-------
R
2.1.5.4 Wettability
Whenever two or more fluid phases occupy pore space, one of the fluids will be ad-
sorbed on the solid surfaces more strongly than the other fluid. The fluid that is most
strongly adsorbed is called the wetting fluid or wetting phase. The displaced fluid is the
nonwetting fluid. In most cases, liquids are adsorbed more strongly than gases (an ex-
ception is a mercury-gas system).
The angle 6 between the interface and the solid surface (measured through the denser
fluid) is called the contact angle. The contact angle of a wetting fluid is 6 <:90° (Bear,
1972).
Wettability represents the preferential spreading of one fluid over the solid surfaces in &
two-fluid system and is a function of the interfacial tension. A wetting fluid will tend to
coat solid surfaces and occupy smaller openings in porous media, and the nonwetting
fluid will tend to be constricted to the larger openings (Cohen and Mercer, 1993).
2.1.5.5 Relative Permeability
Relative permeability is the ratio of the intrinsic permeability for the fluid at a given
saturation ratio to the total intrinsic permeability of the soil. The two or three fluids (air,
water, and NAPL) in the unsaturated zone will compete for space in which to flow,
thereby reducing the total pore space available to either fluid.
Relative permeability curves can be used to describe different types of multiphase flow
regimes.
2.1.5.6 Hydraulic Conductivity
Hydraulic conductivity (K) is the rate of flow of water through a unit cross section under
a unit hydraulic gradient. This conductivity is a function of the porous medium and the
fluid properties. Hydraulic conductivity is an indication of the ease in which fluid will
flow through a porous medium.
Hydraulic conductivity is defined as:
* = ^ (2-13)
where:
K = intrinsic permeability
p = fluid density
g = gravitational acceleration
\i = dynamic viscosity of the fluid.
19
-------
2.1.5.7 Specific Surface
The specific surface (M) is defined as the total surface area of the pores per unit bulk
volume of the porous medium.
j^ _ surface area of pores
bulk volume of porous medium (
The specific surface is also defined with respect to the "unit volume of solid material" or
unit mass of porous medium."
2.1.5.8 Sorption Capacity
Sorption capacity is a function of the cation exchange capacity and specific surface of a
soil. Clays possess an overall net negative charge of their particles. Positive ions in
the soil fluids become associated with the negative charged soil particles in order to
maintain a chemical balance. This process, which is called the cation exchange in
conjunction with the surface area in the soils (very high for clays) available for contact
with the solution will determine the sorption capacity of a soil.
2.2 Sampling Technologies
Once a conceptual model of the chemical presence/transport, and fate is formulated
from the existing or preliminary site information (Figure 2-1), the site characterization
process is applied (Figure 2-2). Site characterization methods include field point mea-
surements and sampling of physical and chemical properties to improve the conceptual
model and facilitate the risk and remedy assessment. Soil, soil vapor, pore water and
groundwater sampling and analyses are the principal components of the site charac-
terization process. The objectives of the media sampling are to define:
Local geology and hydrogeology (i.e., stratigraphy, capillary barriers, and
traps)
Estimated quantities of hydrocarbons released; the source areas and
affected zones
Nature, extent, migration rate, and fate of contaminants
Fluid properties (i.e., density, viscosity, solubility, sorptive properties, etc.)
Media properties (i.e., permeabilities, porosities, organic content).
The remainder of this section presents various soil and groundwater sampling tech-
niques and methods for conducting both on-site field analyses and laboratory analyses
to support the objectives. • •
20
-------
2.2.1 Soil Sampling Techniques
A wide range of techniques is available for soil sampling. The sampling techniques
used to collect samples for measuring hydrocarbon releases in soil differ substantially
depending on the following:
Type of soil being sampled
Anticipated sampling depths
• Soil sampling capabilities
Equipment availability
Cost.
For a more detailed discussion of the sampling techniques, the reader is referred to
Driscoll (1986) or EPA (1993b) for a more thorough treatise on the subject.
Table 2-2 presents typical soil sample collection methods and techniques. Generally,
samples taken from excavated soil or from the upper 3 feet of soil can be collected with
simple hand tools such as trowels, shovels, spatulas, or manual soil boring methods.
These samples are suitable for most analytical parameters and physical parameters not
requiring undisturbed samples (i.e., bulk density, permeability). Samples are readily at-
tainable, and minimal setup and preparation are required between sample locations.
Hydrocarbons that have migrated vertically from the source to depths greater than 3
feet often require techniques such as tube samplers and augers to collect represen-
tative soil samples. Augers consist of a center shaft and a spiral cutting blade that
transports soil cuttings upward. Hand augers are generally used to depths not ex-
ceeding approximately 5 feet. Below 5 feet, hydraulically or mechanically driven
equipment is generally employed. Machine-operated augers are driven by a motor
(sometimes hand-held, but usually rig-mounted). The auger is rotated with downward
pressure to penetrate the soil. Depths of 100 plus feet are possible, with depths being
limited by the drilling rig torque capacity. Two common types of machine-operated
augers are hollow-stem augers and solid-stem augers. The hollow-stem auger allows
access of a sampling tool through its open annulus, whereas the solid-stem augers
require temporary removal of the auger flight for access of the sampling tools (EPA,
1993b). The depth of auger investigations is usually limited by groundwater depth, soil
characteristics, and the equipment used. Augers can be used to provide disturbed soil
samples or for advancing boreholes so that other types of sampling devices can be
used.
Two common sampling devices used in connection with auger drilling are the split-
spoon and the conventional thin-walled tube samplers. These sampling devices work
well in soils that contain sufficient clay or are cohesive enough for the material to
remain stable during sample collection and retrieval. The split-spoon sampler (Figure
2-3) is a thick-walled tube that is split in half longitudinally and can be separated to ex-
pose and remove the soil sample. The Modified California Sampler is a split-spoon that
21
-------
-------
•I
a
c
*3
O
O
CM
e>
Comments on Sampling
and Analytical Methods
•5 2 ~
2 OT 55
§: §
< o
8 1
=3 *
S to
•^ *n
CO "5
* 1
i i
Q. =>
,§•
1
I
£
_«
e
E
•E
=C
To
CO
•g
Sampling Approa
c
o
•B a>
en co
p ro
•S °-
X
8
& .2
il
CO
„
Some field screening methods (e.g., TPI
analyses and mobile laboratory analyse:
are available for assessing hydrocarbon
contamination of soil.
Laboratory analytical methods also are
typically used to confirm contaminant
levels.
X
X X
XXX
X XXX
Jj
1
jbsurface boring using:
Hand auger*
-Piston sampler
-Drive or core barrel si
-Coring
CO
•n
"O -
"H °
J<
co
1
2
13
co
>> n
Samples typically are collected using
bailers or pumps.
Samples can be characterized physicall
and chemically by laboratory analytical
methods. Laboratories can be mobile <
fixed.
X
CO
i
§
s
Kcavations and existing subs
UJ
X
X
X X
XXX
V
1 1
1 I
% t
emporary wellpoints and obs
stalled by:
Hand auger*
In situ sampling (direct-pu
-Drilling technology
I-.S
0.
->
_
f
5
2 -ft
Samples typically are collected using
bailers or pumps.
Samples can be screened effectively in
the field using indicator and portable
laboratory analytical equipment.
X
X
X X
XXX
V
1 I
1 1
™ 51
0> CO
emporary wellpoints and obs
stalled by:
Hand auger*
In situ sampling (direct-pu
-Drilling technology
F.E
Laboratory analytical methods are
commonly used to characterize
contaminants and to verify results of fie
screening efforts.
X
X
X
X
v
"5
c
1
t
i situ sampling using:.
In situ sampling (direct-pu
-Packer assemblies
*
Close support laboratories can be used
on larger projects for rapid turnaround c
final laboratory results.
X
X
X
X
1
I
c
ermanent monitoring wells (i
illing technology)
0- -0
T?
o
8
o
23
-------
Check valve
Splft barrel
Sample retainer
Hardened shoe
Figure 2-3. Split-spoon sampler.
contains several brass sleeves with metal fingers to retain less-cohesive sandy soils
rf^rm tV6S T bet.,USed in 6ither Sampler to maintain samP|e "^ty for laboratory
determination of volatile organic compounds. Both samples are suitable for either
cohesive or noncohesive soils. Although they can be used for soils with grave slarae
gravel or cobbles can obstruct the sample and affect recovery. 9
h £* ^"-walled tube sampler (also known as a push tube or a
Shelby tube) is a long, thin cylinder typically constructed of stainless steel or brass A
30^ ^ht" ^ rr1^ PUShlng the tUbe Under constant Pressure or 24 to
30 inches The thm-walled sampler is not suited for soils with cobbles that could be ob
strucjons because the device will tend to collapse or otherwise bend from the
' " Undisturbed samP|e suitable ^ *" analytical
24
-------
u
Head assembly
Set screw
Sample tube
Figure 2-4. Thin-wall sampler.
Other investigative and sampling techniques that have gained popularity in recent years
are the cone penetrometer and hydraulically. or mechanically driven probe samplers.
Driven probes consist of a cone tip or probe tip attached to a series of push rods driven
into the ground by a truck-mounted hydraulic drive system. A specially equipped truck
or van is used to house, transport, and deploy the driven probe sampler or the cone
penetrometer. Driven probe samplers are suitable to depths of up to 100 feet in soils
free of cobbles. Several makes and models are available that are capable of sampling
soils, soil vapors, and groundwater.
In poorly to moderately consolidated soil or sediment, hydraulically or mechanically
driven probe samples should be used to collect soil samples for residual liquid hydro-
carbon analysis. Soil samples for residual hydrocarbon analysis should be collected
from both above and below the water table. The depth to the water table and the pre-
sence of liquid hydrocarbons should also be documented. These horizons are usually
evident based on the texture, soil color, and odor of the soil. The presence of free-
phase hydrocarbons in the soil boring is clear evidence that a free hydrocarbon plume
has been penetrated.
25
-------
A though not commonly undertaken, test pits and trenches may be excavated at some
sites to allow visual inspection of, and access to, shallow subsurface soils Test pits
and trenches are open excavations that allow observation of the shallow subsurface
conditions at a site. They are excavated manually or more commonly with the means of
a backhoe, trench excavator, or other earth-moving equipment. An excavation is a
relatively inexpensive method for exploring contamination to depths less than 15 feet
however, it is more suited to sites with cohesive soils that will be self supporting and '
able to maintain stable sidewalls. Test pits may require the use of contaminated soil
handling, produce fugitive hydrocarbon vapor emissions, and have safety hazards
2.2.2 Soil Vapor Sampling
The behavior of a contaminant in the subsurface is difficult to predict given the com-
plexity of the soils and the chemical properties of the organic substances. Contami-
nants that have migrated through the soils or that have volatilized into the vadose zone
can produce free and residual hydrocarbons in the soil. Soil vapor sampling may be re-
^1 pefhnettfJe pref®nce and extent of vaP°r Phase hydrocarbons in the soil pore '
spaces Both static and dynamic techniques can be used to conduct soil vapor samp-
Hng Static sampling is accomplished either with an in situ adsorbent (activated char-
coal) sampler left in the ground to adsorb the volatile organic compounds or as a static
ThJ Ja£Phle H S°" f amP'e taken fr°m the 9round
-------
2.2.3 Soil Pore Water Sampling
Contaminants moving downward through the subsurface will pass through the vadose
zone. The contaminants migrate through pore spaces in the soil that are not filled with
water or other fluids and may remain or be adsorbed to soil particles. The vadose zone
intervals are monitored to characterize the site. The vadose zone pore water under
negative pore pressure, or tension, will not readily flow to allow for sample collection.
Soil pore water samples can be collected with suction lysimeters, however.
A suction lysimeter is a common vadose zone sampler comprised of a porous cup
attached to a hollow tube. Flexible tubing attached to the sampler allows sample col-
lection at the surface. A vacuum is applied to the tube and held for a specified period.
A vacuum greater than the negative pore pressure will cause a gradient on the tube,
and pore water will accumulate in the sampler or be pulled to the surface via sample
tubing depending on the configuration of the lysimeter used. The porous cup is con-
structed of ceramics, nylon, PFTE, or fritted stainless steel, depending on sample-
specific requirements. The sample tube is either PVC or stainless steel (Fetter, 1993).
The vacuum-pressure lysimeter is a modified version of the suction lysimeter. This
device is configured with two tubing lines: one for application of a vacuum to draw the
pore water into the sampler and a second to drive the fluid to the surface for collection.
Once the fluid is drawn into the tube sampler, the vacuum line is then pressurized and
the fluid sample is driven through the second line. Part of the sample may be driven
back through the porous cup into the formation when pressure is applied to the sampler
(Everett, etal., 1984).
2.2.4 Groundwater Sampling Techniques
Because some of the components of hydrocarbons are soluble in water, groundwater
sampling is required to characterize the site and determine the nature and extent of
contamination in the subsurface. Groundwater is sampled at permanent or temporary
sample locations by using a variety of extraction pumps, sampling devices, or sensors.
The sampling techniques used to collect samples for measuring hydrocarbon releases
in groundwater differ depending on the sampling depths and the volume of sample
required. Sample locations are permanent, as in the case of constructed screened
wells, or temporary one-time sample points, as are utilized with driven probe samplers.
The number and location of sampling points required for an assessment are site spe-
cific. A minimum of three spatially distributed wells are required to define the
piezometric or potentiometric surface. One point is generally required upgradient to
provide background information. The number of downgradient wells depends on site-
specific conditions and the nature of the contaminant including but not limited to the
following:
• Contaminant properties (solubility, density, reactivity)
• Groundwater system influences (pumping, migration pathways)
27
-------
i:
• Media properties
Location of receptors.
In addition, it may be necessary to assess conditions at variable depths and to develop
a three-dimensional image of the subsurface. This situation could require the use of
cluster, or nested wells. Nested wells are grouped wells that are screened at individual
horizons in one immediate area to monitor vertical gradients and groundwater quality.
Seasonally saturated intervals or water table fluctuations also should be considered in
defining the horizons and spatial distribution of sampling points.
The following discussions will present temporary and permanent sampling point tech-
niques available for groundwater characterization. These techniques are summarized
in Table 2-2.
2.2.4.1. Conventional Sampling Methods
A conventional groundwater sampling approach utilizes constructed wells completed at
specific horizons. A wide choice of well construction materials, including fluoropoly-
mers, metallics, and thermoplastics, is available that should be matched for compati-
bility with known contaminants. It should be noted that each material has its clear
advantage for application to specific contaminant classes.
Wells typically consist of a well casing, a filter pack consisting of a specified gradation
of sand to prevent infiltration of fine-grained formation soils, and an annular seal that
includes expansive clay to isolate a discrete monitoring horizon or prevent vertical
migration of surface water down the well casing (Figure 2-5). Screens may be of the
same construction material as the well casing or, if appropriate for the contaminant, a
hybrid construction with different materials to reduce the cost of construction materials.
Where wells are constructed such that the screen extends across the water table and
contaminants are not likely to contact the well casing, different materials may be used
than those used for the screen interval in order to reduce costs or improve construction
efficiency.
Once the wells are constructed, samples are collected by manual or mechanical tech-
niques. Manual techniques consist primarily of bailers and are most effective for the
sampling of shallow wells of less than 100 feet. Mechanical techniques consist of a
variety of electrical and air-driven pumps. Purnps generally are used on wells requiring
the removal of large volumes of purge waters or for deep wells over 100 feet. Bailers
are better suited for shallow wells or for those requiring only limited purge water
extraction.
Appropriate sampling equipment should be matched to the objectives of the analytical
program. Sample devices are constructed of inert materials suitable for the analytes of
interest, typically Teflon® or stainless steel. If samples are to be collected for volatile
organic compounds, bailers or positive displacement pumps such as bladder pumps
28
-------
•M
Bentonite Seal
Filter Pack
Well Screen
Figure 2-5. Typical monitoring well.
should be used to prevent loss of volatile components. Suction lift (peristaltic and
centrifugal) and some electrically driven impeller pumps suitable for purging and sam-
pling of other compounds should be avoided for VOC sampling because they may
cause volatiles to be lost due to degassing of the water or turbulent sample flow.
Bailers are the simplest and most portable of the sampling devices. They consist of a
rigid tube that fills with water when lowered into a well. As the device is retrieved, one
or both ends of the tube are sealed by a check valve to prevent loss of water back into
the well casing. Bailers are constructed of PVC, Teflon®, or stainless steel. Normally it
is best to dedicate one bailer to a well to prevent cross-contamination. When a bailer is
not dedicated to a single well, it must be thoroughly decontaminated between sampling
29
-------
of individual wells. The representativeness of the sample depends on the user's tech-
nique when using a bailer and could vary between users (Driscoll, 1986).
Bladder pumps or gas-driven piston pumps are simple technical devices that operate as
the name suggests. Water enters a bladder pump through the bottom and is forced
under a continuous column to the surface through contraction of the bladder by com-
pressed air. Water samples are separated from the air by individual bladders, and
cross contamination is avoided. Pumps of this design can maintain very low flows (i.e.,
100 ml/minute) suitable for sampling. Decontamination is difficult and pumps dedicated
to individual wells are recommended (EPA, 1993b).
In a gas-driven piston pump, gas is injected through one of two tubes to lower the
piston in the gas chamber, thereby allowing water to fill the upper chamber. Air under
pressure is applied via a separate tube to push the piston upward to drive the sample to
the surface. These pumps may operate at great depths but at lower flow rates than
submersible pumps discussed below. The pumps' valves and pistons are sensitive to
sediment and require thoroughly developed wells (EPA, 1993b).
The suction lift pump lifts the sample to the surface by applying a vacuum at the sur-
face. Negative pressure is applied by a portable pump through a tube lowered into the
well. Because of limits in the physics of fluid flow, these pumps have a practical lift limit
of 23 to 26 feet. Only the tubing need contact the sample, and decontamination is mini-
mal. These pumps are portable and relatively inexpensive. Negative pressures may
promote degassing and loss of volatiles.
Electric pumps operate with a submerged, motorized pump that drives impellers to
deliver water to the ground surface. A variety of pumps are available, but few are
designed specifically to collect groundwater samples. (The Grundfos® Rediflo II is one
such sampling pump.) Turbulent flow and water agitation within the pump can promote
degassing and volatilization. Although pump designs are changing to reduce agitation
the user is cautioned to select the appropriate pump. Submersible pumps are capable •
of greater pumping heads than most other pump types (in excess of 150 feet) and at
high flow rates. The initial costs of these pumps are greater than for other designs and
decontamination can be difficult. '
2.2.4.2 Driven Probe Sampling Methods
A relatively new type of groundwater sampling technology that has developed in recent
years is the in situ sampling probe or driven method. This technique allows rapid col-
lection of samples without the installation of permanent wells. These methods are de-
veloped from variations of the conventional cone penetrometer drive devices. They are
best employed in the preliminary site characterization stages where shallow and rapid
sampling options best suit the objectives.
30
-------
The soil-probe-type samplers include but are not limited to the Geoprobe®, Hydro-
punch®, and the BAT system samplers and are variations of the cone penetrometer
method originally developed to measure soil physical properties. The probes are
typically hydraulically driven to the horizon of interest, where a sample is collected via a
bailer type device in the push rods or with an evacuated vial configured with a septum.
The Geoprobe and Hydropunch use a sacrificial drive point that remains in the subsur-
face and provides a one-time sample (no additional samples can be collected from the
driven location). Additional samples require additional points to be driven. The BAT
system may be used as a temporary sample point or can be used to construct a perma-
nent sample point that may be revisited for additional samples.
The cone penetrometer test (CPT) is used to stratigraphically log soils and to determine
various hydraulic parameters. Soil permeability, groundwater head, and water-bearing
zones can be derived from the pore pressure data generated during the CPT run. As
the push rods are driven into the ground, excess pore pressure is produced. When
steady penetration is stopped, the excess pore pressure will decrease over time and
provide the information needed to calculate hydraulic conductivity. This method is not
as accurate for clean sands and coarser materials because the excess pore pressure
generated during the penetration of these materials is dissipated almost as soon as it is
produced.
Driven probe samplers may be successfully employed in soils to depths of 100 feet.
Devices that rely on groundwater suction for sampling will be limited to depths of
approximately 25 feet. Success depends on the nature of the soils and the potential for
obstructions to limit penetration. These devices may be used to sample groundwater
for all analytical constituents; where suction-type pumps are used, however, caution
should be exercised. Positive displacement pumps or bailers are best suited for VOC
sampling.
2.3 Analytical Methods
State regulatory programs require laboratory analyses of soil samples as part of the site
assessment and corrective action plan pertaining to soils containing petroleum hydro-
carbons; however, properly applied and performed field measurement techniques can
provide results more rapidly and useful for making on-site decisions. On-site analytical
methods are capable of providing chemical-specific quantitative data in the field or in a
nonlaboratory setting. Results from laboratory analyses provide quantitative data on
petroleum hydrocarbon concentrations. The following sections present the techniques
and methods utilized in both the on-site and fixed-based laboratory analyses.
2.3.1 On-Site Analytical Methods
Properly applied and performed field measurement techniques provide faster results for
making on-site decisions than do laboratory analyses. Because field measurements
are proving to be useful, new and improved instruments and techniques are being
developed. Table 2-3 presents performance information on currently available field
31
-------
0>
I
8
o
•a
n
•a
8
2
Q.
<*rf
0)
I
0)
«B
0)
iZ
o
Q.
0
CO
•o
o
CO
K_ O
o o
j>» fl»
n E
II
3 O
co a.
*?
N
,2
JO
UJ
o
o
o
3
o 2»
N - ID °
»I
o m
•§
•"
.2
•fc
•I
"55
E
•§a
5«
^••8«
u. -D O
o
CO
1.
to
ro
o>
3
o
I
8
Q
a.
90
u. O
0> S
•§ S
-«^ . ^_
<- p
I i
S Q)
D O
I
s-
o>
(D
(O
o>
S'
0)
I
(D
(0 CD
q= Q.
32
-------
techniques. The advantages of field measurement procedures and instruments include
the following:
Reliable qualitative and semiquantitative data that is produced on site can
be used to make immediate decisions regarding the need for further
assessment and ongoing remediation.
• The lower cost of field measurements allows more sampling points to be
included in the site assessment, which results in a more comprehensive
set of data.
• Immediate sample analysis reduces sample handling and eliminates
sample storage, thus minimizing the loss of volatiles.
The disadvantages of field measurement procedures and instruments include the
following:
• Depending on the procedure or instrument used, the results are generally
semiquantitative or qualitative.
• The age, degree of weathering, or type of petroleum hydrocarbons in a
sample determines which field technique will be used. (Some techniques,
such as headspace methods that are less sensitive to nonvolatile constit-
uents, are not well suited for weathered products.)
• Field techniques are subject to procedural errors that can affect the relia-
bility of the results.
• Several state underground storage programs currently do not accept field
measurement results alone (i.e., laboratory results are also required).
Although information collected by field measurement procedures can save time and
money, many state and local agencies require laboratory analyses to verify field infor-
mation, to quantify benzene, toluene, ethylbenzene, and xylenes (BTEX) concen-
trations and total petroleum hydrocarbon (TPH) levels, or to test for less-volatile prod-
ucts (e.g., diesel fuel).
Comparing results obtained by field measurement procedures and instruments with
those obtained by laboratory analyses is difficult. As indicators of the presence of
hydrocarbons, both techniques can provide useful results (given proper performance of
the field technique). The types of results differ, however. Most field procedures and
instruments test for groups of constituents, whereas most laboratory methods can be
selected to analyze for individual constituents or groups of constituents. Also, the dif-
ferent detectors used in the field do not always evaluate the same range of hydro-
carbons as the laboratory methods.
33
-------
A significant variability factor also enters into comparisons of field and laboratory anal-
yses. Field measurement methods can yield variable results because of variable instru-
ment response, variable conditions in the outdoor environment, inconsistent protocol
detector limitations, and inappropriate calibration. Laboratory analyses can be signif-
icantly influenced by sample collection method, holding time, and sample transport For
these reasons, it is important to have a well-thought-out Quality Assurance/Quality
Control (QA/QC) program to help eliminate external influences on analytical results.
Both field and laboratory analyses provide useful information for investigating a release
Field data are most reliable when obtained by a competent, well-trained field analyst
using properly calibrated and maintained field instruments.
A wide range of field analytical instruments is available for determining the presence of
volatile petroleum hydrocarbons in soil. Some simply detect the presence or absence
of unspecified groups of volatile chemicals, whereas other more-sophisticated tools can
identify and quantify specific constituents. The instruments discussed below are those
commonly used to detect volatile organics.
2.3.1.1 Colorimetric Detector Tubes
A colorimetric detector tube is one of the simplest field analytical tools Each tube is
designed to monitor a specific vapor or gas in air. They consist of tubes packed with
admixtures to react with the chemicals of interest. Although these tubes have some
Imitations m terms of accuracy and detection range, they have the advantages of being
inexpensive and easy to transport, use, and interpret.
2.3.1.2 Photoionization Detector
Portable photoionization detectors (PIDs) are relatively easy to use in the field and par-
ticularly sensitive to aromatic hydrocarbon constituents. The PID utilizes an ultraviolet
light to ionize the vapor sample in order to detect and measure the presence of organic
vapors. The detection range for these instruments.is about 0.2 to 2000 parts per million
(ppm). Accuracy varies with the concentration level being measured, type of con-
stituents present in the sample, and amount of moisture drawn into the instrument.
Because PIDs do not detect alkanes such as methane, they can be useful in detecting
aromatic constituents released in areas containing natural methane (such as in septic
fields, sewer lines, and bogs). The responsiveness of PIDs decreases in moist condi-
tions when the relative humidity of the sample or ambient air is high (above 90 percent).
2.3.1.3 Flame lonization Detector
Flame ionization detectors (FIDs) are commonly used to measure the presence of
organic gases and vapors. This instrument uses a hydrogen flame to ionize molecules
of volatile organic constituents (VOCs) in the vapor sample. The ionized molecules pro-
duce a current proportional to that of the sample. The FID will detect the presence of
volatile vapors, including methane, that may yield high readings (false positives) in
34
-------
areas where methane levels are higher than normal (for example, wetlands, sewers,
septic fields, and bogs). A direct-reading colorimetric detector tube specific to methane
can be used in conjunction with an FID to evaluate methane concentrations. The FIDs
are less sensitive than PIDs to environmental conditions such as relative humidity and
temperature; however, winds, excess carbon dioxide, and depleted oxygen can extin-
guish the flame in the instrument. These instruments are also more sensitive than PIDs
to alkanes such as hexane and butane, which make up a higher fraction of gasoline
than do the aromatics.
2.3.1.4 Portable Gas Chromatograph
A portable gas chromatograph (GC) uses a separation column to isolate and analyze
specific constituents in either a liquid or vapor phase in conjunction with a PID or an FID
detection system. A portable GC consists of a sample injection system, a separation
column, an output detector, and a detection system. A GC/FID system contains a com-
bustible gas supply for the flame; a GC/PID system contains an ultraviolet (UV) lamp.
Although this GC is portable, it still requires a stable field location (i.e., air conditioned/
heated office trailer) and it is not suited for hand carrying to individual site locations.
The instrument is fairly accurate with reproducible results.
2.3.2 Laboratory Analytical Methods
State regulatory programs require laboratory analysis of soil and ground water samples
for confirmation sampling as part of the site assessment and corrective action program
for media containing or having the potential for containing petroleum hydrocarbons.
Laboratory analytical results provide quantitative data for determining the presence of
hydrocarbon compounds in soil or groundwater. The methods generally are performed
in accordance with EPA Method SW-846, Test Method for Evaluating Solid Waste;
however, a broader ranger of methods may be applied because of the complex com-
bination of constituents with different physical and chemical properties (e.g., Methods
418.1, 8020, etc.). Table 2-4 presents a general overview of the analytical methods
and their application.
Analytical methods range from the generic Total Petroleum Hydrocarbon (TPH)
methods to the highly selective and sensitive Gas Chromatograph (GC) methods used
to analyze constituents. Indicator parameter methods (i.e., TPH) focus on the common
characteristics of several petroleum hydrocarbon constituents and are used as a
screening method for identifying gross amounts of TPH. These methods specify the
use of organic solvents to remove hydrocarbons from the soil matrix. Such methods
neither accurately measure the lighter fractions nor identify any natural soil organics
derived from biological activity (API, 1993). API has developed a method to prevent the
loss of volatile compounds between sample collection and analysis using either metha-
nol or methylene chloride as a preservative. Table 2-4 lists the methods commonly
used to detect, identify, and quantify indicator parameters and specific constituents in
soils. Table 2-5 lists the methods commonly used for analyzing dissolved
contaminants.
35
-------
Table 2-4. Analytical Methods for Soil Samples
Parameter
Method8
Comment
Benzene, xylene, toluene, and
ethylbenzene (aromatic volatile
organics)
EPA 5030A
EPA 8020ab
EPA 8021Ab
EPA 8240A"
Polynuclear aromatic hydrocarbons EPA 3550A
(PAHs) EPA8270A
Purge-and-trap extration Method GC-PID,
ignores MTBE
GC-ECD/PID in series
GC/MS (typically used for gasoline)
Ultrasonic extraction method
GC/MS (typically used for used-motor oil
and unknown)
Total petroleum hydrocarbons
Naphthalene
Benzene and 1,2-dichloroethane
(TCLP)
Lead (TCLP)
Ignitability/fiash point
Oil and grease
Percent moisture
pH
Organic matter concentration (total
organic carbon)
Grain size analysis
EPA418.1b'°
EPA8015Ab
EPA 81 00
EPA 131 1d
EPA 8240A"
EPA1311d
EPA 6010
EPA 7421A
EPA 7420A
EPA1010A
EPA 1020A
EPA 9071A
ASTM D2216
EPA 9045A
EPA 9060A
ASTM D422
Does not distinguish between naturally
occurring oils and petroleum-based oils
Does not measure lighter fractions, such as
BTEX
Zero headspace extraction
GC/MS analysis
TCLP leaching method
Inductively-coupled plasma (ICP)
Graphite furnace AA
Flame AA
Applies to liquids only, but is used on soils
Soil pH method
High-molecular-weight oils
Sieve and hydrometer analysis
EPA, Test Methods for Evaluating Solid Waste.
b
The method cannot distinguish between soil matrix interferences and the target compounds or
constituents. •
EPA Methods for Chemical Analysis of Water and Wastes.
40 Code of Federal Regulations Part 261, Appendix II.
36
-------
Table 2-5. List of Dissolved Hydrocarbons and Corresponding Methods of
Analysis (From U.S. Air Force, 1993)
Analytical Group
Constituent
Analytical Method*
Gasoline (motor gasoline,
aviation gasoline, and
gasohol)
Middle distillates
(kerosene, diesel fuel, jet
fuel, and light fuel oils)
Other or unknown
1,2-dichloroethane
Benzene
Toluene
Ethylbenzene
Total xylenes
Total volatile organic aromatics
1,2-dobromoethane
Methyl-f-butyl ether
Total petroleum hydrocarbons
Naphthalenes and other
semivolatiles
Benzene
Toluene
Total xylenes
Ethylbenzene
n-propylbenzene
Total volatile organic aromatics
Total organic halocarbons
Total petroleum hydrocarbons
Priority pollutant
Metals
Priority pollutant
Volatile organics
Priority pollutant
Extractable organics
Nonpriority pollutant
Organics (with GC/MS peaks
greater than 10 ppb)
Total petroleum hydrocarbons
EPA Method 8010
EPA Method 8020
EPA Method 8020
EPA Method 8020
EPA Method 8020
All detectable compounds by EPA Method
8020
EPA Method 8010 with ECD" substituted for
Hall detector; 2-column confirmation
EPA Method 8020
EPA Method 418.1 or 8015
EPA Method 8270
EPA Method 8020
EPA Method 8020
EPA Method 8020
EPA Method 8020
EPA Method 8020
All detectable compounds by EPA Method
8020
All detectable compounds by EPA Method
8020
EPA Method 418.1 or 8015
Typically atomic adsorption; particular
method dependent on metal analyzed
EPA Method 8240
EPA Method 8270
EPA Methods 8240 and 8270
EPA Method 418.1 or 8015
* Alternative methods, such as the EPA 500 and 600 series, exist and can be used in lieu of the EPA 8000 series.
These methods have other detection limits or varying quality assurance/quality control criteria or both.
b Notes: ECD = electron capture detector; GC/MS = gas chromatography/mass spectrometry; ppb = parts per
billion; n- = normal; t- = tertiary.
37
-------
Chapter 3
Free Product Migration and Recovery
3.1 Introduction
Groundwater contamination from hydrocarbon spills or leaks is a widespread problem.
The first step in assessing a spill site is to determine the areal and vertical extent of the
contamination and to estimate the spill volume. The spilled product is defined as
follows:
Residual hydrocarbons retained in the unsaturated zone for small-volume
releases or for a large depth to the groundwater surface.
Free product lor larger spill volumes or shallower water tables.
For larger spill volumes or shallower water tables, nonaqueous phase liquid (NAPL)
may reach the groundwater where it will spread laterally if its density is less than water.
Light nonaqueous phase liquids (LNAPLs) include most crude oils and common refined
hydrocarbons including various aromatic solvents, gasoline, jet fuel, and various grades
of fuel oil.
Common techniques for determining LNAPL volumes in the subsurface involve the use
of shallow wells that intercept the water table to delineate the areal extent of free prod-
uct floating on the water table or measurement of the total petroleum hydrocarbon
(TPH) on soil cores. To quantitatively interpret monitoring well data, the product thick-
ness measured in the wells must be converted to hydrocarbon-specific volume, i.e.,
hydrocarbon volume per unit area in the soil. The hydrocarbon-specific volume will
typically be less than the well product thickness. A disadvantage of TPH measurement
is that it yields a single measurement that is not suitable for long-term monitoring.
A theoretically based method for estimating hydrocarbon-specific volume from well
product thickness was developed and reported by Lenhard and Parker (1990) and Farr
et al. (1990). The method is based on the assumption that vertical equilibrium pressure
distributions in the water and LNAPL phases close to the water table can be inferred
from fluid levels in a well. An estimate of the volume of free product is determined on
the basis of the hydrocarbon-specific volume at each monitoring well.
38
-------
Initial remediation steps involve controlling and removing free product by pumping it
from trenches or wells in order to limit the spread of the plume. Pumping rates should
be selected to maintain hydraulic gradient control. Higher rates will lead to lower overall
product recovery as NAPLs are smeared over a larger cone of depression in the water
table and become trapped by capillary forces as residual hydrocarbons. The volume of
free product that is recoverable will rarely reach 50 percent and may be less than 10
percent in unfavorable cases. Careful placement and design of free product recovery
systems can have a major impact on recovery efficiency.
This chapter will present a discussion of the basic concepts of fluid movement and
retention in multiphase porous media systems along with a discussion of the analytical
tools available for assessment and remedial design of hydrocarbon spills as well as
practical aspects of employing these methods to leaking LIST sites. An overview will
also be presented of the applications, limitations, and design considerations for different
NAPL recovery systems.
3.2 Basics of NAPL Movement and Recovery
Theory of immiscible flow in porous media is not included in groundwater hydrology
textbooks. However, knowledge of relationships between fluid pressure (P), saturations
(S), and permeabilities (M) for the fluids of concern is required to predict advective
velocities and saturations. In this section basic equations for multiphase flow and S-P
relations are presented.
3.2.1 Continuity and Darcy Equations for Multiphase Flow
The following equations are used to describe the basic concepts of multiphase fluid
movement and retention in porous media and thereby properly assess and remediate
hydrocarbon spills.
3.2.1.1 Darcy Equation
The Darcy equation for any phase a (for water, a = w; for hydrocarbon, a = o; and for
air, a = a) is given by:
where:
qBj = volumetric flux of phase a in the /-th direction (Ls L-2 T-i)
kro = relative permeability of the soil to phase a (-)
ky - intrinsic permeability tensor of the soil, ij = x,y,z, (L2)
\ia = a-phase dynamic viscosity (MLT1)
P0 = a-phase pressure (ML"1T2)
Xj = ./-direction coordinate (L)
39
-------
p« = density of phase a (ML3)
g = gravitational acceleration (LT2)
Uj = 8z/8Xj, a unit gravitational vector (+ upward) (-).
In groundwater hydrology, water-height equivalent heads rather than pressures are
commonly used, and the equation may be written equivalently as:
5A.
(3-2)
where:
Kswy = kijDwg/r)w, the saturated conductivity for water (L/T)
Hra = Mo/Mw. relative viscosity of phase a (-)
h« = Pa/gpw, the water-equivalent pressure head of phase a (L)
Pm = Pc/Pw. the specific gravity of phase a (-).
The generalized Darcy law can be used to describe the flow of water, NAPL, and air in
soils when one, two, or three phases coexist within the pore space.
3.2.1.2 Mass Conservation Equation
Continuity equations can also be written for each phase. These equations require
mass conservation for each phase; i.e., within a fixed soil volume, the change of mass
within a phase equals the net difference of mass entering or leaving the phase.
For an incompressible fluid and porous medium, the continuity equation for bulk phase
a is:
where:
q«i = volumetric flux of phase a in the /-direction (LT-i)
pc = density of phase a (ML3)
Sa = ' fraction of pore space filled with phase a (saturation) (-)
t = time (T)
4> = soil porosity (-)
*i = /-direction coordinate (L).
40
-------
In order to make use of these equations, the relations between fluid pressures, fluid
saturations, and fluid relative permeability must be understood.
3.2.2 Capillary Retention and Relative Permeability Relations
3.2.2.1 Two-Phase Capillary Pressure Relations
Two-phase capillary pressure relations define the capillary pressure, Pc, between two
phases (e.g., air and water or oil and water) by:
p = p - p
c nw v>
where P^ and Pw denote nonwetting and wetting phase pressures, respectively.
The capillary head, h , is similarly defined by:
*, = *«r ~ h* (3-5)
where hnw and hw denote nonwetting and wetting phase pressure heads, respectively.
The schematic shown in Figure 3-1 of a hypothetical pore cross section illustrates the
difference between the wetting phase and the nonwetting phase in a two-phase system.
The wetting phase may be water, and the nonwetting phase may be oil or air. The
interface curvature is related to the capillary pressure by the Laplace equation of
capillarity, which is given by:
p -L°
P< ~ r "-- (3-6)
where:
6 = contact angle
r = radius of interface curvature (L)
o = interfacial tension between the fluids (MT2)
Pc = capillary pressure (ML~1T"2).
Note that the capillary pressure, Pc, will increase as the fluid retreats into smaller pore
cross sections when the wetting phase saturation diminishes.
Consider an idealized representation of soil as a bundle of variously sized capillary
tubes (see Figure 3-2) at equilibrium with the water table. At the water table, the water
and air pressures are equal, so the capillary pressure is zero. The air-water capillary
pressure head will increase linearly above the water table, and the capillary rise will
depend on the radius of each "pore." The degree of saturation with the wetting phase
41
-------
6 is the contact angle
Figure 3-1. Hypothetical pore cross section with two fluids.
42
-------
aw
Air
Water
Saturation
Figure 3-2. Capillary bundle model of soil pores and corresponding saturation-
capillary pressure curve.
will diminish as the elevation above the water table increases in a manner that depends
on the pore size distribution as shown in Figure 3-2.
Actual soil pores are a mixture of shapes and sizes interconnected in a complex man-
ner that is better represented as a network of pore bodies connected by a system of
pore throats. It is this pore network, rather than a system of capillary tubes, that gives
rise to the phenomenon of hysteresis, which is the variation in the relationship of capil-
lary pressure to saturation depending on whether wetting phase saturation is increas-
ing (imbibition) or decreasing (drainage). A pore network with fluid distribution is pre-
sented in Figure 3-3; the network shown in Figure 3-3a is filled with wetting fluid-for
example, water. Nonwetting fluid-for example, gasoline-is introduced from the upper
surface by in-creasing the nonwetting phase pressure (or decreasing the wetting phase
pressure) thereby increasing the oil-water capillary pressure. The oil will penetrate into
any pores having a diameter larger than a certain value, which may be computed from
the Laplace equation for the applied capillary pressure. Because the wetting phase is
decreasing, the process is normally referred to as drainage. At the end of the drainage
event, the fluid distribution in the pores may look like that shown in Figure 3-3b.
43
-------
a. Wetting phase saturation
b. Wetting phase drainage
c. Wetting phase imbibition
Figure 3-3. Network model of soil pores.
44
-------
A decrease of the oil-water capillary pressure decrease in oil pressure or increase in
water pressure) will cause water to displace oil from pores if the capillary pressure is
less than the Laplace threshold. When the wetting phase saturation increases, the
process is referred to as imbibition. The distribution of pore body and throat sizes will
inevitably cause some of the oil-filled pores to get cut off because of the water displace-
ment of oil in the surrounding larger pores. The net effect is a trapped nonwetting
phase. Now, even when the capillary pressure returns to zero, some of the oil will
remain at a residual saturation. The different path followed during drainage and
imbibition is referred to as hysteresis. In addition to nonwetting phase entrapment,
other factors such as variations in contact angles during wetting and drainage may lead
to macroscopic hysteresis. Typical main drainage and imbibition curves are shown in
Figure 3-4. The nonwetting phase saturation when the main imbibition curve reaches
zero capillary pressure is referred to as the maximum nonwetting phase residual
saturation, Snr. Note that an infinite number of scanning paths can occur between the
main drainage and imbibition paths.
3.2.2.2 Parametric Models for Saturation-Capillary Pressure (S-P) Relations
Measured capillary pressure data are commonly fit to the van Genuchten or the Brooks-
Corey model. Equation 3-7 shows the van Genuchten model.
Sw = [1 + (a he)nrm for he>0 (3-7)
where:
Sw = apparent wetting phase saturation (L°)
a = parameter proportional to mean pore size (L~1)
h,. = air-water capillary head (L)
n - a parameter inversely related to width of pore size distribution (L°)
m = 1 - 1/n.
The Brooks-Corey model is presented below:
f°r h > * (3-8)
where:
h& = • nonwetting fluid entry pressure (L)
A. = pore size distribution parameter (L°).
As the mean pore size increases (coarser grain size), the displacement pressure hd
decreases and the parameter a increases. The parameters n and A are pore size
45
-------
18
-------
Bow = oil-water scaling factor (L°)
oaw= surface tension of water (MT2)
oow = oil-water interfacial tension (MT2).
Air-oil capillary pressures can also be estimated by assuming that oil is the wetting
phase in the air-oil pair. Thus, a similar method is produced, as shown in the following
equations.
S.(P«*«) = S«(haJ (3-11)
where:
B = ^L (3-12)
rao
°ao
where: oao = air-oil interfacial tension.
Figure 3-5 presents wetting fluid saturation-capillary head relationships for air-water,
air-NAPL, and NAPL-water fluid pairs.
3.2.2.4 Three-Phase Capillary Pressure Relationships
The three-phase capillary systems show the behavior of a porous media with three fluid
phases: air, water, and oil. It is assumed that water is the wetting fluid for the water-oil
pair and that oil is the wetting fluid for the oil-air pair.
The idealized pore cross section for the three phases is shown in Figure 3-6. The oil-
water interface radius varies depending on the amount of water saturation in the pores,
which will control the oil-water capillary pressure. The degree of air-oil interface curva-
ture depends on total liquid saturation (oil plus water), which controls the air-oil capillary
pressure. The three-phase capillary pressure relationships are represented by two
functions: Sw vs. /7W and S, vs. /7ao (where St = Sw + S0). Note that the pristine air-water
curve may vary from the contaminated air-water capillary pressure curve because of a
change in the surface tension of contaminated water.
3.2.2.5 Relative Permeability Relationships
A complete description of fluid dynamics in multiphase systems is needed to under-
stand the relative permeability changes under various conditions. The relative per-
meability of a porous media (kj to the wetting phase depends on the media used and
the degree of wetting phase saturation. As saturation decreases, the flow path be-
comes longer and more circuitous, thus resulting in a decrease in permeability as
saturation diminishes. Capillary network models of the soil indicate:
47
-------
160
iZNZYL. ALCOHOL-VATCR-AIR
160
O. 0 O. 2
0. 4 O. 6 O.3
Saturation
1. O
"C* 200
I
0 1SD "
1
31 1QO -
W
UNSCALED S-P RELATICNS
• AIR-WATER DATA
» AIR-TCE DATA
" TCE-WATER DATA
o. o o. 2 o. 4 o. s o. e i.o
Saturation
BENZYL. ALCOMOL-WATEa-AIR
« air-water data
* air-oil data
* oil-water data
O. O. O. 2 Q. 4 Q. 6 Q. B
Effective Saturation
253
to
|
u,
•o
27O -
150 -
%
•o
CD
s
CO
SCALED S-P RCLATIONS
• AIR-WATER DATA
• AIR-TCE DATA
* TCE-WATER DATA
O. 0 Q. 2 O. 4 0.6 0. 8 1.3
Effective Saturation
Figure 3-5. Measured S-PC drainage curves for two soils and three fluid pairs.
48
-------
Oil
Water
Assume wettability in order: water -» oil -» air
Figure 3-6. Schematic of idealized pore cross section with three fluids.
Source: after Parker, 1989.
R
tat
(3-13)
where:
Ka = mean radius of pores filled with °<-Pnase
KMt = mean radius of all pores (L)
T = tortuosity coefficient (L°).
Theoretical and experimental studies indicate:
where b is an empirical exponent (L°).
(3-14)
The van Genuchten model shows that water relative permeability is only a function of
water saturation, and that air relative permeability is only a function of air saturation; the
model further shows that oil relative permeability is a function of both water and air
saturation, if both phases are present.
(3-15)
49
-------
(d -
(3-16)
The effects of hysteresis were not considered in the preceding equations (3-15, 3-16,
3-17). Experimental studies indicate that water relative permeability is minimally
affected by saturation history. Nonwetting-phase relative permeabilities, however, can
be markedly affected by nonwetting fluid entrapment because trapped fluid is
hydraulically discontinuous and essentially contributes nothing to phase permeability.
Typical relative permeability curves for a two-fluid phase system are shown in Figure
.3-7 for wetting phase drainage and imbibition. Note the small degree of hysteresis in
the wetting-phase relative permeability curve. Also, note that wetting-phase per-
meability decreases several orders of magnitude as the wetting-phase saturation
decreases from 100 to 50 percent, whereas the nonwetting-phase permeability
decrease is roughly proportional to the saturation decrease. This change occurs
because the nonwetting phase, by definition, occupies large pores that contribute
disproportionately to permeability.
3.2.3 NAPL Movement and Residual Saturation in the Unsaturated Zone
The behavior of water alone will first be considered before any discussion of the
movement of NAPL in the unsaturated zone. Following a rainfall or other event that
brings the soil surface to a high degree of water saturation, water will continue to move
vertically downward under the influence of gravity in a process referred to as redis-
tribution. As redistribution proceeds, water saturation in the initial wetted zone de-
creases and relative permeability decreases proportionately. With time, redistribution
essentially ceases and water content reaches a quasi-static state referred to as field
capacity. It should be noted that this is not a true hydrostatic equilibrium condition.
NAPL spills or leaks in the unsaturated zone will move downward under the force of
gravity and capillary pressure as well as laterally because of capillary forces. As the
release migrates, the following occurs:
The rate of advancement will be controlled by oil conductivity, which
varies directly with intrinsic permeability, relative permeability, and NAPL
density and varies inversely with NAPL viscosity.
At high degrees of saturation, oil relative permeability will be near unity; at
lower saturation, the relative permeability will be lower.
50
-------
O. O O. 2 Q. 4 Q. 6 O. 8 l.O
ID
Wetting Fluid Saturation
Figure 3-7. Typical wetting and nonwelting phase relative permeabilities for
wetting phase imbibition and drainage paths.
51
-------
For slow leaks, the saturation will adjust to a low value. After the leak ceases,
the front will continue to advance, but at an increasingly slow rate because the
relative permeability begins to diminish as a result of the gradual reduction in
NAPL saturation as the spill volume distributes over a greater soil volume.
Because the NAPL relative permeability decreases approximately exponentially
as saturation decreases, the rate of redis-tribution tends to decrease
exponentially with time. As a result, after a period of weeks or months of
redistribution after a spill event, NAPL fluxes in the unsaturated zone near the
source may be essentially zero and the NAPL saturation is apparently static
Although a true equilibrium condition does not exist because oil will not exhibit a
hydrostatic pressure distribution, a quasi-static state is maintained by negligible*
fluxes under the ambient gradient (due to gravity).
Residual NAPL saturation occurring behind the draining oil front is re-
ferred to as pendular residual oil saturation or unsaturated zone residual
saturation, not to be confused with trapped or insular residual saturation
caused by an NAPL being occluded by water imbibition.
An example of the separate and combined effects of oil entrapment and unsaturated
auTeraround S^* presented in Fi9ure 3"8 for LNAPL redistribution after a spill
riLhtlhJfn H 7 C3Se considered assumed a Doping water table from left to
nght hat gradually rose over a 20-day period by 1 meter. The oil distribution 20 days
after the spill event, simulated without consideration of oil entrapment (see Figure 3-8a)
shows pendu ar residual oil of about 10 percent of pore space near the ground surface
S±t m°VemeHnt 'I,?' °Ver the W3ter table is much 9reater than th* ™* «£,
as arSr^'8 ?0nSldei;ed (see R9ure 3-8b> beca"*e oil transmissivity is diminished
as a result of the large volume of oil trapped (see Figure 3-8c).
The findings based on this example are summarized as follows:
LNAPL spreads laterally after reaching the water table
Oil entrapment markedly affects NAPL plume movement
Diminishing oil permeability near the ground surface leads to unsaturated
zone residual oil ("pendular" residual oil).
3.2.4 Relationship Between Well Product Thickness and Soil Distribution
After an LNAPL spill product moves vertically downward through the unsaturated zone
until it encounters a fine soil layer whose nonwetting fluid entry pressure precludes
further downward movement, it reaches residual saturation, or it encounters a water
table In this section, the behavior of LNAPL that has reached a water table will be
considered.
52
-------
a. Without oil entrapment
Total oil: t - 2O d
Low donclty
b. With oil entrapment
Troppad oil: t « 20 d
c. Trapped oil
Total oil: t - 20 d
Figure 3-8. Oil saturation distributions twenty days after an LNAPL spill.
a) Total oil saturation without considering oil entrapment, b) Total oil
saturation with oil entrapment, c) Trapped oil saturation for case b.
Units on contours are fractions of pore space.
53
-------
3.2.4.1 Vertical Capillary Pressure Distribution
Water and oil flow are controlled by their piezometric heads. The piezometric heads for
water (qjj and oil (tyj are as follows:
. ft, = h* +z (3-18)
#0 =ho +Pnz (3-19)
where:
hw = water pressure in units of water height (L)
h,, = oil pressure in units of water height (L)
z = elevation above an arbitrary datum (L)
pro = oil-specific gravity (L°).
The following equations were derived from the fluid table definitions defined in Fiaure
3-9:
, - '„ + \ (3-20)
= Pro'ao +ha (3-21)
where:
h,, = air pressure in units of water height (L)
zaw = air and water pressure elevation (L)
zao = air and oil pressure elevation (L).
Note: at constant h., areal water flow is controlled bv z,
, ^BW?
and areal oil flow is controlled by gradients in zaw.
Assuming vertical fluid redistributions occur on a sufficiently short time scale so that
vertical pressure distributions approximate hydrostatic conditions, then vertical capillary
pressure distributions are given by:
kao = Pro (* - zoJ (3-23)
Fluid table elevations are related by:
*aw - *ow = Pro Ho (3-24)
54
-------
RELATIONSHIP BETWEEN SOIL AND WELL PRODUCT THICKNESS
VERTICAL EQUILIBRIUM ASSUMPTION
If recharge/leakage fluxes are small compared to regional
velocities, vertical water pressure distributions will approach
hydrostatic
After oil reaches the water table, oil flow is primarily
horizontal and vertical oil pressure distributions will tend
towards equilibrium
Fluid Table Definitions for an LNAPL
elevation where air and oil pressure are equal
elevation where air and water pressure are equal
elevation where oil and water pressure are equal
Fluid table elevations are detected by a monitoring well or piezometer
Figure 3-9. Fluid table definitions for LNAPLs.
55
-------
where: H0 = z^ - zow is the well oil thickness (L).
3.2.4.2 Vertical Saturation Distributions
The following examples assume that two fluid table elevations or one elevation and the
apparent oil thickness (H0) completely define the three-phase static vertical head distri-
butions.
Knowing fluid table elevations zao and z,™ at a specified areal location,
combined with the three-phase saturation-capillary pressure relationships,
water and oil saturations may be computed as a function of elevation
(Figure 3-10).
An oil-water capillary fringe occurs above z^ which is water saturated or
at apparent water saturation if trapped nonwetting fluids are present. The
thickness of the oil-water capillary fringe for the Brooks-Corey model is
given by the following equation:
5
'W
(3-25)
aw
where:
6aw = thickness of air-water capillary fringe in oil-free system (L)
6OW = thickness of the oil-water capillary fringe (L)
BOW = oil-water scaling factor (L°).
A narrower air-oil capillary fringe, which is also liquid-saturated (disregard-
ing trapped fluids), will occur above zao and is defined by:
5 - *•» (3-26)
"° PaoPro
where:
6ao = thickness of the air-water capillary fringe (L)
Bao = air-oil scaling factor (L°).
Because the oil-water capillary fringe thickness is greater than that of the air-oil capillary
fringe, the actual soil thickness over which oil occurs will be smaller than the well oil
thickness.
Trapped oil is not in hydraulic contact with free oil and cannot be detected
by a monitoring well. Computed oil saturation distributions based on
56
-------
Soil Profile
Well
6.00 -
5.00 -
,1 4-°° ~
g 3.00 H
J3
% 2.00 -
I 1.00 -
0.00 -{
\
0.00 6.25 6.50 6!?5 1.00
Saturation
Air
Oil
Water
Figure 3-10. Oil and water saturation distribution in soil and relationship to fluid
levels in a monitoring well.
nonhysteretic saturation-capillary pressure relationships will yield esti-
mates of free oil saturation only. The maximum elevation (zu) at which
free oil will occur in equilibrium with an oil lens is:
where:
p B - (l -
(3-27)
The free oil-specific volume (Vg,) or free NAPL volume per unit area can
be computed from the vertical oil saturation distribution by the following
equation:
z.
of
/»
(3-28)
soil porosity (L°)
free oil saturation (L°)
oil-water table elevation (L)
maximum elevation where free oil occurs (L)
elevation above a datum (L).
57
-------
Integration of Equation 3-24 yields Vof as a function of H0 as shown in Figure 3-11.
1.00 -j
0.00
0.00
1-00 2.00 3X30 '.4.00 5X10
H0(m)
Figure 3-11. Free oil-specific volume versus well product thickness for a
representative soil.
3.2.5 Areal Movement of Floating Product
3.2.5.1 Areal Water and Oil Flow Equations
Once LNAPLs reach the water table, further vertical movement will be limited by buoy-
ancy effects. Flow will occur primarily in the areal (horizontal) direction. Under condi-
tions of constant gas pressure, areal water flow occurs in response to gradients in the
air-water table and oil flow occurs in response to air-oil table gradients.
Assume approximate vertical equilibrium pressure distributions and
uniform gas pressure. Integration of Darcy's law in the z-direction yields
the following areal water and oil flow equations:
G= - T
w, »
J?
'<• dx.
(3-29)
= - T
(3-30)
58
-------
where:
<2a, = flow rate of a-phase (water or oil) in the l-direction
per unit width (lA'T1)
Ta.. = a-phase (water or oil) transmissivity tensor (u = 1 ,2), (LzT-i)
d^fy = air-water table gradient in the j-direction (L L~f)
= air-oil table gradient in the j-direction.
If the gas pressure is nonuniform, the gradient for water flow will be dzaw/dx + dhjdx,
where h& is the water height equivalent air pressure. The gradient for oil flow will be
Bz Idx + (lip )dh Idx. Thus, if a vacuum is imposed at a recovery well, water and oil
GO ^ * FO* Q
removal rates may be increased without resorting to a large drawdown in the fluid
tables in the recovery well. This is commonly referred to as "vacuum-enhanced product
recovery."
Mass conservation for water and oil requires:
BV... d£
(3-31)
dv
o
'O,
(3-32)
+ Jn
dt dXj
where:
Fw and V0 = water and oil volume per unit area (L)
Tw and T0 = vertically integrated source-sink terms (L3L'2T1).
The preceding equations can be solved by examining the expressions presented in
Figure 3-12 for specific volumes and transmissivities as functions of fluid table
eleva-tions.
Oil relative permeability must be determined as a function of elevation. For a given well
oil thickness, vertical oil and water saturation distributions may be evaluated as
previously discussed and oil relative permeability may then be calculated. Figure 3-13
presents oil transmissivity versus free oil-specific volume for gasoline in a sandy soil.
59
-------
Characterization of water-specific volume
where c is the unconfined aquifer specific yied.
Characterization of nil specific volume
V - V f + V + v
o of ot og
where
V0 = total oil per unit area
Vof = free oil per unit area
Vot = trapped oil specific volume
Vog = pendular residual oil specific volume.
Characterization of water transmissivity
zi
where
ZBW- = oil-water table elevation
6OW = oil-water capillary fringe thickness
z, = aquifer effective lower elevation
Characterization of oil transmissivity
r. - p"
where
pro = oil specific gravity
r\TO = oil specific gravity
It,,, = oil relative permeability
Figure 3-12. Vertically integrated relationships.
60
-------
50.00 n
40.00
30.00
2s
O
20.00
10.00
O.CO
VG model
BC model
0.00 1.00 2.00
V0f(m)
3.00
4.00
Figure 3-13. Oil transmissivity versus free-oil-specific volume for a sandy soil
based on the van Genuchten (VG) and Brooks-Corey (BC) models.
3.2.6 Effects of Oil and Water Table Elevation Changes on Free Product
It is important to distinguish between free product (oil) that is mobile and may be detect-
ed in monitoring wells, as opposed to residual oil that is not free to move. Residual oil
occurs as both oil trapped within pores during imbibition (insular residual oil) and as
pendular oil held in the unsaturated zone by capillary forces.
• Oil-specific volume (V0) is the total of Vof, free oil-specific volume, plus V0,,
insular (trapped primarily in the saturated zone) residual oil, plus V^,
pendular residual oil. Vof may be estimated from well oil thickness.
• The residual oil-specific volume in the saturated zone (Vot) results from oil
entrapment when the oil-water table rises, thereby causing water
imbibition as the oil-water capillary pressure decreases. The magnitude
of Vol is pontrolled by:
The magnitude of the rise in the oil-water table (AzoH,) relative to its
historical minimum
61
-------
Maximum residual oil saturation (Sol)
Maximum historical oil thickness (H™*).
An illustrative Vol function is shown in Figure 3-14.
u.iz-
Ortfl _
.Uo
:
- -
0.04-
-
-
-
0.00-
*
*
*
1
*
If ' ' '
/
si 1
!/
/'
c
H0max (ft)
0.60
- - • 0.83
1 20
1.76
0.0
1.0
2.0 3.0 4.0
Figure 3-14. Trapped oil-specific volumes as a function of the oil water table
elevation increase (A zow historical maximum well oil thickness
Unsaturated zone residual oil-specific volume is another source of
residual oil. During periods of falling zao, downward oil redistribution
eventually becomes negligible under gravitational forces as oil saturation
reaches a critical value called the pendular or unsaturated zone residual
saturation. The residual oil-specific volume in the unsaturated zone, V
is described by: ' °9'
V = Min(Vof - K, 4> S' Az )
og no'
(3-33)
62
-------
where
Azao = downward change in the air-oil table as a result of pumping
and/or seasonal fluctuations
S'og = Min (Sog, S0max) in which S^ is the maximum unsaturated
zone residual saturation after drainage from a high oil
content
5'0max = maximum oil saturation at an elevation z=zao for the current
well oil thickness H0
Depending on the direction of water table fluctuations, water table elevations can affect
well product thickness in various ways.
Fluctuations in the water table also have an effect on the thickness of the product
observed in a monitoring well because of deviations from vertical equilibrium. When
water flow is downward, water pressure increases more gradually with depth, and oil-
water capillary pressure will be lower. At a given free oil saturation, this means the well
oil thickness must become greater. Conversely, when water flow is upward, the well oil
thickness will become compressed (Figures 3-15 and 3-16 from Ground Water, 1990).
These effects produce the same trend as that produced by changes in product
thickness; such changes are due to changes in residual saturation associated with
fluctuating water tables and will usually be indistinguishable in practice.
Monitoring Well
Rtprlnted by permission of the Ground Water Publishing Company. Copyright 1990.
Figure 3-15. Hydrocarbon thickness decrease for rising interface (Ground Water,
1990).
3.2.7 Characterization of Soil and Bulk Hydrocarbon Properties
NAPL flow is governed by the fluid and the media through which it is migrating. The
fluid and soil properties that govern NAPL flow are shown in Table 3-1. In this chapter,
a brief discussion will be presented of the methods used to characterize these
properties.
63
-------
Monitoring Well
Reprinted by permission of the Ground Water Publishing Company. Copyright 1990.
Figure 3-16. Hydrocarbon thickness increase for falling interface (Ground Water,
1990).
Table 3-1. Fluid and Soil Properties Governing NAPL Flow
Fluid properties:
pro . Ratio of oil to water density [L°]
Pao Ratio of water surface tension to oil surface tension [L°]
3ow Ratio of water surface tension to oil-water interfacial tension [l_°]
Soil properties:
Ksw Saturated conductivity principal values [L T1]
(J> Total porosity [L°J
Sm Water saturation at field capacity [L°]
Maximum unsaturated zone residual oil saturation [L°]
Maximum saturated zone residual oil saturation [L°]
VG mean pore size parameter [L'1]
VG pore size distribution exponent [L°]
S0
S0
a
n
64
-------
3.2.7.1 Estimation of Fluid Properties
Product-Specific Gravity Oil-specific gravity, pro, will vary significantly for different petro-
leum hydrocarbons depending on the specific chemical composition (see Table 3-2).
Product-specific gravity can be determined as follows:
• Determine a laboratory measurement of the fluid sample collected.
Measurements should be made within 5° to 10°C of the temperature
expected in the field.
• Calculate product-specific gravity from fluid level data.
Table 3-2. Fluid Properties for Various Hydrocarbons
Product
Crude oil
Diesel fuel
Gasoline
Fuel oil No. 1a
Fuel oil No. 2b
Fuel oil No. 4
Fuel oil No. 5
Fuel oil No. 6
Pro
0.70-0.98
0.80-0.85
0.70-0.80
0.81-0.85
0.86-0.90
0.88-0.92
0.92-0.97
0.94-1.05
Mro
8-90
1.1-3.5
0.4-0.8
1.5-2.5
4-9
5-24
53-175
60-150
Pao
2.0-3.5
2.8-3.2
3.0-3.4
2.5-3.0
2.5-3.0
2.5-3.0
2.5-3.0
2.5-3.0
Pow
1.4-2.0
1.4-1.6
1.4-1.5
1.5-1.7
1.5-1.7
1.5-1.7
1.5-1.7
1.5-1.7
" Fuel oil No. 1 = kerosene
b Fuel oil No. 2 « diesel.
A simple field procedure to determine product density in wells with free product is to
measure the water piezometric elevation (zaw) by use of a tube inserted through the oil
layer in the monitoring well and to measure the air-oil and oil-water table elevations
under static conditions. The product density may be computed by:
00 Otf
(3.34)
where it is assumed that equilibrium conditions exist within the well bore and it is best to
wait until fluid levels are stable following the insertion of the piezometer tubes.
65
-------
Fluid Scaling Factors Air-oil and oil-water scaling factors (Bao and B^) must be deter-
mined in order to describe three-phase saturation-capillary pressure relationships. The
scaling factors are estimated from oil surface tension and oil-water interfacial tension
data (Lenhard and Parker, 1987) by use of the following equations:
P~ = °-/0" (3-35)
Pow = °«/°ow (3-36)
where:
ow = surface tension of water (72 dynes/cm)
o0 = surface tension organic liquid
oow = oil-water interfacial tension.
Because interfacial tension is more difficult to measure than surface tension, an alter-
native protocol for determining Bw is to measure the surface tension of water saturated
with dissolved hydrocarbon (i.e., water that has been shaken with hydrocarbon and
decanted to remove all traces of free liquid) and to compute the interfacial tension via
the following equation:
CJ "~ ^J ~~ O ' /O O'T\
«>»• (3-37)
where:
o^ = surface tension of water saturated with dissolved hydrocarbon.
In the absence of measurements of either oow or ow, an approximate value of Bow can
be obtained assuming ow ~ ow, which indicates that:
0 * — (3-38)
ow
Based on surface tension and interfacial tension data for gasolines, Bao=3.2 and
B™* 1.45 ± 10 percent (Weiss, 1990). Approximate values of the fluid properties for
various hydrocarbon products are given in Table 3-2.
For unrefined petroleum hydrocarbons (i.e., crude oil), an estimate of the scaling factors
can be obtained by determining the correlation between oil surface tension and specific
gravity. This correlation is shown by Baker and Swerdloff (1956) as:
66
-------
(3-39>
These equations provide a simple procedure for estimating scaling factors for unrefined
hydrocarbons. The procedures used to estimate surface tension and interfacial
tensions of fluid mixtures have been reviewed by Lyman et al. (1982).
3.2.7.2 Viscosity
Product viscosity can be determined from laboratory measurements of a fluid sample or
it can be estimated based on the type of product present (gasoline, diesel, oil) and on
published values (see Table 3-2).
3.2.7.3 Estimation of Soil Properties
Soil properties include parameters defining the fluid retention properties and soil per-
meability (see Table 3-3). If soil properties exhibit variations in the vertical direction,
parameters relevant to the capillary fringe zone, where most oil occurs, can be used to
accurately predict oil recovery. An under- (or over-) estimate of water transmissivity can
be corrected by adjusting the effective and actual aquifer lower boundaries deeper (or
shallower) in proportion to the error in the aquifer conductivity.
Table 3-3. Representative Soil Properties for Various Soils
Soil type*
Sand
Loamy sand
Sandy loam
Sandy clay loam
Loam
Silty loam
Clay loam
Sandy clay
Silty clay loam
Silty clay
Km
[m/d]
7.1
3.5
1.06
0.31
0.25
0.11
0.062
' 0.029
0.017
0.0048
[-]"
0.43
0.41
0.41
0.39
0.43
0.45
0.41
0.38
0.43
0.36
sm
[-]
0.13
. 0.21
0.24
0.28
0.35
0.43
0.55
0.66
0.68
0.84
a
[m-1]
14.6
12.5
7.6
5.9
3.7
2.2
2.1
3.2
1.2
0.84
n
[-]
2.7
2.4
.2.0
1.5
1.7
1.7
1.7
1.8
1.9
2.8
" U.S. Department of Agriculture (USDA) classification system.
b Dimensionless.
67
-------
3.2.7.4 Field Capacity Water Saturation
The "Field Capacity" water saturation, Sm, can be determined by use of the following
methods:
Measuring water content above the capillary fringe.
Using S-/7aw curves (measured in the laboratory) of soil cores.
Using specific yield, e, in a long-term pump test, to determine
S = 1
Estimating S-/7aw from grain size data and then estimating Sm = Sw at h
1-3 meters.
'aw-~
The parameter Sm represents the minimum water saturation that will occur in the soil
under field conditions. It should be noted that the minimum saturation determined in
the fitting of laboratory moisture retention data will invariably be much smaller than the
minimum field water content because equilibrium conditions are only asymptotically
approached in the field. Estimates of Sm may be made from direct measurements of
the degree of saturation of the soil cores that are taken from the field at elevations
above the "capillary fringe;" this elevation is above the air-water table (or air-oil table if
oil is present) where water saturation drops more or less sharply.
If the specific yield of the unconfined aquifer is known, this parameter can be used to
estimate Sm:
sm=1 - (3-41)
where $ is the total porosity of the soil and e is the specific yield or effective porosity.
Measured specific yields are often observed to increase in relation to the duration of the
pump tests. Because long-term drainage is of concern here (e.g., weeks to months),
specific yields from short-term pump tests may underestimate the "true" effective
porosity.
If laboratory moisture retention data are available, an estimate of Sm may be made by
evaluating the water saturation at an air-water capillary pressure head of 100 cm, which
may be taken as a close approximation of "field capacity" in humid climates.
3.2.7.5 Unsaturated Zone Residual Oil Saturation
The unsaturated zone residual oil saturation, S^, can be determined by the following
methods:
68
-------
• Laboratory column studies
• An estimate based on soil TPH data
• An estimate as a fraction of field capacity.
Laboratory column studies can be conducted to estimate unsaturated zone residual oil
saturation where water in the soil column is drained to the "field capacity" and then the
column is flooded with oil and allowed to drain freely. The oil saturation is then
measured near the top of the column. For columns less than one meter in length,
suction may have to be applied.
Unsaturated zone residual oil saturation can be estimated from the soil TPH data znn as
ao
So _- TPH X V -# *•" * IP" (3_42)
where TPH Is given in mg/kg.
Oil saturation can also be estimated as a fraction of field capacity in the following
empirical relationship:
where^E may range from 0.2 to 0.5.
Fluids with higher viscosities and soils that are more heterogeneous will tend to have
larger f^ values. Theoretical analyses indicate that residual saturation will increase in
approximate proportion to the fourth root of product viscosity; i.e., f^ «nro1M where nro is
the oil-water viscosity ratio.
3.2.7.6 Saturated Zone Residual Oil Saturation
The saturated zone residual oil saturation, Sor, can be determined by the following
methods:
• Laboratory column studies
• An estimate based on soil TPH data
• An estimate as a function of effective porosity.
Laboratory column studies can be conducted to estimate saturated zone residual oil
saturation where the soil column is saturated with water and then is flooded with oil.
The column is then water-flooded to displace the oil. The oil saturation present in the
column would then be measured.
69
-------
Typical values of Sor are given by:
Sor^ford-SJ (3-44)
"^
where/or ranges from 0.2 to 0.5 with a median of about 0.3.
Fluids with higher viscosities and soils that are more heterogeneous will tend to have
larger/or values.
3.2.7.7 Capillary Pressure Parameters
Air-water capillary pressure parameters a, n, and Sm can be determined by the
following methods:
° Estimation from soil cores
0 Estimation from grain size data
Correlation with saturated conductivity
Estimation from TPH data
Estimation from product recovery data.
Capillary Pressure Parameters From Soil Cores Air-water capillary pressure curves arp
often characterized by fitting model parameters (i.e., a, n, and SJ to the water content
versus capillary pressure data obtained in the laboratory on soil cores, which will yield
true equilibrium parameters. In the field, however, equilibrium is never truly attained be-
cause fluid drainage is impeded by low relative permeabilities as wetting phase satura-
tions dimmish. To correct for the deviation from equilibrium conditions, quasi-static
model parameters should be used that yield the correct water saturation distribution
under field conditions when a hydrostatic water pressure distribution is assumed
Procedures used to estimate quasi-static retention parameters from the laboratory data
are described by Lenhard and Parker (1990). The simplest approach is to fix S at a
value corresponding to the minimum field saturation, discard moisture data below about
Sw = 1.1 x Sm, and then fit the parameters a and n to the reduced data set by using a
nonlinear regression method. Typical quasi-static parameters for various soil types are
given in Table 3-3.
Capillary Pressure Parameters From Grain Size Data A theoretical procedure to
estimate air-water capillary pressure parameters (i.e., a, n, and SJ was derived by
Arya and Paris (1981)'based on the proposition that capillary pressure relationships are
related to the pore size distribution of the soil, which may in turn be inferred from the
grain size distribution. The method was calibrated and implemented by Mishra et al
(1988) in the program SOILPROP (ES&T, 1990). The user may either specify Sm as
known (computed, for example, from specific yield), or Sm may be estimated by the
70
-------
program to correspond to "field capacity" for the soil defined operationally as the water
saturation at which the air-water capillary head is 100 cm.
Capillary Pressure Parameters From Saturated Conductivity Correlation Another
method of estimating the mean pore size parameter a is to employ a correlation with
saturated hydraulic conductivity:
(3_45)
Laboratory analysis of vertical conductivity has shown A = 0.5 m3cf1 (±50%). Because
field-measured horizontal conductivities (e.g., from slug or pump tests) are generally
much higher than vertical laboratory values, estimates of a using A=0.5 /773rf1 and field-
measured conductivities may be higher than values estimated from grain size distribu-
tion data. The true field parameter values will probably lie between these estimates.
Capillary Pressure Parameters From TPH Data The most critical parameter in esti-
mating oil saturation distributions and spill volume is generally the capillary curve
parameter a. If independent data are available on oil saturation at points in the field,
these can be used to calibrate the value of a. Because oil saturation can be inferred
from total petroleum hydrocarbon (TPH) data, if the latter measurements and moni-
toring well data are available, they may be used to calibrate a. The method is based on
the premise that oil saturation distributions computed for given well fluid levels from
three-phase saturation relationships should agree with TPH data if the capillary model
is properly calibrated. The method requires as input well product thicknesses H0, and
oil-water table elevations, z^, in monitoring wells at specified coordinates (xw,yw), TPH
measurements from a given depth interval for specified coordinates (xT,yT), and
estimates of total porosity (J>, irreducible water saturation Sm, the van Genuchten
parameter n, oil-specific gravity pro, and fluid-dependent scaling factors &ao and Bow.
The major steps of the algorithm can be summarized as follows:
1. Interpolate z^, and H0 at locations (xT,yT) where TPH is measured.
2. Calculate an average oil saturation 50 from the interpolated fluid levels
over the interval of TPH measurements and use this 3>0 to calculate the
corresponding TPH value.
3. Compare measured and calculated TPH values and iteratively adjust the
value of a to minimize the sums of squares.
2>0 is calculated by averaging S0 computed at midpoint, lower, and upper limits for TPH
measurement intervals. TPH (in ppm) is calculated as:
71
-------
(3-46)
where p0 is the oil density and pb is the soil bulk density.
In Step 3, measured TPH data are used only for intervals within elevations where free
oil saturation is non-zero. This interval is defined by a lower limit z, and an upper limit
3.3 Free Product Control and Recovery
3.3.1 Introduction
The first requirement in spill remediation is to prevent further free product migration.
This generally involves source removal or mitigation and installation of a system of
trenches, sumps, or wells from which free product and sometimes water are skimmed
or pumped.
The following methods are used to contain free product migration:
Skimming or pumping free product from trenches, sumps, and wells
without pumping groundwater can be an effective technique for use on
layers of free product that are relatively static and remain in the vicinity of
the spill or leak.
Trenches installed on the downgradient side of the spill area that only
skim or pump free product can also be effective in mitigating the migration
of the free product layer.
Pumping groundwater along with the skimming or pumping of free product
is an effective approach when a hydraulic control is necessary to move
the free product into the trench, sump, or well and/or to prevent migration
of the free product layer.
Pumping from trenches, sumps, or wells must be carefully controlled to
limit contaminant migration. Increasing water pumping further will
generally diminish recovery because of the increasing volume of residual
product that will become smeared over the cone of depression of the
water table drawdown. Thus, for a given well or trench configuration,
water pumping rates should be determined in order to control product
spreading.
72
-------
Because various pumping configurations can be used to control the plume, additional
criteria must be evaluated to determine the optimum system design. These criteria are
the unit treatment costs and ultimate remediation objectives (removal of free product
only, reduction of soil or dissolved concentrations below threshold values, etc.). The
optimum design involves minimizing water pumping or the duration of the remediation
period, maximizing total product recovery, or maximizing the product recovered per
volume of water pumped (average oil-water cut).
Consideration also must be given to the total water pumping rate and total product
recovery. The analysis of recovery system design involves two steps:
For a given well configuration, pumping rates are determined to control
free product layer movement.
• Recoverable product volume is estimated to determine the design that
yields maximum recovery.
This chapter includes a description of the methodology used in each of these steps of
the design process.
3.3.2 Design Considerations
3.3.2.1 Free Product Migration Control
In the absence of air-pressure gradients, water flow will occur in response to gradients
in the air-water table, zaw, also referred to as the corrected water table.
*«, =
Pro Ho (3-47)
Lateral flow of separate phase hydrocarbons occurs in response to gradients in the air-
oil table elevation. Once the air-water surface is determined for a given well
configuration, the air-oil table (z^) may be estimated.
*ao = 2aw + (l-pro)#o (3-48)
where:
zmvi - post-pumping water table elevation (L)
pro = oil-specific gravity (L°)
H0 = oil thickness location prior to pumping (L).
If oil thickness is small compared to aquifer thickness, water transmissivity is slightly
affected and water flow may be modeled independently of oil.
73
-------
Equation 3-48 yields an approximation of the air-oil table after water flow approaches
steady-state conditions, but before significant oil redistribution has occurred. Because
oil will tend to accumulate at the wells and will gradually diminish in thickness near the
plume perimeter, the oil gradient around the edge of the plume computed by Equation
3-35 should be an upper estimate. If the gradient of zao is flat or inward at all locations
around the plume perimeter, control of plume migration should be assured for the as-
sumed pumping conditions. For effective plume control, the plume perimeter should
first be delineated (e.g., minimum oil thickness contour of 0.01 ft). The first step is to
plot the air-oil table gradient normal to the boundary. The direction and magnitude of
the oil gradient plume control are indicated when all vectors disappear or point inward
around the entire perimeter of the plume.
To determine the product capture zone, any suitable analytical or numerical ground-
water flow model can be used to compute the water table (z^) distribution in response
to water pumping at selected locations at specified rates. Flow vectors or streamlines
can be computed from the water table gradients to determine the direction of ground-
water flow throughout the region with free hydrocarbon. Inward pointing water velo-
cities around the entire plume perimeter indicate control of the dissolved plume.
To ensure control of the free product plume, the air-water table is corrected for oil
thickness to determine the air-oil table using Equation 3-48. Plotting flow vectors or
streamlines of zao will enable an assessment of oil plume control. If vectors of zao are
inward on the entire plume perimeter, free product plume control is indicated.
Figures 3-17a through 3-17c show a hydrocarbon plume (closed contours of H0), air-oil
table contours, and air-oil gradients for a site with a regional flow field (left to right) and
with three wells pumping at 0.2, 0.5, and 0.75 gpm, respectively. Note that the 0.2-gpm
case does not provide plume control on the downgradient (eastern) plume border.
Control is obtained for the 0.5- and 0.75-gpm cases.
3.3.2.2 Estimation of Recoverable Product Volume
Estimates of the volume of free oil recovered from well oil thickness were discussed in
previous subsections. Only a fraction of this volume is recoverable, however, because
of the processes used that lead to the occurrence of residual oil. Residual oil has been
distinguished as in-solar residual oil in the liquid-saturated zone, which occurs as hy-
draulically discontinuous blobs trapped within a continuous water phase, and as
pendular residual oil in the unsaturated zone, which occurs as thin films and as
pendular rings of oil at particle contacts. Recoverable volume, I., is calculated by the
following equation:
y* - y v v
A, - 2-o/ - ^ot-^og (3-49)
74
-------
Graph Models Setup Help
x=962.59 w=828.81 Men=14541O16 Select Well to Adjust Punp Rate
verable volume *
Ftate=0.2D
gallon's
Figure 3-17a.
Pumping rate of 0.20 gpm to control air-oil table gradient on
plume perimeter.
75
-------
' Recoverable volume = 1 a qalloris
Figure 3-17b. Pumping rate of 0.50 gpm to control air-oil table gradient on
plume perimeter.
76
-------
graph HadeIs Setup Help
X=9O2.93 j|=755.7g MaM=146OO76O Select Hell to Adjust Punp Rate
Rate=0.?5 <3P
Recoverable volume'= 29 gallons
Figure 3-17c. Pumping rate of 0.75 gpm to control air-oil table gradient on
plume permieter.
77
-------
where:
Sof = initial free oil volume (L3)
Sot = volume of oil trapped in the unsaturated zone (L3)
Sog = volume of residual oil in the unsaturated zone (L3).
If the spill area is divided into a grid with N blocks of equal area A, the volumes may be
computed as:
^0f = A^Vof. (3-50)
£ *V (3-51)
Eog =A £ r (3-5.2)
/=!
where:
17
initial free oil-specific volume at location / (L)
residual oil-specific volume that may become trapped in the liquid-
saturated zone at location / (L) ^
residual oil-specific volume held against gravitational drainage
in the unsaturated zone at location / (L).
Free oil-specific volumes may be computed from well product thickness data as dis-
cussed in Section 3.2.4. The insular-specific volume is computed as described in
Section 3.2.6 based on estimates of AZOM, computed assuming that oil thickness does
not change much within the period that water pumping reaches steady state. As
product is removed from the formation during steady water pumping, the oil-water table
will increase until well oil thickness (H0) approaches zero. At this point zow = zaw and
A*™ = Pro Hc (3-53)
The value of H™"*, also needed to compute insular residual-specific volume, is taken as
the value of H0 prior to pumping. The pendular residual-specific volume may also be
computed, as described in Section 3.2.6, assuming A*ao = Azow, which is the water
table drawdown.
78
-------
The above estimates of asymptotic recoverable product volume are based on the
assumption that lateral plume spreading is controlled. If this condition is not met, less
product will be recovered because of increases in residual volumes as the plume
spreads.
An example of different pumping rates for product recovery is illustrated in Figures
3-1 7a, 3-1 7b, and 3-1 7c. Note that at the lowest pumping rate of 0.2 gpm, plume
control is not indicated, and thus the assumptions underlying Equations 3-51 through
3-53 are not met and the recoverable volume cannot be estimated. For cases b and c
(pumping 0.5 and 0.75 gpm), the estimated recovery drops from 126 gallons for the 0.5-
gpm case to only 29 gallons for the 0.75-gpm case, thus reflecting greater residual
because of "smearing" in the larger cases of depression caused by pumping. The initial
free oil volume for this problem was about 1,100 gallons prior to the start of recovery.
3.3.2.3 Free Product Migration to Trenches and Sumps
Trenches and sumps are effective in the control of free product migration and the re-
covery of free product. Previous sections covering vertically integrated flow equations
can be used to estimate the rate of oil seepage into a trench. Flow into a large dug
sump is analogous to that of a trench. Figure 3-18 illustrates water and oil distribution
for the special case of a delta-function soil. For this soil, pore size distribution is narrow
and saturation-capillary pressure relationships will be delta-functions resulting in an "oil
pancake" on the water table. In this example, Vof is the free oil-specific volume, T0 is
the oil transmissivity, and Hs is the oil-contaminated thickness of the soil.
If the pore size is narrow, S-P and S-z relationships will be
delta-functions, resulting in an "oil pancake" on the water table.
For the VG model:
6W = I/a and n -»~
For the BC model:
For delta-function soil, soil hydrocarbon thickness
and oil-specific volume are given, respectively, by
Ha = H0
Figure 3-18. "Oil pancake" approximation.
79
-------
The special case of a delta-function soil is characterized by a step-function capillary
pressure curve with an air entry head of ha and a Brooks-Corey exponent A-°°. (Note,
this is the same as the van Genuchten model with hd = I/a and T|-°°.) For this soil, oil
saturation will equal (1 - Sm) over the interval from z^, + 6,^ to zao + 6ao (see Equations
3-27 and 3-28), giving a thickness of Hs = H0 + 6ao - 6^ in which effective oil saturation
is unity (Sy(1 - Sm). The free-oil-specific volume Vof will be equal to 0(1 - SJHS, and oil
transmissivity 7~0 will be equal to
The oil flux into the trench per unit length will be equal to T0 grad zao. If oil is removed
from the trench, the oil transmissivity on the downstream trench face will be controlled
by the product thickness remaining in the trench: Note that free-oil-specific volume and
transmissivity will be zero if the oil thickness in the trench is less than 6^. Thus, if a
trench is used to recover product, and the product level in the trench is maintained
below 5^, product will not enter the trench on the downgradient side.
3.3.2.4 Example Calculations
Example 1 To illustrate the calculations discussed above, soil concentration data for
samples taken from a soil boring to a depth of 45 feet are used in a spreadsheet form
given in Table 3-4 to compute total hydrocarbon-specific volume, benzene mass per
area, and volume of soil per area with TPH greater than 1000 mg/kg. Given multiple
soil borings, similar calculations using Equations 3-50 through 3-52 can be repeated for
other locations and the results interpolated over a spatial domain to determine total
hydrocarbon volume, benzene mass, and contaminated soil volume within the sampled
region. Interpolation may be carried out using commercial or public-domain software
(e.g., GEO-EAS or on a regular grid Surfer). The soil under consideration is assumed
to have a porosity of 0.35 and a bulk density of 1.72 g/cm3. The hydrocarbon density is
assumed to be 0.80 g/cm3. Values in Table 3-4 were computed as follows:
Column A. Sample depth is the distance from the ground surface to the center of the
core sample.
Table 3-4. Example Spreadsheet Calculations from Soil Boring Data _
H
Depth (ft)
5
15
21
28
37
45
dZ
00
5.5
8.0
6.5
8.0
8.5
4.5
TPH
(nw/ke)
542
1180
3937
6836
678
27
6°.
0.0012
0.0025
0.0085
0.0147
0.0015
0.0001
V0
(ftVft)
0.006
0.020
0.055
0.118
0.012
0.000
0.206
Y.
(me/kg)
5
9
37
53
6
0
/P>Y.
(B/ft1)
1.3
3.5
11.7
20.7
2.5
1.2
39.8
«Z)
0
1
1
1
0
0
6(Z)dZ
(fP/fO
o.o
8.0
6.5
8.0
0.0
0.000
22.5
80
-------
Column B. The sample interval, dZ, is half the distance from the current sample to the
next shallower sample, plus half the distance from the current sample to the
next deeper sample, except for the shallowest and deepest samples, in
which case, the interval is half the distance from the sample center to the
next sample, plus 0.5 ft to account for one half of the actual length of the
sample core.
Column C. Measured TPH in the soil core is given as mg hydrocarbon per kg dry soil.
Column D. The average volumetric oil content in the sample interval is computed from
Equation 3-46.
Column E. Each entry in the column is calculated as 60dZ, and the entire column is
summed to obtain the oil-specific volume, V0 = 0.206 ft3 per ft2.
Column F. Measured soil benzene concentration is expressed as mg benzene per kg
dry soil.
Column G. The average benzene mass per area in g/ft2 in the sample interval is
computed from:
(3-54)
where Ya is the soil concentration of species a, z, and zu are the lower and upper
elevations where the contaminants occur, and the factor / = 0.0283 is inserted to make
the proper unit conversions. The column is summed to obtain the mass of benzene per
area over the boring depth, mtonz = 39.8 g/ft2.
Column H. The indicator variable is 1 if TPH <; 1000 mg/kg and 0 if TPH is smaller.
Column I. The volume of contaminated soil per area in each sample interval is
calculated as 5(Z)dZ. The column is summed to obtain the contaminated
soil volume of 22.5 ft3 per ft2.
Example 2 An example calculation is shown in spreadsheet form to compute free-oil-
specific volume, V0/, from well product thickness, H0, in a monitoring well. The well
product thickness for the example is 3 feet and the air-oil and oil-water table elevations
are 93 and 90 feet, respectively. Assumed oil and fluid properties for the problem are
given in Table 3-5. Vertical integration is performed from a lower elevation of Z^, = 90
feet to an upper elevation of Zu = 93.4 feet, computed from Equation 3-27. Calculations
are performed at 35 equal-depth intervals of 0.1 foot to numerically integrate for oil-
81
-------
specific volume. To compute total free-oil-volume, calculations of oil-specific volume
must be repeated at various areal locations. Because wjell product thickness must be a
smooth function in space (discontinuities in the piezome
while oil-specific volume may be discontinuous as a result of soil heterogeneity, it is
preferable to interpolate H0 from monitoring wells onto a computational grid and to
compute V0/ on the grid from interpolated apparent thickness values. The values in
Table 3-6 were computed as follows:
Table 3-5. Soil and Fluid Properties for Example Problem
ric gradients cannot occur),
Pro = 0.8
Hro = 0.6
3ao = 3.2
Pow=1.5
= 8.0 ft day1
4> = 0.35
a = 2.5 ft'1
n=2.0
Sm = 0.15
Sog = 0.06
Sor = 0.20
Column A. The first value of Z is Zu from Equation 3-27,
the final value is Z^, and
intermediate values are incremented in intervals of dZ = 0.1 ft.
Column B. The oil-water capillary pressure h^ is calculated as h^ = (1-pro) (Z-Z^).
Column C. Air-oil capillary pressure is calculated using Equation 3-22 above Zao. At
lower elevations, hao = 0 is employed to compute total liquid saturation.
Column D. Water saturation is calculated using Equation 3-7 given h^ at the specific
depth.
Column E. Free-oil saturation is calculated for each dep
hasS0/ = St-Sw.
Column F. The free-oil volume per unit area for each de Dth interval is computed as
(|>S0/dZ. A sum of all values in Column F gives the free-oil-specific volume
of 0.309 ft3/ft2.
3.3.3 Systems and Equipment
The previous sections have dealt primarily with the soil, clil, and water characteristics in
the aquifer or formation that affect the migration and recovery of free product. This
section presents a discussion on the basic systems and equipment that are used to
recover free product. '
Free-product recovery techniques can be grouped into three basic approaches:
Recovery of free product from open excavations
Recovery of free product from trench and sjjmp systems
Recovery of free product from wells.
82
-------
Each of these basic approaches is presented in this section. An overview of the
different applications and design considerations for equipment commonly used in each
approach will be discussed.
Table 3-6. Spreadsheet for Free-Oil-Specific Volume from Well Product
Thickness
A
Z
(ft)
93.4
93.3
93.2
93.1
93.0
92.9
92.8
92.7
92.6
92.5
92.4
92.3
92.2
92.1
92.0
91.9
91.8
91.7
91.6
91.5
91.4
91.3
91.2
91.1
continued)
B
how
(ft)
0.68
0.66
0.64
0.62
0.60
0.58
0.56
0.54
0.52
0.50
0.48
0.46
0.44
0.42
0.40
0.38
0.36
0.34
0.32
0.30
0.28
0.26
0.24
0.22
C
hao
(ft)
0.32
0.24
0.16
0.08
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
D
sw
(-T
0.46
0.47
0.48
0.49
0.50
0.51
0.52
0.53
0.54
0.55
0.56
0.58
0.59
0.61
0.62
0.64
0.66
0.67
0.69
0.72
0.74
0.76
0.78
0.81
83
E-
So/
(-)
0.00
0.08
0.20
0.38
0.50
0.49
0.48
0.47
0.46
0.45
0.44
0.42
0.41
0.39
0.38
0.36
0.34
0.33
0.31
0.28
0.26
0.24
0.22
0.19
F
4>S0/dZ
(ftVfP)
0.000
0.003
0.007
0.013
0.018
0.017
0.017
0.017
0.016
0.016
0.015
0.015
0.014
0.014
0.013
0.013
0.012
0.011
0.011
0.010
0.009
0.008
0.008
0.007
-------
Table 3-6. (continued)
A
Z
(ft)
91.0
90.9
90.8
90.7
90.6
90.5
90.4
90.3
90.2
90.1
90.0
B
how
(ft)
0.20
0.18
0.16
0.14
0.14
0.10
0.08
0.06
0.04
0.02
0.00
C
(ft)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
D
(%
0.83
0.85
0.88
0.90
0.93
0.95
0.96
0.98
0.99
1.00
1.00
E
So/
0.17
0.15
0.12
0.10
0.07
0.05
0.04
0.02
0.01
0.00
0.00
F
4>S0/dZ
(te/ft2)
0.006
0.005
0.004
0.003
0.003
0.002
0.001
0.001
0.000
0.000
0.000
0.309
3.3.3.1 Open Excavations
Free product is usually recovered from open excavations during tank, soil, or pipeline
removal or replacement at a petroleum-handling facility. Free product is removed by
one of the following ways:
Vacuum truck
Floating skimmers
Portable trash pumps
Absorbent booms
Other direct removal techniques.
These techniques can effectively remove most of the spilled or leaked product at many
sites. In addition, the excavation and treatment of accessible soil laden with product
can significantly reduce the mass of petroleum remaining in the subsurface.
3.3.3.2 Trench and Sump Systems
Trench and sump systems can effectively recover free product that occurs as follows:
At shallow depths(<15 feet)
Pooled or floating on the grounclwater table
Perched above a low hydraulic permeability layer.
84
-------
Sumps are often installed in the old tank pit excavations to recover NAPL not removed
during initial excavation. In very permeable sands and gravel (>10'2 cm/sec) where
there is considerable slope to the water table, trenches should be placed on the down-
gradient edge of the free product area to intercept the migrating NAPL. In less-
permeable soils, the sumps or trenches are often placed in the middle of the quasi-
static layer to provide for the most efficient product recovery.
Trench Construction and Installation Trenches and sumps are installed in excavations
of at least three to five feet below the top of the free product layer and several feet
below the expected lowest seasonal fluctuation of the water table or to the geologic
barrier that is perching the hydrocarbons. Care must be taken not to penetrate the
geologic barrier, which would allow uncontrolled downward migration of liquid hydro-
carbons. A typical trench system (Figure 3-19) consists of the following:
• One or more layers of perforated pipe, often installed at the water
table/free product interface.
• One or more vertical standpipes or sumps installed with the trench to
remove the product and water.
• Trench excavation backfilling with appropriate fill and cover material.
A sump recovery system, which is installed in an existing or excavated hole, consists of
the following:
• Galvanized perforated pipe (24- to 36-inch-diameter) placed vertically in
the excavation.
• Backfilling in the area around the pipe with gravel to above the upper
elevation for the NAPL layer.
Excavation backfilling with appropriate fill and cover material. •
Product is removed from trenches or sumps by routine manual skimming or with the aid
of various removal equipment.
3.3.3.3 Applications
Trench and sump systems are applicable to a wide variety of hydrogeologic settings,
and the only major limitations are the depth to which they can be installed and the
availability of space for the installation. Trenches can be used successfully for recoveiy
in the following:
Heterogeneous earth materials in which fluid migrates randomly through
placed zones such as discontinuous layers of sand.
85
-------
Plan View
Hydrocarbon sourc
Free liquid hydrocarbon
Umr (optional)
Itooovecy wat or sump
Sand or gravel
Groundwater flew
Reprinted courtesy of the American Petroleum Institute.
API Publication 1628, "A Guide to the Assessment and Remediation of Underground Petroleum Releases," Second Edition August, 1989.
The second edition is currently under revision, for information concerning the third edition contact the American Petroleum Institute.
Figure 3-19. Typical trench system.
86
-------
Areas with shallow water tables and relatively low hydraulic conductivity
where interception of liquid hydrocarbon in trenches is a more practical
alternative than numerous closely spaced recovery wells.
Areas where the saturated thickness of the area aquifer is minimal,
making wells ineffective, such as zones along rivers that may be dry
during part of the year.
3.3.3.4 Recovery Well Systems
Recovery wells are useful for both shallow and deep free-product removal where long-
term pumping or skimming is required. For depths greater than 10 to 15 feet, wells are
generally a less-expensive and more flexible alternative to trenches and sumps. At a
given site, if the initial recovery well or wells prove inadequate for controlling and re-
moving the product layer plume, then additional wells can be added with less expense
and site disruption than additional trenches or sumps. Basic component considerations
include well diameter, placement depth, and pumping rates required for control and
recovery of the product layer plume.
Well Diameter The recommended diameter of recovery wells is 4 inches, and a larger
diameter is considered even more efficient. The effectiveness of a well for removing
product, however, is often more related to the amount of silt and clay in the formation
and the manner in which the well is installed. At some sites, more free product has
been successfully removed from 2-inch diameter wells than from large 24-inch-diameter
wells. Nevertheless, 4-inch or larger diameters are preferable simply because of the
room required to install and adjust skimming or pumping equipment. Regardless of the
diameter, though, care should be taken during well construction that the drilling process
does not reduce the permeability of the product/water table zone. During well opera-
tion, the drawdown should be minimized to prevent fouling and plugging of the annular
space around the well screen and the subsequent reduction in the flow of free product
into the well.
Placement Depth Free product recovery wells are designed to remove free product
with minimal water pumping because water treatment greatly increases the cost of the
recovery system operation. Free product control and recovery is governed by the air-oil
table gradient, which is in turn related to the air-water table gradient and product thick-
ness. At a given pumping rate, the drawdown in the air-water table will increase as the
fractional penetration of the well in the aquifer diminishes. Therefore, to maximize
product recovery per volume of water pumped, product recovery wells should be
screened over the shallowest depth possible to maintain plume control. In most cases,
the bottom screen elevation should be less than 10 to 15 feet below the lowest
anticipated level of the pre-pumping water table elevation. Groundwater models can be
used to evaluate the effects of partial well penetration and pumping rate on water table
draw-down. When dissolved plume control is a factor in system design, placement
depth will be controlled by the depth of the soluble plume. Because greater well
87
-------
placement depths may be required to achieve dissolved plume control, a well system
that is optimal for product recovery may be inadequate for dissolved plume control, and
a system that is adequate for dissolved plume control may not be ideal for free product
recovery.
Areal Placement and Operation of Wells The optimal number and placement of wells
will depend on the areal distribution and thickness of the free product plume, perme-
ability of the aquifer, regional flow gradient, and the residual product saturations that v/ill
remain in the saturated and unsaturated zones. Maximum product recovery will be ob-
tained by minimizing the total drawdown over the zone of the free product plume, while
maintaining plume control around the plume perimeter. For the same total water pump-
ing rate, product recovery will generally increase with the number of wells in operation.
The economically optimum number of wells will depend on the tradeoff between the
cost of well installation and operation versus the benefit gained by reducing the residual
product volume. Regional flow gradients and soil anisotropy will distort the zone of in-
fluence to ellipsoids. A systematic approach to evaluating the effects of well placement
and well numbers on product recovery was described in Section 3.3.2.
3.3.3.5 Recovery System Selection
The selection of an appropriate recovery well system will depend on a variety of
interrelated factors:
Depth to groundwater and free product
Age, extent, and mobility of product
Soil and fluid properties governing free product migration (see Table 3-1)
Regional hydraulic gradient
Physical setting (area available for trenches, wells, and pumps, etc.)
Water and air handling and treatment requirements (oil/water separation,
air stripping, carbon treatment, thermal treatment, etc.)
Chemical properties of the water and product that affect the degree or
potential for scaling, corrosion, or fouling of the system.
Recovery system selection also depends on the type of recovery program being imple-
mented, and the equipment and facilities available at the location. For example, an
acceptable and cost-effective alternative for an oil refinery with available hydrocarbon
separation and water-handling facilities may not be acceptable at a retail service station
where space and fluid-handling facilities are limited. The type of liquid hydrocarbon and
88
-------
quality of the produced water will also have a major impact on the type of system
selected.
3.3.3.6 Recovery Equipment
The following equipment is used to remove NAPL from excavations, trenches, sumps,
and wells:
Direct-removal equipment (skimming, pumping, absorbing)
• Skimming equipment
Single-pump equipment
Dual-pump equipment
• Vacuum-enhanced recovery equipment.
Table 3-7 provides a summary of these techniques along with design considerations,
advantages, and disadvantages. The applications, limitations, and design consider-
ations for each of these systems will be briefly discussed below. Common operating
ranges of each pumping system are shown in Table 3-8.
3.3.3.7 Direct Removal
Direct-removal techniques can be used in many product recovery situations. The most
common technique is the use of a vacuum truck to suck water and product from ex-
cavations, sumps, or wells. At some sites the existing wells are periodically pumped
out by a vacuum truck until the free product is removed to a thin layer or sheen.
Periodic hand bailing of wells or sumps is also a viable technique for use on small
localized free-product layers. Absorbent pads and booms are often used in open
excavations and to swab out wells with very thin layers of material. Dedicated product
skimming bailers and absorbent tubes are used as part of a skimming system, which is
discussed further below.
Skimming Equipment The three basic types of skimming equipment are:
• Floating (large saucer type or small float type)
• Floating inlet (floating filter intake within a pump body)
• Absorbent skimming (absorbent in a dedicated bailer and belt skimmer).
Skimming equipment, which is designed to remove product floating on the water table,
operates at generally low removal rates and produces little or no water. Skimming
equipment is available in the following designs.
Large floating skimmers These skimmers, which can remove product at a fairly high
rate (up to 5 gpm), have a large oliophilic screen that allows product into the pump body
but not water. They are generally limited to shallow (<20 feet) applications and require
a 24-inch-diameter area or more.
89
-------
I
I
5
CU
a
I
•o
CO
CO
£
I
o
O
•o
V
15
o
&
•o
8
a.
o
£
u.
j)
JD
&•
) c
\u
J —
«5 i
l-s
I E.S
O c Q
ed radius of
nce, clogging
screen, general
limited to shallo
(<25 ft) applicat
55
55
55
55
55
5°
X X
T T
O O
X X
A TV
°°
X X
T
O
x
in
A
x
^»
A
x x
A A
U)
Floating
- Large Saucer
- Small Float Ty
C S t
*
-------
*|l
3&S
If I
o
at
.,1
.
fl
1 =
111
CO
2
CO
e-S
il
o t=
111
flfflfs
ftjkffl
lllilll
O 1 sL& 2 £.2
511
ctS
1!
i°
fiiifl
if
— ±J
Hf 111 Ilfl
§
a>
« S « ?• § m
_ ~ m = •" "D S
^.i!
J3 10 !
If*
Its
ID
0)
c
33
O
O
T
O
X
/T
X X
^~ *^
A A
s =
Q- 0)
O is
£*
if
IU
I
si
is-s
t3 « c
» i-0
m I E
I f
> S J= >>
§ w.S)e5
JSX5
II II II li
(D
I
en
te fe
•-STSB «
l
S
To
8
tot
S"^ g1
ifi
1 ss
tjf
°s°
o *• in
ves1
n ii H
-JSX
8
10
o
o E
8.1
a -
I Ii
i Jg;
en
H H II il
°-i*>
r-
W A
-j5:c
3
i1
§
o
"
§
ro
§ S
s.S-
o CN
>r- v>
«* A
II II II
o
To
I
O
91
-------
Table 3-8. Pumping System Versus Common Operational Range
Pump Type
Skimming
Down hole
Suction lift
Vacuum-en hanced
Shallow
Deep
Pneumatic single pump
Submersible
Suction lift
Electric single pump
Submersible
Suction lift
Two-pump systems
Submersible electric
Submersible pneumatic
Suction lift
Fluid Production Rate per Well
Low
(<5 gpm)
Medium
(5-20 gpm)
.
High
(>2O gpm)
'
Reprinted courtesy of the American Petroleum Institute.
API Publication 1628, "A Guide to the Assessment and Remediation of Underground Petroleum Releases," Second Edition August, 1989.
The second edition is currently under revision, for information concerning the third edition contact the American Petroleum Institute.
92
-------
Small float systems These systems, which can work in 4-inch or larger wells, are also
limited to depths of 30 feet or less. The floating inlet skimmers use a floating screen
inlet to capture the product and are contained in a pump device or a bailer. These
generally require a 4-inch-diameter or larger well to operate.
Absorbent skimming bailers These bailers, which are very simple, are hung in the well
across the free-product layer. The absorbent material is periodically removed and
disposed of.
Belt absorbent skimmers These skimmers use a loop of material that slowly moves
down into and out of the well, picking up product as it loops through the water level
surface in the well. These systems are very simple mechanical systems that can work
in 4-inch or larger wells, but are perhaps best suited for skimming sumps.
The following single-pump systems are designed to skim free product layers:
• Pneumatic skimming systems with a top intake that skims fluids from the
liquid hydrocarbon/water interface (Figure 3-20).
Pneumatic skimming systems with a density-sensitive float valve that
allows water to pass before the valve seats.
• Floating or depth-controlled skimming systems with conductivity sensors
that activate the surface-mounted pump when liquid hydrocarbons have
accumulated to a sufficient thickness.
• Filter skimming systems with a filter material that preferentially passes
hydrocarbons.
3.3.3.8 Applications
Because fluid is slowly extracted by a skimming system from the aquifer, only a minimal
depression occurs in the air-oil table and the area of influence is minimal within which
hydrocarbons are directed toward the well.
Skimming equipment is used in the following situations:
In trenches, sumps, or wells
In formations with low hydraulic conductivity (<10~2 cm/sec)
Where water table fluctuations are large (> 1 meter or 3 feet).
3.3.3.9 Design Considerations
The following factors should be considered for selecting or designing skimming
equipment:
93
-------
Hydrocarbon/
water separator
AJr Supply
Free hydrocarbon layer
V
Hydrocarbon/water contact
Hydrocarbon
discharge Vne
Afr supply
and exhaust ine
Reprinted courtesy of the American Petroleum Institute.
API Publication 1628, "A Guide to the'Assessment and Remediation of Underground Petroleum Releases," Second Edition August, 1989
The second edition is currently under revision, for information concerning the third edition contact the American Petroleum Institute.
Figure 3-20. Pneumatic skimming pump.
94
-------
• Simple operation and maintenance
• Depth limitations on some of the skimming equipment
• Fouling of oliophilic or filter screens
Little or no water produced
• Generally inexpensive to install and operate.
3.3.3.10 Single-Pump Equipment
Single-pump equipment produces both water and hydrocarbons from a single intake. In
some cases, as discussed above, it can be used primarily to remove hydrocarbons.
This equipment may consist of either single or multiple wells in which several wells are
manifolded together to a single-surface-mounted pump or single common treatment
area. The types of pumps that can be used include diaphragm, centrifugal, submer-
sible, and pneumatic. Examples of single-pump systems are provided in Figure 3-21.
3.3.3.11 Applications
Because of the costs associated with the separation and treatment of dissolved hydro-
carbons, single-pump equipment is normally limited to areas of medium- to low-
hydraulic conductivity where the volume of produced water is small. The application
and limitations of single-pump systems include the following:
• Creates a larger capture zone through depression of the water table.
Useful in formations of low to moderate permeability (10"4 to 10'3 cm/sec).
• Less expensive to install than dual-pump equipment (total costs may be
higher after installation of an air compressor, separator, and treatment
system).
• Requires an oil-water separator at the surface.
Increases the dissolved hydrocarbon component in the pumped water.
• Tends to emulsify hydrocarbons in the water.
3.3.3.12 Design Considerations
For the selection and design of single-pumps, the following factors should be
considered:
• Simple operation and maintenance.
Can operate at flow rates from as low as 0.1 gpm to over 20 gpm.
95
-------
Pump control witch
Hydrocarbon/
water separator
Backfill/grout
Bentonite seal
Reprinted courtesy of the American Petroleum Institute.
API Publication 1628, "A Guide to the Assessment and Remediation of Underground Petroleum Releases," Second Edition August, 1989.
The second edition is currently under revision, for Information concerning the third edition contact the American Petroleum Institute.
Figure 3-21. Single-pump system.
96
-------
Pumps in contact with hydrocarbons require hydrocarbon-resistant seals
and appropriate ratings for explosive environments.
3.3.3.13 Vacuum-Enhanced Recovery Equipment
Vacuum-enhanced recovery equipment use a surface-mounted vacuum pumping
system (vacuum pump or blower) with an in-well pump to simultaneously remove soil
vapors, NAPL, and water from a recovery well. Vacuum-enhanced systems are prin-
cipally designed to operate in medium- to low-permeability soils (10"3to 10~5 cm/sec),
where high gradients may be necessary to achieve reasonable product recovery rates.
To achieve such gradients by means of water pumping, large drawdowns at the well
bore may be required. Large drawdowns, however, will result in large amounts of resi-
dual product becoming tied up in the cone of depression, thereby limiting the ultimate
recovery. A means of overcoming this problem is to simultaneously pump water and
soil air from the well (or from adjacent wells). The procedure is similar to operating a
venting well (to be discussed in detail in Chapter 4) during the pumping of water and
product, except that the system is designed to maximize product recovery rather than
gas flow. The effect of the vacuum is to reduce the gas pressure to below atmospheric
pressure for a radius around the well. Decreasing the air pressure has the effect of
decreasing the air-oil capillary pressure, which will raise the plane of zero air-oil capil-
lary pressure. Because liquid saturation is controlled by the air-oil capillary pressure, a
large drawdown in the air-oil capillary fringe is implemented while the oil pressure
gradient to the well is enhanced, thus increasing the oil recovery rate. Examples of
vacuum-enhanced pumping equipment are shown in Figure 3-22.
3.3.3.14 Applications
The applicability of vacuum-enhanced recovery equipment can be summarized as
follows:
• Can significantly speed up recovery rates and reduce the time for site
remediation.
• May be used in hydrogeologic settings where the effectiveness of single-
pump systems is limited.
• Useful in areas of medium- to low-hydraulic conductivity (10'3 to 10'5
cm/sec) or where saturated zones are thin.
• Can also be used in high-hydraulic conductivity (< 10~3 cm/sec) areas for
controlled skimming.
Used at depths greater than 5 feet.
* Can use 2-inch or larger wells and is easily adapted to the existing
monitoring well network.
97
-------
CO
o
1.
.0
Q.
E
O)
"55
o
CO
3
u
CNI
CO
I
O)
E
98
-------
Requires oil-water separator at surface.
• Higher initial capital cost than some alternatives, but shorter remediation
time may result in lower overall life-cycle costs.
3.3.3.15 Design Considerations
When the use of vacuum-enhanced recovery equipment is evaluated, the following
factors should be considered:
May require permits and/or treatment for discharge air from vacuum
system.
• Requires sealed well system.
• Requires optimization of vacuum and pumping rates to maximize radius of
influence and recovery of product while minimizing total fluid requiring
treatment.
Proper screening interval is a minimum of 5 feet above and 10 feet below
the water table.
3.3.3.16 Two-Pump Equipment
Dual-pump equipment employs separate pumps for product recovery and water table
depression. Water is withdrawn from a pump placed near the bottom of the well to
create a cone of depression as a capture zone, and the product is removed from a
pump with an intake placed slightly below the hydrocarbon-water interface. Originally,
two-pump systems were designed with a water level probe that turned the groundwater
pump on and off to maintain a constant water level in the well. Field experience has
shown, however, that this technique rarely works because of significant mechanical and
electrical problems. Therefore, proper design and sizing of a continuously operating
groundwater system is recommended. In place of a product pump, a skimming device
such as a floating inlet pump float or floating skimmer can be used to remove the free
product. An example of a dual-pump, dual-probe system is shown in Figure 3-23.
3.3.3.17 Applications
The following factors should be considered for dual-pump applications:
• Where water table depression is necessary for recovery.
• When hydraulic conductivities are high (> 10'2 cm/sec) and saturated
thicknesses are large (> 1.0 m or 3.0 ft).
• Product and water are separated in the well.
99
-------
Water discharge -^
Hydrocarbon storage
Hydrocarbon pump
ouoUotc
Hydrocarbon
detection probe
Hydrocarbon
detection probe
Water pump
Fitter pack
Reprinted courtesy of the American Petroleum Institute.
API Publication 1628, "A Guide to the Assessment and Remediation of Underground Petroleum Releases," Second Edition August, 1989.
The second edition Is currently under revision, for information concerning the third edition contact the American Petroleum Institute.
Figure 3-23. Two-pump systems.
100
-------
Higher capital, operation, and maintenance costs.
Initial startup and adjustments require experienced personnel.
Systems require careful monitoring to maximize efficiency of removal and
prevent mixing of water and oil wastes.
3.3.3.18 Design Considerations
For the design of dual-pump systems, the following factors impact system effectiveness
and cost:
• Pump separation should be maximized to minimize dissolved hydrocarbon
component.
• Pumping rates and placement in well must be balanced to maintain
hydrocarbon-water interface at a constant level to the extent possible.
initial setup must be done by experienced personnel. Setting and running
the groundwater pump at a constant rate is the preferred mode of
operation.
• Requires larger diameter wells than single-pump systems.
• Figure 3-24 shows a typical recovery system capture zone for a two-pump
system.
3.3.4 System Operation and Monitoring Requirements
After the design and installation of a recovery system, the operating system must be
monitored to enable adjustments to be made to maintain system effectiveness.
Periodic measurements should be made of the following parameters:
• Dissolved concentrations of influent and effluent from the water-treatment
system to verify function of the treatment system.
• Oil and water levels in monitoring wells.
• Fluid levels (pump switch levels) in recovery wells and trenches.
Total wa,ter and total product pumped for the period to determine recovery
system effectiveness. Gradual reductions in product recovery and pro-
duct thickness are indications that the system is functioning properly.
Table 3-9 lists potential problems and solutions encountered in the operation of
recovery systems.
101
-------
0)
o
N
£
I
TO
U
£
CO
0)
o
o
Cjl
CO
102
-------
Table 3-9. Potential Problems and Solutions During Recovery System Operation
Problem
Verify
Solution
Physical/biological
clogging of well resulting
in low drawdown in
formation due to large
well loss and possible loss
of plume control.
Regional rising or falling
water table resulting in
reduced free product and
increased residual
product.
Pumping rate too high
resulting in excessive
drawdown tying up large
amounts of residual
product in the cone of
depression.
Loss of gradient due to
inadequate recharge to
sustain pumping rates
over area of influence of
recovery system.
Check air-water table in
monitoring wells to
determine if there is
adequate drawdown.
Check changes in air-
water table elevations for
monitoring wells for
measurement period.
Check air-water table in
monitoring wells to
determine if there is
excessive drawdown.
Check current air-water
table distribution and
changes over time to
determine if gradient is
becoming increasingly flat.
Lower pump level in
well, recondition well
using chemical
backflushing, install
new wells.
Change pump levels to
maintain plume control.
Reduce pumping rates,
install additional
recovery wells if
needed to maintain
plume control.
Install recharge
galleries.
3.3.4.1 Regulatory Considerations
Regulatory constraints can delay the full operation of a recovery system. Some of
these constraints involve federal, state, and local regulations pertaining to the following
permits:
• Permits for wells. In states that require well permits, one of the primary
purposes is to develop a database of existing wells and statewide hydro-
geology.- The permit process itself is generally straightforward, relatively
simple, and rapid.
• Discharge permits (National Pollutant Discharge Elimination System)
[NPDES] for pump water. Disposal of pumped water, on the other hand,
can be a very difficult issue. Effluent concentrations must be less than
either the Federal Maximum Contaminant Level, the applicable Method
103
-------
Detection Limit, or a site-specific risk-based criterion. The specific
criterion is dependent on the applicable Federal, State, or local
regulations.
Permits for injection wells utilizing hydrodynamic control. Injection well
permits are in most instances subject to the same permitting issues as
discussed above for disposal of pump water. Aquifer hydraulics (trans-
missivity, storativity, well yield, etc.) also need to be understood prior to
the operation of an injection system.
• Air discharge permits for air strippers and catalytic combustion. For large
systems, the air permitting process may be time-consuming and costly in
order to meet requirements of the Clean Air Act of 1990. Some states do
not require air discharge permits for smaller systems.
As an interim measure, skimming may be used to remove free product from recovery
wells and thereby reduce the spread of contaminants. The successful removal of free
product at many sites has been limited to technical difficulties encountered in locating
free product in the subsurface and in removing product through recovery wells. At shal-
low depths, strategically located trenches may obviate the need for pumping. In addi-
tion, Resource Conservation and Recovery Act (RCRA) and other regulations may
apply to the disposal of recovered product or residues from the treatment of
contaminated water or air (e.g., spent activated carbon). If recovered free product can
be recycled for reuse, it is not considered a RCRA hazardous waste under Subtitle C
regulations. On-site storage of recovered product may also be subject to state or local
fire codes.
3.3.5 Side Waste Stream Treatment
A partial list of available treatment systems for remediating spilled hydrocarbons
includes:
Separators
° Air stripping
° Bioreactors
Carbon adsorption
Catalytic combustion.
3.3.5.1 Separators ,
Hydrocarbon/water separators are often used in the initial treatment phase. Separators
allow the free-phase NAPL to separate from the water and be skimmed off the top while
the recovered water is pumped from the bottom. Most separators are equipped with
vertical tube coalescing tubes designed to agglomerate small entrained oil particles to
larger particles for effective gravity separation. After oil surfaces inside the separator,
the oil is skimmed by a rotary skimmer and transferred to an external storage tank. The
104
-------
greater the difference between the specific gravities of the oil and water, the faster the
rate of separation. Nonemulsified oil is effectively removed down to about 10 mg/liter.
3.3.5.2 Air Stripping
Air stripping is a technique used to remove hydrocarbons from water by transferring
contaminants to an air stream. Volatile organic compounds having Henry's law con-
stants above 0.01 are readily air-strippable at ambient temperatures (EPA, 1993a). A
stripping tower consists of a tall shell filled with either packing material or a series of
perforated plates to promote contact between the air and water streams. The water
stream is introduced at the top through a spray nozzle or distributor. Air is blown in
from the bottom of the tower. As the streams pass each other, the hydrocarbons are
transferred to the air stream, which exits the top of the tower. The cleaned water
leaves from the bottom. If the required effluent concentration in the water is very low, a
carbon bed may be needed to adsorb any remaining hydrocarbons from the exit water.
The primary factors affecting stripper design are:
Characteristics of tower packing, typically 1 to 12 feet in diameter and 5 to
50 feet in height
• Expected water flow rate, typically 5 to 30 gpm/ft2
• Contaminant type and concentration
• Required water effluent concentration.
3.3.5.3 Bioreactors
Bioreactors are used to microbiologically degrade hydrocarbons in a liquid or slurry. A
typical bioreactor involves a suspended-growth activated sludge system, either as a
continuous-flow system (CFS) or as a periodic process called a sequencing batch
reactor (SBR). In either case, a controlled environment is created that is suited to the
maximum degree of biodegradation obtainable. It is generally accepted that most
"hazardous organics" can be treated biologically if a suitable substrate and group of
organisms can be established. In a CFS, pretreated water enters a completely mixed
bioreactor suitable for organism growth and substrate removal. Biomass is separated
from the treated effluent in a clarifier and then returned to the bioreactor to maintain
system operation. Excess biomass is processed as sludge. Clarified effluent is final
treated if necessary. l,n an SBR, a single tank is used to accomplish the same functions
that a conventional CFS carries out in a series of separate tanks. A typical SBR
involves the use of two reactors, with wastewater flow alternating in cycles between the
two reactors. An SBR is considered more flexible than a CFS because the adjustments
are more easily made to the control system within the reactor tank in an SBR than are
changes to the size of a tank in a CFS.
105
-------
3.3.5.4 Carbon Adsorption
Carbon adsorption beds are used to remove hydrocarbons from air or water streams.
They are capable of very high removal efficiencies from either carrier stream. Usually
two or more beds are installed in series with a sample port between the first and second
beds. When breakthrough occurs, the bed is removed from the process and the back-
up bed is placed in the lead position. Because the adsorption process is reversible, the
carbon bed is usually regenerated for reuse by thermal regeneration. Sizing a carbon
system requires a knowledge of the concentration of contaminants and the process flow
rate. Commercial manufacturers provide a range of sizes to accommodate a wide
range of flow rates and loadings. Carbon use is only economical for relatively low mass
removal rates. High mass-removal rates make the cost of replacing/regenerating the
carbon prohibitive.
3.3.5.5 Catalytic Combustion
Catalytic combustion is a process that uses catalysts to increase the oxidation rate of
wastes at lower temperatures than conventional thermal incineration processes. The
use of a catalyst results in lower fuel requirements and reduced construction materials,
but limited applications for liquid wastes. Catalytic combustion is typically used on
vapor streams to bum contaminant materials as completely as possible prior to dis-
charge to the atmosphere. The catalysts are normally metal compounds, such as a
radium-platinum alloy, distributed on the surface of a support or carrier. Carriers are
inert metal oxides, such as alumina or porcelain, with large surface areas. Reaction
data to assist in the choice of a catalyst is often based on pilot-plant testing of the site-
specific waste materials. The exhaust is then discharged through a stack to the
atmosphere.
This technique may be used in the following treatment applications:
• Vapor stream from air stripper
Pump discharge from a soil venting system
Other process off-gases.
3.4 Free Product Recovery Equipment Costs
3.4. f Basic Cost Information
The areal extent and volume of the free product plume will dictate whether a small
simple recovery system or a large high-volume system will be required. The character-
istics of the soil matrix and/or the aquifer, such as permeability and heterogeneity, will
guide the selection of either high-volume two-pump recovery systems or low-volume
skimming systems. The expected recovery rate and the desired remediation time will,
in many cases, dictate the number and spacing of trenches, sumps, or recovery wells.
If expedited free product recovery is the preferred alternative, then a vacuum-enhanced
pumping system might be selected over a gravity drainage pumping system.
106
-------
B
The physical and chemical characteristics of both the free product and the groundwater
can affect the initial selection and cost of equipment and the long-term operation and
maintenance costs. High dissolved iron and solids can create significant maintenance
problems and costs that must be factored in when recovery and treatment equipment is
selected. The depth to water and free product will not only affect the selection of the
type of skimming or pumping equipment, but will also affect the long-term operation
costs. Deeper water will result in greater energy costs for water and product removal.
The site conditions, restrictions, and regulatory requirements can determine the location
and size of equipment that can be used. The methods of storage, treatment, and dis-
posal of water and free product are usually tightly regulated and will to some extent
dictate system design and operation.
3.4.2 Unit Capital, Installation, Operation, and Maintenance Costs
In Section 3.3.2, the different types of free product recovery systems and equipment
were described. In Table 3-7 the different types of equipment, requirements, relative
costs, advantages, and disadvantages were listed. Each type of free product recover/
equipment has associated water and product handling equipment such as holding
tanks, oil/water separators, and transfer pumps. In addition, water and/or air treatment
may be required that use air stripping and thermal treatment devices. Thus, the unit
capital costs associated with a single type of recovery system has numerous compo-
nents that need to be factored in along with installation, operation, and maintenance
costs.
An example spread sheet showing the different basic components for the free product
recovery systems is shown in Table 3-10. Example unit costs are shown for each of the
different types of free product recovery equipment Also provided are high and low
estimates for each of the cost items, time intervals, and typical volume of product. The
costs are presented in thousands of dollars for simplicity in presentation, and total
system installation costs are subtotaled. Yearly estimated costs for operation and
maintenance, disposal, and monitoring are provided along with typical duration of re-
mediation in years. A high and low subtotal for estimated operational costs is provided
along with a high and low total estimate for the life cycle of each type of free product
recovery equipment. The last two columns in the table provide a very rough estimate of
the estimated cost per gallon of free product recovered with each type of equipment.
This table and the basic approach discussed here are intended to provide the basic
components that shoujd be considered in the evaluation and selection of a free product
recovery system. For each site, a detailed cost estimate needs to be developed that
accounts for specific conditions and regulatory requirements. In the next section, an
example of a detailed cost buildup for a free product recovery system installation is
presented.
107
-------
E
£
CO
CO
o
o
ti
1
Q.
0
£
u.
CD
0)
CO
£
«-^
o
o
o
to
o
o
"E
3
0
o
S
ill
l|i
|II
"ll
i
|J
H
£ll
i'4
|f
0|
|I •
°° .
j
js i
° 1
III.1
MM| i
ii! J
§ » i
1 1
•5 «
^ 1
ki 1
•* 1
Products
Water
Handling Eq.
Low High .
"5 — ^
1 c?| x
|II 1
1 M 18 68§ 11 |^ | S |E
1 ^H Ii 2s3 §1 SSi 2' i 1 =
1 If *5 III II III If ,|
! II -8 111 II Hi l| -|
1 5S SS 2SB; Rg — t § . ,8
" *" *" *"
* O ' CO «O ... «• fs. (N
^ «CN® cNiw S«^*S-» SS c=125 iS40 fe10 52
« 1010 ,„
^ CN (si ^ £ So 1C
-u coin *"i-^-«-^. nocttN *r* to"
! ,1 Sg ' 3H gg gM S 2 " „
f
^" inin inm in in in in in in in in ^ *- m m
J no nr> m m m mm mmm g g „„
& 55 cl™ COCOCO COCO 222 2 2 " °
5 OO gg COCDCO CDCO CMCMCM CM CM in «
*~ *" CD O
» "^ S*" CM^^ CMCM COCOCO CM CM g^
I _ ~ _
J 000 <-00~W ^ W ,- "^ 23
o o
i CM
I **^»-T- vnconco CD
I »« g CPO.S! »« 8!no go »
: "" " »- "" cMnn njj minm o>2 >-:2
i JM^ ITT ID CD COCO 225 (B O «^ CO
, ^" d o
5 5
S 8
S52» S« ggS 8 S
«££ co CM !P!5.S S2 S
*o ^ 2 w
-------
3.4.2.1 Sample Cost Estimation
A typical free product recovery system installation involves many tasks and compo-
nents. Table 3-11 provides a detailed list of these typical tasks and components. The
quantities and costs for individual items in this table will vary between sites and depend
on the specific brand and type of free product recovery equipment being installed. This
table is intended to provide a guide as to the level of detail that should be considered
when a cost estimate is developed.
109
-------
Table 3-11. Typical Installation for Groundwater and
Product Recovery
Description
Material
4-inch PVC Pipe
2-inch PVC Pipe
1 -inch Black I ran Pipe
.75-inch Elec. Cond.
.75-inch Elec. 90D
1.0-inch Elec. Cond.
1.0-inch Elec. 90 D
2-inch Ts
2-inch 90 Deg
2-inch 45 Deg
4-inch 90 Deg
12-inch Drain Pipe
12-inch Coupling
12-inch Manhole
2-inch PVC Screen
2-inch Lock Cap
3x3x3 Culvert
Geotex Fabric (6000)
Ballast Rock (Ton)
Fencing
Wire Mesh
Forms
3000 PSI Concrete
Peagravel (Ton)
Glue (PVC)
Earth Anchor
Tower Anchor
Total
Outside Services
Repaying (Per SqFt)
Trucking (Per Hr)
Disposal (Per Load)
Electrical
Total
Permits
SFWMD Water Use
SPWMD Well Construction
Electrical
Building
Plumbing
Project Engineer (LC352)
Total
Quantity
50.00
260.00
260.00
520.00
10.00
570.00
15.00
1.00
20.00
20.00
5.00
35.00
2.00
1.00
10.00
1.00
1.00
0.25
110.00
60.00
0.25
60.00
8.00
30.00
1.00
3.00
3.00
450.00
40.00
10.00
1.00
1.00
2.00
1.00
1.00
1.00
24.00
Cost
101.25
526.50
526.50
171.60
7.50
273.60
18.00
4.50
30.3
18.00
26.63
330.12
18.26
78.00
39.00
22.50
690.00
138.00
805.20
763.20
17.40
50.40
480.00
228.96
13.80
84.28
9.68
5,383.57
1,620.00
1,440.00
1,140.00
5.040.00
9,240.00
360.00
240.00
300.00
60.00
60.00
1.824.00
2,844.00
110
-------
Table 3-11. (continued)
Description
Equipment Rental
Loader/Backhoe
Excavator (% yd)
Crane Truck
Boom Truck
Air Compressor (185)
Hammer/Hose/Chisel
Concrete Saw w/Blades
Plate Compactor 24 inch
Barricades (each/day)
Utility Vehicle
Total Equipment
Recovery System
Water Pump (Sub)
Product Pump (Sub)
Probes & Cable
Tank Full Probe
Product Tank (250 Gal)
Total
Air Stripper
Tower
Extra Port
Demister
Skid
Pressure Switch
Flow Meters
Telemanager Tel-1 00
Multipanel
Fs Controllers
Power Disconnect
Float Switch
Freight
Total
Recovery Wells (Labor)
Well Installation
Vault ,
Cutting Disposal
Project Engineer (LC352)
Total
Quantity
9.00
2.00
1.00
1.00
3.00
3.00
3.00
4.00
450.00
20.00
1.00
1.00
2.00
2.00
1.00
1.00
1.00
1.00
1.00
1.00
3.00
1.00
1.00
1.00
1.00
1.00
1.00
2.00
2.00
15.00
16.00
Cost
2,106.00
1,020.00
390.00
354.00
198.00
126.00
360.00
216.00
270.00
1.700.00
6,740.00
5,920.00
4,480.00
1,170.00
710.00
375.00
12,655.00
10,105.00
665.00
415.00
4,750.00
55.00
1,350.00
2,100.00
4,550.00
850.00
200.00
150.00
720.00
25,910.00
7,140.00
1,200.00
4,500.00
1.216.00
14,056.00
111
-------
Table 3-11. (continued)
Description Quantity Cost
MANPOWER REQUIREMENTS:
Recharge Gallery
Senior Field Technician 40.00 2,160.00
Field Technician 40.00 1,640.00
Technician Support 20.00 760.00
Project Engineer 10.00 830.00
Total 5,390.00
Trenching
Senior Field Technician 60.00 3,240 00
Field Technician 60.00 2,460.00
Technician Support 30.00 1^140.00
Project Engineer 15.00 1.245.00
Total 8,085.00
Equipment Compound
Senior Field Technician 20.00 1,080.00
Field Technician 20.00 '820.00
Technician Support 8.00 304.00
Project Engineer 8.00 664.00
Total 2,868.00
Unload NEPCCO Equipment
Senior Field Technician 8.00 432.00
Field Technician 8.00 328.00
Technician Support 8.00 304.00
Project Engineer 8.00 664^00
Total 1,728.00
Complete Hookup
Senior Field Technician 30.00 1,620.00
Field Technician 30.00 l|230.00
Technician Support 15.00 570.00
Project Engineer 10.00 830.00
Total . 4,250.00
Project 'Management
Project Engineer 4.00 432.00
Project Engineer 4.00 388.00
Project Engineer 32.00 2.848.00
Total 3,668.00
Total Cost For Free Product System Installation $102,817.57
112
-------
-------
Chapter 4
Remediation of Residual Organics Using
Vapor-Extraction-Based Technologies
i
i.
Soil vapor extraction (SVE) based technology primarily includes soil vapor extraction,
bioventing, and a combination of soil vapor extraction and air sparging. This technol-
ogy removes volatile contaminants and to a lesser extent semivolatile contaminants
from the vadose zone and upper part of the saturated interval (primarily in the case of
air sparging). This chapter will briefly discuss the physical processes involved, feasibil-
ity of application, system design, and monitoring requirements for effective SVE
application. j
4.1 Introduction To SVE-Based Technologies !
SVE systems are used to induce airflow through hydrocarbon-contaminated subsurface
soils. Vapors are withdrawn by applying a vacuum to wells, or trenches, which removes
the contaminants from the subsurface. SVE systems are designed to remove residual
contaminants from unsaturated soils and can be combined with groundwater pumping
wells to remediate soils that were contaminated below the water table. Figure 4-1
depicts a typical SVE system. ;
SVE can be used to biovent the soil and to deliver oxygen for enhanced biodegrada-
tion. Bioventing systems utilize air movement induced by the vapor extraction system
to deliver oxygen to normally anaerobic, hydrocarbon-rich areas. In many cases, an
increased oxygen supply stimulates the activity of naturally occurring microorganisms
that convert the hydrocarbons into water and carbon dioxide. Bioventing is further
discussed in Chapters.
i
In-situ SVE/air sparging systems force air into the porous medium in the saturated zone
in an attempt to aerate the water and strip volatile contaminants from this zone. The
vapors move into the unsaturated zone, where they are removed by the vacuum
created in the soil system. Air sparging and relatepl technologies are discussed in
Chapter 6.
113
-------
Air/Water
Separator
Air inlet
Vapor Exhaust
Treatment
Unit
Vacuum
Gauge
W//////7/////////
Vapor
Extraction
Well
Contaminated
Soil
Free-Liquid
Hydrocarbon
Figure 4-1. Components of an SVE system,,
Source: after Hoag, 1991.
114
-------
4.1.1 Typical Applications of SVE-Based Technologies
SVE-based technologies are typically used to achieve one or more of the following
goals:
Vadose Zone Source Mitigation - The removal and/or biodegradation of
residual volatile and semivolatile contaminants that cannot be remediated
by other means for technical or economic reasons, such as contaminants
located underneath structures, excessive volumes of contaminated soils,
etc.
Vapor Migration Control - Creation of contaminant vapor intercepter
system to prevent the movement of vapors into structures, utility conduits,
or sewer systems. Alternatively, air inlet wells can be used to keep vapor-
phase contaminants from migrating onto a site. Vapors from other
sources may mobilize in response to an applied vacuum.
Groundwater Dissolved Contaminant Plume Remediation - SVE/air
sparging techniques may be used to reduce concentrations of volatile
constituents in groundwater. Some SVE-only applications have also
shown reduced concentrations of volatile constituents in groundwater.
The composition and distribution of contaminants in the subsurface at leaking UST sites
will limit the available remedial options for immobile hydrocarbons. Coupling SVE with
other technologies may prove successful in a broader range of applications SVE is
applicable to a broad spectrum of sites contaminated with gasoline, solvents, and other
volatile compounds. SVE can be implemented with minor site disturbance where
normal business operations can often be continued throughout the cleanup.
The three main factors that influence the effectiveness of any in situ SVE operation are
airflow rate, contaminant vapor concentrations, and the vapor flow path relative to the
contaminant location (Johnson et ai., 1990). These factors will be discussed later in
this chapter.
The current practice for implementing SVE systems typically involves the following
activities:
Site Characterization - to collect soil samples and conduct analyses
according to EPA or state protocols.
Installation of Groundwater/Free-Product Recovery Systems - to remove
dissolved and immiscible liquid hydrocarbons.
115
-------
• Soil Gas Sampling - to determine residual contamination in the vadose
zone or in the location of soil borings and groundwater/free product
recovery systems.
• Air Permeability Testing - to determine vapor flow rates, airflow patterns in
the subsurface, and the number of vapor extraction wells required to cap-
ture volatile constituents from contaminated soil.
• Estimates of Radius of Influence - to determine the number and spacing
of vapor extraction wells for containment of hydrocarbon vapors.
• System Design - through intuition or empirical approach or may match
existing site wells and other equipment.
• Permit Applications.
• System Installation and Operation.
• TPH and BTEX Monitoring - to evaluate the progress of the remediation
and the system performance. Soil gas concentration and composition
yields useful information about the residual composition and extent of
contamination.
• Confirmation Testing - installation of soil borings to determine if contami-
nant concentrations in the soil have decreased. Decreases in vapor
extraction well concentrations are not necessarily evidence that soil
concentrations have decreased.
It should be noted that the current practice for SVE is not necessarily the most effective
approach for remediation. Different approaches to SVE design and monitoring will be
discussed later in this chapter.
4.1.2 SVE-Based System Components
An SVE-based system generally consists of the following major components:
• Extraction system
One or more vapor extraction wells or trenches
One or more air inlet or injection wells (optional)
Vacuum pumps and/or blowers
Vapor liquid separator (optional) (necessary when thermal gas
treatment units are used)
Monitoring wells
116
-------
Monitoring equipment (flow meter, pressure gauges, valves, sam-
pling ports, etc.).
• Off-gas treatment system.
These SVE system components will be briefly described below.
4.1.2.1 Extraction System Components
Vapor extraction wells usually consist of a pipe with the screened interval placed in
permeable packing, typically coarse sand or gravel; the remaining portions of the well
are grouted to eliminate short-circuiting of the system. The materials are selected to be
inert and unaffected by the contaminants of concern. Vapor extraction wells are typi-
cally cased not less than 5 to 10 feet below the ground surface to prevent the short-
circuiting of air near the well. Vapor extraction wells may be designed for vertical or
horizontal installation to (1) penetrate the contaminated zone, and (2) allow sufficient
airflow to be induced throughout the zone of contamination to remediate soils within an
acceptable period of time.
Injection and passive inlet wells, when used, are typically constructed in a manner
similar to the vapor extraction wells. Injection wells essentially inject air by means of
blowers and are typically used to control subsurface airflow to specific zones. Passive
inlet wells allow air to be drawn into the ground, but are limited to low airflow rates as a
function of changes in barometric pressure. Passive inlet wells are relatively inexpen-
sive compared with an injection system.
Vacuum pumps and blowers are used to induce subsurface airflow. The following
vacuum pump or blower types are commonly used:
Liquid Ring Pumps - to apply vacuums of up to 29 inches (786 mm) of Hg
(395 inches of water).
« Rotor Lobe Blowers (similar to liquid ring pumps) - typically operate at
lower vacuums and require periodic (monthly) maintenance.
Rotary Vane Blowers - a maximum vacuum of 27 inches (685 mm) of Hg
(368 inches of water) and, when equipped with carbon blades, require low
maintenance.
• Regenerative Blowers - high vapor flow rates with low to moderate
vacuums (<114 inches of H2O) that require minimal maintenance.
• Centrifugal Blowers - very high flow rates, but operate at very low
vacuum.
117
-------
Vapor/liquid separators are used to remove groundwater hydrocarbon condensate from
the vapor stream. This protects the vacuum and vapor treatment equipment. Separa-
tors generally consist of the following forms:
• Knockout drums or tanks - decrease vapor stream velocity due to their
larger volume compared to the transfer piping. Gravity is used to sepa-
rate dirt and liquid from the vapor stream. A demisting fabric is sometimes
placed in the drum to collect separated waste.
• Condensers - using basically the same approach as knockout drums, are
equipped with a refrigerant to condense additional moisture.
• Demisters - screen or mesh used to entrain and remove water from a
vapor stream by mechanical removal.
4.1.2.2 Offgas Treatment
Vapor treatment units are used to remove contaminant from the vapor stream before it
is discharged to the atmosphere. The selection of a treatment unit should be based on
a balance between the amount of vapor concentrations extracted and economics
(Figure 4-2). A wide variety of methods may be used, including thermal destruction
units, carbon adsorption units, vapor condensate units, and bioreactors. Examples of
these units include:
» Granular-activated carbon drums - Contaminant molecules are sorbed to
carbon surfaces as the vapor stream passes through the unit. Because
carbon adsorption is a transfer process rather than a contaminant destruc-
tion process, additional costs are incurred for carbon disposal or regener-
ation. Regenerated carbon may have a shorter useful life than fresh acti-
vated carbon because of the inability of some regeneration processes to
liberate all carbon sites. Vapors may require cooling before entering cani-
sters. Canisters may require cooling because of high heat generated
during water absorption.
• Catalytic oxidation - The vapor stream is preheated and passed across a
catalyst. Oxidation occurs between 120° and 650 °C depending on the
unit. Systems are available that can operate continuously at flow rates in
the 100- to 100,000-scfm range. For vapor concentrations <200 ppmv,
supplemental fuel is required or auxiliary heating (i.e., electrical bed heat-
ers). Oxidation is self-sustaining with vapor concentrations greater than
3,000 ppmv. Vapor concentrations greater than 12,000 ppmv generally
require dilution to prevent sintering of the catalyst. Catalysts can consist
of metallic mesh, a monolithic ceramic honeycomb, or a packed bed of
118
-------
II 1
1 1 1
K
Potentia
•
Gram
(>
•ey
active
lly Cost Prohibitive
ilar Activate*
CB hydroca
1
d Carbon
rbon)
-
'
'
-
.
'
•
'
Ci
Ox
•
-
-
.
•
The
Incine
italytic
idation
C<
™ -
•
Ceramic
Beads
rmal
ration
.
Internal
>mbustion
Engine
•
0.01 0.1 1.0 10 100 1.000 10.000 100,000 1,000.000
Extracted Vapor Concentration (in ppmv)
Figure 4-2. Selection of vapor treatment technology based on extracted vapor
concentrations.
Source: after USEPA, 1991 a.
119
-------
catalyst-impregnated pellets. Metallic mesh can be poisoned by particu-
lates blocking pore openings on the catalyst surface, chemical reactions
between the metal catalyst and reactant in gas, or the chemisorption of
gaseous material on metal. The average efficiency of these systems
tends to be >90 percent.
• Thermal incinerators - These incinerators convert hydrocarbons to CO2,
H2O, and NOxat operational temperatures of 535° to 760°C in one second
or less, either in a combustion chamber or a direct flame. Units can be
constructed for any flow rate range available up to 1000 scfm for LIST
application. Flow rates are generally about 60 scfm. Incineration is self-
sustaining for vapor concentrations greater than 50,000 ppmv. Incinera-
tion efficiency is greater than 99 percent for vapor concentrations exceed-
ing 200 ppmv and about 95 percent for vapor concentrations in the 50- to
200-ppmv range.
• Packed bed thermal processors - In these processors, which are similar to
catalyst oxidation, the vapor stream passes through ceramic beads elec-
trically heated to a fairly uniform temperature, 980° to 1,800°C. Vapor
concentrations of greater than 2,000 ppmv will sustain the operating
temperature at operational flow rates in the 100- to 500-scfm range.
Internal combustion engines - These units are essentially modified auto-
motive engines. At idling speeds, they can accommodate flow rates of
approximately 100 scfm per 300 cubic inches. Internal combustion
engines require labor-intensive equipment monitoring and have relatively
high operational noise levels.
• Vapor-phase bioreactors - Because these bioreactors are relatively new
and therefore have limited site experience, their capabilities are not well
defined. Their low cost suggests that they may have a significant applica-
tion in the future, once their capabilities are known.
4.1.3 Regulatory Considerations
Regulatory considerations fall into the broad categories of timetables and constraints
In the view of the regulatory community, remediation schedules may be driven by pres-
sure from adjoining property owners, the need to protect drinking water resources or
the workload of the regulatory staff. Time required for an SVE to reach target cleanup
goals can be minimized in certain cases through installation of many closely spaced
extraction wells operating under high airflow rates. Physical constraints resulting from
the mass transfer of volatile hydrocarbons, however, may require additional time for the
remediation of certain sites. In addition, SVE at certain sites may cause a significant
decrease in contaminant concentrations, but may not necessarily achieve remediation
goals.
120
-------
Regulatory constraints are far more common and may not be as open to negotiation. In
some jurisdictions, these constraints may result in the ruling out of technical solutions
that would otherwise be the most suitable for a given site. The following regulatory
constraints are typical:
• Emissions/Discharge control requirements are imposed on vapor emis-
sions to the air and liquid releases to sanitary systems or surface water
bodies. Air emission controls are most common for SVE-based technolo-
gies and are normally expressed in terms of an allowable mass per unit
time. Multiple standards are often in place, such as one limitation on total
hydrocarbons and another for a compound of special interest or concern.
Meeting these restrictions may require upgrading vapor treatment capabil-
ities or downgrading operational flow rates. Either alternative is not likely
to optimize system economy.
• Underground injection control requirements are generally intended to con-
trol the injection of liquids into the subsurface. In some jurisdictions, these
regulations may be phrased or interpreted in such a way that they may not
allow the use of SVE-based system options such as air sparging, subsur-
face heating through the reinjection of treated effluent vapors, or the rein-
jection of untreated effluent vapors in uncontaminated areas to stimulate
microbial activity.
• Noise level requirements are common in many light-commercial or resi-
dential areas. These requirements can often be met by enclosing vacuum
pumps or regenerative blowers in structures. Some vapor treatment
options, such as internal combustion engines, may not be able to conform
to local requirements.
4.2 Processes and Parameters
The basic soil venting process involves inducing a vacuum at a well, which produces
vapor flow through the subsurface, which in turn enhances the natural rate of volatiliza-
tion and removes soil contaminants. Higher vapor pressure components are removed
first, with the less-volatile compounds taking more time for removal. Basically, the
vapor flow rate, composition of the residual contamination, and location of the contami-
nation relative to the vapor flow path are the key processes and parameters that deter-
mine the effectiveness of soil venting at a particular site. These can be grouped into
three governing phenomena (Johnson et al., 1990a): (1) vapor flow, (2) chemical parti-
tioning, and (3) contaminant/soil distribution.
4.2.1 Vapor Flow
There are three governing equations for vapor flow porous media (Johnson, et al.,
1990b): Equation 4-1 is the "continuity equation," Equation 4-2 is "Darcy's Law" for flow
121
-------
through a porous media, and Equation 4-3 is the equation of "Ideal-Gas" assumption
relating air density differences to air pressure differences.
•f- (e p) = -V • (p 10 (4-1)
at
-V(P+pg) (4-2)
P = Pain, (— -) (4-3)
aim
where:
p = vapor density (g/cm3)
g = gravity vector (cm/s2)
e = vapor-filled void fraction of soil (-)
n - Darcian vapor velocity (cm/sec)
k = soil permeability (cm2)
H = vapor viscosity (g/cm-s)
P = vapor phase pressure (g/cm-s2) or (atm)
v = gradient operator (1/cm)
t = time (s)
Pttm = air density at atmospheric pressure (g/cm3)
P»tm = atmospheric pressure (1 .01 3 x 1 06 g/cm-s2).
The relationships established in Equations 4-2 and 4-3 can be substituted into Equation
4-1 to produce the governing equation for vapor flow:
d P 2 «2
Given an estimate of soil permeability to vapor, and the air-filled soil porosity, the solu-
tion procedure involves solving for P, using P in Darcy's Law (Equation 4-2) to obtain u_
and integrating u at extraction/injection points. Equation 4-4 can be applied to specific
vapor flow problems by using either numerical or analytical solutions.
122
-------
4.2.1.1 Analytical Solutions
Equation 4-4 can be solved for the general case of radial horizontal flow to a soil vent-
ing well (Johnson et al., 19905). To do this the pressure, P, in Equation 4-4 is ex-
pressed in terms of the ambient air pressure Patm and a deviation P* from this pressure.
P* is equivalent to the vacuum that would be measured in the soil at any given radius
from the vacuum extraction well. The substitution of Patm and P* into Equation 4-4, and
assuming e, k, and u are constant and that the product P2 is negligible compared to the
product P*Patm, provides the resulting equation for radial flow as shown in Equation 4-5.
The resulting time-dependent solution for pressure on vacuum response P* is shown, as
Equation 4-6 (Bear 1979). Equation 4-5 shows that the vacuum P* in the soil at a given
radial distance from the vacuum extraction well will increase as a function of the natural
log of time.
p* = Q [-0.5772 - /«( r E ** ) + ln(fi] (4-6)
- 4 Tl H (K/H) 4 k P
where:
Q = airflow rate (L3/T)
H = thickness of venting zone (L).
The boundary conditions for the steady-state Equation 4-4 (v 2 P = 0) are:
P = Pwforr = Rw
P = Patmforr = RI
where:
Pw = vapor or pressure at the vacuum extraction well with radius Rw
Rj = radial distance from vacuum extraction well at which the vapor or air
pressure in the soil equals ambient air pressure Patm (i.e., vacuum pres-
sure P* = 0)
FL = radius of vacuum extraction well.
123
-------
The steady-state solution to Equation 4-4 yields a solution for P(r) (Equation 4-7) and
for Q (Equation 4-8).
ta(r/*">f° (4-7)
In
Q = H TC (±) P [1 : (P°"n'PJ} (4_8)
VL w In (*JR.)
Equation 4-7 shows the relationship between the pressure distribution and radius of in-
fluence. The radius of influence, R,, is an empirical parameter that is primarily depend-
ent on soil permeability and stratigraphy. R, also depends on boundary effects and the
degree of horizontal flow. Thus, Equation 4-7 appears to be independent of soil type;
however, the R, parameter incorporates soil properties. Figure 4-3 illustrates the
changes in vacuum, P*, as a function of radial distance for three different assumed R,
values.
Equation 4-8 establishes the relationship between vapor flow rate, Q, well head
vacuum, Pw> soil permeability, k, and air viscosity, u. The flow rate, Q, is also some-
what sensitive to the value of R, as shown in Figure 4-4. At R, values above 50 ft, there
is little effect on flow rate. The flow rate is directly dependent on soil permeability and
vacuum at the extraction well (AP or Patm - PJ, as shown in Figure 4-5.
4.2.1.2 Numerical Solutions
Several vapor flow, compositional flow, and transport models currently available use
numerical solutions to analyze site-specific soil venting problems. The vapor-flow
models include MODFLOW or AIR3D, AIRFLOyV/SVE, and CSUGAS. Numerical
solutions provide a way to simulate 2-D and 3-D effects such as site heterogeneity,
vertical flow, and conduits. Vapor flow models allow an analysis of multiple well
systems with both extraction and injection flows, simulations of layered systems, and an
evaluation of complex airflow pathways. Airflow codes alone, however, are not suffi-
cient for design and performance evaluation primarily because they do not allow the
determination of mass removal from a multicomponent contaminant. Compositional flow
and transport models such as VENT2D can also be used to simulate similar vapor flow
problems, composition changes over time, mass removal, and possibly multiple phases.
These models and others are discussed in Section 4.5.
4.2.2 Chemical Partitioning
Hydrocarbon contaminants released from leaking USTs into a soil matrix may partition
into four phases. These phases are: (1) nonaqueous phase liquids (NAPL) or the
124
-------
Patm-Pw= 100 in H2O
rw = 3 in
o
K
100
80
60
& An
1
^ 20
Ri=son
Ridoon
RI = 200ft
0 50 100 150 200
r(ft)
Steady-state radial flow solution
Figure 4-3. Vacuum vs. radial distance for three values of R,.
Source: after USEPA, 1993a.
Patm-PW = 100 in H26
rw = 3 in
0 50 100 150 200
RI(ft)
Steady-state radial flow solution
Figure 4-4. Flow rate (scfm) vs. radius of influence.
Source: after USEPA, 1993a.
125
-------
100
Ftowrate I0.
[SCFM/ft] I
11
Permeability [darcy]
Steady-state radial flow solution
Figure 4-5. Flow rates (scfm) vs. Permeability for two values (AP) of wellhead
vacuum.
Source: after USEPA, 1993a.
"Immiscible phase," (2) the soil moisture or "dissolved phase" in interstitial water, (3) the
"adsorbed phase" or sorbed to soil particles and colloidal material, and (4) "vapor
phase" (Figure 4-5). The partitioning of contaminants into the different phases is
dependent on the chemical and physical characteristics of the hydrocarbon, the degree
of weathering that has occurred, and the soil characteristics. These phases will be
present in and transient between one or more of 13 locations, referred to as physico-
chemical-phase locations or loci (EPA, 1990a). Of these 13 loci, 4 have a high storage
capacity for hydrocarbons. In theory, hydrocarbon partitioning proceeds until equilib-
rium conditions between the four phases are established. The partitioning of petroleum
hydrocarbons in the subsurface can be defined by the relationships described below.
As stated above, components in the residual contaminant partition between vapor,
adsorbed, soluble (dissolved in soil moisture), and free-liquid (or solid) residual phases.
Much of the following discussion of chemical partitioning is from Johnson et al., 1990c.
The mass balance can be described for any component:
M.
= y, I
soil
RTD
toil
M.
soil
M.
soil
M
(4-9)
w, .
126
-------
where:
MJ = total moles of component I in soil matrix
Yi = mole fraction of I in soil moisture phase
«i = activity coefficient for I in water
kj = sorption coefficient for I [(mass-l/mass-soil)/(mass-l/mass-H2O)]
Py = pure component vapor pressure of I (g/cm-s2 or atm)
«-A = vapor-filled void fraction in soil matrix
Psoii = soil matrix density (g/cm3)
R = gas constant (82. 1 cm3-atm/mole-°K)
T = absolute temperature (°K)
MHC = total moles of free-liquid residual contaminant
MK>° = total moles in soil moisture phase
MSOII = mass of soil matrix (g)
M = molecular weight of water (1 8 g/mole).
W, r
The first term on the right-hand side of Equation 4-9 represents the number of moles of
I in the vapor phase, the second represents the number of moles of I in the free-liquid
residual phase, the third term is the number of moles of I dissolved in the soil moisture,
and the last term is the number of moles of I sorbed to the soil particles.
Most of the hydrocarbon contaminant mass is often partitioned into the sorbed phase
for low concentrations. Most of the contaminant mass is in the immiscible phase at high
concentrations, and the mass of the vapor phase is often negligible. The relationships
that approximate equilibrium partitioning among the vapor, sorbed, dissolved, and free-
liquid phases are discussed below.
Raoult's Law is an approximation that is used to relate the partial pressures of a consti-
tuent at equilibrium to its respective liquid phase mole fraction and the vapor pressure
of the pure constituent. Raoult's Law is applicable only for mixtures that approximate
ideal solution behavior.
It is expressed as:
f,= *, P° (4-10)
where:
Pi = partial pressure of constituent I, above mixture containing mole fraction x,
Xj = mole fraction of constituent I
P? = vapor pressure of the pure constituent I as a function of temperature.
127
-------
The concentration in the vapor phase in contact with a free-liquid phase can be approxi-
mated by Equation 4-10.
v X, P. M .
C,.v = ' ' "•' (4-11)
RT
where:
Cjv = concentration in the vapor phase (mg/L or ppmv)
X; = mole fraction of constituent I
v
P5 = vapor pressure of constituent I
MWfl- = molecular weight of constituent.
Henry's Law is an approximation for evaluating the partitioning of hydrocarbons
between "dissolved" and "vapor" phases and can be described as:
where:
Cy = concentration in the vapor phase (mg/L or ppmv)
C\ * concentration of I in the liquid phase (mg/L)
Hj = Henry's Law constant for constituent I as a function of temperature
[cm3 - liquid/cm3 - vapor].
The contaminant concentration dissolved in soil water can be approximated by:
c" = xt si (4-13)
where:
C? = concentration of constituent in water phase (mg/L)
X,- = mole fraction of constituent in contaminant
Sj = solubility of constituent in rinse water (mg/L).
128
-------
Each constituent common to gasoline blends has a unique solubility in water in its pure
form, and the solubilities of these constituents are different when present in a mixture.
As a result, the concentration of dissolved hydrocarbons in water at any given site will
be dependent on the composition of the blend of gasoline involved. Constituents pre-
sent in petroleum products have widely varying solubilities. As shown in Table 4-1,
pure constituents such as phenols and simple aromatic hydrocarbons (benzene and
toluene) are highly soluble compared with alkane constituents.
The capacity of various soil media to sorb different hydrocarbon constituents present in
the liquid phase, vapor, and dissolved phases is described by various adsorption iso-
therms that approximate partitioning between dissolved and sorbed phases. The
Linear and Freundlich isotherms are most commonly used.
Linear isotherms are:
• Appropriate for sorption relationships in which the energetics of sorption are
uniform with increasing concentration.
• Appropriate when sorbent loading is low.
• Most commonly used at very low solute concentrations.
• Used for solids exhibiting low sorption potential.
• Are often the result of the association of neutral, relatively nonpolar organic
molecules with soils.
The advantage of using this isotherm is that it reduces the complexity of mathematical
modeling and is applicable for soils with low organic content. The linear isotherm is
expressed as:
S = Kdc (4-14)
where:
S = solute sorbed per unit of sorbing phase (M/M)
Kj = distribution coefficient or partition coefficient (L3/M)
C = equilibrium concentration (M/L3).
The Freundlich isotherm is used in describing nonlinear relationships between solutes
and the adsorbing media. It may be expressed as:
S = KdCn (4-15)
129
-------
CN CM co ;: ^cdzzzz
o
o>
j ^ in •<-' T-: T
00 *~ eo ^ T- to
CM O> O> O> CO
N; CO CO p 10
CM o o o o
r--
co
s*-^
p
CO
I
I
CO
o
2
I
I*.
o
0)
I
0)
Q.
o
. O) TO •»' CNJ Z Z
CM CM CO TT 10
O5 TO O5 ^- CO CO O
1^ CM IO CO O i- CO
in CD CD CD r^- h-. t^-
o d o c> c> o o
CM oo TO oo co Cr
if> -^ ^ P **: S
!>- CM TO h- T- ^
C3 ° CD C) ° CD CD
5 »0 CM h- «0 ,„ O
O O
3
S°<<< p
co ji
CN ^
co
^- co CN ^- O) o>
T- T^ I- T- O T-
o o o o o o
IO f^ ^- h- IO CM
OO TO TO TO TO TO
op oq oq oq oq oq
CD CD O O CD O
oo ^f co
o o o
z
co
in
in
S
E co
I ro
O O
S
0)
sw
§«s|i{
ggg-t-AQ
S^ES"
^- T- O
_ 1
O CD
c x:
CD _Q.
-tfl
D- E CN" CM" E JE
jo _co _co at
o o o o
c c c c
2 B »
O O O O O O O
co en
I
o
o
TO h- 00 O) T-
O O O O O
130
-------
where:
S = solute sorbed per unit of sorbing phase (M/M)
K,, = distribution coefficient/partition coefficient (L3/M)
C = equilibrium concentration (M/L3)
n = measure of the degree of nonlinearity.
Vapor molecules tend to sorb more strongly when the soil moisture content is less than
that required to provide complete monolayer coverage of water molecules on the soil
particle surfaces (Valsacaj and Thibodeaux, 1988). This corresponds roughly to the
"wilting point" of a soil, and for sandy soils is in the 0,02 to 0.05 g-H2O/g-soil moisture
content range. More often than not, the moisture content of soils greater than one foot
below ground surface will be greater than the wilting point (Johnson et. al., 1 990c).
4.2.2.1 Maximum Vapor Concentration
After Equation 4-8 is solved, the vapor concentration in equilibrium with the con-
taminant/soil matrix, Cf v , [mass-l/volume-vapor] can be obtained from (Johnson, et al.,
1990c):
«• yt *., *; (4-16)
RT
where Mwi = the molecular weight of component I.
In the limits of low-and high residual contaminant soil concentrations, Equation 4-8
reduces to forms that do not require iterative solutions. In the low concentration limit
(i.e., no free-liquid or solid precipitate phase present), Equation 4-16 becomes:
Ceq =
"V
where:
H = Henry's law constant (= a, P,v MwH o/RT)
Q.SOH = residual contamination level of I [mass-l/mass-soil]
0M = soil moisture content [mass-HjO/mass-soil].
In the high residual contaminant concentration limit, the maximum vapor concentration
of any compound that can be achieved in the vapor-filled pore space I is its equilibrium
or "saturated" vapor concentration. Equation 4-16 becomes:
131
-------
v x. P,v M .
C.v = ' ' *•' (4-18)
R T
where:
Cy = vapor concentration (mg/L, or ppmv*)
MWji = molecular weight of constituent, I
Py = pure component vapor pressure of constituent I
R = gas constant (0.0821 l-atm/mole-°K)
T = absolute temperature of residual constituent.
For mixtures composed of constituents with similar molecular weights, xf is roughly
equal to the mass fraction of constituent I.
*Converting mg/m3 to ppmv:
ppmv = (mg/m3) (24.45) (—) (T + 273)
P 298
where:
P = standard pressure (mm Hg)
T = temperature of vapor (°C).
The conversion factor 24.45 is the number of liters per mole at standard temperature
and pressure.
Table 4-2 shows the expected vapor concentrations of select organic compounds
based on Equation 4-17. (Note that the gasoline vapor concentration is obtained by
summing the contributions of all individual mixture components.)
Equations 4-16 and 4-17 are the two most commonly incorporated in vapor transport
models. Equation 4-16 predicts vapor concentrations that are proportional to the
residual soil concentrations of each constituent and are independent of the relative
concentrations of each constituent in the contaminant mixture, while the vapor
concentrations predicted by Equation 4-17 are independent of residual soil con-
centration levels and depend only on the relative concentrations of constituents. It is
important to recognize that these transport models are only valid for specific limiting
conditions, and generalization to other concentration ranges can produce very
misleading results. For example, Equation 4-16 predicts that vapor concentrations
always increase with increasing residual contaminant levels, but realistically the
equilibrium vapor concentration of any compounds cannot exceed its saturated vapor
concentration, PyMw/RT.
132
-------
Table 4-2. Composition of a Regular Gasoline
Compound Name
Propane
Isobutane
N-butane
trans-2-Butene
cis-2-Butene
3-Methyl-1-butene
Isopentane
1-Pentene
2-Methyl-1 -butene
2-Methyl-1 ,3-butadiene
n-Pentane
trans-2-Pentene
2-Methyl-2-butene
3-Methyl-1 ,2-butadiene
Cyclopentane
2,3-Dimethylbutane
2-Methylpentane
3-Methylpentane
n-Hexane
Methylcyclopentane
2,2-Methylpentane
Benzene
Cydohexane
2,3-Dimethylpentane
3-Methylhexane
3-E:thylpentane
n-Heptane
Methylcyclohexane
(continued)
Mw (g)
44.1
58.1
58.1
56.1
56.1
70.1
72:2
70.1
70.1
68.1
72.2
70.1
70.1
68.1
70.1
86.2
86.2
86.2
86.2
84.2
100.2
78.1
84.2
100.2
100.2
100.2
100.2
98.2
Weight
Fraction
0.0005
0.0085
0.0259
0.0019
0.0018
0.0010
0.0916
0.0032
0.0068
0.0068
0.0628
0.0138
0.0129
0.0003
0.0185
0.0111
0.0515
0.0314
0.0411
0.0214
0.0077
0.0172
0.0059
0.0063
0.0099
0.0168
0.0356
0.0055
Mole
Fraction
0.0001
0.0137
0.0415
0.0032
0.0030
0.0013
0.1181
0.0042
0.0090
0.0092
0.0810
0.0184
0.0171
0.0004
0.0245
0.0120
0.0556
0.0340
0.0444
0.0237
0.0071
0.0250
0.0065
0.0058
0.0092
0.0156
0.0331
0.0052
Piv(20°C)
(atm)
8.50
2.93
2.11
1.97
1.79
0.96
0.78
0.70
0.67
0.65
0.57
0.53
0.51
0.46
0.35
0.26
0.21
0.20
0.16
0.15
0.11
0.10
0.10
0.072
0.064
0.060
0.046
0.048
133
-------
Table 4-2. (continued)
Compound Name
2,2-Diniethylhexane
Toluene
2-Methylheptane
3-Methylheptane
n-Octane
2,4,4-Trimethylhexane
2,2-Dimethylheptane
Ethylbenzene
p-Xylene
o-Xylene
n-nonane
3,3,5-Trimethylheptane
n-Propylbenzene
1 ,3,5-Trimethylbenzene
1 ,2,4-Trimethylbenzene
n-Decane
Methylpropylbenzene
Dimethylethylbenzene
n-Undecane
1 ,2,4,5-tetramethylbenzene
1 ,2,3,4-tetramethylbenzene
1 ,2,4-trimethy l-5-ethy I-
benzene
n-dodecane
naphthalene
methylnaphthalene
Mw(q)
114.2
92.1
114.2
114.2
114.2
128.3
128.3
106.2
106.2
106.2
128.3
142.3
120.2
120.2
120.2
142.3
134.2
134.2
156.3
134.2
134.2
148.2
170.3
128.2
142.2
Weight
Fraction
0.0046
0.0899
0.0028
0.0062
0.0647
0.0015
0.0003
0.0205
0.0153
0.0221
0.0155
0.0033
' 0.0346
0.0201
0.0061
0.0343
0.0210
0.0173
0.0078
0.0511
0.0053
0.0191
0.0050
0.0041
0.0061
Mole
Fraction
0.0038
0.0908
0.0023
0.0051
0.0528
0.0011
0.0002
0.0180
0.0134
0.0194
0.0112
0.0022
0.0268
0.0156
0.0047
0.0224
0.0146
0.0120
0.0046
0.0354
0.0037
0.0120
0.0027
0.0030
0.0040
Piv (20°C)
(atm)
0.035
0.029
0.021
0.020
0.014
0.013
0.011
0.0092
0.0086
0.0066
0.0042
0.0037
0.0033
0.0024
0.0019
0.0036
0.0010
0.00070
0.00060
0.00046
0.00033
0.00029
0.00040
0.00014
0.00005
Total
0.996
0.999
134
-------
Figure 4-6 compares vapor concentrations predicted by Equations 4-6, 4-15, and 4-16
for the regular gasoline defined by Table 4-1. The required chemical parameters (vapor
pressures, octanol-water partition coefficients, water solubility values) can be found in
Johnson et al., 1988. As an example, model parameters for a sandy soil are organic
carbon fraction (f^) = 0.002, soil moisture content (eM) = 5%, total void fraction (eT) =
0.35, and soil bulk density (p^,) = 1.60 g/cm3.
too
Benzene Vapor
Concentration
(mg/I)
100 1000 10000
Residual Soil Concentration
(mg-gasoline/kg-soil)
Figure 4-6. Comparison of vapor concentration models.
Source: after USEPA, 1993a.
For these values, Figure 4-6 indicates that Equation 4-16 is applicable below a residual
soil contamination level of about 500 mg of gasoline/kg of soil. Above this residual
concentration level, however, Equation 4-16 predicts increasing vapor concentrations
with increasing residual levels, while the complete model predicts that vapor.
concentrations become independent of the residual concentration level. This limiting
behavior is predicted by Equation 4-17. Care must be taken when using transport
models based on a "three-phase model," such as Equation 4-16, because they will
overpredict vapor concentrations and emission rates for many situations. Usually there
are no internal checking procedures in these models to ensure that unrealistic vapor
concentrations are not being predicted.
Equation 4-17 can be used to approximate maximum equilibrium vapor concentrations
because it is most applicable for the residual concentration levels encountered at
typical UST release sites.
135
-------
4.2,3 Contaminant Distribution in Soil
The distribution of contaminants in different soil types is affected by any one of a
number of processes that limit the subsurface vapor flow from becoming saturated (i.e.,
reaching the equilibrium concentration) with hydrocarbon vapors. These processes are
referred to as "mass transfer limitations."
Mass transfer limitations are most often the result of diffusional "resistances," which
may occur on the micro- (pore size) or macro (larger) scale. With respect to vapor
extraction processes, mass transfer limitations most often arise from subtle permeability
variations that cause air to flow past, but not through, zones of contamination as shown
in Figure 4-7. This is easy to picture if one considers drawing air through a formation
consisting of alternating sand and clay layers. In this case, significant airflow can be
expected to occur only through the permeable sand zones, and remediation of clay
layers can occur only if the vapors somehow travel to the sand zones. This most likely
occurs as a result of "diffusion," or the movement of molecules from one location to
another as a result of concentration changes with distance.
Sandy Soil &;&$!
v •.*" • .• -v •.*.• ;
Clayey Soil
Contaminant
Figure 4-7. Diffusion-limited mass transfer.
Source: after USEPA, 1993a.
The rate of transport by diffusion is most often described by Fick's Law of diffusion:
dx
(4-20)
where
effective diffusion coefficient (cm2/s)
136
-------
&& = concentration gradient.
iix
Deff is estimated by the Miliington-Quirk (Millington and Quirk, 1961) expression:
3.33
D? = D°v -^- (4-21)
er
' e3-33
Df = D ° _L_ (4-22)
Li Li ry X "
where:
DL = diffusion coefficient in water [cm2/s]
DC = diffusion coefficient in air [cm2/s]
eA = vapor-filled porosity
EL = liquid-filled porosity
eT = total porosity.
The effective diffusion coefficient is described in Lyman, et al., 1982. As an example,
for the significant difference between Df and Df, for benzene at 20°C, D'v = 0.087
cm2/s and DL = 1.0 x 10'6 cm2/s. With EA = 0.26, EL = 0.12, and eT = 0.38, the calculated
Df = 7 x 10'3 cm2/s and Df = 6 x 10"8 cm2/s (Johnson and Ettinger, 1991).
Two cases for determining removal rates where mass transfer is limited are discussed
below.
4.2.3.1 Diffusion-Limited Transport
Figure 4-7 illustrates a situation in which vapors flow primarily past, rather than through,
the contaminated soil zone. As described previously, the air flows past the
contaminated zone, such as in the case of a contaminated clay lens surrounded by
sandy soils. Vapor-phase diffusion through the clay limits the removal rate, with
maximum removal occurring when airflow is fast enough to maintain a low vapor
concentration at the permeable/impermeable soil interface. At any time, t, a
contaminant-free or "dried-out" zone of low permeability will exist with a thickness 6.
The removal rate is estimated by the equation (Johnson, et al. 1990a):
R. = it (R? - Rf) C^l £>/5(0 (4-23)
137
-------
For a single component thickness, the dry-zone thickness change with time can be
calculated as:
8(0 =
D t
(4-24)
For Equations 4-23 and 4-24, the following terms are defined:
Q% = vapor concentration in equilibrium with the contaminant/soil
matrix (Equations 4-14, 4-15)
c* = residual level of contamination in the low-permeability
zone
(g-cohtaminated/g-soil)
^t = estimated removal rate (kg/d)
Ri = extraction well radius (cm)
R2 = radial area of contamination (cm)
D = effective soil vapor diffusion coefficient (cm2/s)
P = soil bulk density (g/cm3)
t = time (s).
It should be noted that Equation 4-24 assumes a single-component system Mixture
removal rates for this and the next situation are difficult to estimate because of
continual system changes in composition and liquid-phase diffusion. Although no
simple analytical solutions exist for these more complex situations, the removal rates
would be estimated to be lower than the rates predicted for pure components.
4.2.3.2 Removal From Free-Product Layers
In this case, air flows parallel to, but not through the zone of contamination, and vapor
phase diffusion is the limiting factor. An example case is a layer of liquid hydrocarbon
resting on top of an impermeable stratum or the water table (Figure 4-8) For a single
component, the removal rate can be estimated iby use of the followinq equation
(Johnson et. al., 1990a):
= _L (6 M - (4'25)
where:
138
-------
n
D
n
k
H
RI
"
atm
= efficiency relative to maximum removal rate
= effective soil-vapor diffusion coefficient [cm2/s]
= viscosity of air = 1.8 x 10"4 g/cm-s
= soil permeability to vapor flow [cm2]
= thickness of screened interval [cm]
= radius of influence of venting well [cm]
= venting well radius [cm]
= absolute ambient pressure = 1.016 x 106 g/cm-s2
= absolute pressure at the venting well [g/cm-s2]
= defines region in which contamination is present.
Removal from
Free-Product Layers
R«« = i\ Q C?
C/C .
Figure 4-8. Removal of contaminants from free-product layer.
Source: after USEPA, 1993a.
The efficiency n is inversely proportional to the screened interval thickness H because a
larger interval will pull in unsaturated air passing above the liquid-phase contamination.
D is calculated by the,Millington-Quirk expression as defined in Equations 4-21 and
4-22. As an example, consider a layer of free-phase hydrocarbon overlain by sandy
soil (k = 1 darcy). A vapor extraction well with 2-inch radius and 10 feet of screen is
operating at Pw = 0.9 atm. Contamination extends 30 feet from the well. Assuming
appropriate values for other input parameters, a venting efficiency of about 9 percent is
calculated. This efficiency is much less than the maximum removal rate not limited by
vapor-phase diffusion.
139
-------
4.3 Feasibility Screening
4.3.1 Is Venting Appropriate?
The primary factors governing the behavior of an in-situ soil venting system (previously
discussed) are airflow rate, contaminant vapor concentration, and airflow path relative
to contaminant distribution in the subsurface, in order to determine if in-situ soil venting
is appropriate at a given site, the following questions should be answered:
• What range of airflow rates can be realistically achieved?
• What is the likely maximum contaminant vapor concentration?
• Under ideal airflow paths, is this concentration adequate to yield acceptable
removal rates?
• What vapor composition and concentration changes will occur with time?
What residual, if any, will be left in the soil? How do these values relate to
the regulatory requirements?
• Are there likely to be any negative effects of soil venting?
if site conditions do not produce acceptable removal rates, in-situ soil venting should be
ruled out as a practical treatment method. A detailed discussion of the approach and
steps for addressing these questions is presented in Johnson, et al. (1990a).
4.3.1.1 Airflow Rate
The first step in feasibility screening is to estimate a realistic airflow rate range for the
given site conditions, according to the equation (Johnson, et al., 1990a):
°-* it)'-"*%£? <««>
where:
k = soil permeability to airflow [crn2 or Darcy]
H = viscosity of air = 1 .8 x 1 Q/4 g/cm-s or 0.01 8 cp
Pw = absolute pressure at extraction well [g/cm-s2 or atm]
Patm = absolute ambient pressure - 1.01 x 106 g/cm-s2 or 1 atm
RW = radius of vapor extraction well [cm]
140
-------
RI = radius of influence of vapor extraction well [cm]
Jj = flow rate per unit thickness of well screen [cm3/sj.
The airflow for this equation is assumed to be horizontal. The calculated flow rate
should be considered as a rough initial estimate of flow. If k can be measured or esti-
mated, based on soil grain-size characteristics, then the radius of influence R, is the
only unknown parameter. Because Equation 4-26 is not sensitive to changes in R, it is
suggested that an average value of R, = 12 meters or 40 feet, based on ranges report-
ed in the literature, be used for estimation purposes. Typical vacuum well pressures
can range from 0.50 to 0.90 atm. Flow rates are usually measured in standard
volumetric units (Johnson, et al., 1990a):
[scfm/ft] (4-27)
In this equation, H is the thickness of the screened interval, which is often chosen to be
equal to the thickness of the zone of soil contamination (this minimizes removing any
excess "clean" air). The largest uncertainty in flow-rate calculations is the air perme-
ability, k, which can vary several orders of magnitude at a site.
4.3.1.2 Vapor Concentrations
The likely maximum contaminant vapor concentrations are estimated based on results
of either soil-vapor surveys, analyses of headspace vapors above contaminated soil
samples, or equilibrium vapor models. In the absence of soil-vapor survey data an
estimate of the maximum or "saturated" vapor concentration can be obtained from:
_ 2 P-(T)Mv.
'" = <• *'—W^ (4'28)
where:
Cest = estimate of contaminant vapor concentration [mg/L]
xj = mole fraction of component I in liquid-phase residual (x, = 1 for single
compound)
PI(T) = pure component vapor pressure of constituent I as a function of
temperature [mg/mole] :
Mw>i = molecular weight of component I [mg/mole]
141
-------
R = gas constant = 0.0821 1-atm/mole-°K
T = absolute temperature of residual [°K].
More-sophisticated equations are available for predicting vapor concentrations in soil
systems based on equilibrium partitioning arguments, but input generally includes data
such as organic carbon content or soil moisture, which are not normally available. If a
site requires remediation, total residual hydrocarbons are usually greater than 500
mg/kg, with most hydrocarbons present as a separate phase. If so, vapor
concentrations are independent of residual concentration and Equation 4-28 is applic-
able. It should be noted that all of the above applies to the start of venting, when
removal rates are the greatest. Contaminant concentrations in the extracted vapors
decline with time because of changes in composition, residual levels, or increased
diffusional resistance.
4.3.2 Airflow Conditions and Removal Rates
The concentration estimate C^ is multiplied by a range of reasonable flow rates Q to
yield R^,, an estimated removal rate:
*~ = Cest Q (4-29)
Typical sites report airflow rates in the range of 10 to 100 scfm, although sandy soils or
large numbers of extraction wells can yield flow rates up to 1,000 scfm.
Pore volume calculations are used along with extraction flow rate to determine the pore
volume exchange rate. The exchange rate is calculated by dividing the pore space
within the treatment zone by the design vapor extraction rate. The pore space within
the treatment zone is calculated by multiplying the soil porosity by the volume of soil to
be treated. Some literature suggests that one pore volume of soil vapor should be
extracted at least daily for effective remedial progress. The time required to exchange
one pore volume of soil vapor can be estimated by the following equations:
E = *L (4-30)
(TO 3 vaporlm 3 soil) x (m 3 soil) _ ,
£, — —— - n
(m vaporlh)
where: E = pore volume exchange time (h)
n = soil porosity (m3 vapor/m3 soil)
142
-------
V = volume of soil to be treated (m3 soil)
Q = total vapor extraction flow rate (m3 vapor/h).
At this point, it is necessary to consider decreasing vapor concentrations during venting
because of compositional changes and mass transfer resistance. Removal rates can
be calculated as a function of vapor concentrations for a range of flow rates.
Acceptable removal rates Racc can be calculated by dividing the estimated spill mass
MSPJI, by the maximum acceptable cleanup time t:
Racc = M,pil^ (4-32)
Maximum removal rates are achieved when the induced vapor flow travels only through
the zone of soil contamination and when no mass-transfer limitations are encountered.
That is, all vapor flows become saturated with contaminant vapors and the estimated
removal rate is described by Equation 4-32. Figure 4^9 illustrates the ideal model
predictions, in which the percent residual removed is inversely proportional to the initial
vapor concentration decrease.
% Initial
Vapor
Cone.
too
% Residual
Removed
• SO 100 ISO 200
[L-vapor/g-initial residual]
% Residual
Composition
• 50 100 ISO 200
[L-vapor/g-initial residual]
>n FlowratexTime f
Spill Mass
Figure 4-9. Ideal model predictions and composition changes.
Source: after USEPA, 1993a.
4.3.2.1 Residuals and Air Volume Calculations
As contaminants are removed, residual soil contamination decreases and mixtures
become richer in less-volatile compounds, with consequent decreases in removal rates
143
-------
over time. As venting continues and residuals decrease, it becomes more difficult to
remove the residual contamination.
The maximum efficiency of a venting system is limited by the equilibrium partitioning of
contaminants between the soil matrix and vapor phases. This efficiency changes with
time as a function of the initial residual composition, vapor extraction well flow rate, and
initial soil contamination level.
In addition, there is a practical limit to the amount of residual contamination that can be
removed by venting. In the case of gasoline, for example, after 90 percent of the initial
residual has been removed, the remainder consists of relatively insoluble and
nonvolatile compounds. A much larger airflow volume will be required to remove each
remaining gram in the final 10 percent than in the first 90 percent. This inherent
limitation indicates that soil venting alone may be inadequate to meet a mandated
regulatory cleanup concentration, which often will require enhanced biodegradation or
other applied technologies to achieve lower concentrations than achievable by venting
alone.
Soil venting systems have the following potentially negative effects:
• Off-site contaminant vapors may induce migration toward extraction wells,
especially in more industrialized areas such as an intersection with multiple
gasoline stations. In such cases, a vapor barrier should be established at
the perimeter of the contaminated zone by allowing vapor flow into
perimeter groundwater monitoring wells, which will act as passive air-supply
wells. Trenches may also be installed in lieu of wells.
• The water table may rise below the vapor extraction wells as a result of the
vacuum induced by the extraction well pumping system. A dewatering
system may be necessary to ensure that contaminated soils remain
exposed to vapor flow.
4.3.3 Site Screening, Levels I and II
To summarize, before installation of an in situ soil venting system, it is important to
evaluate the appropriateness of venting for site remediation. Following the
measurement or estimate of the parameters relevant to vapor flow conditions, the first
step is to calculate the maximum potential removal rate (maximum removal rate = flow
rate x maximum vapor concentrations) and to compare that rate with the desired
removal rate (desired removal rate = residual mass/desired cleanup time). If the value
for the de-sired removal rate is less than the range of maximum removal rates, then the
site is a candidate for SVE remediation according to the first screening approximation.
If the desired removal rate is achievable, then the minimum number of extraction wells
needed can be calculated as a function of the vapor flow rate per well, residual
144
-------
contaminant mass, and minimum vapor volume requirement. If that number of
extraction wells is realistic within constraints of site geometry and access limitations and
budget re-straints, then installation and operation of a soil venting system is realistic.
In the remedial design process, the values of air permeability used in calculations to
determine the appropriateness of SVE were estimates derived from the site investiga-
tion hydrogeologic data collected to define the nature and extent of contamination.
During a pilot test program, air permeability values are measured in the field and the
site-specific values are used to revise the calculated values of soil permeability to air.
Based on the revised calculations, the number of vapor extraction wells necessary for
remediation is revised for the system design.
4.4 SVE System Design
4.4. i Determining Remedial Goals For Technology Selection
The goals of a given site remediation that drive the selection of the remedial technology
are based on the following:
• Composition and distribution of contaminants in the subsurface
• Effectiveness (limitations of available technologies)
• Cost
• Regulatory requirements.
A brief discussion will be presented of each of these factors and their influence on the
selection of appropriate and effective corrective action technologies for achieving the
desired remediation goal.
4.4.1.1 Contaminant Composition
The majority of petroleum products released from UST systems include motor fuels
(such as gasoline and diesel fuel), jet fuel (such as JP-4), heating oil, and lubricating oil.
These products are often referred to as single bulk fluids; however, each product
consists of a complex mixture of organic constituents. Gasolines, for example, typically
consist of C4-C12 constituents, whereas diesel fuel consists of C12-C25 constituents. For
aliphatic compounds, low-carbon-number constituents are more volatile and have a
higher water solubility, while higher-carbon-number constituents are less volatile and
more immiscible. Aromatic compounds such as benzene, toluene, ethylbenzene, and
xylene (BTEX) are more soluble in water than aliphatic compounds of the same carbon
number. Figure 4-10 matches the representative range of hydrocarbon constituents in
different petroleum products with the most commonly considered soil treatment
technologies.
Approaches to cleaning up petroleum-contaminated soil do not generally address the
entire mixture of compounds in a bulk blend, but are designed to address specific
constituents or classes of constituents. For example, soil vapor extraction may be
145
-------
effective only for volatile constituents (< C12). In addition, no single in situ technology is
effective in addressing constituents in all phases or subsurface locations. Traditional
pump-and-treat technologies only address light nonaqueous phase liquids (LNAPLs) on
the ground water table and some of the contaminants dissolved in groundwater;
whereas contaminants present in other phases or locations are not addressed.
Consequently, corrective action technologies are being misapplied in many cases, or
the effectiveness of single or multiple technologies is not being optimized.
Boiling
Point, °C
630 Cg,
Vtdose
Zone
Saturated
Zone
In Situ Technologies
• Soil Vapor
Extraction
• Soil Vapor
Extraction and
Air Sparging
• Soil Vapor Extraction
and Bloventlng
• Soil Vapor Extraction,
Air Sparging, and
Biovantlng
Ex Situ Technologies
•Thermal Desorptlon
• Bioremedlatlon
•Soil Washing
— Tr
t
v
L.
Figure 4-10. Range of hydrocarbon constituents in different petroleum products
associated with the most commonly considered cleanup
technologies.
Source: after USEPA, 1993a.
4.4.1.2 Cost Considerations
The level of capital investment for a site is limited based on the level of contamination
present at a site and the solvency of the responsible party (RP) or Trust Fund. These
remedial costs are often tied to an imposed time frame to complete the cleanup.
Remediation by an SVE-based technology commonly requires 3 years or more. This
146
-------
time frame may exceed the O&M commitment of an RP or Trust Fund, or it may not be
rapid enough because of external issues.
4.4.1.3 Regulatory Requirements
The selection of SVE-based technologies may be constrained by remediation targets.
The problem could be a function of the numerical concentration target (risk-based
approach to corrective action, ASTM ES-38) or the analytical parameter or method thait
is required. The rate and degree of soil remediation achievable through SVE
technology is a function of the physical characteristics of the affected soils (primarily
permeability) and the overall carbon number range of the hydrocarbon contaminant.
The low-carbon-number volatile fuels such as gasoline will evaporate readily, and the
rate and degree of achievable remediation is a function of airflow rates through the
affected soils. Higher-carbon-number fuels such as diesel are much less volatile and
may be more slowly consumed through biodegradation. The rate-limiting controls for
biodegradation are primarily oxygen availability, as well as temperature, nutrients, and
moisture. Low-permeability soils cannot be readily remediated through either
evaporation or biodegradation because low airflow rates are achievable through SVE.
In either SVE or bioventing applications, a small fraction of the affected soils is likely to
require long treatment periods to achieve cleanup goals even though the majority of
contaminated soils have been adequately remediated. Existing regulatory
requirements may prolong treatment periods for such sites beyond practical limits.
If the hydrocarbon contaminant mass contains heavy fraction compounds, SVE could
remove all the volatile constituents present in the subsurface, but still not achieve a
remediation goal on the basis of TPH analyses. Available evidence indicates that SVE!
is very effective for removing those fractions of contamination located in the vapor and
free-liquid phases or adsorbed to the external surfaces of the soil matrix. Both
theoretical considerations and field studies, however, indicate that SVE will not be
effective for removing contamination trapped in the interior of the soil matrix. Because
the quantity of contaminants trapped in the interior of the soil matrix may exceed
surface contamination by 1 to 2 orders of magnitude, SVE cannot be relied upon to
return long-contaminated soils to their original condition (Travis and Macinnis, 1992).
4.4.2 Range of Common SVE Design Approaches
In practice, the approaches employed in designing an SVE-based remediation system
are as varied as the sites themselves. The basic design approaches fall into the
following five broad categories: (1) intuition or empirical, (2) matching existing
equipment, (3) radius pf influence analyses, (4) screening model analysis, and (5)
detailed modeling analyses. A summary of these five approaches is provided in Table
4-3 (Johnson et al. 1992).
147
-------
Table 4-3. Soil Vapor Extraction Process Design Approaches
(Johnson, etal. 1992b)
Approach
Required
Information8
Advantages
Disadvantages
"Intuition," or empirical 1,2
approach based on
past experience
System design
matched to existing
equipment
Radius-of-influence-
based approach
1,2, inventory of
existing equipment
1,2, 4,5", 6
Based on screening 1,2, 3b, 4, 5,6,
level model results economic data
Detailed modeling, 1 through 10
numerical optimization economic data
Quick, easy, low skill
level required
Quick, easy, minimizes
new capital expenditure,
maximizes use of exis-
ting equipment
Insures containment of
hydrocarbon vapors
Little effort required,
design based on desired
performance; cost of
analyses not prohibitive
Design can be
optimized and based on
desired performance
Unknown system per-
formance, technology
may not even be appli-
cable
Unknown system per-
formance, technology
may not even be appli-
cable
Unknown system per-
formance, does not
insure remediation in
reasonable time frame
Requires higher level of
expertise and ability to
interpret data
Requires highest level of
expertise and ability to
interpret data; cost may
be prohibitive
* Refers to activities defined in Table 4-4.
b Optional, not always used in this approach.
4.4.2.1 Intuition or Empirical
The intuition or empirical approach is usually applied at small sites with a limited vadose
zone contamination problem. These could be sites where tank replacement operations
have revealed a limited area of soil contamination or where a site assessment has
identified an area of limited soil contamination around one or two wells or borings.
Using an intuitive approach, many contractors simply put a soil vent system in the tank
excavation or hook up to the existing monitor works. A standard 1Vz- to 3-hp soil vent
blower system is typically installed. The systems are permitted as necessary and oper-
ated until vapor concentrations are consistently below detection limits. The systems are
then shut down, and the vadose zone is assumed to have been remediated. For many
small soil contamination sites, this can be a quick, easy, and cost-effective approach.
The limitation of this approach is that in several cases the performance of the System is
unknown and the extent of residual soil contamination is not addressed. In addition, the
SVE systems being employed may be either too large or too small for the site. In some
cases, SVE may not even be applicable for site conditions.
148
-------
4.4.2.2 Matching Existing Equipment
The approach of matching the SVE system to existing site wells and other equipment is
similar to the intuitive or empirical approach. The main difference is that in most cases
the matching approach typically will involve more analyses up-front as to the extent and
characteristics of the soil contamination, analysis of the well head vacuum, total vapor
flow rates, expected maximum discharge concentrations, and selection of appropriate
SVE equipment. The emphasis of this approach is to both utilize existing equipment
and to try and achieve site remediation goals. For many small service station sites with
ample existing monitoring wells and a relatively simple vadose zone, this can be a
practical and cost-effective approach because it minimizes new capital expenditure and
maximizes use Of existing equipment. If an up-front analysis is not conducted,
however, the performance of the SVE system is not necessarily known and (as in the
sampling approach) may not even be applicable.
4.4.2.3 Radius of Influence
The radius of influence approach is perhaps the most common approach currently
being used for SVE design and implementation. In this approach, the site assessment:
has been completed and the extent of soil and groundwater contamination has been
delineated. An SVE system is then designed and/or specified that will have the re-
quired "radius of influence" to encompass the area of soil contamination. The radius of
influence all too often is "assumed" based on previous experience with the equipment
used and the soil types from similar sites in which the SVE wells are screened. This
type of approach is very similar to the intuitive/empirical approach and has the same
limitations and disadvantages.
At many sites, a pilot test is usually run to provide a better estimate of the radius of
influence (see Pilot Tests for SVE-Based Systems). The vacuum readings at different
distances are plotted as a function of the distance from the pilot vapor extraction well.
The radius of influence is then interpreted as the distance at which the vadose zone
vacuum is approximately 0.1 inch of H2O or as a.region contributing 90 percent of total
airflow to a vapor extraction well that roughly corresponds to that area where the mea-
sured soil vacuum is ^ 1 percent of the applied vacuum at the vapor extraction well.
Figure 4-11 presents an example of the plots used to determine radius of influence. In
this example, the radius of influence is different depending on whether it is based on
vacuum or airflow measurements. The SVE system is then designed based on this
radius and the use of enough wells to overlap and encompass the area of soil
contamination. The blower is then sized to pull vapors from the SVE well system. This
approach assumes that a measurable vacuum reading is an indication of vapor flow
and does not ensure that remediation occurs in a reasonable time frame. Problems
with this approach include the following:
• Measurable pressure/vacuum readings are no guarantee of significant
vapor flow or remediation.
149
-------
The current approach is not related to any remedial objective (i.e.,
cleanup time, removal, cost minimization, etc.).
The current approach is based on a containment philosophy.
Gauge
Vacuum
o\/
10 let iMt
Distance From Extraction Well (ft]
Figure 4-11. Field measurements and plots for determining radius of influence.
Source: after USEPA, 1993a.
The "radius of influence" defines a zone of "containment" and not a region of remedia-
tion. The time for remediation is proportional to the ratio of containment mass/airflow
through the region. The approach for evaluating and designing the SVE system should
consider the following:
• The zone of treatment will be a function of geometry, soil characteristics,
and total flow rate.
• Measurable vacuum/pressure readings are not sufficient evidence of
significant vapor flow in a given region.
• The most accurate estimates of flow behavior can only be obtained by
combining modeling with field observations.
• Flow strength can be inferred by pressure gradients (AP/AX) and soil
permeability.
• In the absence of numerical simulations, guess estimates can be based
on extrapolations from results of modeling similar situations where the
radius of influence is equal to the depth to screened interval (no cover), or
the radius of influence is equal to cover width (perfect surface seal).
150
-------
• Soil gas monitoring points (concentration and composition) are used to
monitor remediation effectiveness.
The use of screening models and detailed modeling with field measurements are briefly
discussed below (also see Screening and Modeling Tools).
4.4.2.4 Screening Model Analysis
In this approach, the concentrations and characteristics of the petroleum contaminant
and the site conditions are evaluated by using screening models to (1) determine if SVE
is inappropriate at a given site, and (2) to identify, estimate, and evaluate required site-
specific data. If economic data is available, the feasibility of SVE can be evaluated.
Screening models use analytical solutions for airflow and contaminant-removal
calculations and typically require estimates of the radius of influence, contaminant
distribution/mass, and desired remediation time as input. An overview of these models
is presented in Section 4.4.4, Screening and Detailed Modeling Tools. This approach
requires minimal effort in terms of data requirements and use. For relatively
homogenous sites, this approach can be used to initially evaluate design criteria based
on the desired performance (e.g., achieving site cleanup goals in a reasonable time
frame). Although this approach is relatively inexpensive, it requires a higher level of
expertise and ability on the part of the practitioner than does the previous design
approaches discussed.
4.4.2.5 Detailed Modeling Analysis
The detailed modeling approach is generally used on larger more complex sites.
Detailed modeling simulates vapor behavior in the vadose zone as an aid in designing
and optimizing SVE-based systems. This approach can be used to (1) select design
parameters, (2) determine if SVE is appropriate, if economic data is available, and if
SVE is feasible at a given site, (3) determine the optimum number and location of
extraction wells, (4) size aboveground treatment systems, if required, (5) evaluate
modifications to existing systems, and (6) provide a basis for more realistic cost
estimates.
Two types of models are used in this approach: airflow models and compositional flow
and transport models. Airflow models simulate 2-D or 3-D airflow paths and account for
differences in vertical and horizontal permeability in the subsurface and in boundary
conditions that affect airflow to the extraction wells. Airflow models can only predict
potential airflow and do not estimate the mass removal or effectiveness of SVE
systems. Compositional flow and transport models simulate similar airflow problems,
but they also model the mass transport of multicomponent mixtures and compositional
changes of the residual contaminant overtime in addition to their possible application to
multiple phases.
The readily available screening and detailed models for evaluating SVE systems are
identified in Section 4.4.4, Screening and Detailed Modeling Tools. Use of these
151
-------
models in conjunction with site characterization and pilot test data provide a means to
quantify important SVE operating processes. Site-specific data that is used in these
modeling approaches for evaluating and designing SVE systems is obtained as part of
both the site characterization and from pilot tests. The types of site characterization
and pilot test activities that are conducted as part of the different design approaches are
described below.
4.4.3 Pilot Tests for SVE-Based Systems
4.4.3.1 Purpose of Pilot Tests
A pilot test is a small-scale, short-duration (typically less than 8 hours) test of a basic
SVE system in order to obtain the data required to design an effective large-scale SVE-
based remediation system. For the data to be useful, the test must be conduced long
enough to ensure one "pore volume" of air has been moved through the contaminated
soils to the vapor extraction wells. To be successful, a pilot test must provide accurate
and reliable data to (1) identify sustainable airflow rates, (2) anticipate contaminant
composition and removal rates, (3) determine airflow patterns in the subsurface, and
(4) estimate the number and location of vapor extraction wells that will be required to
capture volatile constituents from the target areas of contamination.
SVE pilot testing is an integral step in the process leading to proper SVE system
design. Table 4-4 presents a comprehensive list of activities that should be considered
during the planning and performance of a pilot test. The preliminary site charac-
terization activities consist of an assessment of (1) the vertical and horizontal
distribution of the hydrocarbon phases and type of hydrocarbon present and (2) the
local geology/hydrogeology (see Chapter 2). The geologic/hydrogeologic assessment
for evaluating the application of SVE systems should identify the different soil strata in
the unsaturated zone, assess the permeability of the soils that are contaminated (by
core tests, sieve analysis, etc.), determine the static groundwater table and seasonal
fluctuations, and identify and delineate any subsurface conduits, piping, tanks, etc., that
may influence airflow.
When modeling is used in combination with field measurements, laboratory column
studies are sometimes used in evaluating the feasibility of SVE systems. These tests
move large pore volumes of air through a soil column in order to evaluate the effec-
tiveness of the SVE system to remove contaminants in the soil, to determine the
residual contaminants that are not readily removed by the SVE system, and to estimate
vapor concentrations.. Column studies cannot simulate nonideal conditions. The
results of these studies can be generally similar to ideal model predictions and are most
effectively used if there is some question of model applicability or the exact nature of
the final leachate from the study.
The activities that can be conducted in a field-scale pilot study are listed in Table 4-4
(Johnson et al., 1992) and include: (1) applied vacuum/pressure and flow rate
152
-------
Table 4-4. Site Characterization and Pilot-Test Activities
(Johnson et al., 1992b)
Activity
Nc>- Description
Preliminary Characterization Activities
Hydrocarbon Assessment
• vertical/horizontal
hydrocarbon characterization (type, boiling point distribution, regulated
component identification)
Geologic/Hydrogeologic Assessment
identification of soil strata
permeability assessment (core tests, sieve analysis, etc.)
static water table determination (and seasonal fluctuations)
• . subsurface conduits, piping, tanks, obstructions, etc.
Laboratory Characterization Activities
Laboratory Soil Column Feasibility Studies (optional)
Field Pilot-Scale Activities
Airflow -vs- Applied Pressure/Vacuum Test
vacuum test for vapor extraction wells
pressure test for air injection wells
Effluent Vapor Characterization -vs- Time
• total hydrocarbon concentrations
• regulated compound speciation
hydrocarbon characterization (i.e., boiling point distribution)
speciation
Subsurface Pressure Distribution
as function of depth and distance
steady-state and transient measurements
Subsurface Vapor Concentration Distribution
• as function of depth and distance
hydrocarbon concentrations and composition
speciation
8 Groundwater Elevation Changes Resulting from Air Extraction/Injection
9 Groundwater Monitoring
• hydrocarbon levels
• dissolved oxygen
10 Tracer Gas Tests
153
-------
measurements that include vacuum and flow rate measurements for vapor extraction
wells, and pressure and flow rate measurements for air injection wells (for air sparging
pilot tests), (2) the off-gas characterization over time where the total hydrocarbons are
measured, the boiling point distribution is determined for the type of hydrocarbons
present, and the speciation of regulated compounds is determined for off-gas treatment
requirements, (3) subsurface pressure distribution as a function of depth and distance
under steady-state and transient conditions, (4) subsurface vapor concentrations, com-
position, and distribution to monitor SVE effectiveness, (5) groundwater elevation
changes from air extraction (for SVE systems) or air injection (for air sparging systems),
and (6) monitoring of groundwater quality primarily for changes in concentrations of
contaminant indicator compounds and dissolved oxygen (for air sparging systems).
Tracer tests using helium or sulfur hexafluoride are beginning to be used to better
define flow paths, flow strengths, and areas of treatment and efficiency for air sparging
and SVE systems. Typically, not all of these measurements and tests are conducted
as part of a field-scale pilot test for SVE. In fact, appropriate procedures and methods
are still being developed for both SVE and air sparging pilot tests. The following dis-
cussion will focus on common practices that are currently used in conducting field-scale
SVE pilot tests (CRTC, 1991).
4.4.3.2 Common Practices, Errors, and Limitations
A pilot test typically involves applying a vacuum to an extraction well and then collecting
vacuum and airflow data at the extraction well, collecting vacuum data at two or more
vacuum monitoring points, and collecting effluent vapor concentrations and
compositions. During the test, the steady-state pressure distribution is measured prior
to the application of pressure/vacuum and measurement of the flow rate versus applied
vacuum. Incremental steady-state pressure data are plotted versus the logarithm of
distance from the extraction well to graphically determine the radius of influence R,.
Well locations are selected to ensure that the calculated radii of influence overlap the
zone to be remediated.
When a pilot test is conducted to collect site data for SVE system design, the following
should be considered:
• Selection of existing wells, or location of newly installed wells, for vapor
extraction
- location and completion interval relative to delineated petroleum
hydrocarbons and soil strata
- location relative to potential observation wells
- location relative to man-made airflow pathways/barriers
- well construction (surface seal, filter pack and screen size, screened
interval)
• Selection of vacuum observation points
- existing wells, temporary vacuum monitoring points, or both
154
-------
• Data collection procedures
- vacuum measurement
- airflow measurement
- effluent vapor concentrations and composition
- monitoring frequency
• Equipment
- vacuum pump/blower
- effluent treatment ,
• Estimated length of pilot test
• Data interpretation
- determination of radius of vacuum influence.
The following sections will discuss the selection of vapor extraction wells, vacuum
monitoring points, pilot-test measurements, and collection of data for evaluatina svstern
design. * '
4.4.3.3 Extraction Wells
The extraction wells used in the pilot test are to be located at or near the center of the
area of highest contaminant concentrations. The screen slot size of the well and filter
pack should be consistent with proper hydraulic design protocols to promote com-
patibility with the surrounding formation and to prevent soil from entering the well
casing. The well should be completed with an adequate grout seal extending from
above the filter pack to the ground surface in order to prevent or reduce short circuiting
of air along the outside of the well casing or bore hole.
An existing groundwater monitoring well can be used as the vapor extraction well durina
a pilot test, providing: a
The screened interval targets the depth interval where contamination is
present.
The target interval is located sufficiently above the water table, so that
vapor extraction from that depth is unaffected by groundwater upwelling.
The monitoring well is constructed so that short circuiting will not occur
through annular materials.
The initial criterion for evaluating a well for purposes of vacuum extraction should be the
length of well screen extending above the water table during the maximum pilot-test
vacuum. The screened interval and filter pack should be long enough to prevent
screen blockage when the water table and capillary fringe rise under the maximum
155
-------
operating vacuum. The rise in water table elevation within the well casing will be
directly proportional to the expected pilot-test operating vacuum as measured in inches
of water. An evaluation of the screened interval of a potential vacuum extraction well
should include an additional 2 to 3 feet of capillary fringe in vadose zone soils outside
the well bore (extra fringe is required in fine-grained soil).
If high vacuums are expected and the well screen extends into or near the water table,
groundwater may need to be pumped during the test to counteract the expected rise in
water table and capillary fringe. Groundwater should be pumped at sufficient rates to
maintain a static water level. If no floating liquid hydrocarbon is present, groundwater
pumping rates may be increased to lower the water table adjacent to the extraction well
itself, or from an adjacent well.
Pumping of groundwater during pilot testing is advisable where the large-scale SVE
system is also likely to require groundwater extraction. When soil contamination is
present at or below the water table, or when water levels fluctuate over a large interval,
optimal SVE performance will depend on groundwater pumping to control and lower the
water table, thereby keeping contaminated soils exposed to airflow.
4.4.3.4 Vacuum Monitoring Points
The vacuum monitoring points for the pilot test are placed at varying distances and
depths and in varying directions outward from the extraction well. Ideally, the number
of vacuum monitoring points should be sufficient and placed to determine vapor flow
patterns that could result from subsurface heterogeneities. The screened intervals of
the vacuum monitoring points will coincide with the depth of contamination.
Many groundwater monitoring wells are suitable for vacuum monitoring in vapor
extraction pilot tests. The current network of groundwater monitoring wells should be
evaluated for this purpose by use of similar criteria as those used for the vacuum
extraction well: screen placement, well construction, and quality of surface seal. Wells
with a minimal screened interval above the water table and capillary fringe should not
be used because of the possibility of submergence during the test. Likewise, a well
with a screened interval extending to within 1 or 2 feet of ground surface may record
little or no vacuum because of its proximity to the surface. Suitable existing wells may
need to be supplemented with additional monitoring wells or temporary soil probes.
The number of groundwater monitoring wells suitable for vacuum monitoring at a
particular site may be limited, and the cost associated with installing new wells prohibi-
tive. Temporary vacuum monitoring probes are an alternative to installing additional
wells to supplement pilot-test vacuum data. Two types of vacuum probes are
commonly used: small-diameter PVC well screen and slotted or perforated steel
probes.
156
-------
4.4.3.5 Lateral and Vertical Placement of Vacuum Monitoring Probes
Radial placement of probes for vacuum measurement is somewhat arbitrary, but
generally they should be located at regular intervals outward from the extraction well
(e.g., about 5 to 10 feet, 10 to 20 feet, 20 to 40 feet, and greater than 40 feet)vlf
directional differences in pressure response are expected, as in extremely anisotropic
or fractured soils, two sets of vacuum monitor points may be oriented radially from the
main vent well and at right angles to each other.
The vertical section of vadose zone soils sampled by a given vacuum monitoring device
may control the measured vacuum value. For example, a pressure response reading
from a temporary probe will be a point value representing vacuum in a limited vertical
section of the unsaturated zone, while a pressure reading from a monitoring well
screened over a vertical section of several feet will be a composite of the vacuum from
the screened interval. Adjacent vacuum measurements from these two types of instru-
ments may vary markedly, especially if the temporary probe is placed in a unit having a
different permeability than the average permeability of the section penetrated by the
well screen. Since permeability affects the time required for vacuum to stabilize,
vacuum monitoring probes placed in low-permeability units may only measure trie
transient vacuum developed during the period of the pilot test, compared with an
adjacent monitoring well that measures stabilized vacuum in more permeable units.
This relationship is common in heterogeneous sites where airflow may be largely
restricted to beds of high-permeability sediments. As a result, if temporary soil probes
are used as the primary measure of vacuum pressure, it is best to install several probes
at different depths in close lateral proximity (i.e., "nested" probes).
Monitoring vacuum near the water table is desirable, but care must be taken not to
place the probes near or within the capillary fringe. When water table and capillary
fringe elevations rise during the pilot test, vacuum probes may fill with water.
4.4.3.6 Pilot-Test Measurements
The measurements required for pilot-test analysis are:
• Extraction well
- vacuum (inches of water)
- airflow rates (scfm)
- effluent vapor concentration and composition (portable field analytical
instrument with Tedlar bag samples for laboratory confirmation)
• Monitoring Points
- vacuum (inches of water)
- time of measurements.
157
-------
The vacuum and airflow measurements at the extraction wellhead, the vacuum distribu-
tion at the vapor monitoring points, and the effluent vapor concentrations and
composition are used to evaluate the design of a full-scale SVE system.
4.4.3.7 Vacuum and Flow Rate Measurements
Vacuum applied to the vapor extraction wellhead should be held constant, with frequent
readings taken to ensure this condition. The associated airflow readings are used to
determine if a given airflow is sustainable at the vacuum extraction well vacuum used
during the test. A procedure for determining airflow versus the applied vacuum and the
extraction wellhead is described by Johnson et al., 1992, and involves the following
steps:
1) Open the air inlet valve.
2) Close the valve leading to the wellhead.
3) Turn on the blower/vacuum pump so that air is being drawn in only
through the air inlet line.
4) Open fully the valve leading to the wellhead.
5) Once the flow rate has stabilized, record the wellhead vacuum and flow
rate from the extraction well.
6) In a series of increments, slowly close the air inlet valve until it is fully
closed.
7) For each increment, allow the flow rate to stabilize and record the well-
head vacuum and flow rate.
These wellhead measurements can be used with those collected from the vapor
monitoring points to evaluate the pressure distribution and airflow in the unsaturated
zone.
At regular intervals during the pilot test, vacuum readings are made at the extraction
well and all soil vacuum monitoring points. Soil vacuum readings should be made at
regular intervals until the vacuum field has stabilized (i.e., vacuum measurements at
observation points show little or no change at successive sampling intervals). Vacuum
readings should be taken frequently early in the pilot test (5- to 10-minute intervals) with
the time between readings lengthening over the course of the test. As shown in Figure
4-12, the stabilized soil vacuum will be a percentage of the vacuum applied at the vapor
extraction well. This percentage decreases with increased distance from the vapor
extraction well.
158
-------
r:
E
=5
3
O
03
CD
CO
LJLJ
O
0)
O)
OB
C
CD
O
CD
Q_
CO
(0
CO
E
O
03
O
CO
0
I
CO
ill
O
CD
O)
I
CD
C£
CD
a.
CO
to
CO
O
CM
CO
v
c
eo
|
3
O
i
O
i
(B
O
§
CO
I
3
O
±
"o
(0
01
^
I
O)
159
-------
Two vacuum levels should be utilized in each pilot test (step testing) in order to
establish a range of operating parameters. The highest pilot-test vacuum setting
should be near the maximum capacity of the blower used on site, and should be within
the likely range of the full-scale system. In a homogenous subsurface environment, soil
vacuum and airflow will stabilize nearest the extraction well first and will stabilize at
progressively greater distances from the well as the pilot test continues. At higher
permeability settings, the vacuum field adjacent to the extraction well typically stabilizes
rapidly and each vacuum level of the step tests may require 2 hours or less to
complete. Longer step tests may be necessary in lower permeability soils.
A source of error in measuring soil vacuum can result from the vertical location of the
vacuum monitoring probe screen. Recognizing that variations in depth can impact soil
vacuum measurements, many contractors try to place the probe screen as deep as
possible. As shown in Figure 4-13, however, field studies (CRTC, 1991) have indicated
that, under isotropic subsurface conditions, contours representing equal values for the
soil vacuum as a percentage of the applied vapor extraction well vacuum are nearly
parallel to the well screen (orthogonal to the water table) at depths equal to 50 to 75
percent of the depth to groundwater. This zone should then be a suitable interval for
the placement of vacuum monitoring probes. When the pilot test starts, groundwater
will respond to the vacuum applied to the vapor extraction well by upwelling. Any
probes too near the water table could then be plugged by the groundwater upwelling.
Vacuum monitoring probes that are plugged during a test must be disregarded, thereby
reducing the number of available data points for evaluating pilot test results. Vacuum
monitoring probes that are not plugged will result in an inaccurate estimation of the
radius of influence.
4.4.3.8 Effluent Vapor Concentrations and Composition
Effluent and subsurface vapors are sampled for total hydrocarbon concentrations by
use of a portable FID (refer to SVE system monitoring). Effluent vapor concentration
readings should be measured with a portable FID with the same approximate frequency
as the time interval for soil vacuum readings. Perhaps one in ten FID readings should
be followed by the collection of samples in Tedlar bags for confirmational analyses at a
laboratory. Of the samples collected in Tedlar bags, the initial, final, and one
intermediate sample should also be subjected to analyses that would identify the
concentrations of the individual constituents in the vapor stream. The concentration
data collected during the pilot test can be used to determine the need for effluent vapor
stream treatment during site remediation. The vapor composition analyses will provide
supporting information on the nature of the contaminant in the subsurface and can be
used in conjunction with some analytical design tools.
In conjunction with project operational system flow rates, the time-series analytical data
will allow calculation of expected hydrocarbon removal rates following start-up, and
determination of the size and nature of any effluent vapor treatment that may be
necessary.
160
-------
CO
£
**
_o
'5.
O)
•c
O)
O
I
3
O
CO
CO
3
2!
I
il I
161
-------
A PID may be used to determine the concentration of volatile aromatics in the system
effluent, but should not be used as a substitute for the FID to measure total
hydrocarbon removal rates. Baseline BTEX concentrations should be collected during
the pilot test and early in the life of the SVE system operation. Periodic measurement
of the BTEX fraction of total effluent serves to confirm system efficiency in contacting
the majority of contaminated subsurface soils.
4.4.3.9 Vacuum Radius of Influence, Airflow, and Hydrocarbon Removal Rates
Pilot tests typically simulate the pressure and airflow regime developed around a single
SVE well during a relatively short operating period. Airflow rates and hydrocarbon
vapor concentration/composition data can be used to calculate an initial hydrocarbon
removal rate per well. The subsurface distribution of pressure developed during the
pilot test is typically used to determine well spacing for the full-scale SVE system. It is
important, however, to understand the relationship between pressure, airflow, and
hydrocarbon removal rates in order to properly design the SVE system. Figure 4-12
illustrates the stabilized vacuum distribution and airflow lines for a hypothetical SVE well
installed in a 10-foot-thick vadose zone. The illustration is derived from the output of a
proprietary airflow model developed by Chevron Research and Technology Company.
The model uses the finite element technique to solve the Laplace equation for pressure,
and a related equation to solve for the discrete airflow line function, important
boundary conditions of the model are that atmospheric pressure exists at the surface
and at arc infinite horizontal distance from the well screen, and that there is circular flow
symmetry around the SVE well. The illustration depicts a two-dimensional view of
vacuum distribution and airflow for uniform soils in which horizontal permeability is twice
that of vertical permeability (Kj/Kv = 2).
The upper portion of Figure 4-12 shows the highest vacuum developed immediately
adjacent to the well screen, with vacuum decreasing exponentially with increasing radial
distance from the well and proximity to the surface. Airflow lines are orthogonal to
vacuum isobars, with the greatest percentage of airflow occurring adjacent to the
extraction well. Based on the boundary conditions of the model, radial vacuum never
reaches zero, but would decline to some low asymptotic value close enough to zero to
be unmeasurable in the field.
The vacuum radius of influence is defined as the radial distance from a vacuum extrac-
tion well at which soil pore pressure is equal to ambient atmospheric pressure. By
measuring soil vacuum at vacuum monitoring points during a pilot test, an approximate
vacuum radius of influence can be determined as the radial distance from the extraction
well at which induced vacuum is too small to be measured. It is important to note,
however, that the vacuum radius of influence does not correlate with the radial dis-
tance within which sufficient airflow is induced to adequately remediate soils within an
acceptable time frame. The lower schematic in Figure 4-12 shows that even though
measurable soil vacuum may extend beyond 60 feet, 80 percent of the airflow
originates in an area within 27 feet of the SVE well and 95 percent is obtained from
162
-------
within 35 feet. Thus, radial distances of 27 to 35 feet more accurately reflect an
"effective or design" radius of influence for soil remediation based on airflow rates.
By means of example, consider the hypothetical situation depicted in Figure 4-14. In
this case, the subsurface soil is uniformly contaminated by gasoline at a concentration
of 1,000 ppm. During the pilot test, a vacuum of 40 inches H2O was applied at the
vapor extraction well. The measurable radius of influence, based on soil vacuum
measurements of 0.01 inch H2O, was 60 feet. Figure 4-15 represents the anticipated
results of operating a vapor extraction system for two years as a function of the
percentage contribution to total vapor extraction well airflow. As shown in this diagram,
more than 90 percent of the initial contaminant mass has been removed from the soils
within =30 feet of the vapor extraction well where 95 percent of the airflow originated.
In that portion of the subsurface responsible for only 4 percent of the airflow to the
vapor extraction well, hydrocarbon concentrations in the soils had decreased only 40
percent to 600 ppm. Similarly, that region contributing <1 percent of the total airflow
realizes removal of only 5 percent of the initial contamination. Finally, soil
contamination levels outside this last zone will be unaffected by the vapor extraction
system.
Figure 4-16 is a plot of the hydrocarbon mass removal rate (as a function of time) at a
hypothetical site. As depicted, hydrocarbon mass removal began to approach
asymptotic conditions after approximately six months of operation. This would normally
be a reasonable indicator that site remediation was approaching closure. Figure 4-15
however, clearly showed that significant residual contamination was present within the
vacuum radius of influence of the vapor extraction well, as defined on the basis of soil
vacuum readings during the pilot test. This contamination results from the failure of
effective airflow to contact all soil within the vacuum radius of influence. If confirmatory
borings are not used as a precursor to closure, remediation at this site would cease with
appreciable contamination still present in the soils. The design flaw in this instance was
the assumption that the soil vacuum radius of influence matched the "effective" or
design radius of influence for effective subsurface airflow.
4.4.3.10 SVE Pilot Test Interpretation
The empirical "radius of influence" design approach relies on the results of a pilot test to
determine the vacuum radius of influence developed during the duration of the test and
uses this data to estimate an "effective" airflow radius of influence to determine extrac-
tion well spacing. It needs to be reemphasized that the vacuum radius of influence
observed during the pilot test rarely corresponds to the effective radius of influence in
terms of airflow contribution. Instead, the soil vacuum readings are used to determine a
conservative approximation of the airflow radius of influence.
The interpretation of the pilot-test vacuum data involves calculating a series of
normalized vacuum values using the following technique:
163
-------
• m
•^H t? ^^r
III
gg2 flj B
^^^ ^^v V
• •
,--„ £1 ^^
£§£
3 O3-
1040 jj
•2 fl) ^f
gZ
V
\
\
T
i
.
i
E
Q.
Q.
O
«—
0)
c
"o
m
%w
O
c
••
x
0
&
•i
•
o
Cl
l<
\
V
fii
D
3
B
^
3
? :
w
s
*
:'
•
;
*—
C^
HI
!•••
Tt
O
%w
DC
1
o
c
c
0)
3
•IBM
t*«
•••
|
03'
-
0
CO 4^
LL
0 ^
c t»
SS 0)
CO *••
•*-» *•
CO £
e t
CO -2
•— 0)
•n >»
1 z
DC >
0 «
w i
1
^
1 1
0 I
oi <
ft
5 i
4 i
Q) eg
O 3 *
J> i
164
-------
o
•*=
2
o
2
(0
£
3
&
CO
UJ
!
n
— o
a
o
c
m
<§
o
0.
•o
£
u
o
£
s:
iX "•'
•" Kl
ffi %
O)
o
CO
165
-------
uu
s>
a
(0
a
2
|
£
0)
10
w
i
TO
O
8
•O
>,
to
O)
fZ
UJ
CO
I
Zl
O
CO
166
-------
• Plot normalized vacuum data on semi-log graph.
• Fitting a straight line to data, determine site-specific vacuum/distance
function.
• Extrapolate radial distance corresponding to 1 percent operating vacuum.
These normalized vacuums are the ratio of the monitoring point vacuum to the vapor
extraction well vacuum.
The normalized vacuum is plotted on semi-logarithmic paper versus radial distance
from the vapor extraction well to the vacuum monitoring point (Figure 4-17). A straight
line is fit to the plotted data representing the radial vacuum distribution for the site, with
the slope of the line representing the horizontal-to-vertical permeability ratio (M
-------
II
s §
a I
11
3 CO
8 H
> 0)
— "O
"o £
CO S
to
oi
u
_co
^o
CO
O
13
I I I I
II I I 1 I I
o
•
o
o
o
(O
o
o
to
CO
o
o -g
CO «
QC
p
o
CM
p
o
fl)
u
d)
I
O
CO
r^
5
<
=>
168
-------
4.4.4 Screening and Detailed Modeling Tools
Commercial models are available to assist remediation contractors and regulatory
officials in determining the appropriateness and potential effectiveness of SVE as the
remediation technology option for a given site. SVE models are used in providing a
structured approach for understanding remedial processes and limiting factors that are
not readily understood by direct observation. Models are particularly valuable in
evaluating the performance of an SVE system prior to construction, thereby minimizing
the cost associated with trial-and-error system design and operation. Modeling leads to
a better examination of process feasibility, a more accurate evaluation of potential
peifonnance, and the development of system engineering design criteria prior to SVE
implementation. Two general types of SVE models are used: screening models and
detailed models. The detailed models include airflow models and compositional flow
and transport models. This section presents a short summary of the types of models
available for evaluating SVE systems. A thorough examination of the types of problems
addressed by SVE models, the need for and selection of appropriate models, and the
data needs and a review of available SVE models are presented by EPA (1994). Table
4-5 presents a summary of the general types of models.
4.4.4.1 Screening Models
SVE screening models are primarily used to evaluate the feasibility of SVE at a specific
site based on limited input data. These models are not intended for evaluating detailed
SVE, although preliminary conceptual design plans can be examined with model
results. Johnson et al. (1990a,b) presented a useful screening approach for determin-
ing the feasibility of SVE at a particular site. This practical approach makes use of
analytical equations that estimate VOC removal rates and pressure distributions for
various SVE design parameters. The two models that were developed based on this
approach are Hyperventilate and VENTING.
Hyperventilate, which was developed independently from VENTING, is designed to be
used as an instructional tool to identify required site data, decide if SVE is appropriate
at a site, evaluate air permeability tests, and estimate the minimum number of wells
needed. It is especially useful for a quick feasibility evaluation based only on soil per-
meability and the thickness of the screened interval. Soil permeability can be
calculated from air-pumping tests or estimated from the types of soils present. The flow
rate is then calculated from the input of the soil permeability, desired radius of influence
of the extraction well, screened interval for the well, and extraction well diameter or
radius. Calculated estimates of extraction flow rates and vapor concentrations can be
used to estimate maximum mass removal rates. The analytical mass removal rates
assume the presence of residual free product in the vadose zone. Two boundary-layer
screening calculations are included for estimating removal rates above a liquid layer of
free-phase NAPL and above low-permeable soil with residual contamination.
Hyperventilate contains "help cards" that define the equations used and also provide
supplementary information. Chemical files for weathered and unweathered gasoline
169
-------
Table 4-5. Summary of the Screening, Airflow, and Compositional Flow and
Transport SVE Models (EPA. 1994) •
Hyperventilate, v2.0 (IBM PC), v1.01 (Apple Macintosh)
Type - Screening
Functionality - Simplistic steady-state, radial-symmetric airflow and transient one-dimensional
multicomponent contaminant transport
Solution Methodology - Analytical solution; finite-difference solution of a one-dimensional mass
balance equation
Assumptions - Two-dimensional, radial, confined airflow to a vapor extraction well; one-
dimensional, mass-balance approach, volatilization based on Raoult's law
Capabilities - Calculates air permeability, well flow rates, mass removal rate; residual leaks
removal rates in 2 ideal mass transfer limited scenarios; calculates contaminant concentrations
over time for multiple constituents
Advantages - Provides rapid estimates for determination of the potential feasibility of SVE;
provides rapid estimates of contaminant concentrations in extracted gas, allows comparison of
removal rates of different constituents
Limitations - Analytical airflow solution; mass removal rates based on advection from free-phase
NAPL, diffusion-limited models for two scenarios are given; should not be used to design SVE
systems
Hardware/Software Requirements - IBM PC or Compatible 80386/80387 coprocessor or
80486,4 MB, RAM, DOS 3.1 or higher, Microsoft Windows 3.x and runtime version of Object
PLUS; Apple Macintosh (Plus, SE, SE230, II, IIX, or portable): 1 MB RAM, Apple HyperCard
Software (v2.0 or greater)
Availability -Available from EPA as EPA/600/R-93/028 (EPA ORD Publications, 513/569-7562),
Price: FREE. Available to Public from NTIS, Price: $22 IBM PC, $17 Macintosh. Object PLUS
available from Object PLUS Corp., 125 Cambridge Park Dr., Cambridge, MA 02140, Price: $100
(run time version)
Venting, v3.1
Type - Screening
Functionality - Transient, one-dimensional multicomponent contaminant transport.
Assumptions - Calculations based on user-defined flow rate, assumes equilibrium partitioning
between phases in a one-dimensional volume of soil.
Limitations - User supplies flow rate to extraction well; simplistic one-dimensional representation
of mass transport; should not be used to design SVE systems
Hardware/Software Requirements - IBM PC/AT or Compatible, DOS, 512 KB RAM, math
coprocessor
Availability - Environmental Systems and Technologies, Inc., 2608 Sheffield Drive, Blacksburg,
VA 24060, 703/552-0685. Price: $400.00 ________________.^_
170
-------
AIRFLOW™, V3.01 Airflow
Type - Airflow
Functionality - Steady-state, radial-symmetric (two-dimensional cross-section) airflow
Solution Methodology - Finite-element solution of the airflow equation
Assiumptions - Based on Darcian flow of an ideal compressible gas in a porous medium
Capabilities - Calculates pressure distribution in a radial domain, calculates airflow pathiines and
velocities
Advantages - Easy-to-use CAD-type graphical user interface which simplifies model input and
setup; rapid setup for simple problems, aids in hypothesis testing; many sample problems
included with the code
Limitations - Only allows for one extraction well; no mass removal
Hardware/Software Requirements - IBM PC or compatible, 80386/80486, 4 MB RAM, DOS 2.0
or higher, mouse and math coprocessor for 80386-based machines recommended
Availability -Waterloo Hydrogeologic Software, 19 McCauley Drive (RR#2), Bolton, Ontario
Canada, L7E SR8, 905/880-2886, Price: $650.00
CSUGAS
Type - Airflow
Funltionality - Transient, two- or three-dimensional airflow
Solution Methodology - Finite-difference solution of the airflow equation
Assumptions - Based on Darcian flow of an ideal compressible gas in a porous medium
Capabilities - Calculates vacuum distribution in the subsurface, in inches of water
Advantages - Allows full, three-dimensional analysis of heterogeneous, multiwell airflow
problems; text-based input/output is flexible and up to the user
Limitations - Lack of easy-to-use input/output interface may intimidate beginners; no steady-
state solution, option; no mass removal
Hardware/Software Requirements - IBM PC AT/XT or compatible, 640 KB RAM, DOS 2 0 or
higher
Availability - Dr. James W. Warner, Department of Civil Engineering, Colorado State University
Fort Collins, CO 80523, 303/491-5048, Price: $125
AIR3D
Type - Airflow
Functionality - Three-dimensional airflow
171
-------
Solution Methodology - Finite-difference solution of the airflow equation posed in terms of the
groundwater flow equation and solved by the MODFLOW code •
Assumptions - Based on Darcian flow of an ideal compressible gas in a porous medium
Capabilities - Calculates pressure distribution in the subsurface
Advantages - Easy-to-use CAD-type graphical user interface which simplifies model setup and
input; allows three-dimensional analysis of complex problems
Limitations - Users need to have an awareness of the operation and limitations of the
MODFLOW code; no mass removal
Hardware/Software Requirements - IBM PC or compatible, DOS 3.3 or higher, 4 MB RAM,
FGA card and color monitor, mouse is highly recommended
Availability - American Petroleum Inst., 1220 L Street Northwest, Washington, DC 20005, Price:
$500.00
VENT2D.V1.3
Type -Airflow and multicomponent contaminant transport
Functionality - Steady-state, two-dimensional airflow and transient contaminant transport
Solution Methodology - Finite-difference solution of the airflow equation, finite-difference
solution of the transport equation
Assumptions - Transport equation is simplified by ignoring mechanical dispersion (includes
diffusion)
Capabilities - Calculates pressure distribution in the subsurface, multicomponent contaminant
constituent concentrations over time in the subsurface
Advantages - Only readily available compositional flow and transport code; source code is
available; text-based input/output is flexible and up to the user
Limitations - Grid size limited to 25 x 25 cells (can be increased with a different version available
from the author)
Hardware/Software Requirements - IBM PC or compatible, 80x86 with math coprocessor, DOS
3.0 or higher, 525 KB RAM
Availability - David A. Benson, 425 Claremont Street, Reno, NV 98502, 702/322-2104, Price:
$495.00
172
-------
are provided, and a customized file can be created for other compound mixtures.
Hyperventilate has a friendly graphic-user interface of useful calculation cards for
estimating airflow rates, soil permeability, and mass removal rates.
Both Hyperventilate and VENTING are based on assumptions of uniform properties
and subsurface geometries and use an analytical steady-state, horizontal radial flow
solution to produce flow to a vertical extraction well (Johnson, et. al. 1990a,b).
Advantages of these models include minimal data requirements and quick setup and
execution. These models can provide initial estimates for determining whether SVE is
appropriate for a given site.
Although relatively simplistic, the analytical solution approach used in these models can
provide reasonable results for sites where airflow is essentially confined and the airflow
patterns are primarily horizontal. Such sites include those with surface seals such as
pavement cover with no short-circuiting in the subgrade, or sites where the zone of
contamination is confined by low-permeability layers. The basic model principles and
computer software structure are discussed further in this section.
VENTING can be used to estimate the rate of VOC removal from the vadose zone
based on user-defined extraction well flow rate from single or multiple wells. This
model assumes a steady-state airflow, equilibrium or diffusion-controlled phase par-
titioning, and complete mixing within the contaminated zone to complete the extracted
mass of each contaminant constituent during the extraction time. The mass balance
considers partitioning among the free-product aqueous, adsorbed, and vapor phases
and assumes that all the contaminant mass is homogeneously distributed at all times in
a defined volume of soil. It also assumes that the aqueous and adsorbed phases make
negligible contributions to the vapor phase. The mass removal rates of the more-
volatile constituents in a multicomponent mixture are iteratively calculated. The
volumetric airflow rate is the key parameter that determines the VENTING modeling
results. The flow rate may either be input directly based on field measurements or may
be estimated based on the permeability of the contaminated soil and the vent pressure.
VENTING also provides a method of estimating permeability by use of permeability test
data. Hydrocarbon composition files can be created to produce common hydrocarbon
blends. Composition files for fresh and weathered gasoline are also provided. The
model can also generate mass-removal-versus-time plots for each constituent in a
multicomponent mixture.
4.4.5 Detailed Models
4.4.5.1 Airflow Models
Subsurface airflow models are used with the two- or three-dimensional flow of air
through a porous medium as a result of the pressure gradient created by an extraction
well. These models do not consider contaminant concentrations in soil vapor, mass
removal, and overall SVE system effectiveness.
173
-------
Airflow models can be used to develop a detailed design for SVE systems involving well
placement/location for multiple extraction wells. Depending on the type of airflow model
used, heterogeneities such as low-permeability zones and layered soils can be simu-
lated in either two or three dimensions. In general, these numerical models can provide
a more detailed analysis of the airflow field because they have fewer restrictive assum-
ptions. Airflow models can account for nonideal conditions such as a leaky ground
surface boundary condition or the presence of low-permeability layers. The potential
pressure distributions, airflow, and different radius or zone of influence for a well system
can be simulated to determine optimum well placement to visualize the differences in
permeability and preferential flow paths in determining the zones that will be most
affected by the SVE system. An overview of the airflow models CSUGAS, AIRFLOW,
and AIR3D is provided below (see Table 4-5).
CSUGAS is a three-dimensional finite difference model (Sabadell, 1988) that
numerically simulates the flow field of a compressible gas in a porous medium as a
result of the influence of an SVE system. The finite differences method is used to
numerically approximate a solution to the system of equations. This method also allows
for use of a heterogeneous and isotropic porous medium with airflow under steady-
state or transient conditions. Model applications include selecting design parameters,
determining feasibility of SVE at a particular site, and evaluating proposed modifications
to existing SVE systems.
AIRFLOW is a two-dimensional finite element radial-symmetric model (Waterloo Hydro-
geologic Software, 1993) that simulates the flow of vapors in the unsaturated zone. It
calculates steady-state pressure distribution, airflow, and pathlines in cross section to
the extraction well for an ideal, compressible gas. It can be an effective tool for ana-
lyzing the effects of layering, surface seals, and low-permeability zones on the flow of
air to an SVE well. Different vapor characteristics can be simulated by using different
vapor pressure, molecular mass, viscosities, and temperatures. The model can simu-
late heterogeneous and isotropic permeability zones. A variety of boundary conditions
can also be imposed.
AIR3D is a finite difference model that calculates steady-state or transient air pressure
distribution and airflow in the vadose zone resulting from inducing vacuum at SVE
wells. AIR3D uses the MODFLOW groundwater code (McDonald and Harbaugh, 1984)
to solve the equation of airflow in the vadose zone (Joss and Baehr, 1994). AIR3D
transforms the input variables in terms of pressures required for MODFLOW. The
model can simulate anisotropic subsurface conditions where the user defines the zones
of different permeability that can be defined independently of the layers.
AIR3D also includes an optimization module as an aid in selecting optimal well
locations and extraction rates. This module is used to select the number of potential
well locations, range of extraction rates, and a set of pressure gradient constraints; the
optimization module then identifies the best configuration.
174
-------
4.4.5.2 Compositional Flow and Transport Models
Compositional flow and transport models are used to evaluate both the airflow field arid
the transport or removal of contaminant constituents. Some of these models simulate
transport of multiple constituents in mixture when multiple phases are present. These
models can be used as an aid to determine residual concentrations of less-volatile
constituents in the mixture. Most available compositional flow and transport models
consider only a single set of vapor properties, thereby requiring a simulation to be run
for each constituent. Compositional flow and transport can be used in evaluating SVE
design and operation from well placement to contaminant extraction rates.
VENT2D incorporates steady-state airflow and transient multiple-constituent transport.
Unlike other models, it simulates advective and diffusive transport for a number of
chemical constituents simultaneously in two dimensions. The model also solves for
equilibrium distribution of each constituent among four phases (vapor, adsorbed,
dissolved, and nonaqueous phase liquid). Nonideal conditions considered in this model
include nonhomogeneous soil permeability, leakage of atmospheric air into the
subsurface, and irregular contaminant distribution of each contaminant constituent.
Contaminant saturation relative to airflow regimes is considered to reflect the depletion
of pore fluids over time.
This model can be used by a large number of practitioners to simulate multiconstituenl
vapor flow in multiple dimensions through unsaturated soils with variable permeability
and contaminant distributions among a number of phases. This simulation is especially
useful in evaluating irregular SVE well-field geometries. Dispersion is considered only
on the macro-scale in order to simplify the execution time and to minimize the data
needed.
4.4.5.3 Detailed Model Applications
The detailed models presented do not represent an all-inclusive list, but are examples
of the types of models currently available. The selection of one of these or similar
models should, in general, be based on the complexity of the site and the questions
that need to be answered concerning SVE system design. The AIRFLOW model is
perhaps the easiest to use of these detailed design models, but it can only provide a
steady-state solution to a radial flow problem in cross-section. CSUGAS and AIR3D
can provide analysis at different time steps and, therefore, they can provide information
on the expected change jn pressures and flow velocities in soils with varying permea-
bilities. This function is particularly important in evaluating field observations in low-
permeability soils where steady-state conditions might take months or years to achieve
CSUGAS and AIR3D can also stimulate the airflow for complex SVE system designs
such as horizontal wells or piping structures. These models, however, do not calculate
chemical partitioning and transport. Although these models are more flexible for the
evaluation of airflow regimes, they may be more time-intensive to set up and execute
may require greater hardware capabilities for complex problems, and may require more
data than is practical.
175
-------
VENT2D is the only combined compositional and steady-state flow model presented
here. As previously discussed, this model can stimulate the chemical and physical
processes affecting the movement of multi-constituent vapor-phase chemical mixtures
(such as gasoline). The principal application of this model is to more accurately
evaluate the rate at which the various chemical constituents will be extracted From the
vadose zone. This model would be appropriate for estimating mass removal rates from
a flow-field in order to determine SVE system performance. Other models with
combined compositional and transient airflow solutions have been developed for
internal use or are currently under development. As these models become available,
they should provide additional tools for evaluating sites with complex physical and
chemical conditions.
Recent SVE system designs for removing volatile organic compounds (VOCs) have
mostly been empirically based because of the simplicity of the process and the lack of
an understanding of the use of SVE models in aiding system design. Compositional
flow and transport models have practical applications in actual field situations that can
be used to evaluate the effectiveness of SVE in removing organic vapors. Sensitivity
analyses can be used to determine the role of soil moisture, temperature, soil
heterogeneity, and other factors in controlling the migration of volatile constituents
through the unsaturated zone. The process of contaminant desorption from soil
particles involves three consecutive mass transport steps during the operation of the
SVE system. This process can be examined when final cleanup efficiency is
determined. It also can result in significant differences in removal rates for the various
types of soils and volatile organic components.
All of the airflow and compositional flow and transport models are useful tools for
estimating well placement; however, they must be used with an understanding of the
model assumptions and limitations. For example, some models are based on the
number of soil pore volumes. Estimates for contaminant removal are then used to
determine appropriate airflow rates. The models should evaluate three-dimensional
airflow (Shan et al., 1992) to account for differences in vertical and horizontal air
permeability and the boundary conditions of air that enters the well through the ground
surface. Other models are based on horizontal airflow only and do not take into
account vertical air recharge from the ground surface. The models are useful for
screening and rough estimates, but are not designed to determine an exact distance for
well placement.
4.5 SVE System Monitoring
4.5.1 Common Monitoring Practices
SVE system monitoring is performed to determine the amount and movements of
contaminants in the subsurface before, during, and after remediation. The overall
objectives of a monitoring program are to: 1) assess site conditions to determine
176
-------
remediation approach, 2) evaluate the progress of in situ treatment, and 3) determine
site conditions following treatment.
In practice, SVE monitoring is typically performed to meet regulatory (compliance)
requirements for site closure and off-gas treatment, and, for a limited number of sites, to
assess or optimize system performance.
SVE performance monitoring of airflow rates and vapor-phase concentrations and com-
position in extracted vapors directly measures the rate of volatile hydrocarbon removal
by the system. All too often, SVE effluent concentration and flow rate data are collect-
ed as a secondary consideration to equipment maintenance, and are not specifically
included in a monitoring program to evaluate system performance. The parameters
typically measured during the monitoring of SVE systems include the following.
Vapor flow rates at each extraction well and injection well. Measurements
can be made by a variety of flow meters, pitot tubes, and orifice plates.
• Vacuum/pressure at each extraction and injection well and at monitoring
points. These readings can be measured with manometers and
magnehelic gauges. Vacuum/pressure should also be monitored at each
soil gas probe location.
• Vapor concentrations and composition from each extraction well. Vapor
concentrations can be measured by an on-line total hydrocarbon analyzer
calibrated to a specific hydrocarbon. This information can be combined
with vapor flow rate data to calculate removal rates (mass/time) and the
cumulative amount of contaminant removed. Soil gas measurements
should be made periodically at different radial distances by using soil gas
probes to monitor the reduction in contaminant vapor concentration.
Temperature of the soil and ambient air. By monitoring soil temperatures,
Conner (1988) predicted that biodegradation was occurring in the zone of
contamination. At locations with large seasonal differences between air
and soil temperatures, extraction air temperature is also a qualitative
measure of air residence time in the soil.
" Water table elevation. For soils with a relatively shallow water table, water
level measurements can be made with electronic sensors located in
airtight monitoring wells.
» Meteorological data. These data also include barometric pressure,
precipitation, and similar data.
177
-------
In addition to these parameters, product thickness on the groundwater should be
measured if more than one-eighth inch exists on the water table. Table 4-6 provides a
comprehensive list of data interpretation options for SVE-based systems and the
corresponding data needed for interpretation and analysis.
4.5.2 Performance Data Quality
The Environmental Group of the Chevron Research and Technology Company con-
ducted a nationwide study to examine and evaluate SVE performance monitoring data
from 143 SVE systems operated by Chevron, USA, Inc. (Buscheck and Peargin, 1991).
The majority of the SVE systems were installed to remediate gasoline releases. Of the
143 SVE systems in operation, 15 newly installed systems were excluded because of a
lack of available information. Figure 4-18 shows the number of SVE systems and their
period of operation that were used as screening criteria for this study. The remaining
SVE systems were further screened to limit the number of systems evaluated based on
the following criteria:
• In operation less than six months.
• Review of site assessment reports, SVE monitoring reports, and other
information.
• Overriding imprints of operational or hydrologic overprints on SVE perfor-
mance data (e.g., periodic opening of effluent dilution valve, fluctuating
water table submerging contaminated soils). Only 26 percent of the data
generated from SVE system monitoring were adequate for evaluating
SVE system performance; 34 percent of the SVE systems had been in
operation for less than 6 months; and 40 percent of the monitoring data
was inadequate. This subsection presents the results from the evaluation
of the SVE system performance data that were determined to be
adequate.
4.5.3 Monitoring Frequency
SVE performance should be monitored frequently enough to accurately represent both
the variability in the data set and the overall decline of hydrocarbon removal rates over
time. Collection of monitoring data on too frequent a basis can generate unneeded
quantities of data and can add to the cost of system operation. Selection of an
appropriate monitoring frequency is a compromise between data quantity and project
costs, and may be influenced by site-specific factors (e.g., location). The majority of
SVE systems are monitored either weekly or monthly. This monitoring interval can
result from the common practice of weekly (sometimes daily) monitoring of SVE during
the period following system start-up, followed by monthly monitoring after several
weeks have passed.
178
-------
Table 4-6. Process Monitoring Options and Data Interpretation
(Johnson, et al., 1992b)
Data Interpretation/Analysis Requirement Data Collection'
Concentration vs. time 1
Composition vs. time
Flow rate vs; time
Applied pressure/vacuum vs. time
Mass removal rate [mass/time] vs. time
Cumulative removed by volatilization [mass]
Identify mass transfer limitations '
Aerobic biodegradation contribution to removal rate [mass/time] vs. 1, 2, 6°
time:
Aerobic biodegradation contribution to cumulative amount removed
[mass]
Total remediation costs [$] vs. time 1, 2b, 3
Cost per mass of hydrocarbon removed [$/kg removed] vs. time j
Effect of environmental factors [qualitative] 1, 2b, 4
In situ assessment of treatment with time [qualitative areal impact] 1, 2b, 4°, 5, 6b, 8°, 9s
Defined zone of vapor containment [qualitative areal impact] 1, 5°, 7, IP
Closure monitoring report 1, 2b, 3°, 4° 5 7 8 9 10 11°
Areal impact of air sparging 1, 2, 4°, 5°, 6°, 7, 8°, 9, 10, 11°
Effect of water table elevation changes 1,2,4, 5, 6, 7, 9, 10
Injection/extraction flow rate optimization 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11
Flow field definition
»Key: b = Applicable for bioventing applications; o = Optional, or as required; s = Relevant to air sparging.
Data Collection Key:
1 = Process monitoring data; extraction/injection flow rate(s) and vacuum(s)/pressure(s), extraction vapor
concentration and composition.
2 = Respiratory gas (O2, CO2) monitoring of extracted vapor stream.
3 = Cost monitoring; capital, operation and maintenance, and utilities costs.
4 = Environmental monitoring; temperature, barometric pressure, precipitation.
5 = In situ soil gas monitoring; vapor concentration and composition.
6 = In situ soil gas monitoring; respiratory gases (CO2 and O2).
7 = Subsurface pressure distribution monitoring.
8 = Soil samples.
9 = Groundwater monitoring.
10 = Groundwater elevation monitoring. '
11 = Tracer gas monitoring.
179
-------
CL
O
0)
GL
180
-------
To illustrate the effects of sampling frequency on performance data interpretation,
Figure 4-19 has been set up to show a complete SVE performance data set from a site
in San Diego, California. These graphs show monitoring results for hydrocarbon mass
removal rates for all raw data as well as results of biweekly and monthly sampling.
These plots show relatively good agreement in data trends between the full (raw) data
set, and data sets comprised of biweekly and monthly sampling frequency data.
Monthly SVE monitoring appears to be sufficient. More frequent sampling (weekly or
biweekly) will generate more data, which increases the confidence level of data analy-
ses, but may not be cost-effective in all cases. :
4.5.4 Airflow Rates
Results from the evaluation of SVE performance monitoring have led to the following
conclusions: : .
• Hydrocarbon mass removal rates (Ib/dayj are sensitive to changes in airflow.
• Airflow should be measured directly and not be estimated from blower
performance curves.
• Performance should be measured directly with a dedicated device.
• Use of an orifice plate and averaging pitot tube are cost-effective and they have
low maintenance requirements.
SVE flow rates expressed as volume of air removed/day are extrapolated from mea-
surements of airflow in standard cubic feet/minute. Therefore, calculated daily
hydrocarbon mass removal rates are sensitive to relatively small changes in measured
airflow rate. At some sites, airflow rates are estimated from an initial system vacuum
measurement taken shortly after installation, or from the theoretical maximum capacity
of the blower. Inaccurate (or nonexistent) airflow rate data severely compromises SVE
performance data analysis.
Airflow rates all too often are derived from measured total system vacuum applied to
blower performance curves. Most manufacturers provide standard performance curves
for their equipment based on testing at standard conditions (70°F and atmospheric
pressure at sea level). The following operating factors, however, will significantly alter
the accuracy of this information:
• Temperature
• Atmospheric pressure
• Inlet vacuum and
• Discharge pressure
• Moisture content.
181
-------
Rosecrans St., San Diego. CA
VES Performance Data Sampling
CO
X 3
CO
CO
CO
IRAWDATA]
12/23/88
7/11/89
1/27/90
• 8/15/90
CO
>
I
CO
1
IBI-WEEKLY!
12^3/88
7/11/89
1/27/90
8/15/90
1
X 3
1.
re
MONTHLY!
12/23/88 7/11/89 1/27/90 8/15/90
Figure 4-19. SVE monitoring data for a San Diego, CA, site.
Source: after USEPA, 1993a.
182
-------
Vacuum pump or blower performance curves provided by the manufacturers should not
be used as a primary source of flow rate measurement. Performance curves are better
employed as checks for the dedicated airflow metering device, rather than as the
primary means to measure flow rate.
4.5.4.1 Dedicated Air Flowmeters
Airflow rate measurement should be performed with a dedicated flowmeter that is
calibrated to the specific dimensions of the vapor extraction system. The proper
selection, calibration, and installation of a flowmeter will provide accurate flow measure-
ment without significantly reducing system performance through excessive flow losses.
The relative cost of installing dedicated flowmeters is nominal when compared to the
cost of designing and installing a full-scale SVE system.
4.6 Effluent Monitoring
SVE effluent concentrations should be measured with a broad-spectrum hydrocarbon
vapor detector with linear response to changes in both hydrocarbon concentration and
composition. Ideally, a hydrocarbon vapor detector should generate an electrical output
that is proportional to changes in both hydrocarbon vapor concentration (hydrocarbon
vapor volume/unit effluent volume) and composition (mass/unit effluent volume).
Explosimeters, flame ionization detectors, and photoionization detectors all exhibit
linear response with respect to hydrocarbon concentration over given ranges, but vary
in their individual response to changes in overall hydrocarbon vapor composition. A
detector with linear response to hydrocarbons is necessary to accurately convert mea-
sured SVE effluent hydrocarbon concentrations to hydrocarbon mass removal rates.
The LEL (lower explosive limit) meter measures hydrocarbon concentrations in SVE
effluent as a function of the heat of combustion of hydrocarbon vapors passing over a
catalyst-coated wire. LEL meter response is linear across a broad range of hydro-
carbon concentrations (100 to 10,000 ppmv), but is not linear with respect to changes in
hydrocarbon vapor composition because changes in average molecular weight of
effluent gases are not proportional to the heat released during combustion. This
limitation means that expected changes in volatile hydrocarbon composition throughout
the life of the SVE system will not be accurately converted to mass removal rate as
measured by the LEL meter. This limitation makes the LEL meter an inferior detector
for SVE effluent monitoring purposes when compared to other devices such as
photoionization and flame ionization detectors.
The gas chromatograph (GC) and PID are sensitive to volatile aromatic constituents but
do not detect methane or other alkanes. Volatile aromatic hydrocarbons represent only
a limited fraction of the total spectrum of volatile hydrocarbon compounds in SVE
effluent, and are not representative of the overall hydrocarbon mass removal efficiency
of the system. Volatile aromatic concentration data cannot be extrapolated to represent
total volatile hydrocarbon concentrations because the ratio of aromatic to total volatile
hydrocarbons decreases throughout the operating life of the SVE system (Johnson, et.
183
-------
al., 1990a). PID readings, however, can provide limited information on the weathering
of the released product and fractionation of volatiles. Aromatic hydrocarbons are
approximately three orders of magnitude more detectable than other hydrocarbons.
Consequently, BTEX constituents measured by a PID can provide limited use for SVE
monitoring.
When calibrated with an appropriate gas (usually hexane), the FID and GC/FID can be
used to quantify mass-per-unit-volume concentrations of hydrocarbon vapors in SVE
effluent, even though hydrocarbon composition changes with time. FID response is
proportional to the number of ionized carbon molecules released by the hydrogen
flame, and is reasonably linear with respect to changes in both hydrocarbon
composition and concentration (EPA, 1990a). The FID detects all hydrocarbons,
including oxygenates, alcohols, and ethers, over a concentration range of
approximately 1 to 10,000 ppmv. FIDs are also more sensitive than PIDs to alkanes
such as hexane and butane, which make up a higher fraction of gasoline than do the
aromatics. The relative differences between hydrocarbon detectors is presented in
Table 4-7.
Table 4-7. Comparison of Hydrocarbon Field Analytical Instruments
Detection
Type of Limit
Detector Hydrocarbons (ppmbyvol- Recommended
Type Cost Detected ume) Use
Portable Flame $6-8,000 All 1 Field TPH
lonization Detector
(FID)
LEL Meter $2,000 All 100 Safety
Portable $5-7,000 Aromatics 0.1 Field BTEX
photoionization
detector (PID)
4.6.1 Periodic BTEX Monitoring
Johnson, et. al. (1990a) documented a shift in SVE effluent hydrocarbon composition
from highly volatile to progressively less-volatile compounds during system operation:
This "chromatographic" shift of hydrocarbon vapor composition is largely a function of
higher vapor pressures allowing more rapid mass transfer of highly volatile hydrocarbon
compounds to subsurface airflow from the sorbed, free liquid, or dissolve phases.
Highly volatile compounds, such as BTEX, should be removed fairly early during
system operation, while less-volatile hydrocarbon fractions will take a significantly
longer time to be removed.
184
-------
nwQ,p °f total SVE effluent lsa useful ^chnique to
confirm whether SVE performance for a given site is likely to be advection- or diffusion-
Advection-controlled sites (i.e., sites where advection is the primary
transport mechanism) should exhibit a rapid decline in BTEX fractions
relative to total hydrocarbon mass removal rates, significantly before '
overall hydrocarbon removal reaches near-zero asymptotic values This
decline occurs for these sites because the majority of hydrocarbon
contaminants are being contacted by airflow.
Diffusion-limited sites should exhibit a rapid decline in BTEX effluent
fraction, followed by stabilized low-BTEX levels persisting in SVE effluent
even as overall hydrocarbon mass removal rates decline to non-zero
asymptotic values. The continuing presence of measurable BTEX at
these sites can be attributed to lower permeability soils yielding a broad
spectrum of hydrocarbon vapor compositions to SVE effluent.
4.6.2 Hydrocarbon Mass Removal Rates
A review of SVE hydrocarbon removal rate plots from 15 sites suggested the existence
llch nf th 9»nef that C°rrelate with Mrogeologic conditions and SVE flow rates
Each of these "categories" is discussed in the following subsections. !
4.6.2.1 Category 1 Sites
SVE performance data for Category 1 sites is best described by the exponential
regression, with an asymptotic mass removal rate near zero as shown Tn pfgure 4-20.
PriVldeS ! summary of the Category 1 sites including the duration of SVE
airflow rates, number of venting wells and exponential rate constant and a
hydrogeolog,c descr.pt.on of each site. In general, these sites are characterized by:
• Typically medium- to coarse-grained sediments with high permeability.
• Per-well airflow rates 25 to 50 scfm or greater.
i
• Exponential rate constants (k) that fall within the narrow range of 0.0045 to 0.0067
With one exception, depth to groundwater varied from 3 to 18 feet at these sites At
Site H, groundwater exceeds 100 feet in depth. Sites B, C, and G are now closed
Two of these sites (Sites B and E) are underlain by finer-grained sediments At Site B
groundwater was pumped at three gallons per minute (gpm) in order totowerihe water
table, concurrent with vapor extraction. During groundwater pumping, sediments S
185
-------
o
*"*
o.
(ft
•o
o
03
CD
0) -
QC
c
CD
c
o
X
lil
(1)
II
J£,
£
'55
3
CO
o
(£>
O
U)
O
CO
o
CM
cejea
CM
4
D)
wj
o
a£
IU
3
&
g
(O
186
-------
are coarser-grained than the overlying silty clays were exposed to vacuum. The largest
exponential rate constant, k, for all the Category 1 sites was calculated for Site B as
0.0091 day1. This system was operated for the shortest period (506 days), before site
closure.
Table 4-8. Summary of SVE System Data for Category 1 Sites
Site
Site A
SiteB
SIteC
Site D
SiteE
Site F
Site G
Site H
Closed
No
Yes
Yes
No
Nob
No
Yes
No
Duration,
Days
886
506 .
' 1131
881
1202
424
997
675
Airflow
Rate, scfm
61-116
34-59
350-380
52-130
75-140
33-83
70-80
85-115
No. of
Venting
Wells
4
1
36
3
3
3
3
1
ka[day1!
0.0060
0.0091
0.0057
0.0049
0.0030
0.0067
' 0.0053
0.0045
a Exponential rate constant.
b Closure petition submitted to state.
The unsaturated zone at Site E consists of sands and clays overlying natural silt loams
(2 to 6 feet deep and 4 feet thick), which in turn overlie igneous bedrock. Uncon-
solidated sediments at Site E are finer-grained than those found at most of the other
Category 1 sites. The smallest exponential rate constant for all the Category sites was
calculated for Site E where k = 0.0030 day1. This site has operated for longer than any
of the other sites in this category. The calculated rate constants and operating histories
for Sites B and E illustrate that cleanup time is apparently inversely proportional to the
exponential rate constant.
The performance of the Category 1 systems is consistent with the expected vadose
zone airflow rates derived from Table 4-8. Venting wells operate at 25 to 50 scfm and
are completed in relatively permeable, medium- to coarse-grained alluvial materials If
liquid hydrocarbon is present, the early operation of an SVE system is characterized by
a "flushing" period, when mass removal rates remain relatively high. As removal rates
decline, the system is limited by advection and evaporation (Hinchee, 1990). Mass
removal rate plotted versus time generally fits an exponential decline, approaching
187
-------
asymptotic values near zero. If an early flushing period is observed, the exponential
regression should be performed on data following that period.
4.6.2.2 Category 2 Sites
SVE performance data for Category 2 sites decline exponentially but clearly demon-
strate a non-zero asymptotic mass removal rate as shown in Figure 4-21. Table 4-9
presents a summary of the Category 2 sites. The exponential rate constant, k, is not
included in this table because the constants calculated for these sites showed con-
siderable variability, making it difficult to demonstrate trends. None of these sites has
been closed. The period of operation for these systems varies between 205 and 784
days. Although these sites contain multi-well extraction systems, regression analysis
for Sites I and O included performance data from only one well (both systems include
more than eight wells). With the exception of Site I (70 to 100 scfm), which
incorporates groundwater pumping, and Site O (26 to 70 scfm), individual venting wells
at Category 2 sites operate below 20 scfm. At Sites J and K, where SVE incorporates
13 to 14 venting wells, total system airflow rates are between 165 and 225 scfm.
Category 2 SVE wells operate at lower flow rates than those of Category 1 systems,
which is consistent with the descriptions in Table 4-9. Category 2 sites are
characterized by typically fine-grained sediments, low permeability, and per-well airflow
rates less than 20 scfm.
As in Category 1, SVE systems at Category 2 sites may also demonstrate an early
flushing period. Early in the life of an SVE system, preferential flow paths develop
within zones of higher permeability where most of the airflow occurs. Hydrocarbon
vapors within the preferential flow paths are removed by advection and evaporation
(similar to the Category 1 sites). Mass removal rates generally follow an exponential
decline during this period. Ultimately, mass removal rates clearly approach a non-zero
asymptote. In this period of asymptotic mass removal rates, hydrocarbon vapors in the
pore spaces of low- permeability sediments (silts and clays) must reach the preferential
flow paths in order to be removed from the subsurface, and are therefore diffusion-
limited.
4.7 Summary
Monitoring the performance of remediation systems and evaluating performance data
are key elements in site cleanup. Frequency and quality of effluent monitoring data is
critical to system evaluation. Plots of hydrocarbon mass removal rate versus time
should be used in performance evaluation. Exponentially declining mass removal rates
do not in themselves guarantee site cleanup. If the SVE system is installed and
operated in such a way as to adequately access residual hydrocarbon, however,
performance plots should be an accurate measure of remediation efficiency.
Advection-controlled and diffusion-limited sites have been defined as two basic SVE
performance categories based on geologic controls of hydrocarbon mass removal
188
-------
o
'55
w
S
O)
0)
O
Q.
oo
o
(O
oo
to
o
CM
o
o>
5
re
o
(0
£
0)
I
CO
o
•s
re
o
S
TT
O Q)
|3
5fr
g?
c o
££:
N
to c
I
'ejey
CNI
4-
I
O)
iZ
LU
£
re
g
CO
189
-------
Table 4-9. Summary of SVE System Data for Category 2 Sites
Site
Sitel
SiteJ
SlteK
SiteL
SiteM
Site N
SiteO
Status
Open
Open
Open
Open
Open
Open
Open
Duration,
(Days)
613
237
408
205
700
784
409
Airflow
Rate (scfm)
70-100
225 (75)a
165 (74)a
6-84
76-120
30
26-70
No. of
Venting
Wells
1
13(3)a
14 (7)a
4
6
7
1
Early operation of system.
rates. Advection-limited geologic settings are characterized by relatively permeable
soils through which airflow contacts the majority of vadose zone contamination. The
rate of hydrocarbon mass removal for advection-limited sites is primarily a function of
hydrocarbon volatilization and airflow rates. As a result, hydrocarbon mass removal
rates decline toward a near-zero asymptote. Once near-zero asymptotic hydrocarbon
mass removal rates have been achieved for advection-controlled sites, they should be
ready for confirmatory soil sampling prior to closure.
Diffusion-limited geologic settings are typically sites with heterogeneous soils (sands,
silts, clays) through which air flows along preferential pathways in the higher
permeability sediments. Soils in the airflow pathways are remediated early in the life of
the SVE system, but hydrocarbon mass transfer from the lower permeability sediments
is controlled by the rate of diffusion of hydrocarbon vapors into the airflow pathways.
Hydro-carbon mass removal rates for these sites decline exponentially to a non-zero,
diffu-sion-limited asymptotic value. Diffusion-limited sites may require significantly
longer SVE operation times to adequately reduce hydrocarbon concentrations in lower
permeability soils. SVE performance for advection- or diffusion-limited sites can be
anticipated based on stratigraphy determined during site investigation.
190
-------
Chapters
Bioventing and Intrinsic Bioremediation
5.1 Introduction
Minimum action technologies such as bioventing and intrinsic remediation are gaining
popularity as more complex remedial technologies are proving to be very costly while
providing marginal results in many cases. Bioventing and intrinsic remediation can be
very cost-effective when applied at appropriate sites. Additionally, these two technolo-
gies are often complementary when both vadose zone soil and groundwater are con-
taminated. This chapter describes the application of bioventing systems for the treat-
ment of vadose zone soil, the approaches that can be taken to demonstrate intrinsic
remediation in the vadose zone and groundwater, and the benefits of using the two
technologies as a total treatment package for soil and groundwater treatment.
5.2 Bioventing Process Overview
In situ treatment of vadose zone soil is very advantageous under a variety of circum-
stances. Treatment of vadose zone soil without excavation is attractive when space is
limited, when soil volume is large, when disposal options are limited, where excavation
invokes the need for special permits, or where excavation is impossible such as under
buildings and other structures or deep vadose zones.
Soil vapor extraction (SVE) is a widely applied technology for the removal of volatile
organic compounds (VOC) from vadose zone soil. There is some debate as to the
actual effectiveness of SVE for remediating soil contaminated with volatiles, however
and there is general agreement that SVE is ineffective for the remediation of semi- and
nonvolatile compounds (Figure 5-1) (Kent and Graves, 1992). A new technology
commonly known as "bioventing" has emerged in the past few years that addresses
some limitations of SVE (Hinchee, 1991; Miller, etal., 1990; and Dupont, 1993).
Bioventing is the in situ biological treatment of contaminated vadose zone soil Durinq
bioventing, blowers or vacuum pumps are used to supply oxygen to the subsurface and
thus promote remediation (Figure 5-2). Bioventing can effectively treat any compound
that is biodegradable, given that bacteria are present in the soil with the metabolic
capability to biodegrade the target compounds. In the case of petroleum contamination
such as that caused by gasoline, jet fuel, diesel, and heating oils, appropriate bacteria
are nearly always present. These bacteria may not be very active, however because
191
-------
Gasoline
Gasoline &
Diesel
Jet Fuel
Fuel Oil
Bioremediation
Volatilization
Figure 5-1. Comparisons of volatilization and bioremediation.
of a lack of oxygen, extreme pH, lack of nutrients, lack of moisture, or cold temperature
(Sims, etal. 1993).
Bioventing systems use the same equipment as SVE; however, because of the differ-
ence in the treatment objectives of biodegradation versus stripping and extraction,
some design and operational details are different (Dupont, 1993). As a "rule of thumb"
bio-venting systems are designed at one-fifth to one-tenth the capacity of an SVE
system designed for the same site (Dupont, 1993). This estimate is based on the
evaluation and/or application of bioventing at numerous U.S. Air Force sites. Because
of the low recovery of soil vapors, the cost for aboveground treatment systems such as
activated carbon units or fume incinerators is reduced. In cases where the principal
contaminants are semi- or nonvolatile compounds, vapor treatment may not be
necessary.
Another operational avenue to reduce capital and operating cost and complexity is to
inject air into the soil to achieve oxygenation (Figure 5-3). This approach should be
evaluated cautiously because toxic or flammable fumes can accumulate in poorly
ventilated structures such as basements or even trenches or ditches. When vapor
192
-------
To Negative-pressure Blower
Concrete Protector
Hydrocarbon Plume
Figure 5-2. Bioventing by air extraction.
193
-------
From Positive-pressure Blower
Concrete Protector
Hydrocarbon Plume
Figure 5-3. Bioventing by air injection.
194
-------
accumulation is not a concern, the diffuse emission of volatiles from the ground surface
is usually undetectable. .
Bioventing system design has many similarities to SVE design. Common predesign
information that is needed includes the zone of influence about an extraction well,
which is determined by measuring vacuum at monitoring points; soil gas composition
including VOC, oxygen, carbon dioxide, and methane; the vertical and horizontal loca-
tion of contamination; type of contamination; and subsurface stratigraphy (EPA, 1993d).
Additionally, a few other pertinent site characteristics should be investigated to specifi-
cally support the design and operation of a bioventing system.
5.3 Laboratory Testing to Support Bioventing
Bioventing is a form of bioremediation that depends on enhancing biological activity in
the subsurface. Therefore, physical, chemical, and microbiological parameters known
to impact biological activity in the environment should be evaluated.
5.3.1 Microbial Population Density
The presence of a viable microbial population is critical to the success of bioventing. A
very small or nondetectable microbial population may indicate vadose zone conditions
that are averse to bioventing. These conditions may be correctable if they are identi-
fied. Determining the size and viability of the indigenous microbial population is a
simple approach for identifying potential problems.
Some practitioners do not recommend microbial enumerations as part of their bio-
venting monitoring program. They opt to measure the performance of the system solely
by in situ respiration tests. Although this approach is simple and may be adequate in
many cases, the status of the microbial population has to be assumed and changes in
performance cannot be correlated with changes in microbial population density. Be-
cause the remediation process is affected by microorganisms, periodic assessment of
the microbial population density can provide useful information, especially at compli-
cated or difficult sites.
!
The number of bacteria in the impacted soil is determined by using accepted methods
for enumerating bacteria. The spread-plate method and the most-probable-number
method are the most common techniques for enumerating bacteria; however, micro-
scopic and gene probe techniques are also being successfully used.
Total aerobic heterotrophs and aerobic contaminant-degrading bacteria should be
enumerated. The results are reported as the number of bacteria or number of colony
forming units (CPU, equivalent to number of culturable bacteria) per gram (g) of dry soil.
Typical microbial densities in soil range from 10,000 to 10,000,000 CFU/g of soil. Total
microbial density lower than 10,000 CFU/g is not necessarily bad, but it may indicate
that some condition in the soil is slowing the growth of the bacteria. Contaminant
degraders usually constitute 50 percent or less of the total population.
195
-------
5.3.2 Microbial Stimulation Testing
A laboratory test evaluating the response of indigenous bacteria to enhanced environ-
mental conditions may be used to test the viability of site bacteria and their potential
activity during remediation. These tests are designed to stimulate the growth of bac-
teria by providing oxygen and nutrients. Although the test can be conducted in several
ways, an important criterion for the test is to conduct an independent evaluation of the
effect of oxygen on microbial growth as well as an independent evaluation of the effect
of oxygen and nutrients on microbial growth. In many cases where the population of
bacteria is low, the addition of oxygen alone is adequate to stimulate microbial activity;
however, in other cases both oxygen and nutrients may be required to stimulate the
growth of bacteria. When oxygen or nutrients limit bacterial activity, the population
density may increase several-fold more within a short time after the addition of oxygen
or oxygen and nutrients. This result strongly suggests that indigenous bacteria will also
respond when site conditions are enhanced by the treatment system. In cases where
the bacteria fail to respond during this test, further investigation is warranted to deter-
mine why growth did not occur because this may also indicate poor performance of the
bioventing system.
Not all practitioners of bioventing advocate the use of microbial stimulation tests to
evaluate population viability and responsiveness to treatment conditions. Bioventing
systems have been successfully installed without such data; however, the risk of instal-
ling an ineffective system is increased. Performance problems can often be anticipated
based on the results of a microbial stimulation test.
5.3.3 Residual Nutrients
Oxygen is the principal limiting factor for in situ biodegradation followed by nutrient
limitations. Fixed nitrogen is the inorganic nutrient required in the greatest concentra-
tion for bacterial activity. Phosphate is also required, but in smaller amounts. Reported
ratios of carbon to nitrogen to phosphate required to support bacterial growth range
from approximately 100:10:1 to 600:10:1 (Kent and Graves, 1992).
Ammonium, the preferred source of nitrogen for most bacteria, is often found at low to
very low concentrations in soil, thus suggesting that microbial activity can be limited by
lack of fixed nitrogen. Phosphate, often found in relatively high concentrations in soils,
is less likely to limit biological activity in soil. Determining the concentration of these
two inorganic nutrients is important for designing a bioventing system that maximizes
biological activity. Laboratory tests frequently indicate that nutrients are beneficial and
will increase the rate pf biodegradation. Field results are mixed, with some published
reports suggesting that nutrient addition does not increase biodegradation while others
indicate some benefit ( Miller et al., 1990; Mark-Brown, 1994; Lakshmiprasad and
Dupont, 1993; and Graves and Leavitt, 1993). Nutrient addition may prove to be bene-
ficial in cases where inorganic nutrients are not available.
196
-------
Although nutrient addition is not recommended a priori, it may significantly improve the
performance of bioventing systems in nutrient-deficient soils or after several months of
operation when residual nutrients have been depleted. Nutrients may be applied in
solution or as vapors. Delivery and even distribution of nutrient solutions is challenging,
especially in deep vadose zones. Field experience is lacking in vapor-phase nutrient
delivery; however, laboratory results show that it is feasible and beneficial (Mark-Brown,
1994; Lakshmprasad and Dupont, 1993; Graves and Leavitt, 1993).
5.3.4 SoilpH
Bacteria generally thrive between a pH of 5.5 and 8.5 (Sims et al., 1993; Dupont, 1993).
When the pH deviates from this range, biological activity may be reduced. A pH greater
than 9.5 and less than 4.5 virtually stops the metabolic activity of most bacteria. Adjust-
ing the pH of vadose zone soil is such a challenging task that an extreme soil pH may
eliminate bioventing as a viable remedial option.
5.3.5 Soil Moisture
An acceptable range of soil moisture that will support bacterial activity in a typical soil is
7 to 20 percent moisture by weight (approximately 25 to 85 percent of the moisture
holding capacity [a.k.a. field capacity] of the soil)(Dupont, 1993). Some cases have
been reported where a much lower water content was adequate. Higher water content
impedes oxygenation and airflow by filling pore spaces; therefore, excessive soil mois-
ture will reduce the effectiveness of the oxygenation process, resulting in reduced acti-
vity by aerobic bacteria.
High rates of soil gas extraction can cause excessive soil drying and limit the biodegra-
dation rate. Adding water back to the vadose zone can be difficult; therefore, the best
approach is to avoid desiccating the soil. This becomes a serious issue when SVE is
used together with bioventing because the operation of an aggressive SVE isystem
before bioventing can dry the soil and inhibit biodegradation.
i
5.3.6 Conducting Laboratory Tests .
A biotechnology laboratory that routinely conducts the indicated tests should be con-
tracted to perform the laboratory tests for the bioventing project. The laboratory should
be selected based on bench scale and field experience.
5.4 Field Investigations to Support Bioventing
Bioventing and SVE share several common elements. Therefore, several aspects of
field investigations supporting both technologies are similar.
5.4.1 Zone of Influence About a Vacuum Well or Trench
The ability to oxygenate the subsurface is critical to the operation of a successful
bioventing program. Therefore, understanding the area that can be oxygenated by a
well or trench is critical for designing a system to effect treatment of the entire area.
The zone of influence can be determined by connecting a blower to the well or trench
197
-------
and measuring vacuum in the ground at various distances from the well. For biovent-
ing, an alternative measurement is oxygen content of the soil gas at various distances
from the well. The time required to achieve steady-state conditions during zone-of-
influence testing varies with soil type and is, therefore, specific to each site (see
Chapter 4.0). Based on zone-of-influence measurements, wells are placed to avoid
gaps in coverage where oxygen will not be replenished.
Figure 5-4 shows a common layout for measuring the zone of influence about a
vacuum well. When a vacuum is applied to the venting well, negative pressure can be
detected at the monitoring points. When the vacuum observed at each monitoring point
is plotted against the distance from the venting well by using a semi-logarithmic coor-
dinate system, a straight line should result. The point at which the vacuum is reduced
by 60 to 90 percent is usually taken as the functional radius of influence for SVE
applications.
For bioventing, the defining parameter is the oxygen content of the soil gas. Airflow
through the subsurface must only meet the oxygen demand of the contaminated soil.
Depending on the permeability of the soil, radial influence frequently ranges from 10 to
100 feet. Achieving a practical zone of influence in shallow soil can be difficult because
of air short-circuiting to or from the surface. An impermeable cover can be placed over
the treatment area to improve the radius of influence. Plastic sheeting, asphalt, and
concrete can be effective barriers to airflow. In cases where extensive surface area is
paved, fresh air sources may be unavailable or poorly located to provide thorough
aeration. Passive air inlets or injection wells may facilitate aeration of the entire
treatment area, thus increasing system effectiveness.
5.4.2 Extent of Contamination
Defining the horizontal and vertical extent of the plume of contamination is critical for
the placement of wells and screened intervals. Differences in the air permeability of
different soil strata are also an important consideration because air will preferentially
flow around less-permeable zones and lenses. This can result in oxygen deficiency in
less-permeable strata. Maximum aeration of the plume is ensured when wells are
screened within the vertical limits of the contamination.
i
Additionally, placement of well screens within the contaminated area insures accurate
respiration measurements. When wells are screened over a wide vertical area that
includes substantial areas of clean soil, the soil gas withdrawn from the impacted area
is diluted with soil gas.from uncontaminated areas. This dilution causes the soil gas to
have a much higher oxygen content and a much lower carbon dioxide and VOC con-
tent. This observation can lead to misinterpretation of the actual performance of the
bioventing system.
198
-------
Contaminant Plume
Zone of Influence
(Aerated Area)
[Usefor Background
I Respiration Test
Extraction Well
Use for
Respiration Test
Piezometers
Figure 5-4. Well location for radius of influence and in situ respiration testing.
199
-------
5.4.3 Soil Gas Composition
Soil gas composition can be used as the principal indicator of the performance of a
bioventing system. Under conditions suitable for bioventing, the oxygen and carbon
dioxide content of the soil gas in the impacted soil will be very different from atmos-
pheric conditions. Air contains about 20.9 percent oxygen and 0.032 to 0.036 percent
carbon dioxide; however, soil gas collected from contaminated soil has been frequently
observed to contain less than 10 percent oxygen and more than 5 percent carbon di-
oxide (Kent and Graves, 1992). These conditions strongly suggest that aerobic bio-
logical activity has affected the soil gas composition because aerobic bacteria consume
oxygen and produce carbon dioxide during respiration. The oxygen partial pressure of
soil gas can be reduced by the presence of VOC; however, the increase in carbon di-
oxide cannot be correlated to partial pressure changes resulting from the VOC content
of the soil gas.
Methane is also commonly detected in soil gas from sites impacted with petroleum
products. The presence of methane indicates that anaerobic conditions exist in the
subsurface, resulting in methanogenesis. This interpretation must be accepted only
after regional geology has been considered because some areas have natural gas
seeps.
The goal of a bioventing system is to move air through contaminated soil to increase
oxygen content, reduce carbon dioxide content, and encourage aerobic biodegradation
of contaminants (Hinchee et al., 1992; Dupont, 1993). Optimum oxygen concentrations
for bioventing have not been determined; however, an oxygen concentration greater
than 5 percent should be maintained to sustain aerobic metabolism. Table 5-1 indi-
cates the relationship between gaseous oxygen and dissolved oxygen. The carbon
dioxide content should be maintained at less than 5 percent to avoid toxicity.
Deciding if the contaminated soil is being adequately aerated requires correct place-
ment of the screened interval in the monitoring wells (vapor monitoring points)
(Hinchee, et al., 1992). The vapor monitoring point should be screened within the
impacted zone only, otherwise soil gas from clean soil will dilute the soil gas'collected
from the impacted area. The soil gas composition results collected from an improperly
screened vapor monitoring point will show a higher oxygen and lower carbon dioxide
content. Erroneous results give the impression that the bioventing system is effectively
aerating the soil when the impacted area may be anaerobic.
5.4.4 Respiration Measurement
Measuring the respiration rate in the impacted soil provides the best indicator of the
potential performance of a bioventing system (Hinchee and Ong, 1992; Hinchee and
Arthur, 1991). In general, a respiration test consists of aerating the soil and measuring
the consumption of oxygen and the production of carbon dioxide over time. The U.S.
Air Force Center for Environmental Excellence (AFCEE) has produced a widely used
guidance document describing in situ respiration testing (Hinchee et al., 1992). Based
200
-------
on the test results, the rate of respiration can be determined and the amount of hydro-
carbon consumed per unit of time can be estimated.
Table 5-1. Gaseous Oxygen, Dissolved Oxygen, and Aerobic Biodegradation.
Gaseous
Oxygen
(Percent)
21a
19a
17a
15a
13a
11a
9a
7a
5a
3
1
0
Gaseous
. Oxygen
(mg/L)
300
271
243
214
186
157
129
100
91
43
14
0
Gaseous
Oxygen
(Torr)
159.6
144.4
129.2
114
98.9
83.6
68.4
53.2
38
22.8
76
9
Dissolved
Oxygen
(mg/L)
8.6
7.8
7.0
6.1
5.3
4.5
3.7
2.9
2.0
1.2
0.4
0
8 Oxygen concentrations that support aerobic biodegradation.
Performing the respiration test in situ gives the most representative results. To conduct
an in situ respiration test, a vapor extraction well or piezometer must be installed with
the screened interval placed within the impacted area. Good results can be obtained
by using a very small vapor monitoring point (0.5-inch diameter). The well or vapor
monitoring point is connected to a vacuum pump.or blower, and the soil gas is evacu-
ated (or fresh air injected) until the gas composition approaches that of air. This pro-
cedure can be done along with radius-of-influence measurements.
After the impacted soil has been aerated, the vapor monitoring point or well is sealed.
Periodic gas samples are collected from the well or from nearby vapor monitoring points
and they are then analyzed for oxygen and carbon dioxide. The first sample should be
collected immediately after shutdown of the pump system. The next sample should be
collected one hour after system shutdown. The next samples should be collected after-
two or three hours. Subsequent sampling times can be projected based on the
changes observed in the first three samples. In situ respiration tests typically last for 24
to 96 hours depending on the level of biological activity in the contaminated soil
(Hinchee et'al., 1992; Hinchee and Ong, 1992). The test should be stopped after the
oxygen concentration drops to 5 to 10 percent. When oxygen consumption is plotted
201
-------
against time, either a linear or semi-logarithmic curve should fit the points. The slope of
the curve represents the in situ respiration rate.
An important control for the in situ respiration test is the measurement of background
respiration. To determine this parameter, a vapor monitoring point is placed in an un-
contaminated area, the soil is aerated, and a respiration test is conducted. The oxygen
consumption and carbon dioxide production rates from this area are subtracted from
the results obtained from the contaminated area. The difference is the respiration re-
sulting from contaminant biodegradation.
A useful modification to the test is the injection of helium (2 percent final concentration)
as an inert tracer (Hinchee et al., 1992; Hinchee and Ong, 1992). The presence of
helium in the air samples collected during the test ensures that the soil gas being
collected is the same gas that was injected. This modification greatly increases the
validity of an in situ respiration test and is very useful if a rigorous defense of test
results is required.
The recommended equipment for the in situ respiration test is a blower (1- to 5-horse-
power) and small-diameter vapor monitoring points because they require less volume
for effective purging (Hinchee et al., 1992). A small pump can be used to withdraw
samples from the vapor monitoring point. Soil gas oxygen and carbon dioxide can be
conveniently determined using several different field methods. Comparisons of field
and laboratory methods for gas analysis show that field measurements are adequate
for the in situ respiration test. Laboratory confirmation is recommended only if quality
control issues warrant.
Several options are available to measure oxygen and carbon dioxide in the field.
Indicator tubes (Drager and Sensidyne) are available for both carbon dioxide and
oxygen. Oxygen indicator tubes span a range of 5 to 23 percent. Several ranges are
available for carbon dioxide indicator tubes. Low-and high-range carbon dioxide tubes,
which usually provide results as parts per million volumetric or volumetric percent,
should be available during the test. The Gastechtor Model 32520X (Gastech, Inc.) is
an electronic instrument that provides both a direct reading oxygen and carbon dioxide
detector in the same instrument. It also has an internal pump that simplifies sample
collection. A confined-space-entry oxygen detector with a range of 0 to 23 percent can
also be used for field oxygen measurements. A helium leak detector can be used when
helium is used as an inert tracer. The precision of all these devices is adequate if a
good sample is collected. A good sample is one that is truly representative of the soil
gas in the impacted zone soil. Purging the vapor monitoring point is critical. Placement
of the well screen within the contaminated area is also essential.
In cases where an in situ respiration test cannot be performed, soil samples can be
collected and shipped for laboratory analysis. Because the test is conducted on dis-
turbed soil, the results are less representative of actual site conditions. The laboratory
202
-------
respiration test, however, has the benefit of allowing a cost-effective evaluation of
nutrient delivery.
5.5 System Design and Installation
The design of the bioventing system is determined by the soil type, depth and extent of
contamination, radius of influence, depth to groundwater, air permeability, and soil
stratigraphy (Hinchee et al, 1992). Although the best design for a site will ultimately be
decided by site conditions, two general design options have merit and should be con-
sidered during the design of a bioventing system.
5.5.1 Bioventing Wells and Well Spacing
Wells suitable for bioventing are typically very simple, consisting of a well screen sec-
tion connected to a solid pipe that continues to the surface. At the surface, the solid
pipe is manifolded to other wells or the air-moving equipment. The selection of well
screen size is usually not critical. A 10- or 20-slot PVC well screen (slot widths of 0.010
or 0.020 inch) is a common choice. The open space in the well screen ranges between
30 and 40 percent. A key consideration in selecting a well screen is to choose a
screen size that does not restrict airflow. Although a wire-wrapped well screen can be
used, the extra cost is usually not justified by a significant performance increase. The
length of the screened section is based on the vertical extent of contamination. In
cases where significant changes in air permeability occur over the vertical extent of
contamination, multiple screened areas can be installed and independently controlled to
encourage air movement through soil strata of differing permeability.
The borehole for the venting well should be drilled by using a technique that does not
alter the cut face of the soil. Drilling a 6- to 10-inch borehole with a hollow-stem auger
is a favored method. For very deep systems air rotary drilling can be used. Mud-rotary
drilling should be avoided because drilling mud alters the permeability of the cut face.
Unlike groundwater wells, a bioventing well cannot be developed to improve the per-
meability of the cut face.
5.5.2 Well Finishing
Well diameter is determined by the volume of air to be delivered or removed and the
pressure drop in the pipe. Two- to four-inch Schedule 40 PVC piping is most commonly
used. The well screen and solid riser are placed in the boring, and the screened area iis
sand packed. A 20- to 30-mesh sand is a common choice for the sand pack. The pur-
pose of the sand is to support the walls of the boring, provide a permeable connection
between the venting well and the soil, and protect the well screen from clogging by soil
fines. The infiltration of fines and clogging of the screen is a minor concern in most
bioventing applications. If the water table rises above the bottom of the well screen,
however, the bottom slots can become clogged and ineffective when the water table
falls.
203
-------
The well is sand packed over the entire screened interval and for another few feet
above the screen. Bentonite or grout is used to seal the boring above the sand pack.
The boring can be completely filled with bentonite/grout or back filled with soil to within
a few feet of the surface. Normally, the last 4 to 6 feet of the well should be grouted to
the surface. The solid piping used to connect the well to the blower can be installed on
the surface or in a shallow trench.
5.5.3 Bioventing Trenches
Shallow vadose zone contamination may be best addressed by using venting trenches.
Backhoes or trenchers are adequate for digging bioventing trenches. The depth of the
trench is determined by the depth of contamination. After digging is completed,
trenches are partially filled with gravel. Slotted pipe or corrugated, perforated, flexible
drainage tubing is placed in the trench and covered with gravel. The piping is posi-
tioned in about the center of the gravel layer. The top of the gravel layer is then
covered with plastic sheeting to prevent air from short-circuiting to the atmosphere.
Finally, the trench is back filled with soil to the surface. The top of the gravel layer
should be at least one foot below the surface. Depending on local weather and the
depth to ground-water, a sump to remove water from the trench may be a useful
addition to the design. The slotted/perforated piping is connected to the blower with
solid PVC pipe at one or several locations in the trench, if the trench network contains
multiple branches, valves to control air flow to each branch can prove useful.
5.5.4 Well Spacing and Airflow Modeling
Wells and trenches are spaced to provide adequate soil aeration. Wells also should be
positioned to provide full coverage of the treatment area.
Several computer models are available that model airflow under user-defined soil per-
meability and injection or extraction pressures. A very useful application of these
models is the estimation of airflow patterns by establishing "no flow" areas within the
treatment zone. For example, the airflow pattern around a foundation, underneath a
paved area, or around an abandoned underground storage tank can be modeled.
Additional wells or trenches can be added to the model to achieve the desired cover-
age. Some models also predict the total airflow into or out of a well or trench system.
This information is useful for sizing a blower or vacuum system.
5.5.5 Air Movement Equipment
Blower or vacuum system size depends on the amount of air that must be moved to
replace a pore volume, of soil gas in the contaminated area. A pore volume replace-
ment rate of once every 1 to 5 days is adequate for most bioventing systems. An
adequate pore volume replacement rate is indicated by the in situ respiration test. If the
test indicates that most of the soil gas oxygen is consumed within 24 hours, a 1-day
pore volume replacement rate should be achieved. The manufacturer's pressure and
airflow tables for various blowers, compressors, and vacuum pumps are compared with
204
-------
the desired flow rate to identify equipment that meets the air movement requirements of
the system.
For example, on a site with 250 cubic yards of contaminated soil and a porosity of 30
percent (0.3), the soil pore space within the treatment area is 2,025 cubic feet. The
pore volume divided by the replacement time will yield an approximation of the air
movement rate required of the blower. Assuming a 1-day pore volume replacement
rate, 2,025 cubic feet divided by 1,440 minutes per day indicates that the blower must
remove or inject only 1.4 cubic feet of air per minute to effect a pore volume exchange
everyday.
5.6 Bioventing Operation and Monitoring
Bioventing is usually a low-maintenance technology. The critical operating parameters
include confirming that the aeration equipment (blower) is working, that air emissions
meet discharge standards, and that the impacted soil is aerated. Monitoring can be
extensive or very limited. A list of common monitoring parameters is shown in
Table 5-2 (Sims et al., 1993; Mark-Brown, 1994; Hinchee et al., 1992). Some combina-
tion of Ithese parameters is usually adopted to monitor the performance of a bioventing
system.
The minimum monitoring protocol involves quarterly to semiannual measurements of in
situ soil respiration. If the respiration test results indicate respiration above background
bioventing should continue. Typical rates range from 1 to 20 mg of hydrocarbon con-
sumed per kg of soil per day (Hinchee and Ong, 1992). When a significant decline in
respiration is observed, soil samples can be collected and analyzed to quantify the
performance of the system. Samples should also be analyzed for soil moisture, pH,
microbial population density, and residual nutrients.
Depending on the contaminant, 3 to 3.5 parts of oxygen are required to bio-oxidize 1
part of hydrocarbon. Using this stoichiometry, an estimate of contaminant biodegrada-
tion can be made. The biodegradation rate for hydrocarbons (based on hexane with
units of mg of hydrocarbon per kg of soil per day) is estimated by using the followinq
equation (Hinchee etal., 1992):
KB = -KoAD0C/100
where:
KB = hydrocarbon biodegradation rate (mg/kg per day)
KO = oxygen utilization rate (percent/day)
A = volume of air/kg of soil (L/kg)
DO = density of oxygen gas (mg/L)
C = mass ratio of hydrocarbon to oxygen required for mineralization.
205
-------
Table 5-2. Suggested Groundwater Monitoring Parameters for Bioventing
Parameter
Method
Contaminant
Oxygen (soil gas)
Carbon Dioxide (soil gas)
Methane (soil gas)
Soil Moisture
pH
Ammonia
Phosphate
Total Kjeldahl Nitrogen
Microbial Density
In Situ Respiration Rate
Varies with contaminant
Field analysis with a direct-reading instrument or
colorimetric indicator
Field analysis with a direct-reading instrument or
colorimetric indicator
Field analysis with a direct-reading instrument or
SW-846 Method 8020
Gravimetry
Field analysis with pH electrode or SW-846
Method 150.1
SW-846 Method 350.1 or 350.2; Standard Method
4500-NH3
SW-846 Method 365.1, 365.2, or 365.3
SW-846 Method 351.1
Spread Plate or Most Probable Numbers (MPM)
Methods (usually modified for environmental
samples)
Field Testing and Analysis in Changes of 02 and
C02 Concentrations Over Time
"Ko" is determined directly from the in situ respiration test. "A" is calculated assuming a
soil porosity of 0.3 and a soil bulk density of 1 ,440 kg/m3, yielding 0.3 L of air/kg of
soil/1 .440 kg/L or 0.21 liter of air per L of soil. "D0" is typically assumed to be 1 ,330
mg/L, although this value varies with temperature, barometric pressure, and altitude.
"C" assumes the following chemical stoichiometry for the biological oxidation of hydro-
carbon. Hexane is routinely used as a typical hydrocarbon in these calculations.
C6H14
9.5 02 - 6C02
7 H2O
206
-------
Given that oxygen (O2) has a molecular weight of 32 and hexane has a molecular
weight of 86, the mass of oxygen required to oxidize one mole of hexane is given by:
C = (Molecular Weight of Hexane)/[(Molecular Weight of O2) 9.5]
C = 86/304
C = 0.28
The mass of hydrocarbon biodegraded within the treatment area is calculated by the
following equation:
B = KBM/106mg/kg
Where:
B = kilograms of hydrocarbon degraded per day
KB = average hydrocarbon consumption (biodegradation) rate in mg/kg per day
M = kilograms of soil within the treatment area.
Most practitioners recognize that bioventing is a slow process. The U.S. Air Force pre-
dicts that up to 10 years will be required to remediate most of its petroleum-impacted
sites using bioventing. The most hazardous constituents (benzene, toluene, ethyl
benzene, and xylenes [a.k.a. BTEX]) are also the most biodegradable, however, and
they should remediate much faster.
5.7 Bioventing Conclusions
Bioventing is an emerging technology that combines features of SVE and in situ bio-
remediation. The technology permits the in situ treatment of vadose zone soil impacted
with any biodegradable contaminant. Bioventing has the potential to meet several
important needs in waste management and remediation. Bioventing should be con-
sidered the primary or supplemental treatment technology for contaminated vadose
zone soil that cannot be excavated.
5.8 intrinsic Bioremediation: Process Overview
Intrinsic bioremediation is the preferred term to describe the natural biological pro-
cesses that lead to contaminant biodegradation. Intrinsic bioremediation can occur in
any environment that supports microbiological activity. The rate of biodegradation may
be slow, however, because of the lack of a suitable respiratory substrate (such as
oxygen) or inorganic nutrients (such as fixed nitrogen), an extreme pH, low soil mois-
ture, or limited contaminant bioavailability. Elimination of the contaminant source is
essential for the successful application of intrinsic remediation. Accurate delineation of
contamination, an understanding of subsurface conditions and characteristics, as well
as an understanding of contaminant migration rates and direction are critical for
evaluating the success of intrinsic remediation and for establishing regulatory support
207
-------
for its use at a site (Davis, Klier, and Carpenter, 1994; Davis et al., 1994; Wisconsin
Department of Natural Resources, 1993).
Several criteria should be considered before choosing intrinsic bjoremediation as the
principal remedial technology (Wisconsin Department of Natural Resources, 1993).
These include:
• Risk of further environmental damage
• Risk of human endangerment
• Detrimental consequences to local flora and fauna
• Technical feasibility, practicality, and effectiveness of other technologies
Site-specific evidence for successful application of intrinsic bioremediation
• Cost of intrinsic bioremediation compared to other options.
When these issues can be addressed in favor of intrinsic bioremediation, the technol-
ogy is a cost-effective and practical remedial alternative for soil and groundwater.
5.8.1 Intrinsic Bioremediation in the Vadose Zone
Intrinsic bioremediation has been shown to be effective in the vadose zone. The
frequent observation of low oxygen and high carbon dioxide and methane in the soil
gas within a hydro-carbon plume is indicative of intrinsic microbial activity.
Without an engineered process to supply oxygen, hydrocarbon plumes become ana-
erobic and the biodegradation rate drops dramatically. Under natural conditions, re-
aeration of the soil is dependent on diffusion, barometric pressure changes, and oxy-
genated rainwater infiltration. In areas where the water table is influenced by tides, the
daily rise and fall of the water table may act as a pump to displace oxygen-depleted soil
gas.
Lacking oxygen, biodegradation can be supported by nitrate and sulfate; however, this
is probably not common in the vadose zone because nitrate and sulfate are very solu-
ble and leach into the groundwater. Methanogenic biodegradation is a common anaer-
obic process in contaminated vadose soil.
The principal difficulty in applying intrinsic bioremediation in the vadose zone is proving
that remediation is occurring. Two factors work together to make definitive proof of
remediation challenging. First, intrinsic bioremediation is a slow process. Second,
variability in soil samples makes analytical proof of reduction in contaminant concentra-
tion difficult until large concentration changes have occurred. This may take several
years and be further complicated by plume dispersion, diffusion, and migration.
Parameters that indicate the occurrence and effectiveness of intrinsic bioremediation in
the vadose zone are ill-defined. The following suggestions, however, highlight some of
the more useful approaches for quantitating intrinsic vadose zone bioremediation.
208
-------
5.8.2 Monitoring Parameters for Intrinsic Vadose Zone Bioremediation
The chemical constituency of the original petroleum product and the chemical profile of
contaminants in the soil evaluated over time can indicate the occurrence and magni-
tude of intrinsic bioremediation. In the case of a gasoline spill, the original product will
contain an identifiable mix of chemicals. Over time, the chemical profile will change.
The earliest and most obvious changes will occur in the volatile organic compounds,
especially benzene, toluene, ethyl benzene, and xylenes (BTEX). In addition to
changes in concentration, the relative ratio of the BTEX constituents will also change.
This preferential loss of chemicals may indicate biodegradation. For example, toluene
and xylene tend to biodegrade faster under anaerobic conditions than do benzene and
ethyl benzene. The hydrocarbon "fingerprint" will also be shifted toward a higher ratio
of heavy hydrocarbons because high-molecular-weight hydrocarbons tend to biode-
grade at a slower rate than the low-molecular-weight compounds. The fingerprint of
most hydrocarbon fuels will change with time. Physical and biological processes con-
tribute to the "weathering" process (Wilson, 1993). i.
The soil gas composition can also indicate intrinsic biodegradation. The occurrence of
methane in subsurface soil is uncommon in most areas; however, methane is often
associated with soil gas within petroleum-contaminated vadose zone soil. Similarly, the
oxygen concentration is frequently very low and the carbon dioxide concentration is
high. Biological activity is responsible for these changes. Additionally, the total petro-
leum hydrocarbon (TPH) and VOC content of the soil gas should eventually decrease
as intrinsic biodegradation proceeds. The long-term rate of petroleum biodegradation
cannot be determined from simple measurements of soil gas composition. Biodegrada-
tion rates can be estimated, however, by determining the rate at which changes in soil
gas composition occur (Hinchee et al, 1992; Hinchee and Ong, 1992).
Restricted plume migration and, especially, demonstration of reduced plume dimen-
sions are also indicative of intrinsic biodegradation. Such observations support intrinsic
biodegradation as a viable approach for limiting plume migration and treating the con-
taminant. The process will be very time consuming, and periodic soil and soil gas
samples should be collected to determine current conditions and contaminant concen-
tration in soil and soil gas. As long as intrinsic biodegradation is preventing further
damage to the environment and monitoring supports contaminant reduction;, intrinsic
bioremediation can be a cost-effective and practical way to treat a site. The principal
cost element during intrinsic bioremediation is monitoring to ensure that biodegradation
is occurring and that further environmental degradation is not occurring.
5.8.3 Intrinsic Bioremediation in the Saturated Zone
In contrast to the vadose zone, a growing body of published literature and field experi-
ence is available to guide the application of intrinsic bioremediation in the saturated
zone (Wisconsin Department of Natural Resources, 1993). Several approaches are
available to demonstrate the effectiveness of intrinsic aquifer bioremediation. Because
these approaches provide indirect or circumstantial evidence, a set of results
209 '
-------
highlighting multiple indicators of intrinsic aquifer bioremediation is usually required to
construct a convincing demonstration.
Site hydrogeology is extremely important in evaluating intrinsic bioremediation in
groundwater. Groundwater flow rate, recharge rate, vertical and horizontal location of
the contaminant within the aquifer, and the geological characteristics of the aquifer
influence the interpretation of data collected to evaluate intrinsic bioremediation.
5.8.3.1 Contaminant Monitoring
A decrease in contaminant concentration is an important indicator of intrinsic bio-
remediation; however, physical losses must be accounted for before biodegradation
can be accepted. Contaminant diffusion, dispersion, and adsorption to soil can be
estimated by identifying a conserved constituent in petroleum fuel spills. Trimethyl
benzenes and 2,3-dimethyl pentane have been used as conservative indicator com-
pounds for jet fuel and gasoline plumes.
Key characteristics of a useful conserved indicator compounds are poor anaerobic
biodegradability and chemical properties similar to target contaminants. Because dis-
solved hydrocarbon plumes undergoing intrinsic bioremediation are anaerobic, com-
pounds that are not susceptible to anaerobic biodegradation persist and are influenced
primarily by physical processes. The change in concentration observed as the dis-
solved contaminant plume moves through the aquifer is used to estimate physical
losses likely to also occur with compounds more susceptible to anaerobic biodegrada-
tion. The difference between the expected physical reduction in concentration and that
observed is attributed to biodegradation.
Another useful comparative analysis involves examining the ratio of one target contami-
nant to another. This approach is commonly applied when BTEX are the principal con-
taminants. Toluene and xylenes biodegrade under anaerobic conditions at a greater
rate than do benzene and ethyl benzene. Toluene-to-benzene ratios at points down-
gradient from the source of contamination typically show a change from a large toluene-
to-benzene ratio (more toluene than benzene) to a smaller ratio (more benzene than
toluene). Fresh gasoline may have up to 10 times as much toluene as benzene, and
groundwater exposed to gasoline may have several times more dissolved toluene as
benzene. Down-gradient portions of a dissolved-phase gasoline plume usually have
more benzene than toluene. This change in the toluene-to-benzene ratio is due to the
more rapid biodegradation of toluene under anaerobic conditions. Benzene biodegra-
dation can also be established by comparing benzene concentrations with compounds
whose anaerobic biodegradation rate is less than benzene, such as 2,3-dimethyl pen-
tane or trimethyl benzenes.
Intermediate biodegradation products can be used to evaluate intrinsic biodegradation
in some cases. Trichloroethylene and other halogenated solvents are anaerobically
210
-------
biodegraded through dehalogenation reactions that yield identifiable intermediates.
Intermediates observed during anaerobic reduction of trichloroethylene (TCE) include
dichloroethylene, vinyl chloride, and ethene (ethylene).
Unfortunately, the chlorinated intermediates associated with the intrinsic biodegradation
of polychlorinated aliphatic hydrocarbons are typically less susceptible to further anae-
robic degradation than is the parent compound. Demonstrating that a poly-chlorinated
aliphatic compound such as TCE is being converted to dichloroethylene (DCE) does not
imply that complete bioremediation will follow. Anaerobic removal of chlorine atoms
from a hydrocarbon usually leads to greater aerobic biodegradability; however, aerobic
biodegradation is impeded by the lack of oxygen in proximity to the newly formed inter-
mediate.
5.8.3.2 Microbially Induced Changes in Groundwater Chemistry
Actively respiring microorganisms alter their environment. Changes may go unnoticed
if the local environment is dynamic enough to compensate for microbially induced
changes. In many cases, the rate of microbial-induced change within a dissolved con-
taminant plume exceeds the capacity of the local environment to compensate; the
result is that the dissolved contaminant plume has chemical characteristics different
from surrounding groundwater (Wisconsin Department of Natural Resources, 1993).
5.8.3.3 Respiratory Substrate Concentrations
Microbes can use several inorganic compounds as the terminal electron acceptor for
their respiratory pathways. Frequently encountered terminal electron acceptors or
respiratory substrates include oxygen, nitrate, sulfate, and carbonate. Iron and
manganese can also serve as respiratory substrates; however, their contribution to
intrinsic bioremediation is ill-defined. Respiratory substrates are used preferentially in
the following order:
O2 > NO3->Mn+4 > Fe+3 > SO/2 > CO2
Differences between respiratory substrate concentration in contaminated groundwater
and nearby clean groundwater can be used as an indirect indicator of intrinsic bioreme-
diation. Much lower dissolved oxygen in the contaminated portion of the aquifer sug-
gests that microbial activity has depleted the oxygen in the groundwater. This interpre-
tation is valid if the aquifer does not have a high chemical oxygen demand because of
contamination. Lower nitrate content in the contaminated area indicates anaerobic
nitrate-reducing activity. Microbes are known to biodegrade hydrocarbons under nitrate
reducing conditions. Sulfate-based respiration is often detected by measuring sulfide,
the product of respiration. Similarly, carbonate-based respiration results in the produc-
tion of methane. Correlation of changes in the groundwater content of any of these
parameters and the location of a contaminant plume can be indicative of intrinsic
bioremediation.
211
-------
5.8.3,4 Redox Potential, pH, and Other Water Quality Parameters
Table 5-3 lists groundwater parameters that may identify microbiologically mediated
changes within the contaminated plume (Wisconsin Department of Natural Resources,
1993). These parameters are not always helpful; however, until the intrinsic bio-
remediation process is better understood, collection of a variety of data is recom-
mended. As intrinsic bioremediation proceeds and key monitoring elements are
identified, less-useful parameters can be eliminated from site-specific monitoring plans.
5.8.4 Intrinsic Bioremediation Rate Estimates
The rate of intrinsic bioremediation can be estimated from the down-gradient change in
contaminant concentration when physical loss can be approximated (Wiedemeier et al.,
1994). For example, if wells are placed within the dissolved contaminant plume in a line
that follows the down-gradient flow path and are spaced to give a 1-year travel time
between each well, the intrinsic biodegradation rate can be calculated. Monitoring data
from the wells provides contaminant concentration change over time. By correcting for
physical losses, the biodegradation rate and contaminant half-life can be calculated by
using a first-order decay equation of the form:
C.-Q.P*
Where:
C, = concentration after time (concentration in down-gradient well corrected for
physical losses)
C0 - starting concentration (concentration in up-gradient well)
t = time (time required for contaminant to travel from the "C0" well to the "C," well)
k = biodegradation rate constant.
Contaminant half-life is given by:
tI/2 = (ln2)/k
5.9 Computer Modeling of Intrinsic Bioremediation
The most widely used model for intrinsic bioremediation is BIOPLUME II, developed at
Rice University and available from the U.S. EPA. BIOPLUME II can provide accurate
two-dimensional simulations of biodegradation if extensive field data are available to
calibrate the model. Unfortunately, one of the most significant uses of computer
modeling for intrinsic bioremediation is to project a relatively accurate long-term model
of biodegradation under natural conditions with limited data.
BIOPLUME II models groundwater and contaminant movement by using aquifer
characteristics, and it separately models the distribution and movement of oxygen
within the aquifer. The oxygen and contaminant plume models are computationally
superimposed, and the contaminant and oxygen are instantaneously consumed in
212
-------
Table 5-3. Suggested Groundwater Monitoring Parameters for Intrinsic
Bioremediation
Parameter
Method
Contaminant
Total Organic Carbon
Methane
Dissolved Oxygen
Oxidation/Reduction Potential
pH
Temperature
Conductivity
Ammonia
Phosphate
Total Kjeldahl Nitrogen
Nitrate
Nitrite
Sulfate
Sulfide
Alkalinity
Chloride
Iron (II)
Microbial Density
Varies with contaminant
SW-846 Method 415.1 or 415.2
SW-846 Method 8020 ;
Field analysis with dissolved oxygen meter
Field analysis with Oxidation/Reduction (Redox)
Potential (ORP) electrode
Field analysis with pH electrode
Field analysis with thermometer
Field analysis with conductivity meter; SW-846
Method 120.1
SW-846 Method 350.1 or 350.2; Standard Method
4500-NH3
SW-846 Method 365.1, 365.2, or 365.3
SW-846 Method 351.1
SW-846 Method 352.0 or 352.2
SW-846 Method 354.1
SW-846 Method 375.2 or 375.4
SW-846 Method 376.1
SW-846 Method 310.1
t
SW-846 Method 325.3; Standard Method 4500-Cr
Standard Method 4500-Fe
Spread Plate or Most Probable Numbers (MPN)
Methods (usually modified for environmental
samples)
213
-------
stoichiometric proportions of three parts oxygen per part contaminant. Oxygen can be
added to the aquifer model from the vadose zone by advective transport from oxygen-
ated groundwater and by direct injection via wells. Anaerobic biodegradation is mod-
eled by a single anaerobic decay constant that does not accurately reflect the complex
processes occurring in a multicomponent hydrocarbon plume. Additionally, the dissolu-
tion of contaminants into the groundwater is not easily incorporated into the model.
Computer simulations specifically designed to estimate and model intrinsic bioremeclia-
tion are being developed but are not currently available.
A one-dimensional numerical computer model occasionally used to model intrinsic
bioremediation is BIO1D (GeoTrans, Inc.). Computer models are presently not avail-
able for vadose zone intrinsic bioremediation.
5.10 Combined Treatment Strategies
An attractive treatment strategy for contaminated vadose zone soil and groundwater
involves bioventing and intrinsic bioremediation. Vadose zone contamination is fre-
quently the source of groundwater contamination. Bioventing in source areas can be
an affective approach to eliminate the dissolution, diffusion, and leaching of contami-
nants into the groundwater. Without a constant recharge of contaminants, intrinsic
bioremediation in the groundwater can limit plume migration and ultimately reduce
contaminant levels to acceptable or even nondetectable levels.
A combined treatment strategy has the potential to be expedient and cost-effective. In
some cases, it may be the only feasible alternative. Approaches applying cost-effec-
tive, minimum-action technologies deserve full consideration during development of site
remediation plans. In the end, they may be just as effective as more expensive and
complex technologies.
214
-------
Chapter 6
In Situ Air Sparging
6.1 Introduction
Soil vacuum extraction is a cost-effective method for removing volatile contaminants
from the unsaturated zone. Much of the residual product remaining in the subsurface
following completion of free product removal, however, is located within the liquid
saturated zone. Several options are available to remove residual product in the
saturated zone, including:
Pump-and-Treat, which involves long-term pumping of contaminated
water to gradually remove soluble components as they become dissolved
in groundwater.
Enhanced Bioremediation, where oxygen and nutrients are added to the
subsurface to stimulate microbial activity in conjunction with pump-and-
treat.
• SVE/Pump-and-Treat, which involves pumping water primarily to draw
down the water table and is used concurrently with SVE to remove
exposed residual product.
SVE/ln Situ Air Sparging, which involves injecting air below the water
table in conjunction with the use of SVE to remove volatile contaminants.
The major limitation of the pump-and-treat technology for remediating groundwater
containing residual hydrocarbon is that removal rates are limited by the low solubility of
some hydrocarbon components. This constraint results in lengthy remediation
programs.
Biodegradation can be used to enhance the pump-and-treat system by remediating
residual hydrocarbons, that are not removed during its use. The occurrence of bio-
degradation in natural groundwater varies depending on the dissolved oxygen content
of the aquifer and the presence of other factors affecting microbial population (e.g.,
oxygen level, pH, nutrients, temperature).
215
-------
The design and operation of SVE/pump-and-treat systems is identical to that for normal
SVE systems, except that water pumping is employed to maximize the air-permeated
zone.
In situ air sparging is a recently introduced technology that utilizes in situ volatilization to
remove volatile components from residual NAPL or dissolved-phase contaminants
present below the water table. Because of the large variation in the systems being
applied and general lack of understanding of the effectiveness of these systems, the
discussion will be limited to the principles of operation of these methods and their
potential advantages and limitations.
As with SVE, air sparging has broad appeal because it is simple to implement and
capital costs are moderate. Air sparging technology is still in its infancy, however. In
addition, air sparging has not been employed on a routine basis, nor has it been sub-
jected to adequate research or rigorous field investigations. A limited number of air
sparging operations and pilot tests have been evaluated; some of these were effective,
while several were not (Marley, et ah, 1992, 1994; University of Wisconsin, 1992;
Hazmat World, 1994).
In situ air sparging involves injecting clean air through wells or horizontal pipes that
have been installed in saturated porous media (i.e., below the water table). As shown
in Figure 6-1, the injected air streams upward from the well screen through contami-
nated groundwater and saturated contaminated soil, displacing groundwater as it
moves upward. Volatile organic compounds (VOCs) in close proximity to the air stream
partition into the vapor phase. The organic vapors are transferred to the vadose zone
where they can be recovered by an SVE system. The volatilization and transfer of
organics from below the water table to the vadose zone is intended primarily for organic
compounds with high vapor pressures. The oxygen introduced to the groundwater
system causes a secondary benefit—the enhancement of aerobic microbial degradation
of organic compounds, including compounds other than VOCs.
A typical air sparging system using a vertical well configuration is shown in Figure 6-1.
The system consists of an oil-free compressor manifolded to one or more air sparging
wells and an SVE system. The following should be considered in selecting system
components:
• Oil-free reciprocating compressors may be used, but are often rated for
intermittent service.
Continuous-duty oil-free reciprocating compressors are substantially more
costly.
216
-------
Gay Icnsc
Reprinted by permission of the Ground Water Publishing Company. Copyright 1993.
Figure 6-1. Typical in situ air sparging - soil vapor extraction system (Ground
Water, 1993).
217
-------
Continuous-duty rotary screw compressors may be used; however,
filtration systems must be employed because these compressors are not
oil-free.
Distribution manifolds should be gauged and regulated at each air spar-
ging well to control air pressure and airflow individually.
• Check valves are needed between the regulator and the sparging well to
ensure that water will not be forced into the manifold by back pressure in
the formation when manifold pressure is decreased.
• The sparging well point may consist of a well screen or air diffuser in a
sand filter pack.
• The sparging well casing should be grouted in with bentonite or non-
shrinking cement to prevent possible short circuiting of air along the well
casing.
The aboveground equipment necessary for an air sparging system includes an air
compressor, a main header pipe for the compressed air, pressure regulators, and
valves to control the rate of airflow into each air sparging well. The number, design,
and spatial arrangement of wells for an air sparging system depend on the type and
concentration of contaminants, spatial heterogeneities of the geologic media, air
permeability characteristics of the geologic media, and cleanup requirements desired
for the groundwater and saturated geological media. These topics will be discussed
further in this Section.
Air sparging wells can be placed below the water table by use of vertical, horizontal, or
angle drilling techniques. Horizontal pipes cari also be placed in trenches dug below
the water table, which are backfilled with coarse sand or gravel. As long as the screen
interval is below the water table and the annulus around the casing above the screen is
sealed properly, air can be forced down the casing, through the screen, and into the
adjacent formation. Well placement/configuration depends on site conditions and re-
mediation goals and will be somewhat unique for each application. In general, air is
injected into the saturated media about 5 feet vertically below the zone(s) of greatest
contamination. This placement allows the injected air to spread laterally away from the
well and to contact contaminated media as it rises.
In vertical wells, air displaces water down the inside of the casing and then exits the
well near the top of the well screen; a similar arrangement occurs for a well placed at an
angle. The buoyancy of the air will not allow it to displace water much below the top of
the screen before it flows laterally through the screen and migrates upward. Thus, only
short well screens (e.g., 1 to 3 feet in length) are needed for air sparging purposes. In
a horizontal well or a horizontal pipe placed in a trench, a long horizontal screen is
218 i
-------
placed below the water table and air can be introduced to the formation over a long
lateral distance (Figure 6-2).
In the selection and design of in situ air sparging systems, the following factors are
critical to an effective application of this technology: air injection pressure, airflow
features, soil heterogeneity and mass transfer, and vertical and horizontal placement of
sparging well points.
6.1.1 Air Injection Pressure
In order for air to enter soil from the sparging well point,' a minimum air pressure, Pmin,
must be maintained. This pressure is described by:
pmin = g P. D + pj +p; (6r1)
where:
g - gravitational acceleration
Pu = density of water
D = depth below the static water table where the sparging well is placed
Pf - air entry pressure for the soil
J°dB = air entry pressure of the sparging point (measured at injection well).
Installation of sparging wells more than a few meters below the water table becomes
increasingly prohibitive because of the high pressure required to overcome the hydro-
static head. (Note that one atmosphere of pressure corresponds to 10.2 meters or 34
feet of water head.)
The sparging air entry pressure is a function of the pore size of the diffuser unit and the
sand psick pore size through which air (or gas) is introduced into the formation. Dif-
fuser/sand pack pore size controls the bubble size of air injected into the soil. Theoreti-
cally, smaller air bubbles should provide greater mass transfer efficiency between
contaminated soil and air. If bubbles coalesce (i.e., create an air mass) in the soil or in
the sand pack between the sparging well and the soil, however, mass transfer effi-
ciency will be significantly decreased. For this reason, very fine-grained sand should be
used for the sand pack to maximize bubble dispersion.
219
-------
Injection Point for Rushing Gas
Extraction of Contaminated Gas
[A
Vactose Zone
Water Table
Contaminated
Zone
Water-
Saturatod
Zone
Trench
B
#$&$$
.vV'V-V-V-V'
VV^*V^Vv\\-\\v'
•&&&ffsf'&
. '^.-'.V'V-V- . •"
........ — ... ....... — ......... — «... water laoie
„ — Gravel Fill Uppet
Sand
^ Horizontal Pipes
SiltyClay
Figure 6-2 Air sparging systems using horizontal wells (A) or pipes placed in
trenches (B).
220
-------
If capillary characteristics of the soil can be described by the Brooks-Corey capillary
pressure model, the soil air entry pressure is defined by:
Pd§=g-p«*d- (6_2)
where hd is the air entry head.
Typical values of the air entry head can range from a few centimeters for sandy soils to
over 1 meter for clayey soils. This technology is not suited for use in fine-grained soils
(e.g., silts and clay) because of the high pressures required to initiate airflow and the
low flow rates that can be achieved. If the soil matrix is layered (e.g., sandy soils with
mterlayers of clays), the relevant soil air entry pressure will be that of the finest soil
layer between the sparging well and the unsaturated zone (i.e., the upper limit of the
natural capillary fringe).
It is imperative that the seal between the sparging well casing and the soil above the
diffuser unit have an air entry pressure higher than the applied pressure. Excessive air
pressures may cause the seal to fail, resulting in short-circuiting of air along the well
bore. Eixcess pressure may also cause secondary permeability channels to develop.
6.1.2 Airflow Rate
Once the minimum air pressure, Pmin, is achieved, the minimum sustainable airflow rate
Qm,n (measured at the injection well), will be obtained. Nonturbulent airflow is described
by Darcy s law. For vertical well air sparging systems, the contaminant removal rate is
directly related to airflow rate, air injector pressure, and the mass transfer rate of con-
taminant from soil. The airflow rate will generally increase nonlinearly with increasing
pressure, and the rate of contaminant removal will be equal to the product of airflow
rate and gas phase concentration.
As air pressure is increased, air pressure gradient and air permeability also increase
The change in air pressure gradient can be described by the regulated air pressure
minus diffuser pressure drop divided by distance below water table. The increase in air
permeability is due to an increase in air-water capillary pressure.
As the airflow rate increases, the rate of contaminant removal increases Because the
rate of mass transfer from the soil to the air is generally diffusionally limited however
the concentration of contaminants in the air will commonly decrease as airflow rate '
increases. For diffusion-limited sites, as the rate of removal increases, the removal
efficiency (i.e., mass contaminant removed per volume of gas injected) diminishes with
flow rate.
Pulsed airflow is sometimes used to overcome mass transfer limitations and provide
greater removal efficiency. It should be recognized that the price of higher efficiency
221
-------
will be a lower mass removal rate because the total flow is lower for pulsed systems
than for continuous flow systems.
6.1.3 Heterogeneity And Mass Transfer Considerations
Because air has a much lower viscosity than water and a much higher relative perme-
ability, except at very low gas saturations, air mobility is much greater than water in
soils. Because of this large contrast in mobility, coupled with the high gradient imposed
on the gas phase, the following phenomena occur:
• Unstable flow during water displacement
• Channeling of the unstable fluid through the soil in fingers
• No uniform movement of air as a front of constant saturation.
Unstable flow results when the displacing fluid (air) encounters differences in soil pore
size and moves rapidly through pores that offer a path of least resistance. Even small
heterogeneities that occur in the most uniform soil will induce such instabilities (see
Figure 6-3). Larger scale heterogeneity, such as lenses of different texture, can
exacerbate the instability and lead to yet more random flow patterns (see Figure 6-4).
The nonuniformity of gas flow patterns has important implications to the efficiency of
contaminant removal during air sparging. Because air velocity in flow channels
(fingers) is quite high and the spacing between fingers may be great, equilibrium will not
be achieved between the contaminant in the flowing gas and that in the stagnant
regions between fingers. Following an initial flush of high concentration as contami-
nants are removed within the flow channels, mass transfer to the flowing gas will be
limited by the rate of aqueous diffusion and convection in the water, to the extent the
latter occurs. The phenomenon of diffusion-limited mass transfer has certain implica-
tions compared with SVE. The processes are identical for air sparging, but the main
factor controlling the mass transfer rate now is the average size and spacing between
fingers (Figure 6-5).
Remediation time is reduced with an increase in the following:
Airflow rate
• Air channel diameter and frequency
• Diffusion gradient.
However, the gain in mass transfer rate will generally be less than proportional to the
increase in flow rate, so that the net efficiency (in terms of mass removed per volume of
gas) will be lower.
222
-------
Overburden with
4-mm beads
Mixture of two
sizes of beads
Reprinted Iby permission of the Ground Water Publishing Company. Copyright 1993.
Figure 6-3. Observed air channel pattern in uniform mixture of 0.75- and 0.3-mm
glass beads (Ground Water, 1993).
Reprinted by permission of the Ground Water Publishing Company. Copyright 1993.
Figure 6-4. Observed air channel pattern in stratified medium (Ground Water
1993).
223
-------
25
a=0.5 in.
a=l in.
a=2 in.
a=3 in.
10 IS
Influence Radius, b (in.)
Rtprtntad by permission of the Ground Water Publishing Company. Copyright 1996.
Figure 6-5. Theoretical dependence of remediation time on air channel radius
"a" and average spacing between channels (b) at a constant flow
rate (Ground Water, 1996).
6.1.4 Contaminant Type
In general, constituents that are easily removed from contaminated groundwater
through traditional air stripping processes are considered to be well suited to removal
by in situ air sparging systems. Interactions (e.g., soil absorption and retention effects)
within the subsurface may limit the effectiveness of this process. Lighter petroleum
constituents are most amenable to air sparging (C3-C10) (Marley et al., 1992).
Enhanced air sparging techniques are being investigated that may influence less-
volatile constituents. Examples of these enhanced techniques include combinations of
air, ozone, and/or hydrogen peroxide as injected gas to provide increased oxidation or
biodegradation potential for semivolatile constituents.
6.1.5 Vertical and Horizontal Placement of Sparging Well Points
The screened interval of sparging wells is usually no greater than 1 to 3 feet in length
because nearly all of,the airflow will occur very close to the top of the screened interval
where the hydrostatic water pressure is lowest. The depth selected for placement of
sparging points below the water table will depend on several factors. In particular,
sparging points should be placed beneath residual product and at the greatest depth
that contamination is found (as determined by soil analysis). In addition, the radius of
224
-------
influence of a sparging well generally increases with height above the sparging point;
thus, the greater the depth of placement below the contaminated zone, the greater the
areal sweep of the well and the fewer wells that need to be installed.
In fairly uniform soils, the zone of influence is roughly parabolic in shape. As a general
rule, the diameter of the zone of influence will be about 0.5 to 1.5 times the distance
above the sparging point. Intermittent layers and other heterogeneities can greatly
affect the pattern of airflow, however, thus making an evaluation of the effective
diameter one of the most difficult issues in system design.
The factor that must be considered in deciding placement depth is the pressure
required to overcome the hydrostatic head of water. For every foot below the water
table that the system is placed, the gas pressure must be increased by 0.43 psi.
6.1.6 Other Factors
In addition to uncertainties concerning the effects of soil heterogeneity and flow chan-
neling on remediation effectiveness, the following little-known factors may also affect
recovery:
• Enhanced biodegradation from increased aeration via air sparging
Potential for remobilizing residual NAPL in groundwater
•• The further enhancement of biodegradation by the "airlift pump effect."
Air sparging will cause an increase in water pressure at the sparging point. This
increase could have various effects. First, it should produce a circulation field in the
groundwater near the sparging well. It also could distribute oxygen and bring dissolved
contamination to the sparge point where volatiles may be stripped. Finally, the effect
could result in migration of volatile contaminants beyond the zone of influence of the
sparging wells. The importance of these effects has not been demonstrated however
and is not well understood. '
Another possible effect of sparging is the potential for remobilizing residual NAPL
present in the saturated zone because of the change in capillary pressures induced by
elevated gas pressures. Pulsed sparging may induce additional dynamic effects that
could cause NAPL movement. Remobilization of NAPL could have the beneficial effect
of enhancing mass transfer processes by increasing the interfacial surface area be-
tween NAPL and water. NAPL mobilization, however, could also result in movement to
previously uncontaminated areas. Field studies by Ground Water (1996), for example,
have found increased concentrations of dissolved contaminants downgradient of air
sparging wells.
6.1.7 Applicability
An air sparging system, if properly designed, installed, and operated, can reduce con-
taminant concentrations in groundwater and saturated soils. Air sparging is most
225
-------
applicable if the contaminants are VOCs and the extent of contamination is limited to
the upper 5 to 30 feet of a shallow, water-table aquifer. Air sparging technology is more
successful and controllable if the aquifer is relatively homogeneous and free of low-
permeability layers of silt, clay, or other geologic materials. The volatilization of
organics is most effective if the VOCs have a high vapor pressure and volatilize easily.
In many cases, indigenous microbes may be stimulated by the introduction of oxygen
during air sparging and cause a significant reduction of hydrocarbons in the ground-
water and on the saturated soils. This result is particularly likely for fuel-related hydro-
carbons.
6.1.8 Potential Limitations and Disadvantages
Air sparging is a technology primarily intended for easily volatilized organic compounds.
It also is appropriate for organic compounds that are biodegraded quickly when oxygen
is supplied. For many nonvolatile organic compounds, such as polynuclear aromatic
hydrocarbons (PAH) and some chlorinated hydrocarbons, air sparging would probably
not be appropriate or effective for remediation.
Because an air sparging well typically has a zone of influence smaller than a pumping
well capture zone, more air sparging wells are needed to cover equal areas. As the
depth to groundwater increases and the thickness of the aquifer decreases, air spar-
ging becomes less cost-effective relative to a pump-and-treat system. Thus, air
sparging is more attractive for use with shallow water-table aquifers.
Air sparging is normally only operated in the region of 0 to 30 feet below the water
table. Each foot of water in a well casing requires 0.433 pound per square inch (psi) of
air pressure to displace. Therefore, 13.0 psi are required at a minimum to displace 30
feet of water from a casing before air can reach the aquifer. Additional air pressure is
needed to overcome friction losses in the airlines and capillary pressures (i.e., air entry
pressure) at the well/formation interface. Therefore, equipment and operating costs
increase as the depth below water table increases. An increase in the air pressure
needed for air sparging is accompanied by an increase in concerns regarding safety
and the integrity of pipes and the outer well seals. Thus, air sparging would not be an
appropriate technology for use on contaminated zones far below the water table (e.g.,
greater than 30 feet below the water table).
Moving air into a geologic media is easier if the material is a coarse, high-permeability
sand or gravel. The pressure necessary to force air into these materials is much lower
than that required for fine-grained materials. In fine-grained clayey sand, silt, or clay,
the permeability of the materials essentially precludes the use of air sparging as a
remediation technology.
The effectiveness of air sparging is sensitive to the lithologic and stratigraphic heteroge-
neities of a remediation site. In highly stratified materials, air may travel far from the
226
-------
well along zones of more permeable strata (i.e., on the underneath side of clay and sift
layers) before reaching the vadose zone (Figure 6-4). This course could have one or
more negative impacts:
Airstreams/channels may not fully contact the target contaminated zone.
Mounding of the groundwater table would result in a lateral flow of ground-
water and the contaminant plume.
Injected air could potentially migrate along low-permeability zones.
Therefore, additional care must be taken during the design and operation of an air
sparging system in stratified materials to ensure that contaminants are contained and
removed from the ground.
6.1.9 System Design and Operation
A typical air sparging system consists of an oil-free air compressor manifolded to seve-
ral air sparging wells. A compressor rated for continuous duty at the maximum ex-
pected flow rate and pressure provides the greatest flexibility during full-scale system
operation. Although oil-free reciprocating compressors are readily available, they are
usually intended for intermittent use only. Continuous-duty reciprocating compressors
are available, but cost nearly twice as much (Marley, et al., 1992).
Pressurized air is supplied to the sparging wells by a header-type pipe distribution
system. Metal pipe or rubber hose rated for pressure may be needed for air distribu-
tion, depending on site-specific conditions and the anticipated pressures. The use of
rigid plastic pipe (e.g., polyvinyl chloride [PVC]) as header pipes should be avoided
because the heat generated during air compression can damage the pipe.
A pressure gauge and regulator should be provided at each sparging well as a means
of measuring and controlling airflow at each injection point. In addition, check valves
should be placed at the entrance to each sparge well. Once the formation is pressur-
ized, air and water might flow back up the well if the air sparging system is turned off.
The check valve prevents the back flow from reaching the pressure regulator, manifold
line, or compressor (Marley, et al., 1992).
As stated previously, air sparging wells can be installed vertically, at an angle, or
horizontally (by use of,directional drilling). In addition, slotted pipe can be placed in the
bottom of dug trenches, which are subsequently backfilled with coarse sand or gravel.
The well casing and screens can be constructed of PVC or metal. Care shpuld be
taken when constructing the annular seal for each well. The grout used to seal the
wells should be nonshrinking because any cracks or bridging in the seal will allow short:
circuiting of air along the outside of the well casing and greatly reduce the effectiveness
of the sparging well. A screen length of 1 to 3 feet is usually sufficient for vertical and
i
227
-------
angle wells. For horizontal wells, small-diameter pipes of different lengths can be
inserted within the horizontal well to ensure that compressed air is distributed along the
entire length of the well.
Air pressure supplied to sparging wells typically ranges from <5 to 40 psi. Air pressure
as high as 60 psi has been used during one pilot-scale test because air was being
injected into fine sandy silt, i.e., low-permeability material (Marley, et al., 1992).
Airflow rates commonly range from 2 to 15 standard cubic feet per minute (scfm) per
sparge well (Marley, et al., 1992; Johnson, et al., 1993). A high flow rate (170 to 270
scfm) was reportedly introduced to a horizontal air-sparge well because of the longer
available screen length (Kaback, et al., 1991). This is one of the advantages of using
horizontal wells or horizontal pipes in trenches.
Air sparging systems can be operated in the following modes:
• Continuous injection of air into all air-sparging wells
• Alternate injection into the wells (e.g., half the wells are used while half
are inactive at any one time)
• Pulse flow into all wells (i.e., on/off), thus allowing the compressor to run
intermittently.
The advantage of operating under the pulse-flow mode is that the flow into each well is
stopped and started in cycles. This causes air channels to collapse during the off
period and new ones to reform when airflow is resumed. Presumably, the new air
channels will not always coincide with the old channels, thus increasing the potential
contact between air and contaminated water/sediment. In other words, channels
reform in random locations, thus increasing the chances that more pore spaces will
eventually be aerated. The advantage to operating under alternate injection is that air
is injected to only half the wells at a time, so the compressor can be sized at roughly
half the capacity than what would otherwise be required.
If an SVE system is used to capture VOCs from the unsaturated zone, then the system
should be designed for a negative pressure in an area larger than the zone where
sparging is occurring, and the extraction flow rate should be greater than the injection
flow rate.
6.1.10 Performance Monitoring
Performance monitoring methods for use with air sparging systems are difficult to deter-
mine and are still being developed. The zone of influence of individual sparging wells
and the overall effects and efficiency of a sparging system can be difficult to measure.
228
-------
In coarser, homogeneous soils in which channel distribution is more regular, the zone of
influence is somewhat symmetrical around the well and the effective radius is
dependent on the air injection pressure and flow rate. In heterogeneous, stratified
materials, sparge air can traverse laterally over fairly long distances in one or more
directions.
Although specific parameters for monitoring system performance have not been fully
determined, Table 6-1 presents recommended parameters for monitoring system
operation and in situ response. These parameters can be used in determining the
following:
Influence of the sparging system based on dissolved oxygen concentra-
tions.
Changes in contaminant concentrations in groundwater during system
operation.
i
Mass removed by the air sparging-soil vapor evaporation (AS-SVE)
system from SVE off-gas concentrations.
Distribution of contamination and SVE performance based on soil data.
Airflow in the subsurface based on vacuum/pressure distribution and
airflow measurements.
As mentioned in Chapter 4, a series of tracer tests (e.g., SF6, He) and in situ aerobic
respiration tests can be performed in addition to tests on those parameters that are
important to successful performance monitoring of AS-SVE systems (e.g., dissolved
oxygen, hydrocarbon concentrations, flow rates, and pressure). At many sites, it may
be difficult to determine and interpret these performance monitoring data because of
the high degree of temporal variability. Therefore, tracer tests can be used to gain a
better understanding of the contributing factors leading to the variability. The tracer
tests can be used to assess airflow in the unsaturated zone and to determine the extent:
to which hydrocarbon vapors are migrating off site. The tracer tests can also comple-
ment aerobic respiration tests that are used to determine oxygen utilization as an
indicator of biodegradation.
229
-------
Table 6-1. Recommended Monitoring Requirements
Parameter
Measurement Method
System Operation
Extraction well flow rate
Injection well flow rate
Extraction well vacuum
Injection well pressure
Extraction gas concentration
Extraction well composition3
O2 and CO2
In Situ Response
Contaminant distribution levels in vadose
zone and saturated zone soil
Soil gas concentrations
Soil gas composition
O2 and CO2
Soil gas pressure/vacuum
Groundwater elevation
Contaminant levels and distribution in
groundwater
Dissolved oxygen levels
In situ airflow rate
Flowmeter measurement methods (pitot tube,
orifice plate, etc.)
Flowmeter (rotameter, orifice plate, thermal
anemometer, etc.)
Vacuum gauge or manometer
Pressure gauge
Flame ionization detector (FID)
Gas chromatography with FID
Electrochemical cell (oxygen)
Infrared detector (carbon dioxide)
Analysis of soil sample by accepted method
FID
Analysis by accepted method
Electrochemical cell (oxygen)
Infrared detector (carbon dioxide)
Pressure/vacuum gauge manometer,b or inclined
manometer
Electronic water level indicator or tape in monitor-
ing well
Analysis of groundwater sample by accepted
method
Analysis of groundwater sample
Flowmeter
Includes compositional analyses of hydrocarbon (boiling point fractionation or individual species).
Requires vadose monitoring installations or soil gas probes.
Source: Modified from Johnson et al., 1993
230
-------
Section 7
Cost Estimates
7.1 SVE Systems
The cost of procuring, installing, and maintaining a soil vapor extraction (SVE) system
can be divided into three general categories, as described in detail in the "Soil Vapor
Extraction Technology Reference Handbook" (EPA, 1991 a):
Site assessment costs, including a site history and detailed site
characterization.
Capital costs, including design and installation of system components,
permitting, and contingencies.
" Operation and Monitoring (O&M) costs associated with continued opera-
tion of the system.
Table 7-1 briefly describes each type of cost along with a summary of example unit
costs. Table 7-2 presents an example cost estimate for a hypothetical SVE system.
These costs are modified from estimates provided in the SVE Handbook.
7.1.1 Site Investigation Costs
A complete site investigation includes a review of site history, on-site field screening,
and, if necessary, geological and geophysical investigations and environmental
sampling. The site historical review, including interviews and file searches, is designed
to minimize the effort necessary on later tasks. On-site field screening often includes
soil gas and groundwater sampling, and should be designed to better define the nature
and extent of any contamination present. Sites with more complex geology, or more
extensive contamination, may require more detailed site characterization, including
geophysical surveys, more detailed soil gas surveys, and detailed geologic sampling
and evaluation.
The cosits for all of the above tasks can vary dramatically, depending on site history
(surface structures, land use), geological characteristics (depth to groundwater, soil
properties, and homogeneity), and nature of contamination (one or more constituents or
product types, single or multiple events). At the low end, for a site with a known history,
231
-------
Table 7-1. Example Unit Costs for SVE Systems
Item
Site Investigation
Soil sampling
Soil gas analysis
Surface geophysics
Analytical
Capital Costs
Extraction well construction
Casing
Screen
Piping
Valves •
Joints
Surface seals
Blower
Air/water separator
Magnehelic gauge
Flow meter
Sampling port
Soil gas probe
Diffuser stacks
Activated carbon canisters
i
Operation and Monitoring Costs
Type
hand auger
drilling rig
KV system
EM
GPR
Seismic
VOC
ABN
TPH
PVC
PVC
PVC
PVC
PVC
PVC
Bentonite for annular
seal, concrete at surface
Intrinsically safe
Knockout pots
Vacuum
Brass T
Aluminum
Carbon steel
Stainless steel
G-series
Size
1- to 2-inch ID
4-inch auger
8-inch auger
4-inch
4-inch
4-inch
4-inch
4-inch
6-inch
1 hp
130 gal.
0.5-inch ID
4-inch
4-inch
1 00-200 cfm
Cost ($)
$10-40/sample
$20-50/ft
$100-3000/day
$1500/day
$2000/day
$3000/day
$200/sample
$500/sample
$100/sample
$20-50/foot
$3-5/foot
$5-7/foot
$3/foot
$300 ea.
$16 ea.
$3/sq. yd.
$1700ea.
$2400 ea.
$50-75 ea.
•
$20-30 ea.
$30-50 ea.
$8/foot
$30/foot
$700-1 000 ea.
Power (continuous operation)
Vapor treatment
Electricity at 100/kWh
Carbon adsorption
Thermal incineration
232
1-hp blower
$18/day
$20/pound gasoline
removal
Supplemental fuel $1/gallon
-------
Table 7-1 (continued)
Item
Monitoring and analysis
• —
1993 dollars
Table 7-2. Hypothetical SVE
Cost Component
Site investigation
Capital costs
Operation and monitoring
(5 years planned)
- • . .
Total Cost
1993 dollars
Type Size
Catalytic oxidation 200scfm
Internal combustion en- Propane
gines Natural gas
Soil or groundwater TPH
TCLP
VOC
BTEX
ABN
System Cost Estimate
Description
Soil gas survey
Soil sampling and well installation
Soil analysis
Data evaluation and interpretation
Total Site Investigation Costs
Recovery well installation
Piping system
Blower system
Carbon adsorption system
License, permit, legal fees
Project management, engineering de-
sign
Total Capital Costs
Power requirements
Maintenance
Water disposal from knockout tank
Monthly monitoring and analysis
Additional carbon
Total O&M
Cost ($)
;$800/month
$1/gallon
$0.75/gallon
'$75-125 ea.
$11 00-1600 ea.
$200-250 ea.
$100ea.
$450-550 ea.
: cost ($)
$4,000
11,000
4,000
9.000
$28,000
$18,000
15,000
! 30,000
10,000
2,000
18.000
1
1 $73,000
i 10,000
: 10,000
20,000
60,000
i 30.000
130,000
$251,000
233
-------
a well-documented recent release of a known contaminant, and a relatively simple
geologic environment, complete site assessment costs could be as low as $2,000
assuming no intrusive field work is necessary to define the problem.
A more typical situation would involve an investigation of site history (including inter-
views with employees and adjacent property owners). In addition, a preliminary soil gas
survey and use of surface geophysical methods may help define the extent of plume
contamination. Confirmatory soil borings and an analysis of soil and groundwater could
bring the cost of this type of site assessment to between $10,000 and $50,000.
For the more complex site, with perhaps multiple industrial users, multiple historical
contamination events, and a heterogeneous hydrogeologic environment, more detailed
historical review and site characterization would probably be required. Including a more
detailed soil gas survey, additional soil borings and laboratory analyses and perhaps
complementary geophysical techniques, a detailed site assessment on such a site
could range between $20,000 and $75,000.
7.1.2 Capital Costs
Capital costs include engineering design, procurement and installation of the SVE
system, permitting, piping, and instrumentation. The primary cost groups include:
• Vapor capture - primarily well installation, surface seals, and groundwater
level control. The typical system consists of PVC wells and a piping
system to connect injection wells and extraction wells to the removal
system. A typical 30-foot-deep extraction well installation may range in
cost from $2,000 to $4,000, depending on the method of construction,
type of geologic conditions encountered, and diameter of piping. Of this
cost, up to one-half may be for well materials. Depending on the extent of
contamination and the properties of the aquifer, the number of wells
needed for remediation could range from as few as 3 to more than 50.
Vapor removal - consisting of pumps, blowers, valves, monitoring devices,
and other equipment used to remove vapors from the ground and trans-
port them to the vapor treatment system. Typically, a vacuum pump or
positive displacement blower provides the power for an SVE system.
Spark and explosion-proof blowers range in price from $1,700 (1-hp) to
$6,000 (30-hp), depending on fan size and flow rating. Monitoring equip-
ment, which measures vacuum airflow and vapor characteristics, includes
a rnagnehelic gauge at each well ($50-$75) to measure vacuum, in-line
airflow meters ($300), and quick-coupling sampling ports ($25) used in
conjunction with a portable organic vapor analyzer (OVA) or similar device
to measure hydrocarbon concentrations ($6,000 to $25,000).
234
-------
Vaoor treatment - including vapor pretreatment, side-stream treatment,
and a vapor treatment device. Pretreatment involves an air-water separa-
tor (often with a demister) ranging from a simple drum to complex level
controls. Typical size ranges from 20 to 130 gallons, with cost in the
$1,000 to $2,500 range. A variety of methods are available to treat
accumulated liquids and vapors, ranging from carbon adsorption (canis-
ters range in cost between $600 and $11,000), catalytic oxidation ($3,000
to $200,000), and thermal incineration ($12,000 to $40,000). In general,
costs are proportional to the flow rate to be treated and influent concentra-
tions. Supplemental fuel may be required to maintain temperatures for
adequate removal at some influent streams.
For a typical site, the range of expected capital costs is between $10,000 and $50,000.
As with site-assessment costs, site and contaminant characteristics can dramatically
affect total capital costs.
7.13 Operation and Monitoring Costs
Costs for system operation and monitoring (O&M) include blower and piping system
power requirements, handling and disposal of recovered liquids, and off-gas treatment.
O&M costs for small systems are low, ranging up to $50/month. Off-gas treatment, if
necessary, can be expensive, ranging up to $2,000/month.
O&M costs primarily include power requirements for blowers and condensers (if
necessary), fuel costs for incineration, monitoring and analyses to measure progress,
and labor costs (if operated manually). The cost for power is calculated using the
formula:
Cost = 0.75 (fan horsepower) (electricity cost, per kWh) (hours of operation)
Vapor treatment costs are a function of the treatment method, contaminant concentra-
tion, and flow rate. Carbon adsorption costs generally increase with concentration;
incineration and oxidation costs generally decrease with concentration, with sustaining
fuel costs dominating the total cost. Example unit costs are shown in Table 7-1.
7.14 Hypothetical Cost Estimate Case
The totsal cost of a site remediation using SVE can be estimated as:
where:
CR = cost of remediation
Cs = cost of site assessment
235
-------
Cc = capital cost
CMO = average monthly operation cost
CMM = average monthly maintenance cost
T = time of remediation in months.
Total cost CR is the sum of the site assessment and capital costs plus the O&M unit
costs multiplied by the estimated time necessary to achieve site remediation. The time
estimate involves selection of target constituents, initial groundwater concentrations,
and ultimate cleanup concentrations.
Overall system costs have been estimated to range from $35 to $100 per cubic yard of
contaminated soil. This range of costs should cover the small- to average-size site
experiencing controlled site and waste conditions. Site assessment and capital costs
are site specific as described above.
The hypothetical site is a service station that reports a 1000-gallon spill of unleaded
gasoline. After the leak is detected, 400 gallons of product are recovered from the silty
sand aquifer underlying the site. The water table is 20 feet below grade, and the
response of the field investigation is able to limit the extent of contamination to an area
above the water table and about 50 feet by 150 feet in extent underlying the site. Six
recovery wells are installed in the vadose zone and an SVE system initiated. The cost
estimate on Table 7-2 presents the hypothetical costing for this site remediation.
7.2 SVE/Air Sparging Systems
The cost for SVE/air sparging systems is similar to that for SVE-based systems,
including the division into the same three general categories-site assessment costs,
capital costs, and operating and monitoring costs.
Table 7-3 briefly describes each type of cost in a summary of example unit costs pre-
sented in similar fashion to that shown for SVE systems (see Table 7-1). This table
also presents an example of a hypothetical SVE/air sparging system cost estimate.
7.2.1 Site Investigation Costs
As with SVE systems, site investigations involving SVE/air sparging can include a wide
variety of site conditions and data requirements. The investigation could range from a
low of about $2000 (no extensive field work) to upwards of $60,000 for a complex site
in a heterogenous hydrogeologic environment. The most typical situations will include
limited field screening-and analyses, with costs ranging between $10,000 and $25,000.
7.2.2 Capital Costs
As with SVE-based systems, capital costs include engineering design, procurement and
installation of the SVE/air sparging system, permitting, piping, and instrumentation. The
primary cost groups include vapor capture, vapor removal, and vapor treatment.
236
-------
Table 7-3. Hypothetical SVE/Air Sparging System Cost Estimate
Cost Component
Description
Cost ($)
Site Investigation
Soil gas survey
Soil sampling and well installation
Soil analysis
Data evaluation and interpretation
Total Site Investigation
$4,000
15,000
8,000
12.000
39,000
Capital Costs
Injection wells
Recovery well installation
Piping system
Blower system
License, permit, legal fee,
project management, engineering de-
sign
Total capital costs
18,000
25,000
60,000
3,000
25.000
131,000
Operation and Monitoring
(2 years)
Total Cost
Power requirements
Off-gas emissions control
Maintenance
Bimonthly monitoring
Labor
Total O&M
Contingency
$16,000
240,000
10,000
68,000
' 30.000
364,000
20,000
$554,000
237
-------
Vapor capture typically includes installation of PVC wells and piping in various dia-
meters (2 to 12 inch). As described under SVE system costs, a typical 30-foot-deep
extraction well may range in cost between $2000 and $4000, depending on the method
of construction, type of geologic conditions encountered, and pipe diameter. Of this
cost, about one-half may be for well materials (casing, screen, plugs, filter pack, bento-
nite, and grout). Costs can be reduced by early implementation of an effective well
configuration. For example, the use of nested wells, including both sparging and
extraction wells in the same hole, can reduce the total number of borings required.
The use' of horizontal wells, although more expensive, may increase the VOC extraction
efficiency and thus the cost-effectiveness of the system. System piping can be placed
aboveground or buried in trenches. Aboveground piping is obviously more economical
if the site is inactive and the piping is secure. Aboveground piping may also require
heat tracing and insulation to prevent freezing.
A vapor removal system consists of pumps, blowers, and other equipment used to
remove vapors from the ground. An air sparging system introduces air into the
saturated zone using mechanical compression equipment. The type of equipment used
depends upon the flow rate and pressure required, which is partially governed by the
static water pressure above the sparge point and the air-entry pressure of the soil. The
injected air must be oil-free. Vacuum pumps or positive-displacement blowers extract
the sparged air in addition to the airflow induced through the vadose zone. Example
costs are described under SVE-based systems costs.
Vapor treatment includes pretreatment, side-stream treatment, and treatment devices
identical to an SVE-based system. If pretreatment is required, options include carbon
adsorption (55-gallon drums and skid-mounted systems), catalytic oxidation (requires
additional treatment of hydrochloric acid generated during the process), and thermal
incineration (may require supplemental fuel to maintain required temperatures). If
pretreatment is not required, diffuser stacks may be designed to direct vapors into the
atmosphere.
In addition to the above costs, engineering and design fees, permit acquisition, and
other miscellaneous costs are often included as capital costs. Because these costs are
highly site-specific, the figures estimated here are arbitrary. A general guideline is that
engineering and design fees should comprise about 10 to 15 percent of total system
cost. For a typical site, the range of expected capital costs is between $20,000 and
$75,000. Site variability and contaminant characteristics are the major factors in
determining the construction requirements of the system.
7.2.3 Operation and Monitoring Costs
Costs required for system operation and monitoring (O&M) include blower and piping
system power requirements, handling and disposal of recovered liquids, and other
ongoing costs such as labor. Table 7-1 presents example unit costs needed for
operating a system. The successful operation of an SVE/air sparging system requires
238
-------
monitoring of the extracted vapor stream (VOCs, O2/CO2), groundwater (BTEX TPH
dissolved oxygen), and soil (VOCs, biological assay tests) to ensure appropriate
contaminant removal during degradation. Air sparging systems are expected to add 20
to 25 percent to the cost of conventional SVE systems.
7.2,4 Hypothetical Cost Estimate Case
Overall system costs are expected to range from $35 to $75 per cubic yard in the
unsaturated/saturated zones of petroleum-related cleanup sites. These costs are
estimated to be approximately 50 percent of those for conventional SVE systems
designed primarily to accelerate cleanup times using air sparging technology.
As an example cost estimate, consider a leaking UST site remediation where air
sparging is used to deal with gasoline contamination in both the unsaturated and
saturated zones. At this hypothetical site, the depth to water is 60 feet and up to
10,000 cubic yards of soil are believed to be contaminated.
Capital costs at this site include those for two vapor extraction wells, one air injection
well, and four groundwater monitoring wells. The system includes a 25-hp vacuum
pump, a 15-hp air injection compressor, two air/water separators, and piping An off-
gas emissions control system, consisting of granular-activated carbon, is required to
capture BTEX compounds. Because carbon regeneration will occur off site, costs are
included under O&M. O&M costs assume bimonthly analysis of extraction well effluent
concentrations with a portable GC and a total cleanup period of two years
239
-------
-------
Appendix A
Chemical Data
List of Tables
Page
A-1 Unweathered Composition of Three Common Hydrocarbon Products 241
A-2 Range of Abundance of Some of the Constituents Typically Found in
Virgin Mixtures of Gasoline 246
A-3 Range of Abundance of Some Aromatic Chemicals Typically Found in
Virgin Mixtures of Diesel Fuel 247
A-4 Major Components of JP-4 248
A-5 Average Composition of Gasoline Vapor Exposures 252
A-6 Physicochemical Properties of Five Common Hydrocarbon Mixtures 254
240
-------
Table A-1. Unweathered Composition of Three Common Hydrocarbon Products
(from USEPA, 1990d).
Selected Representative
Concentrations (% w/w)
Hydrocarbon
Group
n-Alkanes
C4
C5
C6
C7
C8
C9
C10-C14
Branched Alkanes
C4
C5
C6
C7
C8
C9
C10-C14
Representative
Hydrocarbon
n-Butane
n-Pentane
n-Hexane
n-Heptane
n-Octane
n-Nonane
n-Decane
Isobutane
Isopentane
2-Methylpentane
2-Methylhexane
2,4-Dimethylhexane
2,2,4-Trimethylhexane
2,2,5,5-Tetramethyl-
1
Automotive
Gasoline
10.8-29.6
4.8 - 7.0
1.9-4.5
2.0-12.9
0.2 - 2.3
1.3
0.4 - 0.8
0.2 - 0.8
18.18-59.5
0.7 - 2.2
8.6-17.3
4.6 - 9.7
1.4-8.3
1.8-16.7
1.2-2.7
0.5-2.6
2 3
#2 Fuel Jet Fuel
Oil JP-4
0.12
1.06
2.21
3.67
3.80
2.25
8.73
0.66
2.27
5.48
8.82.
3.36
1.35
•hexane
241
-------
Table A-1. Unweathered Composition of Three Common Hydrocarbon Products
(continued).
Selected Representative
Concentrations (% w/w)
Hydrocarbon
Group
Cvcloalkanes
C6
C7
C8
C9
Others
Olefins
C4
C5
C6
Others
Mono-aromatics
Benzene
Toluene
Xylenes
Ethyl benzene
C3-benzenes
Representative
Hydrocarbon
Cyclohexane
Methylcyclohexane
1 ,2,4-Trimethylcyclo-
pentane
1 ,1 ,3-Trimethylcyclo-
hexane
1-Butene
1-Pentene
1-Hexene
Benzene
Toluene
-m-Xylene
Ethyl benzene
1 ,3,5-Trimethylbenzene
1
Automotive
Gasoline
3.2-13.7
0.2
1.0-3.9
0.2-1.4
0.2 - 0.7
1.6-7.5
5.5-13.5
0.9
1.3-3.3
0.8-1.8
2.5-7.5
19.3-40.9
0.9 - 4.4
4.0 - 6.5
5.6 - 8.8
1.2-1.4
3.2-11.3
2 ; 3
#2 Fuel Jet Fuel
Oil JP-4
2.40
3.77
1.35
3.21
i
0.50
1.33
0.07 2.32
0.03 0.37
0.67 ; 3.59
242
-------
Table A-1. Unweathered Composition of Three Common Hydrocarbon Products
(continued).
Hydrocarbon
Group
Representative
Hydrocarbon
Selected Representative
Concentrations (% w/w)
1 2 3
Automotive #2 Fuel Jet Fuel
Gasoline Oil JP-4
Mono-aromatics (continued)
C4-benzenes
Other
Phenols
Phenol
C1-phenols
C2-phenols
CS-phenols
C4-phenols
Indanol
Poly-aromatics
Nitro-aromatics
C1-anilines
C2-anilines
Complex anilines
1 ,4-Diethylbenzene
Phenol
o-Cresol
2,4-Dimethylphenol
2,4,6-Trimethylphenol
m-Ethylphenol
Indanol
Fluorene
Quinoline
2.1-2.6 0.88 3
1.6-5.2
0.001
0.01
0.02
0.02
0.01
0.001
0.57
0.003
0.004
0.002
.98
Di-aromatics
Naphthalene
0.7
3.43
1.59
243
-------
Table A-1. Unweathered Composition of Three Common Hydrocarbon Products
(continued).
Hydrocarbon
Group
Saturated
hydrocarbons
C8
C9
C10
C11
C12
013
C14
C15
C16
C17
C18
C19
C20
C21
C22
C23
C24
Pristane
Phytane
Representative
Hydrocarbon
n-Octane
n-Nonane
n-Decane
n-Undecane
n-Dodecane
n-Tridecane
n-Tetradecane
n-Pentadecane
n-Hexadecane
n-Heptadecane
n-Octadecane
n-Nonadecane
n-Eicosane
n-Heneicosane
n-Docosane
n-Tricosane
n-Tetracosane
Selected Representative
Concentrations (% w/w)
1 23
Automotive #2 Fuel Jet Fuel
Gasoline Oil JP-4
»
0.05
0.20
0.58
0.98 i
1.14
1.20
1.31
1.42
1.53
1.51
1.31
1.16
0.99
0.51
0.29
0.15
0.05
0.52
0.46
244
-------
Table A-1. Unweathercd Composition of Three Common Hydrocarbon Products
(continued).
Selected Representative
Concentrations (% w/w)
Hydrocarbon
Group
Unknowns
Representative
Hydrocarbon
1
Automotive
Gasoline
6.6-13.8
2
#2 Fuel
Oil
3
Jet Fuel
JP-4
NOTE: Blanks indicate the unavailability of data and do not indicate the absence of a
particular compound from the hydrocarbon product.
245
-------
Table A-2. Range of Abundance of Some of the Constituents Typically Found in
Virgin Mixtures of Gasoline. Values are in Percent by Weight.8 (From
Stelljes and Watkin, 1993).
Component Maximum Abundance Minimum Abundance
Aromiatics
Benzene 3.5 0.12
Toluene 21.8 2.73
Ethylbenzene 2.86 0.36
Xylenes (total) 8.31 3.22
Naphthalene 0.49 0.09
2-Methylnaphthalene 3.85 2.91
Benzo(a)pyrene 2.8 xlO"6 1.9 xlO'7
Branched Alkanes
Isopentane 10.17 6.07
n-Alkanes '
n-Butane
n-Penltane
n-Hexane
Additives
Ethylene dibromide
4.70
10.92
3.5
1.77X10-4
3.93
5.75
0.24
.
7x10'7
Percent by weight values can be converted to parts per million (milligrams per
kilogram) by multiplying the values shown above by 106.
246
-------
Table A-3. Range of Abundance of Some Aromatic Chemicals Typically Found
in Virgin Mixtures of Diesel Fuel. Values are in Parts Per Million
(mg/l) by Weight" (From Stelljes and Watkin, 1993).
Component
Benzene
Toluene
Ethylbenzene
Xylenes (total)
Naphthalene
2-Methylnaphthalene
Benzo(a)pyrene
Benz(a)anthracene
Chrysene
Fluoranthene
Phenanthrene
Pyrene
Triphenylene
Cresol
Phenol
Quinoline
Maximum Abundance
82
800
800
800
2,730b
6,700b
0.6
1.2
2.2b
37
1,500b
41b
2.2b
54.3b
6.8b
9.2b
Minimum Abundance
6
100
100
100
2,730b
6,700b
0.006
0.001
2.2b
NDC
1,500"
41b
2.2b
54.3b
6.8b .
9.2b
a Parts per million can be converted to percent by weight by dividing the concentrations
shown above by 106.
Only one concentration was reported for this chemical.
c ND s Not detected.
247
-------
Table A-4. Major Components of JP-4 (From Smith et al., 1989).
Fuel Component
n-Buf:ane
Isobutane
N-Pentane
2,2-Dimethylbutane
2-Methylpentane
3-Me1:hylpentane
N-Hexane
Methylcyclopentane
2,2-Dimethylpentane
Benzene
Cyclohexane
2-Methylhexane
3-Methylhexane
trans-1 ,2-Dimethylcyclopentane
cis-1 ,3-Dimethylcyclopentane
cis-1 ,2-Dimethylcyclopentane
n-Hepltane
Methylcyclohexane
2,2,3,3-Tetramethylbutane
Ethylcyclopentane
2,5-Dirnethylhexane
2,4-Dirnethylhexane
1 ,2,4-TrimethyIcyclopentane
3,3-Dimethylhexane
Kovats Index
400.0
466.3
500.0
527.7
562.4
578.7
600.0
622.0
629.1
644.5
653.6
669.5
677.3
679.6
681.9
684.4
700.0
715.1
720.5
729.8
737.3
738.4
740.8
743.3
Percent by Weight
0.12
0.66
1.06
0.10
1.28
0.89
2.21
1.16
0.25
0.50
1.24
2.35
1.97
0-36
0.34
0.54
3.67
2,27
0.24
0,26
0,37
o;58
0.25
0.26
248
-------
Table A-4. Major Components of JP-4 (continued).
Fuel Component
1 ,2,3-Trimethylcyclopentane
Toluene
2,2-Dimethylhexane
2-Methylheptane
4-Methylheptane
cis-1 ,3-Dimethylcyclohexane
3-Methylheptane
1 -Methy l-3-ethylcyclohexane
1 -Methyl-2-ethylcyclohexane
Dimethylcyclohexane
n-Octane
1 ,3,5-Trimethylcyclohexane
1 ,1 ,3-Trimethylcyclohexane
2,5-Dimethylheptane
Unidentified
Ethylbenzene
m-Xylene
p-Xylene
3,4-Dimethylheptane
4-Ethylheptane
4-Methyloctane
2-Methyloctane
3-Methyloctane
o-Xylene
Kovats Index
748.1
753.0
764.2
772.0
772.7
775.3
778.0
784.1
786.7
788.8
800.0
825.3
831.0
833.6
839.9
844.9
853.9
854.8
859.8
865.0
868.5
869.6
873.9
875.3
Percent by Weight
0.25
1.33
0.71
2.70
0.92
0.42
3.04
0.17
0.39
0.43
3.80
0.99
0.48
0.52
0.98
0.37
0.96
0.35
0.43
0.18
0.86
0.88
0.79
1.01
249
-------
Table A-4. Major Components of JP-4 (continued).
Fuel Component
1 -Methyl-4-ethylcyclohexane
n-Nonane
Isopropylbenzene
n-Propylbenzene
1 -Methyl-3-ethylbenzene
1 -Methyl-4-ethylbenzene
1 ,3,5-Trimethylbenzene
1 -Methyl-3-ethylbenzene
1 ,2,4-Trimethylbenzene
n-Decane
n-Butylcyclohexane
1 ,3-Diethylbenzene
1 -Methyl-4-propylbenzene
1 ,3-Dirnethyl-5-ethylbenzene
1 -Methyl-2-i-propylbenzene
1 ,4-Dirnethyl-2-ethylbenzene
1 ,2-Dirnethyl-4-ethylbenzene
n-Undecane
1 ,2,3,4-Tetramethylbenzene
Naphthalene
2-Methylundecane
n-Dodecane
2,6-Dimethylundecane
Unidentified
Kovats Index
881.3
900.0
905.1
937.2
944.9
946.8
952.8
961.0
975.6
1000.0
1025.6
1031.4
1034.7
1041.6
1049.1
1060.2
1067.1
1100.0
1128.8
1156.5
1166.0
1200.0
1216.1
1262.3
Percent by Weight
0.48
2.25
0.30
0.71
0.49
0.43
0.42
0.23
1.01
2.16
0.70
0.46
0.40
0,61
, \
0:29
0:70
0.77
2032
0.75
0.50
0.64
2.00
0.71
0.68
250
-------
Table A-4. Major Components of JP-4 (continued).
Fuel Component Kovats Index Percent by Weight
2-Methylnaphthalene 1265.7 0.56
1-Methylnapthalene 1276.4 0.78
n-Tridecane 1300.0 1.52
2,6-Dimethylnaphthalene 1379.4 0.25
n-Tetradecane 1400.0 0.73
251
-------
Table A-5. Average Composition of Gasoline Vapor Exposures (From Haider et
al., 1986).
C3
Compound
n-Propane
n-Butane
Isobutane
n-Pentane
Isopentane
Cyclopentane
2,3-Dimethylbutane
2-Methylpentane
3-Methylpentane
Methylcyclopentane
3-Methylhexane
n-Hexane
2,3-Dimethylpentane
2,4-Dimethylpentane
2-Methylhexane
3-methylhexane
n-Heptane
2,2,4-Trimethylpentane
Isobutylene
1-Butene
trans-2 Butene
cis-2-Butene
Compositional Make-Up8
Amoco Oil"
(wt %)
Amoco Oil0
(wt %)
Shell Oil"
(vol %)
Alkanes:
33.7 (7.8)
4.1 (0.8)
8.1 (2.5)
21.6(3.7)
212(10.4)
3.4(1.6)
9.4(1.5)
27.2 (6.7)
1.3(0.7) 3.3(1.8)
3.4(1.3) 4.9(1.4)
2.0 (0.7) 3.2 (0.9)
1.1(0.6) 1.5(0.4)
0.7(0.4) 11.03
1.8(0.7) 3.1(0.7)
0.6 (0.3) 0.9 (0.8)
0.8 (0.5)
0.6(0.3) 1.1(0.3)
0.7(0.4) 11.03
0.7 (0.2)
0.7(0.5) 1.8(1.2)
Alkenes:
1.2(0.5)
0.9 (0.3)
1.0(0.7)
1.2(0.7)
0.8(1.1)
38.1 (5.7)
5.2(1.9)
7.0 (4.0)
22.9 (6.1)
0.7 (0.7)
0.7 (0.5)
2.1 (1.3)
1.6(0.9)
1.3(0.4)
1.5(0.9)
0.7 (0.6)
0
0.5 (0.5)
1.1 (1.5)
API8
(wt %)
10.9(4.2)
1-7(0.9)
252
-------
Table A-5. Average Composition of Gasoline Vapor Exposures (continued).
Compositional Make-Up8
Compound
Cs 2-MethyI-1-butene
2-Methyl-2-butene
1-Pentene
trans-2-Pentene
cis-2-Pentene
Amoco Oilb
(wt %)
0.9 (0.4)
-
-
0.8 (0.6)
-
Amoco Oil0
(wt%)
-
1.5(0.7)
0.7 (0.4)
-
-
Shell Oild
(vol %)
1.6(2.1)
1.7(1.8)
-
-
1.2(1.7)
API6
(wt %)
Aromatics:
C6 Benzene
Cj Toluene
Ca Xylene (p, m, o)
Total Percent
2.2(1.0)
3.1 (1.6)b
0.9 (0.7)
89.7
0.6 (0.3)
4.0(1.8)"
1.5(0.7)
94.1
0.7 (0.4)
1.8(1.3)
0.5 (0.6)
91.7
2.2(1.1)
2.2(1.8)
1.1 (1.5)
a Components listed comprise at least 0.5% by wt. or vol. Composition less than 0.5% denoted by"-",
Composition presented as arithmetic mean (± standard deviation).
N = 12. Bulk terminal exposures.
N = 11. Marine loading exposures.
d N « 95.
6 N = 152.
Toluene coeluted with 2,3,3-trimethylpentane on the analytical column; however, the major proportion is
assumed to be toluene.
253
-------
Table A-6. Physiochemical Properties of Five Common Hydrocarbon Mixtures
(From USEPA, 1990a).
Product
Automotive
Gasoline
#2 Fuel Oil
#6 Fuel Oil
Jet Fu«l (JP-4)
Mineral Base
Crankcase Oil
Air
Saturated
Aqueous Vapor
Liquid Density
(g/cm3)
(0.73)
0.72-0.76 [15.6]
(0.91)
0.87-0.95
(0.91)
0.87-0.95
0.75
0.84-0.96 [15]
-
1
Liquid Viscosity
(cPoise)
(0.45)
0.36-0.49 [15.6]
(1.56)
1.15-1.97 [21]
(254)
14.5-493.5 [38]
0.829 [21]
275 [38]
'-
1
Water Solubility
(mg/l)
(158)
131 -185 [13-25]
3-10 [20-23]
-5
10-20
insoluble
-
Vapor Pressure
(mmHg)
(469)
263-675 [38]
(14.3)
2.12-26.4 [21]
(14.3)
2.12-26.4 [21]
91
N/A
760
17 5
N/A = Not Available.
Notes: All values for 20°C unless noted in brackets [].
Values in parentheses are typical of the parameter ().
Values for air and saturated aqueous vapor are included, where applicable, as a means of
comparison.
254
-------
-------
Appendix B
Recommended Specifications
Contents
Appendix B.1
Appendix B.2
Appendix B.3
Appendix B.4
Well Construction and Specifications
Collection System Design Guidance
Equipment Specifications
Instrumentation and Control
Page
256
274
282
291
255
-------
Appendix B.1
Well Construction and Specifications
Introduction
This Appendix provides guidance on the specification of proper well/trench construction
for multiphase fluid extraction, soil vapor extraction (SVE), and system monitoring. This
guidance is not comprehensive and must be adapted as necessary for site-specific
conditions and objectives. Guide specifications for well construction are available
through the Corps of Engineers' Guide Specification (CEGS) system, including CEGS
02671 Wells for Monitoring Groundwater and CEGS 02670 Water Wells. These can be
modified for typical SVE system or multiphase fluid recovery applications. Guidance for
selecting well location and screen placement is not provided here; refer to the main
text.
Applicable Standards
In the specification, reference is made by title and number of all American Society for
Testing and Materials (ASTM) or other standards for materials and testing procedures
identified within the specification. For example, standards exist for plastic well casing
(ASTM F 480, D 1785, D 2241), cement (C 150), and soil classification (D 2487 and
D 2488). The Environmental Protection Agency (EPA), ASTM, American Water Works
Association (AWWA), American National Standards Institute (ANSI), and National
Sanitation Foundation (NSF) also have applicable standards for materials (e.g., NSF
Standard 14) or well construction (e.g., EPA 570/9-75/001, AWWA A100, ANSI/ASAE
EP400.1, ASTM D 5092) that may be appropriate to reference. Standard texts on well
construction, such as Driscoll (1986), may also be appropriate to reference.
Quality Control/Assurance
Several testing procedures can be performed to ensure that the installation of the wells
or trenches has been successful. Specifications require appropriate quality verification.
Well/Trench Performance Testing
Multiphase Fluid Recovery Well Performance
The performance of the well should be verified. This includes requiring a measurement
of the draw down in the well at various flow rates (specific capacity), measuring free
product recovery rates, and/or performing a pump test in the well. The measurement of
specific capacity or product recovery rates after completion allows a comparison with
the design yield and provides a baseline against which later performance can be
256
-------
compared as a guide for well maintenance. The performance of a pump test, using
nearby observation or monitoring wells, provides data on aquifer transmissivity and
storativity that can be useful in operational modeling of the extraction system. Perhaps
more important, a pump test can provide an estimate of well efficiency to be developed.
This estimate can be used to evaluate the adequacy of the installation. The
specification should describe the testing procedures and duration(s) and the methods of
data analysis. Driscoll or other text on well hydraulics can be used to obtain information
on applicable methods.
SVE Well Performance and Leak Testing
Preliminary testing of the airflow rate and vacuum levels at the well head can indicate if
the well has been adequately installed. Drastically lower-than-expected airflow rates or
much-higher-than-expected vacuum levels, compared to design values, at a well may
suggest damage or improper placement. Either the main blower system or a skid-
mounted blower and treatment system can be used for the test. If the SVE extraction
or vapor monitoring well is not properly sealed, air may "short circuit" from the surface
along to the casing. Monitoring of nearby wells during the performance testing of
extraction may indicate a drastically smaller-than-expected radius of influence for the
well and suggest a leak. Cement-bond geophysical logging may be appropriate if a
poor bond between the casing and grout or voids behind the casing are suspected.
The well can also be pressurized and if air is found leaking around the well head, poor
installation should be suspected.
Recovery Trench Performance Testing
The performance of the trench should be assessed in a manner similar to a vertical
extraction well. The response of vacuum monitoring points or piezometers installed
within and outside the trench backfill should be measured to verify the response along
the trench.
Alignment Verification
Vertical Well Alignment and Survey
For deep (greater than 75 to 100 feet) installations, some wells may significantly deviate
from the vertical. In these relatively rare instances, an alignment test, typically run by a
geophysical firm, may be necessary to verify the actual screen location and attitude.
Specification of maximum deviation from vertical or the target coordinates may be
appropriate.
Plumbness and Pump Placement
Wells in which pumps are to be placed should be tested for plumbness and alignment
to assure that the pumps can be placed in the wells. Usually a well survey will
determine the actual deviation of the well from the vertical. A dummy pipe of a diameter
%- to 1/£--inch larger than the pump can be used to verify easy passage in the well. The
dummy must be decontaminated and disinfected before each use. ;
257
-------
Trench Alignment
A survey may be required of the alignment of the casing and screen to verify its grade,
placement, and condition. Downhole (i.e., crawler-type) cameras can be used to verify
the condition of the pipe and identify breaks or crimping in large-diameter pipe. Various
methods are used for surveying horizontal wells to verify location. These surveys are
strongly recommended because the directional control is highly variable, especially on
long (more than 100 feet) runs.
Trench Backfill Compaction Density Testing
Soil density of the trench backfill is tested to verify adequate compaction for excavated
trenches. Many methods are used to determine soil density, including Esteems D
2922, Test Methods for Density of Soil and Soil-Aggregate in Place by Nuclear Method;
D 2167, Test Method for Density and Unit Weight of Soil in Place by the Rubber
Balloon Method; and D 1556, Test Method for Density of Soil in Place by the Sand-
Cone Method. These tests should be specified at intervals along the trench or at the
discretion of the field representative.
Contractor Qualifications
Competent professionals, drillers, and installers are required for successful installation
of wells and trenches. Minimum criteria for these personnel must be identified in the
specification.
Well Installation
Specify the level of experience of the contractor's well driller and hydrogeologist (or
engineer) directing the well installation. Also specify state registration or certification
where required.
Horizontal Well/Trench Installer Qualifications
Special requirements may be established for the operators of the trenching machine or
horizontal drilling rig, such as a minimum number of months or years of experience. A
registered or licensed driller may also be necessary.
Submittals
The contractor must submit various items to allow evaluation of contractor designs, to
verify contractor performance, and to record installations. The level of approval for the
submittal depends on the needs of the project. Items that are critical to project success
should require approval before the contractor moves forward with the work. Others are
only provided for inforrnation and do not need approval.
Work Plan
The contractor must submit a general plan of action, including materials, locations,
drilling and installation procedures, and schedules. This requirement is particularly
important if much of the design was left to the contractor to meet a performance goal.
This submittal normally requires approval. For efficiency, this plan should include all
258
-------
drilling activities, including those required for installation of multiphase fluid extraction
wells, vapor monitoring points, and groundwater monitoring points. This plan may be
part of a comprehensive contractor work plan governing all cleanup activities.
Permits and Other Documentation
Copies of the permits and clearances should be submitted, including disposal records
pertaining to drill cutting and other waste. Documentation of the contractor's qualifi-
cations is often required as a submittal.
Catalog Data
Catalog information on the materials used or to be used should be submitted.
Boring Logs and As-Built Diagrams
The contractor must submit a log of materials encountered in drilling and a detailed
as-built drawing of each well.
Test Results
The results of testing conducted for quality assurance purposes, such as the survey/
alignment, performance, leak, compaction density or alignment tests, should be
submitted if performed by the contractor. In addition, results of any chemical or
physical testing of soils or other materials should be submitted. !
Development Records
In specifications for multiphase fluid extraction wells, the contractor must submit a
record of the actions taken to develop the well. This requirement includes quantitative
observations of turbidity and sand production, calculations of the total fluid volume
removed, and a description of the development tools used.
Multiphase Fluid Recovery Wells
Multiphase fluid recovery wells are intended to capture any combination of ground-
water, free product, and vapors. This action provides a checklist of topics to be
covered in a specification for such wells. Typical requirements are discussed under
each topic. Measurement and payment, chemical quality management, safety, site
preparation and cleanup, and other topics are not included here but would normally be
addressed in the specifications.
Materials
The materials used for multiphase fluid recovery wells will generally depend on site
conditions and project objectives. Composition of the materials depends on the
subsurface geochemistry uncloaking the natural constituents and contaminants.
Casing
For many applications, Schedule 40 PVC well casing is adequate. The recommended
specification is ASTM D 1785, Standard Specification for Polyvinyl Chloride (PVC)
259
-------
Plastic Pipe, Schedule 40, 80, 120 or ASTM F 480, Standard Specification for
Thermoplastic Water Well Casing Pipe and Couplings Made in Standard Dimension
Ratio (SDR). If high levels of liquid organics are to be encountered by the casing, the
compatibility of the casing material with the fluids must be considered. Stainless steel
(generally Schedule 5S or 10S, type 304) is required if PVC will be degraded by the
product. The recommended specification is ASTM A 312, Standard Specification for
Seamless and Welded Austenitic Stainless Steel pipe.
Alternatively, PVC may be preferred in an environment that is highly corrosive to
metals. The well can be a "hybrid" of PVC casing and stainless steel screen. PVC
casing exposed to sunlight should be protected or treated to withstand ultraviolet
radiation without becoming brittle. Casing diameter generally depends on pump space
requirements. Dual-phase pumps usually require a minimum of 6 inches inside
diameter; larger diameters allow easier pump installation. If only groundwater and
vapors are to be removed, groundwater pumps as small as 2 inches in diameter
capable of pumping 10 gallons per minute are available. Generally, 6-inch-diameter or
larger wells are recommended. The specifications should require casing with
flush-threaded joints and O-ring seals.
Screen
Well screen is usually PVC, but as noted above, other materials may be more
appropriate. The use of continuous-wrap "V-wire" screen is strongly recommended.
Screen slot size is designed based on the formation material gradation established in
the methods outlined in Driscoll (1986) or another similar reference. Different slot sizes
can be used in different portions of the screened interval if the producing formation
varies in soil gradation. If the gradations of the producing formation have not been
determined during design, the contractor should obtain samples during drilling. The
contractor must run gradations according to an appropriate method (e.g., ASTM D 422
Standard Method for Particle-C Size Analysis of Soils) and then size the screen slot
(and filter pack, discussed below) accordingly. Screening with flush-threaded joints and
O-ring seals is preferred.
Filter Pack
The filter pack requirements for this application are generally more critical than for SVE
wells because the filter pack plays a more significant role in reducing entrainment of
fine sands, silts, and clays in the fluid produced. As described above, the filter pack
gradation should be chosen based on the gradation of the producing formation. Design
should follow methods, outlined in Driscoll (1986) or another similar reference. If only
groundwater and vapors are to be recovered, the chosen filter pack must have a
uniformity coefficient of 2.5 or less. A less uniform filter pack may be appropriate if
nonwetting fluids, such as hydrocarbons, are to be recovered. Rounded to subrounded
siliceous particles, free from organic matter and calcareous or elongated particles, are
required. If free product recovery is of primary concern, a special filter pack that
includes hydrophobic materials, such as ground plastic or PTFE, may improve the early
260
-------
' !
rates of product recovery (Hampton and others, 1993). In certain (relatively rare)
circumstances, a well can be designed that does not initially include a filter pack, but
rather develops a natural filter pack. Thorough well development can selectively
remove fines from the native formation material and leave coarser native sands and
gravel around the well as a natural pack.
Seal and Grout
A well seal is necessary to prevent entry of grout into the filter pack and well screen.
Unamended sodium bentonite, as pellets, granules, or a high-solids bentonite grout, is
normally specified for the seal material. Because most applications will involve the
extraction of groundwater and either floating product or soil vapors, the well seal will be
above the water table and pellets or granules must be hydrated with water added to the
annulus. A cement grout is normally required above the bentonite well seal. The
mixture of the grout, which should be specified, is normally one 94-pound bag of
cement (with up to 5 pounds of bentonite powder added to further resist cracking) and
less than 8 gallons of clean water. The recommended specification is ASTM standard
C 150, Standard Specification for Portland Gement. In the event that the seal will be
placed below the water table, the use of bentonite pellets placed by tremie pipe is
preferred.
End Caps and Centralizers
Flush-threaded end caps, consistent with the casing and screen in size andimaterial
should be specified. Centralizers center the well in the borehole and must be the size
appropriate for the casing and borehole. Centralizers must be made of material that will
not lead to galvanic corrosion of the casing. Stainless steel Centralizers are
recommended with PVC or stainless steel casing.
Installation
Acceptable practices for installing the wells should be described in this portion of the
specification.
Test Holes
Careful design of the filter pack, screen slot size, and screen location is based on
site-specific conditions. The contractor may need to drill test holes at the proposed well
locations to obtain boring logs and samples. The number, locations, and depths of test
holes should be specified.
Drilling Methods
There are many methods for drilling. Drilling methods can be proposed by the
contractor or specified. Mud-based drilling fluids should be avoided if possible because
of the difficulty in developing the zone containing floating product. The use of water-
based fluids can also impede product recovery because the water can displace the
hydrocarbon near the well and disrupt continuous hydrocarbon flow pathways. Auger
air-rotary, dual-wall air-casing-hammer, or cable tool drilling may be acceptable,
261
-------
depending on site conditions. All equipment must be decontaminated and disinfected
before drilling is begun at each location.
Soil Sampling and Logging
Sampling of soils encountered during drilling can lead to an increase understanding of
the subsurface and can allow a better decision to be made about well construction
including screen placement. Soils must be sampled at regular intervals, at least every 5
feet; sometimes, continuous sampling is appropriate. Samples should be obtained by
an appropriate method such as a split spoon sampler or thin-walled tube, as specified
in ASTM D-1586, Standard Method for Penetration test and Split-Barrel Sampling of
Soils, or D-1587, Thin-Walled Tube Sampling of Soils, respectively. Sample volume
requirements should be considered when specifying the sampling method. Sampling
for chemical and physical analyses must be done according to an approved sampling
and analysis plan. It is strongly recommended that a drilling log be prepared by a
geologist or geotechnical engineer. Materials encountered should be described
according to a standard such as ASTM D 2488, Standard Practice for Description and
Identification of Soil (Visual-Manual Procedure). Geophysical logging may be
appropriate for borings that extend into the water table. Electrical and gamma ray logs
can help identify coarser materials for screen placement and can supplement or reduce
the amount of soil sampling. This can reduce the time needed to drill and sample the
hole.
Borehole Diameter and Depth
The dimensions of the borehole for well installation should be specified. Normally, the
diameter is at least 4 inches greater than the diameter of the casing and screen to allow
placement of the filter pack. The depth of the borehole should be based on the screen
depth. The borehole should only extend to 1 foot below the projected bottom of the
screen.
Screen and Casing Placement
The casing and screen should be cleaned or decontaminated before placement. The
screen and casing should be joined by flush-threaded joints and suspended in the
center of the borehole. To maintain plumbness and alignment, the string should not be
allowed to rest on the bottom of the hole. Centralizers should be placed on the casing
at regular intervals if the depth of the well exceeds some minimum value such as 20
feet.
Filter Pack Placement
The specification should require the filter pack to be placed by using a decontaminated
tremie pipe. Because much, if not most, of the filter pack is placed below the water
table, the tremie pipe should be kept within 2 to 5 feet of the surface of the placed filter
pack. This prevents the pack material from bridging or segregating by size while falling
through the water column. The level of the pack material should be measured following
placement. Approximately 1 foot of filter pack should be placed in the borehole below
262
-------
the bottom of the screen to act as a cushion for the screen and casing. Filter pack
material should extend 2 to 5 feet above the top of the screen to allow for settlement so
that native material will not collapse around the screen. Gentle agitation of the water
within the well during or after filter pack placement can help in ensuring full settlement
before grouting. The pack material should be stored and handled carefully to avoid
contamination from undesirable materials.
Seal and Grout Placement
The grouting of the well is critical to prevent short circuiting resulting from air leakage
from the ground surface when a vacuum is applied. Normally 3 to 5 feet of a bentonite
well seal is placed above the filter pack. If the well seal is to be placed above the water
table, f:he specification should include a requirement for hydrating the bentonite before,'
placement of the grout. The specification should require the addition of a volume of
distilled or potable water for every 6-inch lift of bentonite pellets or granules. The
bentonite should hydrate for at least 1 to 2 hours prior to placement of the grout. This
can be avoided by specifying the use of a bentonite high-solids grout as the seal. The
high-solids bentonite grout should be placed by tremie pipe. A cement grout should
also be pumped into annular space via a side-discharge tremie pipe, and the pipe
should be kept submerged in the grout during grout placement. If the grout is to be
placed to a depth of less than 15 feet, the grout may be poured into place directly from
the surface. If the well seal is to be placed below the water table, the bentonite pellets
must be allowed to hydrate in place for 2 to 4 hours before the well is grouted. Fine
sand can be placed above the bentonite pellets to further prevent grout intrusion.
Surface Completion
A suitable well head is required to extract multiple phases from a single well. The well
head specifications may require multiple discharge pipes, electrical leads, compressed-
air or vacuum lines, control device leads, and sampling ports. These requirements are
project-specific. If finished above grade, the well may require suitable protection, such
as bollards, to avoid damage to the well from traffic, etc. A well vault may also be
required. Separate specifications may be required for the construction of the vault itself
and for other aspects of the instrumentation, as appropriate. The use of pitiless
adapters, etc., should be described, and reference should be made to appropriate
drawing details.
Well Development
Wei! development is critical to the ultimate performance of the well. A careful specifica-
tion of the acceptable Development methods and development criteria is strongly
recommended. The well must be developed by surging and bailing. In addition, a
suitable size surge block or jetting must be used at appropriate water velocities. Note
that jetting can affect product recovery by disrupting floating hydrocarbon flow path-
ways. The development should continue until the well is producing clear water with less
than 2 to 5 ppm by weight sand and/or other suspended solids. A turbidity criterion
defined as less than 5 nephelometric turbidity units (NTU) determined by a
263
-------
nephelometric turbidity measure merit method can be used. Sometimes, the use of
dispersing agents such as phosphates can help develop wells by breaking down clay
smears on the borehole walls. The regulatory authorities may need to approve
dispersing agents or other additives such as acids. The well is developed after
placement of the filter pack and before or after the well is grouted. Development prior
to the grouting of the well will ensure that the filter pack is fully settled before grout
placement, thus assuring no voids would be created; however, the potential exists for
cross contamination while the well annulus is open above the pack. Normally, the well
should be developed after grouting.
Disinfection
In some cases, biological encrustation has caused severe degradation of the perfor-
mance of the extraction wells. Contaminated sites often provide ample food for
microorganisms that can plug well screens. Disinfection of the drilling tools and the well
itself can help prevent or slow these problems. Disinfection can be accomplished by
various means (Driscoll, 1986), including creating a specified concentration of a strong
oxidizing agent, such as sodium hypochlorite, in the well. The chemical ramifications of
any additives should be considered.
Surveys
A survey should be done to establish the horizontal coordinates of the well. The
elevation of the top of the casing should be surveyed to provide accurate groundwater
elevations*. The accuracy of the surveys depends on the project needs, but generally
they are accurate to the nearest foot for the horizontal coordinates and the nearest 0.01
foot for elevation.
Permits
The contractor may be required to obtain utility clearances and certain permits from the
regulatory agencies and/or the land owner. In addition, the contractor should comply
with local water well installation, abandonment, and reporting requirements for SVE/EV/
air sparging wells.
Well Acceptance
The contractor must provide a well that functions properly, to the extent that site
conditions permit. If the contractor, due to its negligence or error, installs a nonfunc-
tional well, the specification must require the contractor to repair or replace the well at
its expense. If the well cannot be repaired, the contractor must be required to properly
decommission the we[l (see Section on "Well Decommissioning").
Soil Vapor Extraction Wells
Many topics covered in the section on multiphase fluid recovery wells are relevant for
SVE wells. This section outlines differences between the checklist topics described for
multiphase fluid recovery wells and topics appropriate for SVE well construction
specifications. Other topics covered for multiphase fluid recovery wells but not
264
-------
mentioned here are directly applicable to soil vapor extraction wells with the exception
of well development and disinfection. Wells used for passive or active air injection
generally can also be installed according to these requirements.
Materials
The composition, properties, and sizes of materials should be specified as discussed
below.
Casing
New polyvinyl chloride (PVC) pipe, 4 to 6 inches in diameter, is normally used for SVE
well casing. Larger diameters are preferred to increase flow capacity, but require larger
boreholes. Schedule 40 PVC well casing is adequate for most SVE applications
because the casing will generally not be in contact with liquids. Other materials may be
specified if contaminants, at expected vapor/eondensate concentrations, are likely to be
damaging to PVC. PVC casing exposed to sunlight should be protected or treated to
withstand ultraviolet radiation without becoming brittle. The casing must be strong
enough to resist collapse at the expected vacuum levels and grout pressures.
Screen
Well screen is usually PVC with slotted or continuous wrap openings. Continuous-wrap
screen is strongly preferred because the increased open area reduces the pressure
drop across the screen and therefore reduces energy costs for the blower. Slot size is
generally 0.020 inch, but should be as large as possible to reduce the pressure vacuum
drop across the screen. Slot sizes of 0.040 inch or larger may be used. Larger slots
sizes may, in a few cases, lead to increased entrainment of abrasive particles in the
airflow.
Filter Pack ' . j .
Pack material should be a commercially available highly uniform gradation of siliceous
sand or gravel with no contaminants (chemical or physical). A uniformity coefficient of
2.5 or less should be specified. The actual gradation should generally be based on the
formation grain size and the screen slot size. Coarser material may be used; however,
coarser gradations may, in a few cases, lead to increased entrainment of abrasive
particles in the airflow.
Seal and Grout
In essentially all cases, the well seal in SVE wells will be placed above the water table
Guidance on specifying materials in seals or above the water table in multiphase fluid
extraction wells will apply here. A cement/bentonite grout is preferred to fill the annulus
above the seal to the ground surface because it resists desiccation cracking^
Installation
The installation of the SVE well is similar to the installation of multiphase fluid extraction
wells. This section notes the differences.
265
-------
Drilling Methods
Some drilling methods, particularly those using drilling mud, are to be avoided because
of the potential to plug the unsaturated soils. Hollow-stem auger drilling is most
common and is preferred where appropriate.
Filter Pack Placement
Filter pack should be placed around the screen to some level above the top of the
screen, normally 2 to 3 feet. Filter pack is normally placed dry by being poured down a
tremie pipe.
Surface Completion
The completion of the well head will depend on the other features of the design, such
as the piping and instrumentation requirements. Project drawings should be referenced
as appropriate. An appropriate "tee" may be placed below or at grade to establish a
connection with buried or aboveground piping, respectively. A vertical extension from
the tee to a specified level will allow attachment of appropriate instrumentation. If a
surface cover is used, it may be appropriate to describe the means by which the cover
is sealed around the well. This can also be described in a specification for the place-
ment of the cover. Refer to CEGS 02271, Geomembrane Barrier, for further guidance.
Surveys
A survey is used to establish the horizontal coordinates of the well. The elevation of the
top of the casing should be surveyed. If the SVE well intercepts groundwater, the water
elevation would be of interest. Caruso requirements are similar to those for multiphase
fluid extraction wells.
Vapor Extraction Trench
This section provides a checklist of topics to be covered in a specification for vapor (or
liquid) extraction trenches. Such trenches are often used at sites with shallow ground-
water or near-surface contamination; thus, the depth of excavation is often modest. A
horizontal recovery system can be placed by several methods including normal
excavation, trenching machines (which excavate and place pipe and filter pack in one
pass), and horizontal well drilling. Where possible, typical requirements are discussed
under each topic. Many specification topics are very site specific. Additional topics,
such as measurement and payment, chemical quality management, safety, site
preparation and cleanup, are not included here but would normally be addressed in the
specifications. A Corps of Engineers Guide Specification, 02222 Excavation and
Backfilling for Utilities Systems, can be modified for application to vapor extraction
trenches excavated by normal means.
Materials
Materials specified for extraction trench construction are often similar to those specified
for vertical wells. Different materials may be needed if specialized trenching (or
266
-------
drilling/jacking) methods or machines are used. Differences between horizontal and
vertical applications are discussed below.
Casing
Although PVC casing is commonly used, flexible or rigid polyethylene pipe may be
more efficient for certain excavation methods such as trenching machines. The pipe
must resist the crushing pressures of the backfill and compaction equipment. Refer-
ence should be made to appropriate ASTM standards for PVC pipe or ASTM D 3350,
Standard Specification for Polyethylene Plastics Pipe and Fittings Materials. The
specification should allow casing to be joined by threaded coupling or thermowelds, as
appropriate for the material. Pipe sizes of 4 to 8 inches are often used. Larger pipe
sizes allow easier access for surveys and maintenance.
Screen
Given the generally longer screened intervals in horizontal applications, air-entry
vefocities are generally lower and well efficiency is less of a concern. Thus, the screen
open area can be somewhat lower than is needed in vertical wells. Although continu-
ous-wrap screen is still preferred, successful systems have also used slotted pipe. If
slotted pipe is specified or allowed, the specification should require a minimum open
area perfect. Drain pipe wrapped with geotextile must not be used because of the
potential for fine material to plug the geotextile. Slot size can be quite large, 0.040 inch
or larger, because the lower air velocities reduce the potential for entrapment of small
particles. The screen can be joined by threaded couplings or it can be thermowelded.
For some horizontal well applications, a prepacked well screen is appropriate. Pre-
packed screens are really two screens enclosing preselected filter pack material. The
use of prepacked screen can overcome the difficulties of installing filter pack within a
horizontal well.
Bedding Material/ Filter Pack
Generally, the guidance for specifying filter pack in SVE wells applies for trenches, but
somewhat coarser material may be needed for a secure bedding for the pipe and '
screen. A reference to ASTM D 2321, Standard Practice for Underground Installation
of Flexible Thermoplastic Sewer Pipe may be appropriate. Filter material placed above
the water table generally need not be sized for the formation, and can be quite coarse
A reference to ASTM C 136, Standard Method for Sieve Analysis of Fine and Coarse
Aggregate, may be appropriate for verifying the gradation.
Cover and Seal Material
Native material may occasionally be used as backfill above the filter pack in an exca-
vated trench. Given that vapor extraction trenches are typically used at sites with
shallow groundwater, low permeability material is preferable to enhance the lateral
vacuum influence of the trench. The use of clay or a geomembrane is required, if
appropriate. Refer to the CEGS 02271, Geomembrane Barrier for Landfill over for
267
-------
guidance on specifications for geomembranes, or CEGS 2443, Low Permeability Clay
Layer, for specification of clay backfill.
Geotextile
A geotextile may be needed to separate the filter pack from native material or clay
backfill in an excavated trench. Refer to CEGS 02272, Separation/Filtration Geotextile,
for guidance in specifying the geotextile.
Marking Tape and Locator Strips
A locator strip is needed specifically manufactured for marking underground utilities.
This tape is made of colored polyethylene that is backed with foil or contains embedded
wire to allow others to locate the trench at later dates. This would not be applicable for
horizontal well installations.
Installation
Installation methods vary significantly depending on excavation method.
Excavation Methods
Methods used to install trenches or other horizontal installations include standard earth
excavating equipment (e.g., backhoe), trenching machines, horizontal drilling tech-
niques, and pipe jacking/microtunneling. Given this wide variety, it may be desirable to
only specify the pipe, screen, pack materials, and an ultimate pipe alignment and depth.
This would allow the contractor the option to propose what might be the most cost-
effective method; however, the trenching technique used by the contractor must provide
an adequate filter placement around the collector pipe. Note that horizontal drilling,
pipe jacking, etc., reduces the amount of disturbed material and minimizes both the
potential for worker exposure and disruption to surface features. Most horizontal drilling
techniques require drilling fluids that may not be appropriate for vapor extraction
techniques.
Soil Sampling and Logging
If open excavation techniques are used, a graphical log of the materials encountered in
the trench should be prepared, including the description of the materials according to
ASTM D 2488. Other excavation methods will require some log of the materials
encountered at different stations, and would usually be based on cutting returns from
the trenching machine or drilling. Other sampling should be done as needed according
to an approved sampling and analysis plan.
Trench Dimensions
The trench dimension should be wide enough to allow preparation of the bottom of the
trench and placement of the pipe. Normally, the trench width is limited to the pipe
diameter plus 24 inches. If the material to be trenched is contaminated, a smaller
trench reduces the volume of material to be disposed of or treated as waste. Com-
pliance with appropriate OSHA regulations is required for workers who need to enter a
268
-------
trench for installation. If a horizontal drilling method is used, some annular space
between the borehole and the screen should be required in a manner similar to that
used for vertical wells. The use of a prepacked well screen may require less annular
space,
Trench Bottom Preparation and Pipe Placement
The bottoms of the excavated trenches must be prepared prior to placement of pipe
and screen. The trench must be leveled to the required grade to provide uniform
bearing for the pipe. A bedding layer of filter pack material 4 to 8 inches thick should be
placed and compacted prior to pipe and screen placement. All rocks or angular debris
greater than 3 inches in diameter must be removed from the trench bottom to avoid
damage to the pipe. Unstable materials should also be removed. The pipe and screen
should be placed in a way that prevents entrapment of filter pack or native material
inside the pipe. The section of the pipe and screen must be joined in a manner
consistent with the material used and the manufacturer's recommendations. A clean-
out or access port for the pipe should be provided to allow for later surveysiand
maintenance of the screen and casing. This should be shown on the drawings. The
contractor must prevent the run-in of surface water into the trench. If the trench is to be
installed to below groundwater, dewatering may be necessary. Separate dewatering
and waiter treatment spell suffocations may be required.
Soil Stockpiles
The specification should describe what should be done to manage the excavated
material. This is highly site-specific. For example, the soil may be used for; backfill or it
may need treatment. Adequate measures to protect the material from being contami-
nated or from spreading contamination should be required.
Filter Pack Placement
Filter pack placement is relatively simple in open trenches, but much more difficult in
drilling or jacking operations. The filter pack material should not be compacted with
6 inches to 1-foot of the pipe and screen. Some trenching machines place the pipe and
filter pack material as it progresses. In these cases, it is important to verify that the
machine is placing adequate filter pack around the screen. For horizontal drilling
applications, various methods exist for placing the filter pack; the most common and
probably most desirable of these methods is the use of the prepacked screen. In this
method, the native material is allowed to collapse back upon the prepacked ;screen.
Backfilling and Compaction
The remainder of an excavated trench should be backfilled with the specified material
to the grades shown on the drawings. Placement of a geotextile between the filter pack
and backfill may be appropriate if there is a significant difference in grain sizb between
the two materials. Refer to CEGS 2272 Separation/Filtration Geotextile for guidance on
specifying placement of the geotextile. Backfill should be placed in 6- to 8-ihch lifts and
compacted to approximately 90 percent optimum standard density, determined by
i
269
-------
ASTM D 69, Test Method for Laboratory Compaction Characteristics of Soil Using
Standard Effort, if cohesive materials are used. The specification can identify the
acceptable compaction method. The locator strip should be placed within 18 inches of
the surface.
Surface Cover and Seal Placement
The specification should identify the means to place the surface cover, if one is used.
Refer to the CEGS 02271 Geomembrane Liner for guidance on specifying handling,
seaming, penetration, and other practices if a geomembrahe is required. Generally, a
minimum of 6 inches of soil cover is placed over a geomembrane to protect it. Some-
times concrete or asphalt paving is used instead of a geomembrane, and a reference to
separate specifications on base course and pavement construction would be appropri-
ate. As with the vertical wells, an appropriate surface completion will be required, such
as a well vault or below-grade connection.
Repair of Casing or Screen or Removal of Trench
The contractor should be required to repair or replace, at its expense, any segment of
the casing or screen that was damaged because of inadequate installation or materials.
In the event the entire installation fails to function or is no longer needed at the end of
the project, the casing and screen can be pressure grouted or excavated for salvage.
Excavated trenches should be backfilled and compacted according to the appropriate
part of the specification.
Soil, Vapor, and Water Sampling
Extraction Well Sampling
The specifications should require a baseline sample of the groundwater, soil vapor,
and/or product constituents prior to system operation. This can assist in treatment
design and monitoring of overall system performance upon startup. Sampling should
be done according to the approved sampling and analysis plan.
Soil Sampling for Chemical Analysis
In addition to sampling for logging and physical testing purposes, soil samples for
chemical analyses are often required. This information is used to further characterize
the extent of contamination and to establish a baseline against which soil remediation
by vapor extraction is judged. Sampling should be done according to an approved
sampling and analysis plan.
>
Monitoring Point Sampling
The specification should require an initial sampling of the soil gas and/or groundwater
contaminant concentrations from each monitoring point. This should be done according
to the approved site sampling and analysis plan.
270
-------
Soil Gas/Vacuum Monitoring Points
This section outlines differences between the check lists for multiphase fluid recov-
ery/SVE wells and the topics appropriate for monitoring point construction specifica-
tions. Other topics covered for extraction wells are also generally applicable to soil
gas/vacuum monitoring points. The applicability of the topics should be evaluated on a
site-specific basis. Again, this appendix does not address the location or depth
selection for these features.
Materials
Generally, the same materials can be used for the monitoring points as are used for the
soil vapor extraction wells; however, there will be obvious differences in size.
Casing
Generally, 3/4- to 2-inch-diameter PVC pipe is used. Flush-threaded pipe is preferred,
but for smaller diameters, couplings may be needed. For some shallow applications,
flexible Teflon or polyethylene tubing can be used. ;
Screen
Either slotted or continuous-wrap screen can be specified. Slotted pipe is adequate for
monitoring ports. Continuous-wrap screen is not commonly available at the smaller
diameters (less than normal 2-inch-diameter), but can be ordered. Slot sizes smaller
than those typically used for extraction wells may be appropriate for monitoring points
(i.e., 0.010- to 0.020-inch slots). Other "screen" types can be used. Options include
slotted drive points, porous points or, for short-term use, even open-ended pipe.
Filter Pack
Filter pack material should be appropriately sized for the screen slot width. The pack
simply provides support for the screen and is not critical to monitoring point function. In
some cases, no filter pack will be necessary (see Section on "Drilling Methods").
Installation
Drilling Methods
Although the use of a hollow-stem auger is still the primary means of installing monitor-
ing points, direct-push methods can also be used to place slotted-drive points or other
vacuum/soil gas probes at specific depths. Again, mud- or fluid-based drilling methods
are not appropriate for this work.
Soil Sampling and Logging
As with SVE wells, it is appropriate to adequately sample the materials encountered for
logging purposes and physical and chemical testing. Samples must be obtained at
least every 5 feet from holes drilled for monitoring points. If the monitoring point is
located in close proximity (less than 5 to 10 feet away) to another well that has been
logged and sampled, a separate log is usually not required.
271
-------
Borehole Diameter and Depth
The borehole diameter should be approximately 4 inches larger than the screen/casing
to allow placement of the filter pack. This obviously would not apply to points placed by
direct push methods. Multiple wells/tubes can be placed in a single borehole, and
these are typically referred to as multiport completions. Adequate room must be
allowed for proper installation if multiport monitoring systems are to be used. Multiport
monitoring systems are difficult to place, and it may be more time-efficient to drill
separate holes for the points at different depths in a cluster. Monitoring point depth
selection is entirely site dependent, but monitoring of multiple depths within the vadose
zone is recommended. It may be appropriate to extend the monitoring point into the
water table to monitor water table fluctuations resulting from seasonal change or in
response to the SVE system or other remedial actions.
Screen and Casing Placement
Casing and screen are normally placed by methods seemlier to those used to install
SVE extraction wells; however, direct-push techniques are rapid alternatives for placing
monitoring points to the desired depths. Actual means of placement is dependent on
the system, materials used, and site geology.
Seal and Grout Placement
The procedures for sealing the well would generally be the same as those used for the
SVE wells. Points placed by direct-push methods may depend on a tight seal with
native soil to prevent leaks. Multiport monitoring systems require careful placement of
seals between the monitored intervals to prevent leakage of vapors between the
various target intervals.
Surface Completion
The monitoring points can be completed with a suitable barbed/valved sampling port
attached by threaded connection to an appropriate end cap. The cap is attached to the
top of the casing by an air-tight connection. The points can be set above grade with
suitable protection or below grade, typically in a flush-mount valve box. Refer as
appropriate to a drawing detail showing the desired surface completion.
Surveys
Horizontal coordinates are necessary for each point, and vertical coordinates to the
nearest 0.01 foot are necessary if water levels are monitored.
Groundwater Monitoring Well Construction
Consult CEGS 02671 for specification language and design guidance. ASTM D 5092,
Recommended Practice for Design and Installation of Groundwater Monitoring Wells in
Aquifers, is also an excellent reference for specifying monitoring well installation.
272
-------
Drill Cuttings Disposal
The specifications should address the means for disposing of the drill, cuttings and other
potentially contaminated waste. This depends on the project, but actual disposal can
be left to the contractor.
Well Decommissioning |
The specification should describe acceptable methods for sealing wells that are either
no longer needed near the end of the project or do not meet the specification. Individ-
ual states may have specific requirements for well decommissioning. These must be
followed. Normally, wells are sealed to prevent preferential migration of contaminants
via the well. Wells also can be physically removed (by drilling them out if made of
plastic, or by over excavation and pulling the casing and screen) and the hole grouted
Alternatively, the wells can be pressure grouted from bottom to top. If the well is to be
decommissioned because of a questionable grout seal, pressure grouting may not be
effective in preventing preferential contaminant migration. If the decommissioning is
required because of poor contractor performance, the decommissioning and replace-
ment of the well should be at the contractor's expense. The contractor must submit an
alert describing the decommissioning activities.
273
-------
Appendix B.2
Collection System Design Guidance
Introduction
This Appendix provides guidance on the specification, layout, and construction of the
manifold collection system piping and associated appurtenances. This guidance is not
comprehensive and must be adapted as necessary for site-specific conditions and
objectives.
Layout and Plans
A manifold system interconnects the injection or extraction wells into a single-flow
network prior to being connected to the remainder of the SVE/BV system. A manifold
system generally includes a series of flow-control valves, pressure and airflow meters
(liquid flow meters for liquid phase), and VOC sampling ports at each well head; these
devices may be grouped in one central location for convenience. A typical manifold
system is constructed of PVC, high-density polyethylene (HDPE), or stainless steel.
Diameter (SVE and BV)
Manifolds are generally constructed with 4-inch pipe. Although manifolds as large as
24 inches have been installed, these large system have centrifugal blowers that require
a low manifold vacuum. The designer should evaluate pipe friction in the system to
ascertain that the manifold will conduct the desired airflow rate under either of the
following conditions:
• If a 2-inch manifold pipe is used, the airflow rate is over 50 scfm, and any
piping run is longer than 50 feet.
• If a 4-inch manifold pipe is used, the airflow rate is over 300 scfm, and any
piping run is longer than 50 feet.
If the air velocity is lower than a few thousand feet per minute, the manifold may
accumulate condensation. Condensation buildup may be avoided by sloping the
manifold toward the air-extraction wells where it can drain. A second method with
buried manifolds is to use a relatively small-diameter vertical pipe where the direction
changes from horizontal to vertical, allowing the airstream to carry condensation toward
a condensate trap. See Section on Motors and Gages. A third alternative, but less
274
-------
satisfactory, is to use a smaller diameter pipe in order to maintain a high air velocity on
the entire manifold. This alternative is less satisfactory because pipe friction may be
excessive, resulting in added requirements for blower capacity and excessive electrical
costs.
Condensate Traps
A condensate trap is also called a water trap, separator tank, or demister. For SVE/BV
systems, condensate controls are often necessary to prevent unwanted liquids from
accumulating in piping, blowers, or air emission control devices. These controls
remove moisture and store the liquid prior to disposal. In general, a condensate trap
should be included in the design if the up-hole air velocity in the air-extraction wells is
greater than 1,000 feet per minute or if a rotary lobe blower is used.
Condensate trap configuration includes the following:
A vertical pipe, cap, and tee in a manifold that is capable of holding less
than 5 gallons
A large tank in line with the manifold
An engineered trap that uses a cyclone action to separate water droplets
from the air system.
Water that accumulates within these traps needs to be addressed. If a groundwater
extraction system is also used at the site, the condensate water can be added to the
pumped groundwater that is to be treated and/or disposed of. If no groundwater
extraction system is used at the site, the water must be properly disposed of.
Motors and Gages
Flow Meter (SVE/BV)
Regular or averaging pitot tubes are generally used to measure flow. Averaging pitot
tubes are designed to only require a single reading. In general, manufacturers recom-
mend installing these tubes ten or more straight unobstructed pipe diameters upstream
and five or more diameters downstream.
Flow Meter (Free Product Recovery)
A flow meter should be installed on the system to measure the amount of pumping from
each well. It should be a totalizing-flow meter that indicates the total fluid pumped.
275
-------
Vacuum (SVE/BV)
Vacuums should be measured with a manometer, a magnehelic gauge, or a vacuum
gauge. Most SVE systems operate at a low enough vacuum that the measurements
are read in inches of water column.
Temperature
Temperature is usually read with a bimetal dial-type thermometer that is installed
through a hole in the manifold pipe.
Relative Humidity or Dew Point
Relative humidify or dew point measurements are not required, but may be beneficial to
evaluate moisture content of biodegradation or carbon filters. A wet bulb thermometer
or digital meter is used to measure relative humidity or dew point.
Sample Taps
The sample tap design is specific to the sample container and the field procedure used
for collecting samples. It may have a septa fitting for direct syringe insertion, or it may
be as simple as a hose barb for a piece of plastic tubing. The sample ports may have
to be fabricated for the specific sampling device. For aboveground systems, the
sample ports and instrumentation for each well may be located near the well itself. On
buried manifold systems, the sample ports and instrumentation may be located near the
blower system where the manifold pipe exits from the ground.
Configuration
The designer should configure the manifold and place valves in such a way to allow
control and sample collection at each well. The option that places the instrumentation
nearest the well generally provides the best vacuum and temperature information for
the well, but is more likely to freeze up in winter on low-flow systems and systems with
a shallow water table.
Piping and Valves
Materials of Construction
Proper selection and specification of materials plays an important role in the success of
SVE/BV remediation and free product recovery.
Piping (SVE/BV)
Piping for SVE/BV systems generally includes vacuum lines, pressure lines, sampling
lines, and condensate lines. Off-gas treatment, such as catalytic or thermal oxidizers,
may also require fuel supply lines. The design of a piping system must consider the
following major issues: pressure limitations, temperature limitations, insulation, mechan-
ical considerations, pneumatics and hydraulics, and chemical compatibility.
276
-------
Pressure Limitations The design pressure must not exceed the maximum allowable
limits for the piping system minus a factor of safety (e.g., 50 percent). Per ANSI B31.3,
Section 301.2, pressure relief valves should be included where required. It should be
noted that PVC pipe is not appropriate for uses involving high pressures (many
atmospheres) because it cannot safely withstand the stresses that are imposed. It
should be noted that vacuums and pressures exerted during SVE/BV operations are
usually less than one atmosphere, however, and therefore fall well within the safe range
of operation under the provision of appropriate pressure/vacuum relief. When flexible*
hoses are used on the vacuum side of the system, vacuum limits may be far less than
pressure limits.
Temperature Limitations Plastic piping, such as PVC, chlorinated polyvinyl chloride
(CPVC), polypropylene (PPE), or polyvinylidene fluoride (PVDF), is commonly used for
SVE/BV systems. Temperature limitations of the material also must not beiexceeded.
Plastic piping should never be used on the blower discharge because if the blower
overheats, the piping may melt.
Winter Operations In cases where the project must operate all year, the manifold
system should be winterized. Self-regulating heat tape and/or pipe insulation should be
used for aboveground manifolds. Because these aboveground systems are not easily
winteri2:ed, they are usually insulated or installed near or below frost level. If heat tape
is to be used for winterizing, CPVC pipe should be used instead of PVC in order to
provide higher strength at high temperatures.
Mechanical Stress Supports for all piping should be designed and spaced in accor-
dance with ANSI/MSS SP-58, 69, 89, and 90. Supports shall have a nominal diameter
of at least 2 inches.
Pneumatics and Hydraulics The piping system must be sized to be compatible with the
overall pneumatic scheme. Friction losses and settling velocities should be;considered.
Velocities greater than 3 feet per second are recommended for pumped condensate
lines.
Chemical Compatibility A list of acceptable materials is provided in Table 126.1 of
ANSI B31.1. Specifically, chlorinated solvents may degrade plastic piping. Piping that
will be exposed to sunlight must be UV resistant or contain a UV protective coating.
Piping Manifold (Liquid Phase)
The manifold consists of the piping system used to move the pumped liquids to the
storage tanks and/or treatment system. It may be above ground, but in most cases it is
buried. These piping systems must be constructed of a material compatible with the
contaminants that are being pumped. The piping must also be capable of withstanding
the pressure and volume of the pumping system under worst-case scenarios. If the
designer utilizes a pneumatic-pumping system, the lines must be capable of holding the
277
-------
pressure of the regulated compressed air source. If the designer utilizes a submersible
pump, the lines should be able to hold the pump pressure if the flow is blocked at the
treatment location. Designers should use the working pressure rating, and not the burst
pressure rating, when assessing the pressure capability for manifold lines. Steel or
other materials should be used in lieu of PVC if heat tape is used.
Metallic Piping
Unlined steel is not recommended because of the potential for condensation and the
possibility of acid formation. The presence of entrained solids and ambient temperature
fluctuations may limit the use of metallic pipe with its sophisticated lining system.
Although copper is a suitable material for most installations, it is prohibitively expensive
in diameters above 250 mm (1 inch). Stainless steel may be a suitable material for
some installations. Stainless is available in a variety of grades, which are not equally
resistant to corrosion and other physical damage, and has a wide price range.
Double-Walled Pipe
In general, single-walled pipe is preferable to double-walled pipe. The performance of
single-walled pipe, joints, and coupling is superior to the performance of double-walled
pipe under a wide range of conditions. Regulations may require the use of double-
walled pipe.
Fittings
The manifold for plastic piping should be constructed with glued fittings because slip fit
joints may fail with time. A steel wire or similar material should be installed in the trench
along with buried manifolds containing plastic pipe so that a metal detector can detect
its location at a later date.
Valves and Meters
Most of the above considerations that apply to piping also apply to valves. Valves are
used in SVE/BV systems to regulate flow rate and on/off control. A typical SVE/BV
system will be equipped with a flow control valve on each extraction or injection line.
The valves must be chemically compatible with the liquid or air stream; they must
operate safely in the temperature and pressure range of the system; they must not
create excessive frictional loss when fully opened; and in some situations, they must be
insulated and/or heated to prevent condensation. In addition, the operating range of a
control valve must match the flow control requirements of the application. All control
valves must be sized properly. If a valve is sized too large, the valve will operate mostly
in the near-closed posjtion, and will give poor sensitivity and control action. If a valve is
sized too small, the upper range of the valve will limit flow. Formulas and sizing
procedures vary with valve manufacturer. During the layout of the system, the designer
should ensure that the valves are accessible. Valves should be numbered and tagged.
Several valves commonly used for SVE/BV are described below.
278
-------
Plug Valve A plug valve is primarily used for on-off service and throttling applications.
Flow is controlled by a plug with a hole in the center that rotates to align with the flow
path.
Ball Valve A ball valve is also used primarily for on-off control and throttling applica-
tions. Flow is controlled by a plug with a hole in the center that rotates to align with the
flow path.
Butterfly Valve Used for on-off and throttling applications, the butterfly valve controls
flow with a rotating disk or vane. This valve has relatively low friction loss in the fully-
open position.
Diaphragm Valve This multiturn valve is used to control flow in both clean and dirty
services. The diagram valve controls flow with a flexible diaphragm attached to a
compressor and valve stem.
Needle Valve This multiturn valve is used for precise flow control applications in clean
services, typically on small-diameter piping. Needle valves have high frictional losses in
the fully-open position.
Globe Valve A glove valve is used for on-off control and throttling applications. This
valve controls flow with a convex plug lowered onto a horizontal seat. Raising the plug
off the seat allows for fluids to flow through.
Dilution or Bleed Valve This valve is needed on the manifold immediately before air
enters the air filter or blower (if no filter is used). The dilution valve allows atmospheric
air into the blower, when opened, and relieves vacuum to reduce overall air-extraction
rates from the wells. These valves should not be installed between the wells and the
sample ports because the sample results would not represent extracted air concentra-
tions. A dilution valve is more efficient than a throttle valve. In addition to a dilution
valve, an automatic pressure relief valve should be installed if the blower has the
potential to overheat under a blocked flow condition.
General Installation Requirements
Abrasion
It is important to protect against abrasion that weakens the pipe in spots and may
interfere with joint construction. Flexible materials withstand internal abrasion better
than rigid materials.
Bending
It is also important to avoid bending in excess of the deflection recommended for the
pipe material.
279
-------
Pressure and Leakage Tests
Pressure and leakage tests should be performed after joints have cured and before
pipe is buried where possible.
Casings
Casings should be avoided where possible. Entrances and exits from casings are
potential shear points where the shear stresses may exceed the stress allowable for the
pipe.
Aboveground Piping
Support
Minimum support dimensions required for the pipe material and diameter shall be met.
Support shall be provided for valves and other heavy equipment independently of the
pipe.
Restraints
Guides, anchors, and restraints shall be provided within the allowances for the pipe
design. Provision shall be made for longitudinal expansion and contraction.
Protection
Protection from mechanical damage shall be provided.
Insulation
See section on "Winter Operations."
UV Protection
UV protection should be provided for any PVC pipe used.
Below Grade Piping
Loading Conditions
Loading conditions must be designed to avoid excessive point loads. Pipe passing
under or through walls is particularly susceptible.
Burial Depth
Adequate cover for freeze prevention shall be provided in severe climates and for
mechanical protection, in mild climates.
Surge and Thrust
Surge and thrust effects shall be evaluated and pipe restraints provided where neces-
sary.
280
-------
Specification Sources
ASTM D2447 Polyethylene (PE) Plastic Pipe, Schedules 40 and 80 Based on Outside
Diameter.
ASTM D2513 Thermoplastic Gas Pressure Pipe, Tubing, and Fittings.
ASTM D2517 Reinforced Epoxy Resin Gas Pressure Pipe and Fittings.
ASTM D3035 Polyethylene (PE) Plastic Pipe (SDR-PR) based on Controlled Outside
Diameter. !
ASTM D2241 Poly (Vinyl Chloride) (PVC) Plastic Pipe and Fittings (SDR-PR).
ASTM D2740 Poly (Vinyl Chloride) (PVC) Plastic Tubing.
Soil Vapor Extraction and Bioventing Engineer Manual, EM 1110-1-4001. In Press.
Guidance for Design, Installation and Operation of Soil Venting Systems, Wisconsin
Department of Natural Resources, July 1993, PUBL-SW185-93
Guidance for Design, Installation and Operation of Groundwater Extraction and Product
Recovery Systems, Wisconsin.
Department of Natural Resources. August 1993, PUBL-SW 183-93.
281
-------
Appendix B.3
Equipment Specifications
A. Blower/Vacuum Pumps
Introduction
Blower/vacuum pumps should be provided and installed as complete and totally
functional systems, along with all necessary ancillary equipment including a motor,
filtration system, silencers, controls, protective devices, instrumentation, and lubrication
system. Blower/vacuum pumps commonly used in SVE/BV systems include regenera-
tive blowers, rotary lobe blowers, liquid ring vacuum pumps, rotary vane blowers, and
centrifugal blowers. Although many blower/vacuum pumps could be used in SVE/BV
systems, the types listed are frequently encountered.
Regenerative Blower
Regenerative blowers consist of a multistage impeller that rotates in a stationary
housing. The multistage impeller creates pressure through centrifugal force. A unit of
air enters the impeller and fills the space between two of the rotating vanes. The air is
thrust outward toward the casing but then is tuned back to another area of the rotating
impeller. This process continues regenerating the pressure until the air reaches to the
outlet. Regenerative blowers are compact and produce an oil-free airflow.
Rotary Lobe Blower
Rotary lobe blowers consist of a pair of matched impellers rotating in a stationary
housing with inlet and outlet ports. The impellers rotate in opposite directions in the
housing and trap a volume of air at the inlet port and move it around the perimeter to
the outlet port. Rotation of the impellers is synchronized by timing gears keyed into the
shaft. Oil seals are required to avoid contaminating the air stream with lubricating oil.
These seals must be chemically compatible with site contaminants. When a belt drive
is employed, blower speed may be regulated by changing the diameter of one or both
sheaves or by using a variable-speed motor pulley.
Liquid Ring Vacuum Pump
A liquid ring vacuum pump transfers both liquid and gas through the pump casing.
Centrifugal force acting on the liquid within the pump causes the liquid to form a ring
around the inside of the casing. Gas is trapped between rotating blades and com-
pressed by the liquid ring as the gas is forced radially inward toward a central discharge
282
-------
port. After each revolution, the compressed gas and accompanying liquid are dis-
charged. Vacuum levels close to absolute vacuum (i.e., absolute pressure equals zero)
can be generated in this manner. These pumps generate a waste stream of liquid that
must be properly disposed of. The waste stream can be reduced by recycling the
liquid; however, a cooling system for the liquid stream may be required to avoid
overheating the pump.
/?ofa/y Vane Vacuum Pump
A rotary vane vacuum pump consists of multiple vanes in a rotor located on a driven
rotating shaft located inside an eccentric housing. The rotor and vanes rotate and the
volume between adjacent vanes decreases (compression) and increases (suction) as
the vanes sweep across the eccentric housing. The gas enters when the vanes sweep
past the inlet port, is compressed as the volume between the rotor and eccentric
housing decreases, and is discharged out the discharge port when the volume between
the rotor and eccentric housing is at a minimum. The cycle continues as each vane on
the rotor sweeps past the inlet and discharge ports. I
Centrifugal Vacuum Producer
A centrifugal vacuum producer, sometimes called a turbocompressor, belongs to a
family of turbo machines that includes fans, propellers, and turbines. These machines
continuously exchange angular momentum between a rotating mechanical element
(impeller) and a steadily flowing fluid. The suction flow enters the impeller in the axial
direction and discharges radially at high velocity. The change in diameter through the
impeller increases the velocity of the gas flow. The dynamic head is converted into
static head, or pressure, through a diffusion process that generally begins within the
impeller and ends in a radial diffuser and scroll outboard of the impeller Centrifugal
compressors can be single stage, with only one impeller, or can be multistage with two
or more impellers mounted in the same casing. In multistage compressors, the gas
discharged from the first stage is directed to the inlet of the second stage through a
return channel. Once the gas reaches the last stage, it discharges to a volute or
collector chamber and then passes out through the compressor discharge connection.
Accessories
Blower Particulate Filters and Demisters ;
Particulate filters are typically installed between the condensate removal system and
the blower inlet. Although the condensate removal system will decrease the concentra-
tion levels of airborne particulate, the removal efficiency may not be sufficient High
particulaite levels may cause operational problems with the blower, downstream piping
or off-gas treatment equipment. Particulate air filters should be employed to remove
airborne particles down to the 1- to 10-micron range. :
Cartridge air filters are often used in this type of application. Filter elements are
manufactured from a variety of elements including pleated paper, felt, or wire mesh.
283
-------
Paper elements are inexpensive and typically disposable. Felt and wire mesh filters
may be washed. The filter is selected based on airflow rate, desired removal efficiency,
and pressure drop.
Pressure gauges, or a single deferential pressure gauge, should be installed upstream
and downstream of the filter. Filters should be changed at the recommended pressure
difference across the filter.
Demisters are often installed downstream of vacuum units to reduce the entrained
oil/water mists in the vapor stream.
Blower Silencers and Acoustics
Depending on the size and the location of the SVE/BV system, inlet and outlet silencers
may be necessary to reduce blower noise. Blowers present two noise problems: (1)
pulsation within the piping system, and (2) noise radiation from the blower itself.
Pulsation noise peaks can be severe for large blowers and can result in noise dis-
charges in the high decibel range.
Silencers are selected based on flow capacities and noise attenuation properties.
These devices typically contain chambers with noise absorptive elements. Silencer
manufacturers should provide the designer with an attenuation curve, which is a plot of
noise attenuation (decibels) versus frequency (hertz). The objective is to obtain the
greatest noise reduction near the sound frequencies emitted by the blower.
Also, if the SVE/BV system is located within a building, shed, or trailer, the choice of
wall material should be selected taking into consideration acoustical properties.
Complete tables of absorption coefficients of various building materials vs. frequency
may be found in books on architectural acoustics.
Issues concerning hearing protection must be addressed in the site health and safety
plan. The 8-hour time weighted-average (TWA) sound level is 85 decibels. The TWA
represents an action level requiring that workers be provided with hearing protection.
Specification Sources
Antifriction Bearing Manufacturers Association (AFBMA)
AFBMA 9 (1990) Load Ratings and Fatigue Life for Ball Bearings
AFBMA 11 (1990) Load Ratings and Fatigue Life for Roller Bearings
American Society of Mechanical Engineers (ASME)
ASME B40.1 (1991) Gauges - Pressure Indicating Dial Type - Elastic Element
Submittais
Submittals must be limited to those necessary for adequate quality control. The
importance of an item in the project should be one of the primary factors in determining
284
-------
if a submittal for the item should be required. Submittals should include drawings
showing shop and erection details, including catalog cuts, connections, holes, bolts,
and welds; manufacturer's certificates attesting that the blower/vacuum pumps meet the
specified requirements; and manufacturer's operation and maintenance manuals
including startup and shutdown procedures.
Materials and Equipment
Materials and equipment shall conform to their respective reference publications and
specification requirements. Specifications for blower/vacuum pumps for SVE/BV
projects should address the following requirements.
Performance requirements: cfm and pressure
Air quality: oil free, filtration
Drive type: direct drive, belt drive, adjustability, VFD
Motors: starting requirements, speed control
Controls: automatic, manual
Duty cycle: continuous, intermittent, cycles per hour
Pump type: regenerative, rotary lobe, water ring, rotary vane, centrifugal
Construction details: materials, configuration, coatings
Bearing life: L-10 life as defined by AFBMA 9 or 11; 5 years suggested
Lubrication system: pressure, splash
Installation requirements
Field testing .
Equipment painting
O&M manuals
Field training
Accessories: filters, silencers
Pressure gauges: ASME B40.1, locations, range
Thermometers: locations, glass, or deal tepee range
Valve position indicators
Air volume indicator
Protective devices: bearing temperatures surge protection, vibration monitoring.
B. Electric Motors
Introduction :
Alternating Current (AC) motors, fractional and integral horsepower, 500-hp and
smaller, shall conform,to NEMA MG 1 and UL 1004 for motors; NEMA MG 10 for
energy management selection of polyphase motors; and UL 674 for use of motors in
hazardous (classified) locations. Installation shall be limited in accordance with NFPA
70. The horsepower rating of motors should be limited to no more than 125 percent of
the maximum load being served unless a NEMA standard size does not fall within this
range; otherwise, the next larger NEMA standard motor size should be used. Motors of
1 hp or more with open, drip-proof or totally-enclosed fan-cooled enclosures;should be
285
-------
high-efficiency type. Equipment and wiring shall conform to the class, division, group,
and temperature requirements in NFPA 70.
Specification Sources
National Electrical Manufacturers Association (NEMA)
NEMA MG 1 Motors and generators
NEMA MG 10 Energy Management Guide for Selection and Use of Polyphase Motors
NEMA ICS 1 Industrial Controls and Systems
NEMA iCS 2 Industrial Control Devices, Controllers and Assemblies
NEMA ICS 3 Industrial Systems
NEMA ICS 6 Enclosures for Industrial Control and Systems
National Fire Protection Association (NFPA)
NFPA 70 National Electrical Code
Underwriters Laboratories (UL)
UL 508 Industrial Control Equipment
UL 674 Electric Motors and Generators for Use in Hazardous (Classified) Locations
UL 845 Motor Control Centers
UL 1004 Electric Motors
Submittals
Submittals must be limited to those necessary for adequate quality control. The
importance of an item in the project should be one of the primary factors in determining
if a submittal for the item should be required. Submittals should include drawings
showing shop and erection details, including catalog cuts, connections, holes, bolts,
, and welds; manufacturer's certificates attesting that the motors meet the specified
requirements; and manufacturer's operation and maintenance manuals including
startup and shutdown procedures.
Materials and Equipment
Materials and equipment shall conform to their respective reference publications and
specification requirements. Specifications for electric motors for SVE/BV projects
should address the following requirements.
Motor requirements: efficiency, voltage, phase frame, duty temperature reference,
starting characteristics
Motor control requirements: manual, automatic
Reduced-Voltage controllers: auto transformer, reactor, resistor, wye-delta, part
winding
Motor control centers: class, type
Contacts
Safety controls
Motor disconnect means
286
-------
Equipment connections: flexible conduits, liquid tight conduits
Testing.
C. Generator Sets
Introduction
Engine-generator sets should be provided and installed as complete and totally
functional systems, with all necessary ancillary equipment to include air filtration;
starting system; generator controls, protection, and isolation; instrumentation; lubrica-
tion; fuel system; cooling system; and engine exhaust system. ;
Specification Sources
American Society for Testing and Materials (ASTM) !
ASTM 0975(1991) Diesel Fuel Oils
National Fire Protection Association (NFPA)
NFPA 30 (1990) Flammable and Combustible Liquids
NFPA 37 (1990) Installation and Use of Stationary Combustion Engines and Gas
Turbines
Submittals
Submittals must be limited to those necessary for adequate quality control. The
importance of an item in the project should be one of the primary factors in determining
if a submittal for the item should be required. Submittals should include drawings
showing shop and erection details, including catalog cuts, connections, holes, bolts
and welds; manufacturer's certificates attesting that the generator sets meet the
specified requirements; and manufacturers operation and maintenance manuals
including startup and shutdown procedures.
Materials and Equipment
Materials and equipment shall conform to their respective reference publications and
specification requirements. Specifications for generator sets for SVE/BV projects
should address the following suggested items:
Power application: prime, standby
Engine-Generator application: parallel, stand-alone
Engine cooling type: water, air, remote, integral
Governor type: hydraulic, electric-hydraulic, mechanical, electronic-hydraulic
Governor application: Isochronous, droop
Maximum speed: rpm
Frequency: 59, 60 Hz
Voltage
Phases: 3-phase, WYE - 3-phase, delta - single phase
Phase rotation: ABC, ACB
287
-------
Service load: VA, kW
Power factor: 0.8 lagging
Nonlinear loads: kVA
Overload capacity: percent of service load for number of consecutive hours
Motor starting kVA
Max step load increase: percent of service load
Transient recovery time: seconds
with step load increase (voltage)
Transient recovery time: seconds
with step load increase (frequency)
Maximum voltage deviation: percent of rated voltage
with step load increase
Maximum frequency deviation: percent of rated frequency
with step load increase
Max step load decrease: percent of service load
Transient recovery time: seconds
Step load decrease (voltage)
Transient recovery time: seconds
Step load decrease (frequency)
Maximum voltage deviation: percent of rated voltage
with step load decrease
Maximum frequency deviation with step load decrease: percent of rated frequency
Max time To start and assume load: seconds
Max summer indoor temp
Min winter indoor temp
Seismic zone: 1,2,3,4
Installation elevation: above sea level
Max summer outdoor temp
Min winter outdoor temp
Engine generator set enclosure
Vibration limitation
Fuel system: filter, day tank, NFPA 30 and 37, No. 2-D diesel - ASTM D 975.
Lubrication: NFPA 30 and 37
Air intake equipment: filters and silencers
Exhaust system
Emissions: local and federal regulations
Starting system: air, battery, starting aids
Safety system: alarms and action logic
Generator: 1 or 2 bearing
Exciter: type
Voltage regulator
Generator control and protection
Generator and synchronizing panels
Factory inspection
288
-------
Field testing
O&M manuals
Field training.
D. Prime Movers
Introduction
Prime movers should be provided and installed in a complete and totally functional
system, with all necessary ancillary equipment including air filtration, starting system,
controls, protection, instrumentation, lubrication, fuel system, cooling system, and
engine exhaust system.
Specification Sources >
American Society for Testing and Materials (ASTM)
ASTMD975 (1991) Diesel Fuel Oils
National Fire Protection Association
NFPA30 (1990) Flammable and Combustible Liquids
NFPA 37 (1990) Installation and Use of Stationary Combustion Engines and Gas
Turbinejs :
Submittals
Submittals must be limited to those necessary for adequate quality control. The
importance of an item in the project should be one of the primary factors in determining
if a submittal for the item should be required. Submittals should include drawings
showing shop and erection details, including catalog cuts, connections, holes bolts
and welds; manufacturer's certificates attesting that the prime movers meet the
specified requirements; and manufacturer's operation and maintenance manuals
including startup and shutdown procedures.
Materials and Equipment
Materials and equipment shall conform to their respective reference publications and
specification requirements. Specifications for prime movers for SVE/BV projects should
address the following suggested items:
Application: describe usage
Engine cooling type: water, air, remote, integral
Governor type: hydraulic, electric-hydraulic, mechanical, electronic-hydraulic
Maximum speed: rpm
Power: hp
Torque: Ft-lb at rpm
Overload capacity: percent of power for number of consecutive hours
Max time to start and assume load: seconds
Max summer indoor temp
289
-------
Min winter indoor temp
Seismic zone: 1,2,3,4
Installation elevation: above sea level
Max summer outdoor temp
Min winter outdoor temp
Engine generator set enclosure
Vibration limitation
Fuel system: filter, day tank, NFPA 30 and 37, No. 2-D diesel - ASTM D 975
Lubrication: NFPA 30 and 37
Air intake equipment: filters and silencers
Exhaust system
Emissions: local and federal regulations
Starting system: air, battery, starting aids
Safety system: alarms and action logic
Factory inspection
Field testing
Equipment painting
O&M manuals
Field training.
290
-------
Appendix B.4
instrumentation and Control
Introduction
In the design of an SVE/BV system, a good deal of attention must be paid to the
instrumentation and control system. A good instrumentation and control system design
will assure that the individual components are coordinated and operate effectively.
Description of Design Elements
An SVE/BV design will include, at a minimum, the following elements:
P&l Diagrams
Process and instrumentation (P&l) diagrams show the interrelationship between
/^foml A°O P°nentS> Plping and Process control devices. ISA and ANSI standards
(ANSI/ISA-S5.1) govern the preparation of P&l diagrams. These diagrams show all
major process components organized according to process flow. The instrumentation
symbols are shown in "bubbles."
i
Elementary Wiring Diagram
This diagram shows the wiring of all physical electrical devices, such as transformers
motors, and lights. If appropriate, the diagram is organized in ladder logic form.
Description of Components • ;
The specifications must include a description of instrumentation and contror compo-
nents including installation and mounting requirements.
Sequence of Control
The sequence of control must be included in the design submittal and the operation and
maintenance manual Control information concerning system start-up, system shut-
down, and response to malfunctions must be included.
Control Panel Layout,
A control panel layout must be designed. This drawing will show, to scale all electrical
components and associated wiring. Depending on the project, this control item may be
submitted as a shop drawing by the instrumentation and control contractor
291
-------
Logic Diagram
A logic diagram must be included if the process control logic is not apparent from the
P&I Diagram. This diagram shows the logical (and, or, nor, if-then) relationships
between control components, but does not show interconnecting process flow. For
example, the diagram may show that if Switch No. 2 is placed in the on position and
there are no alarm conditions, then the blower will turn on and energize a green
indicator light.
Legend
The set of documents must have a legend to explain the symbols used. Despite the
existence of the legend, standard symbols must be used wherever applicable.
Degrees of Automation
The degree of automation is generally dependent on the complexity of the treatment
system, remoteness of the site, and monitoring and control requirements. Typically,
there is a tradeoff between the initial capital cost of the instrumentation and control
equipment, and the labor cost savings in system operation.
Generally, there are the following three forms of process control: local control, central-
ized control, and remote control. In a local control system, all control elements (i.e.,
indicators, switches, relays, motor starters) are located next to the associated equip-
ment. In a centralized control system, the control elements are mounted in a single
location. These systems may include a hard-wired control panel, a programmable logic
controller (PLC), or a computer. Remote control can be accomplished several ways,
including by means of modems or radio telemetry.
To select the appropriate control scheme, the advantages and disadvantages each
control scheme must be considered. A localized control system is less complex, less
expensive, and easier to construct. For example, if a level switch in a tank is controlling
an adjacent discharge pump, it would obviously be simpler to wire from the tank directly
to the adjacent pump rather than to wire from the tank to the centralized control panel
and then from the panel back to the pump. As the control system becomes more
complex, it quickly becomes advantageous to locate the control components in a
central location. Centralized control systems are also easier to operate. Centralized
data acquisition and control may include the use of computers or programmable logic
controllers (PLC). Automated process control is a complex topic that is beyond the
scope of this document; however, several points are worth considering. The greater
the number of control jnputs, the more worthwhile it is to use a computer or PLC
control. For SVE/BV systems, the inputs may include signals from level indicators,
pressure switches, or thermocouples. The threshold for using PLCs or computers is
generally between five and ten inputs, depending on the type of input and operator
background. Often plant operators will be more familiar with traditional hard-wired
control logic than with control logic contained in software. Process logic contained in
software, however, is easier to change (once you learn the software) than hard wiring.
292
-------
Hard wiring of the process logic should be avoided if extensive future modifications to
the proposed system are anticipated. . ' '
Modems and radio telemetry can be used to control these systems remotely. Radio
telemetry is typically used over shorter distances when radio transmission Is possible.
Modems are used with computerized control systems. Systems can also be equipped
with auto dials to alert the operator of a malfunction by telephone or pager. Again,
considerations such as site location, capital cost, standardization, operator background,
and system complexity govern the selection of these devices.
Minimum Acceptable Process Control Components
For SVE/BV systems, the four major operational"parameters that require control are:
Liquid Collection ;
The condensate collection system accumulates liquid that may overflow. Liquid level
indicators, switches, and alarms are required.
Pressure/Vacuum
Blowers may require vacuum breaking controls to protect the motor units, the system
may also require pressure-relief valves to protect tanks or vessels. :
Flow Rate •
Flow rate monitoring is essential to judge the progress of the SVE remediation effort
and flow control is required to balance multiwell systems. Flow rate can be determined
by orifice plates, flow nozzles, venturi tubes, annular pitot tubes, turbine meters vortex
shedding flow meters, and acoustic flow meters. !•
Temperature ;
Temperature control may be necessary to (1) prevent motor overload on the pumps and
blower, (2) prevent carbon bed fires, (3) safely operate catalytic or thermal oxidation
systems, and (4) protect piping from thermal stresses and melting.
At a minimum, the following process control components are required:
«; Pressure and flow indicators for each well
• Blower motor thermal overload protection
• Run time for indicating the total hours of blower/vacuum pump operation
• Vacuum-relief valve and/or vacuum switch to effect blower shutdown
• Sampling ports before and after air treatment and at each well head
• Pressure indicators at blower inlet and outlet
• High-level switch/alarm for condensate collection system
• Explosimeter for sites with recently-measured LEL levels greater than
10 percent
• UL-listed burner controls for catalytic and thermal oxidizers for SVE systems.
293
-------
Special Instrumentation
Several specific instruments are common to SVE/BV systems. These instruments
include piezometers, LEL meters, organic vapor analyzers, and process gas chromat-
ographs.
Piezometers
Vacuum levels should be monitored at individual wells or at the treatment system.
Pressure transmitters and data loggers can be used.
Explosimeter
Used on sites where high-VOC levels cause a potential explosion hazard, these meters
must be equipped with relays to automatically shut off process component or dilute the
air stream with ambient air. Most explosimeter probes use catalytic combustion as part
of the detection process.
Organic Vapor Analyzers
Used to monitor vapor-phase discharges, these units are typically equipped with flame
ionization (FID), photoionization, thermal conductivity, or infrared detectors. Process
units (as opposed to the hand-held units frequently used in environmental work) can be
rack- or panel-mounted and equipped with control relays.
Process Gas Chromatography (GC)
SVE/BV systems can use GC-FID for on-site monitoring and control (on-line GC and/or
portable field GC). Several vendors manufacture GCs that can be automated for
process monitoring and control; however, laboratory facilities (to prepare standards,
etc.) and trained chemists are also required for GC monitoring.
Specification Sources
American National Standards Institute (ANSI)
ANSI/ISA-S5.I
Code of Federal Regulations (CFR)
CFR 47 PART 15 Radio Frequency Devices
CFR 47 PART 68 Connection of Terminal Equipment to Telephone Network
National Electrical Manufacturers Association (NEMA)
NEMAICS 1 Industrial Control and Systems
Submittals ,
Submittals must be limited to those necessary for adequate quality control. The
importance of an item in the project should be one of the primary factors in determining
if a submittal for the item should be required. Submittals should include equipment
data, system descriptions, system drawings, commissioning procedures, hardware
testing manuals, software manuals, operator's manual; manufacturer's certificates
attesting that the instrumentation and controls meet the specified requirements; and
294
-------
manufacturer's operation and maintenance manuals including startup and shutdown
procedures. .
Materials and Equipment
Materials and equipment shall conform to their respective reference publications and
specification requirements. The instrumentation and control system shall be a complete
system suitable for the process. Specifications for instrumentation and control for
SVE/BV projects should address the following suggested items:
Power-line surge protection
Power-line conditioners
System reliability ;
System accuracy
Field hardware
Instrumentation
Control devices (electric solenoid-operated pneumatic control valves)
Electronic devices
Standardization of signals
Temperature limits
Control panel software
Parameter definition
I/O point database definition to include:
Name
Device or sensor type (i.e., sensor, control, motors)
Point identification number
Area
Sensor range
Controller range
Sensor span
Controller span
Engineering units conversion (scale factor)
High and low reasonableness value (analog)
High and low alarm limit (analog)
High and low alarm limit differential (return to normal)
Analog change differential (for reporting)
High accumulator limit (pulse)
Status description (digital inputs)
Wire and cable , ;
Installation criteria
Contractor responsibilities
Control sequences of operation
Factory testing
Site testing
Performance verification testing
295
-------
Endurance testing
System calibration
Commissioning procedures
O&M manuals
Hardware manual
Software manual
Training
Maintenance and service
Emergency service.
296
-------
Glossary
adsorption: The attraction and adhesion of ions from an aqueous solution to the
surface of solids.
air sparging: The process of injection of air below the water table to strip volatile
contaminants from the saturated zone.
anisotropy:^ The conditions under which one or more of the hydraulic properties
aquifer vary with direction. H«'"«*
aquifer: A geologic formation, group of formations, or part of a formation that contains
saturated permeable material that yields sufficient, economical quantities of
ground water.
aquifer test. A test to determine hydraulic properties of an aquifer, involving the
withdrawal or injection of measured quantities of water from or to a well and the
measurement of resulting changes in hydraulic head in the aquifer.
aquitard: A semipervious geologic formation that can store water but transmits water at
a very low rate compared to the aquifer.
biodegradation: A subset of biotransformation, it is the biologically mediated
conversion of a compound to more simple products.
bioventing: A process by which air is injected into the subsurface to stimulate
biodegradation by microbes.
bubbling pressure: The pressure at which air enters the saturated zone (also known a«=-
air entry value-or threshold pressure).
bulk density: The mass of a soil per unit bulk volume of soil; the mass is measured
after all water has been extracted, and the volume includes the volume of the
soil itself and the pore volume. |
297
-------
capillary forces: Interfacial forces between immiscible fluid phases, resulting in
pressure differences between the two phases.
capillary fringe: The zone immediately above the water table within which the water is
drawn by capillary forces (fluid is under tension). The capillary fringe is saturated
and it is considered to be part of the unsaturated zone.
concentration gradient: The change in concentration with distance across a fluid
medium.
cone of depression: A depression in the groundwater table (or potentiometric surface)
that has the shape of an inverted cone and develops around a vertical discharge
well.
confined aquifer: An aquifer bounded above and below by confining layers of distinctly
lower permeability than the aquifer material and the one containing confined
groundwater. When a well is installed in a confined aquifer, the water level in the
well rises above the top of the aquifer.
conservative solute: A nonreactive constituent that does not undergo chemical reaction
during substance migration.
cosolvency: The interaction of one or more organic contaminants that may cause them
to behave differently in the subsurface than if they were present alone in their
pure form.
Darcy's law: An empirically derived equation for the flow of fluids through porous
media. It is based on the assumptions that flow is laminar and inertia can be
neglected, and states that the specific discharge, q, is directly proportional to the
hydraulic conductivity, K, and the hydraulic gradient, J.
dispersion: The spreading and mixing of chemical constituents in groundwater caused
by diffusion and mixing due to microscopic variations in velocities within and
between pores.
distribution (partitioning) coefficient: Relates the quantity of a solute sorbed per unit
weight of the solid phase and the quantity of the solute dissolved in water per
unit volume ofwater.
DNAPL: Dense Nonaqueous Phase Liquid. A liquid consisting of a solution of organic
compounds (e.g., chlorinated hydrocarbons) and which is denser than water.
298
-------
drawdown: A lowering of the water table of an unconfined aquifer or the potentiometric
surface of a confined aquifer caused by pumping of groundwater from wells. The
vertical distance between the original water level and the new water level.
effective porosity: The interconnected pore space through which fluids can pass,
expressed as a percent of bulk volume. Part of the total porosity will be occupied
by static fluid being held to mineral surface by surface tension, so effective
porosity will be less than total porosity. ;
effective grain size: The grain size corresponding to the 10% fines by weight on the
grain-size distribution curve.
extraction well: A discharge well used to remove groundwater or air.
Pick's law: The mass flux due to the molecular diffusion is proportional to the
concentration gradient and the diffusion coefficient.
gravitational water: Water that moves into, through, or out of a soil or rock mass under
the influence of gravity.
groundwater: The water contained in interconnected pores below the water table in an
unconfined aquifer or in a confined aquifer.
Henry's Law: The relationship between the partial pressure of a compound and its
equilibrium concentration in a dilute aqueous solution through a constant of
proportionality know as the Henry's Law Constant.
heterogeneity: Characteristic of a medium in which material properties vary from point
to point.
homogeneity: Characteristic of a medium in which material properties are
identical throughout. Though heterogeneity or nonuniformity is the
characteristic of most aquifers, assumed homogeneity, with some other
additional assumptions, allows use of analytical models as a valuable tool for
approximate analyses of groundwater movement.
hydraulic conductivity (K): Proportionality constant relating hydraulic gradient to specific
discharge, which for an isotropic medium and homogeneous fluid, equals the
volume of water at the existing kinematic viscosity that will move in unit time
under a unit hydraulic gradient through a unit area measured at right angles to
the direction of flow. The rate of flow of water in gallons per day through a cross
section of one square foot under a unit hydraulic gradient, at the prevailing
temperature (gpd/ft2). In the standard International System, the units are
m3/day/m2 or m/day. A coefficient of proportionality describing the rate at which
299
-------
water can move through a permeable medium. The density and kinematic
viscosity of the water must be considered in determining hydraulic conductivity.
hydraulic conductivity, effective: Rate of water flow through a porous medium that
contains more than one fluid (such as water and air in the unsaturated zone),
which should be specified in terms of the fluid type, content, and the existing
pressure.
hydraulic gradient (J): Slope of a water table or potentiometric surface. More
specifically, change in the hydraulic head per unit of distance in the direction of
the maximum rate of decrease. The difference in hydraulic heads (hrh2), divided
by the distance (L) along the flowpath: J = (hrh2)/L.
hydraulic head (h): Height above a datum plane (such as mean sea level) of the
column of water that can be supported by the hydraulic pressure at a given point
in a groundwater system. Equal to the distance between the water level in a well
and the datum plane.
ideal gas: A gas whose pressure-volume-temperature (P-V-T) behavior can be
described completely by the ideal gas law, PV = nRT, where n is the number of
moles of gas and R is the universal gas constant.
immiscible: The chemical property where two or more liquids or phases do not readily
dissolve in one another, such as soil and water.
intrinsic permeability: Pertaining to the relative ease with which a porous medium can
transmit a liquid under a hydraulic or potential gradient. It is a property of the
porous medium and is independent of the nature of the liquid or the potential
field.
isotropy: The condition in which the properties of interest (generally hydraulic
properties of the aquifer) are the same in all directions.
kinematic viscosity: The ratio of dynamic viscosity to mass density. It is obtained by
dividing dynamic viscosity by the fluid density. Units of kinematic viscosity are
square meters per second (m2/s).
Klinkenberg effect: Gas slippage along pore walls. Darcy's Law assumes that the
velocity of a fluid at the pore wall surface is zero.
laminar flow: Fluid flow in which the head loss is proportional to the first power of the
velocity; synonymous with streamline flow and viscous flow. Type of flow in
which the fluid particles follow paths that are smooth, straight, and parallel to the
300
-------
channel walls. In laminar flow, the viscosity of the fluid dampens out turbulent
motion. •
leaky aquifer: An artesian or water table aquifer that loses or gains water through
adjacent semipermeable confining units.
LNAPL: Lighter-than-water nonaqueous phase liquid.
molecular diffusion: Process in which solutes are transported at the microscopic level
due to variations in the solute concentrations within the fluid phases.
NAPL: Nonaqueous phase liquids.
organic carbon content: The amount of the organic carbon present in a soil Organic
chemicals in soil adsorb to soil organic carbon, and the amount of adsorption can
be related to the soil organic carbon content.
r
i
partitioning: Chemical equilibrium condition where a chemical's concentration is
apportioned between two different phases according to the partition coefficient
which is the ratio of a chemical's concentration in one phase to its concentration
in the other phase.
perched aquifer: A special case of phreatic aquifer that occurs wherever an impervious
(or semipervious) layer of limited areal extent is located between the water table
of a phreatic aquifer and the ground surface.
/- i
permeability: Ability of a porous medium to transmit fluids under a hydraulic gradient
The capacity of a porous rock, sediment, or soil to transmit a fluid; it is a
measure of the relative ease of fluid flow under unequal pressure.
Permeability coefficient: Rate of flow of water through a unit cross-sectional-area under
a unit hydraulic gradient at the prevailing temperature (field permeability
coefficient), or adjusted to 15°C.
permeability, effective: Observed permeability of a porous medium to one fluid phase,
under conditions of physical interaction between the phase and other fluid
phases present.
permeability, intrinsic: Relative ease with which a porous medium can transmit a fluid
under a potential gradient, as a property of the medium itself. Property of a
medium expressing the relative ease with which fluids can pass through it.
301
-------
porosity: Ratio of the total volume of voids to the total volume of a porous medium.
The percentage of the bulk volume of a rock or soil that is occupied by
interstices, whether isolated or connected. Porosity may be primary (formed
during deposition or cementation of the material), or secondary (formed after
deposition or cementation), such as fractures.
pressure head: Hydrostatic pressure expressed as the height (above a measurement
point) of a column of water that the pressure can support.
pumping test: A test that is conducted to determine aquifer or well characteristics. A
test made by pumping a well for a period of time and observing the change in
hydraulic head in the aquifer. A pumping test may be used to determine the
capacity of the well and the hydraulic characteristics of the aquifer. Also-called
aquifer test.
radial flow: The flow of water in an aquifer toward a vertical well.
radius of influence: The radial distance from the center of a wellbore to the point where
there is no lowering of the water table or potentiometric surface (the edge of its
cone of depression). The radial distance from an extraction well that has
adequate air flow for effective removal of contaminants when a vacuum is
applied to the extraction well.
Raoult's Law: A physical law that describes the relationship between the vapor
pressure of a component over a solution, the vapor pressure of the same
component over pure liquid, and the mole fraction of the component in the
solution.
residual saturation: Saturation below which fluid drainage will not occur.
retardation: The movement of a solute through a geologic medium at a velocity less
than that of the flowing groundwater due to sorption or other removal of the
solute.
saturation: The ratio of the volume of a single fluid in the pores to pore volume
expressed as a percentage or a fraction.
saturated zone: Portion of the subsurface environment in which all voids are ideally
filled with water under pressure greater than atmospheric. The zone in which the
voids in the rock or soil are filled with water at a pressure greater than
atmospheric. The water table is the top of the saturated zone in an unconfmed
aquifer.
302
-------
slug test: A test for estimating hydraulic conductivity of an aquifer in which a rapid
water-level change is produced in a piezometer or monitoring well usually by
introducing or withdrawing a "slug" of water or a weight. The rise or decline in
the water level is monitored. i
sorption: Processes that remove solutes from the fluid phase and concentrate them on
the solid phase of a medium; used to encompass absorption and adsorption.
specific surface: The amount of surface area of a dispersed system per gram or per
unit volume of the dispersed phase.
storativity: A dimensionless term representing the volume of water an aquifer releases
from or takes into storage per unit surface area of the aquifer per unit change in
head. It is equal to the product of specific storage and aquifer thickness In an
unconfined aquifer, the storativity is equivalent to the specific yield.
transmissivity: Rate at which water of the prevailing kinematic viscosity is transmitted
through a unit width of the aquifer under a unit hydraulic gradient. It is equal to
an integration of the hydraulic conductivities across the saturated part of the
aquifer perpendicular to the flow paths. The rate at which water is transmitted
through a unit width of an aquifer under a unit hydraulic gradient. Transmissivity
values are given in gallons per minute through a vertical section of an aquifer
one foot wide and extending the full saturated height of an aquifer under a
hydraulic gradient of one in the English Engineering System; in the Standard
International System, transmissivity is given in cubic meters per day through a
vertical section of an aquifer one meter wide and extending the full saturated
height of an aquifer under a hydraulic gradient of one. It is a function of
properties of the liquid, the porous media, and the thickness of the porous
media.
unconfined: Conditions in which the upper surface of the zone of saturation forms a
water table under atmospheric pressure. i
unsaturated flow: Movement of water in a porous medium in which the pore spaces ar<=»
riot completely filled with water.
unsaturated zone: The zone between the land surface and the water table It includes
the root zone, Intermediate zone, and capillary fringe. The pore spaces contain
water, as well as air and other gases at less than atmospheric pressure.
Saturated bodies, such as perched groundwater, may exist in the unsaturated
zone, and water pressure within these may be greater than atmospheric. Also
known as "v/arins^ rnno "
known as "vadose zone."
303
-------
viscosity: The internal friction within a fluid that causes it to resist flow. See Kinematic
and Dynamic viscosity.
VOC: Volatile organic contaminants, typically with a high vapor pressure and a
tendency to evaporate rapidly.
volatilization: The transfer of a chemical from liquid to the gas phase. Solubility,
molecular weight, vapor pressure of the liquid, and the nature of the air-liquid
interface affect the rate of volatilization.
water content: Ratio of the volume of water to the bulk volume in the unsaturated flow.
water table: Upper surface of a zone of saturation, where that surface is not formed by
a confining unit; water pressure in the porous medium is equal to atmospheric
pressure. The surface between the vadose zone and the groundwater; that
surface of a body of unconfined groundwater at which the pressure is equal to
that of the atmosphere.
304
-------
References
American Petroleum Institute. 1989. A guide to the assessment and remediation of
underground petroleum releases, API Publication No. 1628, 2nd ed 81 pp Wash-
ington, D.C. '
American Petroleum Institute. 1993. A guide to the assessment and remediation of
petroleum hydrocarbons in soil, API Publication No. 1629, Washington, D.0.
American Society for Testing and Materials. 1994. Emergency standard guide for risk-
based corrective action applied at petroleum release sites, ASTM Designation- ES 38-
94, July, Philadelphia, PA.
Angell, K.G. 1992. In Situ Remedial Methods: Air Sparging. The National Environ-
mental Journal (January/February), pp. 20-23.
Ayra, L M. and J. F. Paris. 1981. A physicoempirical model to predict the soil moisture
characteristics from particle size distribution and bulk density data. Soil Sci Soc Am
J. 45:1023-1030.
Baker, O.M., and W. Swerdloff. 1956. Finding surface tension of hydrocarbon liquids
Oil and Gas Journal, 125 (January 2).
Bear, J. 1972. Dynamics of fluids in porous media, American Elsevier Publishinq
Company, Inc.
Bear, J. 1979. Hydraulics of groundwater, McGraw-Hill.
Benson, D. A., D. Huntley, and P. C. Johnson. 1993. Modeling vapor extraction and
general transport in the presence of NAPL mixtures and nonideal conditions National
Groundwater Association, Dublin, Ohio, 31 (3).
Buscheck, T. E., and T. R. Peargin. 1991. Summary of a nation-wide vapor extraction
system performance study, Proceedings of Petroleum Hydrocarbons and Organic
Chemicals in Ground Water, National Well Association, Houston, Texas, November
305
-------
Chevron Research and Technology Company (CRTC). 1991. Vapor extraction system
performance study. Chevron Research and Technology Company internal document
written for Chevron USA marketing department.
Cohen, R. M., and J. W. Mercer. 1993. DNAPL site evaluation. CRC Press, Boca
Raton, Florida.
The College of Engineering, University of Wisconsin-Madison. 1993. Passive bioreme-
diation; case history. Underground Tank Technology Update, 7 (6).
Conner, J. R. 1988. Case study of soil venting. Pollution Engineering, July.
Davis, A., et al. 1994. Attenuation and biodegradation of chlorophenols in ground
water at a former wood treating facility. Ground Water, 32 (2).
Davis, J. W., N. J. Klier, and C. L. Carpenter. 1994. Natural biological attenuation of
benzene in ground water beneath a manufacturing facility. Ground Water, 32 (2).
Driscoll, F. G. 1986. Groundwater and wells. Johnson Filtration Systems, Inc., St.
Paul, Minnesota.
Dupont,-R. 1993. Fundamentals of bioventing applied to fuel contaminated sites.
Environmental Progress, 12(1 ):45-53.
Environmental Systems and Technologies Software (ES&T), Ltd. 1988-1993.
VENTING User's Guide, ES&T, Inc., Blackburg, Virginia.
ES&T, ARMOS - A Program to Estimate Unsaturated Soil Hydraulic Properties From
Particle Size Distribution Data, Vers. 2.0 User's Guide, Environmental Systems &
Technologies, Inc., Blacksburg, Virginia. 1990.
Everett, L. G., L. G., Wilson, and E. W. Hoylamn. 1984. Vadose zone monitoring for
hazardous waste sites. Noyes Data Corporation, Park Ridge, New Jersey.
Farr, A. M., R. J. Houghtalen, and D. B. McWorter. 1990. Volume estimation of light
nonaqueous phase liquids in porous media. Groundwater28:48-56.
Fetter, C. W. 1993. Contaminant Hydrogeology. MacMillan Publishing Company, New
York.
Fetter, C. W. 1994. Applied Hydrogeology. MacMillan Publishing Company, New
York.'
306
-------
Graves, D. A., and M. E. Leavitt. 1993. Vapor-phase nutrient delivery system for in situ
biorernediation of soil. U.S. Patent No. 5,178,491.
Ground Water, Vol. 28, pp. 244-252. 1990. Kemblowski, M. W. and C. Y; Chiang.
Hydrocarbon thickness fluctuations in monitoring wells.
Ground Water, 1996, Ahlfeld, D., A. Dahmani, M. Farrell, and W. Jj. Detailed Field
Measurements of an Air Sparging Pilot Test. Environmental Research Institute, U.
Conn., Draft Manuscript Submitted to Groundwater.
Ground Water Monitoring and Remediation 13 (4): 115-126. 1993. Ji, W., A. Dahmani,
D. Ahlfeld, J. Lin, and E. Hill. Laboratory study of air sparging: air flow visualization.
Johnson, P. C. 1992a. Soil vapor extraction workshop figures. Shell Development
Corp. (personal communication).
Hampton, D. R., T. R. Barrett, H. S. Nayyar, and T. P. O'Connell. 1993. "Hydrophobia
gravel packs for product monitoring and recovery wells." In Proceedings of the Seventh
National Outdoor Action Conference, May 25-27, 1993, National Groundwater
Association, pp. 581-595.
Hazmat World. 1994. Air sparging success 7 (5):25.
Hinchee, R. E. 1990. "Soil venting techniques." Paper presented at Annual Chevron
USA Inc., Marketing Environmental Engineering Conference, Denver, CO.
Hinchee, R. E., and M. Arthur. 1991. Bench scale studies of the soil aeratipn process
for bioremediation of petroleum hydrocarbons. Applied Biochemistry and Biotechnology
28/29:901-906.
Hinchee, R. E., and S. K. Ong. 1992. A rapid in situ respiration test for measuring
aerobic biodegradation rates of hydrocarbons in soil. J. Air Waste Management Assoc
42:1305-1312. '
Hinchee, R. E., etal. 1992. Test plan and technical protocol for a field treatability test
for bioventing. Air Force Center for Environmental Excellence, Brooks Air Force Base
San Antonio, Texas.
Hoag, G. 1991. The U. Conn/U.S. EPA draft site remediation course: soil vapor
extraction (in review).
Johnson, P. C., and R. A. Ettinger. 1991. Heuristic model for predicting the intrusion-
rate of contaminant vapors into buildings. Environ. Sci. Technol. 25:1445-1447.
307
-------
Johnson, P. C., M. B. Hertz, and D. L. Byers. 1990c. Estimates for hydrocarbon vapor
emissions resulting from service station remediations and buried gasoline-contaminated
soils. Petroleum Contaminated Soils. 3:295-326.
Johnson, P. C., M. W. Kemblowski, and J. D. Colthart. 1988. "Practical screening
models for soil venting application." In Proceedings ofNWWA/API conference on
petroleum hydrocarbons and organic chemicals in ground water, Houston, Texas.
Johnson, P. C., M. W. Kemblowski, and J. D. Colthart. 1990b. Quantitative analysis for
the cleanup of hydrocarbon-contaminated soils by in situ soil venting. Groundwater8
;41 3-429.
Johnson, P. C., et al. 1990a. A practical approach to the design, operation, and moni-
toring of in situ soil-venting systems. Groundwater Monitoring Review 1 0(2): 1 50-1 78.
Johnson, P. C., et al. 1992b. Vapor extraction-based remedial technologies: soil
venting, air sparging, bioventing, and thermally-enhanced vapor extraction (in review).
Johnson, R. L. 1993. "Enhancing biodegradation with in situ air. sparging: a conceptual
model." In Proceedings of the Conference on In Situ and On-Site Bioreclamation, April
6-8, Battelle Memorial Institute, San Diego, California.
Johnson, R. L., et al. 1992. Experimental examination of integrated soil vapor extrac-
tion techniques. Project Report, U.S. Environmental Protection Agency Risk Reduction
Engineering Laboratory.
Johnson, R. L., et al. 1993. An overview of in situ air sparging. Ground Water Moni-
toring and Remediation 13 (4): 127-1 35.
Joss, J. C., and A. L Baehr. 1994. AIR3D: An Adaptation of the Groundwater Flow
Code MODFLOWto Simulate Airflow in the Unsaturated Zone. U.S. Geological Survey
Open File Report, publication expected 1994.
Kabak, D. S., et al. 1991. Innovative Ground Water and Soil Remediation: In Situ Air
Stripping Using Horizontal Wells. Proc., Fifth National Outdoor Action Conference on
Aquifer Restoration, Ground Water Monitoring, and Geophysical Methods, National
Water Well Association, held in Las Vegas, Nevada, May 13-16.
Kent, B., and D. Graves. 1992. Enhanced Biodegradation of Contaminated Soils
Using Vent Wells. Contaminated Soils - Diesel Fuel Contamination. Eds. P. T.
Kostecki and E. J. Calabrese. Lewis Publishers, Boca Raton. Chapter 6, pp. 81-94.
308
-------
Lakshmiprasad, T., and R. R. Dupont. 1993. Pilot Scale Study of the Enhanced Bio-
-degradation Potential of JP-4 Jet Fuel Under Simulated Soil Vacuum Extraction Condi-
tions. Poster No. 15. Air and Waste Management Association, Denver, Colorado.
Lenhard, R. J., and J. C. Parker. 1990. Estimation of Free Hydrocarbon Volume From
Fluid Levels in Observation Wells. Groundwater, 28: 57-67.
Lenhard, R. J., and J. C. Parker. 1987. Measurement and Prediction of Saturation-
Pressure Relationships in Three-Phase Porous Media Systems. J. Contam Hvdrol 1-
407-424.
Lymari, W. J., W. F. Reehl, and D. H. Rosenblatt. 1982. Handbook of Chemical Prop-
erty Estimation Methods. McGraw-Hill, New York.
Mackay, D. M., and J. A. Cherry. 1989. Groundwater contamination: pump-and-treat
remediation, Environmental Science and Technology 23, (6): 630-636.
Mark-EJrown, N. V. 1994. Aspects of soil venting design, Hydrocarbon Bioremediation,
R. E. Hinchee, B. C. Alleman, and R. E. Hoeppel, eds. CRC Press, Inc , Boca Raton
pp. 362-367. '
Marley, M. C., D. J. Hazelbrouck, and M. T. Walsh. 1992. The application of in situ air
sparging as an innovative soil and groundwater remediation technology. Groundwater
Monitoring Review, 12 (2), 137-145.
Marley, M. C., E. X. Droste, and R. J. Cody. 1994. "Mechanisms that govern the suc-
cessful application of sparging technologies. Paper 94-WP102.05, presented at Air and
Waste Management Association 87th Annual Meeting, Cincinnati, Ohio, June 19-24.
McDonald, M. G., and A. W. Harbaugh. 1988. A modular three dimensional finite
difference ground water flow model, U.S. Geological Society Book 6.
Miller, R. N., et al. 1990. "A field scale investigation of enhanced petroleum
hydrocarbon biodegradation in the vadose-zone at tyndall AFB, Florida." In Petroleum
Hydrocarbons and Organic Chemicals in Ground Water Prevention, Detection, and
Restoration, NWWA Conference, Houston, Texas.
Millington, R. J., and J. M. Quirk. 1961. Permeability of porous solids. Trans Faraday
Soc. 57:1200-1207.
Mishra, S. J., C. Parker, and N. Singhal. 1988. Estimation of soil hydraulic properties
and their uncertainty from particle size distribution data. J. Hydro/. 108:1-18.
309
-------
Morris, D. A. and A. I. Johnson, 1967. "Summary of hydrologic and physical properties
of rock and soil materials." USGS Water-Supply Paper 1839-D.
Sabadell, G. P., et al. 1988. CSUGAS: Flowfield Model for In Situ Volatilization of
Organic Compounds in Soils. Argonne National Laboratory, Argonne, Illinois.
Sabadell, G. P. Eisenbecs, and J. J. Sunada. 1988. "The 3-D model CSUGAS: a man-
agement tool for the design and operation of soil venting systems." Proceedings of the
9th National Superfund Conference (November).
Shan, C., R. W. Falta, and I. Javandel. 1992. Analytical solutions for steady state gas
flow to a soil vapor extraction well. Water Resources Research 28 (4): 1105-1120.
Sims, J. L, et al. 1993. In situ bioremediation of contaminated unsaturated subsurface
soils. EPA Engineering Issue, EPA 540/S-93/501.
Travis, C. C., and J. M. Maclnnis. 1992. Vapor extraction of organics from subsurface
soils: Is it effective? Environmental Science and Technology 26(10): 1885-1887.
University of Wisconsin. 1992. Underground tank technology update. University of
Wisconsin-Madison, College of Engineering. 6(3).
U.S. Environmental Protection Agency. 1990a. Field measurements: Dependable
data when you need it, EPA/530/UST-90-003, pp. 61-74.
U.S. Environmental Protection Agency. 1990b. Assessing UST corrective action tech-
nologies: Early screening of cleanup technologies for the saturated zone, EPA/600/2-
90-027.
U.S. Environmental Protection Agency. 1990c. .Assessing UST corrective action tech-
nologies: Site assessment and selection of unsaturated zone treatment technologies,
EPA/600/2-90-011.
U.S. Environmental Protection Agency. 1991 a. Soil vapor extraction technology refer-
ence handbook, EPA-540/2-91/011.
U.S. Environmental Protection Agency. 1991b. Site characterization for subsurface
remediation. EPA/625-4-91/026.
U.S. Environmental Protection Agency. 1991c. Assessing UST Corrective Action
Technology, A Scientific Evaluation of the Mobility and Degradability of Organic Con-
taminants in Subsurface Environments. EPA/600/2-91/053.
310
-------
U.S. EEnvironmental Protection Agency. 1992. Underground storage tank research
program, Vol. 1 - Draft Report.
U.S. EEnvironmental Protection Agency. 1993a. Free product recovery and residual
hydrocarbon removal. UST Corrective Action Workshop (January).
U.S. EEnvironmental Protection Agency. 1993b. Subsurface characterization and moni-
toring techniques. Volume I: Solids and ground water, EPA/625/R-93/003a.
U.S. Environmental Protection Agency. 1993c. Subsurface characterization and moni-
toring techniques. Volume II: The vadose zone, field screening and analytical
methods. EPA/625/R-93/0036.
U.S. Environmental Protection Agency. 1993d. Remediation technologies screening
matric and reference guide. A joint project of the U.S. Environmental Protection
Agency and the U.S. Air Force.
U.S. Environmental Protection Agency. 1994. Soil Vapor Extraction Technology:
Evaluation of Mathematical Models for Remediation of VOC-Contaminated Sites.
Geotrans, Inc. Prepared for Risk Reduction Engineering Laboratory, Office of Research
and Development.
Valsacaj, K. T., and L. J. Thibodeaux. 1988. Equilibrium adsorption of chemical vapors
on surface soils, landfills, and landfarms-A review. J. Hazard Materials 19:79-99.
Waterloo Hydrogeologic Software. 1993. Airflow user's guide. Waterloo, Ontario,
Canada.
Weis, G. (Ed). 1990. Hazardous Chemicals Data Book, Noyes Data Corp., Park,
Ridge, New Jersey, p. 485.
Wilson, J. T. 1993. Testing bioremediation in the field. In situ bioremediation when
does it work? Natural Academy Press, Washington, D.C.
Wisconsin Department of Natural Resources. 1993. ERRP issue guidance on natural
biodegradation. Release News 3(1).
311
-------
-------