United States
Environmental Protection
Agency
Control Technology Center
EPA-600/R-92-124
July 1992
AIR EMISSIONS FROM THE TREATMENT OF
SOILS CONTAMINATED WITH PETROLEUM
FUELS AND OTHER SUBSTANCES
control
technology center
-------
CONTROL TECHNOLOGY CENTER
Sponsored by:
Emission Standards Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Air and Energy Engineering Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Center for Environmental Research Information
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/R-92-124
July 1992
AIR EMISSIONS FROM THE
TREATMENT OF SOILS CONTAMINATED
WITH PETRpLEUM FUELS AND OTHER SUBSTANCES
FINAL REPORT
Prepared by:
BartEklund
Patrick Thompson
Adrienne Inglis
Whitney Dulaney
Radian Corporation
8501 Mo-Pac Boulevard
P.O. Box 201088
Austin, Texas 78720-1088
EPA Contract No. 68-DO-0125, Work Assignment 25
and EPA Contract No. 68-D1-0117, Work Assignment 31
EPA Project Officer:
Susan A. Thoraeloe
Air and Energy Engineering Research Laboratory (MD-63)
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
Prepared for
U.S. Environmental Protection Agency
Office of Air and Radiation
and
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
This document summarizes available information on air emissions from the
treatment of soils contaminated with fuels. Soils contaminated by leaks or spills of fuel
products, such as gasoline and jet fuel, are a nationwide concern. Air emissions during
remediation are a potential problem due to the volatile nature of many of the fuel
components and the remediation processes themselves, which may promote or result in
contaminant transfer to the vapor-phase. Limited information is also included on air
emissions from the treatment of soils contaminated with hazardous wastes.
The document will allow staff from State and local regulatory agencies, as
well as staff from EPA regional offices, to assess the different options for cleaning up
soil contaminated with fuels. Seven general remediation approaches are addressed in
this report. For each approach, information is presented about the remediation process,
the typical air emission species of concern and their release points, and the available air
emissions data. Control technologies for each remediation approach are identified and
their reported efficiencies are summarized. Cost data are given for each remediation
approach and for its associated control technologies. Emission factors and other
emission estimation procedures for each remediation approach are presented along with
a brief case study.
This report was prepared in fulfillment of Contract No. 68-DO-0125, Work
Assignment 25 by Radian Corporation, under the sponsorship of the U.S. Environmental
Protection Agency. The report was revised under Contract No. 68-D1-0117, Work
Assignment 31.
-------
TABLE OF CONTENTS
ABSTRACT ii
METRIC CONVERSIONS viii
EXECUTIVE SUMMARY ix
1.0 INTRODUCTION 1-1
1.1 Background 1-1
12 Objectives 1-2
13 Approach 1-2
1.4 Frequency of Use of Various Remediation Options 1-3
1J5 Limitations of the Document 1-4
1.6 Organization of the Report 1-8
2.0 SUMMARY OF RESULTS 2-1
3.0 EXCAVATION AND REMOVAL 3-1
3.1 Process Description 3-1
32 Identification of Air Emission Points 3-3
33 Typical Air Emission Species of Concern 3-5
3.4 Summary of Air Emissions Data 3-5
3.5 Identification of Applicable Control Technologies 3-7
3.6 Costs for Remediation 3-12
3.7 Costs for Emissions Controls 3-14
3.8 Equations and Models for Estimating VOC Emissions 3-14
3.9 Case Study 3-17
3.10 References 3-17
4.0 THERMAL DESORPTION 4-1
4.1 Process Description 4-1
42 Identification of Air Emission Points 4-9
43 Typical Air Emission Species of Concern 4-10
4.4 Summary of Air Emissions Data 4-11
4.5 Identification of Applicable Control Technologies 4-23
4.6 Capital and Operating Costs for Remediation 4-29
4.7 Capital and Operating Costs for Emission Controls 4-30
4.8 Equations/Models for Estimating Emissions 4-32
4.9 Case Studies of Remediation and Air Emissions 4-34
4.10 References 4-41
111
-------
TABLE OF CONTENTS (Continued)
5.0 SOIL VAPOR EXTRACTION 5-1
5.1 Process Description 5-1
52 Identification of Air Emission Points 5-7
53 Typical Air Emission Species of Concern 5-7
5.4 Summary of Air Emissions Data 5-7
53 Identification of Applicable Control Technologies 5-9
5.6 Costs for Remediation 5-13
5.7 Costs for Emission Controls 5-14
5.8 Equations and Models for Estimating VOC Emissions 5-16
5.9 Case Study 5-19
5.10 References 5-22
6.0 IN-SITU BIODEGRADATION 6-1
6.1 Process Description 6-1
62 Identification of Air Emission Points 6-4
63 Typical Air Emission Species of Concern 6-5
6.4 Summary of Air Emissions Data 6-5
6.5 Identification of Applicable Control Technologies 6-7
6.6 Costs for Remediation 6-8
6.7 Costs for Emissions Controls 6-10
6.8 Equations and Models for Estimating VOC Emissions 6-10
6.9 Case Study 6-11
6.10 References 6-11
7.0 EX-SITU BIODEGRADATION 7-1
7.1 Process Description 7-1
12 Identification of Air Emission Points 7-8
73 Typical Air Emission Species of Concern 7-8
7.4 Summary of Air Emissions Data 7-9
15 Air Emission Controls 7-9
7.6 Costs for Remediation 7-13
7.7 Costs for Emissions Controls 7-13
7.8 Summary of Existing Air Emissions Data and Models 7-14
7.9 References 7-16
iv
-------
TABLE OF CONTENTS (Continued)
Pag!
8.0 INCINERATION 8-1
8.1 Process Description 8-1
82 Identification of Air Emission Points 8-4
83 Typical Air Emission Species of Concern 8-4
8.4 Summary of Air Emissions Data 8-7
8.5 Identification of Applicable Control Technologies 8-7
8.6 Costs for Remediation 8-10
8.7 Costs for Emissions Controls 8-12
8.8 Equations and Models for Estimating VOC Emissions 8-12
8.9 Case Study: On-Site Incineration 8-14
8.10 References 8-14
9.0 SOIL WASHING/SOLVENT EXTRACTION 9-1
9.1 Process Description 9-1
9.2 Identification of Air Emission Points 9-12
93 Typical Air Emission Species of Concern 9-12
9.9 Summary of Air Emissions Data 9-12
9.5 Identification of Applicable Control Technologies 9-13
9.6 Capital and Operating Costs for Remediation 9-13
9.7 Capital and Operating Costs for Emission Controls 9-13
9.8 Equations/Models for Estimating Emissions 9-14
9.9 Case Studies of Remediation and Air Emissions 9-14
9.10 References 9-14
APPENDDC A: PROPERTIES AND COMPOSITION OF VARIOUS FUEL
TYPES A-l
APPENDIX B: STATE CLEANUP REQUIREMENTS B-l
APPENDIX C: EXAMPLE CALCULATIONS C-l
APPENDIX D: ENGINEERING BULLETIN FOR THERMAL DESORPTION
TREATMENT D-l
APPENDIX E: ARTICLE ON SOIL VAPOR EXTRACTION E-l
APPENDDC F: ARTICLE ON INCINERATION F-l
-------
UST OF FIGURES
1-1 Relative Frequency of Use of Remediation Technologies at UST Sites 1-5
1-2 Relative Frequency of Use at UST Sites by Specific Technology .... 1-6
1-3 Relative Frequency of Use of Groundwater Remediation Technologies
at UST Sites 1-7
3-1 Summary of Air Emission Points for Excavation and Removal 3-4
4-1 Soil Treatment Temperature Guide 4-4
4-2 Generalized Process Diagram for Thermal Screw-Based
Thermal Desorption 4-7
5-1 Simplified Guide to Applicability of Soil Vapor Extraction 5-3
5-2 Generalized Process Flow Diagram for Soil Vapor Extraction 5-6
5-3 Process Flow Diagram for Terra Vac In-Situ Vacuum Extraction
System 5-20
6-1 Generalized Process Flow Diagram for In-Situ Biodegradation 6-3
6-2 Flow Diagram for Off-Gas Treatment System for In-Situ
Biodegradation 6-12
7-1 Slurry Biodegradation Process Flow Diagram ;. 7-3
8-1 Process Flow Diagram for Commercial Rotary Kiln Incinerator 8-3
9-1 Schematic Diagram of Aqueous Soil Washing Process 9-5
9-2 Schematic Diagram of Solvent Extraction Process 9-8
9-3 Generalized Soil Flushing Process How Diagram 9-11
VI
-------
LIST OF TABLES
2-1 Summary of Cost Information for the Treatment of Contaminated Soil 2-2
3-1 Input Variables for Emission Equations 3-16
4-1 Comparison of Features of Thermal Desorption and Offgas
Treatment Systems 4-2
4-2 Summary Data for Asphalt Kilns 4-12
4-3 Summary Data for Mobile Thermal Desorption Units ,4-16
4-4 Estimated Emissions of Selected Compounds for the Cleanup of
PCB-Contaminated Soil using the IT Process 4-24
4-5 Remedial Costs for Various Thermal Desorption Units 4-31
4-6 Cost Information for Fabric Filters 4-31
4-7 Cost Information for Thermal Oxidizers 4-33
5-1 Summary of Emissions Data for SVE System 5-8
5-2 Summary of Capital Costs to Control VOC Emissions from
SVE Systems 5-15
5-3 Estimated Emissions for Terra-Vac's In-Situ Vacuum
Extraction System 5-20
6-1 VOC Emissions from a Refinery Landfarm: Average Measured
Emission Rates by Plot by Half-Day 6-6
6-2 Estimated Costs for Remediation Using Hydrogen Peroxide
To Enhance Biodegradation 6-9
7-1 Applicability of Slurry Biodegradation for Treatment of
Contaminants in Soil, Sediments, and Sludges 7-2
7-2 Desired Inlet Feed Characteristics for Slurry Biodegradation Processes 7-5
7-3 Performance Results for Slurry Biodegradation Process Treating Wood
Preserving Wastes 7-7
7-4 Estimated Volatile Losses from Aerated Wastewater Treatment 7-10
Vll
-------
8-1
8-2
8-3.
8-4
9-1
9-2
9-3
LIST OF TABLES (Continued)
PICs Found in Stack Effluents of Full-Scale Incinerators 8-8
Characteristics of Off-Gas from On-Site Incineration Systems 8-9
Estimated Range of Costs for Off-Site Incineration 8-11
Estimated Range of Costs for On-Site Incineration 8-11
Removal Efficiencies for Remediation of PCB Contamination 9-15
Removal Efficiencies for Remediation of API Separator Sludge 9-15
Summary of Performance Data on Soil Washing 9-16
Metric Conversions
Readers more familiar with the metric system may use the following factors to convert to
that system.
-\ " ' - Non-metric - ^
MMBtu/hr
°F
ft
acfm
dscfm
gal
hp
in
Ib
mil
mile
ton
cuyd
XV -" Multiplied by i
405435
0355556 (T- 32)
03048
0.028317
0.028317
3.78541
746
234
0.453592
0.0254
1609344
0.907185
0.76455
Yields Metric
MMJ/hr
°c
m
acmm
dscmm
L
J/sec
cm
kg
mm
m
metric ton (1,000 kg)
m3
Vlll
-------
EXECUTIVE SUMMARY
This document summarizes the available information on air emissions from
the treatment of soils contaminated with fuels. It is intended to guide State and local air
pollution control agencies in the evaluation of the air emission potential of treatment of
contaminated soil and the cost-effectiveness of applicable emission control technologies.
The scope was limited to the emissions of volatile organic compounds (VOCs); however,
due to the limited data that were available, information was also included for the
emissions of other organic compounds. This additional information is primarily from the
treatment of soils contaminated with hazardous wastes.
Seven general approaches for the disposal or treatment of soils
contaminated with gasoline, oil, or diesel fuel were identified:
• Excavation and removal;
• Thermal desorption;
• Soil vapor extraction (SVE);
• In-Situ biotreatment (e.g. landtreatment);
• Ex-Situ (batch) biotreatment;
• On-site incineration; and
* Soil washing/solvent extraction.
Each general approach may include several specific options. For example, thermal
desorption may be performed in portable units designed specifically for soil treatment or
in rotary drum aggregate dryers that are part of asphalt plants or other industrial
facilities.
Literature pertaining to the emissions of volatile organic compounds
(VOCs) for each remediation approach was identified and reviewed. The summarized
information was organized into the same ten part format for each approach:
• Process description;
• Identification of air emission points;
• Identification of typical air emission species of concern;
• Summary of published air emissions data;
• Identification of applicable control technologies;
• Cost data for the overall remediation approach;
• Cost data for the emission controls;
• Equations and models for estimating VOC emissions;
• Case study of the use of the remediation approach; and
• References.
-------
For most of the technologies examined, VOC emission estimates or measured data were
found. Emission factors, in grams per hour, were identified or developed that are based
on available data as well as assumed "typical11 operating conditions for the remediation of
relatively large sites. Cost data, in dollars per ton or cubic yard of soil treated, were
obtained from a variety of sources, but data prior to 1986 were generally avoided
because of the changes in remediation technology, standard operating practices, and
regulations in recent years. All cost data were converted to 1991 dollars using a 5%
annual escalation factor.
Certain limitations of the data presented in this document should be
considered before extrapolations are made to a specific site under consideration. Any
generalized guidance has inherent limitations due to the variety of site-specific and
process-specific factors that may be encountered. Many of the cleanup processes are
emerging technologies and have short operating histories. For these technologies, data
on air emissions, treatment effectiveness, and costs are very limited. Furthermore, each
site has its own unique obstacles to cleanup that may force modifications to the cleanup
hardware or operating conditions. The development of typical air emission rates and
emission factors applicable to the maximum number of site conditions and site locations
required assumptions regarding the rate and scope of the clean-up effort, the type of fuel
being treated, the number and nature of emission release points, and so on. The more a
specific site differs from the assumed conditions, the less likely the generalized air
emissions data will be applicable.
Data gaps were identified and suggestions for future research topics were
given. In general, only limited information was found for air emissions from the
treatment of contaminated soil. The need for more data is greatest for emerging
technologies and those that are area sources of VOC emissions. The general needs are
for more emissions data, more control cost and effectiveness data, and for accurate
emission models. The most important research needs that were identified during this
study were:
• VOC emission rate data for excavation;
« Theoretical models to estimate VOC emissions from excavation;
• Cost and effectiveness data of area source emission controls; and
• Fate studies for VOCs in biotreatment systems.
-------
1.0 INTRODUCTION
1.1 Background
The Control Technology Center (CTC) at the U.S. Environmental
Protection Agency (EPA) is responsible for supporting State and local air pollution
control agencies in the implementation of their programs. As part of this support, the
CTC provides control technology assessments to evaluate the control technologies
available for reducing emissions from a particular type of source. The CTC typically
provides expertise and information not otherwise available to the State or local agency.
The CTC has received requests from State and local regulatory agencies, as
well as from EPA regional offices, regarding how to assess the different options for
cleaning up contaminated soil. The requests have addressed a number of specific
remediation techniques, such as the clean-up of soils using rotary drum dryers.
Information is needed for estimating the potential air emissions from various types of
processes and for determining what potential controls may be needed. While some
guidance is currently available, it is dispersed among multiple documents.
The purpose of this project is to develop a procedure and guidance
document for use by State and local regulatory agencies for evaluating the air emission
potential and applicable control technologies for the treatment of contaminated soil.
Radian Corporation, under contract to the U.S. EPA (Contract Number 68-DO-0125,
Work Assignment 25 and Contract Number 68-D1-0117, Work Assignment 31) assisted
the CTC in this effort. Existing guidance for how to assess both potential air emissions
and available control technologies was identified. Examples of different clean-up
operations were identified for soils contaminated with gasoline, diesel fuel, or fuel oil.
Information on the kind of control technologies that are available and their expected
range of capital and operating costs was also obtained.
1-1
-------
12 Objectives
The specific objectives of this program were to:
• Identify options for the disposal/treatment of soils contaminated
with gasoline, oil, or diesel fuel;
• Review the available literature pertaining to the emissions of
volatile organic compounds (VOCs) for each clean-up option;
• Develop step-by-step instructions on how to estimate the VOC
emissions from the various clean-up options;
• Identify applicable control technologies and compile ranges of
capital and operating costs for each technology;
• Assess the uncertainty associated with the emission estimates and
the need for any laboratory or field studies to collect data to address
data gaps; and
• Summarize the information in a guidance document
The clean-up options addressed in this document are:
Excavation and removal;
Thermal desorption (includes asphalt plants);
Soil vapor extraction (SVE);
In-Situ biotreatment (landtreatment);
Ex-Situ (batch) biotreatment;
On-site incineration; and
Soil washing/solvent extraction.
L3 Approach
The general approach was to perform a literature search and then to
evaluate the collected documents. Over two hundred publications were reviewed and
evaluated. Additional information was obtained from researchers active in this area.
1-2
-------
For each of the identified remedial options, the literature was reviewed to develop a
process flow diagram and identify emission points, as well as to analyze available air
emissions data. For most of the technologies examined, VOC emission estimates or
measured data were found. Where VOC data were limited, data for other types of
organic compounds were also compiled. Emission factors were identified or developed
that are based on available data as well as assumed "typical" operating conditions for the
remediation of relatively large sites.
Much of the information in this document is based on and taken directly
from the document titled, "Emission Factors for Superfund Remediation Technologies"
(Thompson, Inglis, and Eklund, EPA-450/ 1-01-002, May 1991). Cost data were obtained
from a variety of sources, but data from prior to 1986 were generally avoided due to the
changes in remediation technology, standard operating practices, and regulations in
recent years. All cost data were converted to 1991 dollars using a 5% annual escalation
factor.
Frequency of Use of Various Remediation Otions
The remediation options addressed in this document are all potentially
suitable for use as part of the remediation process for soils contaminated with fuels. The
various options, however, are not necessarily all equally cost-effective nor is their use
equally widespread. EPA's Office of Underground Storage Tanks (OUST) has surveyed
state agencies responsible for the cleanup of leaking underground storage tank (UST)
sites to ascertain the frequency of use of various remediation options. The information
is primarily derived from the remediation of UST sites contaminated with gasoline. This
information is summarized in several figures.
1-3
-------
Figure 1-1 shows the relative frequency of use of the major classes of
remediation options. Landfilling (excavation and removal) is used somewhat more than
half the time, with in-situ methods, thermal treatment, or land treatment also frequently
used. Figure 1-2 provides more detail as to the type of in-situ, land treatment, and
thermal treatment methods that are employed. For sites employing in-situ remediation,
the exact technology used is undefined the majority of the time. It is assumed that soil
vapor extraction is probably used in most of these cases. For applications of thermal
treatment, thermal desorption is almost always employed and incineration is only very
rarely used. Figure 1-3 shows the frequency of use of groundwater remediation
technologies at UST sites. While these technologies are not part of the scope of this
document, they are frequently used in conjunction with soil remediation technologies and
the treatment of groundwater (e.g., by air stripping) may contribute to the overall levels
of air emissions from the site. It is important to address all media when evaluating
possible remediation scenarios.
1*5 Limitations of the Document
The initial review of the existing information showed that there was only
limited published data. There was not adequate VOC air emissions data from
remediation to assess the importance of fuel type, spill volume, the age of the spill, and
the soil type as they relate to the combination of remediation and control technologies
that are applied. Therefore, there was insufficient data to develop step-by-step
estimation procedures and to assess the uncertainty associated with such estimates.
Instead, the limited existing information was compiled to provide users with a summary
of air emissions data. Information was included for VOC air emissions from the
treatment of both soils contaminated with petroleum fuels and the treatment of
hazardous waste to fill as many data gaps as possible.
1-4
-------
In-Situ 19%
Landfilling 54%
Other 2%
Thermal Treatment 13%
Land Treatment 11%
Figure 1-1. Relative Frequency of Use of
Remediation Technologies at UST Sites
Source: EPA-OUST
1-5
-------
In-Sitll
Undefined/Other 89%
Soil Vapor Extraction 9%
Bioremedlation 2%
Land Treatment Technologies
Aeration 50%
Landfarming 36%
Land Application 13%
Thermal
Asphalt Options 61%
Thermal Desorption 39%
Incineration 0.1%
Figure 1-2. Relative Frequency of Use at UST Sites
by Specific Technology
Source: EPA-OUST
1-6
-------
Pump & Treat with Undefined
Treatment Method 37%
Pump & Treat with Carbon
Absorb tion 26%
Bioremediation <1%
Soil Venting 2%
Other 2%
Pump & Treat
with Air Stripping 33%
Figure 1-3. Relative Frequency of Use of Ground water
Remediation Technologies at UST Sites
Source: EPA-OUST
1-7
-------
Generalized guidance for the remediation of soils contaminated with fuels
has inherent limitations. Many of the cleanup processes are emerging technologies and
have short operating histories. For these technologies, data on air emissions, treatment
effectiveness, and costs are very limited. Furthermore, each site was its own unique
obstacles to cleanup that may force modifications to the cleanup hardware or operating
conditions.
The development of typical air emission rates and emission factors
applicable to the maximum number of site conditions and site locations required
assumptions regarding the rate and scope of the clean-up effort, the type of fuel being
treated, the number and nature of emission release points, and so on. Assumptions were
based on what is "typical" and "reasonable" for the remediation of relatively large sites.
Obviously, the diverse nature of sites with fuel contamination will result in the
information presented here being more applicable to some sites than others. A limited
data set must be used to generalize about a wide-spectrum of process conditions.
The VOC air emissions data compiled in this document can be used for
planning purposes and for comparison to permit applications, but the user must take into
account the inherent limitations of the data and the limitations in extrapolating the data
to fit the specific remediation scenario under consideration.
1.6 Organization of the Report
The remainder of this report is divided into nine sections. Section 2
presents a summary of the results obtained in this study. Sections 3 to 9 are each
devoted to one of the seven cleanup technologies examined (excavation and removal,
thermal desorption, soil vapor extraction, in-situ biotreatment, ex-situ biotreatment,
incineration, and soil washing. Each of these sections follow the same ten-part format.
1-8
-------
A number of appendices to the report are also given. Appendix A
summarizes the properties and composition of various fuel types. Appendix B lists state-
by-state clean-up requirements for fuel spills. Example calculations are given in
Appendix C. Copies of selected references are given in Appendices D through F.
1-9
-------
2.0 SUMMARY OF RESULTS
Typical treatment cost data are given in Table 2-1 for treatment operations
with and without emission controls. The emission factors are based on "reasonable"
operating conditions for the remediation of sites contaminated with petroleum fuels.
These estimates may not be applicable to some clean-up programs. A range of costs are
given in most cases and these estimates are considered to be the best available
information in the literature. Since the cost estimates are not all based on the same
remediation scenario, the data may not be directly comparable because the underlying
assumptions of the volume of contaminated soil, the types and mass of contaminants that
are present, the rate of treatment, the type of controls employed, etc. may vary.
There was insufficient data to provide estimated VOC emission factors
based on starting soil contamination levels for the technologies discussed in this report.
The discussions of each technology presented in subsequent sections summarize existing
air emissions data. Concentration data (i.e. mass/volume of air) and emission rate data
(mass/time) are given from various test programs, but these data are not necessarily
directly comparable due to differences in the underlying assumptions.
2-1
-------
Table 2-1.
Summary of Cost Information for the Treatment of Contaminated Soil
Technology
Excavation and Removal
Thermal Desorption
Soil Vapor Extraction
In-Situ Biodegradation
Ex-Situ Biodegradation
On-Site Incineration
Soil Washing
Solvent Extraction
Soil Hushing
Estimated Treatment Cost ($/ton)
Controlled
ND
35-125
52/ton of VOC
NA
ND
390-1020"
NA
NA
NA
Uncontrolled
75 - 500
NA
26/ton of VOC
100
70 - 130
NA
53 - 215
105 - 525
ND
"Assumes a small site and assumes incineration of hazardous waste, as opposed to
incineration of soil contaminated with petroleum fuels.
ND = No estimate
NA = Not applicable
2-2
-------
3.0 EXCAVATION AND REMOVAL
3.1 Process Description
Excavation and removal of soils contaminated with fuels is a common
practice. Excavation and removal may be the selected remediation approach or it may
be a necessary step in a remediation approach involving treatment. If removal is the
preferred approach, the excavated soil is typically transported off-site for subsequent
disposal at a landfill. If the soil contains large amounts of fuel or highly toxic
contaminants, the soil may need to be treated off-site prior to final disposal. Excavation
activities are also typically part of on-site treatment processes such as incineration,
thermal desorption, batch biotreatment, landtreatment, and certain chemical and physical
treatment methods. The soil is excavated and transported to the process unit and the
treated soil is typically put back into place on the site. The information presented in this
section for excavation and removal is generally applicable to all soils handling operations
including excavation, dumping, grading, short-term storage, and sizing and feeding soil
into treatment processes.
The magnitude of VOC emissions depends on a number of factors,
including the type of compounds present in the waste, the concentration and distribution
of the compounds, and the porosity and moisture content of the soil. The key
operational parameters are the duration and vigorousness of the handling, and the size
of equipment used. The longer or more energetic the moving and handling, the greater
likelihood that organic compounds will be volatilized. The equipment size influences
volatilization by affecting the mean distance a volatilized molecule has to travel to reach
the air/solid interface at the surface of the soil. In general, the larger the volumes of
material being handled per unit operation, the lower the percentage of VOCs that are
stripped from the soil. Control technologies for large area sources such as excavation
are relatively difficult to apply and are often much less effective than controls for point
sources.
3-1
-------
The relative advantages of excavation and removal over other remediation
approaches are that:
1) Earth-moving equipment and trained operators are widely available;
2) Large volumes of soil can be quickly moved in a cost-effective
manner; and
3) Residual contamination remaining at the site is minimal.
The major disadvantages of excavation and removal versus other remediation approaches
are that:
1) The magnitude of air emissions may be high;
2) Air emissions from excavation are difficult to control; and
3) The contaminants are only removed, they are not destroyed.
The success of excavation and removal for a given application depends on
numerous factors with the three key criteria being: 1) the nature of the contamination; 2)
the operating practices followed; and 3) the proximity of sensitive receptors. Each of
these criteria is described below.
As previously discussed, spills or leaks, of fuels typically involve liquids
containing dozens of different constituents. Excavation and removal is generally a viable
option, except for those cases where air emissions potentially pose an unacceptable risk.
For example, soil containing percent levels of benzene or other volatile carcinogens
would almost certainly pose a large risk to on-site workers and the surrounding populace
if it were to be excavated. In-situ remediation methods, such as soil vapor extraction,
would be preferable for such a site, either in lieu of excavation or prior to excavation to
reduce the emissions potential.
The magnitude of emissions from soils handling operations will vary with
the operating conditions. Add-on control technologies are available for minimizing VOC
3-2
-------
emissions, but they are relatively ineffective and costly to implement. VOC emission
control can also be achieved by controlling the operating conditions within preset
parameters. The rate of excavation and dumping, the drop height, the amount of
exposed surface area, the length of time that the soil is exposed, the shape of the storage
piles, and the dryness of the surface soil layers will all influence the levels of VOC
emissions. Large reductions in emissions can be achieved by identifying and operating
within acceptable ranges of operating conditions.
Since some release of volatile contaminants is inevitable during excavation
and removal unless extreme measures are taken (e.g. enclose the remediation within a
dome), the proximity of downwind receptors (i.e. people) will influence whether or not
excavation is an acceptable option. Excavation of contaminated areas that abut
residential areas, schoolyards, etc. may require more extensive controls, relocation of the
affected population, or remediation only during certain periods (e.g. summertime for
school sites).
32 Identification of Air Emission Points
VOC emissions from handling operations result from the exchange of
contaminant-laden soil-pore gas with the atmosphere when soil is disturbed and from
diffusion of contaminants through the soil. There are several potential emission points
involved in excavation as shown in Figure 3-1; all are considered to be fugitive area
sources. For excavation, the main emission points of concern are emissions from:
exposed waste in the excavation pit;
material as it is dumped from the excavation bucket; and
waste/soil in short-term storage piles.
In addition, emissions of VOC, paniculate matter, nitrogen oxides, etc. will also occur
from the earth-moving equipment.
3-3
-------
/. Emissions from soil cap removal
® Emissions from soH cap excavation
Q> Emissions from soil cap in bucket
x.xjrii
///. Emissions from transport of soil cap ®
//. Emissions from truck filling with soil cap
IV. Excavation of contaminated soil zone
9 Emissions from exposed contaminated soH zone
-------
33 Typical Air Emission Species of Concern
The emissions of concern from soils handling operations such as excavation
can be any contaminant that is present in the soil. Relatively large amounts of volatile
organic compounds (VOCs) may be released from soil during handling, so VOCs are
typically the emissions of most concern. Emissions of paniculate matter and associated
metals and semi-volatile compounds may be of concern at some sites.
3-4 Summary Of Air Emissions Data
Very limited references with VOC emissions or emission rate data for
excavation were identified in the literature search performed for this project. The
process of measuring emission rates from dynamic processes, such as excavation, is
difficult and costly, and so has rarely been attempted.
Volume m of the Series of Air-Superfund Guidance Manuals (Eklund, et
al, 1989) for estimating clean-up emissions indicates that soils handling operations such
as excavation increase VOC emission rates from contaminated soil over baseline rates.
The increase in emissions is typically a factor of ten or more, and the increased emission
rate decays exponentially back to near the baseline rate over short time periods (e.g. 4
days). A database of baseline emission rate measurement data (Eklund, et al, 1991) is
available.
Emission rate measurements have been made at two sites (Eklund, et al,
1990). Excavation decreased the soil moisture content by 35% to 56% and tended to
somewhat decrease (e.g. -13%) the dry bulk density of the soil. Measured emission rates
were as high as 4 g/min for specific compounds. These emissions were from both the
excavation and dumping processes. Excavation resulted in most of the mass of various
volatile compounds being stripped from the soil, based on a comparison of measured
total emissions versus the mass of these same contaminants in the soil (calculated from
3-5
-------
soil concentration data). This was true for both sites, despite differences in soil
concentrations and soil type. For sites with medium to high ppm-levels of contamination
and/or wet soil, a lower percentage of contaminants would be stripped from the soil
during excavation.
Theoretical models for estimating emissions (Eklund 1992a) indicate that
about 70% of the mass of a volatile compound such as xylene is emitted during
excavation of soil with a starting contaminant concentration of 1 ppm under the assumed
typical conditions. Another theoretical study (Saunders, 1990) of soils handling emissions
estimated that relative to excavation, other soils handling operations would have the
following emissions: 1) Truck Filling - 0.58; 2) Transport - 5.23; and 3) Exposed soil -
1.47 (emissions/excavation emissions).
Field experience indicates that actual emissions may be substantially lower
than the stripping percentages discussed above. For dry, porous soils containing low ppb
levels of contaminants it can be assumed that most or all of the more volatile VOCs will
be lost to the atmosphere during soils handling. For sites with moist soils and ppm
levels of contaminants, however, a reasonable assumption may be that only 5 to 10% of
the VOCs are emitted to the atmosphere during each handling step. More measurement
data are needed to support these assumptions.
No valid emission factors were found. A theoretical study of the emissions
from the clean-up of leaking underground storage tank sites (U.S. EPA-OUST, 1989)
estimated that emissions from storage piles of contaminated soil with a surface area of
186 m2 (2,000 ft2) were:
Average Benzene Emission Rate (lb/hr)*
Total Benzene Emissions (Ib)
Average VOC Emission Rate (Ib/hr)
Total VOC Emissions (Ib)
1
336
50
16,800
* Readers more familiar with the metric system may use the factors listed at the end of
the front matter to convert to that system.
3-6
-------
The total emissions are based on a two-week time period. Emissions from the actual
excavation process, as opposed to soils storage, were not estimated.
3.5 Identification of Applicable Control Technologies
A number of methods are available for controlling VOC and paniculate
matter emissions from soils. In general, any method designed primarily for paniculate
control will also reduce VOC emissions and vice versa. Compared to point source
controls, VOC emission controls for excavation and other area sources are difficult to
implement and only moderately effective. Controls such as water sprays or foams will
alter the percent moisture, bulk density, and average heating value of the soil and may
impact treatment and disposal options.
VOC emission controls for soil area sources are described below including:
• Covers and physical barriers;
• Temporary and long-term foam covers;
• Water sprays;
• Operational controls;
• Complete enclosures; and
• Wind Barriers.
Additional information is given in Eklund, et al., 1992b.
3.5.1 Covers and Physical Barriers
The most commonly used VOC control approach for area sources is the
use of covers to provide a physical barrier to vapor transport. The simplest barrier is the
use of relatively clean soil as a cover for contaminated soil. The soil layer increases the
necessary transport distance for vapor diffusion and thus greatly reduces, at least
temporarily, the emission rate. Soil covers are widely used at sanitary landfills to control
the emissions of odorous compounds and to control wind-borne pollution. The
effectiveness of soil covers will depend on the depth of the cover and the percent of
3-7
-------
contaminated soil that can be covered. Measured emission rates may be substantially
reduced (e.g., >95%) by the addition of compacted soil (Suder and Schmidt, 1992);
however, lateral migration of VOCs may still occur. Soil covers will be less effective
over long time periods and their use will tend to increase the total volume and mass of
material that must be treated.
Synthetic covers are typically used to control VOC emissions from
excavated soil in short-term storage piles. Synthetic covers are also widely used to ,
control VOC emissions during transport by rail or truck. The cover may be thin (4-6
mil) plastic sheeting or relatively thick (30-40 mil) plastic sheeting or geotextile material.
The resistance of various polymers to chemicals, weather, gas permeability, and tears is
documented (Landeeth, et al., 1983). The barrier material is available in large rolls and
can be quickly applied to even large soil piles. The barrier must be secured against
wind. Application of the barrier generally forces on-site workers to come in contact with
the waste. The barrier can be left in place indefinitely, though physical and
photodegradation of the polymer will tend to limit the effective lifetime of thin barriers
to a few weeks. The effectiveness of the cover will depend on its permeability to the
vapors that are present and the percentage of the soil pile that is adequately covered.
Laboratory measurements of a 20 mil PVC membrane showed relatively poor
performance for limiting vapor diffusion (Springer, et al, 1986). The PVC membrane
proved to be only as effective as a covering of a few inches of porous soil.
Numerous mulch materials, such as sawdust, wood chips, straw, and wood
fibers can also be used as a cover for soil undergoing long-term storage (U.S. EPA,
1991). The mulch acts primarily to control diffusion by insulating the soil surface and
thereby lowering the soil temperature. The mulch material will also limit diffusion
somewhat if it is used as a cover, but if mixed in with the contaminated soil the mulch
will generally increase the porosity of the soil and thereby increase the emission rate.
The mulch will also increase the volume and mass of contaminated material to be
treated or disposed.
3-8
-------
3.5.2 Temporary and Long-Term Foam Covers
Modified fire-fighting foams are commonly used to control VOC emissions
during the remediation of hazardous waste sites containing volatile toxic compounds. At
least six types of foam products are available (Evans and Carroll, 1986) from vendors
such as Rusmar and 3M. The different foams vary in their compatibility and
effectiveness for various classes of contaminants. Specialized equipment is available for
applying foams over large areas. The foam is applied to a depth of 6-18 inches and
coverage rates of 100 m2/min are possible. The liquid foam concentrate is applied via
an air-aspirating nozzle or chute. The degree of expansion (how many gallons of foam
produced from a gallon of liquid concentrate) can be high (250:1), low (20:1), or
medium.
Two general types of foams are used: temporary and long-term. The
temporary foams provide coverage for up to an hour, at which time 25% or more of the
liquid incorporated in the foam will have been released. Long-term foams contain a
stabilizing additive to extend the useful life of the foam to days or even months. The
effectiveness of foams is quite high for the areas that are covered. Short-term emission
reductions of 75% to 95% (for total paraffins and total arornatics, respectively) have
been measured in the field over 20 minute time periods (Aim, et al, 1987). Emission
reductions for total VOCs of 99% to 100% using stabilized foam have been measured in
the field over 24-hour time periods (Aim, et al, 1987).
The two primary advantages of foams are that they can be highly effective
and they can be applied directly to the backhoe bucket and the exposed contaminated
soil. There are several disadvantages to foams worth consideration. The thick layers of
foam required for emissions control can be applied more effectively to horizontal
surfaces than to vertical surfaces such as the sides of the excavation pit. Incomplete
coverage of the emitting surfaces will markedly decrease the effectiveness of the controls.
The foam concentrates are usually over 90% water and the addition of this water
3-9
-------
increases the weight of the soil, makes it more difficult to handle, and makes it less
amenable to thermal treatment. Specialized foam application equipment and a large
supply of foam concentrate are needed. The foam is difficult to apply on windy days.
Frequent application or re-application of the foam may be necessary.
3.53 Water Sprays
Water sprays are a commonly used control method for paniculate matter
(PM) emissions. The addition of dust control chemicals such as polymers or acrylics to
the water increases the effectiveness of the spraying. The water added to the soil will
decrease the air-filled porosity of the soil and will also tend to cool the surface soil
temperature. The reduction in vapor transport will diminish VOC emissions, though the
effectiveness of water sprays for VOC control is not documented. Water sprays are
certainly much less effective than water-based foams, and they have essentially the same
limitations as those listed above for foams.
3.5.4 Operational Controls
Operational controls can be effective in minimizing VOC emissions. These
controls may involve controlling the rate of excavation, the amount of contaminated soil
area that is exposed, and the duration that soil piles are left uncovered. The timing of
excavation can also be important. Scheduling excavation during times of the day or
seasons of the year when wind speeds and temperatures are low can reduce emissions.
Stagnant wind conditions, however, may lead to unacceptable ambient air concentrations
at the work site. The work can also be scheduled to avoid seasons with dry soil
conditions to further minimize emissions.
3-10
-------
3.5.5 Complete Enclosures
If warranted, complete enclosure of the excavation site can be
accomplished to minimize VOC emissions. The enclosure acts to collect any emissions,
which can then be vented to some type of control device suitable for point sources (see
Section 5.5). The enclosure may be either air supported or self supported. Self-
supported domes are more practical if trucks or other heavy equipment must regularly
enter and leave the structure. If properly designed and operated, the enclosure may.
reduce VOC emissions to negligible levels.
There are severe limitations that limit the use of complete enclosures to
the few sites where other control options are not acceptable. The capital cost of the
structure is relatively high. Operating costs can also be very high if large volumes of air
must be treated and exhausted to keep the concentrations of contaminants in the
atmosphere within the dome at levels that are safe for the workers. The use of a
complete enclosure will generally require the workers inside the structure to work under
Level B or C safety requirements. Due to the greenhouse effect, the air temperatures
within the structure may be high. The added safety requirements along with the time
delays in transferring trucks in and out of the structure will extend the time to complete
the excavation and thereby increase the cost.
3.5.6 Wind Barriers
For small work areas, the use of wind barriers can reduce VOC emissions
by lowering the effective wind speed at the soil surface. Commercial, porous wind fence
material that is typically used for dust control has been found to be more effective than
solid fence material (Springer, et al, 1986). For larger working areas, fencing is less
practical.
3-11
-------
VOC (and PM) emissions from storage piles can be minimized by
controlling the placement and shape of the piles. When feasible, the piles can be placed
in areas shielded from the prevailing winds at the site. The amount of surface area can
be minimized for the given volume of soil by shaping the pile. The orientation of the
pile will affect the wind velocity across the pile with the lowest windspeed occurring
when the length of the pile is perpendicular to the prevailing wind direction.
3.6
Costs For Remediation
The total costs for the treatment of contaminated soil by excavation and
removal will be the sum of the costs for excavation, transport, and treatment or disposal.
The total costs will vary widely and are primarily dependent on the disposal or treatment
costs. Total costs per ton may range from $75 to $500 or more, for excavation and off-
site disposal. All costs shown below have been converted to 1991 dollars using a 5%
annual escalation factor. The cost per cubic yard will tend to increase for smaller levels
of effort such as the cleanup of a typical LUST site.
Standard costs for earth-moving activities are available (Means, 1991).
Estimates of excavation costs for petroleum contaminated soils are in the range of $230
to $6.00 per ton (Troxler, et al., 1992). The costs of excavation will depend the level of
personal protective equipment required by the operator and on-site workers. Costs to
excavate soil contaminated with hazardous wastes for different safety levels are (Lippitt,
et al, 1986):
No hazard
Level D
Level C
Level B
Level A
$22.37 +/- 18.80 per m3 ($l/m3 = $1.30/yd3);
$74.73 +/- 56.19 perm3;
$9132 + /-83.79 perm3;
$117.10 +/- 8537 per m3; and
$13338 + /- 9638 perm3.
3-12
-------
Published cost estimates for excavation of soil contaminated with hazardous wastes vary
widely. The estimated cost to excavate and load sixteen million cubic yards at the Rocky
Mountain Arsenal was only $6/yd3 (U.S. GAO, 1986). The cost to excavate large
volumes of soils contaminated with explosives has been estimated to be $11.14/ton, or
about $13.92/yd3 (TVA, 1990). This cost comprised 79% labor, 7% operating expenses,
10% equipment expense, and 4% for site reclamation. The cost to excavate large
volumes of soils at another DOD site were estimated per cubic yard to be (Cullinane, et
al, 1986):
Dry excavation
Wet excavation
Site grading and revegetation
Site grading
Backfilling with clean soil
$5.36
(per yd3)
$10.72
$1.66
$1.15
$25.84
Cost estimates for transportation of petroleum contaminated soils range
from $0.08 to $0.15 per ton per mile (Troxler, et al., 1992). Vendor quotes for off-site
transportation of soil contaminated with hazardous wastes typically range from $2.50 to
$5/yd3 per mile, though they may be higher under some site-specific conditions. Costs
for transporting soil will be lower for on-site work and will be lower for off-site transport
if it is not necessary to follow the procedures typically employed for transporting soils
contaminated with hazardous wastes. Published cost estimates for off-site transport
include an estimate of about $3.80/yd3 per mile (Cullinane, et al, 1986) and an average
from ten sites of $025/ton per mile (Yang, et al, 1987). A cubic yard of soil can be
assumed to weigh about 2500 pounds.
Disposal costs are highly dependent on the amount and nature of
contamination present in the soil. Vendor quotes for disposal are typically $250 to
$350/yd3 of soil. Published estimates (Cullinane, et al, 1986) include costs of $38/yd3 for
disposal in a sanitary landfill and $160/yd3 for disposal in a RCRA landfill.
3-13
-------
3.7
Costs For Emission Controls
Costs for VOC controls for excavation are not widely available in the
literature. The following summary information was obtained from Eklund, et al., 1992b:
Control
Clay
Soil
Wood chips, plastic net
Synthetic Cover
Short-term foam
Long-term foam
Wind screen
Water Spray
Additives:
Surfactant
Hygro Salt
Bitu/Adhes.
Material Cost
($/M2 except as
noted)
$4.15
$1.33
$0.50
$4.40
$0.04
$0.13
$40/M
$0.001 (varies)
$0.65
$2.58
$0.02
Comments
Covers, mat, and membrane
Assume 6" deep; does not include soil
transport
Chip costs vary with site
Assume 45 ml thickness
Assume 2.5" thick, $0.7/M3 foam*
Assume 1.5" thick, $33/M3 foam*
Per linear meter
Assuming municipal water cost of
$!/$ 1,000 L. Water requires constant re-
application. Water truck rental:
$500/week.
Costs vary with chemical use
3.8
Equations and Models For Estimating VOC Emissions
The factors that govern excavation emissions are very complex. During
excavation, the physical properties of the soil that control the vapor transport rate (e.g.
air-filled porosity) are changing with time and the concentration of contaminants may be
rapidly decreasing. Predictive equations for estimating VOC emissions from excavation
have been developed by the U.S. EPA. The predictive equations require assumptions
3-14
-------
about the size of each scoop of soil, the dimensions of the soil scoops and the excavation
pit, and the shape of the soil after it is dumped. Further assumptions are required about
the air and soil temperatures and the length of time that dumped soil is exposed before
it is covered with more soil or with an emissions barrier.
Since it is rarely feasible or efficient to dig soil and immediately transfer
the soil directly to transport vehicles or treatment systems, any estimation procedure
must account for each event in which the soil is handled. In most cases, soil will be
excavated and placed into a temporary holding area and then handled one to two more
times on-site. Elevated levels of VOC emissions are possible each time the soil is
handled. When estimating emissions from sequential soil handling steps, it is important
to adjust the starting concentrations for each step to account for contaminants emitted
during prior steps.
The equations used are shown below. The average emission rate (g/sec)
from excavation is equal to the sum of emission rates from the soil pore space and from
diffusion:
ER - ERps + ERD1FF
PMW 106 E. Q ExC (
-------
Table 3-1.
Input Variables for Emission Equations
Variable
P
MW
R
T
E,
Sv
Q
irf
ExC
C
10,000
SA
*«,
Vo
it
t
De
M
Definition
Vapor pressure
Molecular weight
Gas constant
Temperature
Air-filled porosity
Volume of soil moved
Excavation rate
Conversion factor
Soil-gas to atmosphere
Concentration in bulk soif
Conversion factor
Emitting surface area
Equilibrium coefficient
Gas-phase mass transfer coefficient
Pi
Time
Effective diffusivity in air
Total mass of contaminant
Units
mm Hg
g/g-mol
mm Hg-cm3 /g-moF K
Degrees Kelvin
Dimensionless
m3
m3/sec
cm3/m3
Dimensionless
g/cm3
cmVm2
m2
Dimensionless
cm/sec
Dimensionless
sec
cm2 /sec
g
Default Value ;
35
100
62361
298
0.440
150
0.042
—
033
135 x 104
—
290
0.613
0.15
3.14
60
0.0269
_
Other Variables Required to Calculate Certain Variables Listed Above
W
ft
P
D.
U
Ma
P,
4
Time to excavate a given volume of soil
Bulk density
Particle density
Diffusivity in air
Wind speed
Viscosity of air
Density of air
Diameter of emitting area
sec
g/cm3
g/cm3
cm2 /sec
m/sec
g/cm-sec
g/cm3
m
—
15
2.65
0.1
2,0
1.81 x Id4
0.0012
24
'Soil concentration data is typically available as ppm or /*g/g. This value can be
multiplied by the bulk density of die soil (g/cm3) and by a conversion factor of
(g///g) to yield units of g/cm .
3-16
-------
The output from Equation 3-2 should be multiplied by the duration of
excavation and compared to the total mass of contaminants present in the soil:
,3
M = C * Sv * 106 —
m3
where: M = Total mass of contaminant in a given volume of soil (g).
(Eq. 3-4)
If Equation 3-2 gives a value that exceeds one-third of CTOJ, then the following equation
should be substituted for Equation 3-2:
where: t^ = Time to excavate a given volume of soil (sec).
Equation 3-3 can be used to estimate VOC emissions from storage piles.
3.9 Case Study
No suitable case study exists for excavation. Studies that have valid data
for emissions, control efficiencies, and costs are referenced above in the applicable
subsections.
3.10 References
Aim, R.R., ICA. Olson, and R.C. Peterson. Using Foam to Maintain Air
Quality During Remediation of Hazardous Waste Sites. Presented at the
80th Annual AWMA Meeting (Paper 87-183), New York City, June 21-26,
1987.
3-17
-------
Cullinane, M J., et al. Feasibility Study of Contamination Remediation at
Naval Weapons Station, Concord, California. Dept. of the Navy. NTIS
AD-A165623. February 1986.
Eklund, B., et al. 1989. Air/Superfund National Technical Guidance Study
Series, Volume HI: Estimation of Air Emissions from Cleanup Activities
at Superfund Sites. Report No. EPA-450/1-89-003 (NTIS PB 89-180061).
U.S. EPA, Research Triangle Park, NC, January 1989.
Eklund, B. Personal communication. 1990.
Eklund, B., et al. Air Superfund National Technical Guidance Series:
Database of Emission Rate Measurement Projects. EPA 450/1-91-003
(NTIS PB 91-222059). June 1991.
Eklund, B. Personal communication. 1992a.
Eklund, B., S. Smith, and A. Hendler. Estimation of Air Impacts for the
Excavation of Contaminated Soil. EPA 450/1-92-004 (NTIS PB92-171925).
March 1992b.
Evans, M. and H. Carroll. Handbook For Using Foams to Control Vapors
From Hazardous Spills. EPA/600/8-86/019 (NTIS PB87-145660). July
1986.
Landeeth, B., et al. Lining of Waste Impoundment and Disposal Facilities.
U.S. EPA Report SW-870 (NTIS PB86-192796). March 1983. .
Lippitt, J., et al. Costs of Remedial Actions at Uncontrolled Hazardous
Waste Sites: Worker Health and Safety Considerations. EPA/600/2-86/
037 (NTIS PB86-176344). September 1986.
Means, 1991. Means Site Work Cost Data, K. Smit, Sr. Editor. Published
by RJ. Grant. 1991.
Radian Corporation. Personal communication from Richard Pelt, Radian,
Research Triangle Park, NC. 1991.
Saunders, G.L Air/Superfund National Technical Guidance Study Series -
Development of Example Procedures for Evaluating the Air Impacts of
Soil Excavation Associated With Superfund Remedial Actions. EPA-
450/4-90-014 (NTIS PB90-255662). July 1990.
3-18
-------
Springer, C, K.T. Valsaraj, and LJ. Thibodeaux. In Situ Methods to
Control Emissions from Surface Impoundments and Landfills. JAPCA Vol.
36, No. 12, pp!371-1374, December 1986.
Suder, D.R. and CE. Schmidt. Control Efficiencies and Costs of Various
Technologies for Reduction of Volatile Organic Compound Emissions from
Exposed Hazardous and Non-Hazardous Waste. Presented at the 85th
Annual AWMA Meeting (Paper 11.15), Kansas City, MO, 1992.
Tennessee Valley Authority. Economic Feasibility Analysis for
Development of Low-Cost Chemical Treatment Technology for Explosive
Contaminated Soils. USATHAMA (NTIS AD-A223 497). May 1990.
Todd, Q.R., et al. Dust and Vapor Suppression Technologies for Use
During the Excavation of Contaminated Soils, Sludges, or Sediments. In:
Proceedings of the 14th Annual Hazardous Waste Research Symposium,
EPA/600/9-88/021 (NTIS PB89-174403). pp53-64. July 1988.
Troxler, W.L. Personal communication. 1992.
U.S. EPA, Estimating Air Emissions from Petroleum UST Cleanups.
Office of Underground Storage Tanks. Washington, DC. June 1989.
U.S. EPA, Engineering Bulletin: Control of Air Emissions From Material
Handling. EPA/540/2-91/023. EPA/ORD-Cincinnati. October 1991.
U.S. General Accounting Office. Hazardous Waste: Selected Aspects of
Cleanup Plan for Rocky Mountain Arsenal. NTIS PB87-102448. August
1986.
Yang, E.C., et al. Compendium of Costs of Remedial Technologies at
Hazardous Waste Sites. EPA/600/2-87/087 (NTIS PB88-113477).
October 1987.
3-19
-------
4.0 THERMAL DESORPTION
This section discusses both mobile and stationary process units designed for
soil remediation and the use of asphalt aggregate dryers for soil remediation. Data are
included in this section for the treatment of soil contaminated with petroleum fuels and
soil contaminated with hazardous wastes.
4.1 Process Description
In the thermal desorption process, volatile and semi-volatile contaminants
are removed from soils, sediments, slurries, and filter cakes. This process typically
operates at temperatures of 350°-700°F but may operate in the 200°-1200°F temperature
range. It is often referred to as low temperature thermal desorption to differentiate it
from incineration. At these lower temperatures, thermal desorption promotes physical
separation of the components rather than combustion. Contaminated soil is removed
from the ground and transferred to treatment units, making this an ex situ process.
Direct or indirect heat exchange vaporizes the organic compounds producing an off-gas
that is typically treated before being vented to the atmosphere (Vatavuk, 1990). The
best single source of information on thermal desorption is contained in an EPA
Guidance Document currently being prepared (Troxler, et al. 1992). The best currently
available source of information is an engineering bulletin prepared by the U.S. EPA
(U.S. EPA, 1991), which is included as Appendix D to this report.
After it is excavated, the waste material is screened to remove objects
greater than 1.5" to 3.0" in diameter (de Percin, 199 la). In general, four desorber
designs are used: rotary dryer, asphalt plant aggregate dryer, thermal screw, and
conveyor furnace. The treatment systems include both mobile and stationary process
units designed specifically for treating soil and asphalt aggregate dryers which can be
adapted to treat soils. Typical specifications for thermal desorption systems are shown in
Table 4-1.
4-1
-------
Table 4-1.
Comparison of Features of Thermal Desorption
and Offgas Treatment Systems
Estimated number of systems
Estimated number of contractors
Mobility
Typical site size (tons)
Soil throughput (tons/hour)
Maximum soil feed size (inches)
Heat transfer method
Soil mixing method
Discharge soil temperature f F)
Soil residence time (minutes)
Thermal desorber exhaust gas temperature C F)
Gas/solids flow
Atmosphere
Afterburner temperature C F)
Maximum thermal duty (MM Btu/hr^
Heatup time from cold condition (hours)
Cool down time from hot condition (hours)
Total Petroleum Hydrocarbons
Initial concentration (mg/kg)
Final concentration (mg/kg)
Removal efficiency (%)
BTEX
Initial concentration (mg/kg)
Final concentration (mg/kg)
Removal efficiency (%)
Rotary Dryer
4040
20-30
Fixed and mobile
500-25,000
10-50
2-3
Direct
Shell rotation
and lifters
300-6001
600-l^rf5
3-7
500-850*
800-1,00?
Co-current or
counter-current
Oxi dative
1,400-1300
10-100
03-1.0
1.0-2.0
800-35,000
< 10-300
95.0-99.9
NR
99
Cbavcybr Furnace \
1
-
Mobile
500-5,000
5-10
1-2
Direct
Soil agitators
300-800
3-10
1,000-1000
Counter-current
Oxidative
1,400-1300
10
0.5-1.0
Not reported
5,000
<10.0
>99.9
Not reported
<0.01
Not reported
Carbon steel materials ot construction
b Alloy materials of construction
cHot oil heat transfer system
"Molten salt heat transfer system
e Electrically heated system
'Not used on all systems
9Total duty of thermal desorber plus afterburner
h Vendor information: Soil Purification, Inc.
Source: Troxler, personal communication, 1991.
4-2
-------
Because thermal desorbers may operate near or above 1000°F, some
pyrolysis and oxidation may occur in addition to the vaporization of water and organic
compounds. Collection and control equipment such as afterburners, fabric filters,
activated carbon, or condensers prevent the release of the contaminants to the
atmosphere (de Percin, 199 la). Various types of thermal desorption systems can
produce up to nine process residual streams: treated soil, oversized media rejects,
condensed contaminants, water, paniculate control dust, clean off-gas, phase separator
sludge, aqueous phase spent carbon, and vapor phase spent carbon (de Percin, 199 Ib).
The final temperature is a function of residence time and heat transfer and
is the principle variable in controlling effectiveness. A study by the Hazardous Waste
Research and Information Center and the Gas Research Institute showed that
temperatures and residence times effective in bench-scale systems also proved effective
in pilot-scale systems. Such data support the use of a bench-scale test to determine the
best residence time and temperature variables as well as whether the thermal desorption
process will suitably treat the waste (de Percin, 1991a). The typical treatment
temperature range for petroleum fuels from leaking underground storage tanks (LUST)
sites is 400°F to 900T. For the treatment of soils containing pesticides, dioxins, and
PCB's, temperatures should exceed 850°F (de Percin, 1991c). The distillation
temperature range will vary with the type of fuel contaminated soil as shown in Figure
4-1.
Thermal desorbers effectively treat soils, sludges and filter cakes and
remove volatile and semi-volatile organic compounds. Some higher boiling point
substances such as polychlorinated biphenyls (PCB's) and dioxins may also be removed
(if present). Inorganic compounds are not easily removed with this process, although
some relatively volatile metals such as mercury may be volatilized. Temperatures
reached in thermal desorbers generally do not oxidize metals (de Percin, 1991a).
4-3
-------
No. 6 Fuel Oil
No. 4 Fuel Oil
No. 3 Fuel Oil
No. 2 Fuel Oil
No. 1 Fuel Oil (kerosene)
Jet Fuel - A
Jet Fuel - B
Automobile Gasoline
Naphtha (heavy)
Aviation Gasoline
Naphtha (light)
100 200 300 400 500
Distillation Temperature (C)
Figure 4-1. Soil Treatment Temperature Guide.
Source: Troxler, et al. 1990
4-4
-------
The soil is most effectively treated if it contains moisture level within a
specified range due to the cost of treating waste with a high water content. Typical
acceptable moisture ranges for rotary dryers and asphalt kirns are 10-30%, (Troxler,
personal communication, 1991 and SPI, 1991), while thermal screw systems can
accommodate a higher water loading of 30-80%. For VOC removal, soils ideally should
contain 10-15% moisture since water vapor carries out some of the VOCs (de Percin,
1991c).
High molecular weight organic compounds may foul or plug baghouses or
condenser systems, therefore, the types of petroleum products that may be treated by
specific technologies may be limited. Rotary dryers typically can treat soils that have an
organic content of less than two percent. Thermal screw units may treat soils that
contain up to 50% organics. (Troxler, personal communication, 1991).
Thermal desorption has several advantages over other treatment processes.
It treats a wide range of organic contaminants, is mobile, commercially available, and
enjoys more public acceptance than other thermal treatment methods (de Percin, 1991a).
Also, because thermal desorbers operate at lower temperatures than incinerators,
significant fuel savings may result (Vatavuk, 1990). They also produce smaller volumes
of off-gases to treat than incinerators. Thermal desorption also differs from incineration
in the regulatory and permitting requirements and in the partitioning of metals within
the process residual streams.
Potential limitations of the treatment process exist as well. Foremost,
thermal desorption does not destroy contaminants; it merely strips them from the solid
or liquid phase and transfers them to the gas-phase. Therefore, devices to control VOC
emissions are necessary. Since metals (e.g., Pb) tend to remain in the soil after
treatment, further treatment of the soil, such as stabilization, may be required. The
efficiency of the thermal desorption process will vary with the chemical and physical
properties of the specific contaminants.
4-5
-------
A generalized schematic diagram of a thermal screw, thermal desorption
process is shown in Figure 4-2; the system shown most closely resembles Weston's UP
system. Other designs may use different types of control technology. Information about
specific vendor designs is given below. The information in this section is primarily based
upon the use of portable remediation units, but the information should also be generally
applicable to other types of thermal desorption such as rotary dram aggregate dryers.
X'TRAX™ by Chemical Waste Management, Inc.
A transportable, indirectly heated rotary dryer, the X'TRAX™ system treats
up to 100 tons of soil and sediment contaminated with hazardous wastes per day.
Propane fires an outer shell which then heats the soil to 300°-900°F. Nitrogen gas
sweeps the water and organic vapor to gas treatment and mitigates explosion hazards.
Gas treatment comprises condensation, refrigeration, and carbon adsorption. The liquid
water is separated from the liquid organic compounds and used for dust control (de
Percin, 1991b).
Taciuk by SoilTech, Inc.
The Taciuk system is a two-zone, double-shell rotary dryer that treats up to
25 tons of soil and sediments contaminated with hazardous wastes per hour. The solids
enter the first zone of the inside shell where temperatures of 300°F vaporize water and
volatile organic compounds (VOCs). Entry into the second zone of the inside shell
enables additional organic compounds to be volatilized and pyrolyzed at temperatures of
1000°F. These high temperature solids enter the outer shell where they transfer heat to
the inner shell. Fired natural gas or propane heats the annulus between the shells. A
cyclone, baghouse, caustic scrubber, and carbon adsorber treat the combustion gases
while a condenser liquifies gases from both zones. Non-condensable gases from pyrolysis
help to heat the system (de Percin, 199 Ib).
4-6
-------
Classifier
contaminated soil
oversize objects
paniculate
control
dust
exhaust gases
t
vc
1
>Cs
r
Bag
House,
Thermal Processor
^
^
Heating
System
1
treated soil
clean off-gas
Stack
Figure 4-2.
Generalized Process Diagram for Thermal Screw-Based Thermal
Desorption.
4-7
-------
LT3 by Roy F. Weston, Inc.
The Low Temperature Thermal Treatment, or LT3 system, treats up to 20
tons of soil and sediment per hour using two banks of four heated screws. The process is
used primarily for treating hazardous wastes. The combustion of propane heats transfer
oil, which is pumped through the screws, heating the shell to 600°F. The combustion
gases sweep the water and organic vapor to the gas treatment system (de Percin, 1991b).
DAVES by Recycling Sciences, Inc.
The Desorption and Vapor Extraction System (DAVES) treats
contaminated material in a fluidized bed where it is fed along with hot air. Gas-fired
heaters heat the air to 1000°-1400°F. The hot air vaporizes water and organic
compounds and carries them to the gas treatment system. Gas treatment consists of a
cyclone, baghouse, venturi scrubber, chiller, and carbon adsorber (de Percin, 199 Ib).
ReTec by Remediation Technologies, Inc.
The ReTec thermal desorption system operates at capacities of 0.5 - 3.5
tons/hr and is designed to treat soils contaminated with organic compounds and oily
sludges. The process begins with a dewatering step if the waste material has a high
moisture content. Dewatered filter cakes from the press are fed to storage hoppers and
then transported to the dryer by a covered conveyer. The Holo-Flite* Processor consists
of a jacketed trough which houses a double-screw mechanism. The heat transfer fluid is
circulated through the trough jacket can by thermal oil or steam. The material enters
and exits the dryer through rotary air-locks to prevent leakage of ambient air into the
processor. The gas flow from the dryer, which is designed to remove moisture and lower
boiling organic compounds, passes through a particle removal system, quench chamber,
condenser and activated carbon beds (Abrishamian, 1991).
4-8
-------
The partially treated soil leaves the dryer and enters the processor where
the soil is subjected to temperatures up to 900T and shorter residence times (relative to
the drying step) to remove higher boiling organic compounds. Heated by an electric or
fuel-fired heater, molten salt is used as the heating medium for the processor. Molten
salt operates at temperatures between 500° and 950°F, does not produce off-gases, and
has the additional advantages of being non-toxic, non-flammable, and easily cleaned up if
spilled. Inert gas added to the processor inhibits oxidation and enhances vaporization of
contaminants. Off-gases from the processor undergo treatment with cyclones, a semi-
volatile organic separator, chilled condenser, and activated carbon beds. A solids cooler
lowers the temperature of the solids to less than 180°F for safe handling (Abrishamian,
1991).
4.2 Identification of Air Emission Points
The air emissions associated with thermal desorption come from several
sources. Most studies include data about contaminant concentrations in the soil directly
before and after treatment, which can yield information about point source air emissions
from the desorption process itself. These studies do not address, however, the change in
concentration before and after excavation due to volatilization. Similarly, little data is
available on fugitive emissions from individual units in the process other than the
desorption chamber and from the other waste streams.
Point sources of air emissions from thermal desorption vary widely with
each process. The stack of an afterburner vents combustion products, as does a fuel-
fired heating system if the combustion gases are not fed into the desorber. The fuel-
fired heating system typically operates with propane, natural gas or fuel oil. If VOC
emission controls consist of a baghouse, scrubber, and vapor phase carbon adsorber, the
stack will vent small concentrations of the original contaminants, as well as products of
any chemical reactions that might occur.
4-9
-------
The volume of off-gas from a thermal desorption unit depends on the type
of processor. Devices that are heated indirectly have off-gases composed of volatilized
VOCs and water from the soil being treated and possible some sweep gas used to carry
the contaminants out of the device. This volume of gas is typically 1000 to 5000 acfin
(Troxler, personal communication, 1991). In directly heated units, the off-gas contains
volatilized contaminants and water, but also the combustion gases used to heat the soil.
The result is a much larger volume of off-gas that needs to be treated, around 10,000 to
50,000 acfin (Troxler, personal communication, 1991). Therefore, off-gases from
indirectly heated LTTDs, i.e. thermal screws, can be treated with smaller chemical/
physical systems, such as a baghouse, a condenser, and then an afterburner.
Fugitive emissions from area sources may contribute significantly to the
total air emissions from a remediation site. Probably the largest source is excavation of
the contaminated soil. Other sources may include the classifier, feed conveyor, and the
feed hopper. Fugitive emissions from the components of the thermal desorption system
and controls are possible as well. Emissions may also emanate from the waste streams
such as exhaust gases from the heating system, treated soil, paniculate control dust,
untreated oil from the oil/water separator, spent carbon from liquid or vapor phase
carbon adsorber, treated water, and scrubber sludge.
43 Typical Air Emission Species of Concern
The volatile and semi-volative contaminants under remediation are the
species emitted if no destruction or other chemical treatment has taken place. The
sources emitting these VOCs may include excavation, soil handling, classifier, oversize
objects rejected by the classifier, feed conveyor, feed hopper, control stack, and fugitive
emissions from the entire thermal desorption system and from waste streams.
Combustion products are emitted when a destructive control device such as an
afterburner is used and also when the heating system is fuel-fired. In some cases,
pyrolysis occurs to a certain degree in the dryer, so products from these reactions may
also be emitted.
4-10
-------
4.4 Summary of Air Emissions Data
There is little published data of the levels of air emissions from thermal
desorption systems. This reflects the relatively early stage of commercial development
for this technology.
4.4.1 Air Emissions Data for Asphalt Aggregate Dryers
The VOC emissions from asphalt aggregate dryers will vary by several
orders of magnitude depending on whether afterburners are used as a control device.
Soil Purification, Inc. (SPI), a subsidiary of a leading manufacturer of
asphalt plants, has estimated the typical emissions for soil treatment in a modified
asphalt aggregate dryer. This system consists of a direct-fired rotary drum operating at
550-1000°F. A primary cyclonic tube collector and pulse-jet baghouse are used to control
paniculate emissions. A thermal oxidizer (i.e., afterburner) destroys organic compounds
in the off-gas stream (99-99.99% efficiency). Based on a processing rate of 35-60 tons
per hour, typical emissions from this type of unit are:
• Paniculate - 0.02-0.03 gr/dscf; and
Total VOC - 0.1-5 Ib/hr.
Data were identified for three asphalt plants that were modified for the
treatment of petroleum contaminated soils. Afterburners were not used on any of these
systems. A summary of the data is presented in Table 4-2. Each site will be discussed in
more detail below.
Soil contaminated with diesel fuel and gasoline were treated at an asphalt
plant with a 450 tons/hour capacity at 5% moisture (Barr, 1990). Soil enters the dryer
opposite of the burner and flows countercurrent to the combustion gases. This
configuration allows the VOCs which desorb from the soil in the upper portion of the
dryer to exit the system without exposure to the burner flame. The results presented for
this plant, labelled Barr, represent the worst case conditions since all of the VOCs
4-11
-------
Table 4-2.
Summary Data for Asphalt Kilns
Site
Procts* Parameter
.
Value, Units
=====
Contaminant
1. Barr Engineering (Barr, 1990) '"~
Di
Processing Equipment
Control Equipment
Air How
Feed Rate
Slack Exit Gas
Temperature
Soil Characteristics
Initial
Concentration
(ppm)
Final
Concentration
(ppm)
Removal
Efficiency
Ofl-gaa Characteristics
Stack
Concentration
Units
Estimated
Emission
Rate (g/hr)
Asphalt Plant
Wet scrubber
Cyclonic demisler
80,000 acfm at 300°F
280 lons/hr
150'F
Benzene
Toluene
Xylenes-m,p
xylene-o
THC
Naphthalene
Acenaphtylene
Acenaphlhene
Fluorene
Phenanlhrene
Anthracene
Fluoranthene
Pyrene
Bcnzo(a)an(hracene
chrysene
benzo(b)fluoranlhene
benzo(k)fluoranthene
benzo(a)pyrene
indeno(l,2,3-c,d)pyrene
dibenzo(a,h)anlhracene
benzo(g,h,i)perylene
Paniculate
Gasoline Contaminated Soil
Processing Equipment
Control Equipment
Airflow
Peed Rate
Stack Exit Gas
Temperature
Asphalt Plant
Wet scrubber
Cyclonic demisler
80,000 acfm at 300°F
280 lons/hr
150'F
Benzene
Toluene
Xylenes-m,p
xylene-o
THC
Naphthalene
Acenaphlylene
Acenaphlhene
Fluorenc
19.5
<.5
<.8
3.1
39.5
<2
<3
15.6
<.01
0.1
0.2
<.01
99.9
93.299.9
93.2-99.9
f
4.3
0.6
0.42
2590
6757
901
638
763
645
427
135
111
<=9
7.5
<=4.9
<«=9
<-2.0
<=4.5
<=4.4
<=6.0
0.2
ppm,d
ppm,d
ppm,d
ppm
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
gr/dscf
15
2.4
1.92
115200
7.80096
1.03824
0.73782
0.90876
0.74268
0.49032
0.15546
0.12744
0.01008
0.00282
0.00558
0.01008
0.00228
0.00504
0.00492
0.00672
29000
<.01
<.02
1.2
<.01
99.9
93.2-99.9
93.2-99.9
8.6
0.8
3.5
2800
5136
634
317
4OS
ppm,d
ppm,d
ppm,d
ppm
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
30.6
3.36
16.8
140600
5.5644
0.68784
0.35628
0.46206
-------
Table 4-2. (Continued)
Site
Process Parameter
Value, Units
Contaminant
'henanthrene
Anthracene
•luoranthene
'yrene
Benzo(a)anthracene
chrysene
benzo(b)fluoranlhene
benzo(k)(1uoranthene
benzo(a)pyrene
indeno(l ,2,3-c,d)pyrene
dibenzo(a,h)anthracene
benzo(g,h,i)perylene
Paniculate
Petroleum Hydrocarbons
Soil Characteristics
Initial
Concentration
(ppm)
Final
Concentration
(ppm)
Removal
Efficiency
85-94
Off-gas Characteristics
Stack
Concentration
385
<1.4
24
32
<4
<.8
<2.2
<3.9
<.9
<2
<2
<2.6
0.2
Units
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
ug/Nm3
gr/dscf
Estimated
Emission
Rate (g/hr)
0.43656
0.00162
0.02682
0.03474
0.0045
0.0009
0.00252
0.00444
0.00096
0.00222
0.00222
0.003
30400
2. Petroleum Contaminated Soil (Batten, 1987)
Processing Equipment
Control Equipment
Feed Rale
Soil Temperature
Residence Time
rotary drum dryer
baghouse
120 tons/hr
350-400F
2-3 min
Benzene
Toluene
Xylenes
elhylbenzene
total petroleum HC
THC
0.11
0.27
13,1
0.11
440
39-393
0.06
<.01
0.1
•c.Ol
<=5.5
5.7-9.5
84.5
100
100
100
97
61^5
129-175
3. Soil Cleanup System at Kingvale (SCS, 1990)
Processing Equipment
Manufacturer
Control Equipment
Feed Rale
Soil Temperature
Exit Gas Temperature
Estimated Diesel Input
Exhaust Oas 1 low
Soil Processed
rotary kiln, asphalt plant
Earth Purification Engineering, Inc.
cyclones
baghouse
1.8yd3/hr
775F
800-1000F
9.1 Ib/hr
1552 dscfm
I50yd3
non-methane VOCs
Semi-VOAs
VOCs+Semi-VOAs
Paniculate
Diesel
1875
<1
89
71
268
0.1278
ppmv as propane
ppmv
gr/dscf
472
712
— - — — — — .
-------
volatilized from the soil may not be destroyed and no additional VOC control device is
present (Ban, 1990). The hot exhaust gases are routed to a wet scrubber and a cyclonic
demister. The gas is then emitted to the atmosphere through the stack.
The feed rate for the diesel fuel and gasoline contaminated soils for the
test runs were 280 and 255 tons/hour, respectively. The air pollution control equipment
accommodates 80,000 acfrn at 300T (Barr, 1990). The soil headspace concentrations
and removal efficiencies are reported for the remediation tests. The uncontrolled
organic and controlled paniculate emissions are also presented in Table 4-2. Measured
total hydrocarbon (THC) emission rates for these tests were 254 to 310 Ib carbon per
hour (i.e., about ten times the typical emission rate during asphalt production).
Emission rates of paniculate matter were 64 to 67 Ibs/hour.
A second trial burn was performed on a soil contaminated with petroleum
hydrocarbons. The plant treats up to 120 tons/hour of soil at temperatures around 350-
400°F (Batten, 1987). The soil concentrations, removal efficiencies and emissions data
are presented in Table 4-2 for VOCs and other components. The exhaust gas from the
system contained 129 to 175 ppmv of THC above background. THCs were emitted at a
rate between 30.4 and 47.7 Ib/hr. The estimated emission factor for total non-methane
hydrocarbons was 021 to 026 Ibs per ton of soil treated. Based on the results, Batten
(1987) concluded that hydrocarbon controls would be necessary in order for the asphalt
kiln to meet air pollution control requirements.
The Soil Cleanup System (SCS, from Earth Purification Engineering Inc.)
was demonstrated in the treatment of a diesel contaminated soil from a leaking
underground fuel tank in Kingvale, California. The SCS is an asphalt recycling unit
modified to treat contaminated soils. The offgas from the rotary kiln is routed to dual
cyclones, an exhaust cooler, and a baghouse. The soil exits the system at 775°F. The
results for this site are presented in Table 4-2. The emission rates for non-methane
VOCs and semi-volatile organics were 1.04 and 157 Ib/hour, respectively, or 0.44 and
0.67 Ib/ton assuming 1.25 ton/yd3 (SCS, 1990).
4-14
-------
4.42 Air Emissions Data for Mobile Units
Several sets of data were found for the treatment of various contaminated
soils by thermal desorption. Table 4-3 presents a summary of the data for all of the
sites. The data represent the treatment of soils contaminated with petroleum fuels and
soils contaminated with hazardous wastes. Each set of data is described in more detail
below.
According to Weston Services, Inc., the use of their full-scale LT3 system
on the Springfield, Illinois cleanup of gasoline and No. 2 fuel oil contaminated soils
produced stack emissions that were in compliance with federal and state regulations,
including VOCs, HC1, CO, and particulates (Nielson and Cosmos, 1989). The emission
rate of BTEX (benzene, toluene, ethylbenzene, and xylene) was 1055 grams/hour before
controls and 21 grams/hour after the control devices, as shown in Table 4-3.
ReTec's thermal adsorption unit was used hi the remediation of a coal tar
contaminated soil. Molten salt in a thermal screw was used to indirectly heat the soil to
approximately 450°F. The soil was treated at a rate of 100 pounds per hour. The
controlled emissions of BTEX were 0.26 grams per hour or 0.011 Ib/ton (U.S. EPA,
1991). Initial and final soil concentrations for various contaminants are also presented in
Table 4-3.
At the McKin Superfund Site in Gray, Maine, soil containing primarily
trichloroethylene (TCE) was treated by Canonic Environmental Services Corporation.
Temperatures varied between 150 and 380°F, and the capacity was 1-4 cubic yards per
batch. To achieve 0.1 ppm TCE concentration in the treated soil, the temperature was
adjusted to 300°F for 6-8 minutes (Webster, 1986). The concentration reductions,
removal efficiencies, and emission rates for VOCs are presented in Table 4-3. The total
reported emission rates were 24 g/hr.
4-15
-------
Table 4-3.
Summary Data for Mobile Thermal Desorption Units
Contaminant;
Soil Characteristics
Initial
Concentration
Final
Concentration
Units
Removal
Efficiency
Offgas Characteristic*
Control
Efficiency
Stack
Concentration
Units
Uncontrolled
Emission
Rate (g/hr)
Estimated
Controlled
Emission
Rate (|Ar)
I. L13 full scale on no.Z fuel oil and gasoline contaminated soils (Nielson and Cosmos, 1991; U.S. EPA, 1991)
Processing Equipment
Manufacturer
Healing Medium
Control Equipment
Feed Rate
Soil Temperature
Residence Time
Thermal screw
Weston Services Inc.
Oil
Fabric Filter
Condenser
Afterburner
7.5 tons/hr
350-400F
70min
benzene
toluene
xylene
ethyl benzene
naphthalene
Carcinogenic Priority PNAs
benz(a)anthracene
benzo(a)pyrene
benzo(b)fluoranthene
chrysene
dibenzo(ah)anthracene
Noncarcinogenic Priority PNAs
acenaphthene
acenaphthalene
anthracene
benzo(ghi)perylene
benzo(k)fluoranlhene
fluoranlhene
fluorene
indeno(123-cd)pyrene
phcnanlhrene
pyrene
1000
24000
110000
20000
4900
<6000
<6000
<6000
<6000
<6000
890
1200
2700
<6000
<6000
<6000
4900
<6000
2400
<6000
5.2
5.2
<1
4.8
<330
<330
<330
<330
590
<330
<330
<330
<330
<330
<330
<330
<330
<330
430
<330
ppb
ppb
ppb
ppb
ppb
ppb
ppb
ppb
ppb
ppb
ppb
ppb
ppb
ppb
ppb
ppb
ppb
ppb
ppb
ppb
99.5
99.9
>99.9
99.9
>99.3
<94.5
<94.5
<94.5
<90.2
<94.S
>62.9
>72.5
>87.8
<94.5
<94.5
<94.5
<93.3
<94.5
82.1
<94.5
98
98
98
98
98
163
749
136
31.1
3.27
15
2.72
0.622
2. ReTec's Pilot Unit with Coal tar contaminated soils (U.S. EPA, 1991)
Processing Equipment
Healing Medium
Control Equipment
Feed Rale
Thermal Screw
molten salt
cyclones
semi-voa separator
chilled condenser
activated carbon beds
100 Ib/hr
benzene
toluene
ethylbenzene
xylenes
naphthalene
fluorene
phenanlhrene
anthracene
1.7
2.3
1.6
6.3
367
114
223
112
<.l
<.l
<.l
<.3
94
>95
>93
>95
>99
>99
91.9
95
95
95
95
95
0.726
0.999
0.681
2.72
166
0.0363
0.0499
0.034
0.136
8.29
-------
Table 4-3. (Continued)
Treatment Temperature
450F
Contaminant
fluoranthene
pyrene
benzo(b)anlhracene
chrysene
benzo(b)fluoranthenc
bcnzo(k)fluoranlhcne
benzo(a)pyrene
benzo(ghi)perylene
indeno(l 23-cd)pyrene
Soil Characteristics
Initial
Concen(ra(ion
214
110
56
58
45
35
47
24
27
Final
Concentration
15
11
<1.4
3.7
<1.4
<2.1
<.9
<1.1
<6.2
Units
ppm
ppm
ppm
Ppm
ppm
ppm
ppm
ppm
ppm
Removal
Efficiency
93
90
>97
93.6
>97
>94
>98
>95
>77
Offgas Characteristic*
Control
Efficiency
Stack
Concentration
Units
Uncontrolled
Emission
Rate (g/hr)
Estimated
Controlled
Emission
Rate (g/hr)
3. McKin Supcrfund Sile, TCB contaminated soil (Webster, 1986)
Manufacturer
Control Equipment
Residence lime
Soil temperature
Feed Rate
Canonic Environmental
Services Corp.
baghouse
packed lower air scrubber
carbon bed adsorbers
6-8 min
300F
4212 Ib/hr
Trichloroethylene
tetrachloroelhylene
1,1,1-lrichloroethane
1 ,2-dlchlorobenzene
toluene
xylenes
17-115
11-19
.11-.3
3.5-50
1-2
5-69
nd, 0.05
nd, 0.05
nd, 0.05
nd, 1
nd, I
nd, 1
ppm
ppm
ppm
ppm
ppm
ppm
>99
>99
>55
>71
>0
>80
95
95
95
95
95
95
220
36.2
0.478
93.7
19.91
130
11
1.81
0.0239
4.68
0.0956
6.5
4. Thermolech Systems Corporation (Thermotech, 1990-1991)
Control Equipment
Air Row
Afterburner Retention Time
DRB
dust collector
thermal oxidizer
64,000 acfm
2 sec
95-98%
Washington, D.C.
Afterburner Temperature
Row Rate
1408F
39014 acfm; 7484 dscfm
paniculate
benzene
toluene
ethylbenzene
xylene
TPH
0.0077
0.038
0.79
<=.12
<-.03
<-.05
21.7
gr/acf
gr/dscf
ppm
ppm
ppm
ppm
ppm
2.38
0.102
<-.0187
<-.0041
•C-.0077
0.42
-------
Table 4-3. (Continued)
Site
Process Parameter
Grand Rapids, MN
Feed Rate
Afterburner Temperature
Flow Rate
JSSSSSSg^S SSSSSSSS^SSSSSSSSSSSSi
Value, Units
33.8 tons/hr
1445F
51 366 acfm; 9070 dscfm
Contaminant
Paniculate
Soil Characteristic*
Initial
Concentration
Final
Concentration
Unit*
Removal
Efficiency
Of/gas Characteristics
Control
Efficiency
Stack
Concentration
0.0131
0.064
Units
gr/acf
gr/dscf
Uncontrolled
Emission
Rate (g/hr)
• '
Estimated
Controlled
Emission
Rate (g/hr)
4.98
5. Mobile Thermal Processor, Model 100 (Remedial Technology Unll, 1990)
Manufacturer
Control Equipment
U.S. Waste Thermal Proc.
Venturi scrubber
Afterburner
Gasoline Contaminated Soil
Feed Rale
Moisture
Soil Exit Temperature
Afterburner Temperature
Row Rate
5.65 yd3/hr
7.23%
299F
1825F
2491 dscfm
Particulates
Gasoline
Dloxins
Furans
Dlchlorobiphenyl
other PCBs
naphthalene
phenanthrene
anlhacene
fluoranihene
pyrene
other PAHs
5000
Diesel Contaminated Soil
Feed Rate
Moisture
Soil Exit Temperature
Afterburner Temperature
Flow Rate
3.95 yd3/hr
6.34%
450F
1825F
2361 dscfm
Particulates
Diesel
Dloxins
Furans
Dichlorobiphenyl
other PCBs
naphthalene
phenanthrene
anthacene
pyrene
other PAHs
5500
nd
mg/kg
0.0084
nd
nd
0.073
nd
nc
33
1.5
1.3
1.7
nd
gr/dscf
Ug/dscm
ug/dscm
ug/dscm
ug/dscm
ug/dscm
ug/dscm
ug/dscm
ug/dscm
ug/dscm
ug/dscm
4.2
nd
mg/kg
0.0057
nd
nd
nd
nd
6.6
13
0.25
nd
nd
nd
gr/dscf
ug/dscm
ug/dscm
ug/dscm
ug/dscm
ug/dscm
ug/dscm
ug/dscm
ug/dsco
ug/dscm
ug/dscm
2.7
-------
Table 4-3. (Continued)
Process Parameter
Contaminant
Soil Characteristics
Initial
Concentration
Final
Concentration
Units
Removal
Efficiency
Offgas Characteristic*
Control
Efficiency
Stack
Concentration
Units
Uncontrolled
Emission
Rale (g/hr)
Estimated
Controlled
Emission
Rate (g/hr]
16. Todds Lane Soil Remediation Plant (United Engineers and Constructors, 1991)
Control Equipment
Feed Rate
Afterburner Efficiency
cyclone
multiclones
baghouse
afterburner
IBS tons/hr
99%
benzene (max.avg)
toluene (max.avg)
elhylbenzene (max.avg) .
xylene (max.avg)
99
99
99
99
20,3.7
100,24
360,69
130,35
ppm
ppm
ppm
ppm
0.0696
0.348
1.25
0.452
7. Soil Remediation Unit 202 (Air Consulting and Engineering, 1991)
Processing Equipment
Control Equipment
Feed Rate
Afterburner Temperature
Afterburner Retention Time
Slack Row Rate
portable rotary kiln
afterburner
baghouse
25 tons/hr
HOOF
.5 sec
8300 scfmd
VOC
Paniculate
88.3
95.4
98.4
99
99
99
95
95
95
13.6
6.3
2.5
0.0303
0.0266
0.076
ppm dry as propane
ppm dry as propane
ppm dry as propane
gr/scf
gr/scf
gr/scf
0.74
0.29
0.1
2.52
1.89
1.65
8. X'TRAX pilot scale (U.S. EPA, 1991)
Processing Equipment
Control Equipment
Feed Rate
externally fired rotary kiln
wet scrubber
condenser
participate filler
carbon adsorption unit
secondary scrubber
4.5 metric tons/hr at 30% moisture
Clay
Silly clay
clay
sandy
clay
PCBs
VOCs
PCBs
VOCs
PCBs
VOCs
PCBs
VOCs
PCBs
VOCs
5000
1320
2800
1031
1600
530
1480
2950
630
2100
24
57
19
72
4.8
35
8.7
170
17
180
ppm
ppmv
ppm
ppmv
ppm
ppmv
ppm
ppm
99.3
95.6
99.5
93
99.7
93.3
99.1
94.2
97.3
91.4
<. 00056
<. 00055
<. 00051
<. 00058
<. 00052
mg/m3
mg/m3
mg/m3
mg/m3
mg/m3
0.02
0.03
0.01
0.07
0.08
-------
Table 4-3. (Continued)
Site
9. 11
Process Parameter
Value, Units
*~'"""^^^B^BBB^B-B!!!-B»"»»~»«—««»
Contaminant
===^=
Initial
Concenlralion
Soil Characteristics
Final
Concentration
Processing Equipment
Control Equipment
Soil Temperature
Residence Time
Moisture
rotary dryer with nitrogen
scrubber
condenser
HBPA filter
activated carbon unit
secondary scrubber
1025F
lOmin
4.64-6.87%
phenol
2-methylphcnol
4-melhylphenol
2,4-dimelhylphenol
naphthalene
2-melhylnaphthalene
acenaphthylene
acenaphthene
dibenzofuran
fluorene
phenanthrene
anthracene
duoranlheme
pyrene
benzo(a)anlhracene
chtysene
bis(2-ethylhexyl)phthalale
benzo(b) and (k)fluoranlhene
benzo(a)pyrene
indeno( 1 23-cd)pyrene
dibenzo(ah)anlhracene
benzo(ghi)perylene
Mercury
melhylene chloride
acetone
carbon disulfide
2-bulanone
1,1,1-lrichloroclhane
richlorotrifluoroethane
benzene
cylenes
<2.3
<1.3
<1.9
<6.6
130
170
36
320
220
365
965
390
630
550
160
890
150
72
<=24
<«=3.8
<=24
18
BDL
BDL
BDL
nd
nd
BDL
BDL
BDL
BDL
BDL
<.043
<.023
<.033
<.12
•c.016
<.19
<.007
<.21
<.08t
<.02
<.034
<.073
<.01
<.OS2
•c.023
<.12
<.047
<.U
<.035
<.016
<.32
0.77
Units
nig/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
Removal
Efficiency
na
na
na
na
>99.99
>99.89
>99.98
>99.93
>99.96
>99.9
100
>99.98
100
>99.99
>99.98
>99.92
>99.97
>99.84
na
na
na
95.04
Offgas Characteristics
Control
Efficiency
Stack
Concentration
46600
7700
20300
4800
138000
165000
87000
327000
266000
512000
1600000
770000
1040000
770000
255000
325000
2100
276000
104000
39000
20500
32500
5000
27000
410000
450000
220000
Unit*
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
Uncontrolled
Emission
Rate (g/hr)
Estimated
Controlled
EtniMlon
Rate (g/hr)
. ,. 1
L_ i
K)
O
-------
Table 4-3. (Continued)
Process Parameter
Value, Unit*
Contaminant
chloromelhane
1,1-dichloFoethane
1,1-dichloroelhane
chloroform
bromodichloromelhane
Irichloroethene
ethylbenzene
styrenc
Soil Characteristics
Initial
Concentration
Final
Concentration
Units
Removal
Efficiency
Offgas Characteristic*
Control
Efficiency
Stack
Concentration
11000
650
700
930
450
400
47000
67000
Unit*
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
Uncontrolled
Emission
Rate (g/hr)
Estimated
Controlled
Emission
Rate (g/hr)
-------
The Thermotech Systems Corporation's Portable Soil Remediation Unit
was used to treat petroleum contaminated soils in Washington, D.C and Grand Rapids,
MN. The unit has air pollution controls for particulate matter (dust collector) and
organics (thermal oxidizer). The average particulate and VOC emissions of three runs
are presented in Table 4-3. For the Washington, D.C. site, the particulate, BTEX, and
TPH emissions were 5, less than or equal to 0.13, and 0.42 pounds per hour, respectively.
The emission rate for particulates from the Grand Rapids site were 2.4 Ib/hr
(Thermotech, 1990-1991).
U.S. Waste Thermal Processing's Mobile Thermal Processor, Model 100,
was used to treat gasoline and diesel contaminated soils. The transportable treatment
unit consists of a primary furnace with an afterburner to incinerate the combustibles.
The offgas from the afterburner is routed to a wet scrubber for particulate removal. The
soil exit temperature is maintained between 300 and 650°F, and the afterburner operates
at 1800°F, with a minimum residence time of 0.5 seconds. The scrubber is a dual-venturi
collision scrubber. The test was performed on soil at 5000 and 5500 mg/kg
contamination levels of gasoline and diesel fuel, respectively. The particulate emission
rates were 42 and 2.7 Ib/day (Remedial Technology Unit, 1990).
The Todds Lane Soil Remediation Plant handles soils contaminated with
petroleum hydrocarbons. Pollution control devices include a cyclone, multiclones, a
baghouse, and an afterburner. The maximum anticipated concentrations of certain
VOCs in the soil were used to estimate the emission rates of the compounds after the
afterburner, assuming 99% efficiency. The total emission rate for BTEX was determined
to be 2.1 pounds per hour (United Engineers and Constructors, 1991).
The Soil Remediation Unit (SRU) 202 consists of a rotary kiln, an
afterburner and a baghouse. This unit was used to treat contaminated soil, and the
emissions were reported as follows: particulate - 1.7 to 2.5 Ib/hr and VOC - 0.1 to 0.74
Ib/hr (Air Consulting and Engineering, 1991).
4-22
-------
The pilot-scale X'TRAX system uses an externally fired rotary kiln for the
treatment of soils contaminated with hazardous wastes. The offgas is first treated in a
liquid scrubber where paniculate matter is removed. The gas is then cooled further to
allow for condensation of the contaminants. The gas is routed through a paniculate
filter and to a carbon adsorber where most of the remaining organics are removed. The
VOC emissions ranged from 0.01-0.08 Ib/day for the treatment of clay and sandy soils
(U.S. EPA, 1991).
A pilot-scale test was performed by IT Corporation for the treatment of
creosote contaminated soils at the Burlington Northern Superfund Site. The thermal
desorption unit was operated at 1025°F and a residence time of 10 minutes. The LTTD
unit consists of a rotating desorber tube partially enclosed within a gas-fired furnace
shell. Nitrogen is introduced into the system to flush out desorbed contaminants and to
maintain an atmosphere that does not support combustion. The offgas is treated with a
cyclone, a primary scrubber, a condenser, a demisting filter, a paniculate filter and an
activated carbon unit. Finally, the offgas is scrubbed in a secondary scrubber and then
discharged to the atmosphere. Pretreatment and post-treatment soil analyses are
presented in Table 4-3 for VOCs and semi-volatile organics. The removal efficiencies
and stack gas concentrations are also presented. These values represent the average of
six samples (IT, 1991). The IT system was also used to treat PCB contaminated soils at
a rate of 40 to 70 pounds per hour. The result of these tests are presented separately in
Table 4-4; the operating temperature and residence time varied for each run.
4.5 Identification of Applicable Control Technologies
Control of volatile organic emissions is crucial to the overall success of
thermal desorption remediation of contaminated soils. Because the process uses physical
separation driven by heat, the vaporized contaminants would simply be transferred from
one medium (soil) to another (air) if no emission controls were employed. The types of
controls include both destruction and separation technologies. Typically two to six
4-23
-------
Table 4-4.
Estimated Emissions of Selected Compounds for the Cleanup of PCB-Contaminated Soil using the IT Process
Contaminant
PCB's
2,3,7,8-TCDD
2,3,7,8-TCDD
2,3,7,8-TCDD
2,3,7,8-TCDD
2,3,7,8-TCDD
2,3,7,8-TCDD
Residence
Time
(minutes)
19
40
19
10.5
24
5.6
20
Temperatiiire
•F
1022
1040
1040
1040
860
1022
1031
Initial
Concentration
37.5
260
236
266
233
48
56
Final
Concentration
2
0.018
0.018
0.018
0.5
0.084
0.23
Units
ppm
ppb
ppb
ppb
ppb
ppb
ppb
Rate of
Uncontrolled
Emissions g/hr
1.14
0.00832
0.00755
0.00851
0.00744
0.00153
0.00178
Overall Estimated
Percent Efficiency
95%
95%
95%
95%
95%
95%
95%
Estimated
Emissions Rate
g/hr
5.68e-02
4.16e-04
3.78C-04
4.26e-04
3.72e-04
7.67e-05
8.92e-05
-------
controls in series are chosen to suit the specific VOC contaminants present and the other
pollutants of concern. liquid phase and solid waste streams are usually treated on site
or stored for subsequent off-site treatment Typical VOC controls for point sources are
briefly described below; Section 5.5 contains additional information. More detailed
information is available in a recent EPA report (Eklund, et al., 1992).
Asphalt kilns will have similar air emission control devices as for mobile
thermal desorption units, except that no VOC controls are typically employed and the air
flowrates are higher requiring some differences in design parameters.
A majority of LTTD control devices use an off-gas treatment system
consisting of a cyclone, an afterburner, and a baghouse (fabric filter). The cyclone is
used to reduce the paniculate loading on downstream devices. The baghouse may be
placed before or after the afterburner; it efficiently removes particulates to low levels.
The afterburner oxidizes organics and carbon monoxide by thermal destruction. Some
LTTD systems use an afterburner followed by a quench chamber and a venturi wet
scrubber. This system is capable of controlling acid gases if they are a concern. Some
systems collect the organic contaminants, as shown in Figure 4-2, rather than destroying
them.
4.5.1 Particulate Removal
Off-gases from the desorber typically pass first through a paniculate
control device. Particles that become entrained in the off-gas stream may be removed
with cyclones, venturi scrubbers, or fabric filters. Collected particulates are usually
returned to the incoming waste stream and retreated with the soil.
Cyclone collectors remove particles by creating a vortex from the inlet gas
stream velocity. Centrifugal acceleration forces entrained particles outward where they
collide with the wall and fall to a collection point.
4-25
-------
Venturi scrubbers are sometimes used to treat desorber off-gas, and
efficiently remove particles greater than 0.5 ftm in diameter using an aqueous stream.
The performance is not affected by corrosive, sticky or flammable particles, but high
collection efficiencies require a higher pressure drop and thus more costly operation
(Sink, 1991). The scrubber may also serve as the initial condensation stage for water and
organic compounds.
The fabric filter may be a series of fine-mesh synthetic fabric similar to the
type used in asphalt batch plants. An induced draft fan can be used to draw the exhaust
gas through a filter. The filter may be a jet-pulse design such that high pressure (80
psig) air periodically removes accumulated particulates to collections bins. Dust from
the bins may then be combined with the contaminated soil for reprocessing. The
maximum allowable pressure drop across the filter may be 15 inches of water (Weston,
1990), though 3 to 8 inches of water is typical.
4S2 Condenser
The most important aspect of using condensers to remove VOCs from a
vapor stream is designing the condenser to most efficiently remove the specific
contaminants present. This physical separation process operates on the basis of the
contaminants' vapor pressures which vary widely. By reducing temperature or increasing
pressure until the saturation vapor pressure is reached, the vapor condenses to a liquid
phase and is treated accordingly. Contaminants with high vapor pressures require
correspondingly low condensation temperatures. In these situations, a quencher that
removes a large portion of the moisture present often precedes the condenser to prevent
icing. The coolant may be air, ambient water, brine, chilled water, or refrigerants. A
separator directs the vapor/gas and liquid streams to appropriate control systems.
4-26
-------
4.53 Liquid Phase Treatment
The liquid from the condenser is sometimes partially treated on-site. The
liquid is separated by a gravity oil/water separator. The insoluble light organic fraction
is skimmed off the top, placed into 55 gallon drum and stored for off-site treatment.
The contaminated water from the separator is passed through carbon adsorption columns
and then used for dust control (Weston, 1990).
4.5.4 VOC Control by Afterburner
Fume incinerators (i.e., afterburners) are often used for the control of
VOC emissions from thermal desorption systems, especially for systems used for treating
underground storage tank sites. An afterburner used in one system identified in the
literature is a 3.5 million BTU/hr gas-fired fume incinerator, but afterburners may use
up to 40 million BTU gas per hour. Typically, afterburners operate at 1400-1800°F and
have a residence time of 0.5-2 seconds. The air that carries the vaporized contaminants
serves as the combustion air. The flame vortex exposes the VOCs to temperatures and
turbulence necessary for complete combustion. A combustion air fan maintains a
minimum of three percent excess oxygen exiting the afterburner. Exhaust gases leaving
the afterburner are sometimes mixed with ambient air to be cooled and then passed
through a scrubber (Weston, 1990).
4.5.5 VOC Control by Carbon Adsorption
Often used as a polishing process after other treatments, carbon adsorption
works on the principle that contaminants are physically adsorbed onto the activated
carbon. No chemical change or reduction of the waste amount occurs. Absorption
processes can occur in either the liquid or vapor phase. Regeneration or disposal o
spent carbon may also produce emissions, though this is very rarely done on-site.
4-27
-------
A liquid phase carbon adsorption usually treats water with low contaminant
concentrations with a two-stage system. Clean water is often used to cool discharge
solids from the desorber and suppress dust formation (Nielson et al, 1989).
4.5.6 Scrubber
Exhaust gases from thermal destruction processes may be treated in a
scrubber to remove particulates or neutralize acid gases. Wet scrubbers use a liquid to
absorb pollutants from a waste gas stream; the process is enhanced through a large.
liquid/gas contact surface area. Wet scrubbers operate by either chemical, absorption
(reaction between pollutant and liquid), or physical absorption (pollutant trapped by
liquid). Dry scrubbers operate by chemical absorption. Acid gases are not typically a
concern when processing petroleum contaminated soils due to low concentrations of
halogenated compounds (Troxler, et al. 1992). Paniculate scrubbers (Venturis) can be
employed to capture the particles by impingement and agglomeration with liquid
droplets. If a wet scrubber is used to treat the off-gases, a liquid separator is needed
downstream of the scrubber, such as a cyclone or mist eliminator.
4.5.7 Miscellaneous System Adaptations and Control Approaches
Other emissions control techniques include using treated water for dust
control and using ultraviolet light. Ultraviolet rays have been used to destroy dioxin in
the condensate from the thermal desorption of contaminated soils.
While conducting a pilot study of the McKin Superfund site in Gray,
Maine, Canonic Environmental Services Corporation made efforts to control VOC and
dust emissions from excavation and aeration processes. The soil was contaminated with
trichloroethylene (TCE). Excavation down to 40 feet was conducted with a kelly bar
caisson rig fitted with a digging bucket and attached to a 100-foot crane. Soils
discharged from the digging bucket entered a front end loader equipped with a
4-28
-------
removable plastic cover. Cylindrical steel caissons were augered into the deep
excavation holes to prevent further volatilization (Webster, 1986).
4.6 Capital and Operating Costs for Remediation
Because thermal desorption is virtually never used without controls, the
costs reflect emissions controls as well as remediation. Most thermal desorption units
offered by vendors are predesigned systems with VOC and paniculate controls already
installed. This is especially true for portable systems which are typically housed on flat
bed trailers. Asphalt plants do not typically have VOC controls, but if they have been
modified to treat soils, the organic control device may already be added. The costs for
treatment of soils contaminated with petroleum hydrocarbons typically range from $50 -
125/ton for low temperature thermal screw units and from $35 - 100/ton for rotary
dryers (Troxler, et al. 1992).
IT Corporation performed a pilot study cleanup of PCBs on the
Rosemount Research Center site of the University of Minnesota and in its report
estimated that direct operating costs for a full-scale systems would be about $80/ton
based on a 10 ton/hour system treating soil with 20 percent moisture. This figure
includes $60/ton for labor, utilities, fuel, materials and supplies, and administrative costs
as well as $20/ton for depreciation. These are hopper to hopper treatment costs; total
costs are likely to be $175 to $350 per ton (Troxler, personal communication, 1992).
Costs depend on the contaminants present and site conditions. Cost estimations should
take into consideration planning and procurement, permitting, site preparation,
equipment mobilization, equipment erection/startup, operations, equipment
demobilization, and site closure (Fox, et al, 1991).
Soil contaminated with Herbicide Orange at the Naval Construction
Battalion Center (NCBC) in Gulfport Mississippi was treated with the FT Corporation's
pilot-scale thermal desorption/ultraviolet apparatus. The costs summarized below
4-29
-------
include the cost of ultraviolet destruction technology which is not typically a part of the
thermal desorption process (Helsel and Thomas, 1987):
Cost of treating dioxin-contaminated soil with TD/UV Photolysis
Amount of Soil (tons)
10,000
20,000
40,000
Total Cost
(millions of $)
6.002
8.030
11.796
Cost ;
$600/ton
$402/ton
$295/ton
Remediation Technologies, Inc. estimates the cost of treating oily soils and
sludges to be in the range of $100-$300/ton of feed. The costs depend on quantity of
waste, term of the contract, and moisture and organic content of the contaminated soil.
Estimated costs for some other systems are presented in Table 4-5. The installed cost of
complete thermal desorption systems which include treatment of off-gases and
condensates is usually about two to four times the cost of the thermal units themselves
(Abrishamian, 1991).
4.7
Capital and Operating Costs for Emission Controls
Costs for emission controls are included in the remediation costs given in
Section 4.6. The relative contribution of emission controls to the total costs are not
known. This information for selected thermal desorption systems is currently being
collected under the EPA SITE program and should be publicly available in 1992.
A cost estimate was determined for some typical controls used with
thermal desorption units. The cost estimate was calculated from procedures outlined in
various U.S. EPA documents. These values were also compared with vendor quotes.
The estimation was performed for gas flow rates of 5,000, 15,000, and 40,000 acfm. The
total capital investment was determined for both mechanical shakers and pulse-jet fabric
filters and includes equipment and installation costs. Site preparation and construction
costs are not included in this figure. The results are presented in Table 4-6.
4-30
-------
Table 4-5.
Remedial Costs for Various Thermal Desorption Units
System
X*TRAX", Chemical Waste Management
LTTA, Canonic Environmental Services, Corp.
Lf.RoyF. Weston
Cost
($/ton feed)
150-350
80-150
100-120
Soil Characteristics
30% moisture;
<10% organics
—
20% moisture;
10,000 ppm organics
Soil Feed
Rate
—
30-50 tons/hr
20,000 Ib/hr
SOURCE: Johnson and Cosmos, 1989
Table 4-6.
Cost Information for Fabric Filters
Filter Type
Mechanical Shaker
Pulse-Jet Fabric Filter
Flow Rate (acfm)
5,000
15,000
40,000
5,000
15,000
40,000
Estimated Capital Cost6
(1992$) :
159,000
298,000
509,000
124,000
205,000
456,000
Vendor Estimates1*
24,000"
36,400"
30,000"
52,000"
*Dustex Corporation
"Typical cost from Soil Purification Inc. given as $250-350,000 for a pulsed-jet fabric filter
for a "typical" size portable system.
"Estimated capital costs based on correlations given in the OAQPS Control Cost Manual
(Vatavuk, 1990).
4-31
-------
The cost estimation for an afterburner, or thermal oxidizer, was based on
similar flow rates (5,000, 15,000, and 40,000 scfm). Two cases were considered for each
flow rate: a) no heat exchanger (no heat recovery from the thermal oxidizer) and b)
50% heat recovery. The cost estimates and vendor information are summarized in Table
4-7. As stated earlier, thermal desorption units are typically sold as a predesigned unit
that already incorporates the control devices into the total cost of the system.
4.8 Equations/Models for Estimating Emissions
Theoretical models based on fundamental principles have been proposed
for predicting the evolution of volatile compounds from soil in the thermal desorption
process (Lighty, et al, 1990). Both particle desorption and bed desorption were
examined. The models are partial differential equations based on mass and energy
balances and on the Freundlich isotherm equation. In practice, an assessment of the
applicability of thermal desorption for a given site will not be based on modeling
calculations, but will be based on the types of contaminants present in the soil, the
physical properties of the soil, and the results of any bench-, pilot- or full-scale test runs.
In most cases, the process conditions such as temperature and residence time can be
modified to yield the desired removal efficiency, though heavier weight petroleum fuels
such as No. 6 fuel oil may present problems for systems with relatively low operating
temperatures. The cost to operate at these process conditions, however, will dictate
whether or not thermal desorption is competitive with other remediation options.
Using removal efficiencies obtained from test runs, a mass balance yields
the following equation to estimate an emission rate for a volatile compound leaving the
desorber. This estimate does not include emissions from excavation or other handling of
contaminated soil nor does it include fugitive emissions from the desorber system or
from liquid and solid phase waste streams. Combustion gases from the heating system
and exhaust gases from afterburners produce additional emissions not taken into account
by this estimation method.
4-32
-------
Table 4-7.
Cost Information for Thermal Oxidizers
Heat Recovery (%)
0
50
How Rate (scfm)
5,000
15,000
40,000
5,000
15,000
40,000
Estimated Capital Cost0
(1992$)
156,000
209,000
304,000
304,000
437,000
580,000
Vendor Estimates6
100,000a
300,000"
150,000"
450,000a
'Conversion Technology, Inc.
'Typical cost from Soil Purification, Inc. given as $75-300,000 for a thermal oxidizer for a
"typical" size system.
^Estimated capital costs based on correlations given in the OAQPS Control Cost Manual
(Vatavuk, 1990).
4-33
-------
where
ERj =
C =
1000 =
MR =
- CE0/100)
emission rate for contaminant i (g/hr);
concentration of species i in contaminated soil (mg/kg);
conversion factor (mg/g);
mass rate of soil treated (kg/hr);
percentage of contaminant i volatilized; and
percent efficiency of control devices.
Case Studies of Remediation and Air Emissions
The air emissions for one case are summarized below, followed by
treatment effectiveness data for a number of cases. Not all the case studies are specific
to soils contaminated with petroleum products, but they are included for illustrative
purposes.
4.9.1
Case Study of Air Emissions From Thermal Desorption
Canonic Environmental Services Corporation conducted a pilot study of
the McKin Superfund site in Gray, Maine. Air quality was monitored during the study to
estimate emissions from thermal desorption treatment of contaminated soil. The results
are shown below. Treatment results were presented in Table 4-3 (Webster, 1986).
Air Emissions From Thermal Desoi
Location :. •- • .• ' ';: • ;-j : • '
Within 2 feet of caisson bucket and front end loader
20 feet downwind of excavation (5 minute avenge)
Site Perimeter
Upwind of site, background, surrounding residences
Site Perimeter
Site Perimeter
Site Perimeter
Site Perimeter
Site Perimeter
Site Perimeter
Site Perimeter
Site Perimeter
)tion Treatment at McKin Superfund Site
Contaminant
VOC
VOC
VOC
VOC
TOE
COjF
1,2-dichloroethyiene
toluene
ethytbenzene
xylene
total suspended paniculates
total suspended paniculates
'Concentrations
1,000 ppm
5 ppm above background
0 above 2 ppm background
1-5 ppm
0.002 - 0.01 ppm
0.010 . 0.018 ppm
< 0.02 ppm
<0.02 ppm
< 0.02 ppm
< 0.02 ppm
> llOng/af without controls
> 50 jig/of* with controls
4-34
-------
4.92
Case Studies of the Efficiency of Thermal Desorption
Thermal desorption has been found to have high removal efficiencies for
volatile and semi-volatile compounds. Described below are several cases studies with
tables showing removal efficiencies as well as feed and product concentrations.
Use of X*TRAX™ on VOCs and PCBs
A laboratory-scale study of Chemical Waste Management's (CWM)
X'TRAX thermal desorber treated one to two kilograms per hour of various
contaminated soils. The X*TRAX system consists of a rotary kiln, scrubber, condenser,
paniculate filter, and carbon adsorber. The soil concentrations and removal efficiencies
are presented for several contaminants (Ayen and Swanstrom, 1991).
CWM X*TRAX™ Laboratory-Scale Treatment of Contaminated Soils
Soil Type
Petroleum contaminated soil
Contaminated silt, day, gravel
Non-PCB soil, sludges, and mixture day
soil
Soil/sludge
Sludge
Compound
Acetone
Xylenes
ethylbenzene
Styrene
Tetrachloroethylene
Chlorobenzene
12-DichIoroethane
Anthracene
bix(2-ethylhexyl)phthalate
pentachlorophenol
PCBs
Xylenes
1,2,4-Trichlorobenzene
di-N-butylphthalate
pentachlorobenzene
33-dichlorobenzidine
nitrobenzene
azobenzene
2-chloroanaline
benzidine
33-dichlorobenzidine
azobenzene
benzidine
33-dichlorobenzidine
azobenzene
Concentrations (ppm)
Feed
2600
2400
1600
200
150
110
38
4650
2380
497
805
18.8
24.8
132
11.6
1716
42.9
3000
779
792
700
44.6
13
503
16.8
Product
16
9.5
52
<.005
0.094
0.18
0.062
12
<33
2£
172
<.125
<33
<33
<33
<.66
<33
4.9
ND
ND
<.66
ND
ND
<.66
ND
Percent
Removal
9938
99.6
99.68
> 99.99
99.94
99.84
99.84
99.74
> 99.99
99.44
97.9
>993
>98.7
>97.5
>97.1
4-35
-------
The pilot-scale unit, which treats 4.5 metric tons per day at 30 percent
moisture, was used to treat various soils contaminated with PCBs. The following data
were reported for three types of soil (de Pertin, 1991a; Ayen and Swanstrom, 1991).
CWM X*TRAX™ Pilot-Scale Treatment of Sandy Soil with PCFs
Compound
PCB's
1,2,4-Triehlorobenzene
Di-n-Butylphthalate
Bis(2-Ethylhexyl)phthalate
Feed (ppm)
1480
2.9
1.0
9.1
Product (ppm)
8.7
not detected
0.24
0.18
% Removal
99.4
>99.9
76.0
98.0
CWM X'TRAX™ Pilot-Scale Treatment of Clay, Silt, and Gravel with PCB's
Co
PCB's (Arochlor 1254)
1,2,4-Trichlorobenzene
Di-n-Butylphthalate
Bis (2-Etbylhexyl) phthalate
Feed (ppm)
1400
2800
6.9
4.7
Product (ppm)
34
19
not detected
not detected
% Removal
97.6
993
>98.0
>97.2
CWM X'TRAX™ Pilot-Scale Treatment of Contaminated Soil
Compound
Methyl ethyl ketone
Tetrachloroethylene
Clorobenzene
Xylene
1,4-Dichlorobenzene
1,2-Dichlorobenzene
Hexachlorobenzene
Feed .(ppb)
100,900
91,000
61,800
56,400
78,400
537,000
79,200
Product (ppb)
<100
15.6
6.5
2.8
1.4
74.1
300
% Removal ..I
>99.9
99.98
99.98
99.99
99.99
99.99
99.62
4-36
-------
Study of Soil from Two Superfund Sites
In Phase II of an Environmental Protection Agency study on thermal
desoiption, soils from two Superfund sites were processed. A Lindberg furnace (Model
51848) was equipped with an electronic temperature controller and 1600 watt heater.
Berlin-Farro and Old Mill Sites soils were processed at 350T and 550°F with 30 minute
residence times. A summary of results is shown below (Lauch, et al, 1990).
EPA Study of Berlin-Farro Site Soil
Contaminant
2-Butanone (detected in blank)
Trichloroethene
Tetrachloroethene
Toluene
Xylenes (total)
Hexachlorobutadiene
Hexachlorocyclopentadiene
Pentachlorobenzene
Hexachlorobenzene
Containinant * : ''•*•
2-Butanone (detected in blank)
Trichloroethene
Tetrachloroethene
Toluene
Xylenes (total)
Hexachlorobutadiene
Hexachlorocyclopentadiene
Pentachlorobenzene
iexachlorobenzene
Feed fcg/kg)
290
147
280
483
387
1900
46000
10200
105000
Feed fcgfkg)
290
147
280
483
387
1900
46000
10200
105000
Product (Kg/kg)
343
<23
<23
19
<23
430
3050
15100
250000
Product (^gAg)
80
<25
3
27
<25
<3300
<3300
2500
47000
% Removal at 350T
<18>
>84
>92
96
>94
77
93
<48>
<138>
% Removal at 550°F
72
>83
99
94
>91
—
93
75
55
4-37
-------
EPA Study of Old Mill Site Soil
Contaminant
Trichloroethene
Tetrachloroethene
Toluene
Xylenes (total)
Aroclor 1260
Contaminant
Trichloroethene
Tetrachloroethene
Toluene
Xylenes (total)
Aroclor 1260
s i3Feed(^g/kg)?«
2400
362
152
950
2000
FeedM/kg)
2400
362
152
950
2000
Product 0/g/kg)
173
35
43
285
3000
Product (^g/kg)
<25
<25
57
48
not detected
^Removal at 350T
93
90
72
70
<50>
% Removal at 350°F
>99
>93
63
95
>95
Use of LT3 on JP-4 Fuel and Other VOCs at Tinker Air Forces Base
Weston's LT3 thermal desorption system was employed in a pilot study to
clean up contaminated soils at Tinker Air Force Base in Oklahoma. The contaminants
were primarily JP-4 jet propulsion fuel and chlorinated organic compounds. After
modifications, the residence time was set at 40 minutes and the heat transfer oil at
600°F. The system was designed to process 15,000 Ibs/hour but in practice handled 30
percent more than the design specification. The emissions control system included a
fabric filter, condenser, afterburner, and scrubber (Weston, 1990).
Pilot Study of Tinker Air Force Base Site
Contaminant
Vinyl chloride
Dichloromethane
1, 1-Dichlroethene
Chloroform
1,2-Dichloroethane
1, 1, 1-Trichloroethane
Trichloroethene
Tetrachloroethene
Feed (^g/kg)
<3500
<1800
<1800
140°
<1800
<1800
37250
2760*
Product 0*g/L)
0.2
0.1
0.1
0.1
0.1
0.1
03
0.1
% Removal
—
_
_
> 98.57
—
—
99.986
> 99.93
4-38
-------
f^f^*fP|iC6«tamiiiant '
2-Butanone
Benzeneb
Toluene"
Chlorobenzeneb
Ethylbenzeneb
1,2-Dichlorobenzene
13-Dichlorobenzene
1,4-Dichlorobenzene
Fluorantheneb
Benzo(a)anthraceneb
Benzo(a)pyreneb
Benzo(b)£luoranthene
Chryseneb
Dibenzo(a,h)anthraceneb
Acenaphtheneb
Acenaphtyleneb
Anthracene1*
Benzo(g,h,i)peryleneb
Fhioreneb
Indeno(l,2,3-c,d)pyreneb
Phenanthreneb
Pyreneb
Benzo(k)fluorantheneb
Trichlorofluoromethane
trans-l,2-Dichloroethene
m-Xylene
o,p-Xylene
Naphthalene
Feed^l/kg)
< 11000
<1800
<1800
<1800
<1800
35000
__c
8700
<3300
<3300
<3300
<3300
<3300
<3300
<3300
<3300
60*
<3300
<3300
<3300
<3300
7908
280*
<3300
___c
___c
__c
c
Product (Mg/L)
0.6
0.1
0.1
0.1
0.1
6a
c
6a
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<50
<10
<10
<10
c
c
— c
___c
% Removal
—
—
—
—
~_
> 98.46
>857
>6322
>48.48
—
—
—
—
—
—
—
_
—
—
—
>41.77
—
—
> 99.43
> 99.74
> 99.89
> 99.82
> 96.98
"Not detected at the specified detection limit.
Potential constituent of JP-4 fuel.
cOnly % removal data given in report.
4-39
-------
Use of ReTec's Thermal Screw on Semi-Volatile Organics
The ReTec pilot unit was used to remediate refinery filter cake, creosote
contaminated clay, and coal tar contaminated soil, which was presented earlier. The
process uses molten salt to heat the soil to 450 to SOOT, volatilizing the contaminants.
Cyclones, a semi-volatile organic separator, a chilled condenser, and activated carbon
beds are used to control the paniculate and contaminant emissions. The following table
presents the results (U.S. EPA, 1991).
ReTec's Pilot Unit
1
Waste Type
Refinery vacuum filter cake
Creosote contaminated clay
Compound
naphthalene
acenaphthylene
acenaphthene
fluorene
phenanthrene
anthracene
fluoranthene
pyrene
benzo(b)anthracene
chrysene
benzo(b)fluoranthene
benzo(k)fluoranthene
benzo(a)pyrene
dibenz(a,b)anthracene
benzo(ghi)perylene
indeno(123-cd)pyrene
naphthalene
acenaphthylene
acenaphthene
fluorene
phenanthrene,
anthracene
fluoranthene
pyrene
benzo(b)anthracene
chrysene
benzo(b)fluoranthene
benzo(k)fluoranthene
benzo(a)pyrene
dibenz(a,b)anthracene
benzo(ghi)perylene
indeno(123-cd)pyrene
Concentration (ppm)
Feed
<.l
<.l
<.l
10.49
465
9.8
73.94
15837
5633
64.71
105.06
22537
17458
477.44
16353
122.77
1321
<.l
293
297
409
113
553
495
59
46
14
14
15
<.l
7
3
Product
<.l
<-l
•c.l
<.l
<.l
98.9
>993
>96.6
>995
>99S
975
>99J9
97.9
98.4
98.9
975
96.6
96.6
>99.9
>99.96
>99.%
99.6
>99.7
99.7
99.6
>99.99
>995
823
>9957
>99.9
>99.4
>993
4-40
-------
4.10 References
Abrishamian, R. Thermal Desorption of Oily Soils and Sludges. Presented
at the 84th Annual AWMA Meeting, Vancouver, June 1991.
Air Consulting and Engineering, Inc. Personal communication. 1991.
Ayen, R J., and C. Swanstrom. Development of a Transportable Thermal
Separation Process. Environmental Progress, Vol. 10, No. 3. August 1991.
Barr Engineering. Personal communication. 1990.
Batten, R.L. Air Emissions From an Asphalt Plant Rotary Drum
Aggregate Dryer Used to Decontaminate Gasoline Contaminated Soil
(Trial Bum). Presentation at the CAPCOA Engineers Technical Seminar,
December 2-4, 1987.
de Percin, P.R., 199la. Thermal Desorption Attainable Remediation
Levels. In: Proceedings of the 17th Annual Hazardous Waste Research
Symposium, EPA/600/9-91/002, pp511-520. April. 1991.
de Percin, P.R., 1991b. Thermal Desorption Technologies. Presented at
the 84th Annual Meeting of AWMA, Vancouver, BC, June 1991.
de Percin, P.R., 1991c. Personal communication. 1991.
Eklund, B. Personal communication. 1992.
Fox, R.D., E.S. Alperin, and H.H. Huls. Thermal Treatment For the
Removal of PCBs and Other Organics from Soil. Env. Progress, Vol. 10,
No. 1, pp40-44, February 1991.
Helsel, R.W. and R.W. Thomas. Thermal Desorption/Ultraviolet
Photolysis Process Technology Research, Test, and Evaluation Performed
at the Naval Construction Battalion Center, Gulfport, MS, For the USAF
Installation Program - Volumes I and IV. AFESC, Tyndall Air Force
Base, FL. Report No. ESL-TR-87-28. December 1987.
IT Environmental Programs, Inc. and IT Corporation. Personal
communication with John Carroll of IT. 1991.
Johnson, N.P. and M.G. Cosmos. Thermal Treatment Technologies for
Hazardous Waste Remediation. Pollution Engineering. October 1989.
Lauch, R.P., et al. Low Temperature Thermal Desorption For Treatment
of Contaminated Soils - Phase II Results. In: Proceedings of the 15th
Annual Research Symposium, EPA/600/9-90/006 (NTIS PB91-145524),
pp!37-150. Feb. 1990.
4-41
-------
Lighty, J.S., et al. Fundamentals for the Thermal Remediation of
Contaminated Soils. Particle and Bed Desorption Models. ES&T Vol. 24,
No. 5, pp750-757, May 1990.
Nielson, R.K. and M.G. Cosmos. Low Temperature Thermal Treatment
(LT3) of Volatile Organic Compounds From Soil: A Technology
Demonstrated. Env. Progress, Vol. 8, No. 2, pp!39-142, May 1989.
Remedial Technology Unit. Personal communication. 1990.
Sink, M.K. Handbook of Control Technologies for Hazardous Air
Pollutant. EPA/625/6-91/014 (NTIS PB91-228809). June 1991.
Soil Purification, Inc. Vendor Information Concerning Asphalt Kilns
Modified for the Treatment of Contaminated Soils. 1991.
Soil Cleanup System (SCS). Personal communication with John Schmuck
of Soil Recycling Technologies, Inc. 1990.
Thermotech Systems Corporation. Vendor Information Concerning
Laboratory Results, Emissions and Remediation Reports for the Portable
Soil Remediation Unit. Thermotech Systems Corporation, Orlando, FL.
1990-1991.
Troxler, W.L. Personal communication. 1992.
United Engineers and Constructors. Air Toxics Screening Analysis, Todds
Lan Soil Remediation Plant. United Engineers, Philadelphia, PA.
Personal communication with Michael Kendall of the Baltimore County
Department of Environmental Protection and Resource Management.
August 1991.
U.S. EPA. Engineering Bulletin - Thermal Desorption Treatment.
EPA/540/2-91/008 (NTIS PB91-228080). May 1991.
Vatavuk, W.M. OAQPS Control Cost Manual (4th Edition). EPA/450/3-
90/006 (NTIS PB90-169954). January 1990.
Webster, D.M. Pilot Study of Enclosed Thermal Soil Aeration for
Removal of Volatile Organic Contamination at the McKin Superfimd Site.
JAPCA, Vol 36, No. 10, pp!156-1163, October 1986.
R.F. Weston, Inc. Task Order 4, Demonstration of Thermal Stripping of
JP-4 and Other VOCs From Soils at Tinker Air Force Base, Oklahoma
City, OK. USATHAMA Report No. CETHA-TE-CR-90026. March 1990.
4-42
-------
5.0 SOIL VAPOR EXTRACTION
5.1 Process Description
Soil vapor extraction (SVE) is one method used for the treatment of soil
contaminated with volatile hydrocarbons. The process is sometimes also referred to as
soil venting, vacuum extraction, aeration, or in-situ volatilization. In general terms, soil
vapor extraction removes volatile organic constituents from contaminated soil by creating
sufficient subsurface air flow to strip contaminants from the vadose (unsaturated) zone
by volatilization. Complete removal may not be possible unless the source of vapors
(e.g. hydrocarbon lens on groundwater) is also removed. As the contaminant vapors are
removed, they may be vented directly to the atmosphere or controlled in a number of
ways. Soil vapor extraction has been widely used to remediate sites contaminated with
gasoline or chlorinated solvents (e.g. TCE). It is also sometimes used to minimize
migration of vapors into structures or residential areas during other types of remediation.
By its nature, SVE is an on-site, in-situ treatment method. It is often used in conjunction
with or following other remedial measures such as excavation of subsurface waste bodies,
removal (pumping) of any hydrocarbon lens that is present, or air stripping of
contaminated ground water.
A number of reports and articles have recently been published that provide
useful information regarding SVE systems. The best single source of information is a
recent EPA report (Pedersen and Curtis, 1991). Much of the information in this section
was drawn from that report and a second EPA report (Thompson, et al, 1991). Other
key references are three studies that include summarized information about existing SVE
systems in use at field sites (Crow, et al, 1987; Hutzler, et al 1989; and PES, 1989), an
evaluation conducted under EPA's SITE program (Michaels, 1989), and a recent
overview paper (Johnson, et al., 1990). The Johnson, et al. paper is given as Appendix E
of this report.
5-1
-------
The relative advantages of SVE over other remediation approaches are:
1) The equipment is readily available and simple to install and
operate;
2) Large volumes of soil can be treated in a cost-effective manner;
3) Remediation can proceed in many cases without disrupting on-going
commercial activities at the site; and
4) Air emissions are released from a point source and thus can readily
be controlled.
The major disadvantages of SVE versus other remediation approaches are that:
1) The method is not applicable for saturated soils or soils with low
air-permeabilities;
2) The success of the method varies with the volatility (vapor pressure)
of the contaminants present; and
3) Significant residual contamination may remain in the soil after
treatment under some remediation scenarios.
The success of SVE for a given application depends on numerous factors
with the three key criteria being: 1) the nature of the contamination; 2) the behavior of
subsurface vapor flow at the site; and 3) regulatory requirements. Each of these criteria
is described below. A simplified decision guide for judging the applicability of SVE is
shown in Figure 5-1.
Spills or leaks of fuels typically involve liquids containing dozens of
different constituents. For removal by SVE to be effective, the contaminants generally
must have vapor pressures greater than 1.0 mm Hg at 20°F. The tendency of the organic
contaminants to partition into water or to be adsorbed onto soil particles also affects
SVE effectiveness, so the compound's water solubility, Henry's Law constant, and soil
sorption coefficient are of interest. The soil temperature affects each of these variables
and hence the rate of vapor diffusion and transport. The concentration of contaminants
that are initially present will also affect their relative partitioning between vapor and
liquid phases, and the amount that is solubilized or adsorbed. The time that the
5-2
-------
VAPOR
PRESSURE
Butane'
Pentane
Benzene
Toluene
Xylene
Phenol
Naphthalene
AJdicarb
-10'
—10'
K10
—10
-2
-10
—10
SVE
LJKEUHOOD
OF
SUCCESS
SOIL AIR
PERMEABILITY
SUCCESS
VERY
LIKELY
HIGH
(gravel,
coarse
sand)
SUCCESS
SOMEWHAT
LIKELY
SUCCESS
LESS
LIKELY
MEDIUM
(fine sand)
LOW
(clay)
Match Point
TIME
SINCE
RELEASE
Weeks
Months
Years
Months
Years
Weeks
Months
Years
Figure 5-1. Simplified Guide to Applicability of Soil Vapor Extraction.
Source: (Pedersen and Curtis, 1991)
5-3
-------
contamination has been present is also an important factor, as mixtures of contaminants
will generally become depleted of their more volatile components over time through
volatilization. This process, referred to as weathering, will tend to cause SVE to become
progressively less applicable as the site ages. It also affects the operation of the SVE
system, as the more volatile components are typically removed first and the composition
of the vapors collected and treated varies over time.
Although SVE may be used in a variety of soil types, the effectiveness will
depend on the ability of air to flow through the soil. The ability of vapors to flow
through a porous media such as soil is usually defined as the air permeability. Any
factors that influence the air permeability of the soil, such as soil porosity, grain size,
moisture content, depth to ground water, and stratification must be taken into
consideration when planning this type of remediation. The presence of cracks,
inadequately grouted boreholes, or other subsurface conduits will alter the subsurface
flow patterns. Soil vapor extraction may not be practical for sites where the source of
vapors is deep underground (e.g. > 100 feet), in areas with shallow groundwater tables
***•
(e.g. < 10 feet) or at sites where the groundwater level fluccuates greatly over time.
The types of contaminants present and the clean-up criteria will affect the
cost-effectiveness of SVE versus other remediation options. The final cleanup level for
contaminants in the soil will dictate whether or not SVE is a viable option. Very
stringent cleanup levels may dictate excavation of the soil and further on-site treatment.
Air emission regulations may require the use of controls to reduce the level of
contaminants in the exhaust gas. The cost of such emission controls may influence the
overall selection of a remediation approach.
5-4
-------
A few potential problems may arise in implementation of soil vapor
extraction, but effective solutions to most problems exist. When there is some concern
that contaminant vapors from a nearby site may be drawn in by the vacuum, air inlet
wells may be placed around the perimeter of the site to limit remediation to the site
under treatment. To avoid channelized flow, butterfly or ball valves may be placed on
the extraction wells so that they may be shut down if necessary. If contaminated water is
extracted in the process, a liquid phase treatment system is usually installed.
Evidence of unusually high carbon dioxide levels indicates that some sites
may experience enhanced subsurface biodegradation that may be partially or wholly a
result of soil vapor extraction. Subsurface air flow may promote growth of aerobic
hydrocarbon degraders which feed on the organic contaminants by improving the level of
available oxygen for the microbes. At one site, carbon dioxide concentrations in the soil
gas were 8.5%; much higher than the 0.03% or 0.04% typically present in the
atmosphere. Rough calculations indicated that up to 40% of the gasoline was destroyed
by degradation. Other sources of organic material such as co-disposed municipal waste
were not considered and may have been partially responsible for the high carbon dioxide
levels.
Figure 5-2 shows a generalized process flow diagram for the soil vapor
extraction process. Typical systems include extraction wells, monitoring wells, air inlet
wells, vacuum pumps, vapor treatment devices, vapor/liquid separators and liquid phase
treatment devices. Wells are generally 4 to 8 inches in diameter. An option sometimes
employed is to introduce the air at the air inlet well into the saturated zone (i.e.
groundwater table). This technique, referred to as air sparging, acts to strip some of the
volatile and semi-volatile compounds from the ground water. Another option is to heat
the air entering the inlet wells to enhance the volatilization of less volatile, higher
molecular weight contaminants, such as diesel fuel.
5-5
-------
Clean Flue Gas
CLEAN
WATER
— ^
"
LIQUID
TREATMENT
VadoseZooe
Water Table
Figure 5-2. Generalized Process Flow Diagram for Soil Vapor Extraction.
5-6
-------
5.2 Identification of Air Emission Points
The air emissions associated with soil vapor extraction systems come
primarily from the stack. Stack heights are typically 12-30 feet and usually only one
stack is used. Additional releases of volatile organics may occur from the treatment of
any contaminated water that is extracted. Fugitive emissions are considered negligible
due to the negative pressure throughout most of the system.
S3 Typical Air Emission Species of Concern
Emissions include untreated volatile organics from the extraction process.
Removal and emissions of semi-volatile organic compounds will also occur, though with
less efficiency than for VOCs. Lesser amounts of air emissions associated with the
control system may also occur. Due to the variety of technologies used, stack emissions
may include products of incomplete combustion, nitrogen oxides, paniculate matter,
carbon monoxide, acid gases and any other possible products of these technologies. Of
primary concern, however, are the volatile organics emitted from the point sources.
5.4 Summary Of Air Emissions Data
Air emissions data for several SVE systems are summarized in Table 5-1.
The data are from a variety of soil vapor extraction systems. Overall there is little
detailed published information about SVE systems, making it difficult to assess the
representativeness of this sample.
The emission rate of VOC compounds over time from continuously
operated SVE systems tends to show an exponential-type decay curve. If the system is
stopped and then restarted, however, the VOC emission rate returns to near the original
rate unless the remediation is nearing completion. Apparently, shutting off the vacuum
allows the soil-gas equilibrium to become re-established. Due to this behavior, the most
5-7
-------
Table 5-1.
Summary of Emissions Data for SVE System
Source
Crow, eLal.
(1987)
Hutzler, et.al.
(1989)
PES (1989)
No. of
Systems
Surveyed
13
19
17
Parameter
Flowrate per well
Removal
Exhaust Gas
Concentration
Total Flowrate
Treatment:
- None
- Carbon
• Catalytic Incineration
- Combustion
Removal Rate
Total Flowrate
Pollutant Concentration
Control Efficiency
Units
cfm
Ib/day
ppmv
cfm
# systems
Ib/day
cfm
ppmv
%
Range or Value
53-300
2-250
20-350
3 - 5,700
9
6
1
1
4-430
25-11300
150-38,000
90-99
Approximate
Average
80
60
100
800
100
2,200
4,000
95
5-8
-------
efficient method of operation is to run the SVE system only for a part of each day or
week, i.e. operate in a "pulsed" mode.
Published emission factors for SVE systems based on typical operating
conditions are (Thompson, et al, 1991):
Uncontrolled Emissions: 25,000 g/hr
250 kg/day (based on 10 hours of operation)
Controlled Emissions: 1,250 g/hr
0.05 g/g VOC in soil
5.5 Identification of Applicable Control Technologies
As the vapors are removed from the soil they are either discharged to the
atmosphere or treated to reduce air emissions. If the hydrocarbon content is high
enough, direct combustion is theoretically possible. However, because concentrations
typically drop significantly during removal, natural gas or some other fuel will be needed
to maintain combustion. Also, for safety reasons, dilution air is typically added to
maintain the VOC concentration below the lower explosive limit (LEL). In some cases,
the wells may be shut down for a period of time to allow subsurface vapor pressures to
re-equilibrate, thus yielding concentrations sufficient to sustain a flame. For lower levels
of hydrocarbons, catalytic oxidation may be effective. Carbon adsorption systems are
often used but they may be costly to implement and are generally not acceptable for
high-humidity gas streams.
A recent survey indicates that the exhaust from about 50% of SVE systems
is vented directly to the atmosphere with no controls (PES, 1989). The trend, however,
is for VOC controls to be required. For those systems with controls, the most viable
options are:
1) activated carbon adsorption;
2) catalytic oxidation;
3) thermal incineration;
4) internal combustion engine; and
5) miscellaneous control approaches.
5-9
-------
The first three treatment options are the most commonly used for large SVE systems
such as those used at Superfund sites or refineries. Internal combustion engines (ICE)
are a common choice for control of emissions for small systems such as those used at
Leaking Underground Storage Tank (LUST) sites. Removal efficiencies of 95-99% for
VOCs should be theoretically achievable with any of these control options.
No single control method is preferred. Each has advantages and
disadvantages that must be considered for each specific application. Control options are
discussed below. Further information is-available from EPA's Center for Environmental
Research Information (CERI) (Eklund, et al., 1992).
5.5.1 Carbon Adsorption
Carbon adsorption using activated granular carbon (GAC) is the most
common control method for SVE systems. VOCs are removed by being physically
trapped on the surface of the GAC or by chemical reactions with the carbonr- Tie
efficiency of GAC is due to its very large surface area per unit mass. Two options for
GAC systems are available: 1) "throw away" systems, and 2) fixed bed regenerable
systems. In the first option, canisters of GAC are used and disposed of or reactivated
off-site. In regenerable systems, steam or hot air is used to strip contaminants from the
GAC in place. The contaminants are recovered as a liquid. The cost-effectiveness of
regenerable systems will increase as the treatment time and the mass of contaminants to
be treated increase.
Modular, skid-mounted treatment systems are available from numerous
vendors. Prefabricated GAC units containing up to a ton of carbon are available.
Flowrates over 1,000 scfrn can be accommodated.
5-10
-------
The primary advantage of carbon adsorption over other control options is
that the control efficiency of GAC systems is not significantly affected by the changes in
air flowrate and VOC concentration that typically occur at SVE sites. It is applicable to
most contaminants having molecular weights between 50 and 150; lighter compounds
tend to pass through the GAC unabsorbed and heavier compounds tend to bind
permanently to the carbon and cannot be desorbed. GAC tends to be the control
method of choice for SVE systems with low VOC concentrations in the exhaust gas (e.g.
less than 500-1000 ppmv).
Carbon adsorption has several limitations that may be significant for SVE
applications. One, water vapor will occupy adsorption sites and reduce the removal
capacity. It is usually recommended that the gas to be treated have a relative humidity
of less than 50% for GAC to be effective. Two, carbon tends to not retain organics at
temperatures exceeding 150T. This temperature is well below the temperatures of 200
to SOOT in the exhaust gas that can be caused by compression of air caused by the
removal pump. The air can be cooled or pumps used that don't add much heat to the
system (e.g., liquid ring seal pumps). Three, high mass loadings of VOCs in the exhaust
gas will cause the carbon to quickly be exhausted and result in high costs to replace or
regenerate the carbon. The first two limitations can be off-set through modifications to
the system design, but these modifications will increase the cost of remediation.
5.5.2 Thermal Incineration
Thermal incineration can be used to destroy vapor-phase contaminants.
Contaminant-laden vapors are heated to temperatures above 1000°F via a direct flame or
a combustion chamber. The method is applicable to a wide range of compounds and
over a large range of concentrations. It is not, however, widely used for SVE
applications except for large-scale, long-term clean-ups. For the flame to be self-
sustaining, the VOC concentration needs to be at percent levels that may be above the
lower explosive limit for the contaminant of concern. For lower VOC levels, auxiliary
fuel such as methane or propane must be added. The cost of this fuel can be
5-11
-------
prohibitive. The efficiency of the method is also affected by changes in the flowrate. As
the flowrate varies from design conditions, the mixing and residence times in the
incinerator will vary and decrease the destruction efficiency.
5.5 J Catalytic Oxidation
Catalytic oxidation, also called catalytic incineration, is similar in design
and operation to thermal incineration except that a catalyst is present that enhances
combustion. The catalyst is usually palladium or platinum in a metallic mesh, ceramic
honeycomb, or catalyst-impregnated beads in a packed bed. The catalyst allows
destruction to occur at lower temperatures than for thermal incineration (600-900°F).
There is therefore less auxiliary fuel required and commensurate lower fuel costs.
Design efficiencies of 95 to 99 percent are typical. The catalyst can be
damaged by overheating, so the air stream must be diluted, if necessary, to a VOC
concentration below about 3000 ppmv, to maintain acceptable operating temperatures.
Maintenance of this VOC level raises the capital and operating cost of the system since
accurate monitoring of the gas stream is needed as is the ability to control the dilution of
the gas stream. As for thermal incinerators, catalytic oxidation systems function best
when the flowrate is constant.
The catalyst will become less effective over time and can be adversely
impacted by trace contaminants in the gas stream. Depending on the type of catalyst
employed, it can be damaged by chlorinated hydrocarbons, mercury, phosphorus, or
heavy metals.
5.5.4 Internal Combustion Engines (1C)
Industrial or automotive engines have been widely used to control VOC
emissions from SVE systems. Depending on the engine size, air flows of 30 to 100 scfm
have been treated. The effective flowrate is reduced, however, if ambient air must be
5-12
-------
added to the air stream to add sufficient oxygen to support combustion. As with other
thermal treatment methods, supplemental fuel is needed.
Destruction efficiencies of 99+% have been reported for the most common
components of gasoline (Pedersen and Curtis, 1991). Advantages of 1C engines as
controls are that the systems are portable, they can handle very concentrated air streams
without the need for dilution, and the engine can provide power to operate the SVE
system. Disadvantages are that the systems can only treat small flowrates and that
manual supervision is required for a period during start-up to set the flowrates and
operating conditions. Emissions of nitrogen oxides from the engine may be a concern in
some locales.
5.5.5 Miscellaneous Control Approaches
A number of additional control devices may potentially be applicable for
controlling VOC emissions from SVE systems, including condensers, packed bed thermal
processors, and biofilters. Condensers using chilled water or other refrigerants can
remove anywhere from 50 to 90% of VOCs from concentrated streams (>5000 ppmv
VOCs). Packed bed thermal processors consist of a bed of ceramic beads heated to
1800T that is used to destroy organics and chlorinated hydrocarbons. Biofilters consist
of soil beds that trap VOCs in a manner analogous to GAC and then are regenerated by
biological action. Biotreatment requires time to establish an active culture of microbes
and careful control of soil moisture, temperature, and air flow patterns to maintain the
efficiency of the microbial action.
5.6 Costs For Remediation
The costs to install and operate an SVE system will, of course, vary from
site to site. A typical cost to install and start-up a small system is less than $100,000
(Newton 1990). Total capital costs for equipment range from $65,000 to $135,000
excluding the cost of each vapor-recovery well (Cochran, 1987). Typical capital costs for
5-13
-------
the major components of the system (Pedersen and Curtis, 1991) are $2,000-4,000 per
well, $10,000 or more for a large (25hp) blower/fan, $2,500 for an air/water separator,
and $10,000 for a structure to house the system. A major variable is the cost of any
monitoring and control system needed to maintain the VOC level in the exhaust stream
within preset limits.
Typical operation and maintenance costs are $6,000 to $26,000 per year
(Cochran, 1987). The major operating costs (Pedersen and Curtis, 1991) are for power,
VOC controls, monitoring, and labor. Power costs for a lOhp system are estimated to be
about $20 per day. VOC control costs are discussed in the next subsection. Monitoring
and labor costs are highly variable. In general, operating costs in these areas can be
minimized through the use of automated monitoring and control equipment. The
optimal split for this trade-off between capital and operating costs will depend on the
duration of the remediation and the proximity of the site to the labor source.
5.7 Costs For Emission Controls
Equations for predicting the costs of emission controls based on system
design parameters are available (PES, 1989). Typical costs for various types and sizes of
treatment systems are given in Table 5-2. The cost estimates are drawn from a number
of vendors and therefore a range is shown in most cases. The costs from different
vendors may not be directly comparable since the cost basis may vary. For example,
regenerable carbon adsorption systems cost from $22,000 to $55,000 for manually
regenerated systems and $165,000 for a fully-automated equivalent system. The cost data
are intended to show the general level of costs likely to be incurred for various types of
control options.
5-14
-------
Table 5-2.
Summary of Capital Costs to Control VOC Emissions From SVE Systems
Treatment
Carbon Adsorption
(Regenerable)
Thermal Incineration
Carbon Canisters
Catalytic Oxidation
Internal Combustion Engine
Maximum How (scfm)
105
250
500-600
1100
100
570
100
500
1000
4000
100
200
500
1000
5000
45
100
Capital Cost ($)
20,000
24,000
9,000 - 33,000
12,000
25,000
44,000
700
8,000
6,000
23,000
25,000
31,000 - 69,000
44,000 - 86,000
77,000 - 94,000
220,000
+ 20,000 for dilution system
62,000
50,000
Source: Adapted from Pederson and Curtis, 1991.
5-15
-------
Maintenance costs will vary depending on the type of system and may
include power, fuel, activated carbon, and the associated labor. The costs will vary with
the size of the system and the operating rates. Electricity to run the blower motor are
about $600 per month for a 10 hp blower. Fuel costs for thermal incineration and
catalytic oxidation depend on the VOC concentration of the influent air. Typical costs
are $500 to $1,000 per month. Auxiliary fuel costs for 1C engines are also about $500 to
$1,000 per month. Activated carbon will cost from $1 to $2 per pound. Typical carbon
costs are about $25 per pound of hydrocarbons removed (about $160 per gallon).
5.8 Equations and Models For Estimating VOC Emissions
The factors that govern vapor transport in the subsurface are very complex
and no practical, accurate theoretical models for predicting emissions or recovery rates
for SVE systems exist. During operation of SVE systems, the vacuum that is applied to
the soil and the resulting pressure gradient is the dominant factor in determining the
flow rate of vapors. The induced vacuum in the soil decreases with distance from the
extraction well and a radius of influence exists that defines the extent to which vapors
can be drawn to the well. The length of this radius depends on the strength of the
vacuum source; soil properties such as porosity, permeability, and moisture content; and
site properties such as surface coverings. In practice, field tests are typically performed
to evaluate the potential effectiveness of SVE for a given site. The field tests may be
either a pilot-scale demonstrations of SVE or tests of the air permeability. This
information is used to determine the number of wells required to remediate the site and
the spacing of the wells.
Subsurface vapor flow equations based on Darcy's Law have been
published that predict the flow rate of vented gas (Johnson, et al, 1990). The key inputs
are the air permeability of the soil, the air-filled porosity of the soil, the thickness of the
soil layer, the density of the vapor, and the gradients of pressure and vapor
concentration. Methods for measuring the air permeability are based on measuring the
5-16
-------
difference between the ambient atmospheric pressure and the air pressure in the soil
during vapor transport. Johnson, et al (see Appendix E) describe a test where air is
withdrawn from a well at a constant flow rate while the draw-down (vacuum) pressure is
measured in a monitoring well some distance away. The pressure is plotted versus the
log of time and the slope of this equation is the aif permeability. Measurements at
several monitoring well locations are needed. Similar test methods may also be
employed such as air injection tests and oil field tests such as pressure buildup and draw-
down tests (Pedersen and Curtis, 1991).
Using data from pilot-scale tests at the site, air emissions can be estimated
with the following mass balance equation (Ekhmd, et al, 1992):
/ o \ (Eq- 5-1}
ER = (Cg) ^J (ID'6)
where: ER = Emission rate (g/sec);
Cg = Concentration in extracted vapors (^g/m3);
Q = Vapor extraction rate (m3/min);
1/60 = Conversion factor (min/sec); and
10"6 = Conversion factor (g/«g).
The extraction rate, Q, can be estimated from the results of pilot-scale tests at the site if
any changes in pump size and number of wells between the pilot- and full-scale systems
are taken into account. If no pilot-scale data are available, results of field test of soil-air
permeability can be used to estimate Q. If these too are not available, a default value
can be used for the extraction rate. Typical flow rates for Q at Superfund sites range
from 14 m3/min (500 cfrn) to 425 m3/min (15,000 cfm), with a typical default value being
Q = 85 m3/min (3,000 cfm).
The contaminant concentration in the extracted vapors, Cp can also be
estimated from the results of pilot-scale tests at the site. The second best approach is to
estimate Cf by collecting samples of the headspace vapors above the contaminated soil
arid measuring the concentration of the compound(s) of interest. These equilibrium soil-
gas samples can be collected using ground (soil-gas) probes or by transferring soil
5-17
-------
samples from split-spoon samplers (to minimize VOC losses) to sealed containers and
allowing the headspace to equilibrate.
Field data are required to get an accurate value for Cr If no field data are
available, however, a very conservative value for Cg can be estimated by assuming that
the soil-gas is saturated. The maximum vapor concentration of any compound in the
extracted vapors is its equilibrium or "saturated" vapor concentration, which is calculated
from the compound's molecular weight, vapor pressure at the soil temperature, and the
ideal gas law:
(Eq. 5-2)
«R)CD)
C - CPvap)(MW * 109)
"
where: Cg = Estimate of contaminant vapor concentration (ag/m3);
Pvap = Pure component vapor pressure at the soil temperature (mm Hg);
MW = Molecular weight of component i (g/mole);
R = Gas constant = 62.4 L-mm Hg/mole -°K;
T = Absolute temperature of soil (°K); and
109 = Conversion factor (^g-L/g-m3).
Values of molecular weight, vapor pressure at 25°C, and saturated vapor concentration at
25°C are given in Appendix A. It is important to note that Equation 5-2 gives the
theoretical maximum value of Cr It will overpredict Cg for any compounds present in
the soil at relatively low concentrations. Equation 5-2 will also overpredict the long-term
average value of Cg since the concentration of contaminants in the gas extracted using a
SVE system will tend to drop over time. It can drop by more than 95% in the first two
days of operation, though pulsed operation will allow the soil-gas concentration to be
periodically re-established at levels near the initial concentration.
Equation 5-2 assumes that an infinite source of vapors exists and that the
contaminants are present in the soil or ground water at relatively high concentrations
(e.g., total hydrocarbons of 500 ppm in the soil). Therefore, the vapor-phase
concentration for a given compound is independent of the concentration of that same
compound in the soil/liquid matrix.
5-18
-------
Removal rates can be 500-600 kg/day or higher and control efficiencies
(when applicable) range from 60-99%. As previously mentioned, only about half of the
sites listed used any VOC control equipment.
5.9 Case Study
Process Description
Terra Vac Incorporated has developed a vacuum extraction system
designed to remove volatile organic contaminants from the vadose zone. At the
Groveland, Massachusetts Superfund site, the contaminated air stream was treated with
two sets of activated carbon canisters (Michaels, 1989 and Stinson, 1989). Due to
weather conditions, liquid water was extracted as well, so a vapor-liquid separator was
included to remove contaminated water to a holding tank. The process design is shown
in Figure 5-3. No air inlet wells were used at this site.
Characterization of Air Emissions
Because no biodegradation was taking place in this case, the compounds
released to the atmosphere were the same as those found in the soil. By far the most
predominant contaminant removed was trichloroethylene, although 1,1,1-trichloroethane,
trans-l,2-dichloroethylene, and tetrachloroethylene were also extracted.
Emission Factors
Table 5-3 shows emissions factors for each of the four contaminants. The
estimated total VOC peak emission factor is 9.57 g/hr. Based on the field data, the
carbon adsorption control device had an efficiency of better than 99%. Because the
contaminated water was not treated on-site, evaporative emissions from the holding tank
are included in this estimate along with stack emissions from vapor treatment. The total
emissions represent less than 1% of the amount of contaminants removed from the soil.
The removal efficiency for the total mass of contaminants present at the site was not
demonstrated, nor was the associated control efficiency.
5-19
-------
Clean Flue Gas
CARBON CANISTERS
TANK
TRUCK
VAPOR-LIQUID
SEPARATOR
Figure 5-3. Process Flow Diagram for Terra Vac In Situ Vacuum Extraction System.
5-20
-------
Table 5-3.
Estimated Emissions for Terra-Vac's In-Situ Vacuum Extraction System
Pollutant
TCE
DCE
TRI
PCE
Molecular
Weight
g/mol
13129
96.94
133.41
165.83
Totals
Peak
Uncontrolled
Stack
Emissions
g/hr"
5,210.0
274.0
36.0
18.0
5,538.0
Peak
Controlled
• Stack
Emissions
g/hrb
5.91
0.31
0.04
0.02
6.29
Evaporative
Emissions
g/hr<
3.29
0.0
0.0
0.0
3.29
Total
Emissions
g/hr"
920
031
0.04
0.02
9.57
* Uncontrolled emissions equal removal rate of each contaminant.
b Based on measured 99.75% overall control efficiency for two carbon adsorption
canisters in series.
c Estimated evaporative emissions from contaminated water storage.
d Total stack and evaporative emissions.
KEY:
TCE - trichloroethylene
DCE - trans-l,2-dichloroethylene
TRI - 1,1,1-trichloroethane
PCE - tetrachloroethylene
5-21
-------
Costs
The equipment fabrication and construction costs were estimated to be
$55,000 (in 1991 dollars). The total cost to remove 6,000 tons of VOCs from the site
was estimated to be $310,000 or $52 per ton. Of this, costs for activated carbon were
$14 per ton and for waste disposal were $8 per ton. Including power and labor costs, the
VOC control system represents about one-half of the total remediation cost.
5.10 References
Cochran, R. Underground Storage Tank Corrective Action Technologies.
EPA/625/6-87-015 (NTIS PB87-171278). January 1987.
Crow, W.L. Personal communication. 1987.
Eklund, B., et al. Estimation of Air Impacts for Soil Vapor Extraction.
EPA-450/1-92-001 (NTIS PB92-143676). January 1992.
Eklund, B. Personal communication. 1992.
Hutzler, NX, B.E. Murphy, and J.S. Gierke. State of Technology Review -
- Soil Vapor Extraction Systems. Report No. EPA-600/2-89/024 (NTIS
PB89-195184). U.S. EPA, Cincinnati, OH, June 1989.
Johnson, et al. A Practical Approach to the Design, Operation, and
Monitoring of In Situ Soil-Venting Systems. Ground Water Monitoring
Review. Spring 1990.
Michaels, P.A. Technology Evaluation Report: SITE Program
Demonstration Test Terra Vac In Situ Vacuum Extraction System
Groveland, Massachusetts, Volume I. Report No. EPA-540/5-89/003a
(NTIS PB89-192025). U.S. EPA, Cincinnati, OH, April 1989.
Newton, J. Remediation of Petroleum Contaminated Soils. Poll. Eng.. 22
(13) 46-52. December 1990.
Pedersen, T.A. and J.T. Curtis. Handbook of Soil Vapor Extraction
Technology. EPA/540/2-91/003 (NTIS PB91-168476). February 1991.
PES Corp. Soil Vapor Extraction VOC Control Technology Assessment.
EPA-450/4-89-017 (NTIS PB90-216995). September 1989.
5-22
-------
Stinson, M. Applications Analysis Report: Terra Vac In Situ Vacuum
Extraction System Groveland, Massachusetts, Volume I. Report No. EPA-
540/A5-89/003 (NTIS PB90-126665). U.S. EPA, Cincinnati, OH, July
1989.
Thompson, P., A. Inglis, and B. Eklund. Emission Factors for Superfund
Remediation Technologies. EPA-450/1-91-001 (NTIS PB91-190975).
March 1991.
5-23
-------
6.0 IN-SITU BIODEGRADATION
6.1 Process Description
In-situ biodegradation is the term for biological treatment processes that
are performed in place and therefore do not require excavation and removal of the
contaminated soil. This treatment method includes widely used technologies such as
land treatment as well as some emerging technologies that employ subsurface injection
of oxygen or nutrients to promote the biodegradation of contaminants.
The main purpose of in-situ treatment is to employ the natural
microbiological activity of soil to decompose organic constituents into carbon dioxide and
water. Systems that try to enhance this natural biological activity typically use injection
wells to provide an oxygen source (such as air, pure oxygen, or hydrogen peroxide) to
stimulate aerobic degradation or add nutrients to support the growth of waste-consuming
microorganisms. In some cases, microorganisms may be added to the soil that have the
ability to metabolize specific contaminants of interest.
It is important to note that in in-situ biotreatment, biodegradation is
actually only one of several competing mechanisms. In in-situ processes the
contaminants may also be leached, volatilized, undergo chemical degradation, or be
adsorbed onto the soil particles. The overall removal achieved by in-situ biotreatment
processes represents the combined impact of all of these mechanisms. In addition, field
studies (Eklund, et al., 1986) have shown that volatilization may account for the
disappearance of the majority of VOCs being treated.
6-1
-------
Since volatilization makes a potentially large contribution to the overall
removal achieved by most in-situ biotreatment processes, this technology is generally not
suitable for remediating sites which are contaminated with volatile fiiels or other
contaminants, or for remediating sites that are close to sensitive receptors. In-situ
biotreatment is best suited for sites in remote locations and sites that are contaminated
with less volatile fuels (such as JP-4, JP-5, or diesel fuel).
Like all biotreatment processes, in-situ treatment is not applicable for the
remediation of non-biodegradable contaminants such as heavy metals and other
inorganic compounds. Some halogenated organic wastes also are not amenable to
biotreatment. The applicability of advanced treatment methods, which rely on
subsurface injection of oxygen or nutrients, are also highly dependent on the type of soil
in which the contamination has occurred. Studies by the Air Force on the remediation
of soil contaminated with JP-4 and JP-5 showed that in soils with low-permeability there
was reduced delivery of oxygen and nutrients and consequently little biodegradation
occurred (Downey and Elliot, 1990).
The use of landtreatment as a remediation technology is expected to
decline due to recent regulatory requirements that RCRA landtreatment facilities have
no demonstrated migration of contaminants to the surrounding environment.
Figure 6-1 shows a general schematic of an in-situ biodegradation process.
As the figure shows injection wells may be installed around the zone of contamination to
provide oxygen and nutrients to stimulate biodegradation.
6-2
-------
Evaporative losses of volatile
and semivolatile compounds
Top 6"-8" of soil have
highest biological activity
Waste
(Contaminated Soil)
&&&&&&
>*f*f»f*f»f»f»f''r-w ~, -, -»
%«S*%»%*S>%*S*V*S«S>S>S*
•iftfififififififjfififif
>'.>lWlV.>l>?>?H?>tSl>\
j+s+s+t+tti+s+sm
!flfff!ftf}ffflft'
tfffffffff'fffffff'fffffffff'f
Waste will either be spread on surface
or mixed into the top 6"-8" of .oil
Figure 6-1. Generalized Process Flow Diagram for In-Situ Biodegradation.
6-3
-------
As suggested in Figure 6-1, the primary advantages of in-situ treatment are
simplicity and low-cost. The equipment and operating costs for this type of treatment
are very low compared to other technologies. The only heavy equipment that may be
required is a disk or rototiller to occasionally expose subsurface soil to the air. However,
if typical landtreatment techniques are used to treat the contaminated soil a large area
may be required for waste spreading.
In an in-situ biotreatment process, the time required to treat the
contaminated soil will vary greatly depending on a number of factors including the:
• Physical and chemical properties of the soil matrix;
• Physical and chemical properties of the contaminant;
• Initial concentration of the contaminant in the soil; and
* Biodegradability of the contaminants (i.e. biodegradation rate
constants).
Once the concentration of contaminants in the soil has been lowered either
by biodegradation or evaporation, the zone of contamination may be capped or covered
with a new soil layer. Field studies of landtreatment have found no evidence of
anaerobic decomposition (Eklund, et al, 1986).
62 Identification of Air Emission Points
As Figure 6-1 shows, fugitive air emissions from the treated area are the
primary waste stream generated in in-situ biodegradation processes. A more localized
emission source will exist when and where the waste is applied to the soil (or the
contaminated soil pile is being spread). A small amount of additional emissions may
also be generated by heavy equipment used to till or move the soil.
6-4
-------
Typical Air Emission Species of Concern
Typical emissions from in-situ biotreatment process are evaporative losses
of volatile and semi-volatile organic compounds. If the soil is tilled, though, there also
may be some small amounts of paticulate matter (PM) emissions, as well as some
combustion emissions if the tilling device is powered by a gas or diesel engine.
The primary environmental factors, in addition to the biodegradability and
volatility of the soil contaminants, which influence air emissions are wind speed and
temperature. At higher windspeeds the driving force for mass transfer into the gas-phase
increases so evaporative emissions tend to be higher. The microorganisms tend to
perform best within a narrow temperature range. Deviations from these optimal
temperatures (either high or low) will tend to diminish microbial activity and result in
slower biodegradation. As a result, competing mechanisms such as volatilization may
predominate.
6.4 Summary of Air Emissions Data
Since in-situ biodegradation has not been used in the clean-up of a
significant number of fuel spills, there are few data available in the literature on air
emissions for this application of in-situ remediation. However, air emissions associated
with landtreatment refinery and industrial wastes have been more extensively studied.
Table 6-1 shows the fluxes of volatile organic compound emissions from
two plots at a refinery landtreatment facility (Eklund, et al, 1986). The data represent
the total emissions per 12 hour period for test plots that were each 423 m2. As the data
show, air emissions from in-situ treatment initially high and decrease as the
concentration of volatiles in the soil falls off. Tilling of the soil increased the measured
emission fluxes by at least a factor of three. The annual waste loading at this site was
6-5
-------
Table 6-1.
VOC Emissions From a Refinery Landtreatment Facility:
Average Measured Emission Rates by Plot by Half-Day
Sampling Day
0
1
2
3
4
5
6
7
•
9
10
11
12
Emission Rate (kg)
A-PIot (Surface)
1.88
249
142
22.5
32.6
482
146
33.4
463
16.6
15
8.04
49.1
6.41
135
21.8
71.1
2.56
725
124
1234
2.49
6.4
2.97
103
B-Plot (Background)
1.48
24.7
10
4.09
5.11
4.09
9.93
8.18
8.86
2.71
6.13
2.73
10
2.04
5.8
1.32
2.6
12
1.4
1.4
12
4.4
C-Plot (Sub-surface)
1.95
176
52.8
13.8
26.5
120
295
712
67.4
12.4
15.1
10.1
913
12.1
22
13
593
2.93
5.76
3.94
162
113
7.13
4.06
162
6-6
-------
about 16 L/m2 of separator float and sludge. Although refinery and industrial wastes
may differ significantly from the type of contamination present at a fuel spill, these data
may provide an order of magnitude estimate for the emissions that could be expected if
in-situ bioremediation were used to treat a fuel spill
Another field test (Ricardelli cited in U.S. EPA, 1989) demonstrated that
approximately 25% of raw oily refinery waste applied to a landtreatment facility was lost
to the atmosphere. The author found that compounds with boiling points less than 400°F
tended to be volatilized while compounds with boiling points greater than 400°F tended
to remain in the soil and be degraded. A third field test (Dupont and Reineman, 1986)
measured fluxes of BTXE compounds of up to 0.01 ug/cm2-sec immediately after
application. The measured emission fluxes tended to drop by several orders of
magnitude over the first two days after application. Tilling increased the emissions by
approximately a factor of two.
Published emission factors for landtreatment are (Thompson, et al, 1991):
Uncontrolled Emissions:
1,500 g/hr (24-hr average)
188 g/hr (20-day average)
Identification of Aplicable Control Technoloies
In in-situ bioremediation the area over which air emissions are generated is
often quite large. Consequently, it is impossible to apply traditional "end-of-pipe" VOC
control technologies to this type of process. General approaches for controlling VOC
emissions from area sources such as emission-controlling foams and impermeable covers
are discussed in Section 8 on excavation. These approaches, however, are not applicable
to in-situ bioremediation since the controls are designed to inhibit the transfer of gases
between the soil and the atmosphere. While this reduces VOC emissions, it will also
limit the. replenishment of oxygen to the soil and may cause anaerobic conditions to
6-7
-------
develop. Other control approaches such as wind barriers tend not to be applicable for
in-situ biotreatment because of the large treatment areas.
One VOC reduction/control approach that is specific for in-situ
biotreatment is subsurface injection of waste. It is applicable for liquid wastes, but not
for contaminated soil. In this operating practice, a tank truck containing the liquid waste
is driven across the treatment area and the waste is introduced into the soil at a depth of
6-12 inches below groundlevel via a series of hollow tines on a tilling device pulled
behind the truck. Some researchers have claimed that sub-surface injection can reduce
emissions 80% or more over a 24-hour average (Coover, 1990). However, other
researchers contend that over a longer period subsurface injection provides only a minor
reduction in emissions, especially if the soil if the soil is tilled frequently for aeration
(Wetherold, et al, 1985).
6.6 Costs for Remediation
Costs to use traditional land-farming techniques to treat fuel contaminated
soil are very low. The only capital requirements are excavating and tilling equipment.
Operating requirements are also minimal; only fuel for the heavy equipment and labor
may be required to complete the remediation. The total costs for remediating a site are
likely to be on the order of $100,000, excluding the cost of any land which much be
purchased (Newton, 1990). Assuming 10,000 tons of soil, this would be less than
$100/ton.
If some type of advanced in-situ remediation process is applied, e.g. one
using oxygen or nutrient addition, clean-up costs may be significantly higher. Table 6-2
(U.S. EPA, 1987) shows estimated site cleanup costs for hypothetical sites at which
hydrogen peroxide is used for the enhancement of in-situ biodegradation. The cleanup
of 300 gallons of gasoline from a sand/gravel aquifer (Site A) over a period of 6 to 9
6-8
-------
Table 6-2.
Estimated Costs for Remediations Using Hydrogen Peroxide
to Enhance Biodegradation
Contaminant
Formation
Flow Rate
Project Time
Estimated Cost
Site A
300 gallons of gasoline
Sand/Gravel
50 gal/min
6-9 months
$92,000 - $157,000
SiteB
2,000 gallons diesel fuel
Bedrock
10 gal/min
9-12 months
$209,000 - $328,000
: .. :;;-.:.:;:vSite"C
10,000 gallons jet fuel
Fine Gravel
100 gal/min
14-18 months
$525,000 - $786,000
6-9
-------
months is estimated to cost $92,000 to $157,000. Cleanup of 2,000 gallons of diesel fuel
from a fractured bedrock formation (Site B) is estimated to require 9 to 12 months and
cost $209,000 to $328,000. The cost estimate for degrading 10,000 gallons of jet fuel
from a fine gravel formation is estimated to cost $525,000 to $786,000 and take 14 to 18
months.
6.7 Costs For Emissions Controls
The cost for using subsurface injection of the waste soil instead of the
traditional surface application is likely small, on the order of $10,000 to 20,000 for
application equipment and additional labor.
Equations and Models For Estimatin VOC Emissions
Vapor transport and biodegradation in contaminated soil are complex and
competing processes. A PC-based model called LAND7 or CHEMDAT7 (U.S. EPA,
1989) is currently recommended by the EPA to predict the VOC emission rates resulting
from the land treatment of wastes. It is a user-friendly program and a user's manual
accompanies the software1. The model predictions have been compared with available
measurement data and have shown a reasonable agreement with the measurement data
(U.S. EPA, 1989). The model is expected to be accurate to within an order of
magnitude in most cases. Sensitivity studies done using this model under typical
conditions has shown the model predicts that 35-80% of the applied volatiles will be
emitted to the air and the remainder degraded (Coover, 1989). Evaluations of the
Thibodeaux model (also applicable to the LAND7 model) have found it to require
significant amounts of input data about the site and the contamination that is present
(U.S. EPA, 1989). Values for the input parameters may need to be assumed in many
cases; however, default values are available as part of LAND7.
1Can be obtained by calling EPA's Control Technology Center Hotline at
919/541-0800.
6-10
-------
Case Study
As noted previously, traditional landtreatment techniques are not generally
recommended for the remediation of soils contaminated with volatile fuels. For such
sites, the air emissions from landtreatment are usually high enough to eliminate it from
consideration. Air emissions from advanced in-situ bioremediation processes, though,
tend to be lower since the biodegradation is faster, and these types of processes are
being increasingly used. However, limited full-scale data are available on the
effectiveness on these processes. A recent study by the U.S. Air Force indicates that.
while pumping air into contaminated soil does improve biodegradation it also increases
the rate of at which volatile compounds are stripped from the soil. As a result, the
offgas from this type of process will require some sort of treatment as shown in Figure 6-
2. In some applications the cost of treating the offgas may be substantial.
6.10 References
Coover, J.R. Air Emissions From Hazardous Waste Land Farms.
Presented at the Spring National AIChE Meeting, April 2-4, 1989.
Coover, J.R. Reducing Landfarm Emissions of Volatile Hydrocarbons.
Presented at the 83rd Annual AWMA Meeting (Paper 90-188.1),
Pittsburgh, PA. June 24-29, 1990.
Downey, D.C. and M.G. Elliot Performance of Selected In Situ Soil
Decontamination Technologies: An Air Force Perspective. Env. Progress
Vol. 9, No. 3, pp!69-173, August 1990.
Dupont, R.R. and J.A. Reineman. Evaluation of Volatilization of
Hazardous Constituents at Hazardous Waste Land Treatment Sites.
EPA/600/2-86/071 (NTIS PB86-233939). August 1986.
Eklund, B.M., T.P. Nelson, and R.G. Wetherold. "Field Assessment of Air
Emissions and Their Control at a Refinery Land Treatment Facility
Volumes I and II." EPA 600/2-86-086 a&b (NTIS PB88-124540 and
124557), October 1987.
Newton, J. Remediation of Petroleum Contaminated Soils. Poll. Eng.. 22
(13) 46-52. December 1990.
6-11
-------
Thompson, P., A. Inglis, and B. Ekhmd. Emission Factors For Superfund
Remediation Technologies. EPA-450/1-91-001 (NTIS PB91-190975).
March 1991.
U.S. EPA. Underground Storage Tank Corrective Action Technologies.
EPA/625/6-87-015 (NTIS PB87-171278). January 1987.
U.S. EPA. Hazardous Waste Treatment, Storage, and Disposal Facilities
(TSDF) - Air Emission Models. Report No. EPA-450/3-87-026 (NTIS
PB88-198619). U.S. EPA, Research Triangle Park, NC, December 1987.
Wetherold, R.G. et al. Assessment of Volatile Organic Emissions from a
Petroleum Refinery Landtreatment Site. In: Proceedings of the Hazardous
Materials Control Research Institute's 3rd National Conference on
Hazardous Wastes and Hazardous Materials, March 1986. EPA 600/D-86-
074 (NTIS PB86-184603).
Airwiih
stripped volatile:
loconnoi device
Goal of air injection is to supply additional oxygen
to microbes and speed bodegradation. However,
air injection wiB also strip votatites ton sol.
Figure 6-2. Flow Diagram for Off-Gas Treatment System For In-Situ Biodegradation.
6-12
-------
7.0 EX-SITU BIODEGRADATION
7.1 Process Description
Ex-situ biodegradation is the general term for treatment processes in which
an aqueous sluny is created by combining soil or sludge with water and then
biodegraded in a self-contained reactor or in a lined lagoon. This emerging technology,
which is now beginning to move out of the developmental stage, is also referred to as
slurry biodegradation.
There are two main objectives behind using slurry bioremediation: to
destroy the organic contaminants in the soil or sludge, and, equally important, to reduce
the volume of contaminated material. Slurry biodegradation can be the sole treatment
technology in a complete cleanup system, or it can be used in conjunction with other
biological, chemical and physical treatment.
Slurry biodegradation has been shown to be effective in treating highly
contaminated soils that have fuel or other organic contaminant concentrations ranging
from 2^00 nag/kg to 250,000 mg/kg. The slurry process has also shown potential for
treating a wide range of contaminants including pesticides, creosote, pentachlorophenol,
PCBs, and some halogenated organics.
The effectiveness of slurry biodegradation on some general contaminant
groups is shown in Table 7-1. This table is based on current available information and
engineering judgement (U.S. EPA, 1990).
Figure 7-1 shows a general schematic of the slurry biodegradation process.
However, slurry processes may vary significantly among vendors. Each vendor's process
may be capable of treating only certain types of contaminants. Treatability studies to
determine the biodegradability of the contaminants and the solids/liquid separation that
7-1
-------
Table 7-1
Applicability of Slurry Biodegradation for Treatment of
Contaminants in Soil, Sediments, and Sludges
Contaminant
Applicability
ORGANIC CONTAMINANTS:
Halogenated volatiles
Halogenated semivolatiles
Nonhalogenated volatiles
Nonhalogenated semivolatiles
PCBs
Pesticides
Dioxins/Furans
Organic Cyanides
Organic Corrosives
INORGANIC CONTAMINANTS:
Volatile metals
Nonvolatile metals
Asbestos
Radioactive materials
Inorganic corrosives
Inorganic cyanides
REACTIVE CONTAMINANTS:
Oxidizers
Reducers
1
2
1
2
1
2
0
1
0
0
0
0
0
0
1
0
0
KEY:
2 = Demonstrated Effectiveness; Successful treatability test at some scale has been
completed.
1 = Potential Effectiveness; Expert opinion is that the technology will work.
0 = No Expected Effectiveness; Expert Opinion is that the technology will not work.
7-2
-------
Process Vents to
Emission Control
Soil*
Water
lutrients/Additives ^
MDONGTANK
DEWATERING
Oxygen to
Aerators
Waste preparation may be required. This includes excavation as
well as other pretreatment to remove metals & other inorganics.
Figure 7-1. Slurry Biodegradation Process Flow Diagram
7-3
-------
occurs at the end of the process are typically necessary before final selection of ex-situ
biodegradation as a remedy for a given site.
As shown in Figure 7-1, waste preparation is a required first step in
applying slurry biodegradation to a contaminated site. The required preparation includes
excavation and handling of the waste material as well as screening to remove debris and
large objects. Other important waste preparation steps that may be required to meet
feed specifications include particle size reduction, water addition, and pH and
temperature adjustment. Table 7-2 shows the desired feed characteristics for a typical
slurry biodegradation process.
After appropriate pretreatment, the wastes are suspended in a slurry form
and mixed in a tank to maximize the contact between contaminants and microorganisms
capable of degrading those contaminants. From the mix tank the slurry is pumped (using
special slurry pumps) to the bioreactor system. The bioreactor system can either be an
above-ground continuously stirred tank reactor (CSTR) or a lined lagoon. Smc£jierobic
treatment is the most common mode of operation for slurry biodegradation, aeration
must be provided to the bioreactors by either floating or submerged aerators or by
compressors or spargers. Nutrients and neutralizing agents are also supplied to remove
any chemical limitations microbial activity. Other materials, such as surfactants and
dispersants may be used to improve the material's handling characteristics.
In the bioreactor, microorganisms may be added initially to seed the
reaction or added continuously to maintain the correct concentration of biomass. The
required residence time for the waste in the bioreactor will depend on a number of
factors including:
7-4
-------
Table 7-2
Desired Inlet Feed Characteristics for Slurry Biodegradation Processes
Characteristic
Organic Content
Solid Content
Water Content
Solids Particle Size
Feed Temperature
Feed pH
Desired Range
0.025 - 25 wt %
10 - 40 wt %
60 - 90 wt %
< 1/4 in. diameter
15 -35 deg C
4.5 - 8.8
7-5
-------
The physical and chemical properties of the soil or sludge matrix;
The physical and chemical properties of the contaminant, including
its concentration in the waste; and
The biodegradability of the contaminants.
A typical residence time may be as short as several (e.g. 10) days or as long as 8-9
weeks, depending on site conditions.
Once the biodegradation of the contaminants is completed, the treated
slurry is sent to a separation/dewatering system. A clarifier for gravity separation can be
used to remove the solid-phase from the soil.
Slurry bioreactors are generally transportable units which can be brought
on site by trailer. Typically, commercial units require a set-up area of OJ5-1 acre per
million gallons of reactor volume. Water needs at the site can be high since the waste
must be put in slurry form. Large quantities of wastewater may also have to be stored
on-site prior to discharge to allow time for analytical tests to verify that the discharge
standard for the site has been met.
Limited performance data on slurry biodegradation systems are currently
available. The data presented in this report are based on information supplied by
vendors. The validity of these results has not been evaluated.
Table 7-3 shows performance data for a full-scale slurry biodegradation
system, designed by Remediation Technologies, which was used to treat wood preserving
sludges at a site in Sweetwater, Tennessee (U.S. EPA, 1990). As the table shows the
system achieved an overall removal efficiency of greater than 95%. However, a
breakdown of the removal efficiency between biodegradation and volatilization is not
available.
7-6
-------
Table 7-3
Performance Results for Slurry Biodegradation Process
Treating Wood Preserving Wastes00
Compounds
Phenol
Pentachlorophenol
Naphthalene
Phenanthrene & Anthracene
Fluoranthene
Carbazole
Initial
Concentration
Solids
(mg/kg)
14.6
687
3,670
30,700
5,470
1,490
Slurry
(mg/kg)
1.4
64
343
2,870
511
139
Final
Concentration
Solids
(mg/kg)
0.7
12.3
23
200
67
4.9
Shiny
(mg/kg)
< 0.1
0.8
1.6
13.7
4.6
03
Permit
Removal^
Solids
(%)
95.2
98.2
99.3
99.3
98.8
99.7
Sluny
(%)
92.8
92.8
99.5
99.5
99.1
99.8
(a)
Treatment done using a 50,000 gallon reactor supplied by Remediation
Technologies.
Includes the combined effect of volatilization and biodegradation
7-7
-------
Another full-scale test of a slurry biodegradation system was recently
conducted by ECOVA Corporation (U.S. EPA, 1990). In this cleanup effort more than
750 yd3 of soil contaminated with pesticides was treated. Soil pesticide levels were
reduced from 800 mg/kg to less 20 mg/kg (>97.5% efficiency) in 13 days using a 26,000
gallon bioreactor. Residuals of the process were further treated by land application.
12 Identification of Air Emission Points
As Figure 7-1 indicates, there are three primary waste streams generated in
slurry biodegradation processes: the treated solids (sludge or soil), the process water,
and air emissions. The solids are dewatered and may be further treated if they still
contain organic contaminants. Also, if the solids are contaminated with inorganics or
heavy metals, they can be stabilized before disposal. Some portion of the process water
can be recycled, with the remainder treated in an on-site treatment system prior to
discharge.
The air emissions from slurry biodegradation processes can either be area
or point sources. For processes using open lagoons, emissions come from the exposed
surface of the lagoon. On the other hand in systems using above-ground self-contained
reactors, the primary source of emissions is usually a process vent.
13 Typical Air Emission Species of Concern
In bioslurry processes the emissions of concern are usually VOCs. The
soils handling steps required to deliver the contaminated soil to the treatment unit may
also emit significant amounts of VOCs and PM. Emissions from soils handling are
addressed elsewhere in this document.
In open lagoons, the primary environmental factors, in addition to the
biodegradability and volatility of the waste, which influence air emissions are process
7-8
-------
temperature and wind speed. Emissions tend to increase with an increase in surface
turbulence due to wind or mechanical agitation. Temperature affects emissions through
its influence on microbial growth. At temperatures outside the band for optimal
microbial activity, volatilization will increase (U.S. EPA, 1989a). Emissions from self-
contained reactors are also determined by reactor design parameters such as the amount
of air or oxygen used to aerate the slurry. Higher gas flow will strip more volatiles out
of solution and increase air emissions.
7.4 Summary of Air Emissions Data
Little information exists on volatile losses from slurry biodegradation
V
processes. Slurry processes have only recently become commercially available and field
experience to date is limited. However, data on air emissions from wastewaster
biotreatment processes are available. Table 7-4 shows fraction volatilized for different
compounds in an industrial aerated wastewaster treatment tank. Although, both the
mechanisms for volatilization and biodegradation will be different for the treatment of
contaminated soil, this data provides a first estimate of the contribution of volatilization
to overall removal for different types of compounds.
7.5 Air Emissions Controls
When the air emissions from slurry biodegradation processes are released
through a process vent, standard VOC air pollution control technologies can be applied.
Common alternatives for controlling VOC vent emissions include carbon adsorption as
well as thermal and catalytic oxidation. Since the vent stream will likely contain only
dilute amounts of VOCs, auxiliary fuel must be fired in either thermal or catalytic
oxidizers. When the air emissions from slurry biodegradation processes are area air
emission sources, applying air pollution control technologies is more difficult. Two
control options are commonly used. The first alternative is to use a vapor collection
7-9
-------
Table 7-4
Estimated Volatile Losses From Aerated Wastewater Treatment
Compound
Benzene
Toluene
p,m-Xylene
o-Xylene
n-Nonane
n-Decane
n-Undecane
Methylcyclohexane
Total Non-methane Hydrocarbons
Total Mass Loading of
Influent Wastewater
(g/min)
2.5
1.59
2.76
1.24
0.72
0.66
0.68
0.64
30.1
Estimated Volatile
Losses I
(% of Influent)
0.2
3.5
0.4
0.8
21
20
11
6.7
8.4
7-10
-------
hood to capture any VOC emissions and then rout those emissions to a standard control
device. A second, generally less favorable, alternative is to use an oil film or foam on
top of the slurry to suppress evaporative losses.
Control options for processes with point emission sources are discussed in
more detail below. For the relatively low VOC levels and low gas flows from
bioreactors, carbon-based VOC emission controls are generally the best choice.
7.5.1 Carbon Adsorption
Carbon adsorption using granular activated carbon (GAC) is the most
common control method for vent emissions from processes with dilute VOC containing
offgases. In carbon adsorbtion, VOCs are removed by being physically trapped on the
surface of the GAC or by chemical reactions with the carbon. The efficiency of GAC is
due to its very large surface area per unit mass. Two options for GAC systems are
available: 1) "throw away" systems, and 2) fixed bed regenerable systems. In the first
option, canisters of GAC are used and disposed of or reactivated off-site. In regenerable
systems, steam or hot air is used to strip contaminants from the GAC in place. The
contaminants are recovered as a liquid. The cost-effectiveness of regenerable systems
will increase as the treatment time and the mass of contaminants to be treated increase.
Modular, skid-mounted treatment systems are available from numerous
vendors. Prefabricated GAC units containing up to a ton of carbon are available.
Flowrates over 1,000 scfm can be accommodated.
The primary advantage of carbon adsorption over other control options is
that the control efficiency of GAC systems is not significantly affected by the changes in
air flowrate and VOC concentration. It is applicable to most contaminants having
molecular weights between 50 and 150; lighter compounds tend to pass through the GAC
7-11
-------
unabsorbed and heavier compounds tend to bind permanently to the carbon and cannot
be desorbed. GAC tends to be the control method of choice for processes with low
VOC concentrations in the exhaust gas (e.g. less than 500-1000 ppmv).
Carbon adsorption has one limitation that may be significant for bioslurry
applications. Water vapor will occupy adsorption sites and reduce the removal capacity.
It is usually recommended that the gas to be treated have a relative humidity of less than
50% for GAC to be effective. A simple offgas dryer may have to be placed in-line with
a GAC unit to be effective treating the emissions from a bioslurry process.
7.5.2 Thermal Incineration
Thermal incineration can be used to destroy vapor-phase contaminants.
Contaminant-laden vapors are heated to temperatures above 1000T via a direct flame or
a combustion chamber. The method is applicable to a wide range of compounds and
over a large range of concentrations. It is not, however, widely used in applications
similar to bioslurry processes. For the flame to be self-sustaining, the VOC
concentration needs to be at percent levels that may be above the lower explosive limit
for the contaminant of concern. For lower VOC levels, auxiliary fuel such as methane or
propane must be added. The cost of this fuel can be prohibitive. The efficiency of the
method is also affected by changes in the flowrate. As the flowrate varies from design
conditions, the mixing and residence times in the incinerator will vary and decrease the
destruction efficiency.
7.5.3 Catalytic Oxidation
Catalytic oxidation, also called catalytic incineration, is similar in design
and operation to thermal incineration except that a catalyst is present that enhances
combustion. The catalyst is usually palladium or platinum in a metallic mesh, ceramic
honeycomb, or catalyst-impregnated beads in a packed bed. The catalyst allows
7-12
-------
destruction to occur at lower temperatures than for thermal incineration (600-900T).
There is therefore less auxiliary fuel required and comensurate lower fuel costs.
Design efficiencies of 95 to 99 percent are typical. The catalyst can be
damaged by overheating, so the air stream must be diluted, if necessary, to keep the
VOC concentration below about 3000 ppmv. Maintenance of this VOC level raises the
capital and operating cost of the system since accurate monitoring of the gas stream is
needed as is the ability to control the dilution of the gas stream. As for thermal
incinerators, catalytic oxidation systems function best when the flowrate is constant.
The catalyst will become less effective over time and can be adversely
impacted by trace contaminants in the gas stream. Depending on the type of catalyst
employed chlorinated hydrocarbons, mercury, phosphorus, and heavy metals can damage
the catalyst.
7-6 Costs for Remediation
Presently there are little cost data available on slurry biodegradation
processes because of limited experience with this type of remediation. One vendor
estimates the cost of full-scale operation to be $110-210/m3 ($85-160/yd3) of soil,
depending on the initial contaminant concentration and the total amount of soil to be
treated. On a mass basis this cost estimate corresponds to $70-130/ton of contaminated
soil. The process cost will also vary depending on the need for additional pre- and post-
treatment of the soil and on the type of air emission control equipment.
7.7 Costs for Emissions Controls
Equations for predicting the costs of emission controls based on system
design parameters are available (PES, 1989). Section 5 of this document provides typical
costs for various types and sizes of treatment systems which could be applied to an ex-
7-13
-------
situ biodegradation process. As mentioned in Section 5, the cost estimates are drawn
from a number of vendors and therefore a range is shown in most cases. The cost data
are intended to show the general level of costs likely to be incurred.
7.8 Summary of Existing Air Emissions Data and Models
Although no models have been explicity developed for estimating emissions
for ex-situ processes treating contaminated soil, there are currently, several public-
domain PC models that are available for estimating air emissions from a variety of other
biotreatment options, principally surface impoundments. The two most commonly used
models are CHEMDAT-7 (U.S. EPA, 1989b) and the Surface Impoundment Modeling
System (SIMS). Both CHEMDAT-7 and SIMS are based on mass transfer and
biodegradation models developed by the U.S. EPA. The mass transfer model uses two-
film resistance theory, along with the characteristics of the impoundment, to estimate
overall mass transfer coefficients for each pollutant. The biodegradation model assumes
Monod kinetics to estimate a biodegradation rate.
The accuracy of estimating emissions from ex-situ biotreatment processes,
though, is limited by the assumptions inherent in both CHEMDAT-71 and SIMS. Both
models perform all calculations at 25°C and rely on physical property and kinetic data
that are not always readily available for the modeled contaminants. Furthermore, both
models rely on a simple thermodynamic analysis and are only valid in the Henry's Law
regime. Henry's Law is applicable to dilute solutions and may not be applicable to
bioslumes. In addition, neither CHEMDAT-7 nor SIMS use thennodynamic models
that can predict the presence of two liquid phases.
Software and user's guide can be obtained by calling EPA's Control Technology
Center Hotline at 919/541-0800.
7-14
-------
The validity of CHEMDAT-7 and SIMS for modeling emissions from ex-
situ biotreatment processes will depend on the process used and the operating
parameters. Their validity must be evaluated on a case-by-case basis. If these models
prove to be unacceptable for a given application, emissions can be estimated using a
simple mass balance approach. For continuous flow systems, the following correlation is
applicable (Thompson, et al., 1991):
ER; = (Ci/l,000)(Mr)(%Vi/100)
where: ERj = emission rate for contaminant i (g/hr);
Cj = concentration of species i in contaminated soil (mg/kg);
Mr = mass rate of soil treated (kg/hr); and
Vj = percentage of contaminant i volatilized.
The percentage of each contaminant that is volatilized will vary greatly
depending on the physical properties of the contaminant and the design of the treatment
system. Based on field studies of an aerated impoundment treating contaminated water.
as much as 20% of each compound may be volatilized depending on its volatility and
biodegradability (Eklund, et al, 1987).
For batch treatment systems a similar expression can be used to estimate
air emissions:
i = (Ci/l,000)(M)(%Vi/100)/(t)
where: ERj = emission rate for contaminant i (g/hr);
Q = concentration of species i in contaminated soil (mg/kg);
M = mass of soil treated (kg);
Vj = percentage of contaminant i volatilized; and
t = residence time in treatment system.
Again volatilization may be 20% or higher depending on the properties of the compound
of interest.
7-15
-------
Emissions will also occur when excavating the contaminated soil, while
transporting it to the treatment unit, during any soil preparation steps, and when feeding
the soil into the treatment process. These fugitive emissions are not addressed in either
the PC-based models or in the mass balance equations discussed above. Fugitive VOC
emissions from soils handling operations are addressed in Section 8.
References
Eklund, B., et al. Assessment of Volatile Organic Air Emissions From an
Industrial Aerated Wastewater Treatment Tank. In: Proceedings of the
14th Annual Hazardous Waste Research Symposium, EPA/600/9-88/021
(NITS PB89-174403) pp 468-475. July 1988.
Thompson, P., A. Inglis, and B. Eklund. Emission Factors for Superfund
Remediation Technologies. EPA-450/1-91-001 (NTIS PB91-190975).
March 1991.
U.S. EPA, 1989a. Background Document for the Surface Impoundment
Modeling System (SIMS). Report No. EPA-450/4-90-019b (NTIS PB91-
156729). U.S. EPA, Research Triangle Park, NC, September 1990.
U.S. EPA, 1989b. Hazardous Waste Treatment, Storage, and Disposal
Facilities (TSDF) - Air Emission Models. Report No. EPA-450/3-87-026
(NTIS PB88-198619). U.S. EPA, Research Triangle Park, NC, December
1987.
U.S. EPA. Engineering Bulletin - Slurry Biodegradation. EPA/540/2-
90/016 (NTIS PB91-228049). September 1990.
7-16
-------
8.0 INCINERATION
Information is presented in this section on incineration. The use of
incineration to remediate soils contaminated with fuel products is limited and is much
less common than the use of thermal desorption, excavation and removal, etc.
8.1 Process Description
A broad range of technologies fall into the category of thermal destruction
or incineration. The most common incineration technologies include liquid injection,
rotary kiln, and multiple hearth (Lee et al, 1986; Cheremisinoff et al, 1986). However,
for remediation of fuel-contaminated soils, rotary kilns are most often used. In general,
soil remediation by thermal destruction fall into two general categories: 1) on-site
treatment using a transportable incinerator, or 2) off-site treatment where contaminated
soils are shipped to a larger, permanent unit. For the treatment of soils contaminated
with petroleum fuels, on-site incineration using mobile or transportable units is much
more common than off-site incineration. Although incineration is a well-established
technology, the evolution of mobile or transportable incinerators is a relatively new
development. The literature on incineration is very extensive. The best sources of
information on air emissions from incineration are two recent reviews (Oppelt, 1987) and
(Eklund, et al., 1989). The Oppelt article is contained in Appendix F to this report.
In broad terms, thermal destruction of hazardous waste is an engineered
process in which controlled combustion is used to reduce the volume of an organic waste
material and render it environmentally safe. Thermal treatment is a flexible process
capable of being used for many waste types including solids, gases, liquids, and sludges.
The advantages of thermal treatment include the following:
• Demonstrated effectiveness;
• Applicability to a wide range of wastes (can be used on most
contaminant and soil types); and
• High commercial availability.
8-1
-------
However, thermal treatment also has a number of significant disadvantages such as:
• High cost;
• Public resistance to the construction and permitting of incinerators;
and
• The need to meet stringent treatment requirements for process
residuals.
Figure 8-1 shows a generalized process flow diagram for thermal treatment
systems. A typical system includes the waste feed system, primary and (in most cases)
secondary combustion chambers, and exhaust gas conditioning system.
At the front-end of a hazardous waste incineration system is the waste feed
process. The configuration of the waste feed system is determined by the physical
characteristics of the waste. Contaminated soil is introduced to the combustion chamber
by means of screw augers or belt feeders. If liquids are to be treated as well in the
incinerator, they are usually injected into the unit by means of an atomization nozzle(s)
which uses steam or compressed air as an atomization fluid. Liquids with entrained
solids may require screening to prevent clogging of the atomizer nozzle.
The largest part of the waste destruction usually takes place in the primary
combustion chamber. As mentioned earlier, for contaminated soils this chamber is
usually a rotating kiln. Gases formed in the primary combustion chamber are then
routed to a secondary combustion chamber, or afterburner, where any unburned
hydrocarbons or products of incomplete combustion such as CO can be fully oxidized.
After the combustion gases leave the incinerator, they may be routed
through a variety of air pollution control devices including gas conditioning, paniculate
removal, and acid gas removal units. Gas conditioning is accomplished with equipment
such as waste heat boilers or quench units. Typical paniculate removal devices include
venturi scrubbers, wet electrostatic precipitators, ionizing wet scrubbers, and fabric filters.
8-2
-------
SECONDARY
AIR FAN
Auxiliary Fuel
Energetic
Waste
Liquids
QUENCH
SECONDARY
COMBUSTION
CHAMBER
Clean
Hue Gas
i
STACK
PRIMARY AIR FAN
Process Water
To Ash
Disposal
Figure 8-1. Process Flow Diagram for Commercial Rotary Kiln Incinerator
8-3
-------
Acid gas removal units include packed, spray, or tray tower absorbers; ionizing wet
scrubbers; and wet electrostatic precipitators.
82 Identification of Air Emission Points
The air emissions associated with full-scale thermal treatment are primarily
stack emissions of combustion gas. There may, however, be some additional evaporative
emissions from equipment leaks and waste handling. Full-scale, off-site incineration
units may vent all emissions from waste handling and transfer activities to the
combustion chamber as make-up air. The air emissions for on-site incinerators are
similar to off-site units, except that waste handling activities have a greater likelihood of
being uncontrolled. For off-site units, typical incinerator stacks will be 100-200 ft high.
For transportable units stack heights may be in the range 40-100 ft. The fugitive
emissions sources associated with thermal treatment will likely be ground-level.
8.3 Typical Air Emission Species of Concern
Emissions from both on-site and off-site incinerators include: undestroyed
organics, metals, paniculate matter, nitrogen oxides (NOX), carbon monoxide (CO), and
acid gases. The cause of each of these pollutants is discussed below. Fugitive emissions
associated with excavation, storage, and handling of the feed material must also be
considered when assessing potential air impacts from incineration (see Section 8).
Unburned Hydrocarbons
In general, incinerators treating wastes must achieve a required destruction
and removal efficiency of at least 99.99% for RCRA wastes and 99.9999% for PCB- or
dioxin surrogate wastes. The remaining 0.01% or 0.0001% of the waste can be assumed
to pass through the system uncombusted (Eklund, et al., 1989). However, in addition to
unburaed hydrocarbons there may be some additional reactions in the combustion
process that may produce a number of simpler organic compounds, called products of
incomplete combustion (PICs). PICs may include dioxin, formaldehyde, and benzo(a)-
8-4
-------
pyrene and other polynuclear aromatic hydrocarbons. PIC formation is not restricted to
the combustion chamber; the reactions which produce PICs may continue to occur in the
combustion gases as they travel through the incineration system and out the exhaust gas
(Eklund, et al., 1989; Treholm and Oberacker, 1985).
Studies indicate that PIC emissions are a natural consequence of the
kinetically-limited thermal degradation of hazardous wastes. Comparison of PIC
formation/destruction rates based on theory and nominal incineration conditions indicate
that PIC emissions can be several orders of magnitude higher than predicted based on
equilibrium (Dellinger, et al., 1991). This suggests that temporal or spatial excursions
from these nominal conditions are occurring that lead to PIC formation. Possible causes
of PIC emissions include low temperatures due to quenching, residence time short
circuits due to nonplug flow and/or unswept recesses, and locally high waste/oxygen
concentration ratios due to poor microscale mixing.
Metals
The metals introduced to the incinerator via the waste feed stream are not
destroyed. Depending on their boiling point, they can either be volatilized or remain as
solids. Volatilized metals will exit the stack as a gas or they will condense or adsorb
onto particles in the stack gas stream. Metals associated with paniculate matter (PM)
will be captured in the PM control device. Non-volatilized metals can be fluidized and
swept up into the combustion gas or leave the incinerator in the bottom ash.
Paniculate Matter
The waste feed, auxiliary fuel, and combustion air can all serve as sources
for paniculate emissions from an incineration system. Paniculate emissions may result
from inorganic salts and metals which either pass through the system as solids or
vaporize in the combustion chamber and recondense as solid particles in the stack gas.
High molecular weight hydrocarbons may also contribute to particulate emissions if
oxidation is not complete. RCRA requirements for particulate emissions call for a limit
8-5
-------
of 0.08 grains/dscf corrected to 7% O2. A number of potential PM control devices can
be used, including Venturis, wet ESPs, ionizing wet scrubbers, and fabric filters.
Nitrogen Oxides
Achieving high levels of destruction of organic wastes is directly related to
combustion chamber temperature: the higher the temperature, the greater the DRE of
organics. Unfortunately, the fixation of nitrogen and oxygen to form NOX also increases
with combustion temperatures. NOX emissions caused by this mechanism are referred to
as thermal NOr Also if there are bound nitrogen atoms in the waste (e.g. amines),
additional NOX emissions, called fuel NO,, will be formed. In such cases, two stage
combustion or emissions controls may be needed.
Carbon Monoxide
Carbon monoxide emissions are generally low (<100 ppmv) in incinerators
due to the high operating temperatures and excess oxygen maintained in the process.
Acid Gases
Hazardous waste incineration will also produce acid gases. These include
oxides of sulfur (SOX), and halogen acids (HC1, HF, and HBr). The sulfur, chlorine,
fluorine, and bromine contents of the waste and fuel feed determine the emission levels
of their respective acid gases. The concentrations of these elements range widely
amongst different wastes; consequently, the resulting acid gas emissions will also show
wide variability. Acid gas emissions will usually not be a concern for the incineration of
soils contaminated by petroleum fuels. Most incinerators will be equipped with some
type of flue gas treatment system to control acid gas emissions. Control efficiencies will
typically range from 85-99%. Units treating soil contaminated with halogenated solvents
will generally be required to meet RCRA requirements governing HC1 emissions.
8-6
-------
8.4 Summary of Air Emissions Data
The wide variety in design and operation of incinerators makes it difficult
to generalize about air emissions. However, extensive research has been done to
determine the range of unburned hydrocarbon and PIC emissions that can be expected
from full-scale incinerators. Table 8-1 shows the range of PIC concentrations found in
testing of several different full-scale incinerators. Given the volume flowrate of the
incinerator offgas these concentrations can be used to estimate the range of emissions
from a particular incinerator system. The data in Table 8-1 are not necessarily based on
the incineration of fuel contaminated soils, and therefore, may overestimate emissions
from the treatment of fuel contaminated soils.
8.5 Identification of Applicable Control Technologies
Unlike other soil remediation technologies, incineration, which converts
organics into carbon dioxide and water, does not require additional add-on VOC
controls. However, additional controls are usually required to reduce emissions of acid
gases, paniculate matter (PM), and metals.
The two primary alternatives for controlling acid gas and PM emissions are
wet or dry scrubbing systems. Wet scrubbing systems typically use a packed- or spray-
tower scrubber with a caustic scrubbing solution to remove acid gases and a venturi
scrubber or wet electrostatic precipitator to remove particulate matter. Dry scrubbing
systems typically use a spray dryer absorber or dry sprbent injection to remove acid gases
from the waste gas stream. The calcium-based alkali absorbent is usually in the form of
slaked lime. Semi-dry systems inject the alkali as a slurry with water which is then
evaporated off. Dry ESPs or fabric filters are used to remove particulate matter from
the gas stream. Table 8-2 shows typical ranges of emissions and estimated removal
efficiencies for acid gas and PM control systems. The efficiency of PM control systems
will depend on the particle size range present in the flue gas.
8-7
-------
Table 8-1.
PICs Found in Stack Effluents of Full-Scale Incinerators1
im^-:^¥IC
Benzene
Chloroform
Bromodichloromethane
Dibromochloromethane
Bromoform
Naphthalene
Chlorobenzene
Tetrachloroethylene
1,1,1-Trichloroethane
Hexachlorobenzene
Methylene chloride
o-Nitrophenol
Phenol
Toluene
Bromochloromethane
Carbon disulfide
Methylene bromide
2,4,6-Trichlorophenol
Bromomethane
Chloromethane
Pyrene
Fluoranthene
Dichlorobenzene
Trichlorobenzene
Methyl ethyl ketone
Diethyl phthalate
o-Chlorophenol
?entachlorophenol
2,4-Dimethyl phenol
Number of Sites
6
5 '
4
4
3
3
3
3
3
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Concentrations (ng/L); i
12 - 670
1 - 1,330
3-32
1- 12
0.2 - 24
5- 100
1-10
0.1 - 25
0.1 - 1.5
0.5-7
2-27
25-50
4 - 22 /
2-75
14
32
18
HO
1
3
1
1
2-4
7
3
7
2-22
6
1-21
*Data from Trenholm, Gorman, and Jungclaus, 1984.
8-8
-------
Table 8-2.
Characteristics of Off-Gas from On-Site Incineration Systems
Table 8-2a. Typical Properties of Off-Gas from Combustion Chamber*
'ijiM&iParameter
Air flow rate
Temperature of Exit Gas
Oxygen Content
System Pressure Drop
Units
ACFM
oF
%
In.H2O
Value
30,000 - 50,000
1,400 - 1,800
3
10-15
'Based on a limited number of designs
Table 8-2b. Typical Emissions
Paniculate Matter
Hydrogen chloride (Hd)
Sulfur dioxide (SO,)
Sulfuric acid (H2SO4)
Arsenic
Beryllium
Cadmium
Chromium
Antimony
Barium
Lead
Mercury
Silver
Thallium
PCDD/PCDF"
EPA* Conservative
: Estimated
Efficiencies
99+%
—
—
—
95
99
95
99
95
99
95
85-90
99
95
—
Typical Actual
Control
Efficiencies
99.9+%
99+
95+
99+
99.9+
99.9
99.7
99.5
99.5
99.9
99.8
40 - 90+
99.9+
99+
90-99+
Typical Range of
Emission Rates
0.005-0.02 gr/dscf
10-50 mg/Nm3
30-60
2.6
1-5 ^g/Nm3
< 0.01-0.1
0.1-5
2-10
20-50
10-25
10-100
10-200
1-10
10-100
1-5 ng/Nm3
'Based on spray dryer fabric filter system or 4-field electrostatic precipitator followed by
a wet scrubber.
** Total all cogeners.
SOURCE: Donnely, 1991.
8-9
-------
8.6 Costs for Remediation
The costs to use thermal destruction to remediate fuel-contaminated soil
will, of course, vary from site to site and depend on whether on-site or off-site treatment
is used. The choice between off-site and on-site incineration is usually determined by
the volume of soil to be treated and the proximity of full-scale off-site hazardous waste
incinerators. The cost of using a transportable on-site incinerator will only be justified if
the volume of contaminated soil to be remediated is large and/or the expense of off-site
incineration is excessive because of transportation costs or other factors.
Table 8-3 shows approximate costs for off-site incineration. As the table
shows the estimated cost for incinerating contaminated soil is $540 - $1,070 per ton
including transportation costs. It should be emphasized that this costs will vary with type
of contamination and the volume of soil to be remediated (Cochran, R., et aL, 1987).
Table 8-4 shows approximate costs for on-site incineration (U.S. EPA
1990). As indicated in the table, costs may range from $180 to $1,580 per ton depending
on the volume of soil being remediated. Also, as the table shows, on-site incineration is
most economical when a large volume of waste must be treated. The cost of
remediation per ton falls significantly as the volume increases. For large-scale on-site
incineration, capital costs are on the order of 5 to 15 millon dollars. Operating costs
which consist primarily of fuel and labor will also be substantial. Additional costs will
also be incurred for the disposal of the incinerator ash, unless the treated soil can be
backfilled on-site.
8-10
-------
Table 8-3.
Estimated Range of Costs for Off-Site Incineration'
Typesfxf Waste
Drummed Waste
Liquids
Dean Liquids with High Btu Value
Soils and/or Highly Toxic Liquids
Cost Range ($/ton)
$170
$70
$20
$540
$540
$540
$70
$1,070
•Data from Cochran, R., et al., 1987.
Table 8-4.
Estimated Range of Costs for On-Site Incineration8
Site Size (Tons)
Very Small (< 5,000)
Small (5,000 - 15,000)
Medium (15,000 - 30,000)
Large (>30,000)
Cost Range ($/ton)
$530
$390
$260
$180
$1,580
$1,020
$680
$530
"Data from Engineering Bulletin: Mobile/Transportable Incineration Treatment .(U.S.
EPA/540/2-90-014) 1990. Data are for the treatment of hazardous waste.
8-11
-------
8.7 Costs for Emissions Controls
Costs for controlling acid gas and paniculate emissions are substantial.
Depending on the volume gas treated, the installed cost for a wet scrubbing system on a
full-scale (i.e. fixed base) incinerator could be $1,000,000-$3,000,000. Costs for wet
scrubber controls for a mobile system are likely to be on the order of $200,000-
$1,000,000. Similar costs would be expected for dry scrubbing systems.
Equations and Models for Estimatin VOC Emissions
A simple mass approach (Thompson et al, 1991) can be used to estimate
emissions from incineration. Separate correlations for each pollutant of concern are
presented below.
Unbumed Hydrocarbons
An emission rate for unburaed hydrocarbons can be generated from a mass
balance on the incinerator system:
where: ERj = emission rate for pollutant i (g/hr);
DREj = destruction efficiency (assume 99.99% if not known);
n^ = total mass flow rate of waste feed (kg/hr); and
q = waste feed concentration for pollutant i (g/kg).
Products of Incomplete Combustion
Emissions of PICs, both the amount and the type, will vary greatly from
unit to unit depending on design and waste feed. Data is currently unavailable to
generate a single emission factor.
8-12
-------
Metals
Metals are not destroyed in the incineration process. They leave the
system via either the bottom ash, are captured in the air pollution control system, or exit
with the stack gas. There are currently no correlations available for determining the
partitioning of metal emissions in incineration systems. If stack data is available for the
incinerator in question, metals emissions rates can be estimated from:
ERj = (q)(mw)(%MEi/100)
where: ERj = emission rate for metal i (g/hr);
Q = concentration of metal i in the feed (g/kg);
m^ = mass flow rate of waste (kg/hr); and
% MEj = metal emitted to air expressed as a percentage of metal fed
(See EPA, 1989).
Acid Gases
The production of acid gases (HC1, SO2, and HF) is determined by the
respective chlorine, sulfur, and fluorine contents in the waste and fuel feed streams. A
conservative approach to calculating the air emissions of these acid gases is to assume
complete conversion of Q, S, and F into their respective acid gas products and apply a
typical removal efficiency for the air pollution control system. These equations follow
the form:
ERi = (q)(Rw)mw(l.%CEI/100)
where ERj = emission rate for acid gas i (g/hr);
Cj = concentration of element (Cl, S, or F) in waste (g/kg);
RJ/J = stoichiometric ratio of acid gas to element (kg/kg);
n^ = mass flow rate of waste (kg/hr); and
%CEj = control efficiency of acid gas treatment system.
Nitrogen Oxides and Carbon Monoxide
In general, incinerator systems are not considered significant sources of
NOX emissions. NOX is usually only a concern for wastes with high nitrogen content.
Typical NOX emissions for an incinerator may be on the order of 100-200 ppmv (dry
basis), or expressed on a fuel basis, 0.12-033 Ibs NOX per MMBtu. If a low-NOx burner
is used, the emissions may be on the order of 0.05 Ibs of NOX per MMBtu.
8-13
-------
CO emissions from incinerators are also not considered a major problem.
Most systems are designed to be fired with excess air (i.e. oxygen rich) to ensure
complete combustion of organic material to carbon dioxide. Vendors typically guarantee
CO emissions less than 100 ppmv (dry basis). Actual measured CO levels are often
lower.
8.9 Case Study: On-Site Incineration
No suitable case study was found for the incineration of soils contaminated
with petroleum fuels.
8.10 References
Cheremisinoff, P.N. Special Report: Hazardous Materials and Sludge
Incineration. Pollution Engineering, Volume 18, Number 12, pp. 32-38.
December 1986.
Cochran, R., et al. Underground Storage Tank Corrective Action
Technologies. EPA/625/6-87-015 (NTIS PB87-171278). January 1987.
Bellinger, B., P. Taylor, and D. Tirey. Minimization and Control of
Hazardous Combustion Byproducts. EPA/600/2-90/039 (NTIS PB90-
259854). May 1991.
Donnelly, J. Air Pollution Controls for Hazardous Waste Incinerators. In:
Proceedings of the 12th Annual HMCRI Hazardous Materials Control/
Superfund 1991. Hazardous Materials Control Research Institute, Silver
Spring, Maryland. December 1991.
Eklund, B.M., et al. Air/Superfund National Technical Guidance Study
Series, Volume HI: Estimation of Air Emissions from Cleanup Activities
at Superfund Sites. EPA-450/1-89-003 (NTIS PB89-180061). January 1989.
Lee, C.C., G.L. Huffman, and DA. Oberacker. Hazardous/Toxic Waste
Incineration. Journal of the Air Pollution Control Association (JAPCA),
Volume 36, Number 8. EPA, Cincinnati, OH. August 1986.
Oppelt, E.T. Incineration of Hazardous Waste - A Critical Review.
JAPCA, Vol. 37, No. 5, pp558-586, May 1987.
8-14
-------
Thompson, P., A. Inglis, and B. Eklund. Emission Factors for Superfimd
Remediation Technologies. EPA-450/1-91-001 (NTIS PB91-190975). May
1991.
Trenholm, A., P. Gorman, and G. Jungclaus. Performance Evaluation of
Full-Scale Hazardous Waste Incineration Volume I. EPA-600/2-84-181a,
(NTES PB85-129500). November 1984.
Trenholm, A. and D. Oberacker. Summary of Testing Program at
Hazardous Waste Incinerators. In Proceedings of the llth Annual
Research Symposium, Incineration and the Treatment of Hazardous Waste.
EPA/600/9-85/028 (NTIS PB86-199403). September 1985.
U.S. EPA. Engineering Bulletin - Mobile/Transportable Incineration
Treatment. EPA/540/2-90/014 (NTIS PB91-228023). September 1990.
8-15
-------
9.0 SOIL WASHING/SOLVENT EXTRACTION
9.1 Process Description
Three remediation technologies are described below: soil washing, solvent
extraction, and soil flushing. These are all primarily separation processes and further
treatment of the collected contaminants will typically be required.
9.1.1 SoU Washing
Soil washing is an ex situ process in which contaminated soil is excavated
and fed through a water-based washing process. It operates on the principle that
contaminants can be dissolved or suspended in an aqueous solution or removed by
separating out clay and silt particles and the associated adhered contaminants from the
bulk soil. The aqueous solution containing contaminants may be treated by conventional
wastewater treatment methods (U.S. EPA, 1990a).
Most organic and inorganic contaminants bind chemically or physically to
clay or silt soil particles, which in turn adhere to larger sand and gravel particles
primarily by compaction and adhesion. Particle size separation by washing enables the
contaminated clay and silt particles (and the bound contaminants) to be concentrated.
Separating the sand and gravel from the small contaminated soil particles significantly
reduces the volume of contaminated soil, making further treatment or disposal much
easier. The larger particles may be returned to the site (U. S. EPA, 1990b).
If soil washing lowers contaminant concentrations in the soil to acceptable
levels, the only additional treatment to consider would be emissions controls. Soil
washing also serves as a cost-effective pre-processing step for further treatment. It can
potentially be effective for the remediation of soils with a small amount of clay and silt
particles and with a wide variety of organic, inorganic, and reactive contaminants. Large
amounts of clay and silt particles mitigate the effectiveness of soil washing and make it
9-1
-------
inadequate as the only treatment method. Removal efficiencies range from 90-99
percent for volatile organic compounds (VOCs) and 40-90 percent for semi-volative
compounds. Compounds with low water solubilities such as metals and pesticides
sometimes require acids or chelating agents to assist in removal (U.S. EPA, 1990a).
Particle size distribution is a key parameter in determining the feasibility of
soil washing. The relative effectiveness of soil washing for various soil types are shown
below.
Particle size distribution
>2mm
0.25-2 mm
0.063-0.25 mm
< 0.063 mm
Effectiveness i
Oversize pretreatment requirements
Effective soil washing
Limited soil washing
Clay and silt fraction: difficult soil washing
Bench-scale and pilot-scale treatability tests are recommended before undertaking full-
scale operation. Further concerns about feasibility include the fraction of hydrophobic
contaminants that require surfactants or organic solvents for effective removal, the
complexity and stability of the contamination that affect washing fluid formulation, and
the effect of washwater additives on wastewater treatment (U.S. EPA, 1990a).
Excavation and removal of debris and large objects precedes the soil
washing process. Sometimes water is added to the soil to form a slurry that can be
pumped. After the soil is prepared for soil washing, it is mixed with washwater and
sometimes extraction agents. At this point, three separation processes occur: 1) water-
soluble contaminants are transferred to the washwater; 2) contaminants are suspended in
the washwater; and 3) clay and silt particles to which contaminants are adhered separate
from larger soil particles. After separation from the washwater, the soil is rinsed with
clean water and may be returned to the site. The suspended soil particles are removed
by gravity from the washwater as sludge. Sometimes flocculation is used to aid in sludge
removal. This sludge is more highly contaminated than the original soil and undergoes
9-2
-------
further treatment or secure disposal. The spent washwater from which the sludge is
removed is treated and recycled. Residual solids from the recycle process may require
further treatment (U.S. EPA, 1990a).
Soil washing generates four waste streams:
1) contaminated solids separated from the washwater;
2) wastewater
3) wastewater treatment sludges and residual solids; and
4) air emissions.
Any of a number of treatments is feasible for the contaminated clay fines and solids.
They may successfully undergo incineration, low temperature thermal desorption,
solidification and stabilization, and biological or chemical treatment. It is recommended
that as much blowdown water be recycled as possible. Slowdown water released to local
wastewater treatment plants must meet local discharge standards. Sludges and solids
from wastewater treatment require appropriate treatment and disposal. Collected air
emissions from the waste site or soil washing unit can be treated as well (U.S. EPA,
1990a).
Advantages of the soil washing process include:
1) applicability to a wide variety of organic and inorganic compounds.
2) high removal efficiencies for certain soil types; and
3) rninimal fire and explosion hazards.
Some disadvantages as compared to other remediation processes are that
soil washing:
1) is suitable for only certain soil types;
2) does not destroy contaminants; and
3) may require additives that improve removal but compromise
treatment of the waste streams.
9-3
-------
Figure 9-1 shows a process diagram of the soil washing process.
9.12 Solvent Extraction
Solvent extraction differs from soil washing in that it employs organic
solvents rather than aqueous solutions to extract contaminants from the soil. Like soil
washing, it is a separation process that does not destroy the contaminants. The
contaminants will have greater solubility in the solvent than in the soil. The equilibrium
concentration gradient drives the mass transport process such that the contaminant
transfers from the soil to the solvent. When the soil is separated from the solvent, the
soil contaminant concentrations are presumably lower than before contact with the
solvent. Soil washing treats organic compounds much more effectively than inorganic
compounds and metals. It can be used in conjunction with other processes to reduce
remediation costs (U.S. EPA, 1990b).
Sediments, sludges, and soils contaminated with volatile organic compounds
(VOCs), polychlorinated biphenyls (PCBs), halogenated solvents, and petroleum wastes
can be effectively treated with solvent extraction. The removal of inorganic compounds
such as acids, bases, salts, and heavy metals is limited, but the compounds usually do not
hinder the remediation process. Metals may undergo a chemical change to a less toxic
or leachable form but their presence in the waste streams may also restrict disposal and
recycle options (U.S. EPA, 1990b).
The remediation process begins with excavating the contaminated soil and
feeding it through a screen to remove large objects. In some cases, solvent or water is
added to the waste in order to pump it to the extraction unit. In the extractor, solvent is
added and mixed with the waste to promote dissolving of the contaminants into the
solvent Laboratory testing can determine which solvent adequately separates the
contaminants from the soil. Generally, the solvent has a higher vapor pressure than the
9-4
-------
Excavated or Pumped
Contaminated Soil
Figure 9-1. Schematic Diagram of Aqueous Soil Washing Process
9-5
-------
contaminants so that with an appropriate pressure or temperature change, the solvent
may be separated from the contaminants, compressed, and recycled to the extractor (U.S.
EPA, 1990b).
Up to five waste streams may result from the solvent extraction process:
1) Concentrated contaminants;
2) Solids;
3) Wastewater;
4) Oversized rejects; and
5) Treated air emissions.
With the contaminants in a more concentrated form, they may be analyzed and
subsequently designated for further treatment, recycle, or reuse before disposal. Solvent
extraction has presumably improved the condition of the solids but often the solids need
dewatering, treatment for residual organic compounds, additional separation,
stabilization, or other treatment. The water from the dewatering process, the solids, and
the water from the extractor need analysis to choose the most appropriate treatment and
disposal. Typically solvent extraction units are designed to produce negligible air
emissions, but significant levels of emissions may occur during waste preparation (U.S.
EPA, 1990b). The units are a closed-loop design and the solvent is recycled and reused.
The primary advantage of solvent extraction is the treatability of a wide
variety of media. This is in contrast to soil washing, the success of which is heavily
dependent on the particle size distribution.
9-6
-------
Some disadvantages of the process are that solvent extraction:
1) does not destroy the contaminants;
2) may not be appropriate for contaminants with high vapor pressures
because these compounds may be removed with the solvent in the
separation process instead of remaining with the concentrated
contaminant stream;
3) is compromised by the presence of detergents and emulsifiers which
compete with the solvent in dissolving the contaminants;
4) may leave residual solvent and contaminant concentrations in the
treated waste;
5) is not effective for high molecular weight or hydrophilic compounds;
and
6) may use flammable or mildly toxic solvents.
Figure 9-2 shows a process diagram of the solvent extraction process.
A variety of solvent extraction systems have been developed to treat
several types of contamination. Six systems are described below.
CF Systems
Probably the most widely used solvent extraction system is the CF Systems
which uses liquified hydrocarbons such as propane and butane as the solvent to treat soil
and sludge, and carbon dioxide to treat wastewater. Water is added to the waste to
enable pumping of the material through the extraction process. Particles greater than
1/8 inch in diameter are removed. In some cases, oversize particles are reduced in size
for subsequent processing. The pH is adjusted in the feed to minimize corrosion of
metallic components of the treatment system. CF Systems has been used at a 50
tons/day capacity to remediate refinery sludge and at 1.5 gallons/minute to treat
sediments with PCBs (U.S. EPA, 1990b).
9-7
-------
Figure 9-2. Schematic Diagram of Solvent Extraction Process
9-8
-------
RCC B.E.S.T.™
RCCs B.E.S.T.™ system does not need a pumpable waste and uses
aliphatic amines (often triethylamine) as the solvent. Feed pH is adjusted to alkaline
conditions and objects over one inch in size are removed. The process operates at near
ambient temperature and pressure. Due to its high vapor pressure and low boiling point
azeotrope formation, triethylamine is removed with steam stripping. The full-scale
system has treated refinery waste streams at 100 tons/day and PCB-contaminated
petroleum sludge at 70 tons/day. The bench-scale system has treated oil/grease
contamination and sediments with PCBs in a six kilogram batch process.
ENSR
The ENSR system is currently under development and is a mobile solvent
extraction unit that can process contaminated soils and sludges at a rate of five to ten
cubic yards per hour. Using a proprietary solvent and reagent, the system is designed to
operate without significant pretreatment or water use.
Extraksol™
The Sanivan Group in Montreal, Canada, developed the Extraksol™
process in 1984. The system has been used to treat PCBs, oil, grease, polyaromatic
hydrocarbons (PAHs), and pentachlorophenol in a one ton per hour mobile unit that
uses a proprietary solvent. This small unit may treat up to 300 tons of material. Sanivan
is developing a full-scale unit to treat six to eight tons/hour of waste.
Harmon Environmental Services and Acurex Corporation
In a joint venture, Harmon Environmental Services and Acurex
Corporation are developing a solvent soil washer/extractor system for on site
remediation. The U. S. Environmental Protection Agency has sponsored bench-scale
studies of treating No. 2 fuel oil.
9-9
-------
Low Energy Extraction Process (LEEP)
The patented LEEP process treats soils, sediments, and sludges on site
with common hydrophilic and hydrophobic organic solvents to remediate such pollutants
asPCBs.
9.13 Soil Flushing
Soil flushing differs from soil washing and solvent extraction in that it is an
in situ process in which the solvent is sprayed over the contaminated area, percolates
through the soil and dissolves the contaminants. Elutriate is collected in a series of wells
and drains. Flushing solutions may include the following:
1) Water for water-soluble contaminants;
2) Acidic aqueous solutions for metals and basic organic contaminants;
3) Basic aqueous solutions for some phenols, complexing and chelating
agents and for metals such as zinc, tin, and lead; and
4) Surfactants.
Several factors merit consideration in the soil flushing process. Laboratory
tests are essential to determine the best flushing solution for the types of soil and
contaminants present. The flushing solution may affect the soil such that removal is
hindered and it may also alter the soil's physical and chemical properties after
remediation. Aerobic and anaerobic biodegradation may occur, also affecting the soil
and contaminant composition. Solvents and contaminants may migrate into
uncontaminated areas and also be resistant to removal due to soil heterogeneity
(Chambers, C.D., et al., 1990). A process diagram for soil flushing is shown in Figure 9-
3.
9-10
-------
Spray
Application
Lcachalf
Figure 9-3. Generalized Soil Flushing Process Flow Diagram
9-11
-------
Identification of Air Emission Points
In the soil washing process the greatest potential for emissions of volatile
contaminants occurs in the excavation, feed preparation, and extraction process.
Collected emissions from these processes are typically treated by carbon adsorption or
incineration (U.S. EPA, 1990a). Because soil washing occurs in liquid and solid phases,
volatile compounds emitted evolve primarily due to their vapor pressures in these phases.
The waste streams also have the potential to be sources of VOC emissions.
Solvent extraction may also produce emissions during excavation and soil
transport and from contaminated oversize rejects (U.S. EPA, 1990b). Because the
solvent recovery process involves vaporization of the solvent, fugitive emissions are
possible from this as well as other stages of the solvent process, including the waste
streams.
Emissions from soil flushing may emanate from the soil surface, solvent
storage vessels and spray system, and from locations where the contaminant-laden
flushing solution surfaces.
9.3 Typical Air Emission Species of Concern
In addition to the contaminants that may volatilize, the solvents themselves
may be cause for concern. Products of aerobic and anaerobic decomposition are also
possible.
9.4 Summary of Air Emissions Data
No data were identified for the air emissions from soil washing, solvent
extraction, and soil flushing. Information on emissions from excavation may be found in
Section 8.
9-12
-------
9.5 Identification of Applicable Control Technologies
Carbon adsorption and incineration are typical controls used to treat
collected emissions. In solvent extraction, volatile solvents are recovered and recycled.
These control technologies are described in Section 5.5.
9.6 Capital and Operating Costs for Remediation
Cost for remediating contaminated soil by soil washing range from $53 to
$215 per ton of feed soil, according to information from vendors of the equipment. The
more expensive processes included in the cost range cover disposal of soil residue (U.S.
EPA, 1990a).
Solvent extraction costs are most influenced by waste volume, number of
extraction stages, operating parameters, and lost time. Operating parameters include
labor, maintenance, setup, decontamination, and demobilization and lost time may result
from delays in equipment operation. The choice of solvent, solvent/waste ration, feed
rate, extractor residence time and number of passes through the extractor determine the
efficiency of the process. Estimated costs range from $105 to $525. It is not clear
whether this estimate includes emissions controls.
No cost data are available on soil flushing although costs are moderate if
inexpensive flushing solutions are used and no excavation takes place.
9.7 Capital and Operating Costs for Emission Controls
The CF System Organic Extraction Process was used to remediate PCB
contamination in New Bedford Harbor in Massachusetts (Valentinetti, R., 1990a and
1990b). Using liquefied propane and butane, costs were estimated to range between
9-13
-------
$150/ton and $450/ton. These estimates do not include emissions control. General
costs for controlling point source emissions are given in Section 5.7.
9.8 Equations/Models for Estimating Emissions
No equations or models for predicting the air emissions from these
processes were identified..
9.9 Case Studies of Remediation and Air Emissions
Given the lack of air emissions data, no suitable case studies showing
emissions were found for these processes. Remediation performance data, however,
were available and Tables 9-1 through 9-3 show selected results of treatments at several
sites. Further information may be obtained from the relevant documents listed int he
bibliography.
9.10 References
Chambers C.D., et al. Handbook of In Situ Treatment of Hazardous
Waste - Contaminated Soils. EPA/540/2-90/002 (NTIS PB90-155607).
January 1990.
U.S. EPA, 1990a. Engineering Bulletin - Soil Washing Treatment.
EPA/540/2-90/017 (NTIS PB91-228056). September 1990.
U.S. EPA, 1990b. Engineering Bulletin - Solvent Extraction Treatment.
EPA/540/2-90/013 (NTIS PB91-228015). September 1990.
Valentinetti, R. Technology Evaluation Report: CF Systems Organics
Extraction System - Volume 2, New Bedford, Massachusetts. Report No
EPA-540/5-90/002 (NTIS PB90-186503). January 1990a.
Valentinetti, R. Applications Analysis Report: CF Systems Organics
Extraction System - Volume 1, New Bedford, Massachusetts. Report No
EPA-540/5-90/002 (NTIS PB90-186495). August 1990b.
9-14
-------
Table 9-1.
Removal Efficiencies for Remediation of PCB Contamination
Test
2
3
4
Passes
10
3.
6
Initial
Concentration
(ppmPCB)
350
288
2,575
Final
Concentration
(ppmPCB)
40
82
200
Percent
Removal
89%
72%
92%
NOTE: CF Systems Organic Extraction System, New Bedford Harbor,
Massachusetts from (Valentinetti, R., 1990a). This process used liquefied
propane and butane at 240 psi and 69°F to remediate contaminated
sediments.
Table 9-2.
Removal Efficiencies for Remediation of API Separator Sludge
Compound
Benzene
Toluene
Ethylbenzene
Total Xylenes
Anthracene
Benzo(a)pyrene
Bis-(2-ethylhexyl)phthalate
Chrysene
Naphthalene
Phenanthrene
Pyrene
Initial
Concentration
<^g/g)
30.2
16.6
30.4
132
28.3
1.9
4.1
63
42.2
28.6
7.7
Final
Concentration
<^g/g)
0.18
0.18
023
0.98
0.12
0.33
1.04
0.69
0.66
1.01
1.08
Percent
Removal
99%
99%
99%
93%
100%
83%
75%
89%
98%
96%
86%
NOTE: CF Systems Organic Extraction System, Port Arthur, Texas from (U.S.
EPA, 1990b). This process operated at a capacity of 50 tons per day using
liquefied hydrocarbon gases as the solvent.
9-15
-------
Table 9-3.
Summary of Performance Data on Soil Washing
Process
Soil Cleaning of America
Biotrol Soil Treatment System
Biotrol Soil Treatment System
EPA's First Generation Pilot
Drum Screen Washer
MTA Remedial Resources
MTA Remedial Resources
MTA Remedial Resources
Bodemsandering Nederland BV
Bodemsandering Nederland BV
Harbauer of America
Harbauer of America
Harbauer of America
Harbauer of America
Heidemij Froth Flotation
Klockner Umweltechnik
Klockner Umweltechnik
Klockner Umweltechnik
Klockner Umweltechnik
Klockner Umweltechnik
Contaminants
oil and grease
Pentachlorophenol
other organics
oil and grease
—
volatile organics
semi-volatile organics
most fuel producs
aromatics
crude oil
total organics
total phenols
PAH
PCB
oil
hydrocarbons
chlorinated hydrocarbons
aromatics
PAHs
Phenol
Range of
Removal
Efficiencies
50 - 83%
90 - 95%
85 - 95%
90 - 99%
—
98-99+%
98-99+%
98-99+%
>81%
97%
96%
86 - 94%
86 - 90%
84 - 88%
>99%
963%
>75%
99.8%
95.4%
>99.8%
Residual
Concentrations
250 - 600 ppm
<115 ppm
< 1 ppm
<5 - 2400 ppm
—
<50 ppm
<250 ppm
<2200 ppm
>45 ppm
2300 ppm
159 - 201 ppm
7 - 22.5 ppm
91.4 - 97.5 ppm
0.5 - 13 ppm
20 ppm
82.05 ppm
<0.01 ppm
< 0.02 ppm
15.48 ppm
<0.01 ppm
Source: U.S. EPA, 1990a.
9-16
-------
APPENDIX A
PROPERTIES AND COMPOSITION OF VARIOUS FUEL TYPES
Brief descriptions are given below for liquified petroleum gases, gasoline,
diesel fuel, jet fuel, oil, and asphalt and bitumen.
Liquified Petroleum Gases (LPG> comprise ethane, ethylene, propane,
propylene, normal butane, butylene, and isobutane and are typically produced at
refineries or natural gas processing plants. Normal butane added to gasoline helps to
regulate its vapor pressure and isobutane serves as an alkylation feedstock and is sold as
LPG.
Gasoline is petroleum derivative with over 100 components boiling from
90°F to 420°F. Additives that improve gasoline performance can change is physical
properties significantly. Adding normal butane adjusts the Reid vapor pressure (RVP)
so that it varies between about 9.5 psi in the winter and 13.5 psi in the summer.1
Butane accounts for about 75% of the vapor pressure of gasoline with pentanes making
up much of the remainder. Detergent additives acting as surfactants reduce the surface
tension which in turn influences subsurface migration. Hydrophilic additives such as
niethyl-tert-butyl ether (MTBE), methanol, and ethanol boost octane numbers and
considerably increase the solubility of gasoline in water.
Diesel Fuel is used by trucks, railroads, stationary engines, and some
automobiles. The three types of diesel most commonly used are No.l and No.2 for
automobiles and trucks; and No.4, which is heavier and used by large slow-speed
vehicles3. This middle distillate tends to be less volatile, less mobile in soil, and less
water soluble than gasolines4.
Handbook of Energy Technology and Economics, Robert A. Meyer, ed. John Wiley
and Sons, 1983, NY.
2David K. Kreamer and Klaus J. Stetzenback, "Development of a Standard, Pure-
Compound Base Gasoline Mixture for Use as a Reference in Field and Laboratory
Experiments," Spring 1990, GWMR, p. 136.
3Handbook of Energy Technology and Economics, Robert A. Meyers, ed. John Wiley
and Sons, 1983, NY, p. 217-8.
4"A Guide to the Assessment and Remediation of Underground Petroleum Releases,"
API Publication 1628, 2nd Ed., August 1989, p.9.
A-l
-------
Jet Fuels used by commercial and military aircraft resembles kerosene and
has a similar boiling range to light diesel fuels and heating oils. Jet fuels contain no
more than 20% aromatic compounds5. This middle distillate tends to be less volatile,
less mobile in soil, and less water soluble than gasolines6.
Oil - Heating oils No.l and No.2 are used to heat homes and businesses
and the heavier oils, Nos. 4, 5, and 6 are used by shipping and industry, and have higher
viscosity and pour points7.
Asphalt and bitumen are solid phase components of crude oil that remain
virtually immobile hi soil because shallow subsurface temperatures rarely rise above their
melting points8.
5Handbook of Energy Technology and Economics, Robert A. Meyers, ed. John Wiley
and Sons, 1983, NY, p. 217-8.
^'A Guide to the Assessment and Remediation of Underground Petroleum Releases,"
API Publication 1628, 2nd Ed., August 1989, p.9.
'Handbook of Energy Technology and Economics, Robert A. Meyers, ed. John Wiley
and Sons, 1983, NY, p. 217-8.
^A Guide to the Assessment and Remediation of Underground Petroleum Releases,"
API Publication 1628, 2nd Ed., August 1989, p.9.
A-2
-------
APPENDIX B
STATE CLEANUP REQUIREMENTS FOR TOTAL PETROLEUM
HYDROCARBONS IN CONTAMINATED SOIL
State
Soil Cleanup Levels
(ppm Total Petroleum Hydrocarbons)
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
100 if >5 ft to groundwater, 10 otherwise
100
100
case by case basis
10-1,000 for gasoline, 100-10,000 for diesel
case by case basis
10 for gasoline, case by case otherwise
10 for gasoline, case by case otherwise
10-500 for gasoline, 50 for diesel
500
100 (if within 0.5 to 3 miles of well)
50
100 for gasoline, 1,000 for diesel
none for TPH, 0.025 for benzene, 16.025 for BTEX
100
100
100
background concentration
case by case basis
case by case basis, generally 20-50
case by case basis
100
background concentration
10 for gasoline, 1 for diesel (field PID)
100 (BTEX) for gasoline, 100 TPH for diesel
100
10
case by case basis
B-l
-------
State
Soil Cleanup lievels
(ppm Total Petroleum Hydrocarbons)
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
West Virginia
Washington
Wisconsin
Wyoming
100
10 for gasoline, 100 for diesel
100
50 (TAH) for gasoline, 100 TPH for diesel
case by case basis
10
case by case basis
background concentration
50
40-130 for gasoline, 100-1,000 for diesel
case by case basis
50
10 (BTEX) for gasoline, 100 TPH for diesel
10
100-500 for drinking water area, otherwise 250-1,000
100
50
20 (TAH), case by case basis
100
100
100 for gasoline, 200 for diesel
10
10 if <50 ft to groundwater, 100 if >50 ft to groundwater
BTEX = benzene, toluene, ethylbenzene, xylenes
TAH = total aromatic hydrocarbons
TPH = total petroleum hydrocarbons
PID = Photoionization detector reading
Notes: 1) Values in this table are subject to change.
2) Values were taken from several sources including: Steel Tank
Institute. Tank Talk, Vol. 6, Number 1, January 1991.
B-2
-------
APPENDIX C
EXAMPLE CALCULATIONS
CONTENTS
Page
Soil Vapor Extraction (Section 5) C-2
Evaporative Emissions from SVE System at Groveland, MA Site C-3
Emissions from In-situ Biotreatment System (Section 6) C-6
Emissions from Ex-situ Biotreatment System (Section 7) C-7
Emissions from Thermal Destruction (Section 8) C-10
C-l
-------
SAMPLE CALCULATIONS FOR SOIL VAPOR EXTRACTION
(Section 5)
Stack Emissions from Soil Vapor Extraction
Emission Factor
0.05 g VOC Emitted
g VOC Removed From Soil
Hourly Emissions
• Assumptions:
250 kg VOC removed per 10 hr workday
• Emissions:
250 kg VOC x 1 day x 1,000 g 0.05 g VOC Emitted _ 1,250 g VOC Emitted
day 10 hr 1 kg g VOC Removed " hr
C-2
-------
SAMPLE CALCULATIONS FOR EVAPORATIVE EMISSIONS
FROM SVE AT THE GROVELAND, MA SITE
(Section 5)
Estimated Evaporative Emissions from Semi-Closed Tank of Contaminated Water
ClOO
Apply Simple Mass Transfer Model (Two-Film Resistance Theory)8
where: Q = evaporative emissions flux (g/m25);
KJ = overall mass transfer coefficient (m/s);
Q = concentration of contaminant (mol/m3); and
MJ = molecular weight of contaminant (g/mol).
In turn, the overall mass transfer coefficient is given by the equation:
where: K^ = liquid film mass transfer coefficient for contaminant i (m/s);
Kjg = gas film mass transfer coefficient for contaminant i (m/s); ai
Hs = Henry's Law Constant for contaminant i (atm-m3/mol).
j, calculated from the correlations of Mackay and Yuenb.
KJ, = 1.0 x Itr6 + 144 + 10"4 U* Scf0-5
U* = friction velocity
0.01 (6.1 x 0.63 Ujo)0-5
U10 = wind speed at 10 m above water surface
= 0 for semi-covered tank at Groveland
Sc, = Schmidt Number for Liquid Film
Mwater
Mw = viscosity of water
Pw = density of water
D = diffusivity of contaminant i in water
C-3
-------
SAMPLE CALCULATIONS FOR EVAPORATIVE EMISSIONS
FROM SVE AT THE GROVELAND, MA SITE
(Section 5)
(Continued)
K^ calculated from correlation of Mackay and Yuenb.
Kig = 1.0 x 10'3 + 46.2 x 10'3 U* (Scg)-°'
Scg = Schmidt Number for gas film
£air
Pair^air
fi = viscosity of air
p = density of air
Dair = diffusivity of contaminant i in air
Total Evaporative Emissions:
EJ = QA
EJ = emissions from contaminated water
Qi = emissions flux (g/m2-sec)
A = surface area of water (m2)
Estimated Stack Emissions:
Assume uncontrolled VOC emission rate is equal to recovery rate of VOCs:
Uncontrolled Emissions = Recovery Rate
= 5,538 gVOC/hr
C-4
-------
SAMPLE CALCULATIONS FOR EVAPORATIVE EMISSIONS
FROM SVE AT THE GROVELAND, MA SITE
(Section 5)
(Continued)
Calculate controlled VOC emission rate using control efficiency for carbon
canisters:
Controlled Emissions = Recovery Rate *
1 -
eff
100
= 5,538 *
1 -
99.75
100
=6.29 g/hr
Notes:
8 Mackay, D., and P J. Leinonen. "Rate of Evaporation of Low-Solubility Contaminants
from Water Bodies to Atmosphere". Environmental Science and Technology. Vol. 9,
No.13. December 1975.
b Mackay, D., and A.T.K. Yuen. "Mass Transfer Correlations of Organic Solutes from
Water" Environmental Science and Technology. Vol. 17, No. 4. April 1983.
C-5
-------
EMISSIONS FROM IN-SITU BIOTCEATMENT SYSTEMS
(Section 6)
Land
Emission Factor
Hourl Emissions
0.36 g VOC Emitted2
g VOC in Waste
Assumptions
Volume of Waste
VOC Concentration
Treatment Time
1,000 m3
100 g/m3
24 hours
Total VOC Treated (g VOC)
1,000 m3 x -122J = 100,000 g VOC
Emissions
100,000 g VOC x
°'36 £
Emitted
1 = 1,500 g VOC Emitted
g VOC in Waste 24 hours
hour
between impoundment turnovers.
2Based on 24-hour period.
C-6
-------
EMISSIONS FROM EX-SITU BIOTREATMENT SYSTEMS
(Section 7)
Flow-Through Impoundments with Mechanical Aeration
Emission Factor
0.8 g VOC Emitted
g VOC in Waste
Hourly Emissions
• Assumptions
Influent Flowrate 1 m3/min
VOC Influent Concentration 100 g/m3 (100 ppm)
Total VOC Treated (g VOC/hr)
m3 60 min 100 g = 6,000 g VOC Treated
min hr m3 hr
• Emissions
6,000 g VOC Treated 0.8 g VOC Emitted _ 4,800 g VOC Emitted
hrg VOC in Waste hr
Duiescent Impoundments
Emission Factor
0.12 g VOC Emitted
g VOC in Waste
C-7
-------
EMISSIONS FROM EX-SITU BIOTREATMENT SYSTEMS
(Section 7)
(Continued)
Hourly Emissions
• Assumptions
Influent Flowrate 1 m3/min
VOC Influent Concentration 100 g/m3 (100 ppm)
Total VOC treated (g VOC/hr)
m3 60 min ^ 100 g _ 6,000 g VOC Treated
min hr m3 hr
• Emissions
6 000 e VQC Treated x 0.12 g VOC Emitted _ 720 g VOC Emitted
hr g VOC in Waste hr
Disposal Impoundment
Emissions
0.14 g VOC Emitted
g VOC in Waste
Hourly Emissions
• Assumptions
Impoundment Size 15,000 m3
VOC Influent 100 g/m3
Disposal Time 6 months
C-8
-------
EMISSIONS FROM EX-SITU BIOTREATMENT SYSTEMS
(Section 7)
(Continued)
Total VOC Treated (g VOC)
15,000 m3 x 100g = 1,500,000 g VOC
m3
Emissions
000000 e VOC x °-14 8 VOC Emitted 1 x 1 month x 1 day _ 48.6 g VOC Emitted
8 g VOC in Waste 6 months 30 days 24 hour " hr
C-9
-------
EMISSIONS FROM THERMAL DESTRUCTION
(Section 8)
Air Emissions from Off-Site Rotary-Kiln Incineration (Controlled)
Emission Factors:
0.1 g VOC / kg VOC in Waste
50 g Metals / kg Metal in Waste
0.01 g HC1 / g Cl in Waste
0.01 g HF / g F in Waste
0.10 g SO2 / g S in Waste
72 mg PM / m3 flue gas
50 ppmv CO / m3 flue gas
100 ppmv NOX / m3 flue gas
Assumptions:
Incinerator
Stack Gas Flow
Waste Fed
Heat Load
986 m3/min
3,400 kg/hr
63 MM kJ/hr
Waste Characterization
Cl in waste 4.0%
F in waste 1.0%
S in waste 5.0%
Metal in waste 0.1%
Estimated Control Efficiencies for Acid Gas Scrubbing
HC1 99%
HF 99%
SO, 95%
C-10
-------
EMISSIONS FROM THERMAL DESTRUCTION
(Section 8)
(Continued)
Hourly Emissions:
VOC: 1,400^0.999 g Qrganic^.l g
hr g waste kg waste
Metak: MOOJcgx0.001 kg Mx 50 g emitted M =no
hr kg Waste 1 kg M in waste
HQ: 3.400 kg wast^O.04 kgC^O.OU g
hr kg waste g F
3>400kgwaStex0.01kgFx0.11gHF=04 g
hr kg waste g F
,x.;i.jl70
2 hr kg waste g S
PM- 72 mg PMx986 m3x60 minx g ^60 g PM/hr
m3 min hr 1,000 mg
NO - _ x100 ,_=11 g
x" min 0.0236 m3 10* gmole Air
C-ll
-------
APPENDIX D
Unted States
EnvkonmentrfProtacfioR
Agency
Office of Emergency and
fieraedtel Response
Office of
ftanoareh and Development
Ctodnwtf, OH 45268
Superfund
S»A/54u/2-91/!)08
May 1991
?/EPA
Engineering Bulletin
Thermal Desorption Treatment
Purpose
Section 121(b) of the Comprehensive Environmental Re-
sponse, Compensation, and Liability Act (CERCLA) mandates
the Environmental Protection Agency (EPA) to select remedies
that "utilize permanent solutions and alternative treatment
technologies or resource recovery technologies to the maxi-
mum extent practicable" and to prefer remedial actions in
which treatment "permanently and significantly reduces the
volume, toxitity, or mobility of hazardous substances, pollut-
ants and contaminants as a principal element" The Engi-
neering Bulletins are a series of documents that summarize
the latest information available on selected treatment and site
remediation technologies and related issues. They provide
summaries of and references for the latest information to help
remedial project managers, on-scene coordinators, contrac-
tors, and other site cleanup managers understand the type of
data and site characteristics needed to evaluate a technology
for potential applicability to their Superfund or other hazard-
ous waste site. Those documents that describe individual
treatment technologies focus on remedial investigation scoping-
needs. Addenda will be issued periodically to update the
original bulletins.
Abstract
Thermal desorption is an ex situ means to physically
separate volatile and some semivolatile contaminants from
soil, sediments, sludges, and filter cakes. For wastes contain-
ing up to 10% organic* or less, thermal desorption can be
used alone for site remediation. It also may find applications
in conjunction with other technologies or be appropriate to
specific operable units at a site.
Site-specific treatability studies may be necessary to
document the applicability and performance of a thermal
desorption system. The EPA contact indicated at the end of
this bulletin can assist in the definition of other contacts and
sources of information necessary for such treatability studies.
Thermal desorption is applicable to organic wastes and
generally is not used for treating metals and other inorganics.
Depending on the specific thermal desorption vendor se-
lected. the technology heats contaminated media between
200-1 OOO'F, driving off water and volatile contaminants.
^reference number, page number]
Offgases may be burned in an afterburner, condensed to
reduce the volume to be disposed, or captured by carbon
adsorption beds.
Commercial-scale units exist and are in operation. Ther-
mal desorption has been selected at approximately fourteen
Superfund sites [1]* [2]. Three Superfund Innovative Technol-
ogy Evaluation demonstrations are planned for the next year.
The final determination of the lowest cost alternative will
be more site-specific than process equipment dominated.
This bulletin provides information on the technology applica-
bility, limitations, the types of residuals produced, the latest
performance data, site requirements, the status of the tech-
nology, and sources for further information.
Technology Applicability
Thermal desorption has been proven effective in treating
contaminated soils, sludges, and various filter cakes. Chemi-
cal contaminants for which bench-scale through full-scale
treatment data exist include primarily volatile organic com-
pounds (VOCs), semivolatiles, and even higher boiling point
compounds, such as polychlorinated biphenyls (PCBs)
[3][4][5][6]. The technology is not effective in separating
inorganics from the contaminated medium. Volatile metals,
however, may be removed by higher temperature thermal
desorption systems.
Some metals may be volatilized by the thermal desorp-
tion process as the contaminated medium is heated. The
presence of chlorine in the waste can also significantly affect
the volatilization of some metals, such as lead. Normally the
temperature of the medium achieved by the process does not
oxidize the metals present in the contaminated medium [7, p.
85].
The process is applicable for the separation of organics
from refinery wastes, coal tar wastes, wood-treating wastes,
creosote-contaminated soils, hydrocarbon-contaminated soils,
mixed (radioactive and hazardous) wastes, synthetic rubber
processing wastes, and paint wastes [8, p. 2][4][9].
Performance data presented in this bulletin should not be
considered directly applicable to other Superfund sites. A
number of variables, such as the specific mix and distribution
D-l
Printed on Recycled Paper
-------
lotto 1
RCRA Codes for Wastes Treated
by Thermal Desorptton
Wood Treating Wastes K001
Dissolved Air Flotation (DAF) Float K048
Slop Oil Emulsion Solids K049
Heat Exchanger Bundles Cleaning Sludge K050
American Petroleum Institute (API)
Separator Sludge K051
Tank Bottoms (leaded) K052
Table 2
Effectiveness of Thernxil Desorption on
General Contaminant Groups for Soil,
Sludge, Sediments, and Filter Cakes
-H
S
6
|
f
£
g
1
T
0
Contaminant Croups
Halogenated volatiles
Halogenated semivolatiles
Nonhalogenated volatiles
Nonhalogenated semivolatiles
PCBs
Pesticides
Oioxins/Furans
Organic cyanides
Organic corrosives
Volatile metals
Nonvolatile metals
Asbestos
Radioactive materials
Inorganic corrosives
Inorganic cyanides
Oxidizers
Reducers
Demonstrated Effectiveness: Succe
completed
Potential Effectiveness: Expert opini
No Expected Effectiveness: Expert
work
Soil
V
Q
m
Q
Q
Q
Q
0
D
Q
ssful tn
on thai
opinio
flfec
Sludge
T
T
T
V
V
T
T
V
Q
T
Q
Q
Q
Q
Q
Q
Q
aubifity
technc*
n that U
tfvenea
Sedi-
ments
T
Q
T
Q
Q
Q
Q
Q
Q
Q
test at son
agywill we
•chnology
fitter
Cakes
T
T
O
T
0 .
Q
Q
Q
Q
O
O
ne scale
ric
will not
of contaminants, affect system performance. A thorough
characterization of the site and a well-designed and con-
ducted treatability study are highly recommended.
Table 1 lists the codes for the specific Resource Conserva-
tion and Recovery Act (RCRA) wastes that have been treated
by this technology [8, p. 2][4][9J. The indicated codes were
derived from vendor data where the objective was to deter-
mine thermal desorption effectiveness for these specific in-
dustrial wastes. The effectiveness of thermal desorption on
general contaminant groups for various matrices is shown in
Table 2. Examples of constituents within contaminant groups
are provided in Technology Screening Guide For Treatment
of CERCLA Soils and Sludges" [7, p. 10]. This table is based on
the current available information or professional judgment
where no information was available. The proven effectiveness
of the technology for a particular site or waste does not ensure
that it will be effective at all sites or that the treatment
efficiencies achieved will be acceptable at other sites. For the
ratings used for this table, demonstrated effectiveness means
that, at some scale, treatability was tested to show the tech-
nology was effective for that particular contaminant and me-
dium. The ratings of potential effectiveness or no expected
effectiveness are both based upon expert judgment Where
potential effectiveness is indicated, the technology is believed
capable of successfully treating the contaminant group in a
particular medium. When the technology is not applicable or
will probably not work for a particular combination of con-
taminant group and medium, a no expected effectiveness
rating is given. Another source of general observations and
average removal efficiencies for different treatability groups is
contained in the Superfund Land Disposal Restrictions (LDR)
Guide #6A, "Obtaining a Soil and Debris Treatability Variance
for Remedial Actions," (OSWER Directive 9347.3-06FS, Sep-
tember 1990) [10] and Superfund LDR Guide #6B, "Obtain-
ing a Soil and Debris Treatability Variance for Removal Ac-
tions," (OSWER Directive 9347.3-06BFS, September 1990)
[11].
Limitations
The primary technical factor affecting thermal desorption
performance is the maximum bed temperature achieved. Since
the basis of the process is physical removal from the medium
by volatilization, bed temperature directly determines which
organics will be removed.
The contaminated medium must contain at least 20 per-
cent solids to facilitate placement of the waste material into
the desorption equipment [3, p. 9]. Some systems specify a
minimum of 30 percent solids [12, p. 6].
As the medium is heated and passes through the kiln or
desorber, energy is lost in heating moisture contained in the
contaminated soil. A very high moisture content can result in
low contaminant volatilization or a need to recycle the soil
through the desorber. High moisture content, therefore,
causes increased treatment costs.
Material handling of soils that are tightly aggregated or
largely clay, or that contain rock fragments or particles greater
than 1-1.5 inches can result in poor processing performance
due to caking. Also, if a high fraction of fine silt or day exists
in the matrix, fugitive dusts will be generated [7, p. 83] and a
greater dust loading will be placed on the downstream air
pollution control equipment [12, p. 6].
The treated medium will typically contain less than 1
percent moisture. Dust can easily form in the transfer of the
treated medium from the desorption unit but can be mitigated
by water sprays. Normally, clean water from air pollution
control devices can be used for this purpose.
Although volatile organics are the primary target of the
thermal desorption technology, the total organic loading is
limited by some systems to up to 10 percent or less [13, p. II-
D-2
Bulletin:
-------
30}. As in most systems that use a reactor or other equipment
to process wastes, a medium exhibiting a very high pH (greater
than 11) or very low pH (less than 5) may corrode the system
components [7, p. 85].
There is evidence with some system configurations that
polymers may foul and/or plug heat transfer surfaces [3, p. 9].
Laboratory/field tests of thermal desorption systems have
documented the deposition of insoluble brown tars (presum-
ably phenolic tars) on internal system components [14, p.
76]. '
High concentrations of inorganic constituents and/or
metals will likely not be effectively treated by thermal desorp-
tion. The maximum bed temperature and the presence of
chlorine can result in volatilization of some inorganic constitu-
ents in the waste, however.
Technology Description
Thermal desorption is any of a number of processes that
use either indirect or direct heat exchange to vaporize organic
contaminants from soil or sludge. Air, combustion gas, or
inert gas is used as the transfer medium for the vaporized
components. Thermal desorption systems are physical sepa-
ration processes and are not designed to provide high levels
of organic destruction, although the higher temperatures of
some systems will result in localized oxidation and/or pyroly-
sis. Thermal desorption is not incineration, since the destruc-
tion of organic contaminants is not the desired result The
bed temperatures achieved and residence times designed
into thermal desorption systems will volatilize selected con-
taminants, but typically not oxidize or destroy them. System
performance is typically measured by comparison of untreated
soil/sludge contaminant levels with those of the processed
soil/sludge. Soil/sludge is typically heated to 200 -1000° F,
based on the thermal desorption system selected.
Figure 1 is a general schematic of the thermal desorption
process.
Waste material handling (1) requires excavation of the
contaminated soil or sludge or delivery of filter cake to the
system. Typically, large objects greater than 1 j inches are
screened from the medium and rejected. The medium is then
delivered by gravity to the desorber inlet or conveyed by
augers to a feed hopper [8, p. 1].
Significant system variation exists in the desorption step
(2). The dryer can be an indirectly fired rotary asphalt kiln, a
single (or set of) internally heated screw auger(s), or a series of
externally heated distillation chambers. The latter process
uses annular augers to move the medium from one volatiliza-
tion zone to the next. Additionally, testing and demonstration
data exist for a fluidized-bed desorption system [12].
The waste is intimately contacted with a heat transfer
surface, and highly volatile components (including water) are
driven off. An inert gas, such as nitrogen, may be injected in a
countercurrent sweep stream to prevent contaminant com-
bustion and to vaporize and remove the contaminants [8, p.
1][4]. Other systems simply direct the hot gas stream from
the desorption unit [3, p. 5][5j.
The actual bed temperature and residence time are the
primary factors affecting performance in thermal desorption.
These parameters are controlled in the desorption unit by
using a series of increasing temperature zones [8, p. 1 ], mul-
tiple passes of the medium through the desorber where the
operating temperature is sequentially increased, separate
compartments where the heat transfer fluid temperature is
higher, or sequential processing into higher temperature zones
[15][16]. Heat transfer fluids used to date include hot com-
bustion gases, hot oil, steam, and molten salts.
Offgas from desorption b typically processed (3) to re-
move particulates. Volatiles in the offgas may be burned in an
afterburner, collected on activated carbon, or recovered in
condensation equipment The selection of the gas treatment
system will depend on the concentrations of the contaminants,
cleanup standards, and the economics of the offgas treat-
ment system(s) employed.
Figure 1
Schematic Diagram of Thermal Desorption
r
r
Gas Treatment
System
(3)
L-^
Clean Offgas
Spent
~^~ Cartoon
Concentrated Contaminants
Water
Oversized Rejects
Treated
Medium
Bulletin- ThprrnalPgsorr>f>'on
D-3
-------
Process Residuals
Operation of thermal desorption systems typically cre-
ates up to six process residual streams: treated medium,
oversized medium rejects, condensed contaminants and wa-
ter, paniculate control system dust, clean offgas, and spent
carbon (if used). Treated medium, debris, and oversized
rejects may be suitable for return onsite.
Condensed water may be used as a dust suppressant for
the treated medium. Scrubber purge water can be purified
and returned to the site wastewater treatment facility (if
available), disposed to the sewer [3, p. 8] [8, p. 2] [4, p. 2], or
used for rehumidification and cooling of the hot, dusty me-
dia. Concentrated, condensed organic contaminants are
containerized for further treatment or recovery.
Dust collected from paniculate control devices may be
combined with the treated medium or, depending on analy-
ses for carryover contamination, recycled through the des-
orption unit.
Clean offgas is released to the atmosphere. If used, spent
carbon may be recycled by the original supplier or other such
processor.
Site Requirements
Thermal desorption systems are transported typically on
specifically adapted flatbed semitrailers. Since most systems
consist of three components (desorber, paniculate control,
and gas treatment), space requirements on site are typically
less than 50 feet by 150 feet, exclusive of materials handling
and decontamination areas.
Standard 440V, three-phase electrical service is needed.
Water must be available at the site. The quantity of water
needed is vendor and site specific.
Treatment of contaminated soils or other waste materials
require that a site safety plan be developed to provide for
personnel protection and special handling measures. Storage
should be provided to hold the process product streams until
they have been tested to determine their acceptability for
disposal or release. Depending upon the site, a method to
store waste that has been prepared for treatment may be
necessary. Storage capacity will depend on waste volume.
Onsite analytical equipment capable of determining site-
specific organic compounds for performance assessment make
the operation more efficient and provide better information
for process control.
Performance Data
Several thermal desorption vendors report performance
data for their respective systems ranging from laboratory
treatability studies to full-scale operation at designated
Superfund sites [17J[9][18]. The quality of this information
has not been determined. These data are included as a
general guideline to the performance of thermal desorption
equipment, and may not be directly transferable to a specific
Superfund site. Good site characterization and treatability
studies are essential in further refining and screening the
thermal desorption technology.
Chem Waste Management's (CWM's) X'TRAX™ System
has been tested at laboratory and pilot scale. Pilot tests were
performed at CWM's Kettleman Hills facility in California.
Twenty tons of PCB- and organic-contaminated soils were
processed through the 5 TPD pilot system. Tables 3 and 4
present the results of PCS separation from soil and total
hydrocarbon emissions from the system, respectively [4].
During a non-Superfund project for the Department of
Defense, thermal desorption was used in a full-scale demon-
stration at the Tinker Air Force Base in Oklahoma. The success
of this project led to the patenting of the process by Weston
Services, Inc. Since then, Weston has applied its low-tem-
perature thermal treatment (LT3) system to various contami-
nated soils at bench-scale through full-scale projects [19].
Table 5 presents a synopsis of system and performance data
for a full-scale treatment of soil contaminated with No. 2 fuel
oil and gasoline at a site in Illinois.
Canonie Environmental has extensive performance data
for its Low Temperature Thermal Aeration (LTTA5*1) system at
full-scale operation (15-20 cu. yds. per hour). The LTTA5*1 has
been applied at the McKin (Maine), Ottati and Coss (New
Hampshire) and Cannon Engineering Corp. (Massachusetts)
Superfund sites. Additionally, the LTTA*" has been used at
the privately-funded site in South Kearney (New jersey). Table
Table 3
PCB Contaminated Soils
PitotXTRAX™[4]
Matrix
Clay
SiltyClay
Clay
Sandy
day
rCO0
(ppm)
5,000
2,800
1,600
1,480
630
Product
(ppm)
24
19
4.8
8.7
17
Removal
<*;
99.3
99.5
99.7
99.1
97.3
Table 4
HtotXTRAX™
TSCA Testing - Vent Emissions [4]
Total Hydrocarbons
(ppm-V)
Before
Carbon
1,320
1,031
530
2,950
2,100
After
Carbon
57
72
35
170
180
Removal
(%)
95.6
93.0
93.3
94 J!
91.4
VOC
(Ux/day)
0.02
0.03
0.01
0.07
0.08
fee-
(mg/m3)
<0.00056
-------
6 presents a summary of Canonic LTTA5" data [5]. The Can-
non Engineering (Mass) site, which was not included in Table
6, successfully treated a total of 11,330 tons of soil, containing
approximately 1803 Ibs. of VOC [20].
T.D.I. Services, Inc. has demonstrated its HT-5 Thermal
Distillation Process at pilot- and full-scale for a variety of RCRA-
listed and other wastes that were prepared to simulate Ameri-
can Petroleum Institute (API) refinery sludge [8]. The com-
pany has conducted pilot- and full-scale testing with the API
sludge to demonstrate the system's ability to meet Land Ban
Disposal requirements for K048 through K052 wastes. Inde-
pendent evaluation by Law Environmental confirms that the
requirements were met, except for TCLP levels of nickel,
which were blamed on a need to "wear-in" the HT-5 system
[21.p.fi].
Remediation Technologies, Inc. (ReTec) has performed
numerous tests on RCRA-listed petroleum refinery wastes.
Table 7 presents results from treatment of refinery vacuum
Tables
Full-Scale Performance Results
for the IP System [19]
Contaminant
Benzene
Toluene
Xylene
Ethyl benzene
Napthalene
Carcinogenic
Priority PNAs
Non-carcinogenic
Priority PNAs
Soil Range
(ppb)
1000
24000
110000
20000
4900
<6000
890-6000
Treated Range
CPPW
5.2
5.2
<1.0
4.8
<330
<330-590
<330-450
Range of
Removal
Efficiency
99.5
99.9
>99.9
99.9
>99.3
<90.2-94.5
<62.9-94.5
Table 6
Summary Results of the LTTA5"
Full-Scale Cleanup Tests [5]
Site
S. Kearney
McKin
Ottatifc
Goss
Processed
16000 tons
>9500 cu yds
2000 cu yds
4500 cu yds
Contam-
inant
VOCs
PAHS
VOCs
PAHS
VOCs
Soil
(ppm)
177.0 (avg.)
35.31 (avg.)
ND-3310
1500 (avg.)
Treated
(ppm)
0.87 (avg.)
10.1 (avg.)
ND-0.04
<10
<0.2 (avg.)
Table?
ReTec Treatment Results-Refinery
Vacuum Filter Cake (A) [3]
Compound
Naphthalene
Acenaphthyiene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthrene
Pyrene
Beruo(b)anthracene
Chtysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(a)pyrene
Dibenz(ab)antracene
Benzo(ghi)perylene
lndeno(1 2398.9
>99.3
>96.6
>99.8
>99.9
97.5
>99.9
97.9
98.4
98.9
97.8
96.6
96.6
Table 8
ReTec Treatment Results-Creosote
Contaminated Clay [3]
Engineering Bulletin: Thermal Desorption
D-5
Compound
Naphthalene
Acenaphthyiene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthrene
Pyrene
Benzo(b)anthracene
Chtysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(a)pyrene
Dibenzo(ab)anthracene
Benzo(ghi)peiylene
lndeno(1 2399.9
<0.1 -
<0.1 >99.96
<0.1 >99.96
1 .6 99.6
<0.1 >99.7
1.5 99.7
2.0 99.6
<0.1 >99.99
<0.1 >99.8
2.5 82.3
<0.1 >99.8
<0.1 >99.9
<0.1 >99.4
<0.1 >99.3
-------
Table?
ReTec Treatment Results-Coal Tar
Contaminated Soils [3]
Compound
Benzene
Toluene
Ethylbenzene
Xylenes
Naphthalene
Fluorene
Phenanthrene
Anthracene
Fluoranthrene
Pyrene
Benzo(b)anthracene
Chrysene
Benzo(b)fluoranthene
Benzo(k}nuoranthene
Benzo(a)pyrene
1 Benzo(ghi)perylene
lndeno(l 23-cd)pyrene
Treatment Temperature:
Original
Sampk
(pom)
1.7
2.3
1.6
6.3
367
114
223
112
214
110
56
58
45
35
47
24
27
450°F
Treated
Sample
(pom)
<0.1
<0.1
<0.1
<0.3
<1.7
<0.2
18
7.0
15
11
<1.4
3.7
<1.4
<2.1
<0.9
<1.1
<6.2
Removal
Efficiency
(%)
>94
>95
>93
>95
>99
>99
91.9
93.8
93.0
90.0
>97
93.6
>97
>94
>98
>95
>77
filter cake. Tests with creosote-contaminated clay and coal
tar-contaminated soils showed significant removal efficiencies
(Tables 8 and 9). All data were obtained through use of
ReTec's 100 Ib/h pilot scale unit processing actual industrial
process wastes [3].
Recycling Sciences International, Inc. (formerly American
Toxic Disposal, Inc.) has tested its Desorption and Vaporiza-
tion Extraction System (DAVES), formerly called the Vaporiza-
tion Extraction System (VES), at Waukegan Harbor, Illinois.
The pilot-scale test demonstrated PCB removal from material
containing up to 250 parts per million (ppm) to levels less
than 2 ppm [12].
RCRA LDRs that require treatment of wastes to best dem-
onstrated available technology (BOAT) levels prior to land
disposal may sometimes be determined to be applicable or
relevant and appropriate requirements for CERCLA response
actions. Thermal desorption can produce a treated waste
that meets treatment levels set by BDAT but may not reach
these treatment levels in all cases. The ability to meet re-
quired treatment levels is dependent upon the specific waste
constituents and the waste matrix. In cases where thermal
desorption does not meet these levels, it still may, in certain
situations, be selected for use at the site if a treatability
variance establishing alternative treatment levels is obtained.
Treatability variances are justified for handling complex soil
and debris matrices. The following guides describe when and
how to seek a treatability variance for soil and debris:
Superfund LDR Guide #6A, "Obtaining a Soil and Debris
Treatability Variance for Remedial Actions" (OSWER Directive
9347.3-06FS, September 1990) [10], and Superfund LDR Guide
#6B, "Obtaining a Soil and Debris Treatability Variance for
Removal Actions" (OSWER Directive 9347.3-06BFS, Septem-
ber 1990) [11]. Another approach could be to use other
treatment techniques in series with thermal desorption to
obtain desired treatment levels.
Technology Status
Significant theoretical research is ongoing [22][23], as
well as direct demonstration of thermal desorption through
both treatability testing and full-scale cleanups.
A successful pilot-scale demonstration of Japanese soils
"roasting" was conducted in 1980 for the recovery of mercury
from highly contaminated (up to 15.6 percent) soils at a plant
site in Tokyo. The high concentration of mercury made
recovery and refinement to commercial grade (less than 99.99
percent purity) economically feasible [24].
In this country, thermal desorption technologies are the
selected remedies for one or more operable units at fourteen
Superfund sites. Table 10 lists each site's location, primary
contaminants, and present status [1][2].
Most of the hardware components of thermal desorption
are available off the shelf and represent no significant problem
of availability. The engineering and configuration of-the
systems are similarly refined, such that once a system is de-
signed full-scale, little or no prototyping or redesign is required.
On-line availability of the full-scale systems described in
this bulletin is not documented. However, since the ex situ
system can be operated in batch mode, it is expected that
component failure can be identified and spare components
fitted quickly for minimal downtime.
Several vendors have documented processing costs per
ton of feed processed. The overall range varies from S80 to
$350 per ton processed [6][4, p. 12][5][3, p. 9]. Caution is
recommended in using costs out of -context because the base
year of the estimates vary. Costs also are highly variable due
to the quantity of waste to be processed, term of the reme-
diation contract, moisture content, organic constituency of
the contaminated medium, and cleanup standards to be
achieved. Similarly, cost estimates should include such items
as preparation of Work Plans, permitting, excavation, pro-
cessing itself, QA/QC verification of treatment performance,
and reporting of data.
EPA Contact
Technology-specific questions regarding thermal desorp-
tion may be directed to:
Michael Gruenfeld
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
Releases Control Branch
2890 Woodbridge Ave.
Bldg. 10(MS-104)
Edison, NJ 08837
FTS 340-6625 or (908) 321-6625
D-6
Bulletin: Ttyp-mni Qesorption Treatment
-------
Table 10
Superfund Sites Specifying Thermal Desorption as the Remedial Action
Site
Cannon Engineering
(Bridgewater Site)
McKin
Ottati&Goss
Wide Beach
Metaltec/Aerosystems
Caldwdl Trucking
Outboard Marine/
Waukegan Harbor
Reich Farms
Re-Solve
Waldick Aerospace
Devices
Wamchern
Fulton Terminals
Stauffer Chemical
| Stauffer Chemical
Location
Bridgewater, MA (1)
McKin, ME (1)
New Hampshire (1)
Brandt, NY (2)
Franklin Borough, N| (2)
Fairfield, NJ (2)
Waukegan Harbor, IL (5)
Dover Township, N| (02)
North Dartmouth, MA (1 )
New Jersey (2)
Burton, SC (4)
Fulton, NY (2)
Cold Creek, AL (4)
LeMoyne,AL(4)
Primary Contaminants
VOCs (Benzene, TCE &
Vinyl Chloride)
VOCs (TCE, BTX)
VOCs (TCE; PCE; 1, 2-DCA,
and Benzene)
PCBs
TCE and VOCs
VOCs (TCE, PCE, and TCA)
PCBs
VOCs and Semivolatiles
PCBs
TCE and PCE
BTX and SVOCs
(Naphthalene)
VOCs (Xylene, Styrene, TCE,
Ethylbenzene, Toluene) and
somePAHs
VOCs and pesticides
VOCs and pesticides
i
Status i
Project completed 1 0/90 ;
Project completed 2/87 :
Project completed 9/89
In design
• pilot study available 5/91 (
In design
• remedial design complete
• remediation starting Fall '91
In design
In design
• treatability studies complete
Pre-design
In design
• pilot study June/Jury '91
In design
i
j
In design '
• pilot study available 5/91
Pre-design I
j
!
Pre-design
Pre-design ,
Acknowledgements
This bulletin was prepared for the U.S. Environmental
Protection Agency, Office of Research and Development
(ORD), Risk Reduction Engineering Laboratory (RREL), Cin-
cinnati, Ohio, by Science Applications International Corpora-
tion (SAIQ under contract no. 68-C8-0062. Mr. Eugene
Harris served as the EPA Technical Project Monitor. Mr. Gary
Baker (SAIC) was the Work Assignment Manager and author
of this bulletin. The author is especially grateful to Mr. Don
Oberacker, Ms. Pat Lafomava, and Mr. Paul de Percin of EPA,
RREL, who have contributed significantly by serving as tech-
nical consultants during tfie, development of this document.
Thermal Desorption
The following other Agency and contractor personnel
have contributed their time and comments by participating in
the expert review meetings and/or peer reviewing the docu-
ment:
Dr. James Cudahy
Mr. James Cummings
Dr. Steve Lanier
Focus Environmental, Inc.
EPA-OERR
Energy and Environmental
Research Corp.
Ms. Evelyn Meagher-Hartzell SAIC
Mr. James Rawe SAIC
Ms. Tish Zimmerman EPA-OERR
0-7
-------
REFERENCES
1. Innovative Treatment Technologies: Semi-Annual Status
Report, EPA/540/2-91/001, U.S. Environmental Protec-
tion Agency, Technology Innovation Office, Jan. 1991.
2. Personal communications with various EPA Regional
Project Managers, April, 1991.
3. Abrishamian, Ramin. Thermal Treatment of Refinery
Sludges and Contaminated Soils. Presented at Ameri-
can Petroleum Institute, Orlando, Florida, 1990.
4. Swanstrom, C, and C. Palmer. X'TRAX™ Transportable
Thermal Separator for Organic Contaminated Solids.
Presented at Second Forum on Innovative Hazardous
Waste Treatment Technologies: Domestic and Interna-
tional, Philadelphia, Pennsylvania, 1990.
5. Canonic Environmental Services Corp, Low Temperature
Thermal Aeration (LTTA*") Marketing Brochures, circa
1990.
6. Nielson, R., and M. Cosmos. Low Temperature Thermal
Treatment (LT3) of Volatile Organic Compounds from
Soil: A Technology Demonstrated. Presented at the
American Institute of Chemical Engineers Meeting,
Denver, Colorado, 1988.
7. Technology Screening Guide for Treatment of CERCLA
Soils and Sludges. EPA/540/2-88/004, U.S. Environmen-
tal Protection Agency, 1988.
8. T.D.I. Services, Marketing Brochures, circa 1990.
9. Cudahy, J., and W. Troxler. 1990. Thermal Remedia-
tion Industry Update - H. Presented at Air and Waste
Management Association Symposium on Treatment of
Contaminated Soils, Cincinnati, Ohio, 1990.
10. Superfund LDR Guide #6A: (2nd Edition) Obtaining a
Soil and Debris Treatability Variance for Remedial
Actions. Superfund Publication 9347.3-06FS, U.S.
Environmental Protection Agency, 1990.
11. Superfund LDR Guide #6B: Obtaining a Soil and Debris
Treatability Variance for Removal Actions. Superfund
Publication 9347.3-06BFS, U.S. Environmental Protec-
tion Agency, 1990.
12. Recycling Sciences International, Inc., DAVES Marketing
Brochures, circa 1990.
13. The Superfund Innovative Technology Evaluation
Program — Progress and Accomplishments Fiscal Year
1989, A Third Report to Congress, EPA/540/5-90/001,
U.S. Environmental Protection Agency, 1990.
14. Superfund Treatability Clearinghouse Abstracts. EPA/
540/2-89/001, U.S. Environmental Protection Agency,
1989.,
15. Soil Tech, Inc., AOSTRA - Taciuk Processor Marketing
Brochure, circa 1990.
16. Ritcey, R., and F. Schwartz. Anaerobic Pyrolysis of
Waste Solids and Sludges — The AOSTRA Taciuk Process
System. Presented at Environmental Hazards Confer-
ence and Exposition, Seattle, Washington, 1990.
17. The Superfund Innovative Technology Evaluation
Program: Technology Profiles. EPA/540/5-89/013, U.S.
Environmental Protection Agency, 1989.
18. Johnson, N., and M. Cosmos. Thermal Treatment
Technologies for Haz Waste Remediation. Pollution
Engineering, XXI(II): 66-85,1989.
19. Weston Services, Inc, Project Summaries (no date).
20. Canonic Environmental Services Corporation, Draft
Remedial Action Report - Cannons Bridgewater
Superfund Site, February 1991.
21. Onsite Engineering Report for Evaluation of the HT-5
High Temperature Distillation System for Treatment of
Contaminated Soils — Treatability Test Results for a
Simulated K051 API Separator Sludge, Vol 1: Executive
Summary, Law Environmental, 1990.
22. LJghty,)., et al. The Cleanup of Contaminated Soil by
Thermal Desorption. Presented at Second International
Conference on New Frontiers for Hazardous Waste
Management EPA/600/9-87/018f, U.S. Environmental
Protection Agency, 1987. pp. 29-34.
23. Fox, R., et al. Soil Decontamination by Low-Tempera-
ture Thermal Separation. Presented at the DOE Model
Conference, Oak Ridge, Tennessee, 1989.
24. Ikeguchi, T., and S. Cotoh. Thermal Treatment of
Contaminated Soil with Mercury. Presented at Demon-
stration of Remedial Action Technologies for Contami-
nated Land and Groundwater, NATO/CCMS Second
International Conference, Bilthoveh, the Netherlands,
1988. pp. 290-301.
Rtillaiin-
-------
APPENDIX E - ARTICLE ON SOIL VAPOR EXTRACTION
(Source: Pedersen, T.A., and J.T. Curtis. Handbook on Soil Vapor
Extraction Technology. EPA/540/2-91-003. February 1991.
A PRACTICAL APPROACH TO THE DESIGN, OPERATION, AND MONITORING
OF IN-SITU SOIL VENTING SYSTEMS
P. C. Johnson', M. W. Kemblowski», J. D. Colthart*.
D. L. Byers', and C. C. Stanley1"
INTRODUCTION
When operated properly, in-situ soil venting or vapor extraction can be
one of the more cost-effective remediation processes for soils contaminated
with gasoline, solvents, or other relatively volatile compounds. A "basic"
system, such as that shown in Figure E-l, couples vapor extraction (recovery)
wells with blowers or vacuum pumps to remove vapors from the vadose zone and
thereby reduce residual levels of soil contaminants. More complex systems
incorporate trenches, air injection wells, passive wells, and surface seals.
Above-ground treatment systems condense, adsorb, or incinerate vapors; in some
cases vapors are simply emitted to the atmosphere through diffuser stacks.
In-situ soil venting is an especially attractive treatment option because the
soil is treated in place, sophisticated equipment is not required, and the
cost is typically lower than other options.
The basic phenomena governing the performance of soil venting systems
are easily understood. By applying a vacuum and removing vapors from
extraction wells, vapor flow through the unsaturated soil zone is induced.
Contaminants volatilize from the soil matrix and are swept by the carrier gas
flow (primarily air) to the extraction wells or trenches. Many complex
processes occur on the microscale, however, the three main factors that
control the performance of & venting operation are the chemical composition of
the contaminant, vapor flowrates through the unsaturated zone, and the
flowpath of carrier vapors relative to the location of the contaminants.
The components of soil venting systems are typically off-the-shelf
items, and the installation of wells and trenches can be done by most
reputable environmental firms. However, the design, operation, and monitoring
of soil venting systems is not trivial. In fact, choosing whether or not
•Shell Development/bShell Oil Company, Westhollow Research Center
Houston, TX 77152-1380
E-l
-------
Vapor Treatment
Unit
£
Vacuum
Pump
I
1
£T^=
Vapor Extraction Well
Vapor
Flow
Contaminated
Soil
Vapor
Row
Eree-Liquid
Hydrocarbon
Groundwater Table
Figure E-l. "Basic" In-Situ Soil Venting Syst
em
E-2
-------
venting should be applied at a given site is a difficult question in itself
If one decides to utilize venting, design questions involving the number of
wells, well spacing, well location, well construction, and vapor treatment
systems must then be answered. It is the current state-of-the-art that such
questions are answered more by instinct than by rigorous logic. This is
evidenced by the published soil venting "success stories" (see Hutzler et al *
for a good review), which rarely include insight into the design process.
In this paper we suggest a series of. steps and questions that must be
followed and answered in order to a) decide if venting is appropriate at a
given site, and b) to design cost-effective in-situ soil venting systems.
This series of steps and questions forms a "decision tree" process that could
be easily incorporated in a PC-based expert system. In the development of
this approach we will attempt to identify the limitations of in-situ soil
venting, and subjects-or behavior that are difficult to quantify and for which
future study is needed.
THE "PRACTICAL APPROACH"
Figure E-2 presents a flowchart of the process discussed in this paper.
Each step of the flowchart is discussed below in detail, and where
appropriate, examples are given.
The Site Investigation
Whenever a soil contamination problem is detected or suspected, a site
investigation is conducted to characterize and delineate the zone of soil and
groundwater contamination. Often the sequence of steps after initial response
and abatement is as follows:
(a) background review: Involves assembling historical records, plot
plans, engineering drawings (showing utility lines), and interviewing site
personnel. This information is used to help identify the contaminant, probable
source of release, zone of contamination, and potentially impacted areas
(neighbors, drinking water supplies, etc.).
(b) preliminary site screening: Preliminary screening tools such as
soil-gas surveys and cone penetrometers are used to roughly define the zone of
contamination and the site geology. Knowledge of site geology is essential to
determine probable migration of contaminants through the unsaturated zone.
(c) detailed site characterization: Soil borings are drilled and
monitoring wells are installed based on the results from steps (a) and (b).
(d) contaminant characterization: soil and groundwater samples are
analyzed to determine contaminant concentrations and compositions.
Costs associated with site investigations can be relatively high
depending on the complexity of the site and size of the spill or leak. For
large spills and complex site geological/hydrogeological conditions, site
E-3
-------
Process
Leak or Spill Discovered
Site
Investigation
1
Screen Treatment
Alternatives
No
Air Permeability Test
Groundwater Pump Test
No
Yes
System Design
System Operation
& Monitoring
Yes
(target levels based on
exposure assessment)
Output
Site Characteristics:
• soil stratigraphy
• characteristics of distinct soil layers
(permeability estimates)
• depth to groundwater & gradient
• aquifer permeability (estimate)
• residual levels of contaminants
• distribution of contaminant
• composition of contaminant
• soil &. above-ground temperature
• soil vapor concentrations (optional)
1 removal rate estimates
1 vapor flowrate estimates
' final residual levels & composition
• air permeability of distinct, soil layers
1 radius of influence of vapor wells
1 initial vapor concentrations
• aquifer properties (gradient,
transmissiviry, storativiry)
• number of vapor extraction wells
• vapor well construction
• vapor well spacing
* vapor treatment system
• flowrate (vacuum) specifications
• groundwater pumping system specifics
• venting recovery rates
• changes in vadose zone contamination
• "clean" site
System Shut-Off
Figure E-2. In-Situ Soil Venting System Design Process,
E-4
-------
investigation costs are often comparable to remediation costs. In addition
the choice and design of a remediation system is based on the data obtained'
during the site investigation. For these reasons it is important to insure
that specific information is collected, and to validate the quality of the
data.
If it is presumed that in-situ soil venting will be a candidate for
treatment, then the following information needs to be obtained during the
preliminary site investigation:
(a) sice geology - this includes soil type and subsurface stratigraphy.
While they are not essential, the moisture content, total organic carbon, and
permeability of each distinct soil layer also provides useful information chac
can be used to choose and design a remediation system.
(b) sice hydrogeology - the water table depth and gradient must be
known, as well as estimates of the aquifer permeability.
(c) contaminant composition, distribution and residual levels - soil
samples should be analyzed to determine which contaminants are present at what
levels. Recommended analytical methods should be used to identify target
compounds (i.e., benzene, toluene, or xylenes) and total hydrocarbons present.
For soil analyses these methods are:
EPA 8240 - volatile organic chemicals
EPA 8270 - semi-volatile organic chemicals
EPA 418.1 - total petroleum hydrocarbons
The corresponding water analyses methods are:
EPA 624 - volatile organic chemicals
EPA 625 - semi-volatile organic chemicals
EPA 418.1 - total petroleum hydrocarbons
With the current high cost of chemical analyses it is important to
intelligently select which analyses should be performed and which samples
should be sent to a certified laboratory. Local regulations usually require
that a minimum number of soil borings be performed, and target compounds must
be analyzed for based on the suspected composition of the contamination.
Costs can be minimized and more data obtained by utilizing field screening
tools, such as hand-held vapor meters or portable field GC's. These
instruments can be used to measure both residual soil contamination levels and
headspace vapors above contaminated soils. At a. minimum, soil samples
corresponding to lithology changes or obvious changes in residual levels
(based on visual observations or odor) should be analyzed.
For complex contamination mixtures, such as gasoline. diesel fuel, and
solvent mixtures, it is not practical or necessary to identify and quantify
each compound present. In such cases it is recommended that a "boiling point
distribution be measured for a representative sample of the residual
E-5
-------
contamination. Boiling point distribution curves, such as shown in Figure £.3
for "fresh" and "weathered" gasoline samples, can be constructed from GC
analyses of the soil residual contamination (or free-product) and knowledge of
the GC elution behavior of a known series of compounds (such as straight-chain
alkanes). Compounds generally elute from a GC packed column in the order of
increasing boiling point, so a boiling point distribution curve is constructed
by grouping all unknowns that elute between two known peaks (i.e. between
n-hexane and n-heptane). Then they are assigned an average boiling point;
molecular weight, and vapor pressure. Use of this data will be explained
below.
(d) temperature - both above- and below-ground surface.
The cone penetrometer, which is essentially an instrumented steel
rod that is driven into the soil, is becoming a popular tool for
preliminary site screening investigations. By measuring the shear and
normal forces on the leading end of the rod, soil structure, and hence
permeability can be defined. Some cone penetrometers are also
constructed to allow the collection of vapor or groundwater samples.
This tool has several advantages over conventional soil boring
techniques (as a preliminary site characterization tool): the subsurface
soil structure can be defined better, no soil cuttings are generated,
and more analyses can be performed per day.
Results from the preliminary site investigation should be
summarized in contour plots, fence diagrams, and tables prior to
analyses.
Deciding if Venting is Appropriate
As stated above, the three main factors governing the behavior of any
in-situ soil venting operation are the vapor flow rate, contaminant vapor
concentrations, and the vapor flowpath relative to the contaminant location.
In an article by Johnson et al.2 simple mathematical equations were presented
to help quantify each of these factors. Below we illustrate how to utilize
these "screening models" and the information collected during the preliminary
site investigation to help determine if in-situ soil venting is appropriate at
a given site. In making this decision we will answer the following questions:
(1) What contaminant vapor concentrations are likely to be obtained?
(2) Under ideal vapor flov conditions (i.e. 100 - 1000 scfm vapor
flovrates), is this concentration great enough to yield acceptable
removal rates?
(3) What range of vapor flovrates can realistically be achieved?
(4) Will the contaminant concentrations and realistic vapor flovrates
produce acceptable removal rates?
E-6
-------
1.0
Cumulative
Weight 08.
Fraction
0.6^
0.4
0.0
Weathered" Gasoline
-40 0 40 80 120 160 200 240
Tb(°C)
Figure E-3. Boiling Point Distribution Curves for Samples of
"Fresh" and "Weathered" Gasolines.
1-7
-------
(5) What are the vapor composition and concentration changes? What
residual, if any, will be left in the soil?
(6) Are there likely to be any negative effects of soil venting?
Negative answers to questions (2), (3), or (4) will rule out in-situ soil
venting as a practical treatment method.
(1) - What contaminant vapor concentrations are likely to be obtained?
Question (1) can be answered based on the results of soil vapor surveys,
analyses of headspace vapors above contaminated soil samples, or equilibrium
vapor models2. In some cases just knowing which compounds are present is
sufficient to estimate if venting is feasible. In the absence of soil-vapor
survey data, contaminant vapor concentrations can be estimated. The maximum
vapor concentration of any compound (mixture) in extracted vapors is its
equilibrium or "saturated" vapor concentration, which is easily calculated
from knowledge of the compound's (mixture's) molecular weight, vapor pressure
at the soil temperature, residual soil contaminant composition, and the ideal
gas law:
x1
C.,t - I (E-l)
t RT
where:
C.,t - estimate of contaminant vapor concentration [mg/1]
XJL - mole fraction of component i in liquid-phase residual
(xt - 1 for single compound)
P,* - pure component vapor pressure at temperature T [atm]
M^ - molecular weight of component i [mg/mole]
R - gas constant - 0.0821 l-atm/mole-°K
T - absolute temperature of residual [°K]
Table E-l presents data for some chemicals and mixtures often spilled in the
environment. There are more sophisticated equations for predicting vapor
concentrations in soil systems based on equilibrium partitioning arguments,
but these require more detailed information (organic carbon content, soil
moisture) than is normally available. If a site is chosen for remediation,
the residual total hydrocarbons in soil typically exceed 500 mgAg- In this
residual concentration range the majority of hydrocarbons will be present as a
separate or "free" phase, the contaminant vapor concentrations become
independent of residual concentration (but still depend on composition), and
Equation E-l is applicable2. In any case, it should be noted that these are
estimates only for vapor concentrations at the start of venting, which is
when the removal rates are generally greatest. Contaminant concentrations in
the extracted vapors will decline with time due to changes in composition,
residual levels, or increased diffusional resistances. These topics are
discussed below in more detail.
E-8
-------
Table E-l. Selected Compounds and Their Chemical Properties.
Compound
n-pentane
n-hexane
trichloroeinane
benzene
cyclohexane
trichloroethylene
n-heptane
toluene
teaachloroethylene
n-octane
chloro benzene
p-xylene
ethylbenzene
m-xylene
o-xylene
styiene
n-nonane
n-propylbenzcne
1.2,4 trimethylbenzene
n-decane
DBCP
n-undecane
n-dodccanc
napthalene
tetracthyllead
gasoline1
weathered easoiine2
Mw
72.2
86.2
133.4
78.1
84.2
131.5
100.2
92.1
166
114.2
113
106.2
106.2
106.2
106.2
104.1
128.3
120.2
120.2
142.3
263
156.3
170.3
128.2
323
95
111
Tb (1 ami) P
f°O
36
69
75
80
81
87
98
111
121
126
132
138
138
139
144
145
151
159
169
173
196
196
216
218
dec. @200C
.
"
v° (20°C)
(am)
0.57
0.16
0.132
0.10
0.10
0.026
0.046
0.029
0.018
0.014
0.012
0.0086
0.0092
0.0080
0.0066
0.0066
0.0042
0.0033
0.0019
0.0013
0.0011
0.0006
0.00015
0.00014
0.0002
0.34
0.049
* «« A __J &.]_
Csat
(me/I)
1700
560
720
320
340
140
190
110
130
65
55
37
40
35
29
28
22.0
16
9.3
7.6
11
3.8
1.1
0.73
2.6
1300
220
.
v. .. £ 1 4 «t #* vs/\4n^
1 Corresponds to "fresh" 6»—-~~ -
distribution shown in Figure E-3.
dlStriDUtlOn snown in ri$vM.«: s,-... 4H«<, noint
* Corresponds to "weathered" gasoline defined in Table E-2 with boiling point
distribution shown in Figure E-3.
E-9
-------
(2) - Under ideal vapor flow conditions (i.e. 100 - 1000 scfm vapor
flowrates) , is this concentration great enough to yield acceptable
removal rates?
Question (2) is answered by multiplying the concentration estimate CMt,
by a range of reasonable flowrates, Q:
R.,t - Cest Q (E-2)
Here R,,,. denotes the estimated removal rate, and Cest and Q must be expressed
in consistent units. For reference, documented venting operations at service
station sites typically report vapor flowrates in the 10 - 100 scfm range1,
although 100 - 1000 scfm flowrates are achievable for very sandy soils or
large numbers of extraction wells. At this point in the decision process we
are still neglecting that vapor concentrations decrease during venting due to
compositional changes and mass transfer resistances. Figure E-4 presents
calculated removal rates Rest [kg/d] for a range of Cest and Q values. Cejt
values are presented in [mg/1] and [ppmCE4] units, where [ppm^*] represents
me thane -equivalent parts-per-million volume/volume (ppm^) units. The [ppm^]
units are used because field analytical tools that report [ppiaj values are
often calibrated with methane. The [mg/1] and [ppu^] units are related by:
w] * 16000mg-CHt/mole-CH4 * 10"6
[mg/1] -- (E-3)
(0.0821 l-atm/°K-mole) * (298K)
For field instruments calibrated with other compounds (i.e., butane, propane)
[ppmj values are converted to [mg/lj by replacing the molecular weight of CH4
in Equation E-3 by the molecular weight [mg/mole] of the calibration compound.
Acceptable or desirable removal rates RK.^^!.. • can De determined by
dividing the estimated spill mass Msptu, by the maximum acceptable clean-up
time T:
(E-4)
For example, if 1500 kg (*500 gal) of gasoline had been spilled at a service
station and we wished to complete the clean-up within eight months, then
R.cc«pt«bi« - 6-3 kg/d. Based on Figure E-4, therefore, CMt would have to
average >1.5 mg/1 (2400 ppm^) for Q-2800 1/min (100 cfm) if venting is to be
an acceptable option. Generally, removal rates <1 kg/d will be unacceptable
for most spills, so soils contaminated with compounds (mixtures) having
saturated vapor concentrations less than 0.3 mg/1 (450 ppm^) will not be
good candidates for venting, unless vapor flowrates exceed 100 scfm. Judging
from the compounds listed in Table E-l, this corresponds to compounds with
E-10
-------
Removal
Rate
(kg/d)
15300
10
153000
100
Vapor Concentration (mg/1)
* (ppm _, ) - concentration in methane-equivalent ppm (vol/vol.) units
Figure E-4. In-Situ Soil Venting Removal Rate Dependence on
Vapor Extraction Rate and Vapor Concentration.
E-ll
-------
boiling points (Tb)>150°C, or pure component vapor pressures <0.0001 atm
evaluated at the subsurface temperature.
- What range of vapor flovrates can realistically be achieved?
Question (3) requires that we estimate realistic vapor flowrates for our
site specific conditions. Equation E-5, which predicts the flowrate per unit
thickness of well screen Q/H [cm3/s], can be used for this purpose:
Q k
*_Ptf - (E-5)
H M
where :
k - soil permeability to air flow [cm2] or [darcy]
m - viscosity of air - 1.8 x 10"* g/cm-s or 0.018 cp
Pw - absolute pressure at extraction well [g/cm-s2] or [atm]
PAtm - absolute ambient pressure « 1.01 x 106 g/cm-s2 or 1 atm
Rw - radius of vapor extraction well [cm]
Rj - radius of influence of vapor extraction well [cm]
This equation is derived from the simplistic steady-state radial flow solution
for compressible flow2, but should provide reasonable estimates for vapor flow
rates. If we can measure or estimate k, then the only unknown parameter is
the empirical "radius of influence" Rz. Values ranging from 9 m (30 ft) to
30 m (100 ft) are reported in the literature for a variety of soil conditions,
but fortunately Equation E-5 is not very sensitive to large changes in Rz.
For estimation purposes, therefore, a value of Rj-12 m (40 ft) can be used
without a significant loss of accuracy. Typical vacuum well pressures range
from 0.95 - 0.90 atm (20 - 40 in H20 vacuum). Figure E-5 presents predicted
flowrates per unit well screen depth Q/H, expressed in "standard" volumetric
units Q*/H (- Q/H(Pw/PAtm)) for a 5.1 cm radius (4" diameter) extraction well,
and a wide range of soil permeabilities and applied vacuums. Here H denotes
the thickness of the screened interval, which is often chosen to be equal to
the thickness of the zone of soil contamination (this minimizes removing and
treating any excess "clean" air). For other conditions the Q*/H values in
Figure E-5 can be multiplied by the following factors:
R^ - 5.1 cm (2") Rj - 7.6 m (25') - multiply by Q*/H by 1.09
R» - 5.1 cm (2") Rj - 23 m (75') - multiply by Q*/H by 0.90
R. - 7.6 cm (3") Rx - 12 m (40') - multiply by Q*/H by 1.08
R,, - 10 cm (4") Rj - 12 m (40') - multiply by Q*/H by 1.15
R,, - 10 cm (4") Rj - 7.6 m (25') - multiply by Q*/H by 1.27
As indicated by the multipliers given above, changing the radius of influence
from 12 m (40 ft) to 23 m (75 ft) only decreases the predicted flowrate by
10%. The largest uncertainty in flowrate calculations will be due to the air
permeability value k, which can vary by one to three orders of magnitude
across a site and can realistically only be estimated from boring log data
E-12
-------
100
10
(m /m-min)
.1.
.01
.001
.0001
Ru.= 5.!cni(2")
Rj= 12m (40')
P = 0.40 atm = 20.3 ft H,0
W " *•
P = 0.60 atm = 13.6 ft H,0
W
!100
- no
.01
P =0,
W
= 0.80 arm = 6.8 ft H2O
0.90 arm = 3.4 ft H2O
.95 arm =1.7 ft H2O
clayey
sands
fine
sands
sands
coarse
sands
Vapor
Flowrate
(scfm/ft)
- 1.1
_. 0.11
- 0.011
0.0011
.1 1 10 100
Soil Permeabilty (darcy)
1000
[ft H O] denote vacuums expressed as equivalent water column heights
Figure E-5. Predicted Steady-State Flowrates (per unit well screen depth)
for a Range of Soil Permeabilities and Applied Vacuums (Pv).
E-13
-------
within an order of magnitude. It is prudent, therefore, to choose a range of
k values during this phase of the decision process. For example, if boring
logs indicate fine sandy soils are present, then flowrates should be
calculated for k values in the range 0.1
-------
vapor flow
vapor flow
,--—\'
x
/ r-v
vapor flow
side view
top view
b)
vapor flow
vapor concentration
profile
vapor concentration = 0
impermeable layer
liquid contaminant
c)
t
"wet" zone with residual contamination
Figure E-6. Scenarios for Removal Rate Estimates.
E-15
-------
where:
TJ - efficiency relative to maximum removal rate
D - effective soil vapor diffusion coefficient [cm2/s]
M - viscosity of air - 1.8 x 10"* g/cm-s
k - soil permeability to vapor flow [cm2]
H - thickness of screened interval [cm]
R! - radius of influence of venting well [cm]
Rv - venting well radius [cm]
PAUB - absolute ambient pressure - 1.016 x 106 g/cm-s2
Pw - absolute pressure at the venting well [g/cm-s2]
Ri
-------
Figure E-6c depicts the situation in which vapor flows primarily past,
rather than through the contaminated soil zone, such as might be the case for
a contaminated clay lens surrounded by sandy soils. In this case vapor phase
diffusion through the clay to the flowing vapor limits the removal rate. The
maximum removal rate in this case occurs when the vapor flow is fast enough to
maintain a very low vapor concentration at the permeable/impermeable soil
interface. At any time t a contaminant -free or "dried out" zone of low
permeability will exist with a thickness S. An estimate of the removal rate
Rest from a contaminated zone extending from Rx to Rj is:
(E-10)
R.,t - ir(R*- R2)CMtD/«(t)
where D is the effective porous media vapor diffusion coefficient (as
calculated above from Equations E-8 and E-9) and C
-------
1000
R
est
(kg/d)
200 300
Time (d)
200
"Dry" Zone
Thickness
benzene (20 C)
=5.1 cm
R2 = 900 cm
500
Figure E-7.
Estimated Maximum Removal Rates for a Venting Operation
Limited by Diffusion.
E-18
-------
Mixture removal rates for the situations depicted in Figures E-6b and E-
6c are difficult to estimate because changes in composition and liquid-phase
diffusion affect the behavior. Presently there are no simple analytical
solutions for these situations, but we can postulate that they should be less
than the rates predicted above for pure components.
The use of equilibrium-based models to predict required removal rates is
discussed below under the next question.
(5) - What are the vapor composition and concentration changes? What residual,
if any, will be left in the soil?
As contaminants are removed during venting, the residual soil
contamination level decreases and mixture compositions become richer in the
less volatile compounds. Both of these processes result in decreased vapor
concentrations, and hence, decreased removal rates with time. At low
residual soil contamination levels (<500 ppm) Equation E-l becomes less valid
as sorption and dissolution phenomena begin to affect the soil residual -
vapor equilibrium. In the limit of low residual contamination levels,
contaminant equilibrium vapor concentrations are expected to become
proportional to the residual soil contaminant concentrations. As venting
continues and residual soil levels decrease, therefore, it becomes more
difficult to remove the residual contamination. It is important to realize
that, even with soil venting, there are practical limitations on the final
soil contamination levels that can be achieved. Knowledge of these limits is
necessary to realistically set clean-up criteria and design effective venting
systems.
The maximum efficiency of a venting operation is limited by the
equilibrium partitioning of contaminants between the soil matrix and vapor
phases. The maximum removal rate is achieved when the vapor being removed
from an extraction well is in equilibrium with the contaminated soil. Models
for predicting this maximum removal rate have been presented by Marley and
Hoag* and Johnson et al.2 The former considered only compositions in a
residual free-phase, while the latter also considered the effects of sorption
and dissolution processes. A complete discussion of the development of these
models is not appropriate here, but we will discuss use of the predictions.
The change in composition, vapor concentration, removal rate, and
residual soil contamination level with time are functions of the initial
residual composition, vapor extraction well flowrate, and initial soil
contamination level. It is not necessary to generate predictions for every
combination of variables, however, because with appropriate scaling all
results will form a single curve for a given initial mixture e
Figure E-8a presents the results computed with the model presented ^ Johnson
et al 2 for the "weathered" gasoline mixture whose composition is given by
Table'E-2 The important variable that determines residual soil levels vapor
concentrations, aS removal rates is the ratio Qt/M(t-0) which represents the
volume of air drawn through the contaminated zone P« «**•"• « the
contaminant. In Figure E-8, the scaled removal rate (or equivalently the
E-19
-------
Table E-2. Composition of "Fresh" and "Weathered" Gasolines.
Compound Name
propane
isobutane
n-butane
aans-2-butene
cis-2-butene
3-memyl-l-butene
l-pcntcne
2-metnyI-l-butene
2-memyl- IJ-buodiene
n-penooc
craos*2*penicne
2-memyi-2-buiene
t.rrv^^y^l 7-^»ir^ii>n»
33-dim«hyl-l-buiene
ryckipnflitw
3-methyH-pentene
2^3-dimethyibtnane
2-mahylpenane
3-memyiDCBBne
D-hexane
methyicyciopentane
22-dimetnyipenane
benzene
3-metnyihexane
3*cuty ipfjimir.
n-hepBDe
22.4-trinethylpeaaae
meinylcyclohexane
22-diiaetbyihexaae
toluene
2J.4-orimetnytpentane
3-methylhepane
2-flwthyioepciDC
n-ocane
Mw
(V)
44.1
58.1
58.1
56.1
56.1
70.1
72.2
/ 1 MI
70.1
70.1
68.1
722
70.1
70.1
68.1
842
70.1
S62
862
862
862
842
1002
78.1
tAt
O4.X
1002
1002
1AM A
1002
1002
1142
982
1142
92.1
1142
1142
1142
1142
Fresh
Gasoline
0.0001
0.0122
0.0629
0.0007
0.0000
0.0006
0 1049
WB 41^17
0.0000
0.0000
0.0000
O.OS86
0.0000
0.0044
0.0000
0.0049
0.0000
0.0000
0.0730
04ZZ73
0.0000
0.0283
0.0083
0.0076
0.0076
0-0000
04)390
04)000
04)000
04)063
04)121
04)000
04)055
0.0550
04)121
04)000
04)155
00013
Weathered
Gasoline
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0005
04)008
0.0000
0.0095
0.0017
0.0021
0.0010
0.0000
00046
%f»*Ai^^V
0.0000
C.OC44
0.0207
04)186
0.0207
04034
0.0064
04)021
A M«|4*V
04)137
04)000
0X1355
0-0000
0.0447
0.0503
04Q93
04Q07
0.0359
0-fflKW
042343
04)324
0-0300
Approximate
Comoosition
0
0
0
0
0
0
OO1T7
U.wi / /
0
0
0
0
0
0
o
0
00738
w.W / JO
0
0
0
0
0
0
0
0
0.1761
0
0
0
0
0
0
0
0.1926
0
0
0
0
1-20
-------
Table E-2 (continued). Composition of "Fresh" and "Weathered" Gasolines.
2,4,4-trimethylhexane
22-dimetnylhepone
ethylbenzene
p-xylene
m-xylene
33.4-trimethylhexane
o-xylene
22.4-trimethylheptane
n-nonane
33.5-trimeihylheptane
n-propylbenzene
23.4-trimethylheptane
1 3.5-aimethyibenzene
1 2.4-trimethy (benzene
n-decane
methytpiopyibenzene
dimethyieihylbenzene
n^tg^lp^lfff
12,4.5-ienainethyibenzene
123,4-teoanieihylbenzene
1^2.4-trimetnyi-S-etiiylbenzene
ivdodecane
mplialf"^
in c*y iuci uo^
1283
1283
1062
1062
1062
1283
1062
1423
1283
1423
1202
1423
1202
1202
1423
1342
1342
1563
1342
1342
1482
1703
1282
1623
1422
0.0087
0.0000
0.0000
0.0957
0.0000
0.0281
0.0000
0.0105
0.0000
0.0000
0.0841
0.0000
0.0411
0.0213
0.0000
0.0351
0.0307
0.0000
0.0133
0.0129
0.0405
0.0230
0.0045
0.0000
0.0023
0.0034
0.0226
0.0130
0.0151
0.0376
0.0056
0.0274
0.0012
0.0382
0.0000
0.0117
0.0000
0.0493
0.0705
0.0140
0.0170
0.0289
0.0075
0.0056
0.0704
0.0651
0.0000
0.0076
0.0147
0.0134
0
0
0
0
0.1641
0
0
0
0
0
0.1455
0
0
0
0
0
0.0534
0
0
0.1411
0
0
0
0.0357
0
Total
1.0000
1.0000
1.0000
E-21
-------
fl)
QC/QC(t=0) :
.0001
changefl from 4-phasc 10
3-phise system
Weathered Gasoline
T = 20°C
10% moisture content
Ca=0) = 222 mg/1
•80
.01:
.001,
Full Composition
100
% removed
•60
•40
20
100 200
Qt/m(t=0) (1/g)
300
Weathered Gasoline
T = 20°C
10% moisture content
dunged fern 4-ph«e to C(t=0) = 270 mg/1
3-phue system
.001
.0001
Approximate Composidon
100
80
% removed
60
h40
20
100 200
Qt/m(t=0) (1/g)
•o
300
Figure E-8. Maximum Predicted Removal Rates for a Weathered Gasoline.
a) full composition, b) approximate composition.
E-22
-------
vapor concentration) decreases with time as the mixture becomes richer in the
less volatile compounds.
While a detailed compositional analysis was available for this gasoline
sample, an approximate composition based on a boiling point distribution curve
predicts similar results. Figure E-8b presents the results for the
approximate mixture composition also given in Table E-2.
Model predictions, such as those shown in Figure E-8 for the gasoline
sample defined by Table E-2, can be used to estimate removal rates (if the
vapor flowrate is specified), or alternatively the predictions can be used to
estimate vapor flowrate requirements (if the desired removal rate is
specified). For example, if we wanted to reduce the initial contamination
level by 90%, then Figure E-8 predicts that »100 1-air/g-gasoline will be
required. This is the minimum amount of vapor required, because it is based
on an equilibrium-based model. The necessary minimum average vapor flowrate
is then equal to the spill mass times the minimum required vapor flow/mass
gasoline divided by the desired duration of venting. Use of this approach is
illustrated in the service station site example provided at the end of this
paper.
Figure E-8 also illustrates that there is a practical limit to the
amount of residual contaminant that can be removed by venting alone. For
example, it will take a minimum of 100 1-vapor/g-gasoline to remove 90% of the
weathered gasoline defined in Table E-2, while it will take about 200
1-air/g-gasoline to remove the remaining 10%. In the case of gasoline, by the
time 90% of the initial residual has been removed the residual consists of
relatively insoluble and nonvolatile compounds. It is important to recognize
this limitation of venting, and when setting realistic clean-up target levels,
they should be based on the potential environmental impact of the residual
rather than any specific total residual hydrocarbon levels.
(6) - Are there likely to be any negative 'effects of soil venting?
It is possible that venting will induce the migration of off-site
contaminant vapors towards the extraction wells. This is likely to occur at a
service station, which is often in close proximity to other service stations.
If this occurs, one could spend a lot of time and money to unknowingly
clean-up someone else's problem. The solution is to establish a "vapor
barrier" at the perimeter of the contaminated zone. This can be accomplished
by allowing vapor flow into any perimeter groundwater monitoring wells, which
then act as passive air supply wells. In other cases it may be necessary to
install passive air injection wells, or trenches, as illustrated in Figure E-
9a.
As pointed out by Johnson et al.2 the application of a vacuum to
extraction wells can also cause a water table rise. In many cases
contaminated soils lie just above the water table and they become ™ter
saturated as illustrated in Figure E-9b. The maximum rise occurs at the
' well, where thfwater table rise will be equal to the vacuum
E-23
-------
Vapor Extraction
Well
Off-Site
Conmainmion
Passive Air Injection Well
or
Perimeter Groundwater Monitoring Well
b)
Unsaturated
Soil Zone
Vapor Extraction
Well
Saturated
Soil Zone
Water Table Upwelling
by Vacuum
Figure E-9. a) Use of Passive Vapor Wells to Prevent Migration of Off-Site
Contaminant Vapors, b) Water Table Rise Caused by the Applied Vacuum.
E-24
-------
at the well expressed as an equivalent water column height (i.e., in or ft
H20). The solution to this problem is to install a dewatering system, with
groundwater pumping wells located as close to vapor extraction wells as
possible. The dewatering system must be designed to insure that contaminated
soils remain exposed to vapor flow. Other considerations not directly related
to venting system design, such as soluble plume migration control and
free-liquid product yield, will also be factors in the design of groundwater
pumping system.
Design Information
If venting is still a remediation option after answering the questions
above, then more accurate information must be collected. Specifically, the
soil permeability to vapor flow, vapor concentrations, and aquifer
characterics are required. These are obtained by two field experiments: air
permeability and groundwater pump tests. These are described briefly below.
Air Permeability Tests
Figure E-10 depicts the set-up of an air permeability test. The object
of this experiment is to remove vapors at a constant rate from an extraction
well, while monitoring with time the transient subsurface pressure
distribution at fixed points. Effluent vapor concentrations are also
monitored. It is important that the test be conducted properly to obtain
accurate design information. The extraction well should be screened through
the soil zone that will be vented during the actual operation. In many cases
existing groundwater monitoring wells are sufficient, if their screened
sections extend above the water table. Subsurface pressure monitoring probes
can be driven soil vapor sampling probes (for shallow <20 ft deep
contamination problems) or more permanent installations.
Flowrate and transient pressure distribution data are used to estimate
the soil permeability to vapor flow. The expected change in the subsurface
pressure distribution with time P'(r,t) is predicted2 by:
P» . n — dx
4jrm(k/M)
r «
For (r2 /4kPAtmt)<0.1 Equation E-13 can be approximated by:
p. . ___JL_[-0.5772 - ln( ) + ln(t)] (E-14)
4?rm(k/M)
E-25
-------
Pressure
Gauge
Vapor Sampling
Port
Vapor Flowmeter
Vacuum
Pump
Vapor Treatment
Vapor
Flow
Vapor Extraction Well
Contaminated
Soil
Pressure Sampling Probes
Figure E-10. Air Permeability Test System.
E-26
-------
where :
P' - "gauge" pressure measured at distance r and time t
m - stratum thickness
r - radial distance from vapor extraction well
k - soil permeability to air flow
M - viscosity of air - 1.8 x 10'* g/cm-s
6 - air- filled soil void fraction
t - time
Q - volumetric vapor flowrate from extraction well
PAtB - ambient atmospheric pressure - 1.0 atm - 1.013 x 10*
g/cm-s2
Equation E-14 predicts that a plot of P' -vs- ln(t) should be a straight line
with slope A and y- intercept B equal to:
A -- B -- [-0.5772 - ln( - )] (E-15)
4JTm(k/M)
The permeability to vapor flow can then be calculated from the data by one of
two methods. The first is applicable when Q and m are known. The calculated
slope A is used:
k -- (E-16)
4AJTm
The second approach must be used whenever Q or m is not known. In this case
the values A and B are both used:
B
exp( — + 0.5772) (E-17)
Equation E-13 can also be used to choose the locations of subsurface
pressure monitoring points before conducting the air permeability test, given
an estimation of k and the flowrate to be used.
Vapor samples should be taken at the beginning and end of the air
permeability test, which should be conducted for a long enough time to extract
at least one "pore volume" V? of vapor from the contaminated soil zone. This
insures that all vapors existing in the formation prior to venting are
removed. The vapor concentration at the start of the test is representative
of the equilibrium vapor concentration, while the concentration measured after
one pore volume has been extracted gives an indication of realistic removal
rates and the mixing or diffusional limitations discussed in association with
Fi-gure E-6. The time rf for one pore volume to be removed is:
Tp - VP/Q - €AJTR2H/Q (
E-27
-------
where R, H, €A, and Q are the radius of the zone of contamination, vertical
thickness of the zone of contamination, air- filled void fraction, and
volumetric vapor flowrate front the extraction well. For example, consider the
case where R-12 m, H-3 m, €A-0.35, and Q-0.57 m3/min (20 f t3/min) . Then rp-475
m3/0.57 m3/min-833 min-14 h.
Groundwatsr Pmro Tests
To achieve efficient venting the hydrocarbon-contaminated soil has to be
exposed to air flow, which in turn requires that the water table be lowered to
counteract the water up we 11 ing effect caused by the decreased vapor pressure
in the vicinity of a venting well (Johnson et al.2) and to possibly expose
contaminated soil below the water table. Thus the groundwater pumping system
has to have a sufficient pumping rate and be operated for a long enough time
period to obtain the required drawdowns. Since most venting systems are
installed above phreatic aquifers, two aquifer parameters are needed for the
design: average transmissivity T and effective porosity S. These parameters
can be estimated using the results of the standard transient groundwater pump
test with a constant pumping rate (Bear5) . Using the estimated values the
required pumping rate may be calculated as follows:
Q - 4jrTS(r,t)/W(u) (E-19)
where: W(u) is the well function5 of u - SrV4Tt, and s(r,t) is the required
drawdown at distance r and pumping time equal to t.
System Desipn
In this section we discuss the questions that must be answered in order
to design an in-situ soil venting system. It is not our intention to provide
a generic "recipe" for soil venting systems design; instead we suggest a
structured thought process to guide in choosing the number of extraction
wells, well spacing, construction, etc. Even in a. structured thought process,
intuition and experience play important roles. There is no substitute for a
good fundamental understanding of vapor flow processes, transport phenomena,
and groundwater flow.
- Choosing the number of vapor extraction veils
Three methods for choosing the number of vapor extraction wells are
outlined below. The greatest number of wells from these three methods is then
the value that should be used. The objective is to satisfy removal rate
requirements and achieve vapor removal from the entire zone of contamination.
For the first estimate we neglect residual contaminant composition and
vapor concentration changes with time. The acceptable removal rate R»ee.ptabi.
is calculated from Equation E-4, while the estimated removal rate from a
single well R,st is estimated from a choice of Equations E-2, E-6, E-7, or E-
E-28
-------
12 depending on whether the specific site conditions are most like Figure E-
6a, E-6b, or E-6c. The number of wells Nw-u required to achieve the
acceptable removal rate is:
(E-20)
Nw«U "
Equations E-2, E-6, and E-7 require vapor flow estimates, which can be
calculated from Equation E-5 using the measured soil permeability and chosen
extraction well vacuum Ptf. At this point one must determine what blowers and
vacuum pumps are available because the characteristics of these units will
limit the range of feasible (PW,Q) values. For example, a blower that can
pump 100 scfm at 2 in H20 vacuum may only be able to pump 10 scfm at 100 in
H20 vacuum.
The second method, which accounts for composition changes with time,
utilizes model predictions, such as those illustrated in Figure E-8. Recall
that equilibrium-based models are used to calculate the minimum vapor flow to
achieve a given degree of remediation. For example, if we wish to obtain a
90% reduction in residual gasoline levels, Figure E-8 indicates that «100
1-vapor/g- gasoline must pass through the contaminated soil zone. If our spill
mass is 1500 kg (-500 gal), then a minimum of 1.5 x 108 1-vapor must pass
through the contaminated soil zone. If our target clean-up period is six
months, this corresponds to a minimum average vapor flowrate of 0.57 m3/min
(-20 cfm) . The minimum number of extraction wells is then equal to the
required minimum average flowrate/flowrate per well.
The third method for determining the number of wells insures that we
remove vapors and residual soil contamination from the entire zone of
contamination Nala. This is simply equal to the ratio of the area of
contamination Aeonta-lBStloll, to the area of influence of a single venting well
2
1TR
1
This requires an estimate of Rx, which defines the zone in which vapor flow is
induced. In general, Rx depends on soil properties of the vented zone,
properties of surrounding soil layers, the depth at which the well is
screened, and the presence of any impermeable boundaries (water table, clay
layers, surface seal, building basement, etc.). At this point it is useful to
have some understanding of vapor flow patterns because, except for certain
ideal cases6, one cannot accurately predict vapor flowpaths without
numerically solving vapor flow equations. An estimate for RT can be obtained
by fitting radial pressure distribution data from the air permeability test to
the steady- state radial pressure distribution equation :
E-29
-------
PAtm 2 ln(r/RJ
P(r) - Pw[l+(l-( - ) ) - J1'2 (E-22)
where P(r) , PAta, Pw, and R,, are the absolute pressure measured at a distance r
from the venting well, absolute ambient pressure, absolute pressure applied at
the vapor extraction well, and extraction well radius, respectively. Given
that these tests are usually conducted for less than a day, the results will
generally underestimate Rx. If no site specific data is available, one can
conservatively estimate RT based on the published reports from in-situ soil
venting operations. Reported Rt values for permeable soils (sandy soils) at
depths greater than 20 ft below ground surface, or shallower soils beneath
good surface seals, are usually 10 m - 40 m.1 For less permeable soils
(silts, clays), or more shallow zones RT is usually less.
- Choosing veil location, spacing, passive veils, and surface seals
To be able to successfully locate extraction wells, passive wells, and
surface seals one must have a good understanding of vapor flow behavior. We
would like to place wells so that we insure adequate vapor flow through the
contaminated zone, while minimizing. vapor flow through other zones.
If one well is sufficient, it will almost always be placed in the
geometric center of the contaminated soil zone, unless it is expected that
vapor flow channeling along a preferred direction will occur. In that case
the well will be placed so as to maximize air flow through the contaminated
zone.
When multiple wells are used it is important to consider the effect that
each well has on the vapor flow to all other wells. For 'example , if three
extraction wells are required at a given site, and they are installed in the
triplate design shown in Figure E-lla, there would be a "stagnant" region in
the middle of the wells where air flow would be very small in comparison to
the flow induced outside the triplate pattern boundaries. This problem can be
alleviated by the use of "passive wells" or "forced injection" wells as
illustrated in Figure E-llb (it can also be minimized by changing the vapor
flowrates from each veil with time). A passive well is simply a well that is
open to the atmosphere; in many cases groundwater monitoring wells are
suitable. If a passive or forced injection veil is to have any positive
effect, it must be located within the extraction well's zone of influence.
Forced injection veils are simply vapor wells into vhich air is pumped rather
than removed. One must be very careful in choosing the locations of forced
injection wells so that contaminant vapors are captured by the extraction
wells, rather than forced off -site. To date there have not been any detailed
reports of venting operations designed to study the advantages/disadvantages
of using forced injection wells. Figure E-llc presents another possible
extraction/injection well combination. As illustrated in Figure E-9, passive
wells can also be used as vapor barriers to prevent on- site migration of
off -site contamination problems.
E-30
-------
stagnant
air flow
region .
extraction
weiis
b)
vapor flow
lines
injection
well
c)
_ . extraction
.--"" -* wells
Y""
t
f
injection T yX
wells
/l
s
=
1
f
r
i
I
{
=j
1
\
\
mm
Ml
/ extraction
/ wells
A
&
i
3
B
X
=
=
i
Figure E-ll. Venting Well Configxiration.
E-31
-------
For shallow contamination problems (<4 m below ground surface) vapor
extraction trenches combined with surface seals may be more effective than
vertical wells. Trenches are usually limited to shallow soil zones because the
difficulty of installation increases with depth.
Surface seals, such as polymer-based liners and asphalt, concrete, or
clay caps, are sometimes used to control the vapor flow paths. Figure E-12
illustrates the effect that a surface seal will have on vapor flow patterns.
For shallow treatment zones (<5 m) the surface seal will have a significant
effect on the vapor flow paths, and seals can be added or removed to achieve
the desired vapor flowpath. For wells screened below 8 m the influence of
surface seals becomes less significant.
- Well screening and construction
Wells should be screened only through the zone of contamination, unless
the permeability to vapor flow is so low that removal rates would be greater
if flow were induced in an adjacent soil layer (see Figure E-6). Removal rate
estimates for various mass-transfer limited scenarios can be calculated from
Equations E-7 and E-12.
Based on Equation E-5, the flowrate is expected to increase by 15% when
the extraction well diameter is increased from 10 cm (4 in) to 20 cm (8 in).
This implies that well diameters should be as large as is practically
possible.
A typical well as shown in Figure E-13a is constructed from slotted pipe
(usually PVC). The slot size and number of slots per inch should be chosen to
maximize the open area of the pipe. A filter packing, such as sand or gravel,
is placed in the annulus between the borehole and pipe . Vapor extraction
wells are similar to groundwater monitoring wells in construction but there is
no need to filter vapors before they enter the well. The filter packing,
therefore, should be as coarse as possible. Any dust carried by the vapor
flow can be removed by an above-ground filter. Bentonite pellets and a
cement grout are loaded above the filter packing. It is important that these
be properly installed to prevent a vapor flow "short-sircuit". Any
groundwater monitoring wells installed near the extraction wells must also be
installed with good seals.
- Vapor treatment
Currently there are four main treatment processes available. Each is
discussed below.
- vapor combustion units: Vapors are incinerated and destruction
efficiencies are typically >95%. A supplemental fuel, such as propane, is
added before combustion unless extraction well vapor concentrations are on
the order of a few percent by volume. This process becomes less economical as
vapor concentrations decrease below »10,000 ppm,..
E-32
-------
a)
"open" soil surface
b)
Figure E-12. Effect of Surface Seal on Vapor Flowpath.
E-33
-------
cexsr.: can
slotted pipe
section
t
ccment/beotonite
grout
coarse packing
"*" 'material
b)
air-tight monitoring well
cap/water sensor assembly
\
electronic water
sensor
pressure gauge
connection
wire to sensor
double teflon
inner septa seal
monitoring
well cap
Figure E-13.
a) Extraction Well Construction, and b) Air-Tight
Groundwater Level Measuring System.
E-34
-------
- catalytic oxidation units: Vapor streams are heated and then passed
over a catalyst bed. Destruction efficiencies are typically >95%. These
units are used for vapor concentrations <8000 ppn^,. More concentrated vapors
can cause catalyst bed temperature excursions and melt-down.
- carbon beds: Carbon can be used to treat almost any vapor streams,
but is only economical for very low emission rates (<100 g/d)
- diffuser stacks: These do not treat vapors, but are the most
economical solution for areas in which they are permitted. They must be
carefully designed to minimize health risks and maximize safety.
- Groundwater pimping system
In cases where contaminated soils lie Just above or below the water
table, groundwater pumping systems will be required to insure that
contaminated soils remain exposed. In designing a groundwater system it is
important to be aware that upwelling (draw-up) of the groundwater table will
occur when a vacuum is applied at the extraction well (see Figure E-9b).
Because the upwelling will be greatest at the extraction wells, groundwater
pumping wells should be located within or as close to the extraction wells as
possible. Their surface seals must be airtight to prevent unwanted
short-circuiting of airflow down the groundwater wells.
- System integration
System components (pumps, wells, vapor treating units, etc.) should be
combined to allow maximum flexibility of operation. The review by Hutzler et
al.1 provides descriptions of many reported systems. Specific requirements
are:
- separate valves, flowmeters, and pressure gauges for each extraction
and injection well.
- air filter to remove particulates from vapors upstream of pump and
flow meter.
- knock-out pot to remove any liquid from vapor stream upstream of pump
and flow meter.
Monitorinf
The performance of a soil venting system must be monitored in order to
insure efficient operation, and to help determine when to shut-off the system.
At a minimum the following should be measured:
- date and time of measurement.
- vapor flov rates from extraction wells and into injection wells:
these can be measured by a variety of flowmeters including pitot tubes,
E-35
-------
orifice plates, and rotameters. It is important to have calibrated these
devices at the field operating pressures and temperatures.
- pressure readings at each extraction and injection well can be
measured with manometers or magnahelic gauges.
- vapor concentrations and compositions from extraction wells: total
hydrocarbon concentration can be measured by an on-line total hydrocarbon
analyzer calibrated to a specific hydrocarbon. This information is combined
with vapor flowrate data to calculate removal rates and the cumulative amount
of contaminant removed. In addition, for mixtures the vapor composition
should be periodically checked. It is impossible to assess if vapor
concentration decreases with time are due to compositional changes or some
other phenomena (mass transfer resistance, water table upwelling, pore
blockage, etc.) without this information. Vapor samples can be collected in
evacuated gas sampling cylinders, stored, and later analyzed.
- temperature: ambient and soil.
- water table level (for contaminated soils located near the water
table): It is important to monitor the water table level to insure that
contaminated soils remain exposed to vapor flow. Measuring the water table
level during venting is not a trivial task because the monitoring well must
remain sealed. Uncapping the well releases the vacuum and any effect that it
has on the water table level. Figure E-13b illustrates a monitoring well cap
(constructed by Applied Geo'sciences Inc., Tustin, CA) that allows one to
measure simultaneously the water table level and vacuum in a monitoring well.
It is constructed from a commercially available monitoring well cap and
utilizes an electronic water level sensor.
Other valuable, but optional measurements are:
- soil gas vapor concentrations and compositions: these should be
measured periodically at different radial distances from the extraction well.
Figure E-14 shows the construction of a permanent monitoring installation
that can be used for vapor sampling and subsurface temperature measurements.
Another alternative for shallow contamination zones is the use of soil gas
survey probes.
This data is valuable for two reasons: a) by comparing extraction well
concentrations with soil gas concentrations it is possible to estimate the
fraction of vapor that is flowing through the contaminated zone f-C^rw-tioa
w,n/cjoii c*» and °) ic ia possible to determine if the zone of contamination
is shrinking towards the extraction well, as it should with time. Three
measuring points are probably sufficient if one is located near the extraction
well, one is placed near the original edge of the zone of contamination, and
the third is placed somewhere in between.
These monitoring installations can also be useful for monitoring the
subsurface vapors after venting has ceased.
E-36
-------
Ground Surface
1/8" OD Teflon Tubing -
Box Containing Vapor Sampling
Ports ^Thermocouples
rPVCKpe
coarse packing
cement/bentonite
Figure E-14. Vadose Zone Honitoring Installation.
E-37
-------
When To Turn Off The System?
Target soil clean-up levels are often set on a site-by-site basis, and
are based on the estimated potential impact that any residual may have on air
quality, groundwater quality, or other health standards. They may also be
related to safety considerations (explosive limits). Generally, confirmation
soil borings, and sometimes soil vapor surveys, are required before closure is
granted. Because these analyses are expensive and often disrupt the normal
business of a site, it would be valuable to be able to determine when
confirmation borings should be taken. If the monitoring is done as suggested
above, then the following criteria can be used:
• cumulative amount removed: determined by integrating the measured
removal rates (flowrate x concentration) with time. While this value
indicates how much contaminant has been removed, it is usually not very useful
for determining when to take confirmation borings unless the original spill
mass is known very accurately. In most cases that information is not
available and can not be calculated accurately from soil boring data.
- extraction well vapor concentrations: the vapor concentrations are
good indications of how effectively the venting system is working, but
decreases in vapor extraction well concentrations are not strong evidence that
soil concentrations have decreased. Decreases may also be due to other
phenomena such as water table level increases, increased mass transfer
resistance.due to drying, or leaks in the extraction system.
- extraction well vapor composition: when combined with vapor
concentrations this data gives more insight into the effectiveness of the
system. If the total vapor concentration decreases without a change in
composition, it is probably due to one of the phenomena mentioned above, and
is not an indication that.the residual contamination has been significantly
reduced. If a decrease in vapor concentration is accompanied by a shift in
composition towards less volatile compounds, on the other hand, it is most
likely due to a change in the residual contaminant concentration. For
residual gasoline clean-up, for example, one might operate a venting system
until benzene, toluene, and xylenes were not detected in the vapors. The
remaining residual would then be composed of larger molecules, and it can be
argued that these do not pose a health threat through volatilization or
leaching pathways.
- soil gas contaminant concentration and composition: this data is the
most useful because it yields information about the residual composition and
extent of contamination. Vapor concentrations can not be used to determine
the residual level, except for very low residual levels (<500 mg/kg).
Other Factors
- increased blodegradatlon
It is often postulated that because the air supply to the vadose zone is
E-38
-------
increased, the natural aerobic microbiological activity is increased during
venting. While the argument is plausible and some laboratory data is
available7, conclusive evidence supporting this theory has yet to be
presented. This is due in part to the difficulty in making such &
measurement. A mass balance approach is not likely to be useful because the
initial spill mass is generally not known with sufficient accuracy. An
indirect method would be to measure C02 levels in the extraction well vapors,
but this in itself does not rule out the possibility that 02 is converted to
C02 before the vapors pass through the contaminated soil zone. The best
approach is to measure the 02/C02 concentrations in the vapors at the edge of
the contaminated zone, and in the vapor extraction wells. If the C02/02
concentration ratio increases as the vapors pass through the contaminated
soil, one can surmise that a transformation is occurring, although other
possible mechanisms (inorganic reactions) must be considered. An increase in
aerobic microbial populations would be additional supporting evidence.
- in-situ heating/venting
The main property of a compound that determines whether or not it can be
removed by venting is its vapor pressure, which increases with increasing
temperature. Compounds that are considered nonvolatile, therefore, can be
removed by venting if the contaminated soil is heated to the proper
temperature. In-situ heating/venting systems utilizing radio-frequency
heating and conduction heating are currently under study8. An alternative is
to reinject heated vapors from catalytic oxidation or combustion units into
the contaminated soil zone.
- air sparging
Due to seasonal groundwater level fluctuations, contaminants sometimes
become trapped below the water table. In some cases groundwater pumping can
lower the water table enough to expose this zone, but in other cases this is
not practical. One possible solution is to install air sparging wells and
then inject air below the water table. Vapor extraction wells would then
capture the vapors that bubbled up through the groundwater. To date, success
of this approach has yet to be demonstrated. This could have a negative
effect if foaming, formation plugging, or downward migration of the residual
occurred.
Application of the Design Approach to a Service Station
Remediation
In the following we will demonstrate the use of the approach discussed
above and outlined in Figure E-2 for the design operation, and monitoring of
an in-situ venting operation at a service station.
E-39
-------
Preliminary Site Investigation
Prior to sampling it was estimated that 2000 gal of gasoline had leaked
from a product line at this site. Several soil borings were drilled and the
soil samples were analyzed for total petroleum hydrocarbons (TPH) and other
specific compounds (benzene, toluene, xylenes) by a heated-headspace method
utilizing a field GC-FID. Figure E-15 summarizes some of the results for one
transect at this site. The following relevant information was collected:
- based on boring logs there are four distinct soil layers at this site
between 0 - 18 m (0- 60 ft) below ground surface (BGS). Figure E-15 indicates
the soil type and location of each of these layers.
- depth to groundwater was 15 m, with fine to medium sand aquifer soils
- the largest concentrations of hydrocarbons were detected in the sandy
and silty clay layers adjacent to the water table. Some residual was detected
below the water table. Based on the data presented in Figure E-15 it is
estimated that - 4000 kg of hydrocarbons are present in the lower two soil
zones.
- initially there was some free-liquid gasoline floating on the water
table, and this was subsequently removed by pumping. A sample of this product
was analyzed and its approximate composition (-20% of the compounds could not
be identified) is listed in Table E-2 as the "weathered gasoline". The
corresponding boiling point distribution curve for this mixture has been
presented in Figure E-3.
- vadose zone monitoring installations similar to the one pictured in
Figure E-14 were installed during the preliminary site investigation.
Deciding if Venting is Appropriate
For the remainder of the analysis we will focus on the contaminated soils
located just above the water table.
- What contaminant vapor concentrations are likely to be obtained?
Based on the composition given in Table E-2, and using Equation E-l, the
predicted saturated TPH vapor concentration for this gasoline is:
C..t - 220 mg/1
Using the "approximate" composition listed in Table E-2 yields a value
of 270 mg/1. The measured soil vapor concentration obtained from the vadose
zone monitoring well was 240 mg/1. Due to composition changes with time, this
will be the maximum concentration obtained during venting.
- Under ideal flow conditions is this concentration great enough to
yield acceptable removal rates?
E-40
-------
Flow Meter
10 —
20 —
30 —
c5
140.
so-
60—*
apor
ncraior _____
^
•i
•• «•••
m
«
•
W OBM
M
(•
•
V
•
1-L
A,
1
\
Dilution Samplinc
Air Pon "
Inlet N /
Blower
r \
Tank
S^ \ Backfill
-0.02
-0.0
.0.0
t
-0.0
.1.7
1
-24 J
.73
m
Fine to
Coarse Sand
,
Silty Qay
m
m
&
Clayey Silt
m
Medium
Sand
<•
V
_ li I
-**>
-1.7 ^>
-S12
-5.4
-8577
•653
•3267
-1237
-23131
T3319""
.1.7
1
-
Well
Manifold
/ To Water Treating
t jf System
Mi
• '*fl
<•
«•
•H
«•
-
1 '
i 1
i -
\
\
- 0.8
_ 0 J
-8.2
_ 214
- 967
- 971
.28679 J
. 23167
- 0.31
- 0.44
- 0.17
- 8.8
—0.63
. 0.86
- 23
- t.fi
- 32 T
«—
HB-17
Static Ground
Water Table
HB-3
HB-21
[Ground Water
Recovery Welll
SCALE (ft)
10
20
Figure E-15.
Initial Total Hydrocarbon Distribution [tngAS-soil] and
Location of Lower Zone Vent Well.
E-41
-------
Equation E-4 was used to calculate Raee«ptabl« . Assuming Msplll - 4000 kg
and t - 180 d, then:
- 22
Using Equation E-2, C.,t - 240 mg/1, and Q - 2800 1/min (100 cfm) :
R.,t - 970 kg/d
which is greater than Rmec,ptmbu-
- What range of vapor flovrates can realistically be achieved?
Based on boring logs the contaminated zone just above the water table is
composed of fine to medium sands, which have an estimated permeability 1< k < 10
darcy. Using Figure E-5, or Equation E-5, the predicted flowrates for an
extraction well vacuum Pw - 0.90 atm are:
0.04 < Q < 0.4 m3/ni-min R^ - 5.1 cm, RT - 12 m
0.43 < Q < 4.3 ft3/ft-min f^ - 2.0 in, Rx - 40 ft
The thickness of this zone and probable screen thickness of an extraction
well is about 2 m (6.6 ft). The total flowrate per well through this zone is
estimated to be 0.08
-------
contaminated zone to achieve this target. Based on our estimated initial
residual of 4000 kg TPH, 4 x 108 1-vapor are required. Over a six month
period^this corresponds to an average flowrate Q-l-5 m3/nin (54 cfm). Recall
that since this corresponds to the maximum removal rate, it is the minimum
required flowrate.
- Are there likely to be any negative effects of soil venting?
Given that the contaminated soils are located just above and below the
water table, water table upwelling during venting must be considered here.
Air Permeability Test
Figure E-16 presents data obtained from the air permeability test of
this soil zone. In addition to vapor extraction tests, air injection tests
were conducted. The data is analyzed in the same manner as discussed for
vapor extraction tests. Accurate flowrate (Q) values were not measured,
therefore, Equation E-17 was used to determine the permeability to vapor flow.
The k values ranged from 2 to 280 darcys, with the median being -8 darcys.
System Design
. Number of vapor extraction wells:
Based on the 8 darcys permeability, and assuming a 15 cm diameter (6 in)
venting well, a 2 m screened section, Pv - 0.90 atm (41 in H20 vacuum) and
Rj-12 m, then Equation E-5 predicts:
Q - 0.7 m3/min - 25 cfm
Based on the discussion above, a minimum average flowrate of 1.5 m3/min is
needed to reduce the residual to 1000 ppm in 6 months. The number of wells
required is then 1.5/0.7 - 2, assuming that 100% of the vapor flows through
contaminated soils. It is not likely that this will occur, and a more
conservative estimate of 50% vapor flowing through contaminated soils would
require that twice as many wells (4) be installed.
A single vapor extraction well (HB-25) was installed in this soil layer
with the knowledge that more wells were likely to be required. Its location and
screened interval are shown in Figure E-15. Other wells were installed in the
clay layer and upper sandy zone, but in this paper we will only discuss results
from treatment of the lower contaminated zone. A groundwater pumping well was
installed to maintain a 2 m drawdown below the static water level. Its location
is also shown in Figure £-15.
System Monitoring
Three vadose monitoring wells similar in construction to the one pictured
in Figure E-14 were installed so that the soil temperature, soil gas
E-43
-------
a)
Pressure --
Decrease
(inH,0) _
1
-6-
-8
-10
C HB-7D (r=3.4m)
A HB-6D (r=16.m)
C HB-14D (r=9.8 m)
O Q
a a
10 100
Time (min)
1000
b)
Pressure
Increase
(inH,0)
i
40-
'
30-
.
20-
10-
0-
a HB-7D (r=3.4m)
A HB-6D (f=16.m) a^
0 HB-14D (r=9.8 m)
+ HB-10 (r=7.6m) O
+*
U -F*"-C
0* **-*
Igl* ^
-------
concentrations, and subsurface pressure distribution could be monitored at three
depths. One sampling port is located in the zone adjacent to the aquifer. The
vapor flowrate from HB-25 and vapor concentrations were measured frequently, and
the vapor composition was determined by GC-FID analysis. In addition, the water
level in the groundwater monitoring wells was measured with the system pictured
in Figure E-13b. The results from the first four months of operation are
discussed below.
In Figure E-17a the extraction well vacuum and corresponding vapor flowrate
are presented. The vacuum was maintained at 0.95 atm (20 in H20 vacuum), and
the flowrate was initially 12 scfm. It gradually decreased to about 6 scfm'over
80 d. For comparison, Equation (5) predicts that Q-12 cfm for k-8 darcys.
Increasing the applied vacuum to 0.70 atm (120 in H20 vacuum) had little effect
on the flowrate. This could be explained by increased water table upwelling,
which would act to decrease the vertical cross-section available for vapor flow.
The scatter in the flowrate measurements is probably due to inconsistent
operation of the groundwater pumping operation, which frequently failed to
perform properly.
Figure E-17b presents the change in vapor concentration with time. Fifteen
specific compounds were identified during the GC-FID vapor analyses; in this
figure we present the total concentration of known and unknown compounds detected
between five boiling point ranges:
methane - isopentane (<28°C)
isopentane - benzene (28 - 80°C)
benzene - toluene (80 - 111°C)
toluene - xylenes (111 - 144°C)
>xylenes (>144°C)
There was a shift in composition towards less volatile compounds in the
first 20 d, but after that period the composition remained relatively constant.
Note that there is still a significant fraction of volatile compounds present.
Within the first two days the vapor concentration decreased by 50%, which
corresponds to the time period for the removal of the first pore volume of air.
Comparing the subsequent vapor concentrations with the concentrations measured
in the vadose zone monitoring wells indicates that only (80 mg/l)/(240
mg/l)*100-33% of the vapors are flowing through contaminated soil.
Figure E-18a presents calculated removal rates (flowrate x concentration)
and cumulative amount (1 gal - 3 kg) removed during the first four months. The
decrease in removal rate with time is due to a combination of decreases in
flowrate and hydrocarbon vapor concentrations. After the first four months
approximately one-fourth of the estimated residual has been removed from this
lower zone.
Oti day 80 the vacuum was increased from 20 - 120 in H20 vacuum and the
subsequent increase in subsurface vacuum and water table upwelling was monitored.
Figure E-18b presents the results. Note that the water table rise paralleled the
vacuum increase, although the water table did not rise the same amount that the
E-45
-------
HB-25
Vacuum
(inH O)
.
120-
•
;
•
f
0 •. ^
.*•»
80-
OW "
60-
40-
20
0-
*r .
*
Vacuum)
-— •— • Rowme
„
'. ;' '
\ ' "v
*
i , , ,
* * *
V ^ i *
M A * ^
1 v V '
.
/
" " ' ' '
"15
•10
•5
•o
Flowrate
(SCFM)
20 40 60 80 100 120
Time(d)
[in H.,0] denote vacuums expressed as equivalent water column heights
Figure E-17a. Soil Venting Results: Vacuum/Flowrate Data
E-46
-------
b)
100
80-
Vapor
Cone. 60-1
(mg/1)
20
-------
a)
Removal 40'
Rate
(kg/d) 30-
40 60 80
Time(d)
Cumulative
Recovered
(gal)
b)
ftH2O
0.5
0.4
03-
02-
0.1-
0.0
.1
• vacuum incmsc
• water able apwelling
D OP
1 10
Time(min)
100
(ft HO] denote vacuums expressed as equivalent water column heights
Figure E-18.
Soil Venting Results: a) Removal Rate/Cumulative
Recovered, b) Water Table Rise.
E-48
-------
vacuum did.
Figure E-19 compares the reduced measured TPH vapor concentration
C(t)/C(t-0) with model predictions. C(t-O) was taken to be the vapor
concentration after one pore volume of air had passed through the contaminated
zone (-80 mg/1) , m(t-O) is equal to the estimated spill mass (-4000 kg), and V(t)
is the total volume of air that has passed through the contaminated zone. This
quantity is obtained by integrating the total vapor flowrate with time, then
multiplying it by the fraction of vapors passing through the contaminated zone
f (-0.33). As discussed, the quantity f was estimated by comparing soil gas
concentrations from the vadose zone monitoring installations with vapor
concentrations in the extraction well vapors. As can be seen, there is good
quantitative agreement between the measured and predicted values.
Based on the data presented in Figures E-15 through E-19 and the model
predictions in Figure E-8, it appears that more extraction wells (-10 more) are
needed to remediate the site within a reasonable amount of time.
CONCLUSIONS
A structured, technically based approach has been presented for the design,
construction, and operation of venting systems. While we have attempted to
explain the process in detail for those not familiar with venting operations or
the underlying governing phenomena, the most effective and efficient systems can
only be designed and operated by personnel with a good understanding of the
fundamental processes involved. The service station spill example presented
supports the validity and usefulness of this approach.
There are still many technical issues that need to be resolved in the
future. In particular, we must be able to estimate removal rates for ndn- ideal
situations, demonstrate that biodegradation is enhanced by venting, and
investigate novel ideas for enhancing venting removal rates.
REFERENCES
Bear, J., Hydraulics of Groundwater, McGraw-Hill, 1979.
Dev H G C Sresty, J. E. Bridges and D. Downey, Field Test of the Radio
' Frequency in-situ Soil Decontamination Process, in Superfund '88:
Proceedings of the 9th National Conference, HMCRI, November 1988.
Hutzler N J B. E. Murphy, andJ. S. Gierke, State of Technology Review: Soil
Vapor Extraction Systems, U.S.E.P.A., CR-814319-01-1, 1988.
Johnson PC M. W. Kemblowski, and J. D. Colthart, Practical Screening Models
lor Soil Venting Applications, NWWA/API Conference on Petroleum
Hydrocarbons and Organic Chemicals in Groundwater, Houston, TX, 1988.
« i v r »~A c F Hoae Induced Soil Venting for the Recovery /Restoration
^f Gas'olTne Hydroca^oi .in the Vadose Zone, NWWA/API Conference on
Hydrocarbons and Organic Chemicals in Groundwater, Houston, TX,
1984.
E-49
-------
Millington, R. J., and J. M. Quirk, Permeability of Porous Solids, Trans.
Faraday Soc., 57:1200-1207, 1961.
Salanitro, J. P., M. M. Western, and M. W. Kemblowski, Biodegradation of Aromatic
Hydrocarbons in Unsaturated Soil Microcosms. Poster paper presented at the
Fourth National Conference on Petroleum Contaminated Soils, University of
Massachusetts, Amherst, September 25-25, 1989.
Wilson, D. J., A. N. Clarke, and J. H. Clarke, Soil Clean-up by in-situ Aeration.
I. Mathematical Modelling, Sep. Science Tech., 23:991-1037, 1988.
Ca)/C(t=0) prediacd
* C(l)/C(i=0) measured
C(t)/C(t=0).
weathered gasoline
m(t=0)-4000kg
012345 6 789 10
V(t)/m(t=0) (1/g)
Figure E-19. Comparison of Model Predictions and Measured Response.
E-50
-------
APPENDIX F - ARTICLE ON INCINERATION
(Reprinted with the Permission of AWMA)
INCINERATION OF HAZARDOUS WASTE
A Critical Review
E. Timothy Oppelt
Hazardous Waste Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio
Over the last ten years, concern over improper disposal practices of the past has manifested
itself in the passage of a series of federal and state-level hazardous waste clean-up and
control statutes of unprecedented scope. The impact of these various statutes will be a
significant modification of waste management practices. The more traditional and lowest
cost methods of direct landfilling, storage hi surface impoundments and deep-well injection
will be replaced, in large measure, by waste minimization at the source of generation, waste
reuse, physical/chemical/biological treatment, incineration and chemical stabilization/solid-
ification methods. Of all of the "terminal" treatment technologies, properly-designed incin-
eration systems are capable of the highest overall degree of destruction and control for the
broadest range of hazardous waste streams. Substantial design and operational experience
exists and a wide variety of commercial systems are available. Consequently, significant
growth is anticipated in the use of incineration and other thermal destruction methods. The
objective of this review is to examine the current state of knowledge regarding hazardous
waste incineration in an effort to put these technological and environmental issues into
perspective.
Hazardous waste management is the environmental issue of
the 1980s. Discovery of the numerous environmental catas-
trophes resulting from the improper disposal practices of the
past have elevated public awareness and concern. Over the
last ten years, this concern has manifested itself in the pas-
sage of a series of federal and state-level hazardous waste
clean-up and control statutes of unprecedented scope and
impact At the federal level these laws include the Resource
Conservation and Recovery Act of 1976 (RCRA) and its
"cradle to grave" provisons for controlling the storage, trans-
port, treatment, and disposal of hazardous waste. In 1979,
the PCB regulations promulgated under Section 6(e) of the
Toxic Substances Control Act (TSCA), prohibited the fur-
ther manufacture of polychlorinated biphenyls (PCBs) after
July 2,1979, established limits on PCB use in commerce, and
established regulations for proper disposal. Clean-up of the
uncontrolled waste sites created by poor disposal practices
of the past was provided for ;n the Comprehensive Environ-
mental Response, Compensation, and Liability Act of 1980
(CERCLA) which established a national fund (Superfund)
to assist in remedial actions. The 1986 Superfund Amend-
ments and Reauthorization Act (SARA) not only reautho-
rized the Superfund program but greatly expanded the pro-
tons and funding of the initial Act
Co»n|ht 1987—APCA
558
The most significant of all of these statutes were the 1984
amendments and reauthorization of RCRA. Termed the
Hazardous and Solid Waste Act of 1984 (HSWA), these
amendments establish a strict timeline for restricting un-
treated hazardous waste from land disposal. By 1990, most
wastes will be restricted and pretreatment standards will be
established based upon the treatment levels achievable by
Best Demonstrated Available Technology (BOAT).1
The impact of these various statutes will be n significant
modification of waste management practices. T..<. more tra-
ditional and lowest cost methods of direct landfilling, stor-
age in surface impoundments and deep-well injection will be
replaced, in large measure, by waste minimization at the
source of generation, waste reuse, physical/chemical/biologi-
cal treatment, incineration and chemical stabilization/solid-
ification methods.
Of all of the "terminal" treatment technologies, properly-
designed incineration systems are capable of the highest
overall degree of destruction and control for the broadest
range of hazardous waste streams. Substantial design and
operational experience exists and a wide variety of commer-
cial systems are available. Consequently, significant growth
is anticipated in the use of incineration and other thermal
destruction methods.2
While thermal destruction offers many advantages over
existing hazardous waste disposal practices and may help to
F-l
JAPCA
-------
meet the anticipated need for increased waste management
capacity, public opposition to the permitting of new thermal
destruction operations has been strong in recent years.3 The
environmental awareness and activism which spawned the
major hazardous waste laws of the 1980s have in many re-
spects switched to skepticism over the safety and effective-
ness of the technological solutions which the laws were de-
signed to implement. Citizen distrust of the waste manage-
ment facility owners and operators remains. The ability of
government agencies to enforce compliance is questioned.
Reports of trace quantities of chlorinated dioxins, chlorinat-
ed furans, and other combustion byproducts in the stack
•emissions of municipal solid waste and PCB incinerators
have raised questions in the minds of some concerning
whether the RCRA incinerator standards are sufficient to
protect public health and the environment. Yet, waste gen-
erators, faced with the specter of complying with the HSWA
land disposal restrictions or faced with the prospect of fu-
ture multimillion dollar environmental damage settlements
over contaminated groundwater, are looking to ultimate de-
struction techniques such as incineration as their only viable
alternative.
The objective of this review is to examine the current state
of knowledge regarding hazardous waste incineration in an
effort to put these technological'and environmental issues
into perspective. In doing so, it will be important to review:
• Current and emerging regulations and standards for haz-
ardous waste incinerators;
• Current incineration technology and practice;
• Capabilities and limitations of methods for measuring
process performance;
• Destruction efficiency and emissions characterization of
current technology;
• Methods for predicting and assuring incinerator perfor-
mance; '
• Environmental and public health implications of haz-
ardous waste incineration; and
• Remaining issues and research needs.
While the focus of the paper is on hazardous waste inciner-
ation, it is important to understand that many of the same
issues relate to municipal waste incineration and to the use
of hazardous waste as a fuel in industrial boilers and fur-
naces. Where possible and appropriate the performance and
emissions of these systems will be compared and contrasted
with hazardous waste incinerators.
Background
Historical P*rsp«ctlv«
Purification by fire is an ancient concept. Its applications
are noted in the earliest chapters of recorded history. The
Hebrew word for hell, Gehenna, was actually derived from
the ancient phrase ge-ben Hinnom or the valley of the son of
Hinnom, an area outside of Jerusalem which housed the
smoldering town dump and was the site of propitiatory sacri-
fices to Moloch II.4 Today, waste fires on the ground or in
pits are still used by nomadic tribes.
In the Middle Ages an early innovation to waste fires was
the "fire wagon," the first mobile incinerator.5 It was a sim-
ple rectangular wooden wagon protected by a clay lining.
The horse-drawn wagon traveled the streets allowing resi-
dents to throw their refuse into the moving bonfire.
Incineration as we know it today began slightly more than
one hundred years ago when the first municipal waste "des-
tructor" was installed in Nottingham, England.5 Incinera-
tion use in the United States grew rapidly also, from the first
installation on Governor's Island in New York to more than
200 units in 1921. Most of these were poorly operated batch
feed units, some with steam recovery. Until the 1950s, incin-
erators and their attendant smoke and odors were consid-
ered a necessary evil and their operations were generally
undertaken in the cheapest possible manner. However, as
billowing smoke stacks became less of a symbol of prosperity
and air pollution regulations began to emerge, incineration
systems improved dramatically.6 These improvements in-
cluded continuous feed, improved combustion control, the
use of multiple combustion chambers, designs for energy
recovery and the application of air pollution control sys-
tems."
Incineration has been employed for disposal of industrial
chemical wastes (hazardous waste) for slightly more than 20
years. Initial units borrowed on municipal waste technology,
but poor performance and adaptability of these early grate-
type units lead to the subsequent use of rotary kilns. Many
of the earliest rotary kiln facilities were in West Germany.
One of the first United States kiln units was at the Dow
Chemical Company facility in Midland, Michigan.
Regulations
The first U.S. federal standards for control of incineration
emissions were applied to municipal waste combustors un-
der the New Source Performance Standards (NSPS) provi-
sions of the Clean Air Act of 1970. The NSPS established a
time-averaged (2 hours) paniculate emission limit of 0.08
grains per dry standard cubic foot (gr/dscf), corrected to 12
percent COa, for all incineration units constructed after Au-
gust 1971 and having charging rates greater than 50 tons per
day (tpd). Opacity limits were not promulgated at the feder-
al level, but many states now have their own opacity limits
and, in some cases, more stringent paniculate control re-
quirements.
Hazardous waste incinerators were not regulated until the
passage of the Resource Conservation and Recovery Act of
1976 (RCRA). Technical standards for incinerators were
proposed in December 1978 under Section 3004 of RCRA.8
These standards provided both performance and operating
Table I. Quantities of incinerable wastes generated in the
United States. 1983.87
Quantity
Current after
Quantity percent recycled/
generated recycled/ recovered11
Type of waste (MMT)« recovered6 (MMT)
Liquids:
Waste oils
Halogenated solvents
Nonhalogenated solvents
Other organic liquids
Pesticides/herbicides
PCBs
Total liquids
Sludge and Solids:
Halogenated sludges
Nonhalogenated sludges
Dye and paint sludges
Oily sludges
Halogenated solids
Nonhalogenated solids
Resins, latex, monomer
Total sludges/solids
Total incinerable wastes
Total hazardous wastes
14.25
3.48
12.13
3.44
0.026
0.001
33.33
0.72
2.24
4.24
3.73
9.78
4.58
4.02
29.31
62.64
265.60
11
70
70
2
55
0
38
0
0
0
5
0
0
65
10
25
6
12.68
1.04
3.64
3.37
0.012
0.001
20.74
0.72
2.24
4.24
3.54
9.78
4.58
1.41
26.51
47.25
249.28
* MMT = millions of metric tons.
b These waste recycling and recovery practice estimates were
derived by the Congressional Budget Office from information
obtained directly through surveys of industrial waste generators and
the waste recovery industry.25
May 1987 Volume 37, No. 5
F-2
558
-------
requirements. The performance standards included require-
ments for acceptable levels of combustion efficiency, de-
struction efficiency, halogen removal efficiency, and an
emission limit for paniculate matter. Operational standards
required semicontinuous monitoring of process variables
(e.g., CO) and specific minimum temperature and combus-
tion gas residence time levels. During the allowable com-
ment period on the proposed rules, EPA received extensive
comment on the scope of the standards and the adequacy of
the combined EPA and industrial data base used to set the
standards.
Based upon the public comment, EPA subsequently pro-
ceeded down a three-phase regulatory path:
• Phase I (May 19, 1980). Interim status standards were
proposed outlining operating procedures to be followed
by existing incinerator facilities.9
• Phase II (January 23, 1981). Performance standards
were proposed for new incineration facilities, requiring
specific levels of organic hazardous constituent destruc-
tion and removal, exhaust gas HC1 removal, and maxi-
mum paniculate emission concentration.10
• Phase HI (June 24,1982). Interim final standards were
published for both new and existing incinerators, incor-
porating and modifying somewhat the provisions of the
Phase I and Phase II rules.11
The provisions of the final incinerator standards, which
are of most importance to this paper, are the performance
standards which are now listed in the Code of Federal Regu-
lations (CFR) under 40 CFR 264.343. These standards re-
quire that in order for a facility to receive a RCRA permit, it
must attain the following performance levels:
(1) 99.99 percent destruction and removal efficiency
(DRE) for each principal organic hazardous constitu-
ent (POHC) in the waste feed where:
DRE = ((Wm - Wout)/Win) X 100
where: Wj,, = mass feed rate of the principal organic
hazardous constituent (POHC) in the
waste stream fed to the incinerator, and
Wout = mass emission rate of the POHC in the
stack prior to release to the atmosphere.
(2) At least 99 percent removal of hydrogen chloride from
the exhaust gas if hydrogen chloride stack emissions
are greater than 1.8 kg/h.
(3) Paniculate matter emissions no greater than 180 mg/
standard m3 corrected to 7 percent oxygen in the stack
gas. The measured paniculate matter concentration is
multiplied by the following correction factor to obtain
the corrected paniculate matter emissions:
Correction factor = 14/(21 - Y)
where: Y - measured oxygen concentration in the
stack gas on a dry basis (expressed as a
percentage).
The concept and selection of POHCs is an important part
of the incineration regulations. POHCs, which are to be
sampled during "trial burns" to assess attainment of the
standards, are to be selected from the RCRA Appendix VIII
constituents present in the wastes.10 Appendix VIII is a list
of approximately 400 organic and inorganic hazardous
chemicals first published in Part 261 of the May 19, 1980
Federal Register.9 The list is updated semiannually in 40
CFR 261.
POHC selection guidance suggests that Appendix VIII
constituents which are in the highest concentration in the
waste feed and are the most difficult to incinerate are the
most likely and appropriate to be selected as POHCs.12 This
selection approach, particularly the concept of hazardous
compound incinerability, has been the subject of consider-
able scientific debate since the guidance was first proposed
in 1981. These issues will be examined in greater detail later.
It is important to note that EPA chose not to apply the
incineration standards to the practice of disposing of haz-
ardous waste as a fuel in industrial boilers and furnaces.9
This exemption was based upon a lack of sufficient informa-
tion on the practice and the fact that energy recovery consti-
tuted a beneficial use of wastes. Considerable data have been
assembled since the exemption was granted in 1980. EPA
began to control the practice first in 1985 with issuance of
RCRA regulations on the use of waste oil for energy recov-
ery.13 This rule provides a basis for distinguishing between a
used oil and a hazardous waste for energy recovery purposes
and provides a used oil specification that limits the types of
boilers that can burn used oils that fail the specification.
EPA is developing regulations which will cover the disposal
of hazardous waste in industrial boilers and other industrial
process furnaces. These rules are currently scheduled for
proposal in 1987.14
EPA has also promulgated regulations for the incineration
of specific wastes. Incineration of PCBs (polychlorinated
biphenyls) is controlled under TSCA rules promulgated in
May 1979.15 These rules require that whenever disposal of
PCBs is undertaken, they must be incinerated, unless the
PCB concentration is less than 50 parts per million (ppm). If
the concentration is between 50 and 500 ppm, the rule pro-
vides for certain exceptions that allow alternatives to the
incineration requirements, such as use as fuel in high effi-
ciency boilers. Where the concentration exceeds 500 ppm,
PCBs must be disposed of in incinerators which achieve a
99.9 percent combustion efficiency (CE), and meet a number
of specific incinerator operating conditions (combustion
temperature, residence time, stack oxygen concentration).
The incineration of certain wastes containing certain chlo-
rinated dibenzo-p-dioxins, chlorinated dibenzofurans, and
chlorinated phenols is regulated under RCRA rules promul-
gated January 14, 1985. The so-called "dioxin rule" limits
the incineration of these specific wastes (EPA waste codes
F020-F028) to incinerators which have been "certified" as
being capable of achieving 99.9999 percent DRE for chlori-
nated dioxins or similar compounds.16
Current municipal waste incineration standards under the
Clean Air Act provide only limits on paniculate emissions,
as previously stated. The 1984 Hazardous and Solid Waste
Act Amendments (Section 102), however, require EPA to
prepare a Report to Congress on the extent of risks due to
dioxin emissions from municipal waste incinerators and on
appropriate methods for reducing these emissions. EPA also
plans to expand the report to include data on cancer risks
and controls associated with additional pollutants emitted
by these incinerators. EPA is committed by an agreement
with the Natural Resources Defense Council to issue an
announcement by May 1987 on what actions EPA plans to
take regarding risks from municipal waste incineration.
Current Incineration Practice
Incineration Practice
Incineration is an engineered process that employs ther-
mal decomposition via thermal oxidation at high tempera-
ture (usually 900°C or greater) to destroy the organic frac-
tion of the waste and reduce volume. Generally, combustible
wastes or wastes with significant organic content are consid-
ered most appropriate for incineration. However, technical-
ly speaking, any waste with a hazardous organic fraction, no
matter how small, is at least a functional candidate for incin-
eration. For instance, significant amounts of contaminated
water are currently incinerated in the United States.17 Con-
taminated soils are also being incinerated with increasing
frequency. EPA, for example, has employed a mobile incin-
580
F-3
JAPCA
-------
erator to decontaminate 40 tons of Missouri soil which had
been contaminated with four pounds of chlorinated dioxin
compounds."1 Many other designs for mobile incineration
facilities have emerged and are also being applied in the field
for decontamination of soil and debris.19
Since the promulgation of the RCRA interim status incin-
erator standards in 1980, a number of surveys and studies
have been conducted to assess the quantity and types of
hazardous waste generated in the United States as well as
the quantities and types of wastes being managed by various
treatment storage and disposal facilities.20-28 These studies
often reveal significant differences in what would seem to be
relatively straightforward statistics. While frustrating to
those in government and industry who are evaluating waste
management alternatives and economic impacts, these defi-
ciencies in the data base are not surprising. They have re-
sulted from many factors: changes and uncertainties in regu-
latory definitions of hazardous waste terms; differences in
methods and assumptions employed in the various surveys;
and incomplete or inaccurate responses by facility owners
and operators. Continuing changes in waste generation and
the number and permit status of facilities which have oc-
curred in response to regulatory changes and economic fac-
tors have also made it difficult to accurately project waste
management practice from one point in time to another.
In spite of these deficiencies and limitations, it is possible
to construct a reasonable picture of hazardous waste genera-
tion and incineration practice from the aggregate of the
studies. Total annual hazardous waste generation in the
United States appears to be approximately 265 million met-
ric tons (MMT). This number was first projected by EPA in
the so-called Westat mail survey24 and later confirmed in
separate studies by the Congressional Budget Office
(CBO),25 and the Congressional Office of Technology As-
sessment (OTA).26 Only a small fraction of this waste (<1
percent) was believed to have been incinerated. EPA esti-
mated that 1.7 MMT was disposed in incinerators in 198124
and CBO projected this amount at 2.7 MMT in 1983.25
Precise infromation on the exact types of wastes actually
going to incineration facilities is not available. Many facili-
ties operate on an intermittent basis and handle mixtures of
wastes which are difficult to describe in terms of EPA stan-
dard waste codes. A 1983 EPA study examined data on 413
waste streams going to 204 incineration facilities in the Unit-
ed States.17 The major waste streams incinerated were spent
nonhalogenated solvents (EPA waste code F003) and corro-
sive and reactive wastes contaminated with organics (EPA
waste codes D002 and D003). Together, these accounted for
44 percent of the waste incinerated. Other prominent wastes
included hydrocyanic acid (P063), acrylonitnle bottoms
(KOI 1), and nonlisted ignitable wastes (D001).
While only a small fraction of available hazardous waste is
currently managed by incineration, many believe that im-
plementation of the HSWA land disposal restriction regula-
tions and generator concern for long-term liability will result
in increased utilization of incineration for ultimate disposal.
EPA has estimated that nearly five times more hazardous
waste could have been thermally destroyed in incinerators
and industrial furnaces in 1981 than actually was.24 Numer-
ous other studies have indicated that the actual use and
demand for incineration technologies to manage hazardous
waste will increase significantly.2-25-27-29
The CBO study, however, offers the best perspective of
potential hazardous waste incineration practice.25 These
data, which are based upon industrial output models, are the
only available source of comprehensive waste generation
estimates which are aggregated on the basis of waste type.
This allows more precise estimation of incinerable waste
quantities. CBO also examined the potential impact of waste
reduction and recycling activities on waste available for in-
cineration. The results of these analyses (Table I) indicated
that even after recycling and reduction, as much as 47 MMT
per year could have been available for incineration in 1983.
This estimate, however, did not include potentially incinera-
ble wastes from uncontrolled hazardous waste sites.
It is clear that considerable potential exists for expansion
of incineration practice. This assumes, however, sufficient
RCRA-permitted capacity can be made available. This is, of
course, a significant issue and one which has been given
attention in a number of studies.2-20-23 While capacity ap-
pears, to be adequate in the near term, (Table II) waste
quantities received for incineration appear to be increasing
at a faster rate than capacity is being added.20 Beyond this,
future increases in demand will be primarily for organic
solids and sludges (e.g., wastes restricted from land disposal
or resulting from uncontrolled site clean-up). Liquid capaci-
ty will likely remain sufficient for a longer time, especially if
the capacity represented by potential disposal in cement
kilns and industrial boilers is included.2
One of the major barriers to increased incineration capaci-
ty is public opposition to the permitting and siting of new
facilities, especially the off-site commercial facilities which
would be necessary to handle much of the solids and sludges
which will increasingly require suitable disposal. Public op-
position to the permitting of new thermal destruction opera-
tions has been strong in recent years. The normal time re-
quired for permitting new incineration facilities is three
years. This time, as well as the expense of obtaining a permit,
Table II. Estimation of available hazardous waste incinerator capacity by incinerator design.22
Incinerator
design
Rotary kiln
Liquid injection
Fume
Hearth
Fluidized bed
and other
Total or
average values
Number of units
Reported
42
95
25
32
14
208
Projected*
45
101
26
34
15
221
Reported
average
design
capacity*
(MM Btu/h)
61.37
28.26
33.14
22.75
19.29
32.37
Reported
utilization6
(percent)
77
55
94
62
—
67
Projected
available
capacity1
(MM Btu/h)
635
1284
52
294
95I
-------
may be greatly increased if public opposition exists.
This has created considerable uncertainty for waste gen-
erators, equipment manufacturers, and commercial waste
disposers. Since 1981, for instance, almost 100 incinerators
have withdrawn from the RCRA system because they either
ceased operation or decided to no longer handle hazardous
wastes.22 Of the 57 companies identified as marketing haz-
ardous waste incinerators in 1981,23 have either gone out of
business, left the hazardous waste incinerator business, or
put considerably less emphasis on this activity.22
The amount of public opposition to proposed permits for
land-based incinerator facilities varies by location and type
of waste. On-site facilities that directly serve a single waste
generator have greater public acceptance than off-site, com-
mercial incinerators that serve multiple generators in a large
market area. Off-site facilities are often not perceived as
providing sufficient economic benefits to the local communi-
ty to offset the risks associated with the importation of
wastes from other areas. On-site facilities are more clearly
perceived as being linked to businesses that are important to
the local economy, and are generally not perceived as being
importers of hazardous waste. Opposition has tended to fo-
cus primarily on new off-site facilities, including incinerator
ships, and on new applications to burn PCBs, which critics
view as particularly hazardous.
In an effort to assess the dilemma of perceived benefits
versus public concerns, EPA conducted an assessment in
1985 to determine if there was a need for a change in the
approach toward regulating thermal destruction.3 The ma-
jor concern reported by citizens included concern for:
• Hazardous material spills in storage, treatment, and
handling.
• Environmental and health impacts of land-based and
ocean facilities.
• Poor site selection processes.
• Distrust of incinerator owners and operators.
• Inability of government agencies to enforce compliance.
The study concluded that public opposition to both land
and ocean incineration may decline somewhat if regulators
address more fully some citizen concerns regarding national
regulatory strategy, local community impact, equity of facil-
ity siting, public decision-making processes, and especially
enforcement plans and capacity. It was also concluded that
there is a need to better communicate how health and envi-
ronmental concerns and priorities are reflected in regula-
tions and standards. Better communication of regulatory
policy, strategy, and other activity related to decisions on
proposed permits for individual incinerator facilities or ves-
sels is certainly desirable since improved communication
with the public can enhance the credibility of regulatory and
enforcement agencies.
Incineration Technology
Different incineration technologies have been developed
for handling the various types and physical forms of hazard-
ous waste. A recent study identified 221 hazardous waste
incinerators operating under the RCRA system in the Unit-
ed States.21 Some of the results of this study are displayed in
Table II. The four most common incinerator designs (in
order of use) are: liquid injection (sometimes combined with
fume incineration); rotary kiln; fixed hearth; and fluidized
bed incinerators. These units are located at 189 separate
facilities. Only 18 facilities are commercial off-site opera-
tions,31 the balance of incineration practice being located at
the site of waste generation.
The process of selection and design of hazardous waste
incineration systems can be very complex. Fortunately, con-
siderable industrial manufacturing experience exists and
many useful design guides have been published.7-32-34 Thus,
while a detailed examination of design principles is beyond
the scope of this paper, a generalized review of the most
prominent features of incineration systems and important
design factors will be helpful in understanding their opera-
tion and emissions performance.
The major subsystems which may be incorporated into a
hazardous waste incineration system are:
(1) waste preparation and feeding
(2) combustion chamber(s)
(3) air pollution control
(4) residue/ash handling
The normal orientation of these subsystems is shown in
Figure 1 along with typical process component options. The
selection of the appropriate system combination of these
components is primarily a function of the physical and
chemical properties of the waste stream or streams to be
incinerated.
(1) Waste Preparation and Feeding. The physical form
of the waste determines the appropriate feed method.32 Liq-
uids are blended then pumped into the combustion cham-
bers through nozzles or via specially designed atomizing
Waste Preparation
1
i i
Blending Atomization
Screening Ram
Shredding Gravity
Heating Auger
Lance
Waste
Preparation
•^
Waste
Feeding
Combustion
1
1 i •
Liquid Injection .Quench
Rotary Kiln Heat
Fixed Hearth Recovery
Fluidized Bed
Combustion
Chamber! s)
I
Ash
Disposal
Dewatenng
Chemical
stabilization
Secure Landfill
Combustion
— Gas
Conditioning
PO
Air Pollution Control
1
Venturi Packed Tower
Wet ESP* Soray Tower
IWS* Tray Tower .
Fabric Filter IWS t
Wei ESP |
^
rw*
Paniculate
Removal
•
Res*
Treatr
due
nent
Acid
••» Gas -i
Removal
-
— »
Return to
Process
Neutralization -
Chemical Treatment
Demister
•* and
Stack
Residue
-and Ash
Handling
•IWS = Ionizing Wet Scrubber
ESP = Electrostatic Precipnator
POTW = Publically Owned
Treatment Works
Figure 1. General orientation of incineration suosystems and typical process component options.
562
F-5
JAPCA
-------
Table III. Applicability of major incinerator types to wastes of
various physical form.32
Liquid Rotary Fixed Fluidized
injection kiln hearth bed
Solids:
Granular, homogeneous
Irregular, bulky
(pallets, etc.)
Low melting point X
(tan, etc.)
Organic compounds with
fusible ash constituents
Unprepared, large, bulky
material
Gases: •
Organic vapor laden X
Liquids:
High organic strength X
aqueous wastes
Organic liquids X
Solids/liquids:
Waste contains X
halogenated
aromatic compounds
(2,200"? minimum)
Aqueous organic sludge
X
X
X
X
X
X
X
X
X
X
X X
X
X X
X X
X
X
X
burners. Wastes containing suspended particles may need to
be screened to avoid clogging of small nozzle or atomizer
openings. While sustained combustion is possible with waste
heat content as low as 4,000 Btu/lb, liquid wastes are typical-
ly blended to a net heat content of 8,000 Btu/lb or greater.
Blending is also used to control the chlorine content of the
waste fed to the incinerator. Wastes with chlorine content of
70 percent and higher have been incinerated,35 however,
most operators limit chlorine content to 30 percent or less.
chamber selected. Table III provides general selection con-
siderations for the four major combustion chamber (inciner-
ator) designs as a function of wastes of different forms*
Most incineration systems derive their name from the type
of combustion chamber employed.
Liquid injection incinerators or combustion chambers are
applicable almost exclusively for pumpable liquid waste.
These units (Figure 2) are usually simple, refractory-lined
cylinders (either horizontally or vertically aligned) equipped
with one or more waste burners. Liquid wastes are injected
through the bumer(s), atomized to fine droplets, and burned
in suspension. Burners as well as separate waste injection
nozzles may be oriented for axial, radial, or tangential firing.
Improved utilization of combustion space and higher heat
release rates, however, can be achieved with the utilization
of swirl or vortex burners or designs involving tangential
entry.36
Good atomization is critical to achieving high destruction
efficiency in liquid combustors. Nozzles have been devel-
oped to produce mists with mean particle diameters as low as
1 Mm,37 compared to typical oil burners which yield droplets
in the 10- to SO-jim range.38 Atomization may be attained by
low pressure air or steam (1 to lO.psig), high pressure air or
steam (25 to 100 psig), or mechanical (hydraulic) means
using specially designed orifices (25 to 450 psig).
Vertically aligned liquid injection incinerators are pre-
ferred when wastes are high in inorganic salts and fuseable
ash content, while horizontal units may be used with low ash
waste. The typical capacity of liquid injection incinerators is
roughly 28 X 106 Btu/h heat release. Units, however, range as
high as 70 to 100 X 106 Btu/h.
Rotary kiln incinerators (Figure 3) are more versatile in-
cinerators in the sense that they are applicable to the de-
struction of solid wastes, slurries, and containerized waste as
well as liquids. Because of this, these units are most fre-
quently incorporated into commercial off-site incineration
facility design. The rotary kiln is a cylindrical refractory-
lined shell that is mounted on a slight incline. Rotation of
the shell provides for transportation of waste through the
120-250%
Excess Air
Discharge
to Quench or
Waste Heat Recovery
Aqueous
Waste
Steam
Auxiliary
Fuel
Liquid
Waste
Atomizing
Steam or
Air
Y///7/////////////.
Primary
Combustion
Air
2600°F-3000°F
0.3-2.0 Seconds
Mean Combustion
Gas Residence Time
1500°F-2200°F
Figure 2. Typical liquid injection combustion chamber.
Cross Section
Blending to these levels provides best combustion control
and limits the potential for formation of hazardous free
chlorine gas in combustion gases.
Sludges are typically fed using progressive cavity pumps
and water cooled lances. Bulk solid wastes may require
shredding for control of particle size. They may be fea '.. the
combustion chamber via rams, gravity feed, air lock feeders,
vibratory or screw feeders, or belt feeders. Containerized
waste is typically gravity or ram fed.
(2) Combustion Chambers. The physical form of the
waste and its ash content determine the types of combustion
kiln as well as for enhancement of waste mixing. The resi-
dence time of waste solids in the kiln is generally 1 to 1.5
hours. This is controlled by the kiln rotation speed (1-5
revolutions per minute), the waste feed rate and, in some
instances, the inclusion of internal dams to retard the rate of
waste movement through the kiln. The feed rate is generally
adjusted to also limit the amount of waste being processed in
the kiln to at most 20 percent of the kiln volume.
The primary function of the kiln is to convert solid wastes
to gases, which occurs through a series of volatilization, de-
structive distillation, and partial combustion reactions.
However, an afterburner is necessary to complete the gas-
May 1987 Volume 37, No. 5
F-6
593
-------
Discharge to
Quench or
Heat Recovery
Air
Auxiliary \
Combustion
Air
Waste liquids
Auxiliary Fuel
Waste Solids.
Containers or
Sludges
Kiln
Shroud
50-250%
Excess Air
RotBfy Kiln
Figure 3. Typical rotary kiln/afterburner combustion chamber.
120-200%
Excess Air
1.0-3.0 Seconds
Mean Gas
Residence Time
Refractory
REF
phase combustion reactions. The afterburner is connected
directly to the discharge end of the kiln, whereby the gases
exiting the kiln turn from a horizontal flow path upwards to
the afterburner chamber. The afterburner itself may be hor-
izontally or vertically aligned, and essentially functions
much on the same principles as a liquid injection incinerator.
In fact, many facilities also fire liquid hazardous waste
through separate waste burners in the afterburner. Both the
afterburner and kiln are usually equipped with an auxiliary
fuel firing system to bring the units up to and maintain the
desired operating temperatures. Rotary kilns have been de-
signed with a heat release capacity as high as 90 X 106 Btu/h
in the United States. On average, however, units are typical-
ly 60 X 106 Btu/h.
Fixed hearth incinerators, also called controlled air,
starved air, or pyrolytic incinerators, are the third major
technology in use for hazardous waste incineration today.
These units employ a two-stage combustion process, much
like rotary kilns (Figure 4). Waste is ram fed into the first
stage, or primary chamber, and burned at roughly 50 to 80
percent of stoichiometric air requirements. This starved air
condition causes most of the volatile fraction to be destroyed
pyrolytically, with the required endothermic heat provided
by the oxidation of the fixed carbon fraction. The resultant
smoke and pyrolytic products, consisting primarily of vola-
tile hydrocarbons and carbon monoxide, along with products
of combustion, pass to the second-stage, or secondary cham-
ber. Here, additional air is injected to complete the combus-
tion, which can occur either spontaneously or through the
addition of supplementary fuels. The primary chamber com-
bustion reactions and turbulent velocities are maintained at
low levels by the starved-air conditions to minimize particu-
Discharge to
Quench or
Heat Recovery
0.25-2.5 Seconds
Mean Residence Time
100-200%
Excess Air
Steam
Auxiliary Fuel or
Liquid Waste
Auxiliary Fuel
Refractory
Transfer —"*
Ram Ash Discharge
Ram
Ash Discharge
. Typical fixed nearm combustion chamber.
SM
F-7
JAPCA
-------
late emrainment and carryover. With the addition of sec-
ondary air, total excess air for fixed hearth incinerators is in
the 100 to 200 percent range.
Fixed hearth units tend to be of smaller capacity than
liquid injection or rotary kiln incinerators because of physi-
cal limitations in ram-feeding and transporting large
amounts of waste material through the combustion cham-
ber. These lower relative capital costs and potentially re-
duced paniculate control requirements make them more
attractive than rotary kilns for smaller on-site installations.
1050 Seconoi
Mean Combustion
Gas Residence
Tone
Preheat
Burner
Dncriarge
to Cyclone
1400 °F-2000 °F
Auxiliary Fuel
•*• Air Distribution
Manifold
Figure S. Typical fluidLzeO bed combustion chamber.
Fluidized beds have long served the chemical processing
industry as a unit operation. This type of combustion system
has only recently begun to see application in hazardous
waste incineration. Fluidized bed incinerators may be either
circulating or bubbling bed designs.39 Both types consist of a
single refractory-lined combustion vessel partially filled
with particles of sand, alumina, sodium carbonate, or other
materials. Combustion air is supplied through a distributor
plate at the base of the combustor (Figure 5) at a rate suffi-
cient to fluidize (bubbling bed) or entrain the bed material
(circulating bed). In the circulating bed design, air velocities
are higher and the solids are blown overhead, separated in a
cyclone, and returned to the combustion chamber. Operat-
ing temperatures are normally maintained in the 1400 to
1600°F range and excess air requirements range from 100 to
150 percent.
Fluidized bed incinerators are primarily used for sludges
or shredded solid materials. To allow for good distribution of
waste materials within the bed and removal of solid residues
from the bed, all solids generally require prescreening or
crushing to a size less than 2 inches in diameter. Fluidized
bed incinerators offer high gas to solids ratios, high heat
transfer efficiencies, high turbulence in both gas and solid
phases, uniform temperatures throughout the bed, and the
potential for in situ acid gas neutralization by lime or
carbonate addition. However, fluid beds also have the poten-
tial for solids agglomeration in the bed if salts are present in
waste feeds and may have a low residence time for fine
particulates.
Regardless of the incinerator type selected, the chemical
and thermodynamic properties of the waste determine the
sizing of the combustion chamber and its operating condi-
tions (temperature, excess air, flow rates) and determine the
nature of air pollution control and ash/residue handling sys-
tems. Elemental composition and moisture content data are
necessary to determine stoichiometric combustion air re-
quirements and to predict combustion gas flow and composi-
tion. These parameters are important in determining com-
bustion temperature and residence time conditions, the effi-
ciency of waste/fuel/air mixing, and in the type and size of
air pollution control equipment Typical operating tempera-
tures, gas (and solid) residence times and excess air rates for
each of the four major incinerator types are indicated on
Figures 2—5. It is important to understand, however, that
significant deviation from these values has been observed in
actual field practice without detrimental effect on waste
destruction and removal efficiency.40
(3) Air Pollution Control. Following incineration of haz-
ardous wastes, combustion gases may need to be further
treated in an air pollution control system. The presence of
chlorine or other halogens in the waste will generally signal a
need for a scrubbing or absorption step for combustion gases
to remove HC1 and other halo-acids. Ash in the waste is not
destroyed in the combustion process. Depending on its com-
position, ash will either exit as bottom ash, at the discharge
end of a kiln or hearth for example, or as particulate matter
suspended in the combustion gas stream (fly ash). Particu-
late emissions from most hazardous waste combustion sys-
tems generally have particle diameters less than one micron
and require high efficiency collection devices to meet the
RCRA emission standards. In addition, gas cleaning systems
provide some limited additional buffer against accidental
releases of incompletely destroyed waste products. Such sys-
tems, however, are not a substitute for good combustion and
operating practices.
The most common air pollution control equipment em-
ployed in hazardous waste facilities is summarized in Table
IV.21 Most often, several of these devices are employed in
series. The most common system used is a quench (gas cool-
ing and conditioning), followed by high-energy venturi
scrubber (particulate removal), a packed tower adsorber
(acid gas removal) and a demister (visible vapor plume elim-
ination). It is interesting to note, however, that more than
half of the incinerators employ no air pollution control sys-
tem at all (Table II). This could be because these facilities
are handling low ash, low halogen content liquid waste
streams for which such control may not be necessary.
Venturi scrubbers involve the injection of a scrubbing
liquid (usually water or a water/caustic solution) into the
exhaust gas stream as it passes through a high velocity con-
Table IV. Distribution of air pollution control devices (APCD)
among hazardous waste incinerators.*
APCD type
Number
Percent
Quench
Venturi scrubber
Wet scrubber
Wet ESP
Ionizing wet scrubber
Other non-specified scrubber
Packed tower absorber
Spray tower absorber
Tray tower absorber
Other absorbers
None/unknown
Total incinerator
systems surveyed
21
32
7
5
5
12
18
2
1
2
31
90
23.3
35.6
7.8
5.5
5.5
13.3
20.0
2.2
1.1
2.2
34.4
• Total number of APCD types are greater than systems surveyed as
many incinerators report more than one APCD.
May 1987 Volume 37, No. 5
F-8
565
-------
striction, or throat. The liquid is atomized into fine droplets
which entrain fine particles and a portion of the absorbable
gases in the gas stream. The major advantage of venturi
scrubbers is their reliability and relative simplicity of opera-
tion. On the other hand, maintaining the significant pressure
drop across the venturi throat (60 to 120 inches of water
column) required for hazardous waste combustion panicu-
late matter control represents a significant percentage of the
total cost of operation of incineration facilities employing
venturi scrubbing.
Acid gas removal is generally accomplished in packed bed
or plate tower scrubbers. Packed bed scrubbers are generally
vessels filled with randomly-oriented packing material such
as polyethlyene saddles or rings. The scrubbing liquid is fed
to the top of the vessel, with the gas flowing in either concur-
rent, countercurrent, or crossflow modes. As the liquid flows
through the bed, it wets the packing material and thus pro-
vides the interfacial surface area for mass transfer with the
gas phase which is required for effective acid gas absorption.
Like packed bed scrubbers, plate scrubbers also rely on
absorption for the removal of contaminants. The basic de-
sign is a vertical cylindrical column with a number of plates
or trays inside. The scrubbing liquid is introduced at the top
plate and flows successively across each plate as it moves
downward to the liquid outlet at the tower bottom. Gas
comes in at the bottom of the tower and passes through
openings in each plate before leaving through the top. Gas
absorption is promoted by the breaking up of the gas phase
into small bubbles which pass through the volume of liquid
in each plate.
Packed bed or plate tower scrubbers are commonly used
at liquid injection incinerator facilities, where absorption of
soluble gaseous pollutants (HC1, SOZ, N02) is most impor-
tant and paniculate control is less critical. However, at rota-
ry kiln or fixed hearth facilities, or liquid injection facilities
where high ash content wastes are incinerated, venturi
scrubbers are often used in series with packed bed or plate
tower scrubbers.
Many designs in recent years have begun to incorporate
waste heat boilers as a substitute for gas quenching and as a
means of energy recovery.41-42 Wet electrostatic precipita-
tors (ESP) ionizing wet scrubbers (IWS) and fabric filters
are also being incorporated into newer systems43 largely due
to their high removal efficiencies for small particles and
lower pressure drop.
(4) Residue and Ash Handling. The inorganic compo-
nents of hazardous wastes are not destroyed by incineration.
These materials exit the incineration system either as bot-
tom ash from the combustion chamber, as contaminants in
scrubber waters and other air pollution control residues, and
in small amounts in air emissions from the stack. Under
RCRA, residues generated from the incineration of hazard-
ous waste should be managed carefully.
Ash is commonly either air-cooled or quenched with water
after discharge from the combustion chamber. From this
point ash is frequently accumulated on-site in storage la-
goons or in drums prior to disposal in a permitted hazardous
waste land disposal facility. Dewatering or chemical fixa-
tion/stabilization may also be applied prior to disposal.
Air pollution control residues are generated from the com-
bustion gas quenching, particulate removal and acid gas
absorption steps in an incineration system. These residues
are typically aqueous streams containing entrained particu-
late matter, absorbed acid gases (usually as HC1) and small
amounts of organic contaminants. These streams are often
collected in sumps or recirculation tanks where the acids are
neutralized with caustic and returned to the process. Even-
tually, a portion or all of these waters must be discharged for
treatment and disposal (generally when the total dissolved
solids level exceeds 3 percent). Many facilities discharge
neutralized waters to settling lagoons or a chemical precipi-
tation step to allow for suspended contaminants to be con-
centrated and ultimately sent to land disposal. Depending
upon the nature of the dissolved contaminants and their
concentration after treatment, waters may either be re-
turned to the process or discharged to sewers.
Other Hazardow Wast* Thermal Destruction System
Other types of systems are also being employed to ther-
mally destroy hazardous waste. These include: ocean incin-
eration vessels, mobile incinerators, and high temperature
industrial furnaces.
Ocean incineration involves the thermal destruction of
liquid wastes at sea in specially designed tanker vessels out-
fitted with high-temperature incinerators. The principle of
operation of these units is identical to that of land-based
liquid injection incinerators with the exception that current
ocean incinerators are not equipped with air pollution con-
trol systems. Acid gas produced from incinerating chlorinat-
ed wastes is discharged to the air without treatment to be
neutralized by contact with sea water, which has a naturally
high buffering capacity.
Ocean incineration has been routinely used in Europe
since 1969.44 A total of six different vessels have conducted
hundreds of waste burns in the North Sea. Although several
test burns and research studies involving herbicide orange,
mixed organo-chlorine waste and PCBs have occurred under
United States sponsorship ocean incineration has never
been used on a routine commercial basis in the United
States.45 This has been largely due to vocal public concern
over potential environmental effects that could result from
incinerator emissions and spills of hazardous materials dur-
ing loading and transport of wastes.
The importance of ocean incineration, however, is not that
it is currently a major factor in the United States hazardous
waste management but, rather, that it has served as a focal
point for public and scientific debate over the state of knowl-
edge of the character and potential effects of emissions from
incineration processes in general. These issues have been
explored in several major studies on ocean and land-based
incineration conducted over the past two years30-48'47. While
these studies have found incineration to be the most effec-
tive technology currently available for organic hazardous
waste destruction, many of the important technical and poli-
cy issues examined will significantly impact the direction of
incineration practice and research over the next decade.
A number of companies are marketing mobile or trans-
portable incineration systems. Most of these are scaled -
down, trailer-mounted versions of conventional rotary kiln
or fluidized bed incinerators. The thermal capacities of most
mobile systems range from 10 to 20 million Btu/h.
The first mobile rotary kiln incinerator was designed and
tested by EPA as a potential solution to on-sit* clean-up at
uncontrolled waste sites.19-48 Other rotary kiln systems have
since been developed and employed at waste sites. 19-49-fil
Mobile fluidized bed systems are also being marketed.52
Overall, the performance of these mobile systems has been
shown to be comparable to equivalent stationary facilities.
Current experience suggests, however, that waste incinera-
tion is more expensive in mobile units, on a unit cost basis,
than it is in stationary units. The principal advantage of
mobile systems appears to be that they are more socio-
politically acceptable than removal and transportation of
clean-up residues to commercial facilities. In the instance of
soil decontamination, on-site incineration may also be more
cost-effective than transportation of large amounts of con-
taminated material to central incineration facilities.
More substantial than either ocean or mobile incineration
practice is the use of hazardous wastes as fuels in industrial
boilers and furnaces. In 1981, these operations disposed of
more than twice the amount of waste that was disposed of via
incinerators.2 Processes that have burned or do burn hazard-
568
F-9
JAPCA
-------
ous waste materials as fuels include: industrial boilers, ce-
ment kilns, iron-making furnaces, and light-weight aggre-
gate and asphalt plants. The principal attractions to this
approach include: exemption (currently) from RCRA incin-
eration standards, fuel and waste trarspor*ation cost-sav-
ings, and waste disposal cost savings since hazardous waste
used as a fuel can be sold.
The most recent source of information on waste fuel use in
industrial processes was compiled for EPA in 1984.53 The
study synopsizes results of a national questionnaire of waste
fuel and waste oil use in 1983. The study revealed that there
were over 1,300 facilities using hazardous waste-derived fu-
els (HWDF), accounting for a total of 230 million gallons per
year. The chemical industry (Standard Industrial Classifica-
tion 28) accounted for 67 percent of this while operating only
12.4 percent of the facilities using HWDF- Other industries
employing significant quantities of hazardous waste as fuel
included: SICs 26 (paper), 29 (petroleum), 32 (stone, clay,
glass, concrete), and 33 (primary metals). The majority (69
percent) of the waste was burned in large quantities by a few
facilities representing only a small fraction (1.6 percent) of
the 1,300 facilities. These included medium- to large-sized
industrial boilers, cement and aggregate kilns, and iron-
making furnaces.
While waste code-specific data on HWDF are not readily
available, recent data indicate that of the HWDF burned in
1983, 30 percent was organic solvents, and 45 percent was
other hazardous organics.53 Most of this waste was generated
on-site and 74 percent of the balance arrived directly from
an off-site generator rather than through an intermediary.
Measuring Process Performance
Proper and accurate measurement of the emissions from
incineration systems is a critical issue. Great demands have
been placed upon sampling and analysis technology by the
RCRA incinerator regulations. Fortunately, significant pro-
gress has been made in adopting measurement methods to
the rigors of specific compound identification and the level
of detection and accuracy which are often necessary to assess
compliance with the RCRA incineration standards. These
methods will rarely be a limitation in assessing incinerator
performance if proper attention is given to quality assurance
and quality control, adequate advanced planning is conduct-
ed, and experienced personnel are involved in sampling and
analysis activities.
Performance measurement may have any of the following
three purposes:
• to establish compliance with performance standards
(e.g., trial burns)
• to monitor process performance and direct process con-
trol (e.g., continuous monitoring)
• • to conduct performance measurements for research and
equipment development purposes
Methods employed in assessing regulatory compliance are
generally official methods which have been standardized
and published in the Federal Register or in EPA guidance
documents. Routine performance monitoring for process
control often involves the use of continuous monitors for
emissions and facility-specific engineering parameters (e.g.,
temperature, pH, kiln rotation). Research and equipment
assessment investigations may involve any of the above
techniques in combination with standard and often non-
standard sampling and analysis techniques designed for rap-
id screening of performance or for ultra-low detection of
specific materials.
Performance Measurement
Figure 6 illustrates sampling points which may be in-
volved in assessing incinerator performance. In the case of
trial burn activities, the main focus of sampling activities is
on collection of waste feed and stack emission samples. Ash
Solid
Waste
legend
Pressure
temperature
Flow Rate
Differential Pressure
Figure 6. Potential sampling points lor assessing incinerator performance
May 1987 Volume 37, No. 5
567
F-10
-------
iable V. sampling methods and analysis parameters.
Sample
1. Liquid waste feed
2. Solid waste feed
3. Chamber ash
4. Stack gas
Sampling frequency
for each run
Grab sample every 15 inin
Grab sample from each drum
Grab one sample after all
3 runs are completed
Composite
Three pairs of traps
40 min each pair
Composite in Tedlar gas bag
Compounds in mylar gas bag
Continuous (3 h)
Sampling
method*
S004
S006, S007
S006
MM5 (3 h)
VOST (2 h)
son
M3 (1-2 h)
Continuous
monitor
Analysis parameter6
V&SV-POHCs. Cl . ash.
ult. anal., viscosity. HHV
V&SV-POHCs.Cl.ash,
HHV
V&SV.POHCs.TCLI*
SV-POHCs, paniculate.
HjO. HC1
V-POHCs
V-POHCs'
CO2 and O2 by Orsat
CO (by plant's monitor)
• VOST denotes volatile organic sampling train; MM5 denotes EPA Modified Method 5; M3 denotes EPA Method
3; SXXX denotes sampling methods found in "Sampling and Analysis Methods for Hazardous Waste Combus-
tion." **
b V-POHCs denotes volatile principal organic hazardous constituents (POHCs); SV-POHCs denotes semivolatile
POHCs; HHV denotes higher heating value.
c Gas bag samples may be analyzed for V-POHCs, only if VOST samples are saturated and not quantifiable.
d TCLP—toxicity characteristic leaching procedure.58
and air pollution control system residues are also sampled
and analyzed. Sampling of input/output streams around in-
dividual system components (e.g., scrubbers) may also be
conducted In research testing or equipment evaluation stud-
ies.
The main focus of analytical activities is on POHCs. Stack
gas analysis also includes determination of HC1 and particu-
late emissions, and may be extended to a determination of
other organic compound emissions as well as metals of con-
cern. In the case of paniculate emissions, the size distribu-
tion of stack panicles may also be of interest. The size of
emitted paniculate affects its transportation and fate in the
atmosphere and influences the likelihood of inhalation, an
important factor in health effects assessment. Few hazard-
ous waste incinerator tests have actually collected particle
size data, primarily due to time and funding limitations.
EPA has provided guidance on the types and methods of
sampling and analysis to be used in trial burns designed to
measure facility compliance with the RCRA incinerator
standards.54'55 Additional guidance is being prepared. Simi-
lar guidance has also been provided for PCB incinerators.56
Table V outlines sampling and analysis methods typically
involved in RCRA trial burns. For any trial burn, at any one
set of operating conditions and waste feed conditions, three
replicate runs (i.e., identical as possible) are usually recom-
mended to obtain a representative assessment of incinerator
performance.55
The sampling method numbers in Table V refer to meth-
ods identified in a manual of combustion sampling and anal-
ysis methods compiled by EPA.57 This manual expands
upon and augments the information in EPA SW-846, "Test
Methods for Evaluating Solid Waste: Physical/Chemical
Methods"S9 and "Samplers and Sampling Procedures for
Hazardous Waste Streams." m Together, these references
are the best sources from which to identify methods to be
used in incinerator performance evaluations.
Analytical methods for specific hazardous compounds are
often of greatest interest. Analytical methods for Appendix
VIII compounds10 in these references are generally based
upon high-resolution fused-silica-capillary column gas chro-
matography (GO in combination with mass spectrometry
(MS) for specific compound detection. High-performance
liquid chromatography (HPLC) is recommended for deter-
mination of compounds that are inappropriate for detection
by GC/MS. Application of analytical methods has been eval-
uated for 240 of the approximately 400 Appendix VIII Com-
pounds.61-62 The methods showed acceptable precision in
determination of most of the compounds. Detection limits in
synthetic samples were on the order of 1 to 10 nanograms per
injection, but detection in actual waste samples will be de-
pendent upon the nature of interferences in the waste ma-
trix.
While all emissions from hazardous waste incinerators are
important, the greatest interest is most often placed on stack
emissions. The accuracy and reliability of stack sampling
results are central to the entire issue of incinerator perfor-
mance and environmental safety. Existing methods have
been the subject of substantial research, debate and, in some
cases, criticism.63
Stack emissions are sampled to determine stack gas flow
rate, HC1, particulate concentration and the concentration
of organic compounds of interest. Determination of stack gas
flow rate and particulate emissions is performed using the
conventional stack sampling method commonly referred to
as Method 5 (M5). This method encompasses EPA Methods
1 through 5 and is defined in detail in 40 CFR Pan 60,
Appendix A. HC1 emissions are sampled by modifying the
Method 5 train to include a caustic impinger. A specialized
sampling and analytical method has also been developed to
further speciate and quantify hydrogen halide and halogen
emissions.64
The technology of incinerator stack sampling for trace
organic compounds is sophisticated. While the basic tech-
nology is well developed, many pitfalls await those who at-
tempt the job without sufficient knowledge or experience.
Sampling of stack effluent for organics, in order to deter-
mine DRE, may require from one to three separate methods
(or more), depending on the number of compounds to be
quantified and their characteristics, and on the detection
limits that are required to prove a DRE of 99.99 percent or
establish levels of incomplete combustion byproducts. Spe-
cial attention must be given to sampling rate and duration in
planning for emission tests to insure that a sufficient amount
of sample is collected to meet detection limit objectives and
to allow for all necessary analyses to be completed.55
The three methods for hazardous waste incinerator sam-
pling are:
1. Modified Method 5 (MM5)
2. Volatile Organic Sampling Train (VOST)
3. Gasbags
The Modified Method 5 (MM5) train is used to capture
568
F-ll
JAPCA
-------
semivolatile (boiling point 100°C to 300°C) and non-volatile
(boiling point >300°C) organic compounds. The MM5 is
merely a simple modification of the M5 train involving inser-
tion of a sorbent module (XAD-2 resin) between the filter
and the first impinger.65 It is recommended that a separate
MS train for particulate determination be used in tandem
with the MM5 train since drying of the filter for particulate
determination may invalidate analysis of organic com-
pounds on the filter.55 Like the M5, the MM5 involves iso-
kinetic traversing of the stack with a sampling probe. Water-
cooled sample probes are necessary for sampling of hot com-
bustion gases in regions ahead of quenching. Where it is
desirable to collect larger amounts of sample for more exten-
sive analysis or lower detection limits, the much larger
Source Assessment Sampling System (SASS) may be used
instead of the MM5.55 SASS involves single point (pseudo-
isokinetic) sampling at a rate of 110 to 140 L/min (4 to 5 cfm)
compared ~* the 14 to 28 L/min (0.5 to 1 cfm) rate of the
MM5. The iame sorbent resin (XAD-2) is also used. Because
of its more convenient sample size and portability, and its
multipoint isokinetic sampling the MM5 train is generally
preferred over the SASS train.
For volatile organic compounds (boiling point 30°C to
130°C), the Volatile Organic Sampling Train (VOST) is
used. The VOST was developed by EPA in 1981 to enable
detection of stack concentrations of volatile organic com-
pounds as low as 0.1 ng/L.66 This detection limit was deemed
necessary to be able to demonstrate greater than 99.99 per-
cent DRE for volatile organic compounds at concentrations
as low as 100 ppm. in the waste feed. The VOST system
involves drawing a single stack gas sample through two sor-
bent tubes in series. The first tube contains Tenax resin and
the second, Tenaz and activated charcoal. Up to six pairs of
sorbent tubes operating at one L/min for 20 minutes each
may be needed to achieve the lowest detection levels.67 For
higher stack gas concentrations, however, the VOST may be
operated at lower flow rates with less pairs of tubes (longer
sampling times for each pair).
Various types of gas sampling bags may also be used to
sample for volatile organic compounds. These are generally
appropriate only for higher organic concentrations. The ac-
curacy of sampling with this method is a function of the
Table VI. Summary of continuous emission monitors.
Pollutant
Monitor type
Expected
concentration
range
Available
range1
CO-
CO
NO,
SO.
SO,
Organic
compounds
Paramagnetic
Electrocatalytic (e.g.,
zirconium oxide)
5-14%
NDIR
Chemiluminescent
Flame photometry
Pulsed fluorescence
NDUV*
Colorimetric
Gas chromalography
(FID)*
Gas chromatography
(BCD)'
Gas chromatography
(PID)f
IR absorption
UV absorption
GC/MS
2-12%
0-100 ppm
0-4000 ppm
0-4000 ppm
0-100 ppm
0-50 ppm
0-25%
0-21%
0-5000 ppm
0-10000 ppm
0-5000 ppm
0-50 ppm
0-100 ppm
• For available instruments only. Higher ranges are possible through
dilution.
" Nondispersion infrared. ' Electron capture detector.
r Nondispersion ultraviolet. ' Photo-ionization detector.
A Flame ionization detector.
sampling and storage characteristics of the bags.68 The use of
extensive quality assurance and quality control procedures
is required with both plastic bags and the VOST to avoid
sample contamination in the field and in transit69 This
problem was not fully appreciated in some of the early field
tests employing the VOST.
Both the MM5 and VOST sampling methods have been
subjected to laboratory and field validation studies for se-
lected compounds.7*"72 These studies have demonstrated
that excellent results are possible with these methods. It is
important to note, however, that modifications of these
methods may be required for certain POHC compounds
which become chemically or physically altered in the sam-
pling systems. Highly water soluble compounds (e.g., aceto-
nitrile) and water-reactive compounds (e.g., phthalic anhy-
dride), for instance, present special challenges to current
sampling methods.
Process Monitoring
Measurement of a wide variety of incinerator operating
parameters may be necessary to maintain thermal destruc-
tion conditions which are equivalent to those observed dur-
ing a successful trial burn. These measures are used as indi-
cators of the performance of the incineration system and
serve as input to automatic and manual process control
strategies. There are nearly two dozen potential measure-
ments, including such parameters as: combustion tempera-
ture, waste feed rate, oxygen and carbon monoxide (CO)
concentration in the stack, gas flow rate at strategic points,
and scrubber solution pH. These parameters and their use
are described in detail in a number of resource docu-
ments.32-54-55-73
Continuous emission monitors (CEM) are often used or
required in measuring combustion gas components such as
carbon monoxide (CO), oxygen (©2), nitrogen oxides (NOZ),
and total unburned hydrocarbons (TUHC). If properly in-
terpreted, combustion gas components may be indicators of
the completeness of the thermal destruction reaction. These
methods typically require extraction of gas samples from the
gas stream of interest and measurement with a remote in-
strument. Some parameters such as CO and 0? may be
measured in-situ (in the stack). Table VI summarizes moni-
tor types, and available concentration measurement ranges
for a number of CEMs.74
RCRA incinerator operating requirements stipulate that
the permit specify an operating limit for CO concentration
in the stack gas. While continuous CO monitoring is re-
quired, specific CEM requirements are not specified. How-
ever, permit writers frequently cite Reference Method 10
from the EPA New Source Performance Standard (NSPS)
as a guide (40 CFR 60, Appendix A). CO concentration in
stack gas is an indicator of combustion efficiency. However.
this parameter is also being examined as a possible real-time
indicator of DRE for hazardous organic compounds.
Total unburned hydrocarbon (TUHC) emissions are also
being considered as a possible DRE indicator. Three types of
analyzers are available: flame ionization detector (FIDl;
photo-ionization detector (PID); and, the automated total
gaseous nonmethane organics analyzer (TNMO). FID is
used most typically. All three methods have limitations. The
theoretical minimum detectability of all three techniques in
significantly below 10 Mg/dscf (detection in the 10 to 700 «t>
dscf would be required for DRE correlation purposes). How
ever, due to electrical noise, sampling limitations and rum
ponent degradation, the practical field detectability limit "if
from 40 jig/dscf to 200 Mg/dscf. if care is exercised.'1 Thr
individual methods are also limited by the fact thev d<> n»t
respond to all classes of organic compounds of
concern.
Mav1987
Volume 37. No 5
F-12
-------
lab/c Vll. fjicnirrauir periurmanyy and *iack emissions data (data reported as averages for each facility).
Facility tvpe
Commercial rotary kiln/
liquid incinerator
Commercial fixed hearth,
two-stage incinerator
On-site two-stage liquid
incinerator
Commercial fixed hearth.
two-stage incinerator
On-site liquid injection
incinerator
Commercial two-stage
incinerator
On-site rotary kiln
incinerator
Commercial two-stage
fixed hearth incinerator
On-site rotary kiln
On-site liquid injection
incinerator
On-site rotary kiln
incinerator
On-site rotary kiln
incinerator
On-site liquid injection
incinerator
On-site liquid injection
incinerator
On-site fluidized bed
incinerator
On-site fixed hearth
incinerator
On-site liquid injection
incinerator
On-site liquid injection
incinerator
Commercial rotary kiln
incinerator
On-site liquid injection
incinerator
On-site liquid furnace
incinerator
On-site fixed hearth
incinerator
CM*)
10.5
11.4
8.1
11.0
13.2
10.2
9.7
13.4
c
9.7
10.7
14.1
12.4
9.3
3.6
12.9
4.5
3.6
9.4
3.1
6.4
13.5
COippm)
6.2
6.9
9.4
327.7
11.9
1.1
554.0
26.8
794.5
66.3
5.8
323.0
31.9
1.0
67.4
NDd
358.0
28.4
8.0
779.3
56.3
5.0
Tt'HC (ppm)
1.0
1.0
6.0
18.7
1.0
1.3
61.7
1.8
NA*
7.8
NA
NA
1.9
NA
NA
NA
NA
NA
0.5
NA
NA
NA
ORE
99.999
99.994
99.994
99.997
99.999
99.998
99.999
99.996
99.998
99.994
99.996
99.996
99.999
99.996
99.996
99.999
99.995
99.998
99.999
99.999
99.999
99.999
Paniculate
(mg/m-1)
152
400
143
60
186
902
23
168
184
95
404
NA
163
40
259
93
99
12
172
88
4
150
HCI
control
99.4
98.3
99.7
b
b
b
99.9
98.3
99.7
. b
99.9
99.8
98.6
b
b
b
b
99.3
99.9
99.6
99,9
98.4
NA—not available.
" HCI emissions <4 Ib/h.
c Reported only as a range (3.1-16.7%).
d Not detected.
Emissions from Hazardous Waste Incineration
Ideally, the primary products from combustion are carbon
dioxide, water vapor, and inert ash. In reality, what appears
outwardly to be a straightforward, simple process is actually
an extremely complex one involving thousands of physical
and chemical reactions, reaction kinetics, catalysis, combus-
tion aerodynamics, and heat transfer. This complexity is
further aggravated by the complex and fluctuating nature of
the waste feed to the process. While combustion and inciner-
ation devices are designed to optimize the chances for com-
pletion of these reactions, they never completely attain the
ideal. Rather, small quantities of a multitude of other prod-
ucts may be formed, depending on the chemical composition
of the waste and the combustion conditions encountered.
These products along with potentially unreacted compo-
nents of the waste become the emissions from the incinera-
tor.
Hydrogen chloride (HCI) and small amounts of chlorine
(CU), for example, are formed from the incineration of chlo-
rinated hydrocarbons. Hydrogen fluoride (HF) is formed
from the incineration of organic fluorides, and both hydro-
570
gen bromide (HBr) and bromine (Brj) are formed from the
incineration of organic bromides. Sulfur oxides (SOj), most-
ly as sulfur dioxide (SO2>, but also including 1 percent to 5
percent sulfur trioxide (SOs), are formed from the sulfur
present in the waste material and auxiliary fuel. Highly
corrosive phosphorus pentoxide (PsOs) is formed from the
incineration of organophosphorus compounds. In addition,
oxides of nitrogen (NO,) may be formed by fixation of nitro-
gen from nitrogen compounds present in the waste material
or in the combustion air. Suspended paniculate emissions
are also produced and include particles of mineral oxides
and salts from the mineral constituents in the waste materi-
al. A wide range of organic compounds may also be formed
from the incomplete thermal destruction of organic com-
pounds in the waste and auxilary fuel.
Until recently, there were only limited data available on
waste destruction performance and pollutant emissions
from hazardous waste thermal destruction devices. Studies
by EPA and others in the early to late 1970s employed a
variety of evolving trace organic pollutant sampling and
analysis techniques and were often targeted only towards
measuring macro-destruction and combustion efficien-
JAPCA
F-13
-------
Table Yin. Summary of boiler performance.
Facility type
Watertube stoker
Packaged firetube
Field erected watertube
Converted stoker
Packaged watertube
Converted watertube
Modified firetube
Tangentially fired
watertube
Packaged watertube
Packaged firetube
Packaged watertube
Load(%)
100
25
26
78
36-73
53
44
100
65
50-100
82
(>2(%)
6-16
10
4-6
6-7
7-11
8
6
2
3-8
4
Residence
time(s)
1.2
0.8
2
1.1
1.1-0.5
2
0.4
2
1.8
0.7-0.3
1.8
Average
volumetric heat
release rate
(kW/m')
509
739
78
339
960
107
807
180
343
1240
269
DRE*
99.98
99.991
99.999
99.998
99.995
99.98
99.998
99.991
99.998
99.999
99.999
W/F*(%)
40
0.1-0.5
37
18-48
19-66
8.7-10.1
100
2.4-4.3
8.2
100
49
NOX
(ppm)b
163-210
40-65
61-96
193-250
164-492
243-328
67-74
393-466
64-78
410-1125'
85-203
154-278
CO
(ppm)b
900-1200
47-S8
18-21
75-127
83-138
109-139
146-170
142-201
46-750
20-135
102-119
• W/F «= waste heat input as a percent of total heat input
b Range of average values across individual sites and runs including baseline.
' Higher values are for high nitrogen content waste firing.
d Mass weighted average for all POHCs in the waste MOO ppm.
cies,75-77 rather than the performance standards now re-
quired under the Resource Conservation and Recovery Act
of 1976. Since 1981, however, EPA has conducted a substan-
tial program of performance testing at thermal destruction
facilities. The testing was designed to estimate the environ-
mental impact of these operations and to provide informa-
tion on the ability of these facilities to control emissions to
the degree required by the 1982 incinerator performance
standards. The test facilities, test procedures and perfor-
mance results have been summarized78 for the facilities test-
ed (incinerators, industrial boilers and industrial process
kilns). Complete test reports have been published for the
incinerators,79 industrial boilers,80-81 and cement/aggregate
kilns8233 tested. These data as well as trial burn results from
14 additional RCRA incinerators have been summarized
recently in an EPA report, "Permit Writer's Guide to Test
Burn Data—Hazardous Waste Incineration."84
The following sections summarize these data in five areas:
• RCRA-regulated performance and emissions (DRE, par-
ticulate matter, and HC1)
• Metal emissions
• Combustion by-product emissions
• Dioxin and furan emissions
• Ash and air pollution control residue quality
RCRA Regulated Performance and Emissions
Tables VII, VHI, and DC summarize waste destruction
efficiency, HC1 and particulate emissions results for the in-
cinerators, industrial boilers, and cement kilns tested. The
tables also summarize certain process operating parameters
as well as emissions of CO and Oj and, in some instances NOX
and SO,.
These data reveal that well operated incinerators, indus-
trial boilers, and process kilns are capable of achieving a
99.99 percent DRE, the RCRA performance standard. All of
the incinerators tested by EPA achieved this level of perfor-
mance for candidate POHC compounds in concentrations
greater than 1200 parts per million (ppm) in the waste
feed.85 Candidate POHC compounds between 200 and 1200
ppm frequently were not destroyed to a 99.99 percent DRE
and no compounds below 200 ppm in the waste feed met the
RCRA DRE limit. In fact, regression analysis of the pooled
Table IX. Summary of industrial kiln performance and stack emissions data (data reported as averages for each
facility).
Facility type
Wet process cement kiln
(non-atomized waste)
Wet process cement kiln
(atomized waste)
Dry process cement kiln
(non -atomized waste)
Dry process cement kiln
(atomized waste)
Lime kiln
(atomized waste)
Shale aggregrat« kiln
(atomized waste)
Clay aggregate kiln
(atomized waste)
Clay products kiln
(atomized waste)
Test-
W
B
W
B
W
B
W
B
W
B
W
W
W
DRE
(%)
99.200
—
99.996
—
99.998
—
99.992
—
99.997
—
>99.99
99.998
>99.99
Particulate
(kg/MG)«
0.27
0.26
0.27
0.26
—
—
—
0.11
0.10
0.33
0.58
0.002
HC1
(kg/h)
0.36
0.09
2.1
0.6
11.5
1.3
0.47
0.25
0.20
0.09
2.1
0.023
0.84
NOi
(ppm)
68
136
478
371
814
620
486
680
446
386
—
162
—
S02
(ppm)
450
279
265
636
19
7
27
27
596
553
__
1130
—
W/F(%)b
25
—
15
45
—
15
—
30
_
100
59
100
• W = waste testing, B = baseline (fossil fuel only).
b W/F = waste fuel heat input expressed as a percent of total heat input.
c Particulate emissions are expressed as kg particulate per metric ton (MG) of product produced (e.g., cement, lime).
May 1987 Volume 37. No. 5
571
-------
Table X. Average stack emissions ol meiais lr»m five hazardous waste inc.nerators.'
Metals
Plant A
Emission rate (g/kJ) and concentration on paniculate (^g/i
Plant B Plant B
uncontrolled
controlled
Plant C
Plant D
Plant E
-Sb
As
Be
Be
Cd
Cr
Pb
Ng
Ni
Se
Ag
Ti
0.32
(8,380)b
—
0.052
(840)
0.055
(890)
0.14
(2,300)
5.4
(85,500)
d
0.024
(400)
—
0.0008
(11.4)
0.0089
(140)
0.26
(300)
c
0.19
(250)
0.11
(140)
0.73
(950)
2.3
(3,100)
c
0.50
(650)
7.0
(9,200)
c
c
c
c
0.11
(930)
c
0.019
(150)
0.19
(1,500)
0.64
(5,300)
c
0.087
(740)
0.45
(3,700)
c
c
c
c
c
c
c
2.5
(47,500)
c
c
2.7
(49,000)
c
0.33
(620)
c
0.056
(1,490)
0.012
(1,120)
d
0.24
(25,600)
0.052
(4,570)
0.29
(34,600)
0.0076
(917)
d
0.050
(470)
d
0.36
(4.000)
0.094
(1,100)
9.0
(98,000)
d
a Adapted from Wallace et al., 1985.97
b Numbers in parentheses represent values for Mg of metal/g of emitted participate.
' All values below detection limit.
d Some values below detection limits, average not calculated.
* Emissions are reported as gram of paniculate emitted per hour divided by the design heat release rate of the incinerator in
kJ/h.
data suggested that statistically significant correlations
(correlation coefficients were 0.76 and 0.84) existed between
compound penetration (1-DRE) and compound feed con-
centration, showing that DRE increased with waste feed
concentration.85
This phenomenon, which has been observed in tests of
other thermal destruction devices, was not anticipated. A
number of possible explanations have been advanced.2-79
The most frequently stated theory postulates that at the
very low stack emission concentrations (<1 ng/L) necessary
to demonstrate greater than 99.99 percent DRE for a sub-
1200 ppm compound, sufficient amounts of that compound
may actually be formed as an incomplete combustion or
recombination byproduct from other compounds in the
wastes to effect a reduction of the DRE below 99.99 percent.
Others argue that limitations of current stack sampling and
analysis techniques for such low levels of trace organic com-
pounds are responsible.
EPA is conducting research to assess this concentration
phenomenon. From a regulatory standpoint, however, this is
not currently perceived as an issue. Few, if any, of the low
concentration compounds in the wastes identified in the
EPA test program would have actually been selected as
"principal" organic hazardous constituents in trial burns if
existing EPA guidance on POHC selection were employed.
It is also important to note that even though DRE declines
with lower initial compound concentrations in the waste, the
amount absolute of compound emitted also declines. In fact,
the DRE vs. concentration correlation noted above actually
predicts that the net emissions resulting from a reduced
DRE for a 100 ppm compound will actually be slightly less
than those for a 99.99 percent DRE for the compound at
1000 ppm in the waste.
Table VIII indicates that industrial boilers, particularly
the larger water tube units, typically attain 99.99 percent
DRE. Cement kilns, lime kilns, and light weight aggregate
kilns with good combustion control and waste atomization
all met or exceeded the 99.99 percent DRE (Table DC).
All incinerators and industrial process kilns tested met or
approached the RCRA HC1 removal standard of 99 percent.
Industrial boilers typically have no existing controls for HC1,
but none exceeded the 1.8 kg/h emission standard because
wastes with low net chlorine content were employed.
Achieving the RCRA paniculate emission standard of
0.08 gr/dscf was a problem for a number of the incinerators
tested by EPA. Four of the eight units tested failed to meet
the RCRA standard. Two of those facilities were marginally
above the emission limit and could likely meet the standard
with minor operating adjustments. The remaining two facili-
ties appeared to need significant design and/or operational
changes.86 In some cases, failure of the paniculate emission
standard may be attributed to dissolved neutralization salts
in mist carryover from caustic scrubbers.
No significant changes in paniculate emissions were ob-
served for industrial boilers and certain of the industrial
proess kilns when they fired waste fuels compared to emis-
sions for fossil fuels only.87-88 Some increased emissions were
observed in kilns employing electrostatic precipitators for
paniculate control. These increases were attributed to
changes in the electrical resistivity of the particles due to the
presence of increased chloride levels. Adjustments in ESP
operation should correct this in most cases.
Metal Emission*
Metals such as arsenic, barium, beryllium, chromium,
cadmium, lead, mercury, nickel, and zinc are of possible
concern in waste incineration because of their presence in
many hazardous wastes and because of possible adverse
health effects from human exposure to emissions. Incinera-
tion will change the form of metal fractions in waste streams,
but it will not destroy the metals. As a result, metals are
expected to emerge from the combustion zone essentially in
the same total quantity as the input. The principal environ-
mental concern, therefore, centers around where and in what
physical or chemical form the metals end up in the combus-
tion system, i.e., bottom ash, in APCD residues, or stack
emissions.
Most interest has traditionally focused on stack emissions
of metals. Increasing attention, however, is now being given
572
JAPCA
F-15
-------
to the quality of residuals from incineration of metal-bearing
wastes since disposal of these materials may be subject to
tough, restrictions on land disposal under HSWA.
Metals present in the feed to combustion devices are typi-
cally emitted in combustion gases as particles rather than
vapors. However, some of the more volatile elements (e.g.,
mercury and selenium) or their chemical compounds may be
released to the atmosphere partially in the vapor state. The
processes involved in the formation of particles are very
complex and are only partially understood at present. Most
of the current state of knowledge on metal behavior in com-
bustion has come from research on coal combustion.89"92
In general, data on metal emissions and partitioning for
hazardous waste incineration are limited and often incom-
plete. Organic emissions have been the focus of most histori-
cal emissions assessments of these facilities. Data on air
pollution control device effectiveness for metals are even
more scarce.
In 1982, Gorman et al. collected data on metal emissions
from the Cincinnati MSD incinerator.93 Metals data were also
collected in three earlier tests sponsored by EPA in 1976.94"96
The best source of metal emissions data, however, is the
series of incinerator tests more recently conducted by EPA
in support of the Regulatory Impact Analysis for the 1982
RCRA incinerator standards.79 Average emission rates for
these latter tests are shown in Table X. Actual values varied
over a considerable range.
Wallace et al. recently reviewed all of the available metals
emissions data for hazardous waste incinerators and com-
pared them to emissions for other conventional combustion
sources.97 The five most frequently detected metals are Ba,
Cd, Cr, Pb, and Ni. Hg was not found in any of the emissions,
but this was because only paniculate metal and not vapor
was sampled in the studies. The relatively volatile metals
(Sb, Cd, Pb) generally show enrichment in fine particulate
emissions (i.e., higher concentration relative to their concen-
tration on larger panicles). This enrichment phenomenon is
an important consideration for health assessment studies,
since the fine particulates are also more likely to be inhaled.
Metal emissions from the hazardous waste incinerators
tested were equivalent to those reported for municipal solid
waste incinerators.99 The emissions appear to be from two to
20 times higher than those from sewage sludge incinera-
tors.100 It is important to note, however, that it is difficult to
extend the findings from these few tests to hazardous waste
incinerators in general. This is because metal emissions are a
function of the amount of metal input to the incinerator
(which is highly variable day to day and facility to facility) as
well as the efficiency of the incinerator in controlling metal
emissions. Some of the incinerators tested, for instance, em-
ployed no air pollution control equipment.
Even where air pollution control equipment is used, it is
difficult to draw conclusions on air pollution control system
efficiency for metals because of the uncertainty of the sam-
pling methods usually employed for control device input
rates, the lack of panicle size information and the relatively
low quantities of metals available in collection samples. EPA
Table XI. Most frequent thermal destruction process stack
emissions.
Volatile compounds
Semivolatile compounds
Benzene
Toluene
Carbon tetrachloride
Chloroform
Methylene chloride
Trichloroethylene
Tetrachloroethylene
1.1.1 -Trichloroethane
Chlorobenzene
Naphthalene
Phenol
Bis(2-ethylheiyiphthalale
Diethylphthalate
Butylbenzylphthalate
Dibutylphthalate
is in the process of assembling a report on all of the available
air pollution control device efficiency data for metals.
Metals emissions may also be of concern for high tempera-
ture industrial processes employing hazardous waste as a
fuel. Much of the test data in this regard are for waste oil
combustion102-103 in small boilers. Lead is the primary ele-
ment of concern here since 50 to 60 percent of the combus-
tion chamber input generally exits the stack. Stack concen-
trations of lead of 5,000 to 72,000 uglm3 have been observed
in tests of small boilers of sizes ranging from 0.5 to 15 X 106
Btu/h.103 EPA has promulgated rules controlling the metals
level in used oils.13 In addition, restrictions on waste metal
content are being considered in proposed rules for industrial
boilers and furnaces to be published in 1987.14
Combustion Byproduct Emissions
The current RCRA incineration standards regulate de-
struction and removal only for the major hazardous com-
pounds in the waste. However, even under good combustion
conditions, incomplete combustion byproducts may be
emitted. One of the concerns expressed by some scientists
and environmentalists regarding hazardous waste thermal
destruction is the possible impact on human health and the
environment of emissions of potentially hazardous products
of incomplete combustion (often referred to as PICs). While
many of the incinerator field tests conducted to date have
attempted to quantify byproduct emissions, these data have
been criticized as being incomplete and insufficient for the
purposes of a full risk assessment.46 Testing has focused
largely on identification of Appendix VTII organic com-
pounds only. Comparison of total hydrocarbon emissions
with the total quantity of specific organic compounds identi-
fied in the emissions has revealed that only a relatively small
percentage of the total hydrocarbon emissions may have
been identified.104
Incomplete combustion byproducts from hazardous waste
incineration have been recognized for some time. Early pi-
lot-scale studies of the thermal destruction of the pesticide,
Kepone, found emissions of hexachlorobenzene and several
other "daughter products" which had been predicted from
previous laboratory-scale studies.105 Similar thermal decom-
position studies followed for PCBs106 and dozens of other
compounds.107-110
While the RCRA incinerator standards do not currently
regulate incomplete combustion byproducts, the earlier pro-
posals did recognize and discuss this issue. The January 1981
Phase I rule proposed that emissions of incomplete combus-
tion byproducts be limited to 0.01 percent of the POHC
input to hazardous waste incinerators.10 Although no final
action was taken on that aspect of the rule, researchers,
regulators, and environmentalists have pursued the ques-
tion of PICs with great vigor since then, including various
attempts to compare actual field performance results to the
porposed standard.111"113
One of the basic problems in assessing the results of lab-
oratory and, particularly, field studies of PIC emissions is
the fact there is no standardized definition of what a PIC is.
While a POHC is defined in the RCRA regulations, a PIC is
not, in a rigorous sense. Thus, there is often confusion even
among scientists working in the area. Strictly speaking, PICs
are organic compounds which are present in the emissions
from the incineration process, which were not present or
detectable in the fuel or air fed to the incinerator. In EPA's
test program, compounds were considered to be PICs if they
were regulated organic compounds (i.e., listed in Appendix
VIII of CFR 40 Part 261) which were detected in stack
emissions, but not present in the waste feed at concentra-
tions greater than 100 ppm.111
Compounds in the emission stream which are identified as
PICs may actually result from any one of the following phe-
nomena:
May 1987 Volume 37, No. 5
F-16
573
-------
Compounds resulting iroin trie incomplete destruction oi
the POHCs, i.e., fragments of the original POHCs.
New compounds "created" in the combustion zone and
downstream as the result of partial destruction followed
by radical-molecule reactions with other compounds or
compound fragments present. These compounds may
also result from the incomplete combustion of non-Ap-
pendix VIII compounds in the waste. This aspect may be
especially significant where fossil fuel is used in incinera-
tion and where waste is fired into conventional industrial
furnaces as only a percentage of the heat input.
. An Appendix VIII compound originally present in the
feed stream before incineration but not specifically iden-
tified as a POHC.
. Compounds from other sources (e.g., ambient air pollut-
ants in combustion air). In some field tests, compounds
identified in the stack emissions as PICs were actually
found to have come from contaminants (trihalometh-
anes) in the potable water used for scrubber water make-
up.111
latile compounds. The compounds thai occurred most fre-
quently and in the highest concentrations, nine volatile and
six semivolatile, are listed in Table XI. Emission rates for
incinerators, boilers, and kilns are shown in Table XII for 12
of these compounds for which sufficient data are available
for comparison. The data show that values from test run to
test run varied considerably. Thus, these data do not allow
prediction of levels for all three combustion devices. Many of
the volatile compounds showed higher levels for boilers, and
semivolatile compounds tended to be higher for incinera-
tors.
Data were also available from several baseline (no waste
firing) tests on boilers and kilns which allowed comparison
of emissions from hazardous waste combustion with com-
bustion of other fuels. While there was a wide range in values
from test to test, the data suggested there is little inherent
difference between waste and fuel combustion emissions.112
Sufficient data for five semivolatile compounds were
available to compare their emissions when burning hazard-
ous waste versus their emissions from municipal incinerators
Table XII. Emission rates of specific compounds from incinerators, boilers, and kilns, ng/kJ.'
Incinerators
Boilers
Kilns
Mean
Range
Mean
Range
Mean
Range
Benzene
Toluene
Carbon tetrachloride
Chloroform
Methylene chloride
Trichloroethylene
Tetrachloroethylene
1,1,1-Trichloroethane
Chlorobenzene
Naphthalene
Phenol
Diethylphthalate
87
1.6
0.8
3.8
2.2
5.2
0.3
0.3
1.2
44
7.8
3.7
2-980
1.5-4.1
0.3-1.5
0.5-8.4
0-9.6
2.3-9.1
0-1.3
0-1.3
0-6.0
0.7-150
0-16
2.8-4.8
30
280
1.8
120
180
1.2
63
7.5
63
0.6
0.3
0.4
0-300
0-1,200
0-7.2
0-1,700
0-5,800
0-13
0-780
0-66
0-1,000
0.3-2.1
0-0.8
0.04-1.6
580
1.3
2.4
152
0.02
290-1,000
No data
No data
No data
No data
0.7-2.8
No data
(One value)
33-270
No data
0-0.05
No data
* Expressed as ng of emission per kJ of combustor heat input (1 ng/kJ = 2.34 x 10"* Ib/MM Btu).
Given the complexity of sources of potential PIC com-
pounds, it is not surprising that a consensus PIC definition
has been difficult to achieve. Consequently, for the purpose
of this review, it seems more productive to examine the issue
of combustion byproducts separately from any type of spe-
cific definition by ignoring the source or cause of the emis-
sion of particular compounds and considering all organic
compound emissions (including POHCs) as combustion by-
products (CBs). A recent EPA study has examined CBs in
this fashion.112 The study examined field test data from 23
EPA-sponsored emissions tests at thermal destruction facil-
ities. Included were eight incinerators, nine industrial boil-
ers, and six industrial kilns. Organic emissions from hazard-
ous waste facilities were compared to emissions when these
facilities were burning fossil fuel only. The organic emissions
were also compared to organic emissions from municipal
solid waste incinerators and coal-fired utility boilers.
The EPA studies of thermal destruction systems identi-
fied 55 Appendix VIII compounds (28 volatile and 27 semi-
volatile) in stack emissions. These compounds were emitted
at normalized rates that span over five orders of magnitude,
0.09 to 13,000 nanograms of emissions per kilojoule (ng/kJ)
of combustor heat input (one ng/kJ = 2.34 X10"6 Ib/million
Btu).
The greatest number of compounds were emitted in the 10
to 100 ng/kJ range. Only nine of the 23 facilities emitted
identified hazardous compounds at rates exceeding 100 ng/
kJ.
The volatile compounds tended to be detected more often
and in significantly higher concentrations than the semivo-
574
and coal-fired power plants. Sirr.ilar data were not available
for volatile compounds. Table XIII presents this compari-
son. The four phthalate compounds in the table all show
very similar emission rates from all three sources. Naphtha-
lene emissions were lower for power plants than the other
two sources. Again, the data suggest that for these com-
pounds there is little inherent difference between combus-
tion sources.
Dtoxin and Furan Emissions
Without doubt, the greatest amount of scientific and pub-
lic attention has been given to one class of incinerator com-
bustion byproducts, the dioxins and furans. The terms
dioxin and furan refer to families of 75 related chemical
compounds known as polychlorinated dibenzo-p-dioxins
(PCDDs) and 135 related chemical compounds known as
polychlorinated dibenzofurans (PCDFs), respectively.
These compounds are not intentionally made for any pur-
pose; they are unavoidable byproducts created in the manu-
facture of other chemicals such as some pesticides, or as a
result of incomplete combustion of mixtures containing cer-
tain chlorinated organic compounds. Since the first pub-
lished report of PCDD and PCDF emissions from a munici-
pal incinerator by Olie et al.11* & large number of studies
have been carried out to examine this phenomenon, includ-
ing work by Buser et aL,115 Eiceman et ai.,116 Karasek,117
Bumb et oi.,118 Cavallaro et al.,ns and Lustenhouwer et
a/.,120
JAPCA
F-17
-------
Table XIII. Semivolatile compound emission rates (rom hazardous waste combustion, municipal
incinerators, and a coal power plant, ng/kJ.'
Hazardous waste
Naphthalene
Bis(2-ethvlheiyl)phthalate
Diethvlphthalate
Butylbenzyphthalate
Dibutvlphthaiate
Mean
17
4.6
1.2
3.7
0.3
Range
0.3-150
0-21
0.04-4.8
0.7-23
0-1.1
Municipal waste
Mean Range
71
4.6
0.5
3.9
0.4-400
0.4-12
0-0.9
No data
1.5-7.6
Coal power plant
Mean
0.5
7.6
2.8
0.5
3.0
Range
0.06-1.8
0.2-24
0.4-5.7
0.3-1.0
0.09-8.7
* Expressed as ng of emission per kJ of combustor heat input (1 ng/kJ = 2.34 X 10~* Ib/MM Btu).
Most of the interest has been placed on municipal waste
incineration. A number of excellent summaries of field emis-
sion data have been prepared.121"123 EPA has reviewed avail-
able PCDD/PCDF emissions data for a broad range of com-
bustion sources including fossil fuel and wood combustion
and a wide range of industrial furnaces,124 and has reported
the results of recent emissions testing at 11 additional facili-
ties.125
Dioxin/furan emissions data are somewhat less available
for hazardous waste incineration facilities. EPA tests have
examined dioxin/furan emissions at five incinerators,79 six
industrial boilers,126 and three calcining kilns employing
hazardous waste as a fuel.127-129 Data are also available from
test burns at PCB incinerators.130-131 Dioxin/furan emissions
from these data sources are summarized in Table XIV.
Of the 17 facilities only five emitted detectable levels of
PCDD or PCDF. None of the facilities tested was emitting
detectable levels of the most hazardous isomer, 2,3,7,8-
TCDD. The highest PCDD levels reported were for an in-
dustrial boiler using a creosote/PCP sludge as a fuel, where
PCDDs were present in the waste feed.126 In most cases, no
PCDD or PCDF was detected in hazardous waste incinera-
tion emissions for the facilities tested. By comparison,
PCDDs and PCDFs have been found in emissions from 22
municipal waste incinerators for which complete stack emis-
sions data are available.123 Average emissions of PCDD and
PCDF (3,300 ng/m3 and 2,700 ng/m3, respectively), were
nearly three orders of magnitude greater than the highest
values reported for hazardous waste incineration units.
Ash and Air Pollution Control R**ldu« Quality
Facilities which incinerate hazardous wastes containing
significant ash or halogen content will generate combustion
chamber bottom ash and various types of residues collected
by subsequent, air pollution control equipment. Under
RCRA, these ashes and residues are generally classified as
hazardous waste also. Thus, facility operators must assess
the characteristics of these materials to determine the prop-
er method of disposal. The principal contaminants of inter-
est are heavy metals and any undestroyed organic material.
Most, but not all, operating hazardous waste incinerators
which generate combustion chamber ash, quench the ash
(usually in water) before discharge. Since air pollution con-
trol equipment using wet collection methods predominates
incinerator practice, most additional ash and haloacid (e.g.,
HC1) is also collected in aqueous effluent from a scrubber,
absorber, or wet ESP.
Only limited characterization data are available for com-
bustion chamber ash and air pollution control residues.
Table XIV. Dioxin/furan emissions hazardous waste thermal destruction facilities (ing/m3).
Facility type Sample/(waste)« 2.3,7>8TCDD PCDD PCDF Reference
Commercial rotary kiln/
liquid injection
combustion incinerator
Fixed hearth incinerator
Liquid injection incinerator
Horizontal liquid injection
incinerator
Incinerator ship
4 lime/cement kilns
Fixed hearth incinerator
Rotary kiln/liquid injection
Industrial boiler
Industrial boiler
Industrial boiler
Industrial boiler
Industrial boiler
FG/FAb
(HW)
FG/FA
(HW)
FG/FA
(HW)
FG/FA
(HW)
FG/FA
(PCB)
FG
-------
Combustion chamber ash and scrubber waters were ana-
lyzed for several of the incinerators tested by EPA as part of
the incineration Regulatory Impact Analysis (RIA) pro-
gram.79 Recently, ten additional incinerators were sampled
to characterize ash and residues 132
In the RIA study,79 incinerator ash and scrubber waters
were analyzed for organic constituents. Only two facilities
had ash concentrations of organic compounds at levels great-
er than 35 /ig/g. When organic compounds were detected,
they tended to be toluene, phenol, or naphthalene at concen-
trations less than 10 jig/g. The same compounds were also
detected in scrubber waters, usually at concentrations below
Table XV. Parameters typically employed to trigger fail-safe
corrective action for incinerators.
Basis for corrective action
Excess Worker Equipment
Parameter emissions safety protection
The results of the recent ten-incinerator test program
generally confirmed the RIA results.132 While more organic
compounds were detected across all of the facilities (19 vola-
tile and 24 semivolatile compounds), levels in ash were typi-
cally at or well below 30 Mg/g- One facility had a toluene
concentration of 120 ^g/g and phenol concentration of 400
pg/g. These levels were believed to have resulted from the
facility's use of chemical manufacturing plant wastewater
for ash quenching.
More compounds were detected in scrubber waters across
the ten facilities than in the RIA study (nine volatiles and
five semivolatiles) and in higher concentrations. Semivola-
tiles ranged from 0 to 100 #g/L while volatile compounds
were much higher (0 to 32 mg/L).
Combustion chamber ash and scrubber waters were also
analyzed for metals in both the RIA study and the ten-
incinerator study. Detected concentrations varied widely
and were a function of the amount of metal in the input
waste stream at each facility and how the residues were
processed. However, only three of 104 measurements of
these metals in leachates from the residues and ashes ex-
ceeded allowable toxicity characteristic levels using the
EPA's extraction procedure test (EP). If a leachate from a
waste material exceeds these levels (CFR 40 Part 261.24),
the waste will be designated as a hazardous waste.
Overall, the data from both test programs have suggested
that very small amounts of residual organic compounds re-
main in incinerator ash and incinerator air pollution control
residues. Thus, the destruction and removal efficiencies re-
ported for incinerators are almost entirely a result of de-
struction, rather than removal, of organic compounds. Lev-
els of metals in ashes and air pollution control residues
varied widely but appear to generally not exhibit the RCRA
toxicity characteristic. On the other hand, it should be recog-
nized that available data represent short-term samples from
less than 10 percent of the total hazardous waste incinerator
population in the United States. Use of these data to project
residue and ash quality for specific waste/incinerator combi-
nations is not possible. Metal and organic concentrations are
highly waste and facility specific. They will likely be influ-
enced strongly by waste characteristics and by operating
conditions (e.g., scrubber water recycle rate, solids residue
time in the combustion chamber, and contaminants in
scrubber water and quench make-up water).
Predicting and Assuring Incinerator Performance
Existing data indicate that well-operated hazardous waste
incinerators and other thermal destruction facilities are ca-
pable of achieving high levels of organic hazardous material
destruction which equal or exceed current RCRA perfor-
mance standards. Putting aside arguments over the adequa-
cy of the ORE standard for protecting public health and the
environment, virtually any well-designed thermal destruc-
tion unit should be capable of demonstrating high DRE if
sufficient operating temperature, oxygen, and feed control
are provided. While convincing trial burn performance data
can be presented, uncertainty and distrust may exist regard-
ing the reliability of thermal destruction systems in day-to-
High CO in suck gas
Low chamber temperature
High combustion gas flow
Low pH of scrubber water
Low scrubber water flow
Low scrubber AP
Low sump levels
High chamber pressure
High chamber temperature
Excessive fan vibration
Low burner air pressure
Low burner fuel pressure
Burner flame loss
X
X
X
X
X
X
X
X
X
X
X X
X
X
X
X
X
X
day operation after a permit is approved and when regula-
tors are not present. Little is known quantitatively about the
impact of normal process upsets or failure modes upon emis-
sions. This is often a concern of the public in hearings on
permit actions.
Currently, permit conditions are primarily based upon
process operating conditions which are documented during
the conduct of a successful trial burn; i.e., one which demon-
strated that the facility achieved or exceeded the RCRA
performance standards. These operating conditions are then
also used to establish fail-safe controls for the facility, which
designate corrective action to be taken in the event process
operation deviates from the demonstrated set points. Cor-
rective actions could include changing auxiliary fuel addi-
tion rates, shutting off waste feed, increasing combustion air
flow, etc., to control emissions, or other actions to protect
worker safety and process equipment32 Table XV shows
typical shutdown parameters which may be used to trigger
fail-safe controls.
Operating conditions (e.g., combustion temperature, Oj,
and CO in stack emissions, etc.) must be used as surrogates
for continued high destruction performance after the trial
burn since there is currently no real-time method to deter-
mine DRE for specific POHCs. DRE can only be determined
with certainty via expensive ($50,000 to $150,000), often
multiday testing procedures. Analysis results may take
weeks or months to complete. While EPA believes the cur-
rent permit approach is reasonable and protective of public
health and the environment, many argue that the availabil-
ity of a real-time monitoring technique to detect process
upsets and alert operators to automatically take corrective
action would significantly increase public acceptance of
thermal destruction technology.
Two general classes of performance estimation techniques
exist. The first of these involves the use of compounds which
are either identified in the waste or added to it to serve as
"surrogates" for the destruction of other important com-
pounds in the waste. The second approach involves the use
of indicator emissions such as CO or unburned hydrocarbons
to mirror waste destruction efficiency. Both concept* are
used to some extent in incinerator permitting currently.
Surrogates
The surrogates concept involves identifying an easily de-
tected organic compound which is more difficult to thermal -
ly destroy than any of the other hazardous compounds in a
waste mixture. It is then assumed that if destruction effi-
ciency for this compound is known for a given facility then
all other compounds in the waste will be destroyed to at least
that degree. This concept therefore involves developing ar
incinerability ranking of compounds.
F-19
576
-------
The RCRA permit guidance for selecting POHCs in
•vastes actually employs this approach. However, the identi-
fication of compound incinerability has proven difficult and
possibly unreliable. EPA has suggested the use of compound
neat of combustion (AHc) as a ranking of compound inciner-
ability.1'- This ranking method has received considerable
criticism and alternative scales, which have also been criti-
cized, have been proposed. These ranking approaches have
been recently reviewed and compared by Bellinger.133 They
include: autoignition temperature,134 theoretical flame
mode, kinetics,135 experimental flame failure modes,136 igni-
tion delay time,137 and gas phase (non-flame) thermal stabil-
ity.138 The rankings of compounds by each of these indices
were compared to their observed incinerability in actual
waste incineration tests in 10 pilot- and field-scale units.133
Each index failed to predict field results except for the non-
name thermal stability method. This method, based on ex-
perimentally determined thermal stability for mixtures of
compounds under low oxygen concentration conditions,
showed a statistically significant correlation for the com-
pounds evaluated.
While the low oxygen thermal stability concept appears
promising, data are available for only 28 compounds. Corre-
lation for other important Appendix VIII compounds over a
range of compound concentrations will be necessary before
the method can be used reliably for POHC designation or as
a basis for establishing continuous monitoring systems for
specific surrogate compounds.
As a result of the uncertainty over incinerability rankings,
the use of "additives" is being considered for overcoming the
limitations of the single POHC compound approach. This
concept involves the addition of a single, well-characterized
compound or small group or "soup" of compounds to a waste
stream, with subsequent continuous monitoring of the emis-
sions of the compound(s) to serve as a measure of destruc-
tion performance. Compounds such as various freons139-140
and sulfur hexafloride SF6 have been proposed.141-143 Con-
ceptually, these types of materials would be ideal additives
since they rarely occur in hazardous wastes, can be detected
in emissions using on-line instruments and are not likely to
be formed as combustion byproducts.
Combustion byproducts formation has caused difficulty
in interpretation of incinerability data for mixtures of con-
ventional POHC candidate compounds.141-144-148 While lab-
oratory-scale studies have shown some promise, attempts in
correlating field incinerator performance with additives be-
havior results have been inconclusive to date. Additional
testing is needed.
Performance Indicator*
Carbon monoxide (CO) and total unburned hydrocarbons
(TUHC) are emitted from all combustion systems in varying
amounts. Because CO is the final combustion intermediate
prior to the formation of CO2 in the combustion process, it
has been used in the determination of combustion efficiency.
Unburned hydrocarbon emission values do not include all
incompletely combusted hydrocarbons. Rather this is an
instrumentation-derived value resulting from the passage of
gaseous emissions through a hydrogen flame ionization de-
tector (HFID), which is commonly used with gas chromato-
graphs. The HFID responds to the number of carbon-hydro-
gen and carbon-carbon bonds in residuals in the combustion
gas. Because it does not respond to oxidized products such as
O2, CO, C02, and H20, it has been used as an indicator of
residual fuel emissions.
Because CO is an indicator of the degree of completion of
combustion and TUHC may be reflective of the amount of
incompletely combusted material in the exhaust gas, these
measures have been considered as possible indicators of in-
cinerator performance. Continuous monitoring of CO is re-
quired by the RCRA incinerator standards for this reason.
TUHC measurements, however, are not required.
The use of CO and TUHC in hazardous waste incineration
has been studied by several groups136-141-146-149 and criticized
by others.145-148 Waterland obtained pilot-scale data which
indicated correlations of the fractional penetration of
POHCs (1-%DRE/100) with CO and THC.146 Kramlich et al.
and LaFond et a/.136-147 found that increases in CO preceded
increases in the penetration of POHCs in a laboratory-scale
turbulent flame reactor as parameters such as air/fuel ratio,
atomization and degree of thermal quenching were varied.
At the same time, TUHC tended to increase as POHC pene-
tration increased. In a test of a pilot-scale circulating-fluid-
ized bed combustor Chang et al.141 indicated that penetra-
tion of combustion byproducts appeared to be correlated
with TUHC and that there were no instances of high com-
bustion byproducts penetration without a corresponding in-
crease in CO. The converse was not true, i.e., increases in CO
were observed on some occasions without a corresponding
increase in combustion byproduct penetration. POHC de-
struction efficiency was high throughout this series of tests
and did not appear to correlate well with either TUHC or
CO. Daniels et al., although critical of the use of CO as a
surrogate for POHC DRE or as an indicator of incinerator
performance, presented data obtained from a full-scale rota-
ry kiln, which in five out of six cases indicated increased
POHC penetration with increased CO concentration.148
Analysis of the pooled data from the EPA incinerator test
program revealed that there was no absolute level of mean
combustion temperature, mean gas phase residence time or
carbon monoxide emission concentration which correlated
with achieving 99.99 percent DRE.79 Residence times ranged
from 0.1 to 6.5 seconds in the facilities tested. Temperatures
ranged from 648°C to 1450°C. Carbon monoxide (CO) levels
were as high as 600 ppm, but at most plants ranged from 5
ppm to 15 ppm. It was concluded that the relationships
between DRE and these parameters are, in all likelihood,
facility specific and that waste characteristics, waste atom-
ization, and combustion chamber mixing likely play equally
important roles in achieving high DRE. Timing, funding,
and facility constraints, however, did not allow for collection
of sufficient performance data under varying conditions at
each site tested to allow for such relationships to be quanti-
fied. In particular, few of the test conditions produced DRE
significantly below 99.99 percent.
Dellinger145 has suggested that one reason for difficulties
in correlating CO with DRE is that the assumed rapid oxida-
tion of hydrocarbons to CO may not be correct for complex
hazardous wastes containing large halogenated and hetero-
atoro molecules. For these wastes, formation of stable inter-
mediate organic reaction products may delay the production
of CO. This delay would tend to distribute or move the CO
production maximum relative to fuel (waste) destruction
efficiency and tend to negate the usefulness of CO measure-
ment in the region of 99.99 percent DRE. Hall et al. have
conducted laboratory studies of CO formation versus com-
pound destruction for several complex mixtures and found
no correlation.149
Predicting Performance
Based on current knowledge it would appear that no single
performance indicator or surrogate is sufficient as a predic-
tor of organic compound destruction in incinerators. While
low oxygen thermal stability data show promise as a predic-
tor of compound incinerability, the data base is still not
sufficient to extend this concept to POHC selection or to the
development of standard POHC soups for trial burns or
compliance monitoring. Data on additives are also insuffi-
cient to project a DRE correlation. CO may be useful in
setting an upper bound condition on compound penetration
(1-DRE), but there is no demonstrated "correlation" be-
tween CO emissions and DRE. Elevated TUHC emissions
May 1987 Volume 37, No. 5
F-20
577
-------
are indicative of an increase in incomplete combustion by-
product emission, but not necessarily a decline In DRE or
even an increase in hazardous combustion byproduct emis-
sion, in part, because the HFID is less sensitive to halocar-
bon compounds.
One of the additional limitations placed upon attempts to
correlate surrogates and indicators with DRE is the lack of a
significant data base on incinerator operation under failure
conditions. A failure condition can be defined as a normal or
accidental operational deviation which results in failure of
the facility to achieve a 99.99 percent DRE. Most of the field
incinerator data used to make TUHC and CO correlations
has been taken under steady-state operating conditions. The
impact, of .failure modes such as nozzle clogging and kiln
overcharging upon CO and TUHC emissions has not been
adequately quantified, largely because of limitations on test
time and funding and, in particular, permit constraints
which prohibit off-design operation of facilities. EPA is,
therefore, conducting failure-mode testing at its bench- and
pilot-scale research facilities in Jefferson, Arkansas, Cincin-
nati, Ohio, and Research Triangle Park, North Carolina.
Preliminary results from these tests have been reported re-
cently.150-151
Table XVI. Total excess lifetime cancer risk to the
maximum exposed individual for incinerator
emissions.11
,167
POHCs
PICs
Metals
Total
icr7 to io-'°
10-1 to 10~u
10-8 to 10-5
ID-8 to 10-6
EPA has also conducted non-steady-state operational as-
sessments at three boilers employing hazardous waste as a
fuel.152 The impact of typical non-steady-state operating
conditions (e.g., start-up, soot blowing, load change) upon
DRE, combustion byproduct emissions, CO, and TUHC was
studied. While elevated CO emissions were observed at two
of the sites under off-design operation, attempts to correlate
DRE with CO, NOI( and 02 emissions were unsuccessful,
largely because 99.99 percent DRE was achieved under both
good and off-design operation. The testing, however, ac-
knowledged some of the difficulties in conducting off-design
studies. In some cases, the duration of the process transient
to be studied may be shorter than the sampling time re-
quired to collect a sufficient sample to assess DRE. The large
volume and high surface area (boiler tubes) in boilers tended
to delay emissions of organics from one off-design test to the
next, making it difficult to separate cause and effect. This
so-called "hysteresis effect" may also cause difficulty to
some degree in interpreting the results of studies of transient
operation in incinerators.
More testing under non-steady-state is needed, particu-
larly for incinerators. While attempts to correlate perfor-
mance with single indicators and surrogates have been large-
ly unsuccessful to date, taken in some appropriate combina-
tion they may prove useful as real-time indicators of the
onset of process failure.
Environmental and Public Health Implications
Regardless of the apparent capabilities of hazardous waste
incinerators to meet or exceed the RCRA performance stan-
dards, the ultimate public test involves demonstration that
there is no unacceptable increase in public health risk from
the emissions to the environment. While any of the emis-
sions from an incinerator may potentially be of environmen-
tal interest, most attention has been directed toward air
pollution emissions. This is because they appear to represent
the most important source of off-site human exposure and
there is no opportunity for secondary containment or treat-
ment of emissions once they leave the stack. Ash and scrub-
ber residues, however, are lower in volume and can be con-
tained, examined, and if necessary, treated prior to dis-
charge or disposal. In addition to chronic exposure to
recurring emissions, there are also environmental and public
health impacts which could result from potential single
event or catastrophic emissions at incineration facilities.
Risks from Single Event Emissions
As with any industrial facility, there are risks from poten-
tial accidents at incineration facilities such as fires, explo-
sions, spills of raw waste and similar single-point events.
These events are probabilistic in nature and their evaluation
in a risk assessment is handled differently from continuous
pollutant emissions from stacks. For instance, the U.S. De-
partment of Transportation maintains statistics on the fre-
quency of releases of cargo from vehicular accidents involv-
ing trucks. For tank trucks of all types, for instance, this is '
estimated to be 0.35 releases per million miles traveled.47
Similar values may be identified for accidents involving stor-
age facilities and transfer operations.
Little specific information on these types of accidents is
available for hazardous waste incineration facilities.
Ingiwersen et al. evaluated the potential off-site impacts of
five hypothetical accidents at a planned hazardous waste
incineration facility.153 It was estimated that no long-term
adverse effects could be expected from chronic exposure for
the nearest residents (0.5 miles) and that any effects due to
acute exposures to the HC1 emissions from the accidents
were expected to be short-term and reversible. Actual acci-
dents at an operating European facility have been docu-
mented.154 Seven accidents occurred over an 11-year period.
One employee was injured and no off-site effects were re-
ported for any of the incidents, which generally involved
storage and handling operations. EPA has also recently ex-
amined transportation and spill related risks for ocean-
based incineration and found that these risks were greater
than risks from incineration.44 In the absence of accident
data specific for incineration facilities, statistics from relat-
ed industrial practice are probably adequate in assessing
these risks.
Methods lor Assessing Risk* from Recurring Emissions
The major concern of this discussion is the risk associated
with recurring air pollution emissions from incinerators. The
assessment of risk to human health rather than environmen-
tal damage is generally believed to be of greatest interest.
Four general steps are involved in assessing the impact <>n
public health from stack emissions from an incinerator
* Identify the health effects of constituents of concern as a
function of concentration level.
• Predict the concentrations of these constituents to » hich
the public may be exposed.
• Estimate the health impact of these concentration expo-
sures.
• Conduct an uncertainty analysis.
Identification of the constituents of concern in stack emis-
sions and the health effects of these constituents, is, of
course, a function of the waste streams and incineration
facility of interest. In general, any of the constituents on
Appendix VIII of the RCRA standards are of possible inter-
est. However, other organic compounds frequently found in
combustion emissions (certain polynuclear aromatics and
polycylclic aromatic compounds) may be of concern also.
The major health effects of concern are for low-level chronic
exposure to these materials. These effects are generally car-
578
F-21
-------
cinogenicity, mutagenicity, teratogenic or target organ toxic-
ity (e.g., sterility, behavioral effects).
Predicting the potential levels of human exposure to pol-
lutants requires information on the frequency, intensity,
duration, and continuity of exposure.155 Exposure assess-
ment generally requires the use of mathematical models
which simulate the transport and dispersion of emissions
from the stack to the exposed population. Of the air disper-
sion models available, EPA has most often used the Indus-
trial Source Complex Long-Term Model (ISC-LT) for pre-
dicting annual average concentration for hazardous waste
incinerator facility studies.156'157 The Oak Ridge National
Laboratory has linked the ISC-LT model with computerized
meteorology and population data bases and programs to
form the Inhalation Exposure Methodology (IBM).158-159
The IBM employs U.S. population data from the 1980 Cen-
sus and local meteorological data along with ISC-LT to esti-
mate air pollutant concentration^ an'4 human inhalation
exposures in the vicinity of hazardous waste incinerators
located anywhere in the United States. The IEM has been
used extensively by EPA in assessing regulatory alternatives
for hazardous waste incinerators.160'161
The exposure information generated by models such as
the IEM may then be employed to estimate human health
risk. The individuals at highest risk of developing adverse
health effects are of most interest. The risk to this "maxi-
mally exposed population" is estimated from the modeled
exposure at the point of highest annual average pollutant
ground-level concentration outside the facility. For each ex-
posed individual, cancer risk is expressed as the cumulative
risk over a 70-year (lifetime) period of continuous exposure.
A variety of estimators are available to quantify the health
risks of substances. Carcinogen potency factors have been
developed by EPA based on assumed no safe level.162 For
noncarcinogenic effects, no observable adverse effect levels
(NOAELs) have been used to derive reference dose (RfD)
levels.163
There is considerable uncertainty involved in conducting
risk assessment. Numerous assumptions must be made re-
garding pollutant emission leveb, pollutant effects, disper-
sion factors, etc. Only a fraction of the needed tests of the
effects of chronic, low-level exposures to environmental pol-
lutants have been done. There is also considerable uncer-
tainty in extrapolating effects from high doses which cause
effects in animals to low doses in humans. Linearity assump-
tions are typically used in making such extrapolations. Some
investigators have questioned the wisdom of such assump-
tions, however.164
Beyond this, very little is known about how, or even if, this
information can be used to estimate the effects of complex
mixtures of the substances usually present in incinerator
stack emissions. For these reasons and other limitations,
most assessments adopt assumptions and risk estimate val-
ues which produce an estimate of a worst case effect. In order
to promote consistency in risk assessments, EPA has recent-
ly published in the Federal Register a six-part guidance on
risk and exposure assessment methodologies.165 This guid-
ance is an excellent resource to those conducting or evaluat-
ing risk assessment studies.
Overall Risk* from Long-Term Air Pollution Emission* from
Hazardous Waste Incinerator*
Risk assessment and risk management have been used
increasingly by industry and government over the past 10
years in evaluating control technology and regulatory op-
tions for managing hazardous waste.166 The initial 1978
RCRA incineration standards, for instance, were almost en-
tirely design and performance oriented. In the 1981 propos-
al, however, EPA incorporated risk assessment into what
was called the best engineering judgement (BEJ) approach
to regulating and permitting incinerators.10 The operating
and performance standards for incinerators were to apply to
facilities unless a site-specific risk assessment indicated that
a higher degree of control was necessary. The risk assess-
ment proposal, however, was not included in the final rule in
1982, largely because of concern from the regulated commu-
nity over the uncertainty of risk assessment approaches.
Rather, risk assessment and cost-benefit analysis became a
more integral part of the development of hazardous waste
control technology standards through the conduct of Regu-
latory Impact Analyses (RIA) of all proposed standards as
required by Executive Order 12291.
A number of risk assessments have been conducted for
specific hazardous waste incinerators and for incineration on
a national basis. Using the results of emissions data from
nine full-scale incinerator tests,79 EPA conducted a risk as-
sessment as part of its Incinerator RIA in 1982. The objec-
tive was to examine the economic impact of the regulations
on the regulated community, and to estimate the health and
environmental effects of the regulations.167 The risks due to
principal organic hazardous constituents (POHC), combus-
tion byproducts (PICs) and metal emissions were developed
(Table XVI).
While these results show that the human health risks from
most incinerator emissions are low, risks from metal emis-
sions show the greatest potential for exceeding a 10~* cancer
risk. The risks from metals emissions ranged up to two to six
orders of magnitude higher than values for POHCs and
PICs, and dominated the total risk values. Risks from resid-
ual POHC and PIC from the incinerators tested were low
and were essentially equivalent.
Taylor et al. reported the results of a risk assessment for
metal emissions using the same test data,168 but employing
somewhat different assumptions. Using the IEM method-
ology, carcinogenic and noncarcinogenic risks were exam-
ined. Interestingly, these results showed even lower cancer
risks than the EPA study. Individual lifetime cancer risks
for the maximum exposed population ranged from a low of
4.48 X 10~u for beryllium to a high of 3.47 X 1Q-* for
chromium. Noncarcinogenic risks were also smalL All values
v/ere well below the respective ADI (acceptable daily intake)
values. Lead intake was highest, estimated at 2 percent of
theADL
Kelly reported similar conclusions for a risk assessment of
stack emissions from a hazardous waste incinerator in Bie-
besheim, West Germany.169 Maximum ground level air con-
centrations for 24 metals (and for PCB) were estimated
using the IEM. All levels (including PCB) were less than 2
percent of the corresponding continuous exposure limit
(CEL) value.
Holton et al. examined the significance of various expo-
sure pathways for air pollution emissions from three sizes of
land-based incinerators located at three hypothetical sites in
the United States.17(M72 For certain organic chemicals, the
food chain pathway may be an important contributor to
total human exposure. However, the study concluded that
the human health risk from emissions was small for all of the
chemicals studied irrespective of the exposure pathway.
Fugitive emissions from auxiliary facilities at incinerators
(e.g., storage tanks) were also estimated to be an important
contributor to total pollutant emissions.171"172 Few studies
have quantified fugitive emission levels at incinerators. The
studies which have been done, however, have not shown that
ambient levels are a cause for concern.79'173-174
The risks associated with incineration of hazardous wastes
at sea have been recently compared to risks from land-based
incineration.47 While risks of marine and terrestrial ecologi-
cal damage were estimated, the direct human health risks
from stack emissions are of greatest interest in this discus-
sion (Table XVII). The incremental cancer risk to the most
exposed individual was determined for POHCs, PICs, and
metals for two scenarios: a PCB waste and an ethylene di-
Mav 1987
Volume 37 No 5
F-22
579
-------
XVII. Incremental cancer risk to the most exposed
by type of stack release.47
Systems
PCB waste
EDC waste
(H-rmn based
POHCs
PIC.
Metals
Tout suck
Land- based (two sites)
J'OHCs
HICs
Metals
Total stack
1.45 X 10-'°
1.68 X 10-'2
6.37 X lO"7
6.37 X 10-'
5.13 X 10~8
1.79 X 10-«
2.65 X ID'5
2.74 X ID"5
5.51 X 10- 10
3.36 X 10'9
1.06 X 10-«
1.06 X 10-<
1.43 X ID'7
2.59 X 10-«
3.12 X 10-5
3.14 X 10"5
chloride (EDC) waste. Not surprisingly, the human health
risk of stack emissions from ocean incineration were less
than those of land-based systems, largely due to distance
from population. The land-based incinerator risk values
were similar to those estimated in the EPA incinerator RIA.
POHC and PIC releases showed low risk, generally one to
five orders of magnitude less than those for metals. Risks
from metals accounted for from 90 percent to almost all of
the identified risk from either system, and exceeded the 10~6
risk level for only the land-based scenario. The study notes,
however, that the assumptions used in the assessment over-
state the likely levels of carcinogenic metals in the hypo-
thetical wastes used in the assessment and, therefore, likely
overestimate emissions and risk level.
In another study, Holton175 compared the potential differ-
ences in human exposure to emissions from identical PCB
incinerators in ocean and land-based contexts. Land-based
incineration showed higher inhalation exposure. The only
human exposure pathway considered for at-sea incineration
was ingestion of contaminated fish and shellfish. Accidental
spills were not assessed. Inhalation exposure for land units
was two orders of magnitude higher, terrestrial food chain
ingestion exposure was a factor of 20 higher, and drinking
water ingestion exposure was estimated to be about the same
as that for consuming fish and shellfish. No estimate of
absolute risk was attached to any of the estimates.
All of these risk assessment studies point to a conclusion
that stack emissions from incineration of hazardous waste
pose little risk to human health. However, as previously
stated, the emissions data base upon which many of the
assessments were based has been criticized by the EPA Sci-
ence Advisory Board (SAB) as being insufficient.46 SAB has
recommended that a more complete assessment of the quan-
tity and physical/chemical character of incineration emis-
sions be done to provide a basis for a more complete risk
assessment than has been possible to date. The SAB points
to the fact that only a portion of the organic mass emissions
has been identified in past studies. Many believe that most
of this unidentified mass is non-chlorinated Ci-Cg hydrocar-
bons, which are of little concern from a risk standpoint.
However, test data are only just beginning to emerge to
confirm or disprove this belief. EPA has recently completed
a full-scale incinerator emissions test where as much of the
mass emissions as possible will be specifically identified.
The testing included steady-state and typical upset condi-
tions for a large rotary kiln incinerator. Results should be
available by the summer of 1987.176
It is not clear, however, that even this level of emissions
information will really answer the question of how much
absolute risk is associated with incinerator emissions. From
the standpoint of the lay public, it may be more useful and
productive to compare these emissions with other types of
combustion emissions whose risks we have accepted in daily
life. Lewtas and others177 have done interesting work on the
comparative cancer potency of complex mixtures of pollut-
ants (e.g., power plant emissions, automobile exhaust, ciga-
rette smoke). Using short-term bioassays of organics ex-
tracted from actual emissions, the relative cancer potency of
emissions has been estimated.
Comparative mutagenic emissions rates (expressed as re-
vertants per mile or joule) have been determined from test-
ing of mobile sources and stationary sources. Experimental
work to date suggests that the mutagenic emission rates of
wood stoves, for instance, are as much as four orders of
magnitude greater than those for conventional coal-fired
utility power plants.177 It is also apparent from the data base
that variations in organic emission rate affect the net carci-
nogenic emission rate more than variations in the carcino-
genic or mutagenic potency of the emissions. Similar sam-
pling and testing is needed for incineration emissions so that
their potency can be compared to everyday sources such as
wood stoves, oil furnaces, and utility power plants. It should
be noted, however, that use of comparative mutagenic emis-
sions rates alone does not account for variations in potential
human health impact that occur due to differences in expo-
sure level to emissions from sources of different types or as a
result of different routes of exposure.
Conclusions
The body of knowledge concerning hazardous waste incin-
eration has been expanding rapidly since 1980. This review
has examined some of the most significant aspects of this
information. A number of conclusions may be drawn on the
status of incineration technology, current practice, monitor-
ing methods, emissions and performance, and public health
risks. Beyond these, a number of remaining issues and re-
search needs can also be identified.
Based on this review, the following conclusions may be
drawn:
1. Incineration is a demonstrated, commercially available
technology for hazardous waste disposal. Considerable
design experience exists and design and operating guide-
lines are available on the engineering aspects of these
systems.
2. A variety of process technologies exist for the range of
hazardous wastes appropriate for thermal destruction.
The most common incinerator designs incorporate one
of four major combustion chamber designs: liquid injec-
tion, rotary kiln, fixed hearth or fluidized bed. The most
common air pollution control system involves combus-
tion gas quenching followed by a venturi scrubber (for
paniculate removal), a packed tower absorber (for acid
gas removal) and a mist eliminator. However, more than
half of "the existing incinerators employ no air pollution
control equipment at all.
3. Uncertainty exists as to the exact scope of current haz-
ardous waste incineration practice in the United States.
Best information is available for 1983, when between 1.7
and 2.7 million metric tons of hazardous waste is be-
lieved to have been incinerated in 208 incineration units
across the United States. As much as 47 million metric
tons of incinerable waste was generated in 1983.
4. Implementation of HSWA and SARA as well as indus-
trial concerns for limiting long-term environmental li-
ability will encourage increasing amounts of hazardous
waste to be directed to incineration facilities. While cur-
rent capacity appears adequate, a near-term short-fall in
commercial incineration capacity may develop, particu-
larly for facilities which can handle hazardous waste
sludges and solids.
5. The technology of stack sampling for trace organic com-
pounds is relatively sophisticated. Considerable experi-
ence and attention to quality assurance and quality con-
trol are needed. Documented sampling and analysis
methods are available for most of the parameters of
580
JAPCA
F-23
-------
interest in incineration performance assessment. Meth-
ods have been validated for a number of compounds.
With proper planning of test activities, detection limits
are not a limiting factor in assessing incinerator perfor-
mance.
6. Continuous emission monitors are available with ade-
quate operating ranges for many of the combustion
emissions of interest (CO, C02,02, TUHC, NO,). How-
ever, continuous monitors for specific organic com-
pounds are not available. No real-time monitor exists for
measuring destruction and removal efficiency.
7. Incinerators and most industrial processes employing
hazardous waste as a fuel can attain the RCRA destruc-
tion and removal efficiency requirement and the HC1
emission limit.
8. Certain incinerators have had difficulty achieving the
RCRA p»niculate matter emission limit of 180 mg/m3.
Data suggest that improved air pollution control tech-
nology or operating practices will enable these facilities
to be upgraded to meet the standard, however.
9. Insufficient data are available on the fate of heavy met-
als in incineration systems and the efficiency of typical
hazardous waste incinerator air pollution control equip-
ment to control emissions of specific metals and their
salts.
10. Insufficient data are available on the impact of typical
upset or off-design operating conditions on incinerator
and industrial furnace emissions.
11. Considerable uncertainty exists over the definition and
significance of incomplete combustion byproducts. An
insufficient data base exists on the full spectrum of po-
tentially hazardous compounds which may be in inciner-
ator emissions. Comparative emissions data for a limited
number of compounds suggest that incinerator emis-
sions are similar in character and emission rate to emis-
sions from fossil fuel combustion.
12. Based on current data, chlorinated dioxin and furan
emissions are not significant for hazardous waste incin-
erators. The most hazardous dioxin isomer (2,3,7,8-
TCDD) has not been detected in emissions at 17 facili-
ties tested for these compounds. Levels of all PCDD and
PCDF emissions from hazardous waste incinerators are
approximately three orders of magnitude less than those
reported for municipal waste incinerators.
13. Limited data on incinerator ash and air pollution control
residues suggest that organic compound levels are low
and that destruction is the primary reason for high de-
struction and removal efficiencies, not removal. Metal
concentrations in ash and residues vary widely, depend-
ing upon metal input rate to the incinerator and process
operation (e.g., scrubber water recycle and make-up
rates).
14. General process control systems and strategies exist to
control incinerator performance. However, none of the
available real-time monitoring performance indicators
appear to correlate with actual organic compound DRE.
No correlation between indicator emissions of CO or
TUHC and DRE has been demonstrated for field-scale
incinerator operations, although CO may be useful as an
estimator of a lower bound of acceptable DRE perfor-
mance. It may be that combinations of several potential
real-time indicators (CO, TUHC, surrogate compound
destruction) may be needed to more accurately predict
and assure incinerator DRE performance on a continu-
ous basis.
15. Available approaches for estimating compound inciner-
ability have not corelated with field experience. The best
approach appears to involve the use of experimentally
derived non-flame thermal stability data. However, an
insufficient data base is currently available to extend
this method for use in revised guidance on POHC selec-
tion.
16. There appears to be little increased human health risk
from hazardous waste incinerator emissions, based on
assessments done to date. Metal emissions appear to be
most significant in the risk values which have been de-
rived. However, a complete assessment of all of the po-
tentially hazardous materials in incinerator emissions
has not been completed. This information is needed to
enable a comprehensive risk assessment of incinerator
emissions.
17. In spite of the demonstrated destruction capabilities of
hazardous waste incinerators and the apparent low in-
cremental risk of emissions, there is considerable public
opposition to the siting and permitting of these facili-
ties. Permits require three years to finalize, on average.
Uncertainty over permitting and public acceptance will
likely result in a near-term short-fall in needed capacity,
particularly for commercial facilities which could incin-
erate solids and sludges.
Remaining Issues and Research Needs
While thermal destruction represents the most effective
and widely applicable control technology available today for
organic hazardous waste, a number of issues remain concern-
ing its use in the long term. These include:
* Destruction effectiveness on untested/unique wastes
* Detection of process failure
• Control of heavy metal emissions
• Emissions of combustion byproducts
• Real-time performance assurance
• The role of innovative technology
Destruction Effectiveness on Untested/Unique Wastes
All of the performance data which have been used in the
development and assessment of thermal destruction regula-
tions and standards to date have been collected for waste/
thermal technology combinations typical of current prac-
tice. However, the character of wastes which may be subject-
ed to incineration in the near future will begin to change,
perhaps dramatically. These changes will be influenced by
EPA action to restrict many wastes from land disposal under
the Hazardous and Solid Waste Act Amendments of 1984
and by increased emphasis upon remedial action at Super-
fund sites. Incineration will emerge as a feasible technologi-
cal alternative for destruction of many of these wastes and
site clean-up residues. However, EPA and industry will have
considerably less experience in handling these wastes.
Wastes will tend to have higher solids and water content, be
more complex in their physical and chemical composition.
have lower heating value, and/or potentially contain higher
levels of hazardous metals and high-hazard organics com-
pared to wastes which are typically incinerated today.
Consequently, while incineration is capable of achieving
high levels of destruction for today's wastes, future practice
may place new performance demands on current technology.
For instance, 99.9999 percent DRE is now required for
wastes bearing chlorinated dioxin and furan compounds.
Many of these wastes appear at Superfund sites. EPA is also
studying the need for heavy metals control regulations for
incinerators. That effort may suggest the need for improved
air pollution control systems or waste pretreatment.
Thus, performance testing of incinerators and other ther-
mal destruction devices must continue in order to assure
destruction and removal effectiveness for these untested
wastes, to assess process limitations and waste pretreatm*nt
requirements, to determine the safety of process residue.
V/
-------
to improve our ability to predict incinerator perfor-
mance on new waste materials.
Control of Heavy Metal Emissions
While the human health risk of incinerator emissions ap-
pears to be small, metal emissions have been the dominant
component of the risk levels identified thus far. Metal emis-
sions are controlled only indirectly by current standards
through the RCRA incinerator particulate emission limit.
The particulate standard, however, has proven difficult for a
number of operating facilities to achieve. In addition, the
metal content of wastes which may be subjected to incinera-
tion in the future may be higher as a result of the implemen-
tation of the HSWA land disposal restrictions and increased
ultimate clean-up actions at Superfund sites. Consequently,
while metal emissions may not pose a risk now, they may in
future practice.
Insufficient data exist on the physical and chemical char-
acter of particulate matter generated by hazardous waste
incineration systems. Few tests have examined the particle
size distribution of emissions or the specific metal removal
capability of the various air pollution control systems avail-
able. Likewise, insufficient data exist on the fate or parti-
tioning of these materials in incineration systems. This in-
formation is needed to examine the potential impact of met-
als in wastes upon net environmental emissions and to
evaluate various regulatory strategies which may be neces-
sary to control increased emissions. These strategies could
include metal input limits for waste or specific metal emis-
sion limits.
Emissions of Combustion Byproducts.
Current information suggests that organic combustion by-
product emissions identified for incineration of hazardous
waste do not represent a significant risk to public health.
Some, however, have questioned the completeness of emis-
sions data and, therefore, the adequacy of risk assessments
performed using these data. This issue has emerged as a
concern hi numerous public meetings on incinerator permits
and facility siting.
There is little doubt that none of the emissions testing
efforts conducted to date has identified all compounds in
incinerator stack emissions. The same is true, however, for
virtually any other source of air or water pollution. Because
hazardous waste facilities are perceived as being more haz-
ardous than many other types of pollution control or indus-
trial facilities, more attention is given to their emissions.
Thus, while it is unlikely that any major, highly hazardous
components of emissions have been overlooked, the data are
not available to prove this to all who may be concerned. On
the other hand, the task of finding all potentially hazardous
compounds is an open-ended one, ultimately limited by ex-
pense.
Another issue concerning combustion byproducts emis-
sions is that few tests have examined the level and chemical
character of emissions for periods of time facilities may be
operating under upset conditions (transients or failure
modes). More experimental work is clearly needed here.
Some testing has shown that there is little change in POHC
DRE over significant operating ranges or under "apparent"
failure conditions. At the same time, emissions of unburned
hydrocarbons have increased. Research is necessary to de-
termine if these "failure-mode" emissions pose a hazard.
One approach to resolving both the question of data com-
pleteness and failure mode impacts is to examine the relative
potency of emissions using short-term bioassays and to use
bioassay-directed chemical analysis as a means of more cost-
effectively identifying the chemical compounds (perhaps
previously unidentified) which are primarily causing the
potency. While short-term bioassays have their own set of
constraints and limitations, they have proven useful in com-
paring the cancer potency of mixtures of compounds 1 om
other combustion sources. Testing of a reasonable range of
hazardous waste types under "good" and off-design condi-
tions would give an indication of the range of potency of
emissions as a function of operational conditions and in
comparison with conventional combustion sources whose
risks and character are more familiar to the general public.
Real-TUne Performance Assurance
Once the public health significance of incinerator emis-
sions is verified, methods must be available to assure that
effective operation is maintained. A variety of surrogates
and indicators of incinerator performance are being evaluat-
ed. None is fully satisfactory and little evaluation has been
done under true failure conditions.
It may not be possible to find a set of easily monitored
parameters which "correlate" with incinerator performance.
However, it may be possible to identify parameters which
may be sufficient to identify the onset of process failure.
Research is necessary to examine the suitability of existing
real-time monitoring systems and approaches to reliably
predict process failure. Availability of such techniques may
have a significant impact upon public acceptance of these
facilities and form a technical basis for more effective com-
pliance monitoring by regulatory agencies.
Role of Innovative Technology
A wide range of innovative hazardous waste technology
has emerged since the passage of RCRA.178 A number of
these technologies are thermal destruction processes. The
potential destruction capabilities and cost-effectiveness of
these processes has been well publicized, although many of
the techniques must be considered to be only in the develop-
mental stage.
Many argue, particularly in public hearings, that decisions
on permitting conventional incineration facilities should be
postponed in favor of adopting more innovative approaches,
whose inventors often claim higher destruction efficiency at
lower cost than conventional systems. For specific waste
streams (e.g., contaminated soils, PCBs), a number of inno-
vative systems have demonstrated DREs equivalent to those
of conventional systems. Some systems appear to offer ad-
vantages for handling specific (although sometimes limited)
waste streams. Considerable uncertainty exists as to the true
cost-effectiveness of some systems, since practical field ex-
perience is often not yet available to aid in identifying oper-
ating limitations.
Many of these emerging systems will find a role in future
hazardous waste management strategies. Policy makers,
public officials, and industrial decision makers should be
careful, however, hi delaying action on currently available,
demonstrated, thermal destruction systems until the need,
benefit, and operability of such innovative systems are clear-
ly established.
References
Background
1 D C White "EPA program for treatment alternatives for
hazardous wastes," JAJPCA 34:369 (1985).
2 E T Oppelt, "Hazardous waste destruction," Environ. Sci.
' Technol. 20:312 (1986).
3. U.S. EPA, "Assessment of Incineration as a Treatment Meth-
od for Liquid Organic Hazardous Wastes—Background Re-
port V: Public Concerns Regarding Land-Based and Ocean-
Based Incineration," March 1985, U.S. GPO. 1985 .V26 7T8/
30376.
4. The Bible (Old Testament), Kings 23:10.
5 R Hering, S. A. Greelly, Collection and Disposal „/ Municipal
Re/use, McGraw Hill, New York. 1921.
582
F-25
JAPCA
-------
t R. C. Corey, Principles and Practices of Incineration, John
Wiley and Sons, New York, 1969.
7 F. L. Cross, H. E. Hcsketh, Controlled Air Incineration, Tech-
nomic Publishing Co., 1985.
8. U.S. EPA, "Hazardous Waste Proposed Guidelines and Regu-
lations and Proposal on Identification and Listing," Federal
Register 43(243, Part IV):58946-59028, December 18,1978.
9. U.S. EPA, "Hazardous Waste Management System: Identifi-
cation of Listing of Hazardous Waste," Federal Register 45(98,
Part m):33084-33137, May 19,1980.
10. U.S. EPA, "Incinerator Standards for Owners and Operators
of Hazardous Waste Management Facilities; Interim Final
Rule and Proposed Rule," Federal Register 46(15, Part
IV):7666-7690, January 23,1981.
11. U.S. EPA, "Standards Applicable to Owners and Operators of
Hazardous Waste Treatment Facilities," Federal Register,
47(122, Part V):27516-27535, June 24,1982.
12. U.S. EPA, "Guidance Manual for Hazardous Waste Incinera-
tor Permits," EPA-SW-966, NTIS No. PB 84-100577,1983.
13. U.S. EPA, "Hazardous Waste Management System; Burning
of Waste Fuel and Used Oil Fuel in Boilers and Industrial
Furnaces," Federal Register, 50 (No. 230) 49164, November
29,1985.
14. R. Holloway, U.S. EPA, Office of Solid Waste, personal com-
munication, January 1987.
15. U.S. EPA, "Polychlorinated Biphenysl (PCBs) Manufactur-
ing, Processing, Distribution in Commerce, and Use Prohibi-
tions," Federal Register, 44,31514, May 31,1979.
Currant Incineration Practice
16. U.S. EPA, "Dioxin Rule," Federal Register, January 14,1985.
17. G. Vogel, et at., "Composition of Hazardous Waste Streams
Currently Incinerated," Mitre Corp., U.S. EPA, April 1983.
18. S. Tejada, "The blue goose flies!" EPA Journal, October 1985.
19. Camp, Dresser and McKee, Inc., "Superfund Treatment Tech-
nologies: A Vendor Inventory," U.S. EPA, EPA 540/2-80/004,
July 1986.
20. U.S. EPA, "Hazardous Waste Generation and Commercial
Hazardous Waste Management Capacity," SW-894, December
1980.
21. G. Vogel, et al., "Incineration and cement kiln capacity for
hazardous waste treatment," in Proceedings of the 12th Annu-
al Research Symposium, Incineration and Treatment of Haz-
ardous Wastes, Cincinnati, Ohio, April 1986.
22. U.S. EPA, "A Profile of Existing Hazardous Waste Incinera-
tion Facilities and Manufacturers in the United States," NTIS
PB 84-157072, February 1984.
23. U.S. EPA, "Assessment of Incineration as a Treatment Meth-
od for Liquid Organic Hazardous Wastes—Background Re-
port III: Assessment of the Commercial Hazardous Waste In-
cineration Market," U.S. Government Printing Office, 1985-
526-778/30376, March 1985.
24. U.S. EPA, "National Survey of Hazardous Waste Generator
and Treatment, Storage and Disposal Facilities Regulated Un-
der RCRA in 1981," US. Government Printing Office, 055000-
00239-8,1984.
25. U.S. Congress, Congressional Budget Office, "Hazardous
Waste Management: Recent Changes and Policy Alternati-
ves," U.S. Government Printing Office, 1985.
26. U.S. Congress, Office of Technology Assessment, "Technol-
ogies and Management Strategies for Hazardous Waste Con-
trol," U.S. Government Printing Office, Washington, DC,
OTA-196, March 1983.
27. Environmental Resources Management, Inc., "New Jersey
Hazardous Waste Facilities Plan," New Jersey Waste Facili-
ties Siting Commission, Trenton, NJ, March 1985, p. 170.
28. ICF, Inc., "Survey of Selected Firms in the Commercial Haz-
ardous WAste Management Industry: 1984 Update," U.S.
EPA, September 1985.
29. Office of Appropriate Technology, "Alternatives to the Land
Disposal of Hazardous Wastes, An Assessment for California,"
Toxic Waste Assessment Group, Governor's Office of Appro-
priate Technology, Sacramento, CA, 1981.
30. U.S. Congress, Office of Technology Assessment, "Ocean In-
cineration: Its Role in Managing Hazardous Waste," U.S. Gov-
ernment Printing Office, Washington, DC, OTA-0-313, August
1986.
31. U.S. EPA, "1986 National Screening Survey of Hazardous
Waste Treatment, Storage, Disposal and Recycling Facilities,"
December 1986.
32. U.S. EPA, "Engineering Handbook on Hazardous Waste In-
' cineration," SW-889, NTIS PB 81-248163. September 1981.
33. W. R. Niessen, Combustion and Incineration Processes, Mar-
cel and Dekker, 1978.
34. C. A. Brunner. Incineration Systems and Selection and De-
sign. Van Nostrand Reinhold Co., 1984.
35. D. G. Ackerman. J. F. McGaughey. D. E. Wagoner. "At-Sea
Incineration of PCB-Containing Wastes On Board the M/T
Vulcanus," U.S. EPA, 600/7-83-024, April 1983.
36. Y. H. Kiang, "Total hazardous waste disposal through combus-
tion," Industrial Heating 14:9,1977.
37. Y. H. Kiang, A. A. Metry, Hazardous Waste Processing Tech-
nology, Ann Arbor Science, Ann Arbor, MI, 1982.
38. R. D. Reed, Furnace Operations, Third Edition, Gulf Publish-
ing, Houston, Texas, 1981.
39. W. S. Rickman, N. D. Holder, D. T. Young, "Circulating bed
incineration of hazardous wastes," Chemical Engineering Pro-
gress 81:34 (1986).
40. A. Trenholm, "Summary of Testing Program at Hazardous
Waste Incinerators," in Proceedings of the llth Annual Re-
search Symposium, Incineration and Treatment of Hazard-
ous Waste, EPA/600/-9-83/028, September 1985.
41. J. J. Santoleri, "Energy Recovery—A By-Product of Hazard-
ous Waste Incineration Systems," in Proceedings of the 15th
Mid-Atlantic Industrial Waste Conference on Toxic and
Hazardous Waste, 1983.
42. R. G. Novak, W. L. Troxler, T. H. Dshnke, "Recovering energy
from hazardous waste incineration," Chemical Engineering
Progress 91:146 (1984).
43. J. J. Moller, 0. B. Christiansen, "Dry Scrubbing of Hazardous
Waste Incinerator Flue Gas by Spray Dryer Adsorption," in
Proceedings of the 77th Annual APCA Meeting, 1984.
44. A. D. Little, Inc., "Overview of Ocean Incineration," U.S. Con-
gress, Office of Technology Assessment, May 1986.
45. D. G. Ackerman, Jr., R. A. Venezia, "Research on at-sea incin-
eration in the United States," Wastes Ocean 5: (1982).
46. U.S. EPA, Science Advisory Board, "Report on the Incinera-
tion of Liquid Hazardous Wastes," April 1985.
47. U.S. EPA, "Assessment of Incineration as a Treatment Meth-
od for Liquid Organic Hazaroud Waste Background Report
IV, Comparison of Risks from Land-Based and Ocean-Based
Incinerators," March 1985.
48. J. J. Yezzi et al., "Results of the Initial Trial Burn of the EPA-
ORD Mobile Incineration System," in Proceedings of the
ASMS 1984 National Waste Processing Conference, Orlando,
FL, June 1984.
49. G. N. Dietrich, G. F. Martini, "Today's technology offers on
site* incineration of hazardous waste," Hazardous Materials
and Waste Management January-February 1986.
50. N. N. Hatch, E. Hayes, "State-of-the-Art Remedial Action
Technology Used for the Sydney Mine Waste Disposal Site
Cleanup," in Proceedings of the 6th National Conference on
Management of Uncontrolled Hazardous Waste Sites, Wash-
ington, DC, November 1985.
51. J. Noland, W. E. Sisk, "Incineration of Explosives Contami-
nated Soils," in Proceedings of the 5th National Conference
on Management of Uncontrolled Hazardous Waste Sites,
Washington, DC, November 1984.
52. D. L. Vrable, D. R. Engler, "Transportable Circulating Bed
Combustor for the Incineration of Hazardous Waste," in Pro-
ceedings of the 6th National Conference on Management of
Uncontrolled Hazardous Waste Sites, Washington, DC, No-
vember 1985.
53. Westat, Inc., "Used or Waste Oil and Waste-Derived Fuel
Material Burner Questionnaire for 1983," U.S. EPA, 1984.
Measuring Process Performance
54. U.S. EPA, "Guidance Manual for Hazardous Waste Incinera-
tor Permits," March 1983.
55. P. Gorman, et al., "Practical Guide—Trial Burns for Hazard-
ous Waste Incineration," U.S. EPA, NTIS PB 86-190246/AS.
November 1985.
56. D. G. Ackerman, et al., "Guidelines for the Disposal of PCB's
and PCB Items by Thermal Destruction," U.S. EPA, May
1980.
57. J. C. Harris, D. J. Larsen, C. E. Rechsteiner, K. E. Thrun.
"Sampling and Analysis Methods for Hazardous Waste Com-
bustion," U.S. EPA, EPA-600/8-84-002, PB 84-155845. Febru-
ary 1984.
58. U.S. EPA, "CFR 40, Part 268, Appendix I Toxicity Charac-
teristic Leaching Procedures," Federal Register, Vol. 51, No.
216, November 7,1986, pg. 40643.
59. U.S. EPA, "Test Methods for Evaluating Solid Waste, Physi-
cal/Chemical Methods," Second Edition, SW-846, July 1982.
60. E. R. deVera, B. P Simmons, R. D. Stephens, D. L. Storm.
"Samplers and Sampling Procedures for Hazardous Waste
Streams," U.S. EPA, EPA-600/2-80-018, PB 80-135353. Janu-
ary 1980.
61. R. H. James, R. E. Adams, J. M. Finkel. H. C. Miller. L. D.
Johnson, "Evaluation of analytical methods for the determina-
tion of POHC in combustion products,"
-------
a H Bond, "At-sea incineration of hazardous waste " Enui
«"-i>ci. Technol. 18:5(1984).
f> A. Stern. B. M. Myatt, J. F. Lachowski, K. T. McGregor
specution of Halogen and Hydrogen Halide Compound! iii
'•aseous Emissions," in Proceedings of the 9th Annual Re-
"'"J symposium on Incineration and Treatment of Haz-
S' ' EPA-600/*-M-0«. PB 84 234525,
67
65 \&J£* %kenriedej- i W AdHM. K. E. Thrun, "Modified
Method 5 Tram and Source Assessment Sampling System
^sss^KyisJL& EPA- EpA-6oo/8-85-°03 p* 85.
6* <; A. Jungclaus.efai "Development of a Volatile Organic
sampling Tram (VOST)," in Proceedings of the 9th Annual
Research Symposium on Land Disposal, Incineration and
Treatment of Hazardous Waste, EPA-600/9-84-015 PB 84-
234525, July 1984.
K. M. Hansen, "Protocol for the Collection and Analysis of
Volatile POHC's Using VOST," U.S. EPA, EPA-600/8-84-007
PB 84-170042, March 1984.
L D. Johnson, R. G. Merrill, "Stack sampling for organic
emissions," Toxicol. Environ. Chem. 6: 109 (1983).
L. D. Johnson, "Detecting waste combustion emissions," Envi-
n,n. Sci. Technol. 20: 223 (1986).
.1 Prohoska,T. J.Logan, R. G. Fuerst, M. R. Midgett, "Valida-
tion of the Volatile Organic Sampling Train (VOST) Protocol;
Volume I, Laboratory Phase," U.S. EPA, EPA-600/4-86/014a,
PB 86-145547, January 1986.
J. Prohoska, T. J. Logan, R. G. Fuerst, and M. R. Midgett,
"Validation of the Volatile Organic Sampling Train (VOST)
Protocol; Volume II, Field Validation Phase," U.S. EPA, EPA-
600/4-86/014b, PB 86-145554, January 1986.
J. Bursey, M. Hartman. J. Homolya, R. McAllister, J.
McGaughey, D. Wagoner, "Laboratory and Field Evaluation
of the Semi-Vost Method," U.S. EPA. Vol. I, EPA-600/4-85-
075a, PB 86-123551/AS, and Vol. II, EPA-600/4-85-075b, PB
86-123569/AS. September 1985.
U.S. EPA. "Guidance on Trial Burn Reporting and Setting
Permit Conditions," draft report. November 1986.
J. Podlenski, E. Peduto, et al., "Feasibility Study for Adapting
Present Combustion Source Continuous Monitoring Systems
to Hazardous Waste Incinerators," U.S. EPA, EPA-600/8-84-
01 la. March 1984,
Emissions from Hazardous Wast* Incineration
75. U.S. EPA, "Final Standards for Hazardous Waste Incinera-
tion-Background Document," January 1981.
76. U.S. EPA, "Determination of Incinerator Operating Condi-
tions NVoessary for Safe Disposal of Pesticides," July 1975.
77. U.S. EPA. "Destroying Chemical Wastes in Commercial-Scale
Incinerators," 1977.
78. E. T. Oppelt, "Performance Assessment of Incinerators and
High Temperature Industrial Processes Disposing Hazardous
Waste in the U.S.," in Hazardous and Industrial Solid Waste
Testing and Disposal, ASTM Standard Publication 933,
(1986).
79. U.S. EPA. "Performance Evaluation of Full-Scale Hazardous
Waste Incineration," 5 volumes, NTIS, PB 85-129500, Novem-
ber 1984.
80. U.S. EPA, "Engineering Assessment Report—Hazardous
Waste Co-firing in Industrial Boilers," 2 volumes, June 1984.
81. U.S. EPA. "A Technical Overview of the Concept of Disposing
of Hazardous Wastes in Industrial Boilers." June 1984.
82. U.S. EPA, "Draft Summary Report on Hazardous Waste Com-
bustion in Calcining Kilns," March 1985.
83. U.S. EPA. "Evaluation of Hazardous Waste Incineration in
Aggregate Kilns," February 1985.
84. U.S. EPA. "Permit Writers Guide to Test Burn Data—Haz-
ardous Waste Incineration," EPA-625/6-86/012, September
1986.
85. A. Trenholm, P. G. Gorman, B. Smith, D. A. Oberacker, "Emis-
sion Test Results for a Hazardous Waste Incineration R1A,"
U.S. EPA, EPA-600/9-84/015.1984.
86. P. G. Gorman, A. Trenhom, D. A. Oberacker, B. Smith, "Par-
ticulate and Hydrogen Chloride Emissions from Hazardous
Waste Incinerator," U.S. EPA, EPA-600/9-84/015,1984.
87. R. E. Mournighan, J. A. Peters. M. R. Branscome, "Effects of
Burning Hazardous Wastes in Cement Kilns on Conventional
Pollutant Emissions," in Proceedings of the 78th APCA Annu-
al Meeting, VoL 5,85-86.1,1985.
88 R. A Olexsey "Air Emissions from Industrial Boilers Burning
Hazardous Waste Material," in Proceedings of the 77th APCA
g. Vol. 1,84-18.3.1984.
89. L. Edwards, et al., "Trace Metals and Stationary Conventional
Combustion Processes: Volume I, Technical Report," U.S.
EPA, EPA-600/7-80-155a, 1980.
90. J. Kaakinen, et al., "Trace elements behavior in coal-fired
power plants," Environ. Sci. Technol. 9:862 (1975).
91. D. Klein, et al., "Pathways of thirty-seven trace elements
through coal-fired power plant," Environ. Sci. Technol. 9:973
(1975).
92. R. D. Smith, "The trace element chemistry of coal during
combustion and the emissions from coal-fired plants." Pro-
gress in Energy and Combustion Science 6:53 (1980).
93. P. G. Gorman, K. P. Ananth, "Trial Burn Protocol Verification
at a Hazardous Waste Incinerator," U.S. EPA, EPA-600/2-84/
048. February 1984.
94. J. F. Clausen, R. J. Johnson, C. A. Zee, "Destroying Chemical
Waste in Commercial-Scale Incinerators. Facility Report No.
1—The Marquardt Company," U.S. EPA, PB 265-541, April
1977.
95. D. G. Ackerman, J. F. Clausen, R. J. Johnson, C. A. Zee, "De-
stroying Chemical Waste in Commercial-Scale Incinerators.
Facility Report No. 3—Systems Technology," U.S. EPA, April
1977.
96. D. G. Ackerman, J. F. Clausen, et al., "Destroying Chemical
Wastes in Commercial Incinerators. Facility Report No. 6—
Rollins Environmental Services, Inc." U.S. EPA, 1977.
97. D. D. Wallace, A. Trenholm, D. D. Lane, "Assessment of Metal
Emissions from Hazardous Waste Incinerators," in Proceed-
ings, 78th APCA Annual Meeting, Detroit, MI, June 1985.
98. R. L. Davison, D. Natusch, J. Wallace, "Trace elements in fly
ash, dependence on particle size," Environ. Sci. Technol. 8:
(1974).
99. R. Greenberg, W. Zoller, G. Gordon, "Composition and size
distributions of particles released in refuse incineration," En-
viron. Sci. Technol. 12: 566 (1978).
100. R. Bennett, K. Knapp, "Characterization of paniculate emis-
sions from municipal wastewater sludge incinerators," Envi-
ron. Sci. Technol. 16:831 (1982).
101. M. Turgeon, U.S. EPA Office of Solid Waste—personal com-
munication. March 1987.
102. R. E. Hall, M. W. Cooke, R. L. Barbour, "Comparison of air
pollutant emissions from vaporizing and air atomizing waste
oil heaters," JAPCA 33:683 (1983).
103. P. Fennely, M. McCabe, J. Hall, "Environmental characteriza-
tion of waste oil combustion in small boilers," Hazardous
Waste 1:489 (1984).
104. E. T. Oppelt. "Remaining Issues Over Hazardous Waste Ther-
mal Destruction," presented at the 79th APCA Annual Meet-
ing, Minneapolis, MN, June 24,1986.
105. D. S. Duval, W. A. Ruby, "Laboratory Evaluation of High-
Temperature Destruction of Kepone and Related Pesticides,"
U.S. EPA, EPA-600/2-76-299, December 1976.
106. D. S. Duval, W. A. Ruby, "Laboratory Evaluation of High-
Temperature Destruction of Polychlorinated Biphenyls and
Related Compounds," U.S. EPA, EPA 600/2-77-228, Decem-
ber 1977.
107. B. Bellinger, et al., "Determination of the Thermal Decompo-
sition Properties of 20 Selected Hazardous Organic Com-
pounds." U.S. EPA, EPA-600/2-84-138, August 1984.
108. S. Senken, et al., "Combustion/Incineration Characteristics of
Chlorinated Hydrocarbons," AIChE Annual Meeting, New Or-
leans. LA, November 8-12,1981.
109. B. Dellinger, et al., "PIC Formation Under Pyrolytic and
Starved Air Conditions," U.S. EPA, October 1985.
110. J. C. Kramlich, et al., "Laboratory-Scale Flame Mode Hazard-
ous Waste Thermal Destruction Research," in Proceedings of
the 9th Annual Research Symposium on Land Disposal, In-
cineration and Treatment of Hazardous Waste, EPA-600/9-
84/015, July 1984.
111. A. Trenholm, R. Hathaway, "Products of Incomplete Combus-
tion from Hazardous Waste Incinerators," in Proceedings of
the 10th Annual Hazardous Waste Research Symposium, In-
cineration and Treatment of Hazardous Waste, U.S. EPA,
EPA-600/9-84-022, September 1984.
112. Midwest Research Institute, "Products of Incomplete Com-
bustion from Hazardous Waste Combustion," U.S. EPA, draft
final report, June 1986.
113. R. A. Olezsey, et al., "Emission and Control of By-Products
from Hazardous Waste Combustion Processes," in Proceed-
ings of the Ilth Annual Hazardous Waste Research Sympo-
sium, Incineration and Hazardous Waste, U.S. EPA, EPA-
600/9-85-028, September 1985.
114. K. Olie, P. L. Venneulen, 0. Hutzinger, "Chlorodibenzo-p-
diozins and chlorodibenzofurans are trace components of fly
ash and flue gas of some municipal incinerators in the Nether-
lands," Chemosphere 6:455 (1977).
JAPCA
F-27
-------
115. H. R. Buser, C. Rappe, "Identification of substitution patterns
in polychlorinated dibenzo-p-diozins (PCDD's) by mass spec-
trometry," Chemosphere 7:199 (1978).
116. G. A. Eiceman, R. E. Clement, F. W. Karasek, "Analysis of fly
ash from municipal incinerators for trace organic compounds,"
Anal. Chem. 51:2343 (1979).
117. F. W. Karasek, "Diozins from garbage: previously unknown
source of toxic compounds is being uncovered using advanced
analytical instrumentation," Canadian Research, September
1980.
118. R. R. Bumb, et al., "Trace chemistries of fire: a source of
chlorinated diozins," Science 210: (1980).
119. A. Cavallaro, G. Bandi, G. Invernizzi, L. Luciani, E. Mongini,
A. Gorni, "Sampling, occurrence and evaluation of PCDD's
and PCDF's from incinerated solid urban waste," Chemo-
spfcere9:611 (1980).
120. J. W. Lustenhouwer, K. Olie, 0. Hutzinger, "Chlorinated di-
benzo-p-diozins and related coi.ipoui.os in incinerator ef-
fluents: a review of measurements and mechanisms of forma-
tion," Chemosphere 9:501 (1980).
121. A. D. Little, Inc., "Study on the State-of-the-Art of Dioxin
from Combustion Sources," American Society of Mechanical
Engineers, 1981.
122. Camp, Dresser and McKee, Inc., "Dioiin Emissions from Re-
source Recovery Facilities and Summary of Health Effects,"
U.S. EPA, November 1986.
123. Ontario Ministry of the Environment, "Scientific Criteria Doc-
ument for Standard Development No. 4-84, Polychlorinated
Dibenzo-p-diozins (PCDDs) and Polychlorinated Dibenzo-p-
furans," September 1985.
124. Radian Corporation, "National Dioxin Study, Tier 4-Combus-
tion Sources—Final Literature Review," U.S. EPA, Septem-
ber 1985.
125. A. Miles, et al., "Draft Engineering Analysts Report—National
Dioxin Study, Tier 4: Combustion Sources," U.S. EPA, No-
vember 1986.
126. C. Castaldini, "Dioxin Emissions from Industrial Boilers
Burning Hazardous Materials," U.S. EPA, April 1985.
127. M. Branscome, et al., "Evaluation of Waste Combustion in a
Wet Process Cement Kiln at General Portland, Inc., Paulding,
Ohio," U.S. EPA, February 1985.
128. M. Branscome, et al., "Evaluation of Waste Combustion in a
Dry Process Cement Kiln at Lone Star Industries, Ogiesby,
Illinois," U.S. EPA, December 1984.
129. D. R. Day, L. A. Cox, R. E. Mournighan, "Evaluation of Haz-
ardous Waste Incineration in a Lime Kiln: Rockwell Lime
Company," U.S. EPA, EPA-600/S2-84-132, November 1984.
130. L. A. Stretz, et al., "Controlled Air Incineration of PCP-Treat-
ed Wood," U.S. EPA, September 1982.
131. U.S. EPA, "PCB Disposal by Thermal Destruction," NTIS,
PB 82-241860, June 1981.
132. D. Van Buren, G. Ppe, C. Castaldini, "Characterization of
Hazardous Waste Incineration Residuals," U.S. EPA, January
1987.
Predicting and Assuring Incinerator Performance
133. B. Bellinger, et at., "Examination of Fundamental Incinerabil-
ity Indices for Hazardous Waste Destruction," in Proceedings
of the llth Annual EPA Research Symposium on Land Dis-
posal, Remedial Action, Incineration and Treatment of Haz-
ardous Waste, EPA/600/9-85/028, September 1985.
134. J. J. Cudahy, W. L. Troxler, "Autoignition temperature as an
indicator of thermal oxidation stability," J. Hazardous Mate-
rials 8: (1983).
135. W. Tsang, W. Shaub, "Chemical processes in the incineration
of hazardous materials," in Detoxification of Hazardous
Waste, Chapter 2, Ann Arbor Science, Ann Arbor, MI, 1982, p.
41.
136. J. C.Krarnlich, M. P. Heap, W. R. Seeker, G. S. Samuelsen,
"Flame Mode Destruction of Hazardous Waste Compounds,"
in Proceedings of the International Symposium on Combus-
tion, 1985.
137. D. L. Miller, et al., "Incinerability Characteristics of Selected
•Chlorinated Hydrocarbons," in Proceedings of the 9th Annual
Research Symposium on Solid and Hazardous Waste Dispos-
al. EPA 600/9-84/015, September 1983.
138. B. Dellinger. et al., "Determination of the thermal decomposi-
tion properties of 20 selected hazardous organic compounds."
Hazardous Waste I: 137 (1984).
139. Chemical and Engineering News, American Chemical Society,
Washington. D.C., p. 24, January 30, 1984.
140. J. L. Graham, D. L. Hall, B. Dellinger, "Laboratory investiga-
tion of thermal degradation of a mixture of hazardous organic
compounds," Environ. Sci. Technol. 2fc 703 (1986).
141. D. P. Y. Chang, et al., "Evaluation of a Pilot-Scale Circulating
Bed Combustor as a Potential Hazardous Waste Incinerator,"
presented at the 78th Annual Air Pollution Control Associa-
tion Meeting, Detroit, MI, June 1985.
142. M. L. Berger, C. Proctor n, "Sulfur Hexafluoride as a Surro-
gate for Verification of Destruction Removal Efficiency," pre-
sented at the Conference on the Chemical/Physical Processes
of Combustion, 1985.
143. M. L. Berger, C. L. Proctor n, "A Tracer Gas Ratio Technique
to Detenntne.Boiler Suitability for Hazardous Waste Incinera-
tion," presented at the Conference on the Chemical/Physical
Processes of Combustion, 1985.
144. D. C. Wolbach, "Parametric Experimentation with a Pilot-
Scale Boiler Burning Hazardous Compounds," in Proceedings
of the llth Annual Research Symposium on Land Disposal,
Remedial Action, Incineration and Treatment of Hazardous
Waste, EPA 600/9-84/015, May 1985.
145. B. Dellinger, D. Hall, "The viability of using surrogate com-
pounds for monitoring the effectiveness of incineration sys-
tems," JAPCA 36:179 (1986).
146. L. R. Waterland, "Pilot-Scale Investigation of Surrogate
Means of Determining-POHC Destruction," presented at the
77th Annual Meeting of the Air Pollution Control Association,
San Francisco, CA, June 1984.
147. R. K. Lafond, et al., "Evaluation of continuous performance
monitoring techniques for hazardous waste incinerators,"
JAPCA 35:658 (1985).
148. S. L. Daniels, et al., "'Experience in Continuous Monitoring of
a Large-Rotary Kiln Incinerator for CO, COj, and Oj," pre-
sented at the 78th Annual Meeting of the Air Pollution Control
Association, Detroit, MI, June 1985.
149. D. Hall, et al., "Thermal decomposition of a twelve-component
organic mixture," Hazardous Waste and Hazardous Materi-
als, (1985).
150. L. J. Staley, "Carbon Monoxide and DRE: How Well Do They
Correlate" in Proceedings of the llth Annual Research Sym-
posium-Incineration and Treatment of Hazardous Waste,
EPA/600/9-85/028, September 1985.
151. W. P. Linak, et al., "On the Occurrence of Transient Puffs in a
Rotary Kiln Incinerator Simulator," JAPCA 37:54 (1987).
152. H. Lips, C. Castaldini, "Engineering Assessment Report: Haz-
ardous Waste Cofiring in Industrial Boilers Under Non-
Steady-State Operating Conditions, U.S. EPA, August 1986.
Environmental and Public Health Implications
153. J. C. Ingiwersen, et al., "Analysis of the Off-Site Impacts of
Hypothetical Accidents at a Hazardous Waste Incineration
Facility," in Proceedings, 3rd International Symposium on
Operating European Hazardous Waste Management Facili-
ties, Odense, Denmark, September 1986.
154. J. Toffer-Clausen, "Safety Programs," in Proceedings, 3rd In-
ternational Symposium on Operating European Hazardous
Waste Management Facilities, Odense, Denmark, September
1986.
155. K. E. Kelly, "Methodologies for assessing the health risks of
hazardous waste incinerator stack emissions to surrounding
populations". Hazardous Waste 1:507 (1984).
156. U.S. EPA, "Health Risk Assessment Methodologies for RCRA
Regulatory Analysis," draft report November 1983.
157. J. F. Bowers; et-ol,"IndustrialSourace Complex (ISC) Disper-
sion Model User's Guide," U.S_. EPA, EPA-450/4-79-030. De-
cember 1979. "
158. F R. O'Donnell, G. A. Holton, "Automated Methodology for
Assessing Inhalation Exposure to Hazardous Waste Incinera-
tor Emissions," U.S. EPA, EPA-600/9-84/015,1984.
159. F. R. O'Donnell, et al., "Users Guide for the Automated Inha
lation Exposure Methodology (IBM)," U.S. EPA, EPA-600/2
83-29. June 1983.
160. G. A. Holton, C. C. Travis, E. L. Etnier, F. R. O'Donnell.
"Inhalation Pathway Risk Assessment of Hazardous Waste
Incineration Facilities," presented at the 77th APCA Annual
Meeting, Vol. 6,84-102.3,1984.
161. Peer Consultants, "Supporting Documentation for the RCRA
Incinerator Regulations 40 CFR 264, Subpart 0—incmera
tors." U.S. EPA, October 1984.
162. U.S. EPA "Health Assessment Document for Polychlorinated
Dibenzo-p-Dioxins." EPA/600/8-84.914F. p. 11-128. Septem
her 1985.
163. U.S. EPA. "Integrated Risk Information System (IRlSi Ap
pendix A: Reference Dose (RfD): Description and I •* in
10P7
17 No 5
F-28
585
-------
4 RentrRi!!kA^s^ne?ts'"J>g-A1-15-1986-
^feAS^SlI)068 " *• *toOSpheric «*
life™2 Re^ter'VoL 51-No-185' P- 33992-34504, September
167. U.S. pA "Draft Regulatory Analysis for Proposed RCRA
Regulations on Burning of Hazardous Wastes". January 1984
168. C. C. Taylor, et al, "Health Risk Assessment tor MetalsEnuV
sions from Hazardous Waste Incinerators," presented at 79th
APCA Annual Meeting, Minneapolis, MN, June 1986.
169. K..E. Kelly, "Comparison of Metal Emissions Data from Haz-
ardous Waste Incineration Facilities," presented at 78th
APCA Annual Meeting, Detroit, MI, June 1985,
170. G. A. Holton, et at., "Inhalation Pathway Risk Assessment of
Hazardous Waste Incineration Facilities," in Proceedings,
77th APCA Annual Meeting, San Francisco, CA, June 1984.
171. G. A. Holton, C. C. Travis, E. L. Etnier, F. R. O'Donnell, D. M.
Hetrick, E. Dixon, "Multiple-Pathways Screening-Level As-
sessment of a Hazardous Waste Incineration Facility," Oak
Ridge National Laboratory, ORNL/TM-8652, DE85000156.
172. C. C. Travis, E. L. Etnier, G. A. Holton, F. R. O'Donnell, D. M.
Hetrick, E. Dixon, E. S. Harrington, "Inhalation Pathway
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
EPA-600/R-92-124
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Air Emissions from the Treatment of Soils
Contaminated with Petroleum Fuels and Other
Substances
5, REPORT DATE
July 1992
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Bart Eklund, Patrick Thompson, Adrienne Inglis, and
Whitney Dulaney
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Radian Corporation
P. O. Box 201088
Austin, Texas 78720-1088
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-D1-0117, Task 31
68-DO-0125, Task 25
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; 1/91 - 5/92
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES AEERL project officer is Susan A. Thorneloe, Mail Drop 63, 919 /
c/M-oTnn
541-2709.
16. ABSTRACT The report summarizes available information on air emissions from the
treatment of soils contaminated with fuels. Soils contaminated by leaks or spills of
fuel products, such as gasoline or jet fuel, are a nationwide concern. Air emissions
during remediation are a potential problem due to the volatile nature of many of the
fuel components and the remediation processes themselves, which may promote or
result in contaminant transfer to the vapor phase. Limited information is also inclu-
ded on air emissions from the treatment of soils contaminated with hazardous wastes
The report will allow staff from state and local regulatory agencies, as well as staff
from EPA regional offices, to assess the options for cleaning up soil contaminated
with fuels. Seven general remediation approaches are addressed. For each, infor-
mation is presented about the remediation process, the typical air emission species
of concern and their release points, and the available air emissions data. Control
technologies for each approach are identified and their reported efficiencies are sum1
marized. Cost data are given for each approach and f or its associated control tech-
nologies. Emission factors and other emission estimation procedures for each re-
mediation approach are presented along with a brief case study.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution Organic Compounds
Petroleum Products
Emission Volatility
Soils
Contamination
Wastes
Pollution Control
Stationary Sources
Contaminated Soils
13B 07 C
11G
14G 20M
08G.08M
8. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
267
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
F-30
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Center for Environmental Research Information
Cincinnati, Ohio 45268
OFFICIAL BUSINESS
PENALTY FOR PRIVATE USE S3OO
AN EQUAL OPPORTUNITY EMPLOYER
If your address is incorrect, please change on the above label
tear oft: and return to the above address.
If you do not desire to continue receiving these technical
reports, CHECK HIRE D. tear off label, and return it to the
above address,
Publication No. EPA-eoo/R-92-i24
------- |