800 for the 24-gpm system.
TiOa-bonded fiberglass mesh requires replacement over
time. According to Matrix, the mesh requires changing
about every 2 years. TiOa mesh for each cell costs about
$75. The treatment system in the base-case analysis
uses 144 cells. The 12-gpm and 24-gpm systems are
assumed to use 864 and 1,728 cells, respectively. Total
TiQa mesh costs on an annual basis are $5,400 for the 2-
gpm system; $32,400 for the 12-gpm system; and $64,800
for the 24-gpm system.
The UV lamp assembly also requires replacement over
time. This analysis assumes that the U V lamp assembly
requires replacement every year. The 2-gpm treatment
system in this analysis uses 144 75-watt UV lamps. The
12-gpm and 24-gpm systems use 864 and 1,728 lamps,
respectively. Each lamp costs about $50. Total annual
are $7,200 for the 2-gpm system; $43,200 for the
12-gpm system; and $86,400 for the 24-gpm system.
Spent TiO2 mesh, spent UV lamps, used filters, and
disposable PPE are assumed to be'hazardous and need
to be disposed of in 55-gaIlon, drums. Most of the
wastes placed in the drums on an annual basis for each
system will consist of spent UV lamps. As a result,
disposal drum costs are assumed to be attributable to the
direct costs of operating the Matrix system. This analysis
assumes that one drum will be filled every 4 months, for
a total of three drums peryear forthe2-gpm system, the
12-gpm and 24-gpm systems are assumed to fill 18 and
36 drums per year, respectively. Each drum costs about
$25. Total annual drum costs are $100 for the 2-gpm
system; $500 for the 12-gpm system; and $900 for the
24-gpm system.
This analysis assumes that cartridge filters are used for
both the pretreatment system and Matrix system influent
filter. The costs and consumption rates for the two filter
applications are assumed to be the same. This analysis
assumes that for the base-case scenario, two cartridge
filter units in parallel are needed to remove solids larger
than 3-microns in size from the groundwater. This dual-
unit system allows one unit to be used while the other is
being changed. The units are installed upstream of the
ion-exchange system and contain four filters each. The
Matrix influent filter unit contains one filter. Replacement
frequency of the filters depends on the quality of the
groundwater and the flow rate. Used filters are assumed
to be hazardous and require proper storage and disposal.
This analysis assumes that the filters will be changed
once every month for a total of five filters per month for the
2-gpm system. The 12-gpm and 24-gpm systems are
assumed to use 360 and 720 filters per year, respectively.
Each filter costs $10, including delivery. Total annual
filter costs are $600 for the 2-gpm system; $3,600 for the
12-gpm system; and $7,200 for the 24-gpm system.
This analysis assumes that an ion-exchange system will
be used to remove manganese and iron from the
groundwater. This system is installed downstream of the
-------
pretreatment cartridge filter units and upstream of the
treatment system. Replacement frequency of the ion-
exehange resins depends on theconcentrations of metals
in the groundwater and the groundwater flow rate. This
analysis assumes that the firm that designs the system
will provide routine maintenance of the system, replace
the resins when necessary, and regenerate the resins off
site. Based on observations made during the SITE
demonstration, total annual resin changeout costs are
about $18,000 for the 2-gpm system; $108,000 for the
12-gpm system; and $216,000 for the 24-gpm system.
Disposable PPE typically consists of latex inner gloves,
nitrile outer gloves, and safety glasses. This PPE is
needed during periodic sampling maintenance
activities. Total annual disposable PPE costs for the
primary operator are assumed to be about $600 and the
same for all three treatment systems.
Sampling supplies consist of sample bottles and
containers, ice, labels, shipping containers, and laboratory
forms for off-site analyses. The actual number and types
of sampling supplies needed depend on the analyses to
be performed. This analysis assumes that the treatment
process effluent will be sampled monthly for VOC analysis.
Costs for laboratory analyses are discussed in
Section 3.3.10. Annual sampling supply costs are
estimated to be $400 for each system,
Propane fuei delivery service is needed to provide propane
for heating the shelter building. Annual propane usage is
based on the square footage of the shelter building,
number of cold days, building materials, and other
variables. Annual propane costs are assumed to be $800
for the 2-gpm system. Propane consumption for the
larger systems is assumed to increase linearly with the
shelter building size. As a result, for the 12-unit and 24-
unit systems, propane costs are expected to be two and
fourtimes the cost of the 2-unit system; therefore, annual
propane costs are $1,600 for the 12-gpm system and
$3,200 for the 24-gpm system.
Total annual supply costs are estimated to be $34,700 for
the 2-gpm system; $199,600 for the 12-gpm system; and
$398,100 for the 24-gprn system.
3.3.7 Utilities Costs
Electricity is the only utility used by the Matrix system.
Electricity is used to run the Matrix treatment system and
shelter building air conditioning. Electricity costs can vary
considerably depending on the geographical location of
the site and local utility rates. Ultimately, the consumption
of electricity varies depending on the total number of
Matrix units, the total number of pumps, and other electrical
equipment used.
This analysis assumes a constant rate of electricity
consumption based on treatment system design
specifications of 11 kW for the 2-unit system. However,
during the SITE demonstration, actual electrical usage
was observed to be 9 kW per hour, probably as a result
of operating the UV lamps at less than maximum power.
Other electrical equipment is assumed to draw an
additional 20 percent of electrical energy. As a result, the
entire 2-gpm system operating for 1 hour draws about 11
kW-hours (kWh) of electricity. The total annual electrical
energy consumption, considering 10% equipment
downtime, is estimated to be about 86,720 kWh for the 2-
gpm system, The 12-gpm and 24-gpm systems are
assumed to draw 66 kWh and 132 kWh of electricity,
respectively. Electricity is assumed to cost $0.09 per
kWh, including demand and usage charges. Total annual
electricity costs are estimated to be about $7,800 for the
2-gpm system; $46,800 for the 12-gpm system; and
$93,700 for the 24-gpm system.
3.3.8 Effluent Treatment and Disposal
Costs
The treated effluent is assumed to be discharged directly
to a nearby surface water body, provided appropriate
permits have been obtained Section 3.3.2). During
the SITE demonstration, the Matrix system did not meet
target treatment levels for most of the VOCs. Depending
on the treatment goals for a site, additional effluent
treatment will probably be required; therefore, additional
treatment or disposal costs may be incurred. Because of
the uncertainty associated with the need for additional
treatment, this analysis does not estimate effluent
treatment or disposal costs.
3.3.9 Residual Waste Shipping and
Handling Costs
The residuals produced during Matrix system operation
include 55-gallon drums containing spent TiOa mesh,
spent UV lamps, used cartridge filters, used PPE, and
waste sampling and field analytical supplies. These
wastes are considered hazardous and require disposal
at a permitted facility. Most of the waste generated by
each system will consist of spent UV lamps. As a result,
the waste shipping and handling cost is assumed to be
attributable to the direct cost of operating the Matrix
system. This analysis assumes that wastes are disposed
of at a commercial hazardous waste landfill located 200
miles from the site. The costs of loading, transportation,
disposal of the drums and one-time waste stream analysis
is estimated to be about $1,000 per drum. Total annual
drum disposal costs are $3,000 for the 2-gpm system;
$18,000 for the 12-gpm system; and $36,000 for the 24-
gpm system.
3.3,10 Analytical Services Costs
Required sampling frequencies and number of samples
analyzed are highly site-specific and are based on permit
requirements. Analytical costs associated with a
groundwater remediation project include the costs of
laboratory analyses, data reduction, and QA/QC. This
analysis assumes that one sample of treated water and
associated QC samples (trip blanks and field blanks) will
be analyzed for VOCs every month. Monthly VOC analysis
costs including data reduction (also known as the
documentation package) costs are assumed to be $300.
Total annual analytical services costs are estimated to be
$3,600 for all three treatment systems.
43
-------
3.3* 11 Equipment Maintenance Costs
This analysis assumes that annual Matrix system
maintenance costs are about 5% of the capital equipment
costs. This cost covers the costs of equipment
replacement and repair. Maintenance labor is discussed
In Section 3,3.5, Total annual equipment maintenance
costsare about $3fOOOforthe2-gpm system; $18,000 for
the 12-gprn system; and $36,000 for the 24-gpm system.
3.3,12 Site Demobilization Costs
Site demobilization activities include utility disconnection,
treatment system shutdown, decontamination, and
disassembly costs. Salvage value of the Matrix system
components can be used to offset a portion of
demobilization costs. Utility disconnection costs are about
$1,000. To decontaminate and disassemble the treatment
systems, a two-person crew will work about two 8-hour
days for the 2-gpm system, four 8-hour days for the 12
§pm system, and six 8-hour days for the 24-gprn system.
The labor and equipment costs for decontamination and
disassembly are$2,100;$4,200;and $6,300, respectively.
Site cleanup and restoration activities, such as piping
removal, shelter building demolition, regrading, and
materials disposal, may also be conducted at this time;
however, these costs are not estimated in this analysis.
This analysis assumes that the equipment will have a
salvage value of 5% of the original equipment purchase
cost. As stated earlier, the salvage value is not for tax or
depreciation purposes, A cash receipt is assumed to be
realized by the equipment owner at the time of
demobilization. For this analysts, the salvage value is
$3,000 for the 2-gpm system; $18,000 for the 12-gpm
system; and $36,000 for the 24-gpm system. Total cost
of demobilization In current dollars is $100 for the 2-gpm
system. A total credit In current dollars of $12,800 will be
realized for the 12-gpm system, and a total credit in
current dollars of $28,800 will be realized for the 24-gpm
system.
The costs of demobilization, however, will occur at the
end of the remediation project, inthis analysis, the 2-gpm
project will take 30 years, the 12-gpm project will take 5
years, and the 24-gpm project will take 2.5 years to
complete. This analysis calculates the future value of the
current dollar demobilization costs discussed above in
order to present adjusted costs expected to be Incurred
at the end of the groundwater remediation project. At that
time, the 2-apm system will cost $400; the 12-gpm
system credit will be $16,400; and the 24-gpm system
credit will be $32,500 for site demobilization.
3.4 Conclusions of Economic Analysis
This economic analysis considers a base-case scenario
where the Matrix treatment system is used to treat
groundwater contaminated with VOCs at a flow rate of 2
gpm for 30 years. The base-case scenario assumes that
the total volume of groundwater to be treated is 28.4
million gallons. Table 3-1 presents a breakdown of costs
forthe 12 cost categories evaluated inthis analysis. The
table also provides costs forfull-scale systems operating
at 12 and 24 gpm.
Figure 3-1 shows the distribution of one-time and annual
O&M costs forthe 2-, 12-, and 24-gprn systems. Costs
presented in Figure 3-1 are derived from Table 3-1 and
present total costs and percentages for most cost
categories. When cost categories are not presented,
such as permitting and regulatory, and effluent treatment
and disposal, it is because they represent less than
1 percent of either the total one-time or annual O&M
costs. Site demobilization costs are not presented because
they represent less than 1% of the total one-time cost (2-
gpm system) or they represent a credit (12- and 24-gpm
systems).
Forthe 2-gpm base case, total estimated one-time costs
are about $275,800, Of this total, $175,900, or 64%, is for
site preparation activities. About 77% of the site
preparation costs are for constructing a shelter building
forthe treatment system. Although this cost is not directly
attributable to operating the treatment system, it is
necessary for protecting the system from inclement
weather. Equipment costs total $88,000, or 32% of the
one-time costs, of which $60,000 isforthe Matrix treatment
system. Total estimated annual O&M costs are about
$62,200. Supply costs comprise about 56% of this total.
Most of the supply costs areforcomponentsof the Matrix
system that need periodic replacement (such as TIO2
mesh and UV lamps) or feedstocks consumed during
treatment (such as HsOg and distilled water). The annual
O&M costs are incurred for 30 years and adjusted by an
annual inflation rate of 5%. When the annual O&M costs
are added to the one-time cost, the total cost for the 2-
gpm groundwater remediation project is estimated to be
over $4.6 million. The net present value of this figure is
about $1.8 million, which results in a treatment cost of
$64.74 per 1,000 gallons treated.
Costs per 1,000 gallons of groundwater treated increase
with the size of the treatment system because most costs
in this analysis increase in linear proportion as treatment
system flow rates increase, while the total volume of
watertreated remains constant. Thus, inthis instance, no
economies of scale are realized. For a groundwater
remediation project using the Matrix treatment system,
the cost per 1,000 gallons treated is estimated to be
$78.78 forthe 12-gpm system and $108.56 forthe 24-
gpm system.
Table 3-1 highlights in bold italics Matrix treatment system
direct costs, which are summarized in Table 3-2. Figure
3-2 shows the distribution of one-time and annual O&M
direct costs for the 2-, 12-, and 24-gpm systems. This
analysis is provided to segregate the direct costs of
procuring and operating the Matrix system from the total
costs of a groundwater remediation project. For the 2-
gpm base case, the total one-time direct costs are
estimated to be $69,900, and total annual O&M direct
costs are estimated to be $28,100. The direct cost per
1,000 gallons of groundwater treated is estimated to be
$28.53 for the 2-gpm system. Compared to the 12- and
24-gpm systems, no cost savings result from economies
44
-------
Site Preparation
f 175,900
(63.8%)
Permitting and
Regulatory $5,000
(1.8%)
Mobilization and
Startup $8,800
(2.5%)
Equipment
S88.000
(31.9%)
Supplies $34,700
(55.8%)
Equipment
Maintenance
$3,000 (4.8%)
Residual Waste
Shipping and
Handling $18,000
Labor $10,000
" (16.1%)
Utilities $7,800
(12.5%)
Analytical
Services $3,600
(5.8%)
Total One-Tirne Costs
$275,800
Total Annual Q&M Costs
$62,200
2-gpm
Equipment
$441,500
(58.8%)
Site Preparation
$304,200
-- (40.5%)
Supplies $199,600
(67.4%)._,.,..
Mobilization and
Startup $13,600
(1,8%)
Equipment
Maintenance
$18,000 (6.1%)
Residual Waste
Shipping and
Handling f 18,000
(6.1%)
Labor $1 0,000
(3.3%)
., Analytical
"^* Services
Utilities $46,800 $3,500 (1 2%)
(15.8%)
Total One-Time Costs
$751,500
Total Annual O&M Costs
$296,000
12-gpm
Site Preparation
$560,800
(39.9%)
Equipment
$849,000
(60.4%)
Supplies $398,100
(69.0%)
Mobilization and
Startup $20,400
(1.5%)
Equipment
Maintenance
$36,000 (S.2%)
• Labor $10,000
(1.7%)
Residual Waste
Shipping and
Handling $38,000
(6.2%)
Utilities $93,700
(16.2%)
Total One-Time Costs
$1,406,400
Total Annual O&M Costs
$577,400
24-gpm
Figure 3-1. Distribution of One-Time and Annual O&M Costs for a Groundwater Remediation Project.
NOTE; The sums of the tote! one-time cost percentages for the 12- and 24-gpm systems do not equal 100 because the costs include a
credit for site demobilization. The sum of the total annual O&M cost percentages for the 24-gpm system does not equal 100 because it
excludes analytical services costs, which represent less than 1 percent of the total annual O&M costs.
45
-------
Tabfo 3-2, Matrix Treatment Sptem Direct Costs®
2-gprn
Cost Category Itemized
Site Preparation"
TreataMity study and 3,000
system design
Mobilization and Startup"
Equipment and personnel 3,000
mobilization
Assembly and shakedown 3,800
Shipment1'
Matrix treatment system 60,000
Supplies'
HO. 100
Distilled water 1,500
TKXmesh 5,400
UV tamp assembly 7,200
Disposal drums 100
Utilities'
Matrix treatment system 7,800
Residual Waste Shipping and
Handling'
Equipment Maintenance*
Matrix treatment system 3,000
Site Demobilization''''' 400
Total One-Time Costs
Total Annual O&M Costs
Total direct costs *
Net present value'
Costs per 1 ,000 Gallons1-'
SystgjTj
Total
$ 3,000
6,800
60,000
14,300
7,800
3,000
3,000
100
$ 69,900
28,100
$2,058,300
810,300
$28.53
12-ppmJ3vstem
Itemized
3,000
6,000
7,600
360,000
400
S.iOO
32,400
43,200
500
46,800
18,000
(16,400)
Total
$3,000
13,600
360,000
86,400
46,800
18,000
18,000
(12,800)
$ 363,800
168,200
$1,507,900
1,220,200
$ 42.96
24-gpm System
Itemized
3,000
9,000
11,400
720,000
800
17,800
64,800
86,400
900
93,700
36,000
(32,500)
Total
$ 3,000
20,400
720,000
170,700
93,700
36,000
36,000
(28,800)
$714,600
336,40
$1,967,400
1,441,500
$50.76
* Costs are in July 1996 dollars and are rounded to the nearest $100,
11 Qna-tima costs
* Annual O&M costs
* The values presented in the itemized columns represent the cost in future values. The
the appropriate 1996 current dollars. Values in parentheses represent
* One-time and annual O&M costs combined
We! present value calculated using tha same
« ToW of 28,4 million gallons treated
a credit value.
total columns show
assumptions used in Table 3-1
of scale, again because the total volume of water treated
stays constant and because the Matrix system direct
costs increase In linear proportion with system size.
Figure 3-3 shows the relative distribution of direct costs
per 1,000 gallons treated.
For the most part, costs increase in linear proportion in
this analysis because economies of scale cannot be
realized. In practice, however, economies of scale will
most likely be realized on annual supply prices for the
larger two systems. Further, Matrix does not yet have
attractive options for short-term projects. This
requires system users to purchase the treatment
equipment, which results in a high capital equipment cost
over a short period of time.
To eliminate this short-term impact on the costs and
further illustrate the linear relationship of costs among the
three treatment systems, an additional analysis is provided
for the three systems operating over 30 years at their
designed flow rates. This analysis is presented in Table
3-3. Operating for 30 years with an annual downtime of
10%, the 12-gpm system treats a total of 170 million
gallons and the 24-gpm system treats about 341 million
gallons of groundwater. The groundwater remediation
cost per 1,000 gallons treated is $50,24 for the 12-gpm
system and $48.91 for the 24-gpm system. These figures
show a significant reduction in cost compared to the
short-term projects analyzed In Table 3-1. The cost
reduction probably results from the largertotal volume of
water treated under the 12-gpm and 24-gpm systems
operating for 30 years. The Matrix direct costs per 1,000
gallons treated are approximately $28.50 for all three
treatment systems. This similarity shows the linear
relationship of the treatment costs to the size of the Matrix
treatment system used.
46
-------
Table 3-3, Costs Associated with the
Matrix Technology for Projects Lasting 30 Years8
2-gpm System
Cost Category* Detail
Site Preparation5
Administrative
Treatment area
preparation:
Shelter building 136,300
construction
Piping installation 4,400
Effluent discharge 5,500
line Installation
electrical service 7,900
extension
Treatablllty study
and system design
Permitting and
Regulatory"
Mobilization and Startup"
Equipment and
personnel
mobilization
Assembly and
shakedown
Equipment"
Matrix treatment system
Cartridge filter
Ion-exchange
system
Labord
Supplies"
"A
Distilled water
T/O2 mesh
UV lamp assembly
Disposal drums
Filters
Ion-exchange resins
Disposable PPE
Sampling supplies
Propane gas service
Utilities"
Matrix treatment system
Effluent Treatment and
Disposal"
Residual Waste
Shipping and Handling"
Analytical Servicesd
Equipment Maintenance"1
Matrix treatment system
Site Demobilization''*
Total One-Time Costs
Total Annual O&M Costs
Itemized Total
175,900
18,800
154,100
3,000
5,000
6,800
3,000
3,800
88,000
60,000
4,000
24,000
10,000
34,700
100
1,500
5,400
7,200
100
800
18,000
600
400
800
7,800
7,800
0
3,000
3,600
3,000
3,000
400 100
275,800
62,200
1 2-gpm System
Detail Itemized Total
304,200
18,800
282,400
264,600
4,400
5,500
7,900
3,000
5,000
13,600
6,000
7,600
441,500
360,000
24,000
57,500
10,000
199,600
400
8,900
32,400
43,200
500
3,600
108,000
600
400
1,600
46,800
46,800
0
18,000
3,600
18,000
18,000
(52,700) (12,800)
751,500
298,000
24-gproJSystem
Detail Itemized
18,800
539,000
521,200
4,400
5,500
7,900
3,000
9,000
11,400
720,000
48,000
81,000
800
17,800
64,800
88,400
900
7,200
216,000
600
400
3,200
93,700
93,700
36,000
36,000
(118,500)
Total
560,800
5,000
20,400
849,000
10,000
398,100
0
36,000
3,600
(28,800)
1 ,408,400
577,400
47
-------
Tablo 3,3, Costs Associated with the Matrix Technology for Projects Lasting 30 Years (Continued)
Cost Category
Groundwater
Remediation:
Total costs1-9-11
Net present value*
Costs per 1,000 Gallons"
Matrix-Specific
Treatment Costs:
O na-llme costs
Annual O&M costs
Total Costs'-*'"
Net Present Value'
Costs per 1,000 Gallons'
2-gpm System
Detail Itemized Total
4,670,100
1,838,500
64.74
69,900
28,100
2,058,300
810,300
28.53
1 2-gpm System
Detail Itemized Total
21 ,696,700
8,541,500
50.24
376,600
168,200
12,278,600
4,833,800
28.43
24-gpm System
Detail Itemized Total
42,263,700
16,638,400
48.91
743,400
336,400
24,547,300
9,663,800
28.41
All costs are In July 1996 dollars and are rounded to the nearest $100.
Cost categories appearing in bold italic font are Matrix treatment system direct costs.
One-Urns costs.
Annual O&M costs.
The values presented in the itemized columns represent the cost in future values. The total columns show the appropriate 1996 current
collars. Values in parenthesis represent a credit value.
One-time and annual O&M costs combined.
Future value of O&M costs using annual inflation rate of 5%.
The total volume of water treated is 28.4 million gallons by the 2-gpm system, 170 million gallons by the 1q2-gpm system, and 341 million
gallons by the 24-gpm system with an annual downtime of 10% for all three systems.
Annual discount rate of 7.5%.
Net present value.
48
-------
Experiment $60,000
(85.8%)
\
Site Preparation
$3,000
(4.3%)
Equipment
Maintenance
$3,000
Mobilization and
Startup $6,800
(9.7%)
Supplies $14,300
(50.9%)
Residual Waste
Shipping and
Handling $3,000
(10.7%)
Utilities
$7,800
(27.8%)
Total One-Time Costs
$69,900
Total Annual O&M Costs
$28,100
2-gpm
Eqi
$360,000
(99.0%)
Mobilization and
Startup $13,600
(3.7%)
Supplies $85,400
(50.8%)
Equipment
Maintenance
$18,000
(10.7%)
Residual Waste
Shipping and
Handling $18,000
(10.7%)
Total One-Time Costs
$363,800
Utilities
$16,800
(27.8%)
Total Annual O&M Costs
$168,200
12-gpm
Equipment
$720,000
(100.8%)
Mobilization and
Startup $20,400
(2.8%)
Supplies $170,700
(50.7%)
Equipment
Maintenance
$36,000
(10.7%)
Residual Waste
Shipping and
Handling $36,000
(10.7%)
Total One-Time Costs
$714,600
Utilities
$93,700
(27,8%)
Total Annual O&M Costs
$336,400
24-gpm
Figure 3-2. Distribution of One-Time and Annual O&M Matrix Treatment System Direct Costs.
NOTE: The sums of the total one-time cost perpcentages may not equal 100 because (1) site demobilization costs for the 2-gpm system and
site preparation costs for the 12- and 24-gpm systems are not presented because they represent less than 1 percent of the total one-time
cost of (2) include a credit for the 12- and 24-gpm systems.
49
-------
Mobilization Ec»ment gjte Prepara,ion
Supplies
S13.31
and Start-up
$0.23
$2.03
$0.10
Residual Waste
Handling
" $2.79
•- —, Utilities
$7.26
Equipment
Maintenance
$2.79
2 gpm
Total direct costs per 1,000 gallons treated are $28.53.
The project lasts 30 years.
Mobilization "WS!"'
and Start-up »'•*•''*
$0.52 •
Site Preparation
$0.11
Residual Waste
Handling
""" $3.11
•~~, Utilities
$8.09
X Equipment
Maintenance
$3.11
12 gpm
Total direct costs per 1,000 gallons treated are $42.96,
The project lasts 5 years.
Supplies
$14.77
Equipment
Maintenance
$3.11 ™
Mobilization
and Start-up
$0.52
Equipment
$13.74
Utilities s
$8.09
Residual Waste Site Preparation
Handling $0.11
$3.11
24 gpm
Total direct costs per 1,000 gallons treated are $50.76.
The project lasts 2.5 years.
Figure 3-3. Distribution of Matrix Treatment system Direct Costs per 1,000 Gallons Treated.
NOTE: Tha sum of the individual cost components may not equal the total direct cost per 1,000 gallons treated because site demobilization
costs or credits are not shown in the pie diagrams.
50
-------
Section 4
Technology Status
The Matrix technology is available in custom-made
systems that are generally skid- or trailer-mounted. The
mounting of the system depends on the client's
requirements. Trailer-mounted systems may be suitable
for rough or remote terrains where lifting is unadvisabje.
Also, trailer-mounted units are most suited for sites with
small contaminant zones because these units can be
secured and do not require a paved surface for operation.
Trailer-mounted units are housed in either single-car
carriers or semitrailers. Skid-mounted systems are
generally the most economical packaging for the Matrix
system and are suitable for a site where structural,
weathertight housing is available. Skid-mounted systems
can be temporarily or permanently installed. Each skid-
mounted Matrix system is custom fabricated to meet
specific client and location requirements.
The Matrix system is generally shipped complete with
"no extras needed" and ready to install. Minimal set-up
time is required to begin operation. The system can be
operated in flow-through or batch modes. The physical
design of the system is very adaptable. The Matrix
system can be installed and operated with existing
treatment units.
Matrix often follows initial client contact with an "in-
house" laboratory-scale treatability study. For each
application, the Matrix staff will and analyze 15
gallons of waste to determine the technology's applicability
and generate preliminary cost estimates. The second
phase of the process is an on-site, larger-scale
demonstration where Matrix treats waste at the client's
site to achieve a higher degree of system optimization
and to assess the specific costs involved with system
installation. The Matrix team works closely with the client
or its consultants to optimize system design and
configuration.
51
-------
Section 5
References
AI-Ekabl, H,, B. Butlers, D, Delany, J. Ireland, N. Lewis,
T, Powell, and J. Story, 1993. "Titanium Dioxide
(TIO2) Advanced Photo-Oxidation Technology; Effect
of Electron Acceptors." Photocatalytic Purification and
Treatment of Water and Air, Edited by D.F. OIlis and
H, Al-Ekabl Elsevler Science Publishers B.V,
Amsterdam. Pages 321 to 336.
American Public Health Association (APHA), American
Water Works Association, and Water Environment
Federation. 1992. Standard Methods for the Exami-
nation of Water and Wastewater (Standard Methods).
18th Edition. Washington, DC.
Boltz and Howell. 1979. Colorimetric Determination of
Nonmetals. John Wiley and Sons. New York.
Crystal! Ball®, 1996. Forecasting and Risk Analysis for
Spreadsheet Users. Version 4.0. Decisioneering, Inc.
Aurora, Colorado.
Environmental Cost Handling Options and Solutions
(ECHOS). 1996a. Environmental Restoration Assem-
blies Cost Book. Published by R.S. Means Company,
Inc., and Delta Technologies. Inc.
ECHOS. 1996b. Environmental Restoration Unit Cost
Book. Published by R.S. Means Company, Inc., and
Delta Technologies, Inc.
Evans, G. 1990. "Estimating Innovative Technology
Costs for the Superfund Innovative Technology Evalu-
ation (SITE) Program." Journal of Air and Waste
Management Association, Volume 40, No. 7. Pages
1047 to 1051. July.
Gellert, W., S, Scottwald, M. Heilwich, H. KSstner, and
H. KQstner. 1989. The VNR Concise Encyclopedia of
Mathematics. Second Edition. Van Nostrand Reinhold.
New York, New York.
Glaze, W., G, Peyton, F. Huang, J. Burieson, and P.
Jones, 1980. Oxidation of Water Supply Refractory
Species by Ozone with Ultraviolet Radiation. Munici-
pal Environmental Research Laboratory, Cincinnati,
Ohio. EPA-600/2-80-110. August.
Grundfos. 1996, Price List of Frundfos Products. Effec-
tive Date June 15.
Kletnbaum, D., L. Kupper, and K. Mulier. 1987. Applied
Regression Analysis and Other Multivariable Meth-
ods. PWS-Kent Publishing Company. Boston, Mas-
sachusetts.
Lyman, W., W. Reehl, and D. Rosenblatt. 1990. Hand-
book of Chemical Property Estimation Methods. Ameri-
can Chemical Society. Washington, DC.
Matthews, R. 1988. "Photocatalytic Reactor Design: An
Example of Mass-Transfer Limitations with an Immo-
bilized Catalyst." Journal of Physical Chemistry. Vol-
ume 92, No, 23. Pages 6853 to 6854.
Matthews, R., and S. McEvoy. 1992. "A Comparison of
254-nm and 350-nrn Excitation of TiO2 in Simple
Photocatalytic Reactors." Journal of Photochemistry,
Photoblology, and Chemistry. Volume 66. Pages 355
to 366.
McArdle, J., M. Arozarena, and W. Gallagher. 1988.
"Treatment of Hazardous Waste Leachate." Pollution
Technology Review. No. 151. Noyes Data Corpora-
tion.
Perry, R., and C. Chiton. 1973. Chemical Engineers
Handbook. Fifth Edition. Published by McGraw-Hill,
Inc.
Pichat, P., C. Guillard, C. Mailiard, L. Amalrlc, and J.
D'Oliveira. 1993. "T1O2 Photocatalytic Destruction of
Water Aromatic Pollutants." Photocatalytic Treatment
of Water and Air. Edited by D.F. OIlis and H. AI-Ekabi.
Elseview Science Publishers B.V. Amersterdam.
Pages 207 to 219.
PRC Environmental Management, Inc. (PRC). 1995.
"Matrix Photocatalytic Oxidation Technology Demon-
stration, Finai Quality Assurance Project Plan." Sub-
mitted to U.S. Environmental Protection Agency (EPA)
Office of Research and Development, National Risk
Management Research Laboratory (NRMRL), Cincin-
nati, Ohio. August.
PRC. 1996a. Record of Telephone Conversation Re-
garding Ozone Injection. Between Mary
Wojciechowskl, Environmental Scientist, and Doug
Garside, Matrix Photocatalytic, Inc. April 18.
52
-------
PRC.1996b. Record of Telephone Conversation Re-
garding Ion-Exchange System Pricing, Between Jeff
Swano, Economic Analyst, and Steve Schlatre,
Culligan Water Conditioning of Knoxville, June 6.
PRO. 1996c. Facsimile Regarding Matrix Photocata-
lytlc, Inc., Formaldehyde Study. To Kumar Topudurti,
Project Manager, from Doug Qarslde, Matrix Photo-
catalytic, !nc. September 26.
R,S. Means (Means). 1995, Electrical Cost Data. 18th
Annual Edition.
Means. 1996. Building Construction Cost Data. 54th
Annual Edition.
State of California Water Resources Control Board.
1990. "California Ocean Plan, Water Quality Control
Plan, Ocean Waters of California." October 18.
State of Connectteul Department of Environmental Pro-
tection (CDEP). 1993. Personal Communication Re-
garding the Acute Toxicity of Hydrogen Peroxide to
Freshwater Organisms. From A. facobucci. To N.
Lewis, EPA NRMRL.
Tanaka, K,, T. Hisanaga, and A. Rivera. 1993. "Effect of
Crystal Form of TiO2 on the Phoiocatalytic Degradation of
Pollutants," Photocatalytlc Treatment of Water and Air,
Edited by D.F. Ollis and H. AI-Ekabi. Elsevler Science
Publishers B.V. Amsterdam. Pages 169 to 178.
Topudurti, K., N. Lewis, and S. Hirsh. 1993. "The Appli-
cability of UV/Oxidation Technologies to Treat Con-
taminated Groundwater." Environmental Progress.
Volume 12, No. 1. Pages 54 through 60. February,
Turchi, C., and D. Ollis. 1988. "Photocatalytte Reactor De-
sign: An Example of Mass-Transfer Limitations with an
Immobilized Catalyst." Journal of Physical Chemistry. Vol-
ume 92, No. 23. Pages 6852 and 6853.
U.S. Department of Energy (DOE). 1988. Radioactive
Waste Management Order, DOE Order 5820.2A. Sep-
tember.
U.S. Environmental Protection Agency (EPA). 1983.
Methods for Chemical Analysis of Water and Wastes.
Environmental Monitoring and Support Laboratory,
Cincinnati, Ohio. EPA/600/4-79/020. March.
EPA, 1987a. Alternate Concentration Limit (ACL) Guid-
ance. Part 1: ACL Policy and Information Require-
ments. EPA/530/SW-87/017,
EPA. 1987b. Joint EPA-Nuclear Regulatory Agency
Guidance on Mixed Low-Level Radioactive and Haz-
ardous Waste. Office of Solid Waste and Emergency
Response (OSWER) Directives 9480.00-14 (June 29),
9432.00-2 (January 8), and 9487.00-8. August.
EPA. 1988a. Protocol for a Chemical Treatment Dem-
onstration Plan. Hazardous Waste Engineering Re-
search Laboratory. Cincinnati, Ohio. April.
EPA, 1988b. CERCLA Compliance with Other Environ-
mental Laws: Interim Final. OSWER. EPA/540/G-89/
006. August.
EPA. 1989. CERCLA Compliance with Other Laws
Manual: Part II. Clean Air Act and Other Environmen-
tal Statutes and State Requirements. OSWER. EPA/
540/G-89/006. August.
EPA, I994a. Information Collection Rule. 59 Federal
Register 6332. February 10.
EPA. I994b. Test Methods for Evaluating Solid Waste,
Volumes 1A through 1C, SW 846, Third Edition.
Updates II and IIA. OSWER. Washington, DC. Sep-
tember.
53
-------
Appendix A
Vendor's Claims for the Technology
The need for effective groundwater remediation
technologies has become more critical as discharge
regulations become more stringent and conventional
groundwater containment and remediation methods
become outdated; however, regulatory compliance is
often very expensive. These factors contribute to the
need for the advancement of Innovative groundwater
treatment technologies. The viability of an individual
innovative treatment technology depends on the
technology's capability and efficiency.
A.1 Introduction
Matrix Photocatalytic, Inc. (Matrix), began developing its
titanium dioxide (TiOa) technology in 1988, which has
allowed time to develop the technology to successfully
treat a wide variety of hazardous organic contaminants.
Matrix has produced a line of highly efficient, commercially
ready TiOa photocatalytic treatment systems that can
treat organic contaminants in both air and liquids.
A.2 Technology Description
The Matrix photocatalytic oxidation system utilizes an
illuminated TtOa matrix. The baste component of the
system is a photocatalytic reactor cell composed of an
outer stainless steel jacket that contains an internal
photocataiytic matrix, a quartz sleeve, and an ultraviolet
(UV) lamp. The lamp emits low-intensity (normally 254
nanometer) UV light and is mounted coaxially within the
Jacket in a quartz sleeve wrapped with a special fiberglass
mesh bonded with TiOa, which forms the catalyst matrix.
The TiOg catalyst is activated by UV light, resulting in a
momentary shift of an electron in the catalyst to a much
higher energy orbital, creating a reducing environment.
The hole left momentarily by the shift of the electron
exhibits a powerful oxidation effect that breaks down and
mineralizes (destroys) organic molecules in a true
reduction-oxidation reaction. Contaminated air or liquid
flows into the reactor cell and passes through the catalyst
matrix, where organic contaminants are oxidized and/or
reduced (if applicable, as for halogenated organics) into
nonhazardous products.
A.3 Advantages of the Matrix
Photocatalytic Oxidation
Technology
The Matrix TiCfe photocatalytic oxidation technology offers
many benefits over other technologies that treat air or
liquids contaminated with organics. These benefits are
listed below.
• Ability to achieve low part per billion (ppb) or part
per trillion (ppt) contaminant concentrations
Highly effective against polychlorinated biphenyls,
dioxins, furans, and other compounds
Destruction of organic pollutants at the source loca-
tion to eliminate waste handling
No long-term disposal requirement or liability
• Quiet, low-profile, aesthetic solution to environmen-
tal problems
• Ambient temperature process (no Ignition source)
Meets current environmental standards
Attractive acquisition costs because system com-
posed of many "off-the-shelf* components
System's form has very low life-cycle costs
because of high reliability, minimal maintenance,
and no consumable chemical requirements
Primary power source is electricity (220 volts)
System material is recyclable
Functions over a broad range of temperatures,
pressures, and pHs
* Air treatment systems do not generate nitrous ox-
ides or phosgene and therefore do not generate
permitting problems associated with thermal (high-
temperature) treatment systems
54
-------
Air treatment systems not adversely affected by
humidity
Able to destroy highly resistive organics such as
dinitrotoluene and carbontetrachloride
Light-weight system offers lowest cost method for
mass production of hydroxyl radicals compared to
other advanced oxidation processes
TiO2 harmless to human and other life forms and its
toxicity is among the lowest known
Easy installation
Modular system construction adjusts to flow re-
quirements
No operator required
Portable, weather-resistant construction available
A.4 Treatment Systems
The Matrix system is modular in design to allow complete
scaling capabilities to accommodate individual waste
streams. The waste stream's flow rate, contaminant
concentration, andtargetconcentrationdetermine system
sizing. The modulardesign also allows increased flexibility
if the parameters discussed above change over time.
The system can be set up for either flow-through (in-line)
or batch treatment mode operation.
A.5 System Applications
The Matrix photocatalytic technology can be used to
destroy chlorinated or unchlorinated organic contaminants
and reduce total organic carbon (TOC) in the following
applications:
Liquids
Ultrapure Water
Drinking Water
Groundwater
Plant Process Water
Mineralization of TOC source
to low ppb or ppt range
Color removal
Odor removal
Trihalomethane removal
Destruction of organics
Destruction of organics
Mineralization of TOC source
Biochemical oxygen demand
and chemical oxygen
demand reduction
Color removal
Odor removal
Destruction of organics
Mineralization of TOC source
"Closed loop" applications
Air
(works extremely well on oxygenated compounds such
as ethers, alcohols, and acetone)
Ultrapure Air
Ambient Air
Soil Remediation
Plant Air Emissions
Remediates air to quality
suitable for use in instruments
and semiconductor rooms
Odor destruction
Soil venting emissions
Air stripping emissions
Absorbent regeneration
Dry cleaning emissions
Degreasing facility emissions
Incinerator emissions
A.6 Cost Considerations
Treatment cost depends on contaminant concentration,
treatment flow rate, and target removal concentration on
a contaminant-specific basis. Matrix offers "in house"
applicability studies and on-site demonstrations as
evaluation precursors to system installation. Matrix
systems are availablethroughleasingoptionsorpurchase.
A.7 Bibliography
AI-Ekabi, H., and others. 1993. "Titanium Dioxide (TiO2)
Advanced Photo-Oxidation Technology: Effect of Elec-
tron Acceptors." Photocatalytic Purification and Treat-
ment of Water and Air. Edited by D.F. OIlis and H. AI-
Ekabi. Eisevier Science Publishers B.V. Amsterdam.
Pages 321 to 335 and References Contained Therein.
AI-Ekabi, H., and others. 1993. 'The Photocatalytic
Destruction of Gaseous Trichloroethylene and Tetra-
chloroethylene Over Immobilized TiO,." Photocata-
lytic Purification and Treatment of water and Air.
Edited by D.F. OIlis and H. AI-Ekabi. Eisevier Science
Publishers B.V. Amsterdam. Pages 719 to 725 and
References Contained Therein.
AI-Ekabi, H., and others. 1992. "Water Treatment By
Heterogenous Photocataiysis." Chemical Oxidation
Technologies for the Nineties. Edited by W. Wesley
Eckenfelder and others. Technomic Publishing Co.,
Inc. Pages 254 to 263 and References Contained
Therein.
55
-------
Appendix B
Case Study
An extensive independent research study was conducted
by Atomic Energy Canada Laboratories (AECL) to find an
alternate method of waste treatment that does not
generate secondary wastes. The test was performed at
Chalk River Laboratories in Chalk River, Ontario, Canada,
on an aqueous, radioactive waste stream contaminated
with bitumen. AECL prepared a report based on the
study. This appendix breifly summarizes the report
(Sen Gupta and others 1994).
B.1 Site Conditions
Two photochemical oxidation systems were comparatively
tested: an unnamed, ultraviolet (UV)/ozone-oxidation/
carbon reactor system and the Matrix Photocatalytic, Inc.
(Matrix), system. The waste stream was pretreated
because of its high oil and grease concentration of 8.000
milligrams per liter (mg/L). An oil coalescer was placed
upstream of the treatment systems and removed more
than 96% of the oil and grease and saturated aliphatic
compounds in the waste stream. Post-coalescer stream
contaminants consisted mainly of substituted derivatives
of benzene, such as naphthalene and its derivatives (200
to 500 parts per billion [ppb]), and saturated aliphatic
hydrocarbons and their substituted derivatives (each
ranging from 250 to 4,000 ppb). The stream also contained
a maximum concentration of 1,000 mg/L of bicarbonate
at its original pH of 7.
B.2 System Characteristics
System characteristics of the two technologies are listed
below.
Matrix System
• Titanium dioxide (TiO2) catalyst activated by UV
light at 254 nanometers (nm)
• Flow-through mode (single pass) flow rate varied
from 1 to 3 liters per minute (L/min) (50 L at a time)
Optimizing options - compressed air, compressed
oxygen, and hydrogen peroxide (H2O?) injection, all
individually (1 L/hour maximum injection rate)
Carbonate/bicarbonate reduction
UV/Ozone-Oxidation/Carbon Reactor
System
UV light at 254 nm with ozone gas
Semibatch mode (100 L) flow of 22 l_/min
• 25% of effluent passed through carbon filter before
recirculation to feed tank
Paniculate strainers upstream
Optimizing options - pH adjustment and ozone feed
(10 milligrams per kilogram maximum ozone feed
concentration)
B.3 Test Parameters
The systems were tested individually on similar waste
streams with fluctuating contaminant concentration levels.
For the UV/ozone-oxidation/carbon reactor system,
samples were collected from the inlet feed and from the
system effluent after 60 and 120 minutes of contact time.
For the Matrix system, samples were collected from the
inlet feed, after module 1 (after 20 seconds of contact
time), and module 2 (after 40 seconds of contact time),
and after module 3 (after 60 seconds of contact time).
Target contaminants analyzed for included oil and grease,
dissolved organic carbon, and U.S. Environmental
Protection Agency (EPA) Methods 624 and 625 priority
contaminants (primarily benzene and polyaromatic
compounds such as naphthalene and its derivatives).
B.4 Test Conclusions
The Matrix treatment system is capable of reducing the
concentrations of both EPA Methods 624 and 625 priority
compounds to below method detection limits. In addition,
the Matrix system can effectively reduce phenol ics
concentrations to well below the Canadian federal
discharge limit of 20 ppb; therefore, effluent from the
Matrix system could be discharged directly to the Chalk
River. Other aromatic compounds, including naphthalene
and substituted naphthalene, are also effectively removed.
Oxidation can be achieved in the absence of chemical
additives but is accelerated by all the oxidants evaluated,
including air, oxygen, and H2O2. All dissolved organic
56
-------
carbon is not converted to carbon dioxide in the reactor.
Some intermediate oxidation products are formed that
are thought to include aldehydes and organic acids.
Visible concentrations of oil and grease did not foul the
catalyst or reduce the throughput rate of the system. Oil
and grease concentrations of 150 mg/Lwere reduced to
below 3 mg/L after treatment in the first module, and the
effluent was clear after subsequent passage through two
additional modules.
The UV/ozone-oxidation/carbon reactor system was
effective in removing all EPA 624 and 625 priority
compounds present in the waste stream, but the rate of
oxidation was somewhat lower than observed for the
Matrix TiO2 catalytic reactor, partially because the UW
ozone-oxidation/carbon reactorsystem is batch operated,
while the Matrix system operates in continuous, ptug-
f lowf ashion. The rate of dissolved organic carbon removal
seemed optimal at an ozone concentration of about 5 mg/
L at a neutral pH, with a contact time of about 1 hour.
Straight-chain aliphatic compounds were oxidized less
rapidlybythe UV/ozone-oxidation/carbonreactor system
than by the Matrix catalytic oxidation reactor.
B.5 Recommendations
The matrix system proved to be reliable in reducing the
concentrations of many EPA 624 and 625 priority
contaminants. Oil and grease concentrations were
reduced to less than 3 mg/L, and phenolics concentrations
were reduced to below the detection limit.
The UV/ozone-oxidation/carbon reactorsystem, although
effective in removing most priority contaminants (including
phenolics) to concentrations below 20 ppb, seemed
sensitive to pH and was not as effective at removing oil
and grease as the Matrix system. The end-products of
oxidation in the UV/ozone-oxidation/carbon reactor
seemed to include a large assortment of acids and
aldehydes, although their distribution was not quantified.
The process was operated in a batch mode because of
the long contact time required for oxidation (1 to 2 hours),
while the Matrix system was operated in a continuous
fashion (60 seconds contact time),
A two-step process that uses the oil coalescer followed
by the Matrix TiC>2 catalytic reactor is recommended for
treating Chalk River Laboratories aqueous radioactive
wastes contaminated with bitumen; however, more
research is required to determine the toxicity of the Matrix
system effluent and the speciation of the organic
compounds.
According to Sen Gupta and others (1994), Matrix is
dedicated to product development. System enhancement
and research is continuous, which makes current
capability reporting for the system relatively difficult. The
current Matrix technology has enhanced reactor and
catalyst capabilities and would probably out-perform the
Matrix systemtested during this case study by a factor of
2; therefore, the current Matrix technology's capabilities
are not fully represented by this case study.
B.6 Estimated Costs
The cost of "pilot-scale size" treatment for this study
would be $2.27 per 1,000 gallons of waste treated. This
cost includes only an electricity cost of $0.08 per kilowatt-
hour and an oxygen cost of $0.21 per 1,000 gallons of
oxygen. Improvements made to the Matrix system since
the study date have significantly reduced the cost of
treatment per 1,000 gallons.
B.7 Reference
Sen Gupta, S., R. Peori, and S. Wickware. 1994. "De-
struction of Organic Contaminants in Industrial Waste-
water Using Oil Coalescence and Photochemical Oxi-
dation Technologies." EPRI 1994 Conference Pro-
ceedings. Norfolk, Virginia. July 25-27.
*U.S. GOVERHMEHP HOKEING OFFICE: 1998-651-418
57
------- |