EPA/600/R-23/062
June 2023
vvERA ^[j'Q (UJ//])CfWClf©/r IS$dJJ<&
Groundwater Issue Paper
In Situ Thermal Remediation
Eva L. Davis
Contents
Section 1.0 Introduction 2
Section 2.0 In Situ Thermal Technology Descriptions
and Their Implementation 3
Section 2.1 Steam Enhanced Extraction 3
Section 2.2 Electrical Resistance Heating 5
Section 2.3. Thermal Conductive Heating 7
Section 2.4. Self-Sustaining Technology for Active
Remediation (STAR) 9
Section 2.5. Other Thermal Technologies 10
Section 2.6. Which technology is Appropriate for
a Given Site? 11
Section 2.7. Above Ground Treatment of
Extracted Contaminants 12
Section 2.8. Monitoring of Thermal Remediation
Systems 13
Section 2.8.1. Temperature Monitoring 13
Section 2.8.2. Pneumatic and Hydraulic Control
Monitoring 14
Section 2.8.3. Contaminant Mass Recovery
Rate 15
Section 3.0 Mechanisms for Increased Contaminant
Recovery during Thermal Remediation 16
Section 3.1. Increased Vaporization and Co-
boiling of NAPL and Groundwater 16
Section 3.2. Decreased viscosity of groundwater
and NAPL 18
Section 3.3. Other Mechanisms 19
Section 4.0 Defining the Area to be Treated by
Thermal Remediation 20
Section 4.1. Characterization Techniques to
Determine NAPL Presence 21
Section 5.0 Treatability Studies 24
Section 5.1. Laboratory Testing 25
Section 5.2. Field Testing 26
Section 6.0 Thermal Remediation Services and the
Superfund Process 28
Section 6.1. Remedial Goals 29
Section 6.2. What to Look for in a Remedial
Action Work Plan/Remedial Design 30
Section 6.3. How Does Thermal Remediation
Affect the Community? 31
Section 7.0 How Do We Know When We are Done?
33
Section 7.1. Should We Specify Soil or
Groundwater Cleanup Criteria? 33
Section 7.2. What Happens to Groundwater
Concentrations After Heating is Terminated?....34
Section 7.3. Can Thermal Remediation Lead to
Site Closure? 34
Section 8.0 Factors Affecting Costs 36
Section 8.1. General Cost Information for Thermal
Remediation 36
Section 9.0 Life Cycle Analysis of Thermal
Remediation Technologies 37
Section 10.0 References 38
SEE Case Study 43
ERH Case Study 44
TCH Case Study 45
-------
Section 1.0 Introduction
In situ thermal remediation technologies rely on the
addition of energy to the subsurface to change the
phase distribution and other physical properties of
volatile and semivolatile organic contaminants to
mobilize them and aid in their recovery. The most
commonly used in situ thermal remediation
technologies today are Steam Enhanced Extraction
(SEE), electrical resistance heating (ERH), and
thermal conductive heating (TCH; sometimes called
in situ thermal desorption, ISTD). These three most
commonly used technologies are applicable to and
have been proven for a wide variety of organic
contaminants and in a wide variety of hydrogeologic
settings, both above and below the water table. In
addition, there is an emerging thermal technology
called Self-sustaining Technology for Active
Remediation (STAR), that results in the in situ
combustion of liquid contaminants that have
significant fuel value in porous media. Each of
these technologies will be summarized.
Thermal remediation technologies are very
aggressive and are most applicable for
contaminated sites (or portions of sites) where
contaminant concentrations are the greatest,
generally areas where nonaqueous phase liquids
(NAPLs) are present. The NAPLs can be either
lighter than water (LNAPLs, such as petroleum
hydrocarbons) or denser than water (DNAPLs, such
as chlorinated solvents or coal tar-based products).
LNAPLs can be volatile organic compounds (VOCs)
such as gasoline, semivolatile organic compounds
(SVOCs) such as diesel fuel, or longer chain
hydrocarbons that are less volatile such as oil and
grease compounds. Chlorinated solvents are
generally VOCs, while coal tar and its derivative
creosote are SVOCs. NAPLs, and in particular
DNAPLs below the water table in heterogeneous
hydrogeologic settings which include low
permeability soils or fractured bedrock can be some
of the most difficult organic contaminants to
remediate. At the same time, these scenarios can
create significant risk of exposure to contaminated
groundwater, elevated contaminant concentrations
in indoor air, or discharge of contaminants to
surface waters, thus increasing the need for
effective remediation. Thermal remediation
technologies have the advantages of being fast,
with most VOC remediations requiring six months
or less of operation. Large percentages of the
contaminant mass can be recovered, leaving behind
only dissolved and/or adsorbed phases. Using
temperature measurements to ensure treatment of
all the target area greatly increases the certainty of
the remediation.
The purpose of this paper is to briefly describe
these commonly used in situ thermal remediation
technologies, and how they are deployed to
remediate VOC and SVOC contaminated sites. The
effects of temperature on the physical properties of
common organic contaminants that result in the
effectiveness of these technologies for the recovery
or destruction of contaminants are also briefly
described. A case study is provided for each of the
thermal technologies discussed. In addition,
guidance is provided on evaluating, contracting, and
implementing thermal technologies within the
Superfund process of evaluating alternative
technologies, including preparing bid documents for
thermal remediation services, reviewing designs,
and monitoring the implementation of the remedy.
Most of this information will also apply to sites that
are covered by other regulatory programs or are
outside of the regulatory process. Factors that
affect the costs of thermal remediation are also
discussed. Advice based on experience gained from
sites where these technologies have been
implemented is provided on soliciting thermal
remediation services, how to determine the
area/volume to be treated, and when to terminate
the heating portion of the remediation. This paper
includes both information gathered from other
published papers and knowledge gained from the
author's extensive experience of technical support
for thermal remediation.
2
-------
Section 2.0 In Situ Thermal
Technology Descriptions and Their
Implementation
Steam enhanced extraction, electrical resistance
heating, and thermal conductive heating are all
adaptations of technologies that have been used in
the oil industry for enhanced oil recovery. While all
these technologies have the objective of increasing
the mobility of contaminants either by reducing the
viscosity of a liquid phase for SVOCs or by causing a
phase change to a vapor for VOCs, they differ in the
manner in which the energy is injected into the
subsurface to bring about these changes. Some
chemical reactions may also be enhanced at the
temperatures used for thermal remediation. All
these technologies include extraction of the
contaminants in the vapor and/or liquid phase via
soil vacuum extraction (SVE) or multiphase
extraction (MPE), and above ground treatment of
the effluent stream(s). Each of the methods of
heating the subsurface has certain advantages or
disadvantages in certain situations; however, there
is also overlap between the applicability of the
technologies to specific sites. In this section, each
of the three heating techniques that are commonly
used today will be described, as well as the
emerging STAR technology, along with some
advantages and disadvantages of each. Also
discussed is when combining thermal technologies
may be the most effective means to treat a site.
Section 2.1 Steam Enhanced Extraction.
Steam Enhanced Extraction (SEE) involves the
injection of steam under pressure with the
concurrent extraction of groundwater, NAPL, and
vapors (Davis, 1998). Ideally, steam injection wells
are constructed in relatively clean soils surrounding
the contaminated area to be treated, while
multiphase extraction (MPE) wells are constructed
centrally located of the steam injection wells within
the contaminated area (Figure 1). For sites that are
larger than the radius of influence of the injected
steam, a 7-spot pattern is recommended (Figure 2).
Thus, each multiphase extraction well, which
extracts ground water, NAPL, and vapors, ideally is
surrounded by four to six injection wells. As steam
is injected under pressure into the subsurface
through vertical injection wells, initially the steam
condenses and gives up the heat of condensation to
the soils. When the soil reaches steam
temperature, steam will begin to move radially into
the subsurface from the injection well. With
continued steam injection, three different
temperature zones form: closest to the injection
well is the steam zone which is at steam
temperature; next is a narrow variable temperature
zone which goes from steam temperature to
ambient temperatures, and further from the well is
the ambient temperature zone (Udell, 1996; Udell
and Stewart, 1989; 1990). Pore fluids (water and
NAPL) will be displaced by the steam and
condensate, pushing it towards the extraction wells.
Figure 1. This figure illustrates the development of a
steam zone as steam is injected into the subsurface
using vertical injection wells that are installed in a 7-
spot pattern (see Figure 2). The boundary of the NAPL-
contaminated area (shown by the dotted line) is
surrounded by the injection wells, and the sharp steam
front displaces the NAPL to the centrally-located
multiphase extraction well. Steam enters the formation
radially from the injection well, so part of the steam
does not enter the thermal treatment area. Each
injection well may accept differing amounts of steam.
Volatile components of the residual liquid saturation
left behind the displacement front are volatilized when
the steam front reaches it, and the vapors are
transported to the steam front for extraction. As steam
injection continues, steam will breakthrough at the
extraction well. After breakthrough, the steam injection
rate is reduced to decrease the pressure in the
subsurface and allow more of the contaminates to
volatilize from low permeability zones and be
transported to the extraction wells. Several cycles of
decreased and then increased steam injection (called
pressure cycling) are normally performed until reducing
the injection pressure does not significantly increase
contaminant extraction rates.
3
-------
A--A o A--A
7 \ *
* Y /
\ ° A""A o
\ / \
Figure 2. When the NAPL-contaminated area to be
treated is larger than the maximum spacing for steam
injection wells, a 7-spot pattern is generally used, as
illustrated, where the triangles represent steam injection
wells and the circles represent multiphase extraction
wells. This pattern with multiple injection wells
surrounding the extraction wells improves the distribution
of steam within the treatment area. The outer periphery
of wells outside of the treatment area should be steam
injections wells, and they should be outside of the NAPL-
contaminated area.
Liquid NAPLthat is displaced will leave behind
residual liquid saturation, which is typically 10 to 20
percent of the pore space. As the steam front
reaches this residual saturation, it will volatilize the
more volatile components of the residual liquids
and these vapors will be transported to the steam
front. When the contaminant is a mixture of
volatile and semivolatile compounds, such as diesel
fuel, creosote, or spent chlorinated solvents, the
volatile components of the mixture (which would
include benzene, toluene, ethylbenzene, and xylene
(BTEX) compounds in fuels or naphthalene in
creosote) will be volatilized, and transported to the
front of the steam zone, where they are recovered
with the rest of the mobilized NAPL. Volatile
compounds with a boiling point of less than
approximately 150°C will be essentially fully
recovered by the passage of the steam front (Yuan
and Udell, 1993), while higher boiling compounds
are likely to remain in the soil pores. With
continued steam injection, essentially all the
volatiles will eventually be stripped from the
residual NAPL.
The steam injection technology employs all the
mechanisms for contaminant mobilization that
come from raising the temperature of the
subsurface (Section 3), and also employs a
displacement mechanism that can be very effective
for recovery of NAPLs composed of semivolatile
compounds such as creosote (Udell and Stewart,
1989). SEE is most effective in more permeable
soils with hydraulic conductivity greater than about
10"3 centimeters per second (cm/sec). Steam
propagation is governed by heat transfer to the
formation, and thus, injection of steam into the
subsurface is a stable and predictable process
(Heron et al., 2005). A potential disadvantage of
steam injection is that when the preferred
arrangement of injection and extraction wells is
employed, with steam injected around the outside
of the area to be treated, a portion of that steam
will heat the area outside of the treatment zone.
Thus, not all the energy injected goes within the
treatment zone due to the radial flow of steam at
the injection wells. These injection wells are
generally in contaminated groundwater, and these
dissolved phase contaminants can be pushed away
from the treatment area and recovery wells.
However, when pressure cycling (terminating or
reducing steam injection while continuing to
aggressive extract liquids and vapors) is performed,
the collapse or reduction in the steam zone will pull
these contaminants back to the treatment area
where they can be recovered. Pressure cycling will
also enhance the vaporization of contaminants in
low permeability zones and their transport to the
more permeable zones so that they can be
recovered via vacuum extraction (Itamura and
Udell, 1995).
Steam, due to its high specific enthalpy, is a very
efficient means of heating the subsurface (Class and
Helmig, 2002). Steam injection is the most cost
effective technology for large, deep sites with
sufficient permeability, due to the fact that greater
spacings between injection and extraction wells can
be used. However, SEE has also been successfully
used at shallow sites where there is sufficient
permeability to allow sufficient steam injection at
low injection pressures. Steam injection below the
water table will impede the inflow of ground water,
4
-------
allowing steam temperatures to be reached even in
high permeability aquifers. Still, to maintain
hydraulic control, the groundwater extraction rate
must exceed the steam injection rate by a factor of
1.5 to 2.5 in order to maintain hydraulic control by
recovering also the groundwater displaced by the
growth of the steam zone. The heating rate is
dependent on the ability to inject steam, which is
dependent on the hydraulic conductivity of the
formation and the steam injection pressure. The
steam injection pressure in turn is dependent on
the depth of injection; the injection pressure cannot
exceed the overburden pressure. Because
overburden pressure increases with depth, higher
injection pressures can be used at deeper sites,
which will allow a greater spacing between injection
and extraction wells, and reduce the amount of
drilling and above ground infrastructure required.
However, the effects of soil anisotropy on steam
migration may necessitate closer injection well
spacing. Typical steam injection well spacing for
shallow sites are between 20 and 30 feet, while full
scale applications at large, deep (treatment depths
of 100 to 200 feet below ground surface) sites have
been implemented with wells spacings exceeding
80 to 90 feet.
Superfund sites where steam injection remediation
has been used include Southern California Edison's
Visalia Pole Yard Superfund Site to recover creosote
(now deleted from the National Priority List), the
former Williams Air Force Base to recover jet fuel
(the largest at 413,000 cubic yards and deepest at a
maximum of 240 feet below ground surface steam
injection remediation to date), and at the Beede
Waste Oil Superfund site to recover waste oil and
chlorinated solvents (see SEE Case Study). Also, a
steam injection research project was performed in
fractured rock at the Quarry site at the former
Loring Air Force Base (Davis et al., 2005).
Section 2.2 Electrical Resistance
Heating. The Electrical Resistance Heating (ERH)
technology was first developed by Pacific Northwest
National Laboratory in the early 1990s. Generally,
the electrodes are installed vertically in a hexagonal
pattern, spaced 12 to 20 feet apart (Figure 3).
Originally, 6-phase current was used with a
different phase applied to each of the 6 electrodes
in the hexagonal array. Eventually 3-phase
alternating current was adopted for most
application, as it allows for a better current
distribution in irregularly-shaped treatment areas.
ERH relies on water in the pore spaces to carry
current between the electrodes. Soils are naturally
resistant to the current flow, and this resistance
dissipates the current energy as heat. Volatile
contaminants are vaporized and extracted via soil
vapor extraction (SVE). The presence of ions in clay
soils can make clays more conductive to electricity
than more permeable sandy soils, which ensures
that current flows through clay zones and they are
heated. Temperatures equivalent to the boiling
point of water can be achieved with this
technology, but water must be maintained in the
pores for current to continue to flow. Thus, in low
permeability soils or when treating soils above the
ground water table, water or an electrolyte solution
is usually added at the electrodes to ensure that
contact is maintained between the electrodes and
thus heating is maintained (Morgenstern et al.,
2007). Some versions of the ERH technology also
rely on water injection to convectively carry energy
from the electrodes, and thus greater water
injection rates are used (McGee et al., 2000; Mejac
et al., 2008). In this case, ground water extraction is
always used in conjunction with vapor phase
extraction.
A variety of electrode constructions have been
used, including steel pipe or copper plate, with a
backfill of conductive materials such as graphite or
steel shot in the borehole annulus to increase the
effective diameter of the electrode (typically 10 to
5
-------
iHIectrical Resistance Heating
Line Power
Power Control Unit
Power to Electrodes
Electrode/Vapor
Extraction Well
Computer controls
power to electrodes
& collects subsurface
temperature data
Vapor Collection
Manifold
Effluent Treatment
System
Condenser/Moisture
Separator
Vapor Phase GAC
or Thermox
10
Liquid Phase GAC
C;
Figure 3. This figure illustrates an ERH system. Three phase alternating electrical current from the power grid is
distributed to electrodes installed vertically (or at angle to reach under a building or other infrastructure) in the subsurface
in a triangular pattern. Current flows within the soil pore water connecting electrodes of different phases. The natural
resistance of the soil to current flow causes the current to be dissipated to the soils as heat. Current flow will stop if the
pore water is totally removed, limiting the temperature that can be achieved with this technology to the boiling point of
water. Vapors are collected by vacuum extraction, and transported to the effluent treatment system, where the condensible
vapors (steam) are separated from the air and volatile contaminants. The noncondensible vapors are then treated through
vapor phase granular activated carbon (GAC) and discharged. At large sites the vapors can be destroyed on site in a
thermal or catalytic oxidizer. Condensed steam is treated by liquid phase granular activated carbon before discharge.
12 inch diameter). Sheet piling has also been used
for electrodes in relatively permeable soils, which
has the advantage of a greater surface area for
contact with the soil and thus greater current flow
(Cacciatore et al., 2008). For deep sites or sites with
changes in geology and therefore electrical
resistivity parameters with depth, the installation of
stacked electrodes is required to properly heat the
full treatment depth. Bored electrodes can also be
vapor recovery wells, and vapor recovery is
accomplished using conventional SVE techniques.
Vapor recovery can also be done by the installation
of separate vapor extraction wells between
electrodes. ERH can be performed either above or
below the water table, and can be used to treat
both zones simultaneously. A major advantage of
ERH is its ability to treat low permeability silts and
clays effectively, reaching contaminants that are
difficult to contact and treat through the injection
of liquids or conventional SVE. Pressures created by
the vaporization of ground water and contaminants
will force the flow of vapors to the extraction
points. Drying of the soils will also aid in increasing
its permeability to vapors, however, as mentioned
6
-------
previously, some water must be maintained in the
pore spaces to carry current between the
electrodes. Ground water flow rates greater than
approximately 1 foot per day are challenging for
ERH systems and may limit the system's ability to
reach the boiling point of water and/or can lead to a
loss of hydraulic control. In these situations
groundwater extraction upgradient of the
treatment area with reinjection downgradient may
flatten the hydraulic gradient and extend the use of
this technology to more permeable saturated
formations. Another option may be to limit
groundwater flow into the treatment area by
installing an upgradient hydraulic barrier such as a
sheet pile wall or slurry wall
Superfund sites where ERH has been used include
East Gate Disposal Yard at the Fort Lewis Army
Depot where waste oil and chlorinated solvents
were recovered, Cleburn Street Well site where
chlorinated solvents were recovered (see ERH Case
Study), the South Municipal Well Superfund Site
where chlorinated solvents were recovered, and the
Hamilton-Labree Superfund site where chlorinated
solvents were recovered (TerraTherm, 2022).
Section 2.3. Thermal Conductive
Heating. As the name implies, this technology
relies on heat conduction through the soil to heat
the target treatment area. This technology was first
developed primarily by Shell Oil Company in the
1990s with the participation of a variety of other
companies (Stegemeier and Vinegar, 2001). The
basis for using heat conduction for applying energy
to the subsurface is the fact that the thermal
conductivity of different soil types is fairly uniform,
varying only by approximately a factor of two to five
(dry sandy soil will have a thermal conductivity of
0.5 - 1.0 Watt per meter-Kelvin (W/mK) while wet
silt/clay is closer to 2.0 - 2.5 W/mK). This allows for
a relatively more uniform heating rate in highly
heterogeneous soils than can be achieved by the
injection of energy as a fluid, such as by hot water
or steam injection. However, the thermal
conductivity of soils is low, which necessitates a
steep temperature gradient (thus high
temperatures) at the points of energy application in
order to transport energy away from the points of
application (Heron et al., 2015). Thus, heater
temperatures in the range of 500 to 800°C are
typically used (Figure 4), with heater spacings on
the order of 6 to 20 feet, depending on time
available for remediation (larger spacing will require
longer remedial timeframes to reach the same
endpoint).
Figure 4. This figure illustrates an electrical thermal
conduction heating (TCH) heating well installed in
heterogeneous soils. The temperature of the heater well is
in the range of 700°C in order to increase the conduction of
heat into the soils. The thermal conductivity of soils is
relatively uniform, thus different soil types will heat at
about the same rate. Most TCH systems rely on electrical
energy, however, one version of the technology uses
natural gas or propane combustion at the well heads to
provide the energy. The heater wells are installed in a
triangular pattern at spacings of 12 to 20 feet with
extraction wells located throughout the treatment area,
generally at the midpoints between the heater wells.
When installed above the water table, treatment
temperature up to 350 - 400°C can be reached to treat
higher boiling compounds.
7
-------
TCH is the only one of the three commonly used
technologies that can reach temperatures above
the boiling point of water, and in the higher
temperature applications (with target temperatures
up to 350 - 400°C) the heater spacing is typically
between 8 to 12 feet. Also, multiple well patterns
are generally used to increase heating effectiveness
and efficiency through superposition of
heat/energy. Vapor extraction is used to recover
the vaporized contaminants. A significant
advantage of this technology is that because it does
not rely on water for energy application or
convection, temperatures in excess of the boiling
point of water can be achieved when the soils are
unsaturated. This allows not only VOCs but also
SVOCs to be treated by this technology to achieve
very low residual soil concentrations by raising the
temperature to 250 - 350°C (Baker and Heron,
2004). Semivolatile contaminants such as
polyaromatic hydrocarbons (PAHs), polychlorinated
biphenyls (PCBs), pesticides, and dioxins and furans
have been treated using TCH technologies.
Various types of TCH systems available from the
thermal technology vendors. In general terms, the
TCH technology is implemented in situ by installing
heater wells in vertical boreholes. Angled or
horizontal wells can also be used, particularly for
extending under buildings or other structures. The
wells are installed in triangular patterns, creating a
repeating series of hexagons. Energy applied to the
heater wells creates temperatures of 500 - 800°C at
the wells, but due to the steep temperature
gradient, only soils within approximately 6 to 12
inches of the heater wells reaches temperatures
above the boiling point of water in a typical VOC
treatment application. This heat is conducted into
the soil, resulting in a relatively uniform radial
heating pattern even in heterogeneous soils.
Contaminants are vaporized and recovered by SVE.
Vacuum extraction can be built into the heater
wells, so that all wells are dual purpose, or they may
be adjacent to (within three feet of) or at centroid
locations between the heater wells. Most
variations of TCH use electrical energy, however,
one variation of the TCH technology uses natural
gas, propane or other fuel combustion at the well
heads as the energy source. This can be
advantageous in areas where electrical power is not
readily available, however, natural gas may also be
limited in the winter months in colder areas.
Generally treatment durations range from several
months to about a year, depending on the
temperature required for volatilization of the
contaminants and the selected well spacing. When
treating below the water table, higher permeability
soils - and the resulting higher ground water flow
rates - will result in longer treatment times or
necessitate a tighter well spacing, as water flow
through the treatment area will cause heat to
migrate from the treatment area. The TCH
technology is applicable for groundwater flow rates
less than 0.1 to 1 foot per day. Where higher flow
rates are encountered, barriers such as sheet piles
or slurry walls or pumping wells to flatten the
hydraulic gradient have been used to reduce the
effects of groundwater flow.
For the recovery of SVOCs such as polychlorinated
biphenyls (PCBs) or coal tar contaminants, the
target temperature may be as high as 350°C,
depending on the boiling point of the contaminant
and the degree of treatment desired (Stegemeier
and Vinegar, 2001). These high treatment
temperatures can only be achieved above the water
table, or when the water table is controlled to
eliminate ground water flow into the treatment
area. Recently, the treatment of Per- and
Polyfluoroalkyl Substances (PFAS) soil
contamination using TCH was demonstrated (ESTCP
project ER20-D1-5198). PFAS concentrations in soil
were reduced from an average of 232 micrograms
per kilogram (jug/kg) to 4.1 |a,g/kg (range of 20 |a,g/kg
to nondetect at 0.5 |a,g/kg). PFAS compounds were
observed to react and breakdown over a wide
temperature range (Heron, 2023).
Superfund sites where TCH has been used include
Solvent Recovery Services of New England (see TCH
Case Study), Memphis Depot Superfund site (Heron
et al., 2009), the Velsicol Superfund Site where
chlorinated solvents were recovered, and Mare
Island Naval Shipyard where PCBs were recovered.
8
-------
Section 2.4. Self-Sustaining Technology
for Active Remediation (STAR) is an
emerging technology that may be applicable for
sites that are heavily contaminated with NAPLs that
have low volatility and significant fuel value, which
includes longer chain petroleum hydrocarbons, coal
tar, and creosote. This technology relies on
smoldering combustion processes, where a liquid or
solid fuel within a porous matrix is slowly
combusted. The combustion processes, which are
dominated by pyrolysis and oxidation, are initiated
by the addition of heat (typically approximately 250
to 400°C) at the ignition point and sustained by the
heat generated by the combustion process itself
and the injection of air (Figure 5).
Ignition Point
Air Injection .
Vapor Extraction
and Treatment
Vapor Collection
Point*
Vapor Collection
Points
Figure 5. This figure illustrates the emerging smoldering combustion technology called Self-Sustaining Treatment for
Active Remediation (STAR). This technology is applicable to SVOCs such as coal-tar based products or heavy petroleum
hydrocarbons that have more limited volatility but significant fuel value. At the ignition point, a heater is inserted to start the
combustion process. The combustion process itself produces the energy required to continue the combustion of the
contaminants as the combustion front migrates away from the ignition point. Air must be injected to sustain the process.
Tliis process works above and below the water table. Volatile compounds are collected via vacuum extraction along with the
products of combustion, which are treated through vapor phase GAC before discharge.
Smoldering combustion requires that the fuel have
a large surface area exposed to the oxidizer (air), a
condition which can be attained when the fuel is
contained within a porous matrix such as soils. The
smoldering combustion process destroys the higher
boiling point NAPL components in situ, producing
carbon dioxide, carbon monoxide, and water vapor.
Lighter components of the NAPL, such as benzene,
toluene, ethylbenzene, and xylene (BTEX) are
generally volatilized before the combustion front
reaches them. The combustion products as well as
the volatile components of the NAPL (typically less
than two percent of coal tar) are extracted via
vacuum extraction. Typically, this technology
requires NAPL concentrations of the higher boiling
compounds in excess of 3,000 to 5,000 milligram
9
-------
per kilogram (mg/kg) total petroleum hydrocarbon
(TPH) in order to produce the heat required in situ
to sustain the combustion front as it migrates from
the ignition point. Due to limitations with injecting
air, this technology requires a more permeable soil,
with grain sizes ranging from silty sands to gravels,
but even then the typical radius of influence of an
ignition point is around 10 feet, depending
somewhat on lithology. The combustion process
can be sustained below the water table because the
heat produced will dry the nearby soils, and vapors
generated by the process displace the groundwater,
allowing the combustion to proceed.
Oxygen is required to sustain the process, so the
smoldering combustion process can be extinguished
by terminating air injection. Heterogeneity of the
permeability to air can cause reduced air flow in
lower permeability soils that can limit the
combustion process. Final soil concentrations are
generally reduced by 99 percent, with
concentrations ranging from nondetect to a few
tens to hundreds of mg/kg of TPH (Scholes et al.,
2015; Grant et al., 2016).
Section 2.5. Other Thermal
Technologies. A number of other methods for
heating the subsurface for enhanced oil recovery
and/or remediation have been the focus of research
by a variety of researchers. Radio Frequency (RF)
heating has been applied to remediate a limited
number of sites, but is not currently commercially
available. More recent research has shown that RF
heating is not as energy efficient as ERH, due to the
energy losses (on the order of 40 to 50 percent) to
convert power line frequency to RF (Roland et al.,
2011). Microwave heating has been studied
extensively in the laboratory where it has been
found to be successful for heating the soil and
removing volatile contaminants, however,
microwave energy has a very limited penetration
into soils (Falcigilia et al., 2016) that so far has
limited its applicability in the field. Hot air injection
and hot water injection for remediation have also
been the subject of research and attempts in the
field (Davis and Lien, 1993; Johnson, 1994). While
hot water injection can aid in the reduction of the
viscosity of NAPLs which may aid in their recovery,
these low energy methods are not as effective as
other thermal remediation technologies that are
more energy intensive. Low energy thermal
methods are not generally used today.
Other low energy methods with objectives such as
enhancing biodegradation or other chemical
reactions such as hydrolysis have received
considerable interest by researchers but have been
found to be only marginally less costly than thermal
remediation at temperatures of the boiling point of
water when applied to source areas. Also, there is
less certainty in treatment performance and life
cycle costs associated with low temperature
applications meant to increase biotic and/or abiotic
reaction rates to destroy contaminants in situ
(Macbeth et al., 2012). Moderate temperature
increases have been used enhanced biodegradation
in the dissolved phase plume downgradient of
thermal treatment areas (Heron, 2023), however,
temperatures above approximately 35°C have been
found to decrease or eliminate biological
dechlorination of some chlorinated solvents
(Pennell et al., 2009). Low energy, low temperature
thermal systems usually try to reduce costs by not
extracting vapors, which increases the risk of
fugitive emissions. Even at ambient temperatures,
soil gas concentrations of volatile contaminants can
pose a risk to indoor air. Any increase in
temperature will increase the partitioning of
volatiles to soil gas, possibly causing fugitive
emission to the atmosphere or to indoor air if vapor
extraction is not used.
ERH and TCH can also be used to treat soils that
have been excavated. Ex situ application of these
technologies takes place within an engineered,
above ground, fully enclosed treatment pile
structure (Figure 6). Ex situ treatment has been
used for contaminated shallow soils that cannot be
practically treated in situ. Ex situ treatment may be
more appropriate for high boiling compounds such
as PCBs, dioxins and furans, or PFAS, that are more
easily excavated without causing excessive fugitive
emissions of volatile contaminants, but for which
off site disposal is not an option or is very costly.
The principles of the system are the same as for in
situ treatment, in that electrodes or heater wells
10
-------
are installed to heat the soil, and vaporized
contaminants are extracted under vacuum. For
high temperature ex situ TCH applications at
temperatures greater than 200°C, degradation
processes such as pyrolysis and oxidation have also
been found to be important. Treatment piles
typically range in size from 50 cubic yards (yd3) to
70,000 yd3, and treatment times will depend on the
treatment temperature required and the remedial
goals. A noteworthy example is the treatment of
dioxin-contaminated soils and sediments at the
Danang airport in Vietnam (Sorensen et al., 2018).
Steel Sheeting
Insulated
Sidewall Panel
Insulated
Floor
Insulated Surface
Cover
Horizontal Vapor
Extraction Well
Grade Level
Drain Piped
to Sump
Drainage
Layer
Well
Figure 6. This is an example of how excavated soils or sediments can be thermally treated. In this particular
application, ex situ treatment was used to treat dioxins. The use of herbicides during the Vietnam war left surface soils in
parts of the country contaminated with highly toxic dioxins. Contaminated surface soils and sediments were excavated,
then placed in the concrete foundation to create an ex situ thermal treatment pile. The soils were treated using thermal
conductive heating (TCH) at temperatures exceeding 300C to reduce the 2,3,7,8-TCDD concentrations from as high as
157,000 parts per trillion (ppt) to less than 150 ppt (Sorensen et al., 2018).
Section 2.6. Which technology is
Appropriate for a Given Site? while erh
and TCH are more applicable in lower permeability
soils, and SEE requires more permeable soils, there
is overlap in the applicability of these technologies
depending on the contaminant to be remediated,
the geologic and hydrogeologic setting, and the
remedial action goals. Best practice is to allow the
thermal vendors to propose how they would
address the site during the bidding process,
allowing them to determine whether or not they
believe their technology, or which of the
technologies they offer, is best suited for the site.
When treating VOCs such as chlorinated solvents,
their volatility can be exploited to recover them in
the vapor phase with any of the three commonly
used in situ thermal technologies, and the choice of
technology will be more dependent on the
hydrogeologic setting.
For SVOCs, the permeability of the soils, the
hydrogeologic setting, and the remedial goals will
all be important in selecting the technology. If the
remedial goal is to recover mobile NAPL, either ERH
or TCH may be used effectively to reduce the
viscosity of the NAPL to enhance its migration to
extraction wells and increase its ability to be
recovered by pumping. Recovery as a NAPL will
generally aid in recovering more of the higher
boiling components of the NAPL with less energy
input. When more stringent cleanup criteria are to
be achieved, once the recoverable NAPL and
groundwater has been extracted, the temperature
can be raised further using TCH to recover the
contaminants in the vapor phase (Baker et al.,
11
-------
2006). TCH, however, can only achieve
temperatures above the boiling point of water
when implemented above the water table, since a
continuous inflow of groundwater would buffer the
temperature at the boiling point as the water is
boiled off. In permeable soils, SEE has been used
successfully to recover contaminants such as
creosote. Where there are higher levels of
contamination in more permeable soils either
above or below the water table, STAR may be
applicable and able to achieve stringent soil cleanup
criteria. However, the STAR technology may not be
able to treat zones with source material at TPH
concentrations less 3,000 mg/kg.
For sites where the contamination resides both in
high permeability and low permeability soils, a
combination of thermal technologies may be used.
SEE for the more permeable zone can be combined
with either ERH or TCH in the low permeability zone
for volatile contaminants (Newmark et al., 1994;
Newmark and Aines, 1997; Heron et al., 2005;
Heron et al., 2012), while a combination of TCH and
STAR may be applicable at sites of varying hydraulic
conductivity that are contaminated by semivolatile
contaminants. It may be advantages at some sites
to combine ERH and TCH. For example, ERH may be
used at sites with large zones with VOCs where the
boiling point of water is sufficient for treatment,
and TCH can be used for smaller areas with SVOCs
needing treatment at higher temperatures.
Another example would be shallow PCBs in topsoil
which could be treated at high temperatures using
TCH and solvents at deeper depths can be treated
with ERH. ERH is more cost-effective for the VOC
source zones in most cases, particularly for clay and
silt soils and where the subsurface is really wet
(vadose zone and saturated zone). ERH can also
tolerate a lot more water flow before the cooling
reduces efficiencies. At one site ERH was used to
treat a large area and TCH to heat below a creek, as
not having electrodes close to the surface water
reduced the electrical risks (since that time,
methods have been developed for grounding
electrodes and prevention of electrical shock
hazards when operating close to surface water or
other electricity- conducting materials) (Heron,
2023).
SEE, ERH and TCH have all been successfully used to
remediate fractured rock. Again, the appropriate
technology for the site is heavily influenced by the
permeability of the fractured rock. Highly fractured
rock with significant groundwater flow will favor
SEE, while competent rock with relatively few
fractures will favor TCH or ERH. If the rock is of low
porosity and thus low water content, the electrical
resistivity of the rock may be high, making ERH
application more difficult, and favoring TCH.
Sampling of the rock matrix at various sites has
shown that contaminants generally do not migrate
significantly from the fracture surface into the rock
matrix (Davis et al., 2005), and steam temperatures
can be conducted considerable distances into the
rock matrix (Stephenson et al., 2006), so steam
injection may be very effective for recovering the
contaminants from fracture zones without heating
the entire rock matrix (Kluger and Beyke, 2010;
Lebron et al., 2012; Beyke et al., 2014).
Section 2.7. Above Ground Treatment of
Extracted Contaminants. All of the thermal
remediation technologies include above ground
treatment of the extracted vapors and groundwater
to separate the contaminants from the air and
water before discharge. The applied vacuum to the
subsurface varies by application but vapor
extraction mechanisms cannot be compared with
traditional ambient temperature SVE systems,
especially not in low permeability geology settings.
During the thermal remedy the extraction points
can be considered pressure release points and
vapor pathways out of the tight formations, rather
than active extraction points to pull air and vapors
through the soil with a given radius of influence. For
that reason the extraction point density is more
dense than for traditional SVE systems, with well
spacings typically between 15 and 30 feet.
Commonly used treatment trains include vapor-
water separators, heat exchangers, oil/water
separators, and granular activated carbon for both
the vapors and water, in that order. For larger sites
where the use of carbon may be cost-prohibitive, a
thermal oxidizer, catalytic oxidizer, or thermal
accelerator may be used to destroy the
contaminant vapors. Steam-regenerated granular
12
-------
activated carbon (GAC) systems with multiple
adsorption vessels can be used to minimize GAC
usage. These systems that are used on site can
regenerate spent GAC in 4 to 8 hours while
adsorbing and treating extracted vapors. This
produces liquid wastes (NAPL and water
condensate) for treatment and disposal.
Alternatively, a cooling, compressing, and
condensing system has also been used at some sites
to condense the vapors to a liquid which is then
disposed off site. Fuel hydrocarbons may have
recycle value, making one of these alternatives
which convert vapors to liquids especially attractive.
However, for many types of mixed waste
contaminants, this may not be advantageous. Some
communities have objected to having a thermal
oxidizer or one of its variations at a local site to
destroy the recovered contaminants, but where
these systems can be used, the onsite destruction
of contaminants will eliminate the need to
transport the liquid wastes via truck through the
community, which may be inherently safer.
Initially contaminant mass recovery rates may be
small, however, at large sites, hundreds of pounds
of contaminants may be extracted per day as the
site reaches treatment temperature and recovery
rates reach their peak. If the contaminants are
VOCs, most of the mass will be recovered in the
vapor phase, while SVOCs may be recovered as a
liquid. SEE remediation systems will also tend to
recover more of the contaminants as a liquid due to
the displacement of liquids ahead of the steam
front. Robust above ground treatment systems are
required even if the estimates of mass to be
recovered are low or moderate, as estimates of
contaminant mass in the ground are notoriously
inaccurate. Contaminants recovered as a NAPL may
be emulsions in groundwater due to biological
growth, requiring biocides or some other emulsion
breaker to treat the emulsions. Organoclay filters
may be used to improve the separation of the NAPL
from water. Heating the aquifer may also increase
the dissolved phase concentrations of naturally
occurring elements such as arsenic, which may
require treatment before the water can be
discharged.
For Superfund sites undergoing remediation, air and
water discharge permits are not required.
However, non-condensible vapors that are
discharged to the air and water that is discharged to
a publicly owned treatment works (POTW), a storm
sewer, to surface water, or reinjected to the
aquifer, must meet state and/or local discharge
criteria. Weekly or monthly monitoring of the air
and water discharge streams is generally required
to ensure that the air and water discharge criteria
are met.
Section 2.8. Monitoring of Thermal
Remediation Systems, in addition to
monitoring the air and water discharges to ensure
the above ground treatment system is containing
the contaminants, subsurface monitoring is
required to ensure that the entire target treatment
area is being heated, and to verify that hydraulic
and pneumatic control are being maintained so that
contaminants are not lost as fugitive emission or by
migrating away from the treatment area. The
extraction rate of contaminants is also monitored
throughout the remediation to aid in determining
when the remediation reaches 'diminishing
returns'. This section describes the commonly used
means to monitor subsurface conditions and the
rate of contaminant mass recovery.
Section 2.8.1. Temperature Monitoring.
Subsurface temperatures are used to verify that the
energy injected to the subsurface is reaching the
entire target treatment area. All in situ and ex situ
thermal remediation systems include temperature
monitoring points (TMPs) in the treatment zone.
Temperature monitoring strings are inserted in
boreholes throughout the treatment areas, with
thermocouples, fiber optic temperature measuring
systems, or other temperature measuring devices
spaced every 3 to 5 feet vertically, starting a little
above the treatment area and commonly extending
about 1 to 5 feet below the treatment area.
Temperature monitoring strings generally are
placed at the centroids between wells where the
energy is applied to the subsurface, one string for
every 1,000 to 3,000 square feet of treatment area
depending on the heterogeneity of the subsurface
and the thermal technology used. Temperatures
13
-------
are normally measured and recorded automatically
at pre-set intervals, which may be a short as 15
minutes or less. Sometimes real time subsurface
temperature data is available to the site owner on
the vendor's website. The vendor should be
expected to monitor the temperature data on a
regular basis (usually daily) to ensure that the
subsurface is heating at the desired rate, and that
the target temperature for the remediation is
achieved throughout the treatment area.
Section 2.8.2. Pneumatic and Hydraulic
Control Monitoring. Subsurface monitoring
should also include monitoring to verify that
hydraulic and pneumatic control are being
maintained throughout the remediation to ensure
that contaminants are being captured and are not
migrating away from the treatment area. Best
practice is to start the extraction system(s) at least
several days before heating is initiated to
demonstrate that the vacuum extraction system
creates a vacuum throughout the treatment area,
however, it is not always possible to measure
vacuum at pressure monitoring points before
heating starts in low permeability soils. If
multiphase extraction is to be used, the amount of
extracted groundwater during this time should be
documented along with an estimate of the water
extracted as steam during the remediation. It is not
possible to open monitoring wells within the
thermal treatment area during heating to measure
the groundwater elevation within the treatment
area, so groundwater elevations cannot be used to
determine if hydraulic control is being maintained.
Pressure buildup in the subsurface due to heating
and the generation of vapors makes it dangerous to
open monitoring wells to the atmosphere while the
groundwater temperature is elevated. The sudden
release of pressure caused by the opening of the
well can allow hot groundwater to flash to steam
which can escape through the well casing. Pressure
transducers or bubblers within the monitoring wells
can sometimes be used, however, the harsh
conditions in the hot groundwater can damage the
transducers.
A more reliable and safer method of determining if
hydraulic control is being maintained is to measure
temperatures surrounding the treatment area with
thermocouple strings similar to those used within
the treatment area. Some temperature increase
outside of the thermal treatment area can be
expected during heating due to conductive heating,
but sudden increases in groundwater temperature
are likely to be due to the migration of hot
groundwater from the treatment area (Figure 7),
which may carry elevated contaminant
concentrations with it. Thus, thermocouple strings
located exterior to the treatment area and co-
located with monitoring wells are recommended
surrounding the treatment area.
Vapor monitoring points that allow measurement of
soil gas pressure/vacuum should also be included at
these exterior monitoring points - as well as
interior to the treatment area - to demonstrate that
pneumatic control is being maintained. It should be
noted that a small pressure build-up within the
treatment zone is expected especially in low
permeability soils due to the steam production and
transport of vaporized contaminants in the steam
phase to the extraction points.
Section 2.8.3. Contaminant Mass Recovery
Rate. Monitoring the extraction rate of
contaminants as a function of time is critical for
aiding in determining when the thermal system has
accomplished what it reasonably can in terms of
recovery of contaminants. When a majority of the
contamination is located above the water table in
higher permeability soils, vapor extraction rates
may be significant right from the time that vapor
extraction is initiated. When most of the
contamination is located below the water table,
initial extraction rates may be low. In both cases,
the extraction rate should be expected to increase
as the subsurface is heated. The maximum
extraction rate may correspond roughly to around
the time that temperatures in the subsurface
approach the co-boiling point of the NAPL (see
Section 3.1). Extraction rates can then be expected
to decrease gradually as the contaminants within
the treatment area are depleted. The extraction
rate can be expected to reach a continuous low rate
when the majority of the subsurface mass has been
depleted, at which point the remediation is
14
-------
Figure 7. For this small scale pilot ERH, temperature monitoring points (TMPs) were placed surrounding the thermal
treatment area. A thermal model was used to predict the temperature at these exterior TMPs over time due to
conductive heating from the treatment area as the site heated; these results are shown with squares in the graph. The
actual temperatures recorded at these TMPs were lower than predicted during the first four months of operation, likely
as a result of groundwater flow towards the treatment area carrying the heat back towards the treatment area. After that
time, the temperature graphs show rapid temperature increases at some depths, indicating that hot water migrated from
the treatment area. This graph illustrates the fact that generally TMPs exterior to the treatment area provide a reliable
means of determining if hydraulic control is being maintained, and the difference between the expected conductive
heating and loss of hydraulic control of hot water is clearly differentiated. Soil samples were taken adjacent to these
exterior TMPs post treatment, and it was found that contaminants had not been transported outside of the treatment area.
considered to have reached 'diminishing returns'. If
target temperatures have been reached throughout
the treatment area and groundwater concentration
have decreased to low levels, then the remediation
has likely accomplished most of what it can, and
heating can be terminated. Extraction should
continue during at least the early stages of cool
down, while subsurface temperatures remain above
the boiling point of water.
To determine the extraction rate, photoionization
detector (PID) or flame ionization detector (FID)
readings of the vapor flow into the vapor treatment
system are normally made every day that an
operator is on site (normally 5 days a week). Flow
rates are also measured, so that mass extracted can
be calculated. Also, weekly summa canister
samples are obtained and analyzed for the site
contaminants to verify the PID or FID readings.
Groundwater concentrations should also be
monitored during the remediation. As with vapor
extraction rates, the concentrations should be
expected to increase during heating, then decrease
as the contaminant mass in the subsurface is
depleted. Monitoring wells must be constructed to
be able to obtain groundwater samples without
opening the wells during heating due to safety
concerns (see Section 2.8.2). If multiphase
extraction wells are installed as part of the thermal
system, sample ports at the wellhead may allow for
groundwater sampling. Hot groundwater sampling
techniques should be used, which generally involve
running the groundwater through a stainless steel
or copper coil immersed in an ice bath to cool it to
ambient temperatures before obtaining the sample
for analysis. Hot soil samples techniques can also
be used to obtain interim and/or confirmation soil
samples by obtaining the samples within a stainless
steel sleeve, immediately capping both ends tightly,
and then cooling the soil on ice before opening the
sleeves to obtain samples for analysis (Gaberell et
al., 2002).
15
-------
Section 3.0 Mechanisms for
Increased Contaminant Recovery
during Thermal Remediation
Generally the most important property of an
organic compound that determines whether or not
it is amenable to thermal remediation is the boiling
point, which is directly related to its vapor pressure
(Hunt et al., 1988; Davis, 1997). The boiling point or
vapor pressure will also influence the temperature
required for the effective recovery of the
compound. Volatile organic compounds (VOCs)
which are only slightly soluble in water and which
have boiling points less than about 150°C are
generally amenable to volatilization at
temperatures less than or equal to the boiling point
of water (100°C at standard atmospheric pressure)
and can be readily recovered in the vapor phase
using thermal technologies. Some compounds with
higher boiling points, such as naphthalene (boiling
point 218°C), have also been recovered at
temperatures around 100°C. For semivolatile
organic compounds, such as coal tar or creosote,
viscosity reduction and displacement of the liquid
phase may be more important as a recovery
mechanism during thermal remediation at
temperatures up to approximately 100°C, and
vaporization only becomes important at
significantly higher temperatures.
Section 3.1. Increased Vaporization and
Co-boiling of NAPL and Groundwater.
The vapor pressure of organic compounds is an
exponential function of temperature. For VOCs, as
the temperature is increased above about 50 -
60°C, the vapor pressure of most of these
compounds will increase rapidly (Figure 8),
facilitating their recovery in the vapor phase via
vacuum extraction. When two separate liquid
phases are present such as when VOC NAPLs are
present at or below the water table, the vapor
pressure from each liquid contributes to the overall
vapor pressure (Figure 9), and the combined liquids
will boil when the total of their vapor pressures
equals the local pressure (Atkins, 1986; DeVoe and
Udell, 1998). This temperature can be called the co-
boiling temperature. The temperature at which co-
boiling of a VOC NAPL occurs is always less than the
boiling point of water, and the lower the boiling
point of the NAPL, the lower the co-boiling
temperature (Table 1).
Table 1. Boiling points and co-boiling points for
some organic compounds of enviromnental interest at
atmospheric pressure (760 mm Hg)
Chemical
Boiling Point (9C)
Co-boiling Point {^C)
Methylene Chloride
40
38
cis-l,2-Dichloroethylene
60
55-56
1,2-Dichloropropane
96.8
78
Tetrachloroethylene
121.3
88
Benzene
80.1
69
Trichloroethylene
87.3
73-74
Toluene
110.6
84-85
Naphthalene
218
99
Xylenes
138-144
92-94
Chlorobenzene
131.7
91
Carbon Tetrachloride
76.8
67
1,1,2-Trichloroethane
114
86
1,1,2,2-Tetrachloroethane
148
94-95
Thus, when VOC NAPL is present in porous media
along with water, the NAPL and water will boil
when their combined vapor pressures equal the
local pressure. When temperatures exceed the co-
boiling temperature, the NAPL phase has been
removed, however, dissolved and adsorbed phase
contamination will remain in the subsurface.
Maximum removal rates and removal efficiency
requires increasing the temperature to at least the
boiling point of groundwater under the conditions
of the remediation, which maximizes steam
stripping, the process by which steam aids in
transporting the vapors to the extraction points.
The same principles apply to SVOC NAPLs such as
the PAHs that comprise creosote or coal tar, but
higher temperatures are required to create the
significant increase in vapor pressure (Figure 8).
Thus, for SVOC NAPLs, their vapor pressure is not
high enough at 100°C to cause a decrease in the co-
boiling point with groundwater. Significant
volatilization of SVOCs in order to recover them
effectively requires higher temperatures. Coal tar,
creosote, PCBs, and dioxins have all been
successfully recovered in the vapor phase at
temperatures in the range of 300 - 375°C. The
volatilization behavior of per- and polyfluoroalkyl
compounds, commonly referred to as PFAS, varies
16
-------
depending on their alkyl chain length and their
functional group. Laboratory testing found that 10
to 14 days of heating at temperatures of 350°C and
above reduced soil concentrations of PFAS by as
much as 99 percent (Crownover et al., 2019). A
recent field demonstration of TCH treatment of
PFAS-contaminated soils showed that similar
reductions are possible in the field (Heron, 2023).
1,4-Dioxane is a contaminant that is commonly co-
located with chlorinated solvents, having been
added to products such as 1,1,1-trichloroethane in
low concentrations as a stabilizer. 1,4-Dioxane has
a boiling point of 101°C, but is much more soluble in
water than most chlorinated solvent VOCs, giving it
a very low Henry's constant at ambient
temperatures. However, under the thermal
treatment conditions normally used for the
recovery of chlorinated solvents, 1,4-dioxane in soil
and groundwater have been found to be reduced by
as much as 99 percent (Oberle et al., 2015).
Increased vaporization of contaminants and
groundwater due to increases in temperature in low
permeability soils will increase the pressure if the
vapors cannot readily migrate from the low
permeability soil. This increase in pressure can
create microfractures and/or vugg porosity in clay
soils (Figure 10), which increases the permeability
of the soils and allows the vapors to be collected via
vacuum extraction. This increase in pressure in low
permeability soils as the soils are heated makes
thermal remediation of low permeability soils much
more effective than SVE at ambient temperatures,
which relies on sufficient permeability in order to
pull air and vapors through the soil pores (Heron et
al., 2013).
1000
900
800
700
600
500
400
300
200
100
0
Vapor Pressure vs Tempera
ure
Figure 8. This figure shows the vapor pressure as a function of temperature for a variety of organic chemicals of
environmental interest. Volatile compounds such as benzene and TCE can accumulate in soil gas and intrude into buildings
at concentrations that are hazardous to health even at ambient temperatures. As vapor pressures increase exponentially with
temperature, driving these chemicals into the vapor phase, there is greater potential for vapor intrusion from these chemicals.
For compounds like benzene and TCE, the increase in vapor pressure even at more moderate temperature increases is
substantial. However, it is also easier to recover these chemicals using soil vacuum extraction when they are driven into the
vapor phase. SVOCs such as the higher-molecular weight PAHs require significantly higher temperatures to volatilize them
to a significant extent.
Benzene
TCE
PCE
Chlorobenzene
Naphthalene
Phenanthrene
Pyrene
Benzo(a)pyrene
50 100 150 200 250 300
Temperature C
17
-------
1
.8
1
.6
1
.4
1
.2
1
0
.8
0
.6
0
.4
0
.2
0
Temperature vs. Vapor Pressure
40 60 80
Temperature (C)
H20
-PCE
-Total Vapor Pressure
Co-boiling point = 88C
Figure 9. This figure illustrates the additive effect of vapors coming from separate liquid phases to the total vapor pressure^
which creates boiling of the two liquid phases when their combined vapor pressures equal the ambient pressure. When water and
a separate phase volatile organic compound are both present, the boiling point of the combined liquids will always be less than the
boiling point of water. This temperature can be called the co-boiling temperature. During thermal remediation of a site
containing NAPL, when the temperature reaches the co-boiling point, it will remain at that temperature until the NAPL is
removed by converting it to vapor. When the temperature reaches the boiling point of water alone, the NAPL lias been boiled off.
This figure shows that the co-boiling point of PCE DNAPL and water at atmospheric pressure is 88C. Below the water table,
where hydrostatic pressure causes overall higher pressures, the co-boiling temperatures is higher.
Section 3.2. Decreased viscosity of
groundwater and NAPL As the temperature
increases, the viscosity of water arid organic liquids
will typically decrease exponentially. Many of the
chlorinated VOCs and petroleum hydrocarbons
have relatively low viscosities even at ambient
temperatures, ranging from about 0.5 to about 1.5
centipoise, which is similar to that of water at the
same temperature. Thus, even the exponential
decrease in viscosity with temperature may not be
significant with respect to flow rate of these liquids.
However, many SVOCs, including creosote and coal
tar, may have a very high viscosity at groundwater
temperatures which decrease exponentially as the
temperature increases (Figure 11). For these
contaminants, the decrease in viscosity at
temperatures of 50 to 80°C can significantly
increase the mobility of the liquid, allowing it to be
more readily recovered in the liquid phase. This
approach, which has been used in enhanced oil
recovery, has also been used to recover coal tar and
creosote in the liquid phase, which has the
advantage of being able to recover more of the
higher boiling compounds that are difficult to
vaporize, and to recover them with less energy
input.
Figure 10. Sample of clay recovered from 25 feet below
ground surface after 30 days heating by ERH. Steam
bubble formation and escape created microfractures and
vuggv porosity. It was noted that the vacuum pressure on
the subsurface decreased and the flow rate increased as
heating of the clays progressed, with concurrent release of
steam and vaporized contaminants (McGee et al., 2006).
18
-------
Creosote Viscosity as a function of Temperature
Temperature (C)
Figure 11. Viscosity as a function of temperature for cresote product recovered from a Superfund site. Generally, the more
viscous a liquid is at ambient temperatures, the greater will be the viscosity reduction as the temperature increases, allowing
the liquid to flow more readily to wells. This reduction in viscosity will also allow it to be pumped to the surface more
readily.
Section 3.3. Other Mechanisms, other
mechanisms that may aid in the thermal
remediation process include increased solubility
and solubilization rates (Imhoff et al., 1997), and
increased Henry's constant which increases the
volatilization from water (Heron et al., 1998).
Research has also shown that there may be
increased desorption of chemicals of environmental
interest from soils (Cornelissen etal., 1997; Pennell
et al., 2009). Increases in solubility and
solubilization rates generally cause increases in
groundwater concentrations during the initial
stages of heating as NAPL dissolves and/or
contaminants are desorbed from soil surfaces. The
groundwater concentration increases may be more
than an order of magnitude. Once the NAPL has
been essentially depleted, decreasing groundwater
concentrations trends will be observed, although it
can be expected that there will be considerable
variation in concentrations during heating.
Heating contaminated groundwater causes
evolution of gases from the groundwater as well as
the formation of VOC and water vapors. The gases
and vapors are buoyant relative to the
groundwater, and as the bubbles migrate upward
they can carry NAPL with them at the interface of
the bubble and water (Figure 12). This basic
principle is used in Dissolved Air Floatation, a well-
known process which is used to separate solids and
nonaqueous phase liquids from water (Metcalf &
Eddy, Inc, 1972). This process has been
demonstrated at ambient temperatures when an air
bubble is present at the interface of a volatile
compound and water in open water, in capillary
tubes, and in porous media. When the liquids are
heated, it has been observed that PCE and water
vapors generated at the interface of the water and
liquid-phase PCE in a 0.5 millimeter capillary tube
had liquid PCE attached to the bubble, and the
combined PCE liquid and vapor migrated slowly
upward (Udell, 2006).
19
-------
Figure 12. In this laboratory experiment chlorinated
solvent DNAPL (dyed red) was emplaced in a two
dimensional model containing glass beads and air-free
water. Once an air bubble was introduced into the model,
it was observed to grow in size as the DNAPL volatilized
into it. When the bubble was large enough to overcome
the capillary pressure, part of the bubble broke off and
migrated upward in the porous media, forming the
residual DNAPL blobs shown (labeled a in the photo).
DNAPL was also dragged upward as a thin film at the
interface between the bubble and water as shown in the
inset picture (Roy and Smith. 2006)
Pakkala (2007) demonstrated that trichloroethene
can attach to air bubbles and migrate upward in a
packed bed of glass beads 2 millimeters diameter.
The bubbles can be formed by introducing air at the
bottom of the packed bed or by heating the water
and causing the liberation of air and other dissolved
gases. Insoluble compounds such as DDT have been
shown to be carried upward when dissolved in a
solvent such as chlorobenzene that has
accumulated at the air-water interface of bubbles
(Valsaraj et al., 1986). This type of 'bubble
floatation' was believed to contribute to the
coalescing of fuel oil in the subsurface during an
ERH remediation, allowing more of the liquid fuel to
be recovered (Beyke and Fleming, 2002), and likely
helps to account for the lack of downward
migration by DNAPL during heating.
Increased rates of biotic and/or abiotic reactions
may also be observed for some contaminants. The
dechlorination daughter products of chlorinated
solvents may be observed during or after thermal
remediation. Hydrolysis rates can increase
exponentially as the temperature increases
(Washington, 1995). Hydrolysis of chlorinated
ethanes 1,1,1-trichloroethane or 1,1,2-
trichloroethane will form 1,1-dichloroethene (1,1-
DCE). 1,1-DCE does not hydrolyze further, and is
more toxic than the ethanes from which it is
formed. 1,1-DCE is also more volatile than the
parent compounds, will readily volatilize from
water, and can readily be recovered during thermal
remediation.
Carefully performed and controlled laboratory
experiments have shown that chlorinated
compounds such as TCE or PCE do not readily break
down at temperatures used for their recovery via
thermal remediation (Costanza et al., 2007), unless
an appropriate form of iron is present to aid in
abiotic degradation processes. However, laboratory
experiments have shown that when polynuclear
aromatic hydrocarbons (PAHs) are treated by
thermal remediation, oxygenated and/or
hydroxylated PAHs may be formed (Trine et al.,
2019).
During TCH, temperatures near the heater casings
can be as high as 400 to 800°C. This creates a zone
immediately surrounding the heater wells where
oxidation and pyrolysis may occur. Systems can be
designed to minimize contaminant presence in the
vicinity of the heaters, thus avoiding heated vacuum
wells. If the heating and extraction wells are co-
located, contaminants are pulled to the very high
temperature zones, and some chemical degradation
would be expected. In the case of halogenated
contaminants, some acid production can occur,
leading to corrosion of wells and piping.
Section 4.0 Defining the Area to
be Treated by Thermal
Remediation
Appropriately defining the area to be treated by
thermal remediation is essential to the overall
success of the project, and in making the remedial
action cost effective. These aggressive, more costly
technologies are generally applied to source areas,
defined as the areas containing appreciable NAPL,
although there are times when these technologies
have been applied to lower levels of contamination,
particularly in low permeability soils and/or below
the water table. The cost of applying these
20
-------
technologies is generally proportional to the size of
the area and volume that must be treated, so areas
that are only lightly contaminated generally are not
included in the treatment zone. Examples of
exceptions to that are when high dissolved phase
concentrations are in low permeability soils that are
not amenable to other types of remediation such as
pump-and-treat, when a rapid and complete
remediation is required in order to transfer the
property or for reuse of the site, or any other time
that a rapid remediation is preferred. Multiple lines
of evidence of NAPL, including soil borings, are
generally the best way to determine where NAPL is
present in the soils, while groundwater data
provides an additional line of evidence as to
whether NAPL is present or not. The extent of the
NAPL should be fully defined with soil borings
exhibiting only low concentrations - or
concentrations below the cleanup criteria - just
outside of the delineated treatment area.
NAPL presence in wells is an obvious indication of
mobile NAPL in the vicinity, but should not be
interpreted as the full extent of NAPL. Thermal
treatment zones must extend outside of the area
where NAPL exists in wells. For DNAPLs, wells that
do not contain sumps may not be adequate for the
detection of mobile DNAPL, as it may migrate
through the well at times when the well is not being
gauged for product. Where DNAPL is suspected or
known to be present, monitoring wells should be
constructed with sumps to allow detection of
DNAPL migrating into the well.
At sites where a high concentration dissolved phase
plume exists adjacent to the NAPL contaminated
area, it may be appropriate to extend the thermal
treatment area to include this area. This may be
especially true where the high concentration
dissolved phase plume is in low permeability soils,
which may be difficult to treat using pump-and-
treat or where it would be difficult to inject
treatment fluids.
Section 4.1. Characterization
Techniques to Determine NAPL
Presence. While it is not the purpose of this
paper to discuss characterization techniques in
detail, some discussion of NAPL detection
techniques and potential problems with
characterization efforts is provided based on
experiences in the field. It has been found that not
all characterization tools are equally effective in all
hydrogeologic settings and/or when searching for
different types of NAPLs (i.e., chlorinated solvents
versus coal-tar based products). Field work for
characterization is often delegated to the least
experienced environmental professionals.
However, this does not mean that this work is not
critical to the overall remedial efforts. Proper
sampling procedures for soil and groundwater
sampling must be followed, and proper
documentation must be provided of the work that
was performed, including boring logs, well sampling
logs, and groundwater gauging reports. It must be
ensured that the tools being used for
characterization are capable of detecting the
contaminants that are believed to be present. The
Triad approach (EPA, 2004) is recommended to
allow the investigation to be driven by the results
that are being found as the investigation proceeds.
Briefly, with a Triad approach, the investigation
would start where it is known that there is NAPL, to
verify that the tool being used can detect the
contaminant(s) of concern. Then the investigation
continues by stepping out in all directions from
there until the extent of contamination is
determined. Generally 40 to 50 foot spacings are
used for characterization locations for moderate to
large sites, smaller sites may use closer spacings.
When little is known about the location of NAPL on
a site, it may be appropriate to obtain screening
data using a membrane interface probe (MIP) or
laser induced fluorescence (LIF), as these tools
provide continuous inference of contaminant
presence and relative permeability of the soils.
However, this screening data alone should not be
relied on to determine if NAPL is present.
Correlating MIP response to soil concentrations has
been found to be problematic, thus the instrument
response must be verified by comparison with soil
samples (Myers et al., 2002; Mumford et al., 2022).
One aspect of the problem is that the PID detectors
employed in MIPs have a maximum value which
does not allow it to distinguish between high
21
-------
aqueous concentrations and the presence of NAPL
(Mumford et al., 2022).
Some NAPLs such as creosote or coal tar may be
readily visible as a dark liquid or sheen in soil cores,
and the NAPL-contaminated area may be defined
based on visual observation of NAPL in the soil core.
LIF can be a powerful screening tool for screening
soils for the presence of PAHs such as creosote and
coal tar. The fact that these NAPLs have a low
interfacial tension and thus can penetrate a wider
range of pore spaces, leading to greater saturations
and areas where greater pools area formed, may
make them more likely to be detected by an LIF
instrument. Chlorinated solvent NAPLs such as TCE
or fuels such as gasoline are generally much more
difficult to see visually in soils, and the neat liquids
do not fluoresce, and due to their higher interfacial
tension they may have a more heterogeneous
distribution in the soil. DyeLIF is a relatively new
method for chlorinated solvent NAPL investigations,
but successful use of this screening tool relies on
additional lines of evidence of NAPL occurrence,
generally comprised of benchtop testing of the
DyeLIF response to the NAPL, PID readings, and
analytical samples (Einarson et al., 2018; Dakota
Technologies, 2022).
A photoionization detector (PID) or flame ionization
detector (FID) should be employed to screen the
length of the soil core at closely spaced intervals to
determine where the most contaminants are
located within the soil column. Soil samples should
be obtained by the appropriate method
approximately every five to ten feet along the soil
core where the PID/FID shows the maximum
concentrations may reside, and submitted to the
laboratory for analysis. When the contaminant is a
volatile compound, care should be taken to scan the
core quickly, then open the core to the center to
quickly obtain the analytical sample in order to try
to minimize vaporization loses. An additional line of
evidence of the presence of chlorinated compound
NAPLs may be obtained by inserting a piece of
FLUTe ribbon in the soil core; a reaction can be seen
on the ribbon when DNAPL is present. However, the
reaction between the NAPL and the ribbon only
occurs when there is direct contact between them,
as the ribbon will not wick the NAPL. Soils can be
screened for NAPL using oil red dyes, however, for
chlorinated solvents such as PCE false negative
results may be obtained (EPA, 2004).
Soil gas samples may also be used for screening,
however, soil gases may migrate much more readily
than the liquid contaminant, spreading a significant
distance from the NAPL. Additionally, the soil gas
may linger after the NAPL has vaporized. Soil gas
data must be confirmed by soil or groundwater
data. Direct push technology (DPT) may be
applicable at some sites for obtaining soil cores,
however, the small diameter of the core may mask
the presence of NAPL that can be detected in larger
diameter cores as can be obtained using rotosonic
drilling. For some soils, the recovery using DPT may
be poor. When the contaminant is a DNAPL, DPT
may meet refusal before the vertical extent of the
DNAPL has been defined. This is particularly true
when the DNAPL has reached a zone of weathered
bedrock and migrated into the bedrock itself.
Rotosonic drilling can produce heat which
potentially can vaporize VOC contaminants,
however, the larger diameter cores produced by
this drilling method may be better in many soils for
detecting the presence on NAPL. The cores can be
halved lengthwise to expose a fresh surface for
screening with a PID. The other half of the core can
be covered to reduce vaporization of contaminants
until the screening is completed, then soil samples
can be collected when the PID response was the
greatest. Generally rotosonic drilling includes an
outer casing from the core barrel, so if DNAPL is
drilled through it cannot migrate down the
borehole. When drilling through DNAPL or through
a low permeability zone into more permeable soils,
a bentonite plug can be set in the casing to ensure
that DNAPL does not migrate downward through
the borehole. Polypropyline liners within the sonic
core barrel are not recommended.
The continued discharge of highly viscous, low
density coal tar and creosote NAPLs to surface
water demonstrates that these NAPLs can continue
to migrate in the subsurface for more than a
hundred years (Gerhard et al., 2007). However, a
commonly held belief is that low viscosity NAPLs
22
-------
released to the ground surface do riot continue to
migrate over long time periods, but will reach
residual saturations rather quickly. This is thought
to be especially true for less viscous and more
dense chlorinated solvent DNAPLs. The assumption
seems to be that the NAPL will not have sufficient
head to continue migrating long term, but will pool
on top of low permeability zones and not migrate
further. In contrast to this belief, numerical
modelling has suggested that these types of NAPLs
can continue to migrate vertically in heterogeneous
porous media for decades (Reynolds and Kueper,
2004). Observations at a couple sites where large
amounts of NAPL were discharged to the subsurface
seem to support these modeling results. Examples
of continued migration of NAPLs include the
appearance of chlorobenzene and DDT DNAPL in a
well at least 300 feet from the known source zone
40 years after the closure of the DDT manufacturing
facility (Figure 13), and jet fuel migration over a 10
year period between when the characterization
work was done and when the full scale SEE
remediation system was constructed (Figure 14).
Thus, the age of the characterization data used to
delineate the NAPL-contaminated are to determine
the thermal treatment area must be considered.
Also, it is possible that where there are NAPL
saturations greater than residual saturation,
investigation activities (i.e., drilling, open boreholes)
or groundwater pumping can cause or allow NAPLs
Lr ' -itrmm-w-k-.
Figure 13. This Superfund site is a former DDT manufacturing facility which ceased operating in 1982. An investigation to
determine the extent of DNAPL was carried out in 2005. The green line delineates the area below the water table where
DNAPL was found within an aquitard. The pink areas show where mobile DNAPL was found to exist based on the
occurrence of DNAPL in wells and/or soil concentrations above 53,000 mg/kg. In 2017. DNAPL was first detected in the
well shown by the blue and white circle approximately 200 feet to the north of the previously-defined DNAPL contaminated
area, and DNAPL continues to be periodically recovered from this well, demonstrating the continued migration of DNAPL
for years after the discharge occurred.
23
-------
Figure 14. At this former Air Force Base, a large jet fuel leak migrated to depths as great as 235
feet bgs. As the water table subsequently rose, jet fuel was trapped below low permeability zones.
More than 30 years after the leaking was terminated, jet fuel continues to migrate to the southeast
where it is periodically collected from wells. Two and a half million pounds of jet fuel were
recovered using SEE (the SEE treatment area is shown in light blue). The dark blue areas continue
to contain mobile LNAPL that is collected routinely from wells that were installed after the
termination of SEE.
Legend
Additional Characterization Wdl Location
EBR WeB Location
SEE Remediation Well Location
Grouidwaler Monrtonno Well Location
Additional Characterization Soil Boring
Location
Mass Extent Atlrtiuled to Additional
Characterization
Model Extent of Residual LNAPL
Thermal Treatment Zone
Thermal Influence Zone
Radius of influence Zone
to migrate. Sometimes early investigation activities
did not totally delineate the extent of NAPL (Horst
et al., 2021). If significant time has elapsed since
the NAPL delineation was carried out, additional
delineation can be carried out during the
installation of the thermal remediation system to
ensure that the treatment area contains essentially
all of the NAPL.
Section 5.0 Treatability Studies
According to EPA guidance (EPA, 1992), laboratory
scale treatability studies can be conducted during
the remedial investigation/feasibility study stage to
indicate whether a given technology can meet the
expected cleanup goals, while field scale treatability
studies can be performed during the remedial
design/remedial action stage to establish the design
and operating parameters for a sound, cost
effective implementation of the remedy. The Guide
recommends a three-tiered approach, the first of
which is bench-scale tests to determine
qualitatively is a certain technology can achieve the
performance goals. The second tier is remedy
selection testing which can be done at either (or
both) the laboratory and/or field scale to "provide
quantitative data for use in determining whether a
technology can meet the operable unit's cleanup
criteria and at what cost." The third tier of RD/RA
testing is performed by the technology vendors to
pre-qualify them, or to support the detailed design
specifications.
While laboratory and field testing may be very
important and informative for some situations, in
situ thermal technologies have been used
24
-------
extensively enough in a wide variety of geologic and
hydrogeologic conditions (Horst et al., 2021) that
they can be efficiently applied to many common
sites without prior laboratory and/or field scale
testing. When laboratory or field testing is to be
performed, the testing should be carefully designed
in order to obtain valid data. Problems with non-
representative conditions during laboratory testing
can start with the normal procedures of
homogenizing soil samples brought in from the field
for testing. When the contaminants are volatile,
much of the contamination may be lost during the
sampling and homogenizing process, thus the soils
tested may not represent the 'worst case' or even
common contaminant concentrations. In some
cases, enough of the contaminants can be lost
during the sampling and homogenization process
that the pre-test soils already meet the soil cleanup
criteria. For thermal remediation, there can be
tradeoffs between treatment time and temperature
to reach cleanup goals - lower temperatures may
meet the same treatment goal if a longer treatment
time is allowed. Testing time at temperature in the
laboratory is generally on the order of a couple
hours to a couple days, while treatment times in the
field are generally several weeks or more at
temperature. Thus, laboratory testing may not
reveal the optimum combination of treatment time
and temperature for meeting the treatment goals.
One approach to better correlate laboratory testing
with field conditions is on the basis of energy
density, which translates into the amount of pore
volumes of steam generated and removed.
Section 5.1. Laboratory Testing. For
thermal remediation, bench scale tests that would
fall into the category of tier one testing have been
performed by heating a contaminated soil sample in
an oven to determine if the contaminants can be
volatilized at the thermal treatment temperature,
or conversely to determine the treatment
temperature required to meet the remedial goals.
Steam injection treatability tests have also been
performed in the laboratory that would be
considered tier one testing. In general, tier one
bench scale treatability testing is only
recommended for semivolatile contaminants, and
the purpose would be to determine the
temperature that is needed to effectively volatilize
the contaminants. When use of the STAR
technology is being evaluated for the remediation
of SVOCs, one dimensional column studies are used
to determine if the contaminants can sustain in situ
combustion (Grant et al., 2016). For VOCs,
adequate experience has been obtained during field
scale remediations that thermal remediation
systems can be designed with confidence without
site specific treatability studies.
When steam injection testing using a one
dimensional column is performed, it must be
ensured that a steam front is formed in the column.
The formation of a steam zone is dependent on the
pressure and thus the temperature of the steam
being high enough to overcome the back pressure
created by the permeability of the soils - the lower
the permeability of the soils, the higher the
temperature and pressure must be to form a steam
front. The steam injection should be performed
with the steam injected at the top of the column
and with the effluent recovered at the bottom. One
reliable way to demonstrate that a steam front has
been formed within the column is to track the
amount of effluent condensed and collected and to
compare that to the amount of steam injected on a
condensed basis (Figure 15). Steam has a volume
approximately 1500 times that of liquid water.
When a steam front forms in the column, it will
displace in front of it the water from the pores of
the soil sample, and more effluent volume will be
collected than the amount of water injected as
steam. If the amount of effluent liquid collected is
the same as the volume of water injected as steam,
a hot water flood was conducted, not a steam flood.
One dimensional column studies in the laboratory
have demonstrated that hot water flooding is much
less effective at recovering contaminants than
steam. Thus, if a laboratory steam injection is not
performed correctly, erroneous conclusions may be
drawn from the laboratory experiment.
25
-------
Steam injected (condensed) versus effluent collected for the steam injection experiments.
800
100
200
300 400 500
Mass Injected (gms)
600
700
800
Figure 15. This graph displays data from one dimensional column steam injection treatability tests. The formation of a
steam front in the column is indicated for the column experiments where the mass of effluent collected exceeds the mass
injected. Steam has a volume approximately 1500 times that of water, thus the formation of a steam front within the column
displaces in front of it the water from the pore spaces of the soil, and more water is collected in the effluent than steam
injected (considered as condensed to water). If the pressure in the column is sufficient to condense the steam to hot water, the
mass of effluent collected will be essentially the same as the mass of steam injected, as shown for column #2 and #3, which
track the line showing the mass of water injected as being the same as the mass extracted. Steam injection was shown to
recover much more of the contaminant mass than hot water flooding.
The results of one dimensional steam injection
column experiments - even when a steam front is
formed in the column - cannot be directly
extrapolated to the field in terms of the amount of
steam injection required to reach a remedial
endpoint. In the field, steam injected into a well
will flow radially from the well. Thus, as the steam
front expands from the well, the radius gets larger
but the amount of steam reaching a given area is
less. The area closer to the steam injection well will
receive more treatment than the areas further
away.
Section 5.2. Field Testing. The epa guidance
states, "Field testing, however, is important for an
adequate evaluation of in situ treatment. Because
of the unique difficulties associated with simulating
in situ conditions and monitoring effectiveness of in
situ treatment in the laboratory, field testing often
may be the only way to obtain the critical
information needed for the detailed analysis of
alternatives . ..". Tier two/three testing for in situ
thermal technologies would generally consist of a
field test on a portion of the area to be remediated
in order to determine the amount of energy that is
required to reach the target treatment
temperature. This test might also be used to
confirm that the contaminants can be effectively
recovered at the target temperature. Valuable
information may be obtained from these tests to
determine the required electrode or heater well
spacing and vacuum extraction requirements.
However, due to the high costs associated with field
scale pilot testing, it only recommended when the
26
-------
field conditions are somehow unusual from the
more commonly treated sites. Experience obtained
at a large number of field sites and the use of
models to predict heating rates as a function of
energy input and the configuration of the site
allows the thermal remediation vendors to design
and implement these remedies at most sites
successfully without obtaining site specific field data
from a pilot.
Whether or not a pilot scale is recommended for a
particular site can be discussed with the vendors of
the technologies. Examples of when field testing
may be recommended include when the
contaminants at a particular site have not been
treated previously using thermal technologies, for
contaminants such as PFAS for which all the
removal mechanisms are not understood, sites with
high permeability water bearing zones where it
must be determined if steam injection or
groundwater extraction is required, and sites where
the possibility of subsidence must be evaluated.
For unusual contaminants that have not been
treated by thermal remediation previously,
laboratory and/or field testing may be
recommended before thermal remediation is
chosen as the remedy. Often the questions about
treatment of exotic contaminants can be answered
by less-costly laboratory tests rather than field
tests. For design questions about the need for
steam injection or groundwater extraction , the
field scale test may occur after thermal remediation
has been chosen as the remedy. In some cases, a
field test is desired by the site owner as 'proof of
concept' before investing in a full scale remediation
at a large site.
When steam injection is the preferred technology,
steam injection rate as a function of injection
pressure of the steam as well as steam migration
through the formation is usually evaluated through
field scale treatability tests. This type of data can be
very useful for designing SEE remediation systems,
especially in setting such as fractured rock, but
steam injection pilot tests have also been used to
determine the extent of steam buoyancy in highly
permeable sand and gravel aquifers. Generally one
or two steam injection wells are constructed with
several thermocouple strings surrounding them at
varying distances from the injection wells. The use
of temperature monitoring points surrounding the
injection wells allows the vendors to compare
results with the radius of influence (ROI) and steam
shape obtained from steam modeling and back-
calculate the anisotropy ratio, which is then used
for scale-up modeling. Steam may exhibit very
irregular flow patterns in the subsurface depending
on the distribution of hydraulic conductivity.
Temperature is the best means of determining
where the steam will flow spatially and vertically.
Steam injection rates as a function of injection
pressure is also critical to determine the well
spacing and expected breakthrough time of the
steam at extraction wells. These short term steam
injection tests, with a typical duration between 2
and 7 days, can be performed without simultaneous
groundwater extraction. The wells and TMPs used
for the pilot test are then incorporated into the full
scale design.
When field treatability testing is desired, often the
approach proposed by the site owner is to conduct
the pilot in the middle of the most contaminated
area of the site. This can be problematic for several
reasons. Performing a pilot in the most
contaminated area can demonstrate that the
thermal technology can recover the contaminants
effectively, but it cannot demonstrate that soil
cleanup criteria can be met because contamination
from outside the treatment area will continuously
be pulled into the pilot area if hydraulic and
pneumatic control of the pilot area are maintained.
If hydraulic and pneumatic control are not
maintained, contaminants may be spread to
previously uncontaminated areas. Then, due to
recontamination, the pilot area would have to be
re-treated during the full scale remediation.
Another potential problem is the recovery of
significant quantities of contaminants that must be
treated at the surface.
It is recommended that when a field scale pilot
treatability study is to be performed, an upgradient
area of the contaminated site be chosen to lessen
the likelihood of recontamination and enhance the
probability of reaching cleanup criteria during the
27
-------
pilot. Starting in an area with less contaminant
mass increases the likelihood of being able to
successfully treat the contaminants that are
recovered aboveground, while learning more about
how to treat the recovered wastes. This will help in
designing and sizing the treatment system for the
full-scale remediation.
The size of a field scale treatability study should also
be carefully considered along with the objectives of
the pilot. For example, the in situ thermal
technologies generally rely on superposition of heat
or energy, meaning that heat reaches an area from
all sides. Thus, if the pilot scale is to determine if
the site can be heated to the desired temperature,
or how much energy is required to heat the area,
the pilot scale should be large enough that multiple
heat injection points are used to simulate what
would occur at full scale. Thus, thermal technology
pilot scale implementations that rely on one
hexagonal array of the wells providing energy to the
subsurface, whether they are electrodes, heater
wells, or steam injection wells, may not provide
data that can be directly scaled up to the full scale
implementation.
Section 6.0 Thermal Remediation
Services and the Superfund
Process
Successful implementation of thermal remediation
technologies requires substantial expertise in the
design and operation of these technologies as well
as specially designed equipment. Several specialty
vendors offer thermal remediation services, and
most now offer more than one thermal technology.
This section discusses a general procedure for
procuring thermal remediation services through the
Superfund process, but most of the discussion will
also apply to sites under other regulatory programs
and even sites that are not regulated.
When thermal technologies are being evaluated for
a particular NAPL-contaminated site, the evaluation
process and decision documents, such as a
Feasibility Study (FS) and Record of Decision (ROD)
or Engineering Evaluation/Cost Analysis (EE/CA),
best practice is for the document(s) to specify
thermal remediation in general but not a particular
thermal technology. There is some overlap in
applicability of the thermal technologies,
particularly for low permeability soils where both
ERH and TCH are both generally applicable. When a
particular thermal technology is not specified, the
thermal vendors can then propose during the
bidding process the technology they believe is best
suited for a particular hydrogeologic setting,
contaminant, and remedial goal. This allows the
vendors to put forward their best technical proposal
for the site. This will also allow more of the vendors
to bid on the project if they believe their technology
is applicable.
The contract for the thermal vendor must be a
design/build/operate contract. While there is more
than one vendor for each technology type, each
vendor applies the technology somewhat
differently, and they have their own specialty
equipment designed for their process. The vendors
generally will not bid on a design produced by
others, or if they do, the design will be redone. A
thermal vendor will not generally guarantee that
performance standards such as temperature goals
or soil cleanup standards can be met when working
with a design they did not develop themselves.
A basis of design (BOD) should be prepared that is
part of the package that will be used to solicit bids
from the thermal vendors for the
design/build/operate contract. The BOD must not
include a conceptual design - as stated previously,
each of the vendors has their own method of
designing and operating their technology, and they
will not generally bid on a design produced by
others. The BOD should contain a concise but
complete conceptual site model (CSM) for the site,
including a description of the site geology and
hydrogeology, the contaminant(s), their distribution
in the subsurface, and their concentrations in soil
and groundwater along with a breakdown and best
estimate of the total contaminant mass. Plane
views and cross sections containing all of this
information are a good way to present this
information in a readily understandable way. Three
dimensional models or presentations are less
valuable for bidding and design purposes as they
28
-------
may leave openings for interpretation of the
footprint and treatment depth of the target
treatment area, and as a result may make it difficult
to compare the bids that are prepared by different
vendors. Surface structures and subsurface
infrastructure within and immediately surrounding
the area to be treated should be identified on
figures. Access to and size of electrical, water,
sewer, and natural gas utilities should be discussed.
It should be noted that the large but temporary
power demands of these technologies, electrical
and/or natural gas (or possibly propane),
particularly for larger sites, may be problematic.
Electrical power requirements may have a long lead
time to get the power drop. Discussions with the
utility companies should be initiated even as
thermal remediation is evaluated for its applicability
for a given site to determine the availability of the
power required for the remediation.
Section 6.1. Remedial Goals. The bod
should clearly define the remedial goals. For
Superfund sites, Remedial Action Objectives (RAOs)
are defined by the ROD, and generally include
statements of protection of human health and
restoration of groundwater resources. At sites
where NAPL is present, particularly for chlorinated
solvents, there is usually a downgradient dissolved
phase plume that is not being addressed by the
thermal remediation. A remedial goal should be
defined to specifically state what the thermal
remediation is intended to accomplish. Where soil
remedial goals have been established which are
meant to protect groundwater by eliminating or
reducing leaching from soils, these soil cleanup
criteria may be appropriate endpoints for the
thermal remediation within the thermal treatment
zone. Soil concentration goals are to date the most
common hard criteria for thermal remediation
completion. However, if these criteria are
extremely low, as in the tens of parts per billion
range, long thermal treatment times may be
required due to slow desorption rates, and it may
not be cost effective to continue heating until these
very low concentrations are achieved.
For most Superfund sites, restoration of
groundwater is one of the ultimate goals of the
remedial activities. Maximum concentration levels
(MCLs) for chlorinated solvents such as PCE and TCE
or volatile petroleum hydrocarbons such as
benzene are very low (5 micrograms per liter, ug/l),
which is five to six orders of magnitude less than
the solubility limit for the compounds. Reaching
these concentrations in a groundwater system that
has been contaminated by NAPL for a long period of
time, which allows adsorption onto the soil particles
and diffusion into low permeability soil strata, will
likely require extended treatment times due to the
slow desorption process. Also, experience at
thermal remediation sites has shown that
groundwater concentrations will vary significantly
during thermal remediation both spatially and over
time, making it difficult to use a groundwater
concentration as a remedial objective.
Percent mass recovery or a percent reduction in soil
and groundwater concentrations have also been
used or proposed as remedial goals for thermal
remediation. The heterogeneity of NAPL
distribution in the subsurface and our limited
abilities to determine its actual distribution and
saturation within the soil pores even with our most
advanced and/or most prolific sampling procedures
frustrates attempts to calculate the mass of
contaminants in the subsurface. Thus, clean up
criteria tied to a percent mass recovery or a percent
reduction in soil concentrations can be extremely
difficult to verify and document.
For many sites, particularly larger sites, a more cost
effective approach may be to terminate the thermal
remediation system when the system reaches
'diminishing returns'. This remedial endpoint is
especially appropriate if the remedial objective is to
remove contaminant mass to reduce the
contaminant mass flow into the downgradient
plume. To reach 'diminishing returns', first a site
must reach the target temperature throughout the
treatment area. For chlorinated solvents or light
petroleum hydrocarbons, such as BTEX compounds,
the target temperature should be the boiling point
of water at the local pressure condition. As the site
heats up, the mass recovery rate is monitored, and
29
-------
an increase in the recovery rate is generally noted.
When the site temperature exceeds the co-boil
point for the NAPL, the NAPL has been eliminated,
and mass recovery rates will start to decrease.
Groundwater samples analyzed during the heatup
phase would be expected to increase due to
enhanced dissolution rates during heating, and then
decrease when the NAPL is eliminated and only
dissolved and adsorbed phases remain. Once
recovery rates have reached a low rate, and
groundwater concentrations are well below
concentrations that would be indicative of NAPL
presence, it may be most cost effective to terminate
the thermal remediation at that time and to reduce
groundwater concentrations further by
groundwater extraction. Extracting the heated
groundwater will recover contaminant mass that is
desorbing from the soil particles, and aid in
reaching MCLs. Pump-and-treat systems in general
are effective for recovery of contaminants that are
in the dissolved phase, and the process will likely be
enhanced by higher temperatures.
For SVOCs, including PAHs, PCBs, and dioxins, the
target temperature may be 300°C or greater.
Temperatures in excess of the boiling point of water
can only be achieved by the TCH technology or by
STAR, the in situ combustion technology. For TCH,
these higher temperatures can only be achieved
when groundwater is not present. Table 2 shows
pre- and post-treatment PAH concentrations for a
coal tar site treated by TCH at temperatures of
325°C. At the Mare Island Naval Shipyard
Superfund site, TCH reduced PCB concentrations by
as much as four orders of magnitude, to
concentrations less than 0.033 milligrams per
kilogram (mg/kg).
Table 2. Pre- and Post-Treatment Soil Concentrations
of Coal Tar Components Within the Construction
Worker Exposure Depth
Sampling Depth: 6- 14' Average Concentrations
... ^ Pre-Treatment Post-Treatment
Constituent mg/kg mg/kg Reduct.on%
Section 6.2. What to Look for in a
Remedial Action Work Plan/Remedial
Design. Normally during the remedial design (RD)
phase, several rounds of review by the regulatory
agency are conducted. These reviews should
inform the responsible party or regulatory agency
of the details of the design and operation of the
thermal remediation system, however, it should not
be viewed as opportunity to tell the vendor how to
design or operate their system - the vendors are
the experts on these systems. Attempts to dictate
to them the design or operation of the system may
result in liability on the part of the responsible party
or regulatory agency. However, what should be of
particular interest to the site owner and regulatory
agency is the monitoring system. Monitoring
should include performance monitoring to show
progress toward meeting the remedial goals, and
operational monitoring to show that the above
ground treatment system is operating as designed
and is meeting the discharge criteria for air and
water (Section 2.8). At sites where the community
is close by, or near occupied facilities, ambient air
monitoring and/or indoor air monitoring may also
be recommended.
Performance monitoring to show progress toward
meeting the remedial objectives includes
subsurface temperature monitoring, monitoring to
show that pneumatic and hydraulic control are
being maintained, and contaminant extraction rates
overtime. Most thermal remediations will also
Benzene 2068 0.35 99.98%
Anthracene 19 0.48 97.47%
Benzo(a)anthracene 20 0,51 97,45%
Benzo(a)pyrene 20 0.33 98.35%
Chrysene 20 0.71 96.45%
Fluoranthene 43 1.02 97.63%
Naphthalene 679 5.7 99.16%
Phenanthrene 107 3.82 96.43%
Pyrene 65 1.12 98.28%
C11-C22 Aromatics,
unadj.
4000 43.15 98.92%
30
-------
include groundwater concentrations over time to
track remedial progress. Interim soil samples may
also be used to monitor the progress of the system,
especially when the target treatment zone is above
the water table and the cleanup criteria are soil
concentrations. Subsurface temperatures are
monitored automatically on a continuous basis by
thermocouples or by other temperature measuring
techniques such as fiberoptic temperature
monitoring systems that are installed in the
subsurface in vertical borings with the temperature
sensor spaced vertically every 3 to 5 feet. These
temperature monitoring points are often co-located
with piezometers to measure subsurface vacuum to
demonstrate pneumatic control. Weekly progress
reports from the vendor should include figures
showing the current subsurface temperatures
spatially and vertically. Demonstrations of
pneumatic and hydraulic control are best achieved
by including temperature and pressure monitoring
points around the thermal treatment area, as well
as within the treatment area. While some
temperature increase is expected outside of the
thermal treatment area (due to heat conduction
from ERH and TCH systems or radial injection of
steam during SEE remediation), higher rates of
temperature increase than expected are a reliable
indication of the loss of hydraulic control (Figure 7).
These vacuum and temperature measurements
should be made and reported on a weekly basis.
How many monitoring points are installed will
depend on the system size, but a general guideline
may be one per every 1000 to 2000 square feet.
The heterogeneity of the system and the cost of
installing monitoring points should also be
considered when determining the frequency of the
temperature monitoring points. Where there are
sensitive areas adjacent to the treatment area,
additional monitoring may be warranted.
Contaminants extracted in the vapor, liquid
(dissolved) and NAPL phase should be measured
and reported at least on a weekly basis. Commonly,
PID readings are taken of the vapors as they enter
the final treatment stage (thermal oxidizer or vapor
phase granular activated carbon) on a daily basis.
Summa canister samples for laboratory analysis are
generally obtained on a weekly basis. Water
entering the final treatment phase (commonly
liquid phase granular activated carbon) is generally
sampled weekly as well. NAPL accumulation is also
measured weekly to determine the total
contaminant recovery rate and the cumulative
contaminants recovered. When the NAPL is
eliminated, this should be reflected in overall
decreases in groundwater concentrations, although
there may be significant variations in groundwater
concentrations even during this stage of the
remediation. When the concentrations remain well
below the concentrations that are indicative of
NAPL presence and the mass extraction rates are
low, diminishing returns have been achieved and
the application of heat can be terminated. As
groundwater concentrations and mass recovery
rates decrease, it may also be desirable to obtain
soil samples for analysis. Note that groundwater
and soil samples obtained during and after thermal
remediation need to follow hot groundwater and
soil sampling standard operating procedures (SOPs)
that are available from the thermal vendors.
Section 6.3. How Does Thermal
Remediation Affect the Community?
Thermal remediation technologies are frequently
implemented in commercial, industrial and
residential areas (Figure 16).
Figure 16. Thermal remediation technologies have been
used in residential, commercial, and industrial setting. In
most cases, the remediation has been able to proceed
without displacing the residents.
In a few cases, sensitive possible receptors have
been relocated during construction and operation
of the thermal remediation, but in many cases the
thermal system design, construction and operation
31
-------
can be conducted in a manner to reduce the impact
to the residents or business to accommodate them
remaining in their residence or business. Design
elements to reduce the impact to existing
infrastructure include angled or horizontal
boreholes under buildings when the building cannot
be accessed by drill rigs to install borings vertically
(Figure 17), and subsurface completions of wells to
allow safe access to the area. Construction
schedules can be altered to accommodate
businesses, for example by doing the construction
at night when the business is not open. In contrast,
in a residential area, construction activities may be
restricted to normal business hours so that
residents experience less disruption in the early
morning and evenings when they are more likely to
be home. Constraints on construction and where
access to the public must be maintained must be
detailed in project bid documents.
While construction can be limited to specific times
of the day or week, operation of thermal
remediation systems is normally continuous, and
some of the components, such as blowers, can
produce considerable noise. Noise-producing
equipment can be placed away from the property
boundary and/or within structures, or sound
barriers can be built around them to reduce the
noise level at the property boundary. Monitoring of
noise continuously or on an as needed basis can
also be included in the project remedial action work
plan (RAWP).
Ambient and/or indoor air monitoring is usually
warranted when the public will have access to areas
adjacent to or within buildings above the
remediation area. Continuous monitoring via PID or
FID readings at the perimeter of the property
boundary is commonly used when a residential area
is adjacent to the site, however, these readings do
not distinguish between the different types of VOCs
that may be present and the differences in their
toxicity. Summa canister samples of air can
determine what specific chemicals are present and
can be used periodically to support the PID or FID
data, but these samples are generally short term
and thus not representative of the long term air
quality. Also the turnaround time on the analysis
can be as much as three weeks. Absorbent medium
samplers which average the concentrations over
the time that they are deployed can be used for air
samples to provide data on the specific compounds
that are present, but again, this is not real time
data. When there is significant risk of exposure in
buildings that overlay or are adjacent to the
treatment area, real time data can be obtained by
an automated gas chromatograph system. This
system can be set up to obtain air samples from a
variety of locations on a rotating basis and provide
essentially real time data on VOCs concentrations in
air in sensitive areas, The system can also provide
notifications when concentrations of specific
compounds exceed a threshold (Kram et al., 2019).
0) TERRA SONIC
Figure 17. Thermal remediation systems have been installed vertically in warehouses where there is access for a drill
rig and under active manufacturing buildings using angled borings.
32
-------
Some communities may object to having a thermal
oxidizer or catalytic oxidizer on site during
remediation to destroy the vaporized contaminants
as they are recovered from the subsurface. Vapor
phase activated carbon (VGAC) can be used on the
back end of these systems to help alleviate the
concerns. Destruction of the vapors on site may be
a more environmentally safe option for the
community than to have the vapors condensed to a
liquid hazardous waste that is stored on site until
being are transported by truck through the
community to an off site disposal facility.
Section 7.0 How Do We Know
When We are Done?
When to turn off the thermal treatment system - or
at least when to terminate heating - is a commonly
asked question. Regardless of the objectives of the
thermal remediation, there can be benefits to
continuing the remediation until 'diminishing
returns' have been met. When the objective of the
remediation is mass recovery to the extent
practical, 'diminishing returns' would be the best
indication that the remediation has reached that
point. The first criteria that must be met is that the
target temperature should be met throughout the
treatment area (Heron et al., 2006). The design
energy input for the system should be compared to
the amount of energy that has been input. If the
design energy - or at least enough energy to heat
the entire area plus some extra for heat losses - has
not been applied, that may be an indication that
there are significant cold areas in between the
temperature monitoring points. Once the energy
demand has been achieved, then the mass recovery
rate and groundwater and/or soil concentrations
should be considered in determining if 'diminishing
returns' has been achieved. 'Diminishing returns' is
indicated by mass recovery rates and groundwater
concentrations that are low and remain low (Heron
et al., 2023). Groundwater samples can be
obtained from monitoring wells or if groundwater
extraction is being used, samples can be collected
from the multiphase extraction wells which would
already have high temperature pumps installed an
operating. Several rounds of groundwater samples,
spaced over a couple weeks, should indicate that
the concentrations are well below that which would
indicate the presence of NAPL. Then confirmation
soil sampling can also be performed.
Section 7.1. Should We Specify Soil or
Groundwater Cleanup Criteria? How to
specify cleanup criteria is another frequently asked
question. There are advantages and disadvantages
to using either media for the cleanup criteria. Soil
concentrations, such as the concentrations for
leaching to groundwater, have been used at many
thermal sites. Drawbacks to the use of soil criteria
include the fact that extremely small soil samples
are analyzed, which makes it very difficult to
account for the natural heterogeneity of
contaminant distributions. This is particularly true if
a percent reduction in soil concentrations is used as
the cleanup criteria (which is not recommended). If
soil samples are utilized to document cleanup
levels, enough samples should be obtained for the
results to be statistically valid (EPA, 2002). When
the contaminants are VOCs, the soil samples should
be obtained from the sections of the core that
screening with a PID shows has the highest
remaining contaminant concentrations. Standard
operating procedures (SOPs) have been developed
for obtaining hot soil samples and obtaining valid
soil concentration data (Gaberell et al., 2002), but a
drill rig must be mobilized to the site, and there will
likely be issues with access to some sections of the
site due to the site infrastructure.
Groundwater concentrations have an advantage
over soil concentrations in that they interrogate
concentrations over larger areas. As discussed
previously, groundwater concentrations can be
expected to vary considerably during a thermal
remediation, making trying to meet groundwater
concentration criteria a 'moving target.' A
groundwater criteria would be even more
problematic if there were upgradient sources of
contamination that could recontaminate the
treatment area. When groundwater criteria are
used, consideration must be given to the fact that
the treatment area will remain hot potentially for
years after heating has been terminated, and the
solubility of some contaminants is greater at higher
temperatures. Overall, contaminant concentrations
33
-------
should be expected to decrease as the site cools. If
the groundwater concentration goals are met while
the groundwater is still at elevated temperatures,
the goals should continue to be met as the site
cools, as long as there are no upgradient sources.
Other criteria that have been proposed for shut
down of a thermal system are that a specified
amount of mass be recovered, or that the mass in
the ground be reduced by a certain percentage.
Both of these criteria suffer from the fact that it is
extremely difficult to determine the mass of
contaminants in the subsurface (unless a spill of a
known size recently occurred) due to the
heterogeneity of soil and thus contaminant
distribution. The mass of contaminants in the
subsurface has both been over and under estimated
when compared to the amount of mass that was
recovered via thermal remediation. Another closely
related issue is that the mass of contaminants
recovered is also an estimate.
Most of the thermal vendors have performed
guaranteed remediations to meet low soil or
groundwater concentrations. This can have the
advantage of a set cost up front to complete the
remediation. However, when the risk of being able
to meet the cleanup goals is placed on the vendor in
this manner, the costs will be increased over a
remediation that is not guaranteed. Thus, the
remediation is more costly whether it needs to be
or not. The least costly implementation is typically
achieved when a risk sharing and partnering
approach are considered between the responsible
party and the thermal vendor.
Section 7.2. What Happens to
Groundwater Concentrations After
Heating is Terminated? After an aggressive
thermal remediation of the NAPL-contaminated
subsurface, groundwater concentrations have been
shown to have an overall downward trend, even as
some variability in concentrations can be expected.
Figure 18 shows groundwater concentrations of TCE
at the East Gate Disposal Yard after ERH
remediation of two source areas, and documents
the increase in concentrations in the first treatment
area during heating, then the decrease in
concentrations that occurred after heating was
terminated. The figure shows that in this high
permeability sand and gravel aquifer, TCE MCLs
were reached in the source zones in a few years
after thermal treatment. Heron et al. (2016)
documents the decreasing dissolved phase plume
concentrations of PCE and its daughter products
after thermal remediation. Natural attenuation of
PCE in the dissolved phase plume had been
occurring before thermal remediation was
implemented in the source zone, and continued at a
rapid rate after thermal treatment. Five years after
thermal treatment was completed, only one
monitoring well still had vinyl chloride
concentrations above the MCL. The authors
concluded that back diffusion is not necessarily a
barrier to reaching groundwater goals.
Baker et al. (2016) document the effects of thermal
remediation of five source zones on groundwater
concentrations, showing that the mass flux of
contaminants from the source zone into the
downgradient plume can be reduced sufficiently so
that natural groundwater flushing can lead to
restoration of the plume within the time frame of a
decade or so. These five case studies also
demonstrate that back diffusion is not necessarily a
barrier to groundwater restoration, despite
common belief.
These examples show that aggressive source zone
remediation is justifiable, as it can lead to
restoration of groundwater, both within the
treatment zone and within the downgradient
plume. Aggressive treatment of essentially the
entire NAPL source zone can lead to groundwater
restoration in a reasonable timeframe.
Section 7.3. Can Thermal Remediation
Lead to Site Closure? Thermal remediation of
the source zone can in some cases be the sole
remedial action that is needed to close a site with a
notice of No Further Action (NFA) or de-listing from
the National Priorities List (NPL), if there are no
other contaminants (for example, metals) that are
not addressed by thermal remediation, and if there
is no dissolved phase plume that is above
34
-------
NAPL Area 1
Pre-Treatment
TCE in Shallow Groundwater
September-November 2003
<» Lower Vashon Welt TCE (ugri) not
NAPL Treatment Area
NAPL Areas 1 & 2
Post-Trealinent TCE in Shallow Groundwater
March 2006
LEGEND:
) TCE(ugfl)
a Upper Vashon Welt TCE lug/!),
a Lower Vashon Wei; TCE (ug/l> no
HAPL Treatment Area
NAPL Areas 1 & 2
Post-Treatment TCE in Shallow Groundwater
April 2007
J Tcecusn)
9 Upper vashon A'eJ; TCE (ugfl).
o Lowef Vashon Welt TCE (ugfl) n« contoured
NAPL Treatment
Figure 18. In one of the earliest deployments of thermal technologies at a Superfund site, ERH was used to remediate TCE
from a disposal facility for waste oils. Figure 18A shows the baseline TCE concentrations within and immediately
downgradient of the first NAPL-contaminated area to be treated. Note that the highest pre-treatment groundwater
concentration was found just outside of the treatment area. Figure 18B shows the groundwater concentrations after heating
was initiated. Heating increases the solubility and solubilization rate of chlorinated solvents, thus increasing groundwater
concentrations are typically seen after heating is initiated in the immediate vicinity of NAPL. Heating was terminated when
groundwater concentrations were less than about 400 ug/L. Figure 18C shows TCE concentrations 2 years after heating was
terminated in the first treatment area, and 1 year after treatment was terminated in the second adjacent treatment area. In this
sand and gravel aquifer with a relatively high groundwater flow rate, the dissolved phase TCE that remained in the first
treatment area was rapidly flushed from the area. Figure 18D shows that a year later, most of the remaining TCE was
flushed from the second treatment area as well.
groundwater goals. An example of this is a former
wood treatment site with creosote contamination
in the vadose zone only, which was comprised of a
tight clay. Because the contamination was in the
vadose zone and groundwater was not affected,
TCH remediation of the creosote in the vadose zone
was sufficient to reduce contaminants of concern
concentrations in the soil to below the cleanup
goals, and no further action was necessary In other
cases, treatment of the source zone has allowed for
re-use of the site.
Many Superfund sites, however, have more than
one contaminant of concern, and many have
downgradient dissolved phase plumes that are not
cost effective to treat with thermal remediation. In
these cases, multiple treatment methods may be
necessary to achieve site closure. Another former
wood treater site with creosote below the water
table and a downgradient dissolved phase plume
used a combination of steam injection below the
water table to recover the creosote NAPL, pump-
and-treat to recover downgradient dissolved phase
contaminants, and a soil cover to inhibit leaching of
contaminants from the vadose zone to groundwater
to achieve the groundwater cleanup criteria
throughout the former source zone which allowed
the site to be deleted the site from the National
Priorities List (NPL).
NAPL Area 1
Peak-Thermal
Shallow Groundwater
April 2004
35
-------
Section 8.0 Factors Affecting Costs
Many factors come into consideration when
determining the costs of thermal remediation. The
size and dimension of the treatment area (surface
area and depth), the depth to groundwater and the
groundwater flow velocity, type and quantity of
contaminants, wellfield access, and the project
goals will all affect the total project costs. Some of
the largest sites treated (250,000 to 400,000 cubic
yards (yd3)) have had the lowest costs on a per
cubic yard basis ($75/yd3 and $60/yd3, respectively),
while small pilot scale sites may approach
$2000/yd3. The reasons for this include the fact
that some costs, such as mobilization and
demobilization, costs of document production
(design reports, completion reports) are not
scalable, as well as factors such as greater
percentage heat losses on small sites. Sites above
the water table or with limited groundwater flow
rates generally require less energy than sites which
include significant depths below the water table
and significant groundwater flow. An energy usage
of approximately 220 to 250 kiloWatt-hour per
cubic yard (kWh/yd3) is typically required for VOC
treatment at the boiling point of water no matter
which heating technology is used. Treatment of the
lower-boiling SVOCs at their boiling point typically
requires an energy usage between 300 and 400
kWh/yd3. For higher temperature (300°C and
higher) TCH SVOC treatment, energy requirements
may be as much as 500 to 600 kWh/yd3. When
treatment is required to the ground surface, there
can be significant heat losses to the atmosphere
and the added expense of a cap at the surface.
SVOC sites typically have unit costs that are 50 to
100 percent higher than costs for VOCs due to the
greater energy requirements to reach higher
temperatures. The cleanup criteria will also affect
the costs of thermal remediation. For one site, cost
estimates were obtained from the vendors for two
different remedial end points, which showed that
reducing the groundwater cleanup goal from 300
ug/L to 3 ug/L increased the estimated energy costs
by approximately 30 percent.
Section 8.1. General Cost Information for
Thermal Remediation. The costs associated
with in situ thermal remediation were discussed
during a seminar in July 2021 put on by one of the
thermal vendors, and the following information
comes from that presentation. Costs for ERH and
TCH are generally similar, and will vary based on the
size and depth of the site, the amount of
groundwater flow through the site, the
recalcitrance of the contaminants to be treated, and
the remedial goals. Large, deep sites allow the fixed
costs to be spread over a larger volume, and costs
may be in the range of $85/yd3. Significant
groundwater flow into the treatment area is a heat
sink, and will tend to increase treatment costs
either because more energy is used to offset the
heat losses or because a combination of
technologies (TCH and SEE or ERH and SEE) are
required in high groundwater flow regimes to meet
the remedial goals. Higher boiling point compounds
or more stringent remedial goals will also increase
the costs per unit volume. Smaller sites with
volatile compounds may have treatment costs in
the range of $150/yd3 to 350/yd3.
Several contracting strategies can be used to reduce
overall costs. One approach is to pay for as many
costs as possible directly rather than passing the
cost to the thermal vendor. This includes the cost
for power (which is generally approximately 10 to
30 percent of the total costs - the larger the site
volume, the higher the percentage); the costs of
establishing temporary electrical, natural gas, or
potable water service; permitting; waste
management and disposal, including granular
activated carbon; and the costs of laboratory
analysis. Another strategy to reduce costs is to
reduce the risk to the vendor. The vendor takes on
risk when stringent remedial criteria must be met,
as by a guaranteed remediation. Also, the more
uncertain the site characteristics are, the greater
the uncertainty in achieving temperature goals and
thus cleanup criteria, and the greater the costs.
36
-------
Section 9.0 Life Cycle Analysis of
Thermal Remediation
Technologies
According to Lemming et al. (2013), "The LCA
translates the environmental exchanges during the
life-cycle of the remediation project (use of finite
resources, emissions to air, soil and water) to a
number of environmental impacts including global
warming, ozone formation, acidification,
eutrophication, respiratory impacts, human- and
ecotoxicity and resources depletion." However,
that definition fails to mention the primary
environmental impacts from the contamination
itself. From a groundwater contamination and
remediation perspective, primary environmental
impacts are defined as the local impacts related to
the contaminants in groundwater before and after
remediation. Cadotte et al. (2007) found that
primary impacts can represent a very significant
portion of the total global impacts, however,
impacts to groundwater are not included in
established impact analysis methods (Godin et al.,
2004; Lemming et al., 2010).
According to this definition, secondary impacts arise
from the remedial efforts, which includes resource
use and emissions from all the remedial stages
(Volkwein et al., 1999). Most studies comparing the
Life Cycle Analysis (LCA) of remedial technologies
have not evaluated the effects of different remedial
timeframe or different remedial efficiencies
(Cadotte et al., 2007), thus, caution must be used
when comparing the LCA of remedial technologies.
The results of an LCA will depend on a large number
of site specific factors, including the size and depth
of the contamination, the location of the site, its
accessibility, and the cleanup criteria. LCA of
thermal technologies found that in situ thermal
treatment becomes more environmentally efficient
for larger sites. SEE was found to have lower
environmental impacts per unit volume of soil
remediated then the other commonly used thermal
technologies due to the wider spacing of wells that
is possible when SEE is used at large, deep sites.
Non-toxic environmental impacts are generated
due to the energy consumption necessary to heat
the soils, however, these impacts are highly
dependent on the mix of sources for energy
production in the area where the remediation takes
place. Additional impacts are due to the above
grade materials used to treat the recovered
contaminants (Lemming et al., 2013).
Lemming et al. (2013) also evaluated which aspects
of in situ thermal technologies had the highest
secondary environmental impacts. Concrete use,
particularly as a vapor cap when treating close to
the surface, activated carbon for liquid and vapor
treatment, and the materials for wells and above
grade manifolds all contribute to the environmental
footprint of these technologies. Substitutions of
cap materials and the use of bio-based granular
activated carbon will help to reduce these
environmental impacts. Re-use of above ground
treatment systems at a significant number of
remedial sites also helps to reduce the
environmental footprint of these remediations.
This evaluation also concluded that delineating the
contaminated area carefully so that only the soil
volume that needs treatment is remediated is
important for reducing the overall environmental
impacts of remediation.
Acknowledgements. The author would like to
thank the peer reviewers, Dr. Gorm Heron of TRS,
Steffen Griepke of TerraTherm, and Kathy Davies of
EPA Region 3 for their careful reviews and
numerous suggestions to improve the content of
the paper. In addition, I want to thank Gavin Grant
of Savron for reviewing Section 2.4 on the STAR
technology to ensure the accuracy of the
information provided.
Author's contact information: Eva L. Davis, PhD,
can be reached at davis.eva@epa.gov.
37
-------
Section 10.0 References
Atkins, P. W., Physical Chemistry, Third Edition, W.
H. Freeman, New York, 1986.
Baker, R. and G. Heron, In Situ Delivery of Heat by
Thermal Conduction and Steam Injection for
Improved DNAPL Remediation, Paper 2B-18, in: A.R.
Gavaskar and A.S.C. Chen (Eds.), Remediation of
Chlorinated and Recalcitrant Compounds—2004.
Proceedings of the Fourth International Conference
on Remediation of Chlorinated and Recalcitrant
Compounds, Monterey, CA; May 2004.
Baker, R.S., S.G. Nielsen, G. Heron, and N. Ploug,
How effective is thermal remediation of DNAPL
source zones in reducing groundwater
concentrations?, Groundwater Monitoring & Reme-
diation 36(1):38—53, 2016.
Baker, R.S., D. Brogan, and M. Lotti, Demonstration
of three levels of in-situ heating for remediation of
a former MGP site, Paper L-74, Remediation of
Chlorinated and Recalcitrant Compounds - 2006:
Proceedings of the Fifth International Conference,
May 22-25, 2006.
Beyke, G., and D. Fleming, Enhanced removal of
separate phase viscous fuel by electrical resistance
heating and multi-phase extraction, The 9th Annual
International Petroleum Environmental Conference,
Albuquerque, NM, The Integrated Petroleum
Environmental Consortium, University of Tulsa, 22-
25 Oct 2002.
Beyke, G.L., B.A. Hodges, and G.N. Jones, Electrical
resistance heating of volatile organic compounds in
sedimentary rock, Remediation Journal 25(1):53—
70, 2014.
Cacciatore, D. A., N. Kong, and S. A. Peck, Electrical
resistance heating for DNAPL source removal at
Alameda Point, Remediation of Chlorinated and
Recalcitrant Compounds - 2008, Proceedings of the
Sixth International Conference on Remediation of
Chlorinated and Recalcitrant Compounds,
Monterey, CA, May 2008.
Cadotte, M., L. Deschenes, and R. Samson, Selection
of a remediation scenario for a diesel-contaminated
site using LCA, International Journal Life Cycle
Analysis, 12(4):239-251, 2007.
Class, H., and R. Helmig, Numerical simulation of
non-isothermal multiphase multicomponent
processes in porous media. 2. Applications for the
injection of steam and air, Advances in Water
Resources, 25:551-564, 2002.
Cornelissen, G., P. C. M. Van Noort, J. R. Parsons,
and H. A. J. Govers, Temperature dependence of
slow adsorption and desorption kinetics of organic
compounds in sediments, Environmental Science
and Technology, 31:454-460, 1997.
Costanza, J., J. Mulholland, K. D. Pennell, and E. L.
Davis, Effects of thermal treatments on the
chemical reactivity of trichloroethylene, Roberts.
Kerr Environmental Research Center, National Risk
Management Research Laboratory, Ada, Oklahoma,
EPA 600/R-07/091, August 2007.
Crownover, E., D. Oberle, M. Kluger, and G. Heron,
Perfluoroalkyl and polyfluoroalkyl substances
thermal desorption evaluation, Remediation
Journal, 29:77-81, 2019.
Dakota Technologies, DyeLIF Investigation, Paducah
Gaseous Diffusion Plant, Paducah, Kentucky,
February 2022.
Davis, E. L., How Heat Can Enhance In-situ Soil and
Aquifer Remediation: Important Chemical
Properties and Guidance on Choosing the
Appropriate Technique, Ground Water Issue Paper,
R. S. Kerr Environmental Research Laboratory, Ada,
Oklahoma, EPA/540/S-97/502, April 1997.
Davis, E.L., Steam injection for soil and aquifer
remediation, Ground Water Issue Paper, R. S. Kerr
Environmental Research Laboratory, Ada, Oklahoma
EPA/540/S-97/505, 1998.
38
-------
Davis, E. L., and B. K. Lien, Laboratory study on the
use of hot water to recover light oily wastes from
sands, EPA/600/R-93/021, R. S. Kerr Environmental
Research Laboratory, Ada, Oklahoma, February
1993.
Davis, E., N. Akladiss, R. Hoey, B. Brandon, M.
Nalipinski, S. Carroll, G. Heron, K. Novakowski, and
K. Udell, Steam enhanced remediation research for
DNAPL in fractured rock, Loring Air Force Base,
Limestone, Maine, US Environmental Protection
Agency, Office of Research and Development,
Cincinnati, Ohio, EPA/540/R-05/010, 2005
DeVoe, C., and K.S. Udell, Thermodynamic and
Hydrodynamic behavior of water and DNAPLs
during heating, In Proceedings from the First
Conference on Remediation of Chlorinated and
Recalcitrant Compounds, Monterey CA, Battelle
Press, May 18-21, 1998.
Einarson, M., A. Fure, R. St. Germain, S. Chapman,
and B. Parker, DyeLIF: A new direct-push laser-
induced fluorescence sensor system for chlorinated
solvent DNAPL and other non-naturally fluorescing
NAPLs, Groundwater Monitoring & Remediation,
28-42, Summer 2018.
EPA, Guide for Conducting Treatability Studies
under CERCLA, EPA/540/R-92/071a, November
1992.
EPA, Calculating Upper Confidence Limits for
Exposure Point Concentrations at Hazardous Waste
Sites, OSWER 9285.6-10, December 2002.
EPA, Site Characterization Technologies for DNAPL
Investigations, Office of Solid Waste and Emergency
Response, EPA 542-R-04-017, September 2004.
Falciglia, P. P., A. Bonifacio, and F. G. A. Vagliasindi,
An overview on microwave heating application for
hydrocarbon-contaminated soil and groundwater
remediation, Oil and Gas Research, 2:1, 2016.
http://dx.doi.org/10.4172/ogr.100011Q.
Gaberell, M., A. Gavaskar, E. Drescher, J. Sminchak,
L. Cumming, W.-S. Yoon, and S De Silva, "Soil Core
Characterization Strategy at DNAPL Sites Subjected
to Strong Thermal or Chemical Remediation." in:
A.R. Gavaskar and A.S.C. Chen (Eds.), Remediation
of Chlorinated and Recalcitrant Compounds—2002.
Proceedings of the Third International Conference
on Remediation of Chlorinated and Recalcitrant
Compounds, ISBN 1-57477-132-9. Battelle Press,
Columbus, OH, Monterey, CA; May 2002.
Gerhard, J. I., T. Pang, and B. Kueper, Time scales of
DNAPL migration in sandy aquifers examined via
numerical simulation, Ground Water, 45(2):147-
157, 2007.
Godin, J., J-F. Menard, S. Hains, L. Deschenes, and R.
Samson, Combined use of life cycle assessments
and groundwater transport modeling to support
contaminated site management, Human and
Ecological Risk Assessment, 10:1099-1116, 2004.
Grant, G.P., D. Major, G.C. Scholes, J. Horst, S. Hill,
M.R. Klemmer, J.N. Couch, Smoldering Combustion
(STAR) for the treatment of contaminated soils:
Examining limitations and defining success,
Remediation, 27-51, Summer 2016.
Heron, G. J. Bierschenk, R. Swift, R. Watson, and M.
Kominek, Thermal DNAPL Source Zone Treatment
Impact on a CVOC Plume, Ground Water Monitoring
and Remediation, 36(l):26-37, Winter 2016/pages
26-37, 2016.
Heron, G, K. Parker, S. Fournier, P. Wood, G. Angyal,
J. Levesque, and R. Villecca, World's Largest In Situ
Thermal Desorption Project: Challenges and
Solutions, Groundwater Monitoring & Remediation,
35(3):89-100, Summer 2015.
Heron, G., R. S. Baker, J. M. Bierschenk, and J. C.
LaChance, Heat it All the Way—Mechanisms and
Results Achieved Using In-Situ Thermal
Remediation, Remediation of Chlorinated and
Recalcitrant Compounds—2006, Proceedings of the
Fifth International Conference on Remediation of
Chlorinated and Recalcitrant Compounds,
Monterey, CA; May 2006.
39
-------
Heron, G., S. G. Nielsen, G. Crisp, D. Rentschler, J.
Galligan, K. Parker, D. Phelan, J. LaChance, and R.
Baker, Combining Thermal Technologies for
Effective Treatment of Complex Sites, Remediation
of Chlorinated and Recalcitrant Compounds—2012,
Eighth International Conference on Remediation of
Chlorinated and Recalcitrant Compounds,
Monterey, CA, May 2012.
Heron, G., J. LaChance, and R. Baker, Removal of
PCE DNAPL from Tight Clays using In Situ Thermal
Desorption. Ground Water Monitoring and
Remediation, 33(4):31-43, 2013.
Heron, G., K. Parker, J. Galligan and T.C. Holmes,
Thermal Treatment of 8 CVOC Source Areas to Near
Nondetect Concentrations, Ground Water
Monitoring and Remediation, 29(3):56-65, Summer
2009.
Heron, G., S. Carroll, and S.G. Nielsen, Full-scale
removal of DNAPL constituents using steam-
enhanced extraction and Electrical Resistance
Heating, Groundwater Monitoring & Remediation,
25(4):92-107, 2005.
Heron, G., T.H. Christensen and C.G. Enfield, Henry's
Law Constant for Trichloroethylene between 10 and
95 C. Environmental Science and Technology,
32(10):1433-1437, 1998.
Heron, G., personnel communication, 2023.
Heron, G., C. Crownover, C. Thomas, M. Boulos, R.
Glass, G. Crisp, B. Kennington, S. Tarmann, L.
Hidalgo, Thermal Conduction Heating for a 125 ft
deep TCE source - multiple lines of evidence for
verification of remedial goals, Groundwater
Monitoring & Remediation, in press, 2023.
Hunt, J.R., N. Sitar, and K.S. Udell, Nonaqueous
phase liquid transport and cleanup 1. Analysis of
mechanisms. Water Resources Research,
24(8):1247-1258, 1988.
Horst, J., J. Munholland, P. Hegele, M. Klemmer,
and J. Gattenby, In sit thermal remediation for
source areas: Technology advances and a review of
the market from 1988-2020, Groundwater
Monitoring & Remediation, pg. 1-15, 2021.
Imhoff, P.T., A. Frizzell, and C.T. Miller, Evaluation of
thermal effects on the dissolution of a nonaqueous
phase liquid in porous media. Environmental
Science and Technology, 31:1615-1622, 1997.
Itamura, M. T., and K. S. Udell, An analysis of
optimal cycling time and ultimate chlorinated
hydrocarbon removal from heterogeneous media
using cyclic steam injection, Proceedings of the
ASME Heat Transfer and Fluids Engineering
Divisions, HTD-321, 651-660, 1995.
Johnson, Jr., L. A., Development of the CROW™
Process, Topical Report, DOE/MC/11076-3847, May
1994.
Kluger, M., and G. L. Beyke, Electrical Resistance
Heating of Volatile Organic Compounds in
Sedimentary Rock, Remediation, pg. 69-82, Autumn
2010.
Kram, M. L., B. Hartman, and N. Clite, Automated
continuous monitoring and response to toxic
subsurface vapors entering overlying buildings -
Selected observations, implications and
considerations, Remediation, 29:31-38, 2019.
Lebron, C. A., D. Phelan, G. Heron, J. LaChance, S.
Nielsen, B. Kueper, D. Rodriguez, A. Wemp, D.
Baston, P. Lacombe, and F. H Chapelle, Dense Non
Aqueous Phase Liquid (DNAPL) Removal from
Fractured Rock Using Thermal Conductive Heating
(TCH), Naval Facilities Engineering Command
Contract Report CR-NAVFAC ESC-Ev-1202, August
2012.
Lemming, G., M.Z. Hauschild, P.L. Bjerg, Life cycle
assessment of soil and groundwater remediation
technologies: literature review, International
Journal of Life Cycle Assessment, 15:115-127, 2010.
Lemming, G., S. G. Nielsen, K. Weber, G. Heron, R. S.
Baker, J. A. Falkenberg, M. Terkelsen, C. B. Jensen,
and P. L. Bjerg, Optimizing the environmental
performance of in situ thermal remediation
technologies using life cycle assessment, Ground
Water Monitoring & Remediation, 1-14, 2013.
40
-------
Macbeth, T., M. Truex, T. Powell, and M. Michalsen,
Combining low-energy electrical resistance heating
with biotic and abiotic reactions for treatment of
chlorinated solvent DNAPL source areas, ESTCP
Project ER-200719, December 2012.
McGee, B. C. W., B. Nevokshonoff, and R. J. Warren,
Electrical heating for the removal of recalcitrant
organic compounds, in Wickramanayake, G. B., and
A. R. Gavaskar, eds., Physical and Thermal
Technologies: Remediation of Chlorinated and
Recalcitrant Compounds, The Second International
Conference on Remediation of Chlorinated and
Recalcitrant Compounds, Monterey, California, May
22-25, 2000.
McGee, B. C. W., H. Huang, and R. Westbrook, The
Thermodynamic Bubble Problem and Its Relevance
For In-situ Thermal Remediation, Battele Monterey
Conference, 2006.
Mejac, M. M., K. L. Brehm, and J. M. Tarvin, In situ
Electro-Thermal Remediation of chlorinated VOCs:
Full-Scale Evaluation, Remediation of Chlorinated
and Recalcitrant Compounds - 2008, Proceedings of
the Sixth International Conference on Remediation
of Chlorinated and Recalcitrant Compounds,
Monterey, CA, May 2008.
Metcalf & Eddy, Wastewater Engineering:
Collection, Treatment, Disposal, McGraw-Hill Book
Company, New York, 1972.
Morgenstern, M. R., J. A. Amari, A. MacMurray, M.
E. Farrar, T. P. Killeen, and R. F. Blundy, Electrical
Resistance Heating of soils at C-Reactor at the
Savannah River Site, WSRC-STI-2007-00488, 2007.
Mumford, K. G., S. Bryck, B. H. Kueper, S. Mancini,
M. Kavanaugh, and D. Reynolds, Virtual site
investigation to evaluate conceptual site model
development at DNAP-impacted sites, Groundwater
Monitoring & Remediation, 42(3):44-58, Summer
2022.
Myers, K. F., W. M. Davis, and J. Costanza, Tri-
Service Site Characterization and Analysis
Penetrometer System Validation of the Membrane
Interface Probe, U.S. Army Engineer Research and
Development Center, ERDC/ELTR-02-16, July 2002.
Newmark, R. L. and R. D. Aines, Dumping Pump and
Treat: Rapid Cleanups Using Thermal Technology,
Lawrence Livermore National Laboratory, UCRL-JC
126637, 1997.
Newmark, R.L. et al., Demonstration of Dynamic
Underground Stripping at the LLNL Gasoline Spill
Site, Final Report UCRL-ID-116964, Vol. 1-4.
Lawrence Livermore National Laboratory,
Livermore, California, 1994.
Oberle, D., E. Crownover, and M. Kluger, In situ
remediation of 1,4-dioxane using electrical
resistance heating, Remediation Journal, 25:35-42,
2015.
Pakkala, K., Vertical Mobilization of a Residual Oil
Phase in a Bead Pack due to Flow of Discrete Gas
Bubbles, A project submitted to the faculty of The
University of Utah in partial fulfillment of the
requirements for the degree of Master of
Engineering, Department of Mechanical
Engineering, University of Utah, October 2007.
Pennell, K. D., F. E. Loffler, J. Costanza, K. E.
Fletcher, N. S. Ramaswamy, G. Otano, J. Callaghan,
Investigation of Chemical Reactivity, Mass Recovery
and Biological Activity During Thermal Treatment of
DNAPL Source Zones, SERDP Project ER-1419,
October 2009.
Reynolds, D. A., and B. H. Kueper, Multiphase flow
and transport through fractured heterogeneous
porous media, Journal of Contaminant Hydrology,
71(l-4):89-110, 2004.
Roland, U., F. Holzer, and F.-D. Kopinke, Combining
different frequencies for electrical heating of
saturated and unsaturated soil zones, Chemical
Engineering Technology, 34(10):1645-1651, 2011.
Roy, J. W., and J. E. Smith, Multiphase flow and
transport caused by spontaneous gas phase growth
in the presence of dense non-aqueous phase liquid,
Journal of Contaminant Hydrology, 89:251-269,
2006.
41
-------
Scholes, G.C., J.I. Gerhard, G.P. Grant, D.W. Major,
J.E. Vidumsky, C. Switzer, and J.L. Torero,
Smoldering remediation of coal-tar-contaminated
soil: Pilot field tests of STAR, Environmental Science
& Technology, 49:14334-14342, 2015.
Sorenson, K., J. Bamer, A. Lopez, P. Chenevey, R.
Chichakli, R. Baker, G. Heron, T. Boivin, M.
Patterson, Full-Scale Thermal Treatment of Agent
Orange Dioxins at Danang Airport, Vietnam,
Presentation, 2018.
Stegemeier, G. L., and H. J. Vinegar, Thermal
conduction heating for in-situ thermal desorption of
soils, Hazardous & Radioactive Waste Treatment
Technologies Handbook, Chapter 4.6-1, Oh, C. H.,
ed., CRC Press, Boca Raton, Florida, 2001.
Stephenson, K. M., K. Novakowski, E. Davis, and G.
Heron, Hydraulic characterization for steam
enhanced remediation conducted in fractured rock,
Journal of Contaminant Hydrology, 82(3-4): 220-
240, January 2006.
TerraTherm, 2022. https://terratherm.com/wp-
content/uploads/2022/09/Hamilton-Labree-
Superfund-Site.pdf
Trine, L.S.D., Davis, E.L., Roper, C., Truong, L.,
Tanguay, R.L., Simonich, S.L.M., Formation of PAH
Derivatives and Increased Developmental Toxicity
during Steam Enhanced Extraction Remediation of
Creosote Contaminated Superfund Soil, Environ. Sci.
Technol., 53:4460-4469, 2019.
https://doi.org/10.1021/acs.est.8b07231
Udell, K.S., and L.D. Stewart, Field study of in situ
steam injection and vacuum extraction for recovery
of volatile organic solvents. UCB-SEEHRL Report No.
89-2, University of California, Berkeley, CA, 1989.
Udell, K. S., and L. D. Stewart, Combined steam
injection and vacuum extraction for aquifer
cleanup, In: Subsurface Contamination by
Immiscible Fluids, Weyer, K. U., ed., Proceedings of
the International Conference on Subsurface
Contamination by Immiscible Fluids, Calgary,
Canada, April 18-20, 1990.
Udell, K.S., Heat and mass transfer in clean-up of
underground toxic wastes. In Annual Reviews of
Heat Transfer, Chang-Lin Tien, Ed.; Begell House,
Inc., New York, Wallingford, UK, (7):333-405,1996.
Udell, K. S., Upward mobilization of DNAPLs due to
soil heating, Final Report, Cooperative Agreement
82947501, US Environmental Protection Agency,
2006.
Valsaraj, K., J. L. Porter, E. K. Liljenfeldt, and C.
Springer, Solvent sublation for the removal of
hydrophobic chlorinated compounds from aqueous
solutions, Water Research, 20(9):1161-1175, 1986.
Volkwein, S., H-W Hurtig, W. Klopffer, Life cycle
assessment of contaminated site remediation,
International Journal Life Cycle Assessment,
4(5):263-274, 1999.
Washington, J. W., Hydrolysis rates of dissolved
volatile organic compounds: Principles,
temperature effects and literature review, Ground
Water, 33(3):415-424, 1995.
Yuan, Z.-G., and K. S. Udell, Steam distillation of a
single component hydrocarbon liquid in porous
media, International Journal of Heat and Mass
Transfer, 36(4):887-897, 1993.
42
-------
Steam Enhanced Extraction (SEE) Case Study
The Beede Waste Oil Superfund Site is located in
Plaistow, New Hampshire, within a predominantly
residential area. Prior commercial operations at the
site from 1926 to 1994 included storage and
distribution of fuel oil and recycling of used oil.
Spills, leaks from storage tanks, and discharges to
lagoons on the site created subsurface plumes of
light nonaqueous phase liquids (LNAPL) that
contained a wide variety of petroleum
hydrocarbons, PCBs, and chlorinated solvents and
which covered approximately three acres. Between
2001 and 2005, a vacuum extraction system
recovered approximately 90,000 gallons of LNAPL.
The Record of Decision (ROD) chose soil vapor
extraction (SVE) to remediate the smear zone of
LNAPL, with a contingency for thermal
enhancements if it was determined during the
design stage that this was needed in order to meet
the site soil cleanup goals. The ROD also included a
groundwater extraction system to extract the
downgradient dissolved phase plume.
Bench scale treatability studies demonstrated that
thermal remediation of the soils was capable of
reducing contaminant concentrations to meet the
cleanup criteria, however, SVE at ambient
temperatures was not. Subsequently, Steam
Enhanced Extraction (SEE) was chosen as the
remediation technique for two LNAPL-
contaminated areas. In 2015 - 2016, SEE was used
to successfully meet the soil cleanup criteria in the
13,300 cubic yard Phase 1 area, a former lagoon
(Phase 1 treatment area and above ground
treatment system are shown in the photograph).
More than 150,000 pounds of contaminants were
recovered by the injection of 28.7 million pounds of
steam. In Phase 2, 66.3 million pounds of steam
were injected to recover approximately 37,000
gallons of contaminants from 21,456 cubic yards in
the former landfill.
Groundwater from the downgradient extraction
system was used for the steam. Most of the
contaminants were recovered as an LNAPL which
was then shipped off site for disposal, while
recovered contaminant vapors were destroyed in a
thermal oxidizer. In addition to the petroleum
hydrocarbons and chlorinated VOCs contaminants,
the recovered groundwater also contained naturally
occurring arsenic and bromide which were
mobilized by the heat. Additional treatment for
these compounds was required before the water
could be reinjected.
After the soil cleanup criteria were met in Phase 1,
it was found that a small amount of LNAPL
remained in the treatment area; approximately 80
gallons were recovered by bailing. This can be
attributed to the fact that the pumps have to be set
deep enough in the extraction wells to provide the
amount of recovery needed to maintain hydraulic
control, which does not always allow all of the
LNAPL to enter the pump so it can be extracted. To
alleviate this problem, Phase 2 included a 'slurper'
system at the wells, so that any LNAPL floating on
the groundwater table above the pump intakes
could be recovered by inserting a slurper tube and
applying vacuum extraction.
LNAPL from the Phase 2 area had been discharging
to surface water at Kelley Brook. In order to reduce
the effects of heat on the surface water to the
extent practicable, the SEE treatment area stopped
approximately 100 feet from the Brook. A sheet
pile wall was constructed to separate the portion of
the landfill that was to be treated by SEE from the
43
-------
portion closer to the Brook that would be
excavated. Extraction wells on the Brook side of the
sheet pile wall helped to remove LNAPL and heat
from that area. While the sheet pile wall aided in
protecting the Brook from heat and LNAPL, it would
have been more effective had it been extended
further past the steam injection area, and if the
joints had been sealed.
Even after the strict soil cleanup criteria that were
meant to be protective of groundwater were met
and the SEE system was terminated, groundwater
concentrations of naphthalene remained above the
New Hampshire State standard. The elevated
temperature of the groundwater following SEE
likely increased the solubility of naphthalene.
Pumps were re-installed in several of the extraction
wells with higher naphthalene concentrations and
extraction continued until the groundwater
standard was met.
Electrical Resistance Heating (ERH)
South Municipal Water Supply Well Superfund Site
is located in Peterborough, New Hampshire. In
1982, more than 100 parts per billion (ppb) total
chlorinated VOCs were found in the South Well,
causing it to be taken out of service. The nearby
New Hampshire Ball Bearing (NHBB) manufacturing
facility was found to be the source of the
contaminants. Chlorinated solvents including 1,1,1-
trichloroethane (1,1,1-TCA), trichloroethene (TCE),
and tetrachloroethene (PCE) were used at the
facility for degreasing during the manufacturing
process, and had been discharged to the ground.
NHBB had been established in 1957 and expanded
several times, necessitating the relocation of a
creek on site and changes to outfall locations, which
contributed to widespread contamination. Soil
Vapor Extraction (SVE) was implemented in the
vadose zone soils, but in 1997 an Explanation of
5 Study.
Significant Difference (ESD) was signed that issued a
Technical Impracticability (Tl) waiver for the site
due to the presence of DNAPL below the water
table that cannot be remediated by SVE. The ESD
called for the groundwater contamination to be
contained on site using extraction wells, however,
biofouling of the wells reduced their efficiency and
the extraction rates needed to contain the plume
could not be maintained.
In 2009 an amended Record of Decision (ROD) was
signed which called for in situ thermal treatment of
DNAPL source areas. A pre-design investigation
identified a mainly TCE source area under the
central part of the building, and just immediately to
the north was a PCE DNAPL source zone which
extended into the parking lot. An SVE system was
installed to address the TCE mass and to control
vapor intrusion into the building, and ERH was used
to remediate the PCE DNAPL. Angled, bored
electrodes were used under the building, and sheet
piles were used as electrodes in the parking lot. A
few of the electrodes were completed below grade
so traffic flow around the building could be
maintained. The ERH system recovered
approximately 4,500 pounds of contaminants,
including the 1,4-dioxane which was co-located with
the PCE DNAPL.
Indoor air was initially monitored using summa
canisters, which showed elevated contaminant
concentrations within the building soon after
heating started. A well within the building was
resealed, and changes were made to the HVAC
44
-------
system to bring the indoor air levels to acceptable
concentrations. An automated gas chromatography
system was installed that provided real time air
quality data from various locations within the
building, and it was found that the ERH system did
not contribute unacceptable levels of contaminant
vapors to the indoor air throughout the remaining
treatment time.
ERH was terminated in November 2016 after six
months of heating based on 'multiple lines of
evidence' - the treatment area was heated to and
maintained at the target temperature of the boiling
point of water, mass recovery rates had declined to
low levels, and groundwater concentrations had
decreased significantly and were no longer
indicative of DNAPL presence. Post-treatment
groundwater sampling, however, showed increasing
concentrations PCE and its break down products in
some treatment area wells within a year after
treatment was terminated. The wells affected by
these elevated concentrations sometimes varied
from one quarterly sampling period to the next. A
limited investigation to the northwest of the
treatment area located a smaller area which
appears to contain PCE DNAPL, which appears to be
the source of the post-ERH elevated concentrations.
Continued water table elevation measurements
seem to show that a water table mound sometimes
exists in the area of an outfall that is between the
parking lot and a wetlands, which may be the cause
of the changes in which wells are affected by this
dissolved phase plume coming from the northern
source zone. The recontamination of the treatment
area by elevated dissolved phase concentrations
illustrates the importance of addressing all of the
source zone to prevent recontamination.
Thermal Conductive Heating (TCH) Case Study.
Actions completed in the 1990s provided hydraulic
containment of the plume, but non-aqueous phase
liquid (NAPL) wastes remaining on site in the
overburden were a continuous source of
contamination to the groundwater. Technical
support efforts by ORD personnel on remedial
technologies resulted in in situ thermal remediation
being chosen by the Region in the 2005 Record of
Decision (ROD) for remediation of the overburden
soils. The objective of the thermal remediation was
to reduce the soil concentration of contaminants to
below levels that indicated the presence of NAPL.
The Solvent Recovery Services of New England
Superfund Site located in Southington, Connecticut,
is a former waste oil recovery facility that operated
from 1955 to 1991. The facility redistilled
approximately 100 million gallons of solvents. Spills
and discharges of solvents to lagoons during
operations discharged more than 500,000 pounds
of contaminants to the subsurface, which included
chlorinated solvents and petroleum hydrocarbons.
A dissolved contaminant plume reached
downgradient municipal water supply wells, causing
them to be shut down. Non-Time-Critical Removal
In order to determine the area that required
thermal treatment, Direct Push Technology (DPT)
was used to obtain continuous soil cores which
were visually inspected to determine if NAPL was
present. When NAPL was present, a step out boring
was completed further out from the area of known
NAPL presence until the extent of the NAPL was
delineated, an area of approximately 1.7 acres. TCH
was chosen as the treatment technology for the low
permeability soils. It was known that DNAPL was
pooled on top of the bedrock, so a drilling protocol
was developed to avoid the downward migration of
DNAPL into bedrock during the installation of the
TCH system. Sonic drilling was used, with outer
45
-------
casing installed ahead of the core barrel to a depth
of two feet into the bedrock. The bottom of the
borehole was then checked for DNAPL, and if it
was present, it was recovered. The heater casing
was then installed before the outer casing was
pulled. The DPT rig used for characterization had
met refusal at the top of weathered bedrock, while
the sonic rig used for system installation allowed
the actual depth to top of bedrock to be
determined and included within the treatment
area. Ultimately this added another 20 percent to
the total volume to be treated and caused some
adjustments to the design.
Due to the large volume of contaminants present,
the treatment area was divided into two phases,
and heating of the second phase was started only
after the mass of contaminants being recovered
from the first phase had decreased from the peak
extraction rate. This extended the total operational
time, but allowed a smaller above ground
treatment system to be used. The recovered vapors
were destroyed in a thermal oxidizer on site.
Despite the use of a phased startup of heating, the
large amount of petroleum hydrocarbons entering
the thermal oxidizer caused combustion in the pre-
oxidizer heat exchanger which damaged the daisy
wheel at the oxidizer inlet. To avoid future
occurrences, an organoclay filter was added after
the oil-water separator, a temperature sensor was
added at the oxidizer inlet, and the heat exchanger
temperature set point was reduced. During the five
weeks that it took to repair the system, heating was
discontinued, but vapor extraction was continued
with the use of the backup vapor phase granular
activated carbon that was already on site.
The successful thermal remediation was completed
in 2015, and recovered an estimated 496,400
pounds of contaminants. Final soil concentrations
were significantly below concentrations that would
be indicative of NAPL presence. Recent microbial
data collected from the site show that remaining
contaminants are being degraded, and contaminant
concentrations in the hydraulic control wells have
decreased substantially. A cap was constructed on
the site in 2017, and the site, along with the
adjacent former railway right of way, has been
converted to trails, as part of the "Rails to Trails"
program for former railroad lines.
46
------- |