United States
Environmental Protection Agency, Region III
Office of Research and Development
KERL and RREL Laboratories July 27, 1993
A ONE-DAY
SEMINAR ON
BIOREMEDIATION
APPLICATIONS
Slide Copies /
Handouts
-------
TP
0^1
Disclaimer
10
1 Any mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
U.S. EPA Region HI
Regional Center for Environn>cf»t«i
Information
1650 Arch Street (3PM5SII
Philadel phia, PA 1910-3
-------
Contents
Speaker Biographies
Session Slides and Handouts
Microorganisms and Bioremediation I
Hugh Russell, Ph.D.
Bioremediation Technologies II
Daniel F. Pope, Ph.D.
Selection of Bioremediation Technologies III
Daniel F. Pope, Ph.D.
Treatability Studies IV
Hugh Russell, Ph.D.
Regulatory Review V
David E. Giamporcaro, Esq.
Biological Treatment of Contaminated Water VI
David K. Stevens, Ph.D.
Case Studies VII
Hugh Russell, Ph.D.
Case Studies on Bioventing VIII
Gregory D. Sayles, Ph.D.
-------
Speaker Biographies
DAVID E. GIAMPORCARO, ESQ.
David Giamporcaro is a Section Chief in charge of the
Biotechnology Program in the Office of Pollution Prevention
and Toxics (OPPT) at the Environmental Protection Agency
(EPA) in Washington, D.C.
David has practiced environmental law in both the public
and private sectors for the past 13 years. David received
his J.D. from the National Law Center at George Washington
University in Washington, D.C., in 1979. For the past five
and a half years, David has specialized in the regulation of
biotechnology products.
His other principal areas of expertise include the Toxic
Substances Control Act, the Federal Insecticide, Fungicide
and Rodenticide Act, and the Federal Plant Pest Act.
DANIEL F. POPE, Ph.D.
For ten years Dr. Pope conducted laboratory and field
research for the USDA on the physiology and toxicology of
biochemical interactions between the macroflora of agricul-
tural ecosystems. He then spent four years at the Missis-
sippi Forest Products Laboratory conducting laboratory and
field research on biological transformation of polynuclear
aromatic hydrocarbons and chlorinated phenols in soil eco-
systems, and designing industrial biological treatment
systems for remediation of soil and water contaminated with
wood preserving wastes. Currently he is employed by Dynamac
Corporation as a senior environmental scientist furnishing
technical assistance to the U.S. EPA on site characteriza-
tion and remediation problems and overseeing technical
research subcontracts.
-------
HUGH H. RUSSELL, Ph.D.
Dr. Hugh H. Russell is an ecological microbiologist with
the Superfund Technical Support Team at the Robert S. Kerr
Environmental Research Laboratory in Ada, Oklahoma. In
addition to providing support for in-house research pro-
jects, he is also responsible for providing technical assis-
tance to EPA Regional Offices and states in the remediation
of soil and ground water contaminated with hazardous wastes.
This includes the design and implementation of site charac-
terization activities as well as the selection and develop-
ment of remediation technologies. He lectures extensively
at EPA seminars and scientific conferences on soil and
subsurface bioremediation and has numerous publications in
these areas.
GREGORY D. SAYLES, Ph.D.
Greg Sayles is currently a Biochemical Engineer for the
Risk Reduction Engineering Laboratory of the U.S. EPA in
Cincinnati. He received his B.S. from the California Insti-
tute of Technology, M.S. from the University of California
at Davis, and Ph.D. from North Carolina State University,
all in Chemical Engineering. Dr. Sayles did Postdoctoral
research at Duke University's Center for Biochemical Engi-
neering before beginning work at the U.S. EPA in 1990.
Dr. Sayles' research interests are the development of
innovative microbial processes for the cleanup of hazardous
waste. He is particularly interested in bioremediation of
soils using sequential anaerobic/aerobic treatment and
bioventing. His research involves lab-scale studies as well
as field-scale demonstrations.
Dr. Sayles is a member of the American Institute of
Chemical Engineers, the American Chemical Society, and Sigma
Xi.
DAVID KING STEVENS, Ph. D., P.E.
Dr. Stevens received his B.S. in civil engineering from
Tufts University in 1976, and his Ph.D. in Civil and
Environmental Engineering from the University of Wisconsin -
Madison in 1983. He is currently Associate Professor in the
Department of Civil and Environmental Engineering at Utah
State University. His research and field activities include
projects concerning mass transfer and other transport pro-
cesses in bioreactors and soil, biotransformation process
kinetics of organics and inorganics, use of the white rot
fungus for biotransformation, and modeling of biotransforma-
tion and contaminant transport processes.
-------
A One-Day Seminar On
Bioremediation Applications
July 27, 1993
Session I
Microorganisms and
Bioremediation
Presented By
Hugh Russell, Ph.D.
-------
Microorganisms and Bioremediation
Bioremediation is the art of utilizing biological agents to
bring about desired changes within a controlled environment. The
desired changes we wish are the immobilization, detoxification or
destruction of contaminants of interest. These goals can be met in
a number of ways. Immobilization may be accomplished through
biologically induced changes in the valence state of metals, or
structural changes in organic chemicals which bring about
irreversible binding. Detoxification is accomplished by
microbially induced changes in an organic chemical which removes
the toxicity. For instance the dechlorination of trichloroethylene
to ethylene under anaerobic conditions. Destruction also referred
to as mineralization results in the transformation of organic
compounds to inorganic products. Organic chemicals are degraded in
stepwise manner to carbon dioxide, water, and cell biomass. Under
anaerobic conditions, methane may also be produced.
The most common agents used in bioremediation are bacteria.
These unicellular organisms produce enzymes for the degradation of
organic chemicals. These enzymes have either evolved overtime or
merely incorrectly recognize the organic chemical as some other
natural product. Other agents include fungi and actinomycetes.
Microbial agents require for growth a compound which accepts
electrons and is responsible for the regeneration of cellular
reducing power. These compounds are known as "electron acceptors".
Perhaps the most well known of electron acceptors is oxygen.
Oxygen is used by bacteria in much the same way as humans. The
final reaction is the reduction of molecular oxygen to water.
Other bacteria most notably the anaerobes utilize a variety of
compounds as alternate electron acceptors. Nitrate, iron,
manganese oxide and organic compounds are examples of electron
acceptors that may be utilized by bacteria. The use of electron
acceptors may be determined by not only availability, but
physiology.
Microbial agents which utilize oxygen as a terminal electron
acceptor are known as aerobes. The potential energy available from
oxygen is considerably higher than other electron acceptors.
However the difficulty in delivering high concentrations of oxygen
in aqueous solution to subsurface environments may lower the
available energy. If oxygen is unavailable in a system aerobic
bacteria may switch to nitrate as an electron acceptor. As nitrate
accepts electrons it is first reduced to nitrite and finally
molecular nitrogen. If nitrate is unavailable in a system then
iron and/or manganese may replace nitrate. The presence of sulfate
but lack of oxygen or nitrate affords growth conditions for a
specialized group of bacteria known as sulfate reducers. The end-
product of sulfate reduction is hydrogen sulfide. Finally, if no
other electron acceptors are present in a system, methanogens which
utilize organic compounds as electron acceptors may thrive.
-------
The degradation of organic chemicals depends on the energy
available in the system as well as the structure and solubility of
the chemical. Under aerobic conditions almost any soluble compound
can be degraded, depending on its structure and location of
halogens. Notable examples of which are degraded under aerobic
conditions include; monoaromatics such as benzene, toluene,
polyaromatic compounds such as naphthalene, flouranthene (4
condensed ring and higher pahs are usually degraded through co-
metabolism) , alkanes, alkenes and cyclic paraffins. Some
contaminants which may not be subject to aerobic bioremediation are
perchloroethylene, PCBs, dioxins and highly structured pesticides
such as DDT.
Presently nitrate reduction can cause the removal of aromatic
compounds such as toluene and xylene as well as naphthalene. The
use of nitrate reduction for the removal of benzene has not been
proven. Reports of both recalcitrance and degradation of benzene
appear in the literature. Future research should determine the
range of chemicals which can be degraded under nitrate reducing
conditions.
The depth of the ability of microbial consortia under sulfate
reducing conditions to remove chemicals of environmental concern is
unknown. To date reports appear on the degradation of toluene and
chlorinated solvents. Consortia under methanogenic conditions have
been reported to degrade mono-aromatics (less benzene) and cause
the reductive dehalogenation of chlorinated solvents.
In addition to the concentration of electron acceptor,
microbial growth may be limited by a number of environmental,
chemical and biochemical factors.
Environmental factors which may limit biodegradation are:
1) Temperature- an optimum temperature exists for the growth
and metabolism of bacteria.
2) pH- in most instances a pH within the range of 5-9 (with
an optimum of 7) is required for bioremediation.
3) Heavy metals- excessive concentrations of heavy metals
such as mercury or lead may inhibit microbial growth.
4) Salts- in addition to the toxicity of ions such as
sodium, increased concentrations of salt may bind water which
causes free-water limitations within a system.
5) Nutrients- in addition to a carbon and energy source,
microbial agents require the presence of additional nutrients, most
notably nitrogen and phosphorous.
Chemical factors which may limit biodegradation are:
1) Toxic concentrations- high concentrations of chemicals
-------
are toxic to microbial agents.
2) Too low concentration- concentrations of chemicals while
still above MCL may be too low to support active microbial growth.
3) Sorption- chemicals that are hydrophobic in environments
with high organic carbon will bind and may be unavailable for
microbial processes.
4) Structure- highly complex structured molecules are
difficult to degrade.
Biochemical factors which may limit biodegradation are:
1) Lack of proper enzymes- the microbial population may lack
the necessary enzymes or the necessary enzymes are only induced in
the presence of a particular compound which is missing from the
environment.
2) Presence of other more favorable carbon sources- the
presence of other carbon sources which are more readily utilized in
a system may preclude the degradation of COIs.
3) Toxic intermediate or end-products- some intermediate or
end-products of microbial metabolism feedback and cause an
inhibition of growth.
In summary bioremediation requires that the proper environment
in respect to electron acceptor be maintained and that the proper
nutrients are delivered to both the microorganisms and
contaminants. In addition, limitations to microbial growth such as
pH, toxic chemicals, must be removed to all practical extent.
-------
A One-Day Seminar On
Bioremediation Applications
July 27, 1993
Sessions II & III
Bioremediation Technologies
and
Selection of Bioremediation
Technologies
Presented By
Daniel F. Pope, Ph.D.
-------
Bioremediation
Technologies
Daniel F. Pope
Dynamac Corporation
Robert S. Kerr Environmental Research Lab
245- '
Soil
Bioremediation
245-2
Land Treatment - LT
• Bioremediation of soil at
ground surface
• In-situ, ex-situ
245- 3
-------
Land Treatment
• Applying water, nutrients,
electron acceptor, other
additives
• Provide appropriate
environment for
microorganisms
245- 4
Equipment &
Facilities For LT
• Excavators, haulers
• Tractor, tiller
• Fertilizer spreader, injector
245- 5
Equipment &
Facilities For LT
• Berm
• Liner
• Sump
• Irrigation
• Leachate treatment
• Cover
245- 6
-------
Electron Acceptors
For LT
• Oxygen in air
• Hydrogen peroxide
• Solid peroxides
Disposal Of Land
Treated Soil
• Spread on site
• Returned to excavation
• Containment cell
245- 8
Soil Heaps
• Soil piled up
• Air forced through
• Usually covered
245- 9
-------
Soil Heaps
Soil screened to improve
uniformity of air flow
245- 10
Equipment & Facilities
For Soil Heaps
• Excavators, haulers
• Screens, separators
• Blowers
• Piping
• Cover
• Liner or pad
• Irrigation
245- 11
Disposal Of Heap
Treated Soil
• Spread on site
• Returned to excavation
• Containment cell
245- 12
-------
Soil Slurry
Reactors - SSR
Soil suspended in
tank of water
245- 13
Soil Slurry Reactors
• Soil screened or washed to
remove fines for treatment
• Large particles hard to
maintain in suspension
245- 14
Equipment &
Facilities For SSR
• Excavators, haulers
• Screens, separators
• Tanks
• Spargers
• Mixers
• Dewaterers
• Water Treatment
245- 15
-------
I
I
Supplies For SSR
• Fertilizer
• Surfactants
245- 16
Composting
• Organic additions for
microbial substrate
• Heating stage, usually
245- 18
Disposal
• Spread on site
• Returned to excavation
• Containment cell
245- 17
-------
Composting Types
• Static pile
• Windrow
• Vessel
245- 19
Equipment & Facilities
For ComDOStin2
• Excavators & haulers
• Separators & Screens
• Mixers
• Windrowers
• Pad
• Covers
• Blowers & Piping
• Vessel
• Irrigation
245- 20
Supplies
• Bulking agent
• Fertilizer
245- 21
-------
Disposal
High Organic Content
24&- 22
Bioventins
Moving air through vadose zone
at low rate to supply oxygen
without large volatile loss
245-23
Equipment & Facilities
For Bioventing
• Blowers
• Injection/withdrawal wells
• Monitoring wells
• Off gas treatment
2
45- 24
-------
Bioremediation In
Aquifers
245-25
Air Injection
Air Sparging
246-26
Air Spar pin 2
• Injection of air into the
ground water
• Waste must be volatile or
biodegradable
>45- 27
-------
Equipment & Facilities
For Air Snareine
• Injection wells
• Blowers
• Venting wells
245- 2B
Water Injection
Injection of water into the
aquifer, usually with electron
acceptor and nutrients
245-29
Equipment & Facilities
For Water Iniection
• Injection wells or galleries
• Monitoring wells
245- X
-------
Electron Acceptors For
Water Injection
• Hydrogen peroxide
• Air saturated water
• Membrane
• High pressure U tube
• Pure oxygen
• Nitrate
245- 31
Bioremediation Of
Water
245-32
Fixed Film
Bioreactors - FFB
Film of microorganisms
growing on solid media
245- 33
-------
Equipment &
Facilities For FFB
• Tank
• Pumps
• Media
• Rotating disks
• Packing media, biorings
• Spargers, blowers
• Clarifiers
• Polishing
245
»- 34
Supplies For FFB
Nutrients
245- 35
Disposal
• Water
• Biosolids
245- 36
-------
Biological Activated
Carbon - BAC
• Fixed film bioreactor with
carbon adsorbent
• Powdered activated carbon
245- 37
Equipment & Facilities
For BAC
• Tank
• Pumps
• Media - Carbon packing
media
• Spargers, blowers
• Clarifiers
• Polishing
245-38
Disposal
• Water
• Biosolids
• Carbon
245- 39
-------
BAC Example:
The BIFAR System
245- 40
BIFAR
• Adsorption/biodegradation in
single vessel
• Adsorption stabilizes
contaminant concentration in
contact with microorganisms
245- 41
BIFAR
• Operated as a expanded,
fluidized bed reactor
/
• Influent through header into
bottom, effluent through top
245- 42
-------
Recycle/Aeration Vessel
• Collects effluent from reactor
• Aeration
• Some wasted, rest recycled to
reactor
245
- 43
BIFAR
Operational Variables
• Amount of activated carbon
• Bed expansion
• Influent flowrate
• System recycle flowrate
245- 44
Suspended Growth
Bioreactors - SGB
• Clumps of microorganisms
suspended in water
• Recycling of biosolids
• Batch or continuous
245- 45
-------
Equipment For SGB
• Tank
• Pumps
• Spargers
• Blowers
• Clarifiers
245- 46
Disposal
• Water
• Biosolids
245- 47
Selection Of
Bioremediation
Technologies
Daniel F. Pope
245-48
-------
Waste Location
245- «
Waste Location
Surface Soil
• LT
• Soil heaps
• Soil slurry
• Composting
245- SO
Waste Location
Surface Water
Bioreactors
245- 51
-------
Waste Location
Vadose Zone Soil
•LT
• Soil heaps
• Soil slurry
• Composting
• Bioventing
• Hydrogen peroxide injection
245-
52
Waste Location
Saturated Zone
• P&T
• Air sparging
• Water Injection
245-53
Waste Location
Active Site/
Under Structures
Bioventing
245- 54
-------
Geology
Homogeneous Sandbox
245- 55
Geology
Heterogeneity
• Sand, clay lenses
• Preferential flow paths
245- 56
'(o'^c^n fY""0
&4t
rot/aA si[( -{or
Cofl&Jhjon
Geology
• Low permeability materials
• 10"4cm/sec hydraulic
conductivity
245- 57
-------
Excavation of
Contaminated Material
Volatiles
245- SB
Contaminated
Matrix
245-59
Soil & Aauifer Solids
• Strongly adsorbed materials
may be difficult to
bioremediate
• PCBs, 5-6 ring PAHs, dioxins
245- eo
-------
Water
• Extremely soluble materials may
be remediated by pump & treat
• P&T will not completely remove
most materials
• May be used for part of
remediation, and to control site
245- 61
Soil Gas
Very volatile materials may be
removed by venting or sparging
245-62
NAPL
• Bio does not work well on NAPL
• Other technology needed to
remove or disperse NAPL
245- 63
-------
Waste Type
Recalcitrant Compounds
• 5-6 ring PAHs
• PCBs
• Dioxins
245- «
Waste Type
PCBs
• Strongly adsorbed to soil
- no diffusion to microbes
• Recalcitrant to biodegradation
• May be subject to anaerobes
245- 65
Waste Type
PAHs
• 2, 3 ring readily biodegradable
• 4 rings moderately easily biodegraded
• 5-6 rings hard to biodegrade, strongly
sorbed, very low water solubility
245- 66
-------
Waste Type
Dioxins
245- 67
Waste Type
Pentachlorophenol (PCP)
• Quite toxic to microbes
• Reasonably readily biodegraded
at lower concentrations
• Soil less than 1000 mg/kg
• Water at moderate pH
245- ffi
Waste Type
BTEX
• Readily degraded
• Very volatile
245- 60
-------
Waste Type
Linear Hydrocarbons
• Short chain
• Readily degraded
• Very volatile
245- 7D
Waste Type
Linear Hydrocarbons
• Medium chain
• Degradable
• Some mobility
245-71
Waste Type
Linear Hydrocarbons
• Long chain
• Slowly degradable to very
slowly
• Very low mobility unless in
NAPL
245- 72
-------
Final Levels
Very waste, matrix,
technology dependent
245- 73
Final Levels
• Water
• Soil slurry
• Bioventing
• Soil piles
• Pump & treat, injection
• Composting
• Land treatment
245- 74
Time Frame
Water
Fast - hours usually
245- 75
-------
Time Frame
Soil
• LT - months per lift
• Slurry - weeks
• Composting - weeks to months
• Soil piles - weeks to months
• Bioventing - weeks to months
245-
76
Time Frame
Aquifer
• P&T - years
• Injection - years
245-77
-------
Soil Heap Bioremediation
Visqueen Cover
Asphalt
Side
Nutrients
Aeration
Miroorganisms
Plastic Piping
(compatible with contaminants)
Top V
145-74
-------
Soil Slurry Bioreactor
-------
r--
-------
Unscreened
or Screened
Compost
Bulking Materials
and Sludge
Perforated Pipe
Trap for Water
Filter Pile
Screened
Compost
-------
OUTFEED
MIXER 5 ^ INFEED
> ~
-------
ECONOMIC FACTORS
Composting
VOC Control Costs
Energy Input
Technical Oversight
Reactor Costs
Infrastructure Costs
Handling Equipment Costs
~ In-Vessel
H Static Pile
¦ Turned Pile
EXPECTED COST
-------
Typical Bioventing System
Injection
Well
Knock-
By-Pass
-------
~ ~ ~ ~ ~
Vapor
Treatment
Pumping
Well
~
Vacuum
Pump
~ ~ ~ ~
Compressor
-------
Monitoring
Probe
Vapor
Extraction
Well
Air Sparger Monitoring
Well Probe
Vent
Radius
V
Sparge
Radius
°°° CD° O
Vent Radius = f(Vacuum)
Sparge Radius = f(Depth)(Pressure)
oil Parti
ei
Contaminated
Soil
Transient Air
Filled Porosity
Diagram of Air Sparging System
-------
Channeled Air Flow Through
V
V
High Permeability
"* Zone \
\
Air/Contaminant r
Migration - ~
-------
Impervious Barrier
Dissolved Particles
CD o o
°o°°
Air/Contaminant
Migration
-------
Air Sparging Removal Mechanisms as a Function of Product Volatility
Oil Oil Spirits
Volatility
-------
Air
Compressor or
Hydrogen Peroxide
Tank _
Nutrient Addition
Infiltration Gallery
Trapped Hydrocarbons
Recirculated Water
and Nutrients -
Water Table
Monitoring Weil
Recovery Well
Use of Infiltration Gallery For Recirculation of Water and
Nutrients in In Situ Blorestoration
-------
~ ~~~~~~~~~~~~~
~ ~
Air Compressor
Coarse
Sand
-<— Production Well
Water Table
Spilled Materials
To Sewer or
Recirculate
Nutrient
Addition
Tank
Water
Supply
171-67
-------
Blower
Effluent
Air Line 1
Spargers with
Antibackflow
Valves
-------
-------
Effluent
Influent
Bioreactor
Rn
°°°o
Aeration
Tank
Qn°oO°g°ob£P
Recirculation
Biological Activated
Carbon Bioreactor
-------
Influent
Return Biosolids
Effluent
Suspended Growth
Bioreactor
-------
Wate
T
Residual
Saturation
Saturated Zone
Unsaturated Zone
Vapor Phase
Capillary
Fringe
Dissolved
Contaminants
-------
SUBSURFACE MASS BALANCE
CONCEPTUAL FRAMEWORK
Organic
Water
NAPL
Inorganic
-------
DNAPL, WATER, AND AIR IN POROUS MEDIA
-------
Ranges of Porosity in Soils
o
20
40 60
Porosity %
80
100
-------
Permeability
Clay
Silt
/ j
Sane
00
1
_ ©
O
I
~ ©
/^1 1 V" V" V" %¦ V« % ¦ %¦ V- V> V- v. Vi v- v. ¦
1 yfQ\/P ¦": ¦": ¦"¦¦ ¦": •":-V -V ¦".
«, .%-\¦ % . % .% ¦ s ¦1 s ¦% ¦i % ¦% ;S ¦ %«1 ¦
1 1 1 1 1
106 104 102 10° 102
cm/sec
-------
Range of Effectiveness of
Selected In Situ Techniques
SVE Enhanced Bioremediation
Soil Vacuum Extraction
Technically Feasible
Most Cost Effective
In Situ Bioremediation
In Situ Soil Washing
Water Flooding
Decontaminated
Adsorbed
Contaminants
Up to > 10,000 ppm
Residual Oil
High Contaminant Levels
TO to 25% Oil by Wt.
Mobile, Free
Phase Oil
-------
DRAFT
BIOREMEDIATION
USING THE
LAND TREATMENT CONCEPT
Daniel F. Pope
Senior Staff Environmental Scientist
Dynamac Corporation
USEPA Robert S. Kerr Environmental Research Laboratory
Ada, OK 74820
John E. Matthews
Supervisory Environmental Scientist
USEPA Robert S. Kerr Environmental Research Laboratory
Ada, OK 74820
March 23, 1993
-------
INTRODUCTION
Definition of Land Treatment
Land treatment involves use of natural biological, chemical and physical processes in the soil to
transform organic contaminants of concern. Biological activity apparently accounts for most of the
transformation of organic contaminants in soil, although physical and chemical mechanisms may
provide significant loss pathways for some compounds under some conditions. Degradation by
ultraviolet light may serve as a loss pathway for certain hydrophobic compounds at the soil surface.
Volatilization of low molecular weight compounds also takes place at the soil surface and provides a
significant loss pathway for such compounds. Certain chemical reactions such as hydrolysis can play
an important role in transformation of some compounds. The relative importance of these processes
varies widely for different compounds under different circumstances. The land treatment concept
serves as the basis for design and operation of soil bioremediation technologies at a large number of
waste sites requiring cleanup.
Microorganisms and Bioremediation
Bioremediation is carried out by means of microorganisms. Both bacteria and fungi have been
shown to be important in bioremediation processes. Most research in bioremediation has centered on
bacteria, but some investigators have found that fungi can play an important role in bioremediation
processes, especially with halogenated compounds (e.g., pentachlorophenol, a wood preservative). It
is important to realize, however, that in almost all cases bioremediation relies on communities of
microorganism species, rather than one or a few species.
Bioremediation consists of utilizing techniques for enhancing development of large populations of
microorganisms that can transform the pollutants of interest, and bringing these microorganisms into
intimate contact with the pollutants. In order to do this efficiently, necessary provisions for growth
and reproduction of the microorganisms must be maintained.
Life processes for all known living creatures are carried out in water. Some organisms, such as
human beings, can maintain an internal water environment while moving about in a relatively dry
outer environment. Many microorganisms, however, cannot maintain an appropriate inner
environment without being in a relatively wet outer environment. Most microorganisms that are
active in bioremediation must live in water. This water may be in tank reactors or an aquifer, or it
may be a thin film of water on a soil particle.
Microorganisms are sensitive to the osmotic potential of the solution in which they live. The
osmotic potential affects the ability of the microorganism to maintain itself with a desirable amount
of water internally. If the environment is too dry or if the water in the microorganism's environment
contains too high a concentration of dissolved salts, the microorganism cannot maintain the proper
amount of water internally. This factor can be a problem for bioremediation schemes where, for
example, process waters or contaminated soils have high levels of dissolved salts. Sudden changes in
osmotic potential can seriously inhibit microbial activity, often resulting in lysis (disintegration of cell
walls). Microorganisms can adapt to environmental changes within limits if such changes are not
induced rapidly.
1
-------
Specific microorganisms are active within a relatively narrow range of temperatures. Most
bacteria that carry out bioremediation processes are mesophiles ("middle lovers") and are most active
from about 18 to 30 degrees centigrade. Temperatures significantly higher or lower than this range
will significantly limit their activity. Within this range, activity will usually be higher at the higher
temperatures. Outside this range, activity decreases as the temperature moves further away from
these limits. At lower temperatures, activity does not usually stop completely until the freezing point
is reached.
Most microorganisms active in bioremediation processes are aerobic, that is, they require free
(uncombined) oxygen. Some treatment processes make use of anaerobic microorganisms that do not
require free oxygen; however these processes are not yet widely used in environmental cleanups.
Microorganisms living in aqueous reactors, aquifers, or in the subsoil may be supplied with oxygen
by pumping air or oxygen-supplying compounds (e.g., hydrogen peroxide) into the environmental
system. Microorganisms growing in surface soil are usually supplied with oxygen by tilling the soil
to allow air to enter. In many remediation situations the essential problem is the balance between
water and oxygen: the more water, the less oxygen, and vice versa. In the soil environment, for
example, the oxygen supply and the water supply are essentially inversely related, since the pore
space in soil is occupied by either air or water.
In general, microorganisms are active within a relatively broad pH range. The pH is a measure
of the acidity or basicity of the environment. In general, few microorganisms do well below pH 5 or
above pH 9. Although many microorganisms can adapt to pH levels within that range, fungal species
tend to be the dominant members of the microorganism community below pH 6, and bacteria tend to
dominate above pH 7.
The pH range within which bioremediation processes are considered to operate most efficiently is
6 to 8. The optimum pH range for a particular situation, however, is influenced by a complex
relationship between the microorganisms, pollutant chemistry and external environment, and thus is
site-specific. If deemed necessary, the pH can be adjusted to the desired range by the addition of
acidic or basic substances, i.e., sulphur or lime.
Microorganisms are sensitive to the presence of a wide variety of compounds and elements. High
concentrations of heavy metals, certain highly halogenated organics, some pesticides and other
exogenous materials can inhibit bioremediation. Effects of these inhibitors vary so greatly with
concentration, environmental factors, rapidity of contact with the inhibitors, and time of contact that
it is difficult to set any definite concentration limits above which bioremediation is precluded.
Generally, laboratory treatability studies can be used to provide data necessary for decisions in
doubtful cases.
Metals often are present in soils contaminated with organic wastes. These metals will not be
treated (transformed or degraded) in the same sense as the organic materials. However, valence states
may be changed and chemical bonds may be broken so as change the toxicity or mobility of the metals.
Addition of manures and other complex organic materials often used in land treatment may tend to
reduce the mobility of many metals by increasing the ion exchange capacity or adsorption capacity of
the soil.
Microorganisms must have carbon sources and mineral nutrients (nitrogen and phosphorus, for
example) in order to live and reproduce. In many cases, the pollutants themselves will supply the
carbon source and some nutrients; however, usually mineral nutrients and sometimes a supplemental
carbon source must be supplied. Mineral nutrients are usually supplied as soluble salts (fertilizers).
If necessary, carbon may be supplied as animal manures (which will also supply many mineral
nutrients), molasses, glucose, wood chips or a variety of other carbon containing materials.
2
-------
There must be a balance between the various mineral nutrients and the carbon source or the
microorganisms will not be able to make optimum use of the carbon source. For most bioremediation
situations, biodegradation is optimal at carbon/nitrogen ratios in the range of 10 - 30 to 1 and
nitrogen/phosphorus ratios of about 10 to 1.
There are fifteen or more other mineral nutrients that must be available in appropriate amounts,
but these minor nutrients are usually present in the environment in sufficient amounts except in some
process waters and ground waters. The amounts of foods and nutrients to be added should be based
on results from laboratory and site treatability studies.
The availability of nutrients to microorganisms is strongly influenced by pH, but since soil pH
generally is maintained or adjusted to the 6-8 pH range in a land treatment scenario, limited
availability of nutrients caused by pH is rarely a problem.
Since microorganisms are responsible for transformation of pollutants during bioremediation, it
would seem reasonable to assume that the greater numbers of microorganisms present, the faster the
transformation would occur. This has not been found to be generally true; however, it is still
advantageous to have some quantitative measure of the microorganism population.
Several techniques have been devised to determine the microorganism population. One commonly
used method is standard plate counting. Samples of soil, water, or other matrix in which
microorganisms are growing are applied to Petri plates containing a nutrient media. The plates are
incubated for several days to allow microorganisms to grow, after which the number of microorganism
colonies growing on the plates are counted. These counts can then be related to the numbers of
microorganisms present in the original matrix. Such counts can be useful for noting effects of
environmental factors and for adjusting food and nutrient inputs; however, the relationship between
population counts and pollutant transformation rates is not well defined. Often, there are many types
of microorganisms in the bioremediation environment that will be counted in plate counts, but only
a few of these types may actually be involved in transformation of the contaminants of concern. The
nutrient media may be spiked with waste compounds of interest; microorganisms that grow on such
media are considered to be tolerant to the spiked compounds. The use of spiked media yields an
indication of population levels of tolerant microorganisms in the soil tested. Note that this procedure
does not give an indication of the microorganisms present that can degrade the spiked compound, only
those that can survive and grow in the presence of the spiked levels of the compounds.
There are several physical constraints on the use of microorganisms in remediation of soil
contaminants. These are generally related to the problem of getting contaminants and microorganisms
together in close contact under environmental conditions desirable for microbial activity. Generally,
a contaminant must be able to move through the waste/soil matrix and pass through the
microorganism's cell membrane in order for microorganisms to transform the contaminant. In some
cases contaminants can be transformed by extracellular enzymes without entering into the
microorganisms.
Waste compounds that have low solubility in water (for example, 4, 5, and 6 ring polycyclic
aromatic hydrocarbons [PAHs]) move slowly from soil adsorption sites or free phase droplets into the
soil water and from there into the microorganism. Wastes in solid matrices (soil) will have less solvent
(water) in which to be dissolved, will be more likely to have highly variable concentrations throughout
the matrix, will be harder to mix thoroughly for even distribution throughout the matrix, and often
will have a relatively high tendency to be adsorbed onto matrix solids.
3
-------
All of these factors tend to limit accessibility of contaminant compounds to the microorganisms;
therefore it is often easier to achieve biodegradation of a given contaminant in water than in soil.
Also, soil treatment processes where soil is suspended in water and constantly mixed (soil slurry
bioremediation) will usually have faster biotransformation rates than simple solid phase soil
bioremediation processes.
LAND TREATMENT TECHNOLOGY
Land treatment techniques for bioremediation purposes most often are used for treatment of
contaminated soil, but certain petroleum waste sludges have long been applied to soil for treatment.
Ideally, the contaminated soil can be treated in place (in-situ). Often, however, the soil must be
excavated and moved to a location better suited to control of the land treatment process (ex-situ).
In-situ land treatment is limited by the depth of soil that can be effectively treated. In most soils,
effective oxygen diffusion sufficient for desirable rates of bioremediation extends only a few inches to
about twelve inches down into the soil. Air can be pushed or pulled through the soil, but this is
characteristic of bioventing technology and not land treatment. Usually when it is desired to treat
soil in-situ to depths greater than twelve inches, the surface layer of soil is first treated to the desired
contaminant levels, and then the surface layer is removed, or tilled so that lower layers of soil are
moved to the surface for treatment. Most tractor mounted tilling devices can till only to a depth of
about twelve inches. Large tractors with specialized equipment that can till to depths of three feet
or more have been used for in-situ land treatment. Large augers are now available that can move soil
from 50-100 feet depths to the surface, but the practicality of this technique for in-situ land treatment
has not been demonstrated.
Ex-situ treatment generally involves application of lifts of contaminated soil to a prepared bed
reactor. This reactor is usually lined with clay and/or plastic liners, provided with irrigation, drainage
and soil water monitoring systems, and surrounded with a berm. The lifts of contaminated soil are
usually placed on a bed of relatively porous non-contaminated soil.
Whether practiced in-situ or ex-situ, effective land treatment is generally limited to the top 6 to
24 inches of soil. Twelve inches or less is the preferred working depth. Below twelve inches the
oxygen supply is generally inadequate for useful rates of bioremediation using standard land
treatment practices. Tilling is used to mix the soil and increase the oxygen levels, but tilling is
usually limited to 12 inches or less unless specialized equipment is available. Soil that is
contaminated to depths greater than 12 inches may be bioremediated by removing successive soil
layers as they are treated to the required levels, or by moving layers of soil (lifts) from one location
to another for treatment.
The land treatment process may be severely limited in clayey soils, especially in areas of high
rainfall. This limitation is primarily related to oxygen transfer limitations and substrate availability
to the microorganisms. Clayey soils should be applied in shallower lifts than sandy soils, preferably
no more than 9" in depth. Tilth ("workability" of the soil) can often be improved by adding organic
matter or other bulking agents to the soil. If high sodium content causes the soil to have poor tilth,
gypsum (calcium sulfate) can be added.
4
-------
In a prepared bed system, lightly contaminated soil should be applied to the land treatment unit
(LTU) at the beginning of operation before any heavily contaminated soil is placed on the treatment
unit. Manure, nutrients and water should be applied to this initial lift until total microorganism
populations are built up to 106-107 colony forming units per gram of soil After microorganism
populations are high and constituents of interest are reduced to acceptable levels, more heavily
contaminated lifts can be applied.
The soil should be screened before application to remove any debris greater than one inch in
diameter, especially if significant amounts of debris or rocks are present. Any large debris that may
adsorb the waste compounds (i.e., wood), should be removed if possible and treated separately. Small
rocks and other relatively nonadsorptive wastes can be treated if they do not interfere with tillage
operations.
After application to the LTU, each lift should be tilled at intervals to enhance oxygen infiltration
and contaminant mixing with the microorganisms. The soil should be near the lower end of the
recommended soil moisture percentage range before tilling. Tilling when the soil is very wet or
saturated tends to destroy the soil structure, reducing oxygen and water intake and movement,
causing reduced microbial activity. Tilling should not begin until at least 24 hours after irrigation or
a significant rainfall event. Tilling more than is necessary for enhanced oxygen infiltration and
contaminant mixing may be counter-productive since tilling tends to destroy the soil structure and
compact soil below the tilling zone. Tillers tend to mix the soil only along the tractor's line of travel,
so tillage should be carried out in varying directions, i.e., lengthwise of the LTU, crosswise, and on the
diagonal.
A tractor mounted rotary tiller is recommended for tilling and mixing purposes. Occasionally a
subsoil plow or chisel plow should be used to break up any hardpan or zone of compaction created by
tillage and passage of equipment across the LTU. If the tiller does not till deeply enough to mix soil
from the last lift into the top few inches of soil from the preceding lift, a turning plow may be used
to achieve desired mixing of the two lifts. Use of the turning plow should be followed immediately by
tilling to complete the mixing action.
Each subsequent lift, usually six to twelve inches in depth as applied, should be tilled into the top
two or three inches of the previous lift. This tilling will mix populations of well acclimated
microorganisms from the treated lift into the newly applied lift, and help reduce the length of time
for high populations of active degraders to be built up in the new lift.
Timing of application of succeeding lifts should be based on reduction to defined levels of particular
compounds or categories of compounds in the preceding lift. For instance, the goal might be to reduce
total petroleum hydrocarbons (TPH) to less than a regulatory or risk calculated limit in the current
lift before application of a new lift.
Once desired target levels of compounds of interest are established, data obtained from the LTU
monitoring activities can be statistically analyzed to determine if and when desired levels are reached
and the LTU is ready for another lift of soil to be applied.
Nutrients, Carbon Sources, and Other Additives
Microorganisms in land treatment units require carbon sources and nutrients. Fertilizers can be
used to supply the nutrients, while wood chips, sawdust or straw can supply carbon. Various animal
manures are often used to supply both carbon sources and nutrients. High organic levels in manures,
wood chips and the other organic amendments increase sorptive properties of soil, thereby decreasing
mobility of organic contaminants.
5
-------
Organic amendments will also increase water holding capacity of soil. A higher water holding
capacity is desirable in sandy soils, but it can cause difficulty when land treatment is conducted in
areas of high rainfall and poor drainage. In an excessively wet soil the oxygen supply is reduced,
which may essentially stop or severely limit transformation of waste constituents. Drainage must be
carefully designed in these situations.
Animal manures can provide desirable organic supplements. Manure should be applied to each
lift at the rate of about 3-4% by weight of soil. Chicken, cattle, horse or sheep manure will be
effective. The manure selected should be analyzed for nitrogen and phosphorus to determine if any
additional amounts of these nutrients (above the amounts found in the manure) need be applied. The
manure should be in small particles and should be thoroughly tilled into the soil lift.
Manures are often mixed with sawdust or straw, since these materials are used as bedding in
stock facilities. This is acceptable and even desirable since these materials act as bulking agents in
soil. However, the high percentage of cellulosic material in these materials will usually exert a high
nitrogen demand, thereby reducing the amount of nitrogen available for microorganisms transforming
contaminants of concern. If necessary, available nitrogen can be increased with addition of
appropriate inorganic fertilizers.
Additional nitrogen or phosphorus, if required, may be applied as fertilizer grade ammonium
nitrate (for nitrogen), triple superphosphate (for phosphorus), or diammonium phosphate (for both
nitrogen and phosphorus). Nitrogen fertilizers can cause soil pH to be lowered due to formation and
leaching of the nitrate ion coupled with soil cations.
Agricultural fertilizer is usually supplied in prilled or pelleted form (the fertilizer compounds
formed into pellets with a clay binder) suitable for easy application over large areas of soil. Technical
grade, unformulated fertilizer compounds (for instance, ammonium chloride crystals) are difficult to
spread evenly over the land surface. The pelleted fertilizers may be applied with a hand operated or
tractor operated cyclone spreader. Soluble fertilizers are available that can be applied through
irrigation systems. With the use of irrigation systems, application rates may be closely controlled,
applications can easily be made as often as irrigation water is applied, and the fertilizer (since it is
soluble) is immediately available to the microorganisms. Equipment is available to meter concentrated
nutrient solutions into the irrigation system on a demand basis.
The nutrient requirements for biodegradation in the field have not been thoroughly studied, and
detailed information is not available to indicate the optimal levels of particular nutrients in field
situations. The amount of nitrogen and phosphorus needed may be estimated by having
representative soil samples from the cells analyzed by the state agriculture department or university
laboratory, or by a private agricultural soil testing laboratory. General fertilizer recommendations for
vegetable gardens should be followed. The analytical laboratory should be made aware that the soil
contains hazardous materials.
Sometimes inorganic micronutrients, microbial carbon sources, or complex growth factors may be
needed to enhance microbial activity. Animal manures generally will supply these factors.
Proprietary mixtures of various of these ingredients are sometimes offered for sale to enhance
microbial activity. Proof of the efficacy/cost effectiveness of these mixtures is lacking in most cases.
6
-------
Microorganism cultures are often sold for addition to bioremediation units. Two factors limit use
of these added microbial cultures in LTUs: (1) nonindigenous microorganisms rarely compete well
enough with indigenous populations to develop and sustain useful population levels, and (2) most soils
with long term exposure to biodegradable wastes have indigenous microorganisms that are effective
degraders the LTU is managed properly. If the use of proprietary additives is proposed, results of
well-designed treatability studies with appropriate controls should be provided by the vendor to
support such use.
Certain soil factors may interfere with microbiological activity in the LTU soil. High salt levels,
indicated by high electrical conductivity (EC) readings, may reduce or stop useful microbiological
activity. If levels are too high, it may be necessary to leach the soil with water to remove excess salts
before biodegradation can occur. High levels of sodium may be detrimental to soil structure. Sodium
levels may be reduced by applying calcium supplements (usually gypsum, CaS04) and leaching.
Leaching of contaminants may also occur at the same time.
Soil Moisture Control
Soil moisture should be maintained in the range of 40-70% of field capacity; recent experience
indicates that 70-80% of field capacity may be optimum. A soil is at field capacity when soil
micropores are filled with water and soil macropores are filled with air. This condition allows soil
microorganisms to get air and water, both of which are necessary for aerobic biodegradation to occur.
Maintaining soil at somewhat less than field capacity (40-70% of field capacity) allows more rapid
movement of air into the soil, thus facilitating aerobic metabolism without seriously reducing the
supply of water to microorganisms. If soils are allowed to dry excessively, microbial activity can be
seriously inhibited or stopped.
Field capacity of a soil may be determined by saturating a sample of the soil with water and
allowing the soil to drain freely for 24 hours. The soil is weighed, oven dried at 105° C until constant
weight, then weighed again. The difference between the weight of drained soil and the oven dried soil
gives the weight of water in that amount of soil at field capacity. The weight of water divided by the
dry weight of the soil gives the percentage of water in the soil at field capacity. A sandy soil might
hold as little as 5% of its dry weight in water at field capacity; a clay soil might hold as much as 30%
of its dry weight in water at field capacity. ,
Soil moisture may be monitored with a porous cup tensiometer. Instructions supplied with the
tensiometer detail the use of tensiometer for scheduling irrigation for field crops. These instructions
are generally adequate for scheduling irrigation for LTUs.
Continuous maintenance of soil moisture at adequate levels is of utmost importance. Either too
little or too much soil moisture is deleterious to microbial activity. Monitoring soil moisture and
scheduling irrigation is important, and requires considerable, constant attention to detail, and is
perhaps one of the most neglected areas of LTU operations.
Moisture enhancement can be accomplished with by traveling gun or similar irrigation systems
that can be removed to allow easy application of lifts. Hand-moved sprinkler irrigation systems are
more often used, although they are usually more expensive. Sprinkler systems can be designed with
quick detach couplings to facilitate movement when placing lifts of contaminated soil. Permanently
installed sprinkler systems with buried laterals and mains may be used, but the sprinkler uprights
must be avoided when placing lifts and during other LTU operations. The uprights may need to be
lengthened if many lifts are placed during the operating life of the LTU.
7
-------
If a permanently installed, buried line system is used, the uprights should be connected to the
buried lateral lines with a short piece of plastic pipe. Some of the uprights will be hit by field
equipment during operations, and the plastic pipe will break before the lateral line or other parts of
the piping system. The plastic pipe can be easily repaired, while a bent or broken lateral line or
upright can be difficult to repair.
The operating pressure for most sprinklers ranges from 30 to 50 lb/in2. Sprinklers may have two
nozzles, one to apply water at a distance from the sprinkler (range nozzle) and one to cover the area
near the sprinkler (spreader nozzle). Sprinklers may be static or rotating, with a hammer or other
device to cause the sprinkler to rotate. Since one sprinkler will not apply water uniformly over an
area, sprinkler patterns should be overlapped to provide more uniform coverage. The usual overlap
is around 50%; that is, the area covered by one sprinkler reaches to the next sprinkler. Highly
uniform coverage is difficult to achieve in the field, especially in areas where winds of more than 5
mph are common.
Small LTUs can be covered with sprinklers set only on the sides of the LTU. Sprinklers can cover
full, half or quarter circles so that sprinklers on the sides or in the corners of the LTU will cover only
the LTU and not the berm or outside the LTU.
The irrigation system should be sized to allow application of at least one inch of water in 10-12
hours. The rate of water application should never be more than the soil can absorb with little or no
runoff. Since LTUs consist of bare soil, runoff can rapidly cause significant erosion. Generally, coarse
(sandy or loamy), deep soils can take up water at a faster rate than fine clay or clay loam, shallow
soils. Usually application rates of more than 0.5 inches of water per hour are not recommended;
clayey soils with slopes greater than 0.2-0.3% will require lower rates of water application. A water
meter to measure the volume of water applied is helpful in controlling application.
Surface drainage of the LTU can be critical in high rainfall areas. If soil is saturated more than
an hour or two aerobic microbial action is reduced. The LTU surface should be sloped 0.5-1.0%.
Greater slopes will allow large amounts of soil to be washed into the drainage system during heavy
rains. Even a slope of 0.5-1.0% will allow soil to be eroded; therefore the drainage system should be
designed to allow collection and return of eroded soil to the treatment unit.
Underdrainage is generally provided by a sand layer or a geotextile/drainage net layer under the
LTU. The system should be designed so that excess water in soil pores over field capacity will be
quickly drained away so microbial activity will not be inhibited. The lifts of contaminated soil are
usually placed on a bed of sand or other porous soil. This gives a "perched" water table - the
contaminated soil lift will take up water from irrigation or rain until the soil nears saturation, then
the lift begins to drain excess water into the treatment unit drainage system. The interface between
the lift and the coarse layer underneath should be composed of well graded materials so that the
transition from the (usually) relatively fine soil texture of the lift to the coarse texture of the drainage
layer is gradual rather than sudden. Movement of water through the interface will be enhanced,
improving drainage. Grading of the materials reduces the tendency for the soil lift to become
saturated with water before drainage occurs, which would inhibit aerobic biological activity. The
change in texture at the interface can be made more gradual by tilling the lift into the top few inches
of the drainage layer.
Some storage capacity should be provided so that runoff and leachate water can be recycled onto
the LTU. A one inch rain might give a combined runoff and leachate of 10,000 to 27,000 gallons per
acre of LTU, if the LTU is being maintained at the proper (relatively high) moisture content.
Therefore, it may not be practical to provide storage capacity for large rainfall/runoff events.
8
-------
In many cases leachate/runoff water cannot be discharged without treatment. Biological reactors
are commonly used to treat this water prior to discharge. Alternatively, effluent from the biological
treatment unit may be applied to the LTU through the irrigation system. Nutrients and
microorganisms from the biological treatment system may enhance the microbial activity within the
LTU.
9
-------
BIBLIOGRAPHY
-------
BIBLIOGRAPHY
Land Treatment Concept References
Bulman, T.L., S. Lesage, P.J.A Fowlie, M.D. Webber. November 1985. The Persistence of
Polynuclear Aromatic Hydrocarbons in Soil. PACE Report No. 85-2. Petroleum Association For
Conservation of the Canadian Environment. 1202 - 275 Slater Street. Ottawa, Ontario.
Loehr, R. 1989. Treatability potential for EPA listed hazardous chemicals in soil. U. S.
Environmental Protection Agency, Robert S. Kerr Environmental Research Laboratory, Ada, OK,
EPA/600/2-89/011.
Lynch, J., and B. R. Genes. 1989. Land treatment of hydrocarbon contaminated soils. In
Petroleum Contaminated Soils. Vol. 1: Remediation Techniques. Environmental Fate, and Risk
Assessment. P. T. Kostecki and E. J. Calabrese, Eds., Lewis Publishers, Chelsea, MI, p. 163.
Park, K. S., R. C. Sims, and R. R. Dupont, W. J. Doucette, and J. E. Matthews. 1990. Fate of
PAH compounds in two soil types: influence of volatilization, abiotic loss and biological activity.
Environ. Toxicol. Chem.. 9, 187.
Rochkind, M. L., J. W. Blackburn, and G. S. Sayler. 1986. Microbial Decomposition of Chlorinated
Aromatic Compounds. EPA/600/2-86/090, Hazardous Waste Engineering Research Laboratory, U.
S. Environmental Protection Agency, Cincinnati, OH.
Ross, D., T. P. Marziarz, and A L. Bourquin. 1988. Bioremediation of hazardous waste sites in
the USA: case histories. In Superfund '88. Proc. 9th National conf.. Hazardous Materials Control
Research Institute, Silver Spring, MD, p. 395.
Sims, J.L., R.C. Sims, and J.E. Matthews. Bioremediation of Contaminated Surface Soils. August
1989. U. S. Environmental Protection Agency, Robert S. Kerr Environmental Research Laboratory,
Ada, OK, EPA-600/9-89/073.
Sims, R. C. 1990. Soil remediation techniques at uncontrolled hazardous waste sites. J. Air &
Waste Management Assoc. 40
Sims, R. C., D. L. Sorensen, J. L Sims, J. E. McLean, R. Mahmood, and R. R. Dupont. 1984.
Review of in-place treatment technologies for contaminated surface soils - Volume 2: Background
information for in-situ treatment. U. S. Environmental Protection Agency, Risk Reduction
Research Laboratory, Cincinnati, OH, EPA-540/2-84-003b.
Sims, R. C., W. J. Doucette, J. E. McLean, W. J. Grenney, and R. R. Dupont. 1988. Treatment
Potential for 56 EPA Listed Hazardous Chemicals in Soil. U. S. Environmental Protection Agency,
Robert S. Kerr Environmental Research Laboratory, Ada, OK, EPA/600/6-86/001, April.
St. John, W. D., and D. J. Sikes. 1988. Complex industrial waste sites. In Environmental
Biotechnology - Reducing Risks from Environmental Chemicals through Biotechnology. G. S.
Omenn, Ed., Plenum Press, New York, NY, p. 163.
R-l
-------
U. S. EPA. 1989. Guide for Conducting Treatability Studies under CERCLA. U. S.
Environmental Protection Agency, Office of Solid and Emergency Response and Office of Research
and Development, Washington, DC, Contract No. 68-03-3413, November.
U. S. EPA. 1990. Handbook on In Situ Treatment of Hazardous Waste-Contaminated Soils. U. S.
Environmental Protection Agency, Risk Reduction Research Laboratory, Cincinnati, OH,
EPA/540/2-90-002, January.
U.S. EPA. 1986. Permit Guidance Manual on Hazardous Waste Land Treatment Demonstrations.
EPA-530/SW-86-032, Office of Solid Waste and Emergency Response, U.S. Environmental
Protection Agency, Washington, DC.
U.S. EPA. 1991. On-Site Treatment of Creosote and Pentachlorophenol Sludges and
Contaminated Soil. EPA/600/2-91/019. Extramural Activities and Assistance Division, Robert S.
Kerr Environmental Research Laboratory, Ada, OK May.
Soil Properties References
Dragun, J. The Soil Chemistry Of Hazardous Materials. Hazardous Materials Control Institute,
Silver Spring, Maryland.
Foth, H.D. 1990. Fundamentals Of Soil Science, Eighth Edition. John Wiley & Sons, New York,
NY.
McLean, Joan C., and Bert E. Bledsoe. Behavior of Metals in Soils. Ground Water Issue.
EPA/540/S-92/018. October 1992. Superfund Technology Support Center For Ground Water,
Robert S. Kerr Environmental Research Laboratory, Ada, OK
Paul, E.A., and F.F. Clark. 1989. Soil Microbiology and Biochemistry. Chapter 2 - Soil as a
Habitat for Organisms and Their Reactions. Academic Press, San Diego, CA.
Monitoring References
Blackwood, Larry G. Assurance Levels Of Standard Sample Size Formulas: Implications For Data
Quality Planning. Environmental Science And Technology, Vol. 25, No. 8, 1991. ( A short article
discussing the shortcomings of standard formulas for calculating the number of samples to be
taken to achieve a desired confidence limit, and giving a reference for a better method.)
Eklund, Bart. Practical Guidance for Flux Chamber Measurements of Fugitive Volatile Organic
Emission Rates. J. Air Waste Management Association 42:1583-1591. December 1992.
Hawley-Fedder, Ruth; and Brian D. Andresen. Sampling And Extraction Techniques For Organic
Analysis Of Soil Samples. UCRL-ID-106599. February 12, 1991. Lawrence Livermore National
Laboratory. (A short discussion of various extraction and analytical techniques.)
Gilbert, R.O. Statistical Methods For Environmental Pollution Monitoring. 1987. Van Nostrand
Reinhold. ISBN 0-442-23050-8. (A textbook for environmental monitoring statistics, with worked
examples.)
R-2
-------
Gilbert, R.O.; and J.C. Simpson. An Approach For Testing Attainment Of Soil Background
Standards At Superfund Sites. (American Statistical Association 1990, Joint Statistical Meetings,
Anaheim, CA. August 6-9, 1990.) Pacific Northwest Laboratory, Richland, WA 99352. (A long
article discussing two statistical tests, and giving examples of their application to various scenarios
at Superfund sites).
Keith, Lawrence H., Ed. Principles of Environmental Sampling. 1988. American Chemical
Society. (A collection of articles on various aspects of sampling techniques and statistics.)
T.E. Lewis, A.B. Crockett, R.L. Siegrist, and K. Zarrabi. Soil Sampling And Analysis For Volatile
Organic Compounds. Ground-Water Issue. EPA/540/4-91/001. February 1991. Superfund
Technology Support CenteT For Monitoring And Site Characterization. USEPA Environmental
Monitoring Systems Laboratoiy. Las Vegas, NV. (An extended discussion of soil sampling and
shipping techniques and devices for minimization of volatiles loss.)
U.S. EPA. A Guide: Methods For Evaluating The Attainment Of Cleanup Standards For Soils And
Solid Media. Quick Reference Fact Sheet. Publication: 9355.4-04FS. July 1991. Office Of
Emergency And Remedial Response, Hazardous Site Control Division OS-220W.
U.S. EPA. Permit Guidance Manual on Unsaturated Zone Monitoring for Hazardous Waste Land
Treatment Units. EPA/530-SW-86-040. Environmental Monitoring Systems Laboratory, Las
Vegas, NV 89114. October.
U.S. EPA. 1991. Handbook of Suggested Practices For The Design and Installation of Ground-
Water Monitoring Wells. EPA/600/4-89/034. Environmental Monitoring Systems Laboratory, Las
Vegas, NV 89193-3478. March.
Petroleum Contaminant References
Drews, A. W., Ed. Manual on Hydrocarbon Analysis: 4th Edition. ASTM Manual Series: MNL 3.
ASTM, 1916 Race Street, Philadelphia, PA. 19103.
Hoffman, H. L. Petroleum - Petroleum Products. Kirk-Othmer Encyclopedia of Chemical
Technology. 3rd Ed. Vol 17. Gulf Publishing Company.
Miller, Michael W. and Dennis M. Stainken. An Analytical Manual for Petroleum and Gasoline
Products for New Jersey's Environmental Program. In Petroleum Contaminated Soils. Volume 3.
Paul T. Kostecki and Edward J. Calabrese. Technical Editor Charles E. Bell. Lewis Publishers.
1990.
Wood Preserving Contaminant References
USDA. 1980. The Biologic and Economic Assessment of Pentachlorophenol, Inorganic Arsenicals,
Creosote. USDA, Number 1658-11. Washington, DC.
R-3
-------
A Technology Overview of Existing and Emerging Environmental Solutions for Wood Treating
Chemicals. December 1990. National Environmental Technology Applications Corporation.
University of Pittsburgh Applied Research Center.
Becker, G. 1977. Experience and Experiments with Creosote for Crossties. Proc. Am. Wood-Pres.
Assoc. 73:16-25.
Bevenue, A. and H. Beckman. 1967. Pentachlorophenol: A Discussion of Its Properties and Its
Occurrence as a Residue in Human and Animal Tissues. Residue Rev. 19:83-134.
Buser, H. R. 1975. Polychlorinated Dibenzo-p-dioxins, Separation and Identification of Isomers by
Gas Chromatography - Mass Spectrometry. J. Chromatog. 114:95-108.
Buser, H.R. 1976. High Resolution Gas Chromatography of Polychlorinated Dibenzo-p-dioxins and
Dibenzofurans. Anal. Chem. 48:1553.
Crosby, D. G. 1981. Environmental Chemistry of Pentachlorophenol. Pure Appl. Chem. 53:1052-
1080.
DaRos, B., R. Merrill, H. K. Willard, and C. D. Wolbach. Emissions and Residue Values from
Waste Disposal During Wood Preserving. Project Summary. EPA-600/S2-82-062. August 1982.
USEPA. 1978. Report of the Ad Hoc Study Group on Pentachlorophenol Contaminants.
Environmental Health Advisory Committee. Science Advisory Board, Washington, DC.
Hoffman, R.E., and S. E. Hrudley. Evaluation of the Reclamation of Decommissioned Wood
Preserving Plant Sites in Alberta. Waste and Chemicals Division and H. E. L. P. Project, Alberta
Environment.
JRB Associates, Inc. Wood Preserving: Preliminaiy Report of Plants and Processes. 1981.
National Institute for Occupational Safety and Health.
Lorenz, L. R. and L. R. Gjovik. 1972. Analyzing Creosote by Gas Chromatography: Relationship
to Creosote Specifications. Proc. Amer. Wood Pres. Assoc. 68:32-42.
Micklewright, James T. Wood Preservation Statistics, 1988. A Report to the Wood-Preserving
Industry in the United States. 1990 Proceedings of the American Wood Preservers' Association.
Nicholas, Darrel D., Ed. Wood Deterioration and Its Prevention by Preservative Treatments.
Volume II: Preservatives and Preservative Systems. Syracuse University Press. 1973.
Thompson, Warren S., and Peter Koch. Preservative Treatment of Hardwoods: A Review. USDA
Forest Service. General Technical Report SO-35. 1981.
R-4
-------
APPENDIX A
SOIL PROPERTIES
AND
BIOREMEDIATION
-------
SOIL PROPERTIES
Soils are composed of organic matter, inorganic solids (sand, silt, clay, and larger fragments), air,
and water. Organic matter may range from less than one percent in many soils, especially those in
hot or cold desert climates, to 50% or more in the peat soils found in peat bogs. Soils are generally
classified according to their sand, silt and clay content; the ratios of these components may vary in
almost any proportion. Air and water occupy the pore spaces among the sand, silt and clay particles.
Pore space occupies about 20-60% of most uncompacted soils.
Soil parameters important in land treatment include: soil horizons, depth, texture (grain size
distribution - sand, silt, clay proportions), bulk density, porosity (effective, total), hydraulic
conductivity, permeability, tilth, cation and anion exchange capacity, organic matter content, pH,
water content and water holding capacity, nutrient content, salinity, redox potential, color, and
biological activity.
Soil Horizons
Soil horizons are the various layers present in most soil columns. Physical and chemical
differences between soils in the layers affect movement of contaminants through the soil profile.
Organic matter from dead plants and animals accumulates in the upper , "A" horizons. A top layer
of almost pure organic matter is called the "0" horizon. As soils mature, clay particles are moved
along with water down through the soil. The clay tends to accumulate in lower soil layers. The zone
of clay accumulation is called the B horizon. The C horizons comprise the relatively undifferentiated
material, often distinguished from the parent rock only by the lack of consolidation.
Depth
Soil depth to bedrock or ground water affects the volume of soil that may be contaminated,
potential directions of contaminant movement, and the difficulty of access to the contaminated volumes
of soil. Land treatment uses the cocept of a "treatment zone", meaning a zone in which migrating
contaminants are adsorbed and degraded. The depth and activity of this zone affects the potential for
migration of contaminants to ground water.
Texture
Soil texture (as defined by the proportions of sand, silt, clay) influences porosity, hydraulic
conductivity, permeability, tilth, cation exchange capacity (CEC), and sorption capacity for
contaminants. Finer textured soils have greater surface areas per unit volume. The differences
between the chemical and physical properties of the various sands and silts is largely due to the
different particle sizes. Clays are not only much smaller than sands and silts but also are quite
different in chemical composition. Clay particles have negatively charged surfaces that attract and
hold cations or other materials with a positively charged portion, giving rise to a cation exchange
capacity. The edges of the clay particles may also have a positive charge, giving rise to a anion
exchange capacity. Clay particles are flat, platelike structures, with a very high surface area. Clay
particles have interior layers that can separate enough to allow water and many ions to enter and be
held. "Shrink-swell" clays allow much water to enter these interior areas, causing the clay particles
to change greatly in volume as the moisture content changes.
A-l
-------
Table A-l. Size, Number And Surface Area Of Soil Particles
Diameter
Particles
Surface Area
Particle Size
(mm)
per gram
(cmVg)
Very Coarse Sand
2.00-1.00
90
11
Coarse Sand
1.00-0.50
720
23
Medium Sand
0.50-0.25
5,700
45
Fine Sand
0.25-0.10
46,000
91
Very Fine Sand
0.10-0.05
722,000
227
Silt
0.05-0.002
5,776,000
454
Clay
<0.002
90,260,853,000
8,000,000
Bulk Density
The soil bulk density is the mass of dry soil per unit bulk volume. The bulk volume is determined
before drying the soil to obtain the mass. Bulk density is used in most soil transport and fate models.
Porosity, Hydraulic Conductivity, and Permeability
Porosity, hydraulic conductivity, and permeability are three parameters that are closely related
and commonly confused. The terms describe the soil characteristics and rate of water movement
through the soil.
Porosity is the ratio of the volume of void spaces in a soil to the total volume of the soil. The void
spaces may be occupied by air, water, or other fluids, such as contaminants. The effective porosity
represents the interconnection between the void spaces and is defined as the volume of void spaces
through which water or other fluids can travel, divided by the total volume of the soil.
Primary porosity is a characteristic of the original soil or rock matrix. Secondary porosity is
caused by weathering or fracturing processes occurring after the soil or rock was emplaced. Secondary
porosity can greatly enhance the effective porosity of the soil or rock.
Typically, more rounded particles such as gravel, sand, and silt have lower porosities than soils
rich in the platy clay minerals. Soils containing a mixture of grain sizes will also exhibit lower
porosities. The smaller particles tend to fill in void spaces between the larger ones.
Porosity can be an important controlling influence on hydraulic conductivity, which is a
proportionality constant describing the rate at which water can move through the soil. Hydraulic
conductivity is a function of the properties of both the porous medium and the fluid passing through
it. Typically, the hydraulic conductivity has higher values for gravel and sand and lower values for
clay. Thus, even though clay-rich soils usually have higher porosities than sandy or gravelly soils,
they usually have lower hydraulic conductivities, because the pores in clay-rich soil are much smaller.
A-2
-------
The hydraulic conductivity can vary over 13 orders of magnitude, depending on the type of
material and whether the measurement was made in the field or in the laboratory. The methods of
measurement differ significantly, and interpretations placed on the values may be dependent on the
type of measurement. In practical terms, this implies that an order-of-magnitude knowledge of
hydraulic conductivity may be all that is attainable, and that decimal places beyond the second
probably have little significance.
Pore spaces may be classified according to size as micropores and macropores. Porosity of sandy
soils largely consists of macropores, while porosity of clay soils is largely micropores. The ratio of
micropores to macropores influences the movement of soil gases and water in the soil and is of
particular importance for bioremediation, since the ratios and interactions of soil gas and water greatly
influence microbial activity.
Permeability describes the conductive properties of a porous medium independently of the fluid
flowing through it. It includes the influence of media properties that affect flow, including the grain
size,distribution and roundness, and the nature of their arrangement. Permeability is widely used in
situations where multiphase flow systems (vapor, water, and non-aqueous phase liquids) are present.
These conductive properties determine the feasibility of adding or removing materials such as
water, air, and nutrients to the soil. Soil hydraulic conductivities of about 7.0 X 10"3 to 7.0 X 10 5
cm/sec are favorable for adding or removing materials. Soils with conductivities above this range may
require careful management to prevent excessive drainage or contaminant mobility for some
remediation technologies: in soils with conductivities below this range it may be difficult to add or
remove materials for remediation.
The hydraulic conductivity of saturated soils is a function of the grain size and sorting of the
particulate materials, and therefore, is somewhat stable over time. Hydraulic conductivity in
unsaturated soil is not only influenced by grain size and sorting but also is strongly influenced by
water content of the soil. At low soil water content, soil water moves largely in response to adhesive
and cohesive forces in the soil, which are measured as matric potential. Soluble contaminants in
unsaturated soil move in the thin films of water surrounding the soil particles. The thicker the film
of water (e.g., the wetter the soil), the larger the conduit for contaminant movement, and more of the
contaminant that can move in a given period of time.
Movement of contaminants in the vadose zone is usually in the soil gas, pore water or as non-
aqueous phase liquids (NAPLs) with water vapor and other gases generally at atmospheric pressure.
Soil gases may move into the atmosphere, ground water, soil pore water, be adsorbed on soil particles
or undergo chemical/biological transformation. Dissolved contaminants in soil water undergo similar
changes. NAPLs move in response to gravity and changes in soil permeability.
Soil Moisture and Water Holding Capacity
Soil moisture holding capacity is determined by the proportion of clay and organic matter in the
soil. Clays and organic matter tend to hold larger amounts of water relative to their volume than do
the coarser grained silts and sands. When a soil is saturated with water, then allowed to drain freely
for 24 hours, the soil is said to be at field capacity. Essentially this means that the soil micropores
are filled with water, and the macropores are filled with air.
A-3
-------
The ratio of air and water in the soil strongly influences many important processes in the soil.
Aerobic microbial activity in the soil is usually optimum when soil moisture is about 70 - 80% of field
capacity; the higher end of the range is more desirable for coarser soils. Relatively dry soils tend to
adsorb many contaminants more strongly than wetter soils, since water competes with the
contaminants for soil adsorption sites. When the soil is not saturated, water and water soluble
compounds may move in any direction in the soil in response to matric potential, whereas water in
saturated soils moves largely in response to gravity. Water and water soluble compounds move faster
through wetter soils than drier soils. Very dry soils, especially soils with high organic matter content,
may be very difficult to wet since dry organic matter tends to be hydrophobic. NAPLs may move
through moderately wet soils faster than either dry or very wet soils, since dry soil tend to adsorb
much of the NAPL and the pores full of water in wet soils inhibit NAPL movement.
Tilth
Tilth refers to ability of the soil to undergo manipulation (plowing, tilling) and retain a desirable
loose, friable structure that promotes ready movement of water and air into the soil. Structure refers
to the tendency of soil to agglomerate into structural units called peds, granules or aggregates. In
surface soils, this agglomeration is influenced by the microbial secretion of polymeric materials that
cement soil materials together into small particles. High levels of sodium in the soil (measured as
exchangeable sodium percentage (ESP) or sodium absorption ratio (SAR)) may disperse the soil
particles causing a loss of structure. Sodium absorption ratios higher than 15 may indicate a problem,
as do ESP values greater than 10% of the CEC in fine textured soils and greater than 20% of the CEC
in coarse textured soils.
Sorptive and Exchange Capacity
Clay materials in soils generally have a high adsorptive capacity for many organic and inorganic
materials. Coarse, sandy soils may allow rapid movement of relatively small amounts of contaminants
into lower soil layers and aquifers, while soils high in clay may significantly retard movement of many
contaminants. Inorganics and some organics may be influenced by the cation exchange capacity
(CEC), which denotes the capacity of the clay particle for adsorption of positively charged materials
on the negatively charged surfaces of the clay. Mobility of metals in the soil may be greatly affected
by the CEC. Clays may have an anion exchange capacity due to the positive charge on the edges of
some clay particles. The anion exchange capacity is usually less than the cation exchange capacity.
Clays also will adsorb uncharged molecules due to Van der Waals interactions of of the uncharged
materials with clay particles.
Organic Matter
Soil organic matter is generally composed of 25-35% polysaccharides and protein-like compounds
which are readily decomposed by microorganisms and therefore have a short halflife in soils. About
65-75% of the soil organic matter is composed of humic materials, which are complex mixtures of high
molecular weight organics and are resistant to degradation. These percentages do not include those
organic compounds that may be present as contaminants; i.e., oil and grease, volatile organics, etc.
Soils high in organic matter will adsorb significant quantities of organic contaminants, since organic
compounds have a strong tendency to adsorb onto soil organic matter, thereby slowing movement. Soil
organic matter usually has a relatively high CEC and may have a significant anion exchange capacity,
although anion exchange capacity is usually much less than the CEC. Increased soil organic levels
are generally favorable to microbial activity, due to increased CEC, tilth, water holding capacity, and
available carbon. Soil organic matter levels tend to be lower in warm, moist climates, since these
conditions allow rapid microbial oxidation of the organic matter. Soil organic matter may be increased
by addition of straw, hay, sawdust or wood chips, manures, and many other organic materials.
A-4
-------
Addition of easily transformed organic materials may cause shortages of nutrients (particularly
nitrogen and phosphorus) due to the increased microbial population feeding on the added organic
matter.
pH
The pH of the soil affects microbial activity, availability of nutrients, plant growth, immobilization
of metals, rates of abiotic transformation of organic waste constituents, and soil structure. A pH range
of 6-8 is considered optimum for bioremediation in most cases. Most metals tend to be less mobile in
high pH soils (arsenic is an exception), but acidic organics such as pentachlorophenol are more mobile.
Soils with high sodium levels and high pH (most often found in dry climates) tend to deflocculate and
crust, limiting oxygen diffusion and water uptake. Soil pH may be lowered by addition of ferrous or
aluminum sulfate, elemental sulfur or sulfuric acid; soil pH may be raised by addition of agricultural
lime.
Nutrients
Nutrient content relates to the concentration of nutrients available for use by microorganisms.
Nitrogen and phosphorus often limit microbial activity in soils. An organic carbon:nitrogen:phosphorus
ratio of 100-300:10:1 is recommended to stimulate microbial activity, with the lower C:N ratios
recommended when most of the carbon is in a readily degradable form. The percent base saturation,
a general indicator of soil fertility, is defined as the total of the four principal exchangeable bases
(calcium, magnesium, sodium, potassium) divided by the total exchange capacity of the soil. A base
saturation of about 80% is desirable, with calcium comprising about 60-70% of the CEC and potassium
about 5-10% of the CEC.
Most soils have low levels of nitrogen, although soils with high levels of organic matter may have
significant amounts of nitrogen as part of the organic matter; this nitrogen is usually released slowly
as the organic matter decomposes. Inorganic nitrogen in the soil is usually quite water soluble and
therefore readily lost to leaching, which may cause ground water pollution problems. Since
microorganisms benefit from a steady supply of nitrogen, it is advantageous to supply nitrogen either
at a slow steady rate or in a form (e.g., as organic fertilizers or "slow-release" inorganic fertilizers) that
supplies nitrogen to the microorganisms slowly.
Many soils contain significant quantities of phosphorus, but the phosphorus may be strongly bound
in the soil, and little may be readily available to the microorganisms. Usually bound phosphorus is
in equilibrium with phosphorus dissolved in the soil water; the equilibrium is heavily weighted toward
the bound form. For this reason, phosphorous fertilizers are often applied to raise the amount of
phosphorus in the soil water.
Salinity
The electrical conductivity (EC) of a soil reflects the soluble salt content (salinity). An EC of 2 or
less indicates that salinity is not a problem in most instances. An EC of 2-4 may inhibit activity of
very salt-sensitive microorganisms, while an EC of 4-8 may restrict activity of many microorganisms.
An EC greater than 8 will restrict activity of most microorganisms.
A-5
-------
Redox Potential
The redox potential of the soil (oxidation-reduction potential, reported as Eh) is controlled by the
concentration of 02 in the gas and liquid phases. The 02 concentration is a function of the rate of gas
exchange with the atmosphere and the rate of respiration in the soil. Respiration in the soil may
deplete 02, lowering the redox potential and creating anaerobic (reducing) conditions. These conditions
are unfavorable to aerobic biotransformation, but may promote anaerobic processes such as reductive
dehalogenation. Many reduced forms of polyvalent metal cations are more soluble (and mobile) than
their oxidized forms. Well aerated soils have an Eh of about 0.8 to 0.4 volts; moderately reduced soils
are about 0.4 to 0.1 V; reduced soils are about 0.1 to -0.1 V; and highly reduced soils are about -0.1
to -0.3 V. Redox potentials are difficult to measure and are not widely used in the field.
Color
The color of soils is largely due to chemical changes and organic matter content. Dark colors in
soil are caused by highly decayed organic matter. Reds and yellows are caused by oxidized and
hydrated iron in soil minerals. Uniform reds, yellows, and browns indicate that a soil is well drained.
Mottled grays or blues may indicate poor drainage. The location of any mottled layers may indicate
the level of the seasonal high water table.
Biological Activity
Biological activity in the soil is affected by all of the soil characteristics discussed in this Appendix.
Biological activity apparently accounts for most of the transformation of organic contaminants in soil.
Both bacteria and fungi have been shown to be important in bioremediation processes. Most
research in bioremediation has centered on bacteria, but some investigators have found that fungi can
play an important role in bioremediation processes, especially with halogenated compounds (e.g.,
pentachlorophenol). In most cases bioremediation relies on communities of microorganism species,
rather than one or a few species. Bioremediation consists of utilizing techniques for enhancing
development of large populations of microorganisms that can transform pollutants of interest, and
bringing these microorganisms into intimate contact with the pollutants. In order to do this efficiently,
necessary conditions for the growth and activity of the microorganisms must be maintained.
Microbial activity in the soil can be estimated by using plate counts, most probable number (MPN)
counts, direct microscopic counts, respiration measurements, ATP activity measurements, and others.
Unfortunately, the relationship of these measurements to practical use of bioremediation techniques
is unclear, at best. Generally use of these measurements is limited to determining if soil conditions
or waste characteristics are suitable for microbial activity, and whether particular management
techniques have enhanced microbial activity.
By culturing soil microorganisms on special media, counts of "specific degraders" can be
determined. For instance, if PAHs are added to a media with no other carbon sources present, any
microorganisms that grow can be assumed to have the capability of using PAHs as a sole source of
carbon. Again, the relationship of these counts to actual biodegradation in the field is unclear.
A-6
-------
If biodegradable contaminants have been present in the soil for more than a few months or years,
and microorganisms are able to grow and reproduce in the contaminated soil, microorganisms that can
transform the wastes are likely to be present. Treatability studies can be used to determine
techniques that might be appropriate to optimize their transforming activity, as well as determine if
the microorganisms are capable of transforming the wastes to acceptable levels of acceptable end
products in an acceptable time frame. There is little or no well documented evidence to show that
addition of cultured microorganisms (bioaugmentation) to the soil in field situations enhances natural
bioremediation processes.
Bioaugmentation commonly takes two forms. Microorganisms may be isolated from the site in
question, cultured in quantity and added to the site soil, or microorganisms isolated from other sites
may be cultured and added to the site soil. It is very difficult to show that added microorganisms
survive and grow in the soil, and even more difficult to show that the added microorganisms have any
significant affect on transformation.
Metals in Soils
The mean concentrations of metals commonly found in uncontaminated soils are shown in Table
A-2. High concentrations of certain metals (particularly the "heavy" metals lead, mercury, cadmium,
chromium and others) are known to inhibit microorganism activity in laboratory studies, but the
particular levels of metals that would be of significance in field bioremediation are not known with
certainty. The influence of metals concentration on bioremediation appears to be site, contaminant,
and microorganism dependent. In cases where high concentrations of metals appear to be of concern,
treatability studies should be conducted to determine the influence of metals concentrations on
bioremediation.
Table A-2. Mean Concentration (mg/kg) of Metals In The Earth's Crust and Soils a
A1
Fe
Mn
Be
Cu
Cr
ca
Zn
As
Se
Ba
Ni
Ag
Pb
Hg
Soils
72000
26000
550
0.92
25
54
0.35
60
7.2
0.39
580
19
0.05
19
0.09
Crust
82000
41000
350
2.6
50
100
0.11
75
1.5
0.05
500
80
0.07
14
0.05
A-7
-------
United States Office of Office of Solid Waste EPA/540/4-90/053
Environmental Protection Research and and Emergency October 1990
Agency Development Response
&EPA Ground Water Issue
Basic Concepts of Contaminant
Sorption at Hazardous Waste Sites
Marvin D. Piwoni* and Jack W. Keeley"
Introduction
The Regional Superfund Ground Water Forum is a group of
ground-waterscientists, representing EPA's Regional Superfund
Offices, organized to exchange up-to-date information related to
ground-water remediation of Superfund sites. One of the major
issues of concern to the Forum is the transport and fate of
contaminants in soil and ground water as related to subsurface
remediation. Processes which influence the behavior of
contaminants in the subsurface must be considered both in
evaluating the potential for movement as well as in designing
remediation activities at hazardous waste sites. Such factors not
only tend to regulate the mobility of contaminants, but also their
form and stability. Sorption is often the paramount process
controlling the behavior of contaminants in the subsurface. This
paper summarizes the basic concepts of sorption in soil and
ground water with emphasis on nonpolar organic contaminants.
For further information contact: Joe Williams, FTS 743-2246;
Bert Bledsoe, 743-2324; or DomDiGiulio, 743-2271 atRSKERL-
Ada.
The Concept of Sorption
Sorption can be defined as the interaction of a contaminant with
a solid. More specifically, the term can be further divided into
adsorption and absorption. The former refers to an excess
contaminant concentration at the surface of a solid while the latter
implies a more or less uniform penetration of the solid by a
contaminant. In most environmental settings this distinction
serves little purpose as there is seldom information concerning
the specific nature of the interaction. The term sorption is used
in a generic way to encompass both phenomena.
There are a number of factors which control the interaction of a
contaminant and the surface of soil or aquifer materials. These
/ T ^
U -Lochmcal £
Muppon
| 'ojecl .
include chem ical and physical characteristics of the contaminant,
composition of the surface of the solid, and the fluid media
encompassing both. By gaining an understanding of these
factors, logical conclusions can often be drawn about the impact
of sorption on the movement and distribution of contaminants in
the subsurface. The failure to take sorption into account can
result in a significant underestimation of the amount of a
contaminant at a site as well as the time required for it to move
from one point to another.
In introducing sorption theory it is necessary to define the terms
sorbate and sorbent. The sorbate is the contaminant that
adheres to the sorbent, or sorbing material. In this discussion the
sorbate will usually be an organic molecule and the sorbent will
be the soil or aquifer matrix.
This Issue Paper is condensed from a presentation given at the
EPA/EPRI Workshop on Leachate Testing Methods in Houston,
Texas, in January 1989.
Factors Influencing Sorption
The properties of a contaminant have a profound impact on its
sorption behavior. Some of these include:
• Water Solubility
• Polar/tonic Character
• Octanol/Water Partition Coefficient
• Acid/Base Chemistry
• Oxidation/Reduction Chemistry
* Laboratory Services Manager, Hazardous Waste Research and
Information Center, Illinois Department of Energy and Natural
Resources.
" Environmental Engineer, Dynamac Corporation, Robert S. Ken-
Environmental Research Laboratory.
Superfund Technology Support Center for Ground Water
Robert S. Kerr Environmental
Research Laboratory
Ada, OK
Printed on Recycled Paper
-------
Contaminant Characteristics
In discussing sorption it is useful to divide chemicals into three
groups. Although there are many ways to divide chemicals into
subgroups, for this purpose three categories are presented
which transcend normal boundaries between inorganic and
organic species. These are: (1) ionic or charged species; (2)
uncharged polar species; and, (3) uncharged nonpolar species.
Most inorganic chemicals in aqueous solution will occur as ionic
or charged species. This applies to metals and metalloids, and
to other molecules such as cyanide and ammonia. However, in
contaminated water, metals and other inorganic constituents can
exist as polar or nonpolar neutral species. In any event, the
chemical form of a contaminant will have a profound effect on its
sorption and, therefore, its environmental mobility.
Organic contaminants have representatives in all three of the
sorption categories. Many of the more common organic ground-
water contaminants are of the nonpolar species, including
trichloroethene (TCE), tetrachloroethene (PCE), the chlorinated
benzenes, and the more soluble components of hydrocarbon
fuels such as benzene, toluene and xylene. Other important
organic contaminants including many of the pesticides, phenols
and dyes exist in solution as either charged or polar molecules.
Still other, larger organics, such as surfactants, can have both
polar and nonpolar ends within the same molecule. The
environmental mobility of contaminants with these distinctive
properties has been less thoroughly studied than nonpolar
organics; therefore, site-specific investigations may provide the
most reliable information for their transport characteristics.
Soil Characteristics
If one avoids the difference between positive and negative
charges, a simple rule of sorption might be: for charged species,
"opposites attract" and for uncharged species, "likes interact with
likes." Likes refers to the three categories of contaminants and
to the properties of the soil matrix. Some of the most important
characteristics of soil affecting the sorptive behaviorof subsurface
materials include:
• Mineralogy
• Permeability/Porosity
• Texture
• Homogeneity
• Organic Carbon Content
• Surface Charge
• Surface Area
Soil, in its natural state, is primarily composed of sand, silt, clay,
water, and a highly variable amount of natural organic carbon.
The latter profoundly complicates a soil's sorptive properties.
The combination of these characteristics describes the surfaces
offered as sorptive sites to contaminants in water passing through
the subsurface matrix. For example, silts and clays have much
higher surface areas than sand, usually carry a negative charge,
and almost invariably associate with natural organic matter.
It can be deduced that sandy materials offer little in the way of
sorptive surfaces to passing contaminants while silts and clays,
particularly those having substantial amounts of organic matter,
provide a rich sorptive environment for all three categories of
contaminants. Even the most porous and highly productive
aquifers, composed of sands and gravels, usually have some
fine grained material, and a few percent of silts and clays can
result in a substantial increase in the sorptive behavior of the
aquifer material.
Fluid Media Characteristics
Under most contamination situations the primary transporting
fluid is water. One of the most important properties of this solvent
phase is pH for rt dictates the chemical form and, therefore, the
mobility, of all contaminants susceptible to the gain or loss of a
proton. As an example, pentachlorophenol will primarily be an
uncharged polar molecule in an aqueous solution whose pH is
below about 4.7 and an anion when the pH is above that value,
increasing its solubility from 14 to 90 mg/l.
Other characteristics of water that can influence the behavior of
contaminants include the salt content and the dissolved organic
carbon content. Chlorides, for example, which are not usually of
much concern when dealing with organic contaminants, can
have an important effect on the mobility of various metals.
Dissolved organic matter, at relatively high concentrations found
in many leachates, has a significant effect on the mobility of most
nonpolar organics.
implications of These Characteristics
Although somewhat simplified, it can be assumed for purposes
of this discussion, that charged and polar species tend to interact
with charged and polar surfaces, and nonpolar compounds
interact with nonpolar components of soil, usually the natural
organic carbon, in order to make af irst estimate of the significance
of sorption at a site, it is necessary to determine the polar and
nonpolar nature of the material with which the contaminant will
come into contact. This is usually done by measuring the cation
exchange capacity and the natural organic carbon content,
respectively.
The cation exchange capacity (CEC) provides an estimate of the
total negatively-charged sites on the surface of the soil. It is
determined by measuring the mass of a standard cation, usually
ammonia, that displaces another cation held by the soil. Under
normal field conditions these sites will be occupied by cations
common to the f lowing or percolating water, such asNa*, K', Ca2*,
and Mg2*. Larger organic cations and highly-charged metal ions
like Hg2* or Cr** will be preferentially retained at these sites by
"exchanging" with their normal occupants. Thus large organic
cations and heavy metals would not normally be expected to
move far through soils with a measurable cation exchange
capacity.
At contaminated sites, however, conditions may not be "normal"
and Hg2* may be codisposed with high levels of chbride salts. In
the complexation chemistry shown in Figure 1, Hg2* may be
replaced by the neutral complex HgCI2orthe negative ion HgCI^,
both of which move through the soil more quickly than the cationic
form.
Sorption of Nonpolar Organics
As mentioned above the chemicals at many contaminated sites
are nonpolar organics. h was representatives of these types of
compounds (DDT and other chlorinated hydrocarbon pesticides)
2
-------
that first focused attention on the potential hazards of chemicals
in the environment because of their widespread use, potential
human toxicity, and recalcitrance.
Transport and fate characteristics of these compounds have
been well studied, first by the agricultural community and later by
environmental scientists. As a result, an understanding of the
sorptive behavior of these compounds has evolved which can be
used to assess the environmental consequences posed at a
waste disposal site.
Many organics of environmental concern have a limited solubility
in water because of their nonpolarity and molecular size: that is,
the solubility of an organiccontaminant decreases with decreasing
polarity and increasing molecular size. But even with limited
solubilities, many hazardous chemicals at equilibrium are at
measurable, and sometimes toxic concentrations in water. Polar
molecules, such as ethanol, are compatible with water. Their
combination results in a homogeneous solution regardless of the
proportions that are mixed.
Nonpolar organic compounds interact with soil organic matter
through a process known as "hydrophobic sorption" which can be
explained as the affinity of organic compounds for phases other
than water. For example, water being a polar molecule is not
compatible with other nonpolar molecules, such as DDT, which
is immiscible with water.
Octanol-Water Partitioning
Organic molecules of increasing size, decreasing polarity and
therefore water solubility, are said to exhibit increasing
"hydraphobicity" which can be quantified by their octanol-water
partition coefficient. It is a measure of the distribution of the
chemical between a water and an organic (octanol) phase with
which it is in contact. The more hydrophobic the contaminant, the
more likely it is to partition into the octanol phase. The partition
coefficient provides a fairly accurate understanding of the sorptive
process occurring between water and the soil, more specifically,
the soil organic matter.
The octanol-water partition coefficient, expressed as in
Figure 2, is determined by measuring the concentration of a
particular compound in the water and the octanol phases after a
period of mixing. It is important to note that the more hydrophobic
the compound the less accurate the test, and the results should
Octanol-Water Partition Coefficient:
Concentration
Concentration watr
Almost always presented as Log lc because the
numbers are so large for hydrophobic compounds.
Sorption Coefficient:
Concentration
Concentration
mg/kg
Units are , which is Ukg.
mg/L
Carbon Normalized Sorption Coefficient:
Sorption Coefficient, Kp
Fraction Organic Carbon
Figure 2. Relationships Pertinent to Nonpolar Organic
Contaminant Transport
be viewed accordingly. It is often sufficient to know that an
extremely high coefficient means that the compound is very
hydrophobic. Since measured values can be in the millions
for important environmental contaminants (PCB's, chlorinated
pesticides, dioxins and furans), they are often expressed as the
baselO logarithm, Log K^.
The has two attributes that make it especially useful in
environmental assessments. First, it varies in a predictable way
within classes of organic compounds. For example, as shown in
Figures 3 and 4, if is known for one member of a class of
compounds it can be used reasonably well to estimate a value for
other members of the same family. In the examples shown, the
can be correlated to the number of chlorine atoms or the
number of rings in the molecular structure of classes of
contaminants.
The second attribute results from the work of a number of
agricultural and environmental researchers who correlated
sorption on the organic matter of soils with the K of the
compounds involved. By using these attributes of the"k , it is
possible to estimate thepotential sorption of organic contaminants
based on the structure of the compounds and the organic carbon
content of the soil or aquifer material.
Sorption to Soils
Thus far it has been suggested that nonpolar organic compounds
are sorbed by soils as a function of their hydrophobicity (KJ and
the organic carbon content of the soil. There has°"been
considerable research which suggests that the sbw kinetics of
the sorption process may be significant in swiftly moving ground
water. Sorption studies using flow-through columns produce
results sensitive to the flow rate, and batch tests indicate that
Hg2+ +CP *4
~ HgCI+
HgCI+ + CI*
~ HgClj
HgCI° + CI"
~ HgClj
^ iiyCi4
The complexation reactions are driven right and down by
increasing chloride concentrations, often characteristic of
waste waters. Increased complexation produces increased
environmental mobility of the mercury.
Figure 1. Mercury Ion Complexatlon in Chlorlde-Rlch Water
3
-------
J
I
Chlorinated
Banzsrw*
2 3 4 5 6
* of Chlorlns Atoms
Figure 3. Relationship of Molecular Structure to
Hydrophobic Character
increased sorption occurs with longer exposure times. The
practical implication of these findings may be that sorption is
overestimated in aquifer systems with relatively high flow rates.
Sorption is expressed in terms of a partition coefficient K , which
is defined in Figure 2 as the ratio of the concentration of
contaminants associated with the solid phase to that in solution,
and is, therefore, conceptually similar to Km. The usefulness of
Km in estimating sorption stems from the fact that the soil organic
matter serves the same function as octanol in the octanol-water
test. As a result, there have been many empirical relationships
developed for estimating sorption from the Km and the soil
organic carbon content. One expression, developed in the
laboratory by Piwoni and Banerjee, 1989, for the sorption of
common environmental contaminants with a low aquifer organic
carbon, is:
Log Kk - 0.69 Log + 0.22
When applying such a relationship, it is important to select a
study in which the compounds used are similar to those of
interest at the site under investigation. However, as shown in
Figure 5, even when applying the empirical relationship to a
structurally dissimilar compound such as anthracene, if it is a
nonpolar organic, the error of estimate should be less than a
factor of five.
These estimates of sorption are based, in large measure, on a
good evaluation of the soil organic carbon content at a site which
is obtained from the degradation of naturally occurring organic
matter. In this regard it is important to realize that soils and aquifer
materials are very heterogeneous and the organic carbon content
can vary considerably both in the vertical and horizontal dimension.
Fortunately, this variability tends to be the greatest in the vertical
soil profile while most site investigations are concerned with
contaminant movement in the ground water away from the
source. While the soil organic carbon content in the horizontal
plane usually differs by a factor of ten or less, it can vary by a
factor of 10 to 100 in the vertical dimension.
?
Benzo(a)pyrene
Benzanthracene
Oj Naphthalene
I Benzene
Increasingly Hydrophobic
~
3
Rings
Figure 4. Relationship of Molecular Structure to Octanol Water
Partition Coefficient
$
1 3
Chlorinated
Benzenes
Anthracene
0 1 2 3 4 5 6
Log Kow
Figure 5. Partitioning on Soil Organic Carbon as Function of
Octanol-Water Partition Coefficient
4
-------
In order to determine the soil organic carbon content at a site,
samples are usually obtained using split spoon sampler or other
standard soil sampling devices. Representatives portions of the
soil are then burned in an 02 atmosphere and the produced CO.
is measured by IR spectrophotometry. Before burning son
samples must be acidized to remove inorganic carbon. The
accuracy of measuring organic carbon content can also be
questionable, particularly at low levels and in carbonate soils.
Existing analytical methods for measuring soil organic carbon
were developed for the higher concentrations found near the
surface. Therefore, at the low levels found in deeper soils and
ground water, the same quality assurance procedures used in
determining contaminant levels in water should be followed in
determining the subsurface organic carbon content.
The processes driving hydrophobic sorption are nonspecific and
depend upon small amounts of energy gained by moving
contaminants out of the aqueous phase. The extent to which the
process proceeds is dependent upon how receptive the soil
matrix is to the organic molecule, which is a function of the
organic content. But even when the organic carbon content is
very low, some sorption of the most hydrophobic molecules
continues because of the soil's mineral surfaces.
Sorption Estimation
In order to use the information provided above in estimating the
amount of a contaminant associated with the aqueous and solid
phases of an aquifer, it is necessary to develop a contamination
scenario. To that end it is assumed that the contaminant at an
industrial landfill is 1,4-dichlorobenzene, and there is sufficient
data to indicate that: (1) most of the contamination is below the
water table; (2) the contaminant concentration in ground water
averages 1 mg/l; (3) the measured soil organic carbon is 0.2
percent; and (4) the pore water occupies 50 percent of the aquifer
volume. Steps leading to an estimate of the contaminant's
distribution between the aqueous and solid phases are:
Field Measurements:
Average contaminant concentration
in monitoring wells » 1.0 mg/l
Soil organic carbon » 0.2 percent,
therefore f = 0.002
OC
Pore water occupies 50 percent of
the aquifer's volume.
From The Literature:
Log (1,4-dichlorobenzene) = 3.6
Piwoni and Banerjee Regression,
Log = 0.69 K,,, + 0.22
Calculated:
Log = 0.69(3.6) + 0.22 - 2.70
therefore: K^. = 506
Kp- KocCoc) = 506 <0002) = 1-°° Sorbed C
Solution C
Conclusion;
The contaminant, equally distributed between each phase,
is expressed as mg/kg (soil) and mg/l (water). Since soil is
about 2.5 times more dense than water, 2 liters of aquifer
would contain 1 liter of water and 2.5 kg of soil. Therefore,
1.0 mg/l of the contaminant would be associated with the
water and 2.5 mg (70 percent) would be sorbed to the
aquifer's solid phase.
As can be seen from this example, sorption tends to complicate
remediation techniques that require pumping water to the surface
for treatment. The desorption process has kinetic constraints
that can render a pump-and-treat system ineffective. Slow
desorption kinetics result in progressively lower contaminant
concentrations at the surface, and less cost-effective contaminant
removal. It is not uncommon to pump a system until the
contaminant concentration in the pumped water meets a
mandated restoration level, while the aquifer's solid phase still
contains a substantial mass of contaminant. If the pumps are
turned off, concentrations in the ground water will soon return to
their equilibrium level.
Measuring Sorption
h is preferable to obtain the best information possible on which
to base an estimate of sorption. Therefore, tests should be made
with the contaminants of concern, as well as soils and aquifer
material from a specific site. The goal is to obtain a partition
coefficient, K^, for use in the prediction of contaminant movement.
There are essentially two methods for measuring the partition
coefficient, those being batch and dynamic techniques. Batch
techniques are quicker and easier to perform and, therefore,
more amenable to replication and quality control. Dynamic or
flow through techniques offer the advantage of more closely
representing processes occurring in the field.
The standard approach to determine the partition coefficient is to
generate a sorption isotherm, a graphical representation of the
amount of material sorbed at a variety of solute concentrations.
The Freundlich isotherm, S - KpC1/n, is the representation most
often used for the sorption of nonpolar organics to soils and
aquifer materials. In this equation, S is the mass sorbed per mass
of sorbent (mg/kg), C is the solute concentration at equilibrium
(mg/l), Kp is the Freundlich partition coefficient, and 1/n is a fitting
factor. The equation can be expressed in a linear form for
convenience:
Log S «• Log Kp + 1/n Log C
As shown in Figure 6, Log Kp can be estimated by determining
the intercept of the regression of a Log-Log plot of S and C.
Summary
This has been a discussion of the concepts involved in estimating
contaminant sorption, particularly nonpolar organics, at hazardous
waste sites. After determining the types of contaminants present
at a site, it is possible to estimate K using K values from the
literature, an appropriate sorption coefficient/K^ regression
equation, and some organic carbon values.
H sorption determinations are within the scope of the project, site
representative soil samples and contaminantsshould be selected
5
-------
1.00
*— Kp-0.74.Intercept c/
0.50
/
€ 0,10
/
E
VtL
c
.S
~ 0
£ 0.05
%
Jro
°as
0.01
ii ii
0.01 0.05 0.10 0.15 1.00
C-Solution Conoentralon mg/L
Figure 6. Sorption of 1,4-Dlchlorobenzene
from the tests. The measured sorption information is best used
to evaluate the validity of preliminary estimates. If the measured
partition coefficients differ from the estimates by more than a
factor of 2 or 3, it may be useful to select other contaminants from
the site and determine Kp values the same soil samples. A plot
of K values versus values will provide a useful guide for
preaicting the sorption characteristics of other contaminants at
the site.
Selected References
American Society of Agronomy Soil Sciences of America. 1982.
Methods of Soil Analyses. Part 2-Chemical and Microbiological
Properties-Second Addition. Agronomy No. 9, Part 2.
Ballard, T.M. 1971. Role of Humic Carrier Substances in DDT
Movement Through Forrest Soil. Soil Sci. Soc. Am. Proc. 35:145-
147.
Banerjee, P., M.D. Piwoni, and K. Ebeid. 1985. Sorption of
Organic Contaminants to Low Carbon Substrate Core.
Chemosphere 14:1057-1067.
Bouchard, D.C., R.M. Powell, and D.A. Clark. 1988. Organic
Cation Effects on the Sorption of Metals and Neutral Organic
Compounds on Aquifer Material. J. Environ. Sci. Health Part A
23:585-601.
Bouchard, D.C., and A.L. Wood. 1988. Pesticide Sorption on
Geologic Material of Varying Organic Carbon Content. Toxic.
Industr. Health 4:341-349.
Briggs, G.G. 1981. Theoretical and Experimental Relationships
Between Soil Adsorption, Octanol-Water Partition Coefficients,
Water Solubilities, Bioconcentration Factors, and the Parachor.
J. Agric. Food Chem. 29:1050-1059.
Brown, D.S., and E.W. Flagg. 1981. Empirical Prediction of
Organic Pollutant Sorption in Natural Sediments. J. Environ.
Qual. 10:382-386.
Carlson, D.J., LM. Mayer, M.L. Brann, and T.H. Mague. 1985.
Binding of Monomeric Organic Compounds to Macromolecular
Dissolved Organic Matter InSeawater. Mar. Chem. 16:141-163.
Carlson, R.M., R.E. Carlson, and H.L. Kopperman. 1975.
Determination of Partition Coefficients by Liquid Chromatography.
J. Chromatogr. 107:219-223.
Caron, G., H. Suffet, and T. Belton. 1985. Effect of Dissolved
Organic Carbon on the Environmental Distribution of Nonpolar
Organic Compounds. Chemosphere 14:993-1000.
Carter, C.W., and I.H. Suffet. 1982. Binding of DDT to Dissolved
Humic Materials. Environ. Sci. Technol. 16:735-740.
Chin, Y., W.J. Weber, and T.C. Voice. 1986. Determination of
Partition Coefficients and Aqueous Solubilities by Reverse Phase
Chromatography-ll. Water Res. 20:1443-1450.
Chiou, C.T., R.L. Malcolm, T.I. Brinton, and D.E. Kile. 1986.
Water Solubility Enhancement of Some Organic Pollutants and
Pesticides by Dissolved Humic and Fulvic Acids. Environ. Sci.
Technol. 20:502-508.
Dragun, James. 1988. The Soil Chemistry of Hazardous
Materials. Hazardous Materials Control Research Institute.
Silver Spring, MD.
Enfield, C.G. 1985. Chemical Transport Facilitated by Multiphase
Flow Systems. Water Sci. Technol. 17:1-12.
Enfield, C.G., D.M. Walters, R.F. Carsell, and S.Z. Cohen. 1982.
Approximating Transport of Organic Pollutants to Groundwater.
Ground Water 20:711 -722.
Garbarini, D.R., and L.W. Lion. 1986. Influence of the Nature of
Soil Organics on the Sorption of Toluene and Trichloroethylene.
Environ. Sci. Technol. 20:1263-1269.
Gamerdinger, A.P. R.J. Wagonet, and M. th. van Genuchten.
1990. Application of Two-Site/Two-Region Models for Studying
Simultaneous Transport and Degradation of Pesticides. Soil Sci.
Soc. Am. J. 54:957-963.
Gauthier, T.D., W.R. Seitz, and C.L. Grant. 1987. Effects of
Structural and Compositional Variations of Dissolved Humic
Materials on Pyrene K Values. Environ. Sci. Technol. 21:243-
248.
Griffin, R.A. and W.R. Roy. 1985. Interaction of Organic Solvents
with Saturated Soil-Water Systems. Open File Report prepared
for the Environmental Institute for Waste Management Studies,
University of Alabama.
Gschwend, P.M., and S. Wu. 1985. On the Constancy of Sediment-
Water Partition Coefficients of Hydrophobic Organic Pollutants.
Environ. Sci. Technol. 19:90-96.
Hassett, J.P., and M.A. Anderson. 1982. Effect of Dissolved
Organic Matter on Adsorption of Hydrophobic Organic
Compounds by River-and Sewage-Borne Particulate Matter.
Water Res. 16:681-686.
6
-------
Karickhoff, S.W. 1981. Semi-Empirical Estimation of Sorption of
Hydrophobic Pollutants on Natural Sediments and Soils.
Chemosphere 10:833-846.
Karickhoff, S.W., D.S. Brown, and T.A. Scott. 1979. Sorption of
Hydrophobic Pollutantson Natural Sediments. Water Res. 13241-
248.
Landrum, P.F., S.R. Nihart, B.J. Eadie, and W.S. Gardner. 1984.
Reverse-Phase Separation Method for Determining Pollutant
Binding to Aldrich Humic Acid and Dissolved Organic Carbon of
Natural Waters. Environ. Sci. Technol. 18:187-192.
McCarty, P.L., M. Reinhard, and B.E. Rittman. 1981. Trace
Organics in Groundwater. Environ. Sci. Technol. 15:40-51.
Morrow, N.R., and I. Chatzis. 1982. Measurement and Correlation
of Conditions for Entrapment and Mobilization of Residual Oil.
DOE/BC/10310-20.
Wilson, J.L., and S.H. Conrad. 1984. Is Physical Displacement of
Residual Hydrocarbons a Realistic Possibility in Aquifer
Restoration? p. 274-298. In Proc. Petroleum Hydrocarbons
Organic Chemicals Groundwater, Houston, TX.
Wolfe, T.A., T. Demiral, and E.R. Baumann. 1985. Interaction of
Aliphatic Amines with Montmorillonite to Enhance Adsorption of
Organic Pollutants. Clays and Clay Minerals. 33:301-311.
Mortland, M.M., S. Shaobai, and S.A. Boyd. 1986. Clay-Organic
Adsorbents for Phenol and Chlorophenol. Clays and Clay Minerals.
34:581-585.
Piwoni, M.D., and P. Banerjee. 1989. Sorption of Volatile
Organic Solvents from Aquifer Solution onto Subsurface Solids.
Journal of Contaminant Hydrology. 4:163-179.
Poirrier, M.A., B.R. Bordelon, and J.L. Lasetor. 1972. Adsorption
and Concentration of Dissolved Carbon-14 DDT by Coloring
Colloids in Surface Waters. Environ. Sci. Technol. 6:1033-1035.
Rao, P.S.C., and R.E. Jessup. 1983. Sorption and Movement of
Pesticides and other Toxic Substances in Soils, p. 183-201. In
D.W. Nelson et al. (ed.) Chemical Mobility and Reactivity in Soil
Systems. SSA Spec. Publ. II, Madison, Wl.
Schwarzenbach, R.P. and J. Westall. 1981. Transport of Nonpolar
Organic Compounds from Surface Water to Groundwater.
Laboratory Sorption Studies. Environ. Sci. Technol. 15:1360-
1367.
Thurman, E.M. 1985. Humic Substances in Groundwater, p. 87-
103. In G.R. Aiken et al. (ed.) Humic Substances in Soil, Sediment,
and Water. Wiley-lnterscience, New York.
Voice, T.C., C.P. Rice, and W.J. Weber, Jr. 1983. Effect of Solids
Concentration on the Sorptive Partitioning of Hydrophobic
Pollutants in Aquatic Systems. Environ. Sci. Technol. 17:513-
518.
Weber, W.J., Y. Chin, and C.P. Rice. 1986. Determination of
Partition Coefficients and Aqueous Solubilities by Reverse Phase
Chromatography-I. Water Res. 20:1433-1442.
Wershaw, R.L, P.J. Burcar, and M.C. Goldberg. 1969. Interaction
of Pesticides with Natural Organic Material. Environ. Sci. Technol.
3271-273.
Whitehouse, B. 1985. The Effects of Dissolved Organic Matter on
the Aqueous Partitioning of Polynudear Aromatic Hydrocarbons.
Estuarine Coastal Shelf Sci. 20:393-402.
7
-------
BULK RATE
POSTAGE & FEES PAID
EPA
Pemtit No. Q-35
Official Business
Denatty for Private Use $300
Jnited States Center for Environmental Research
Environmental Protection Information
Agency Cincinnati OH 45268
EPA/540/4-90/053
-------
Environmental Protection Research and Waste March 1991
Agency Development and Emergency
Response
&EPA Ground-Water Issue
CHARACTERIZING SOILS FOR
HAZARDOUS WASTE SITE ASSESSMENTS
R. P. Breckenridge1, J. R. Williams2, and J. F. Keck1
INTRODUCTION
The Regional Superfund Ground Water
Forum is a group of ground-water scientists
representing EPA's Regional Offices, orga-
nized to exchange up-to-date information re-
lated to ground-water remediation at hazard-
ous waste sites. Soil characterization at
hazardous waste sites is an issue identified by
the forum as a concern of CERCLA decision-
makers.
To address this issue, this paper was pre-
pared through support from EMSL-LV and
RSKERL, under the direction of R. P.
Breckenridge, with the support of the
Superfund Technical Support Project. For
further information contact Ken Brown, EMSL-
LV Center Director, at FTS 545-2270 or R. P.
Breckenridge at FTS 583-0757.
Site investigation and remediation under the
Superfund program is performed using the
CERCLA remedial investigation/feasibility
study (RI/FS) process. The goal of the RI/FS
process is to reach a Record of Decision
(ROD) in a timely manner. Soil characteriza-
tion provides data types required for decision
making in three distinct RI/FS tasks:
1. Determination of the nature and extent of
soil contamination.
2. Risk assessment, and determination of
risk-based soil clean-up levels.
3. Determination of the potential effective-
ness of soil remediation alternatives.
Identification of data types required for the first
task, determination of the nature and extent of
contamination, is relatively straightforward.
The nature of contamination is related to the
types of operations conducted at the site.
Existing records, if available, and interviews
with personnel familiar with the site history are
good sources of information to help determine
the types of contaminants potentially present.
This information may be used to shorten the
list of target analytes from the several hundred
contaminants of concern in the 40 CFR Part
264 list (Date 7-1-89). Numerous guidance
documents are available for planning all
1 Idaho National Engineering Laboratory, Environmental Science and Technology Group, Idaho Falls, ID 83415.
2 Soil Scientist, U.S. EPA/R. S. Kerr Environmental Research Laboratory, Ada, OK 74820
-A l^rojed
Vsv*/
Superfund Technology Support Center for Monitoring
and Site Characterization, Environmental Monitoring
Systems Laboratory Las Vegas, NV
Superfund Technology Support Center for
Ground-Water Fate and Transport, Robert S. Kerr
Environmental Research Laboratory Ada, OK
Technology Innovation Office
Office of Solid Waste and Emergency Response,
U.S. EPA, Washington, D.C.
Waiter W. Kovalick, Jr., Ph.D„ Director
Printed on Recycled Paper
-------
aspects of the subsequent sampling effort (US EPA, 1987a,
1988a, 1988b, and Jenkins et al., 1988).
The extent of contamination is also related to the types of
operations conducted at the site. Existing records, if available,
and interviews with personnel familiar with the site history are
also good sources of information to help determine the extent of
contamination potentially present. The extent of contamination
is dependent on the nature of the contaminant source(s) and the
extent of contaminant migration from the source(s). Migration
routes may include air, via volatilization and fugitive dust emis-
sions; overland flow; direct discharge; leachate migration to
ground water and surface runoff and erosion. Preparation of a
preliminary site conceptual model is therefore an important step
in planning and directing the sampling effort. The conceptual
model should identify the most likely locations of contaminants
in soil and the pathways through which they move.
The data type requirements for tasks 2 and 3 are frequently less
weii understood. Tasks 2 and 3 require knowledge of both the
nature and extent of contamination, the environmental fate and
transport of the contaminants, and an appreciation of the need
for quality data to select a viable remedial treatment technique.
Contaminant fate and transport estimation is usually performed
by computer modeling. Site-specific information about the soils
in which contamination occurs, migrates, and interacts with, is
required as input to a model. The accuracy of the model output
:s no better than the accuracy of the input information.
The purpose of this paper is to provide guidance to Remedial
Project Managers (RPM) and On-Scene Coordinators (OSC)
concerning soil characterization data types required for
decision-making in the CERCLA RI/FS process related to risk
assessment and remedial alternative evaluation for contami-
nated soils. Many of the problems that arise are due to a lack of
understanding the data types required for tasks 2 and 3 above.
This paper describes the soil characterization data types re-
quired to conduct model based risk assessment for task 2 and
the selection of remedial design for task 3. The information
presented in this paper is a compilation of current information
from the literature and from experience combined to meet the
purpose of this paper.
EMSL-Las Vegas and RSKERL-Ada convened a technical
committee of experts to examine the issue and provide technical
guidance based on current scientific information. Members of
the committee were Joe R. Williams, RSKERL-Ada; Robert G.
Baca, Robert P. Breckenridge, Alan B. Crockett, and John F.
Keck from the Idaho National Engineering Laboratory, Idaho
Falls, ID; Gretchen L. Rupp, PE, University of Nevada-Las
Vegas; and Ken Brown, EMSL-LV.
This document was compiled by the authors and edited by the
members of the committee and a group of peer reviewers.
Characterization of a hazardous waste site should be done
using an integrated investigative approach to determine quickly
and cost effectively the potential health effects and appropriate
response measures at a site. An integrated approach involves
consideration of the different types and sources of contami-
nants, their fate as they are transported through and are parti-
tioned, and their impact on different parts of the environment.
CONCERNS
This paper addresses two concerns related to soil characteriza-
tion for CERCLA remedial response. The first concern is the
applicability of traditional soil classification methods to CERCLA
soil characterization. The second is the identification of soil
characterization data types required for CERCLA risk assess-
ment and analysis of remedial alternatives. These concerns are
related, in that the Data Quality Objective (DQO) process
addresses both. The DQO process was developed, in part, to
assist CERCLA decision-makers in identifying the data types,
data quality, and data quantity required to support decisions that
must be made during the RI/FS process. Data Quality Objec-
tives for Remedial Response Activities: Development Process
(US EPA, 1987b) is a guidebook on developing DQOs. This
process as it relates to CERCLA soil characterization is dis-
cussed in the Data Quality Objective section of this paper.
Data types required for soil characterization must be determined
early In the RI/FS process, using the DQO process. Often, the
first soil data types related to risk assessment and remedial
alternative selection available during a CERCLA site investiga-
tion are soil textural descriptions from the borehole logs pre-
pared by a geologist during investigations of the nature and
extent of contamination. These boreholes might include instal-
lation of ground-water monitoring wells, or soil boreholes. Typi-
cally, borehole logs contain soil lithology and textural descrip-
tions, based on visual analysis of drill cuttings.
Preliminary site data are potentially valuable, and can provide
modelers and engineers with data to begin preparation of the
conceptual model and perform scoping calculations. Soil tex-
ture affects movement of air and water in soil, infiltration rate,
porosity, water holding capacity, and other parameters.
Changes in lithology identify heterogeneities in the subsurface
(i.e., low permeability layers, etc.). Soil textural classification is
therefore important to contaminant fate and transport modeling,
and to screening and analysis of remedial alternatives. How-
ever, unless collected properly, soil textural descriptions are of
limited value for the following reasons:
1. There are several different systems for classification of soil
particles with respect to size. To address this problem it is
important to identify which system has been or will be used
to classify a soil so that data can be properly compared.
Figure 1 can be used to compare the different systems (Gee
and Bauder, 1986). Keys to Soil Taxonomy (Soil Survey
Staff, 1990) provides details to one of the more useful
systems that should be consulted prior to classifying a site's
soils.
2. The accuracy of the field classification is dependent on the
skill of the observer. To overcome this concern RPMs and
OSCs should collect soil textural data that are quantitative
rather than qualitative. Soil texture can be determined from
a soil sample by sieve analysis or hydrometer. These data
types are superior to qualitative description based on visual
analysis and are more likely to meet DQOs.
3. Even if the field person accurately classifies a soil (e.g., as
a silty sand or a sandy loam), textural descnptions do not
afford accurate estimations of actual physical properties
required for modeling and remedial alternative evaluation.
2
-------
such as hydraulic conductivity. For example, the hydraulic
conductivity of silty-sand can range from 105 to 101 cm/sec
(four orders of magnitude).
These ranges of values may be used for bounding calculations,
or to assist in preparation of the preliminary conceptual model.
These data may therefore meet DQOs for initial screening of
remedial alternatives, for example, but will likely not meet DQOs
for detailed analysis of alternatives.
DATA QUALITY OBJECTIVES
EPA has developed the Data Quality Objective (DQO) process
to guide CERCLA site characterization The relationship be-
tween CERCLA RI/FS activities and the DQO process is shown
in Figure 2 (US EPA, 1988c, 1987a). The DQO process occurs
in three stages:
The types of decisions vary throughout the RI/FS process, but
in general they become increasingly quantitative as the pro-
cess proceeds. During this stage it is important to identify and
involve the data users (e.g. modelers, engineers, and scien-
tists), evaluate available data, develop a conceptual site
model, and specify objectives and decisions.
Stage 2. Identify Data Uses/Needs. In this stage data uses
are defined. This includes identification of the required data
types, data quality and data quantity required to make deci-
sions on how to:
- Perform risk assessment
- Perform contaminant fate and transport modeling
- Identify and screen remedial alternatives
Stage 1. identify Decision i ypes. in thus stage the types oi
decisions that must be made during the RI/FS are identified.
PARTICLE SIZE LIMIT CLASSIFICATION
esse
ISSS
ASTM (unified)
0.0002. - uj
0.001
0.002
0.003
0.004
0.006
0.008
0.01
0.02
0.03
0.04
— 0.06
| 0.08
~ 0.1 4-
UJ
N
55 0.2
w 0.3
u 0.4
0.6
0.8
1.0
2.0
3.0
4.0
6.0
8.0
10
20
30
40
60
80
CD ~
li
300
270
200
140
60
40
-- 20
10
4
1/2 In.
3/4 In.
FINE CLAY
CLAY
COARSE
CLAY
COARSE
CLAY
FINE
SILT
FINES
(SILT AND
SILT
MEDIUM
SILT
SILT
CLAY)
COARSE
SILT
VERY FINE
SAND
VERY FINE
SAND
FINE
SAND
FINE
SAND
FINE
SAND
FINE
SAND
MEDIUM
SAND
MEDIUM
SAND
COARSE
SAND
COARSE
SAND
COARSE
SAND
MEDIUM
VERY COARSE
SAND
VERY COARSE
SAND
SAND
FINE
COARSE
SAND
GRAVEL
GRAVEL
FINE
GRAVEL
COARSE
GRAVEL
GRAVEL
COARSE
GRAVEL
COBBLES
COBBLES
COBBLES
M
SA
1
USDA-US DEPARTMENT OF AGRICULTURE, (SOIL SURVEY STAFF, 1975)
CCS - CANADA SOIL SURVEY COMMITTEE (McKEAGUE, 1978)
ISSS - INTERNATIONAL SOIL SCI SOC (YONG AND WARKENTIN, 1966)
ASTM - AMERICAN SOCIETY FOR TESTING & MATERIALS (ASTM, D-2487,1985a)
Figure 1 Particle-size limits according to several current
classification schemes (Gee and Bauder, 1986).
• Siage3. Design Data Coliecuon Program. Aher Stage 1 and
2 activities have been defined and reviewed, a data collection
program addressing the data types, data quantity (number of
samples) and data quality required to make these decisions
needs to be developed as part of a sampling and analysis
plan.
Although this paper focuses on data types required for decision-
making in the CERCLA RI/FS process related to soil contami-
nation, references are provided to address data quantity quality
issues.
Data Types
The OSC or RPM must determine which soil parameters are
needed to make various RI/FS decisions. The types of deci-
sions to be made therefore drive selection of data types. Data
types required for RI/FS activities including risk assessment,
contaminant fate and transport modeling and remedial alter-
native selection are discussed in Soil characteristics Data Types
Required for Modeling Section, and the Soil Characterization
Data Type Required for Remedial Alternative Selection Section.
Data Quality
The RPM or OSC must decide "How good does the data need
to be in order for me to make a given decision?". EPA has
assigned quality levels to different RI/FS activities as a guide-
line. Data Quality Objectives for Remedial Response Activities
(US EPA, 1987a) offers guidance on this subject and contains
many useful references.
Data Quantity
The RPM or OSC must decide "How many samples do I need to
determine the mean and standard deviation of a given param-
eter at a given site7", or "How does a given parameter vary
spatially across the site?". Decisions of this type must be
addressed by statistical design of the sampling effort. The Soil
Sampling Quality Assurance Guide (Barth etal., 1989)and Data
Quality Objectives for Remedial Response (US EPA, 1987a)
offer guidance on this subject and contain many useful refer-
ences.
3
-------
Rl
PHASE 1
FS
PHASE 1
Rl
PHASE II
FS
YES
DQO
Stage
ll/lll
I
PHASE II
Figure 2. Phased RI/FS approach and the DQO process (EPA, 1987a).
4
-------
IMPORTANT SOIL CHARACTERISTICS IN SITE
EVALUATION
Tables 1 and 2 identify methods for collecting and determining
data types for soil characteristics either in the field, laboratory,
or by calculation. Soil characteristics in Table 1 are considered
the primary indicators that are needed to complete Phase I of the
RI/FS process. This is a short, but concise list of soil data types
that are needed to make CERCLA decisions and should be
planned for and collected early in the sampling effort. These
primary data types should allow for the initial screening of
remedial treatment alternatives and preliminary modeling of the
site for risk assessment. Many of these characteristics can be
obtained relatively inexpensively during periods of early field
work when the necessary drilling and sampling equipment are
already on site. Investigators should plan to collect data for all
the soil characteristics at the same locations and times soil
boring is done to install monitoring wells. Geophysical logging of
the well should also be considered as a cost effeclive method for
collecting lithologic information prior to casing the well. Data
quality and quantity must also be considered before beginning
collection of the appropriate data types.
The soil characteristics in Table 2 are considered ancillary only
because they are needed in the later stages and tasks of the
DQO process and the RI/FS process. If the site budget allows,
collection of these data types during early periods of field work
will improve the database available to make decisions on
remedial treatment selection and model-based risk assess-
ments. Advanced planning and knowledge of the need for the
ancillary soil characteristics should be factored into early site
work to reduce overall costs and the time required to reach a
ROD. A small additional investment to collect ancillary data
during earty site visits is almost always more cost effective than
having to send crews back to the field to conduct additional soil
sampling.
Further detailed descriptions of the soil characteristics in Tables
1 and 2 can be found in Fundamentals of Soil Physics and Ap-
plications of Soil Physics (Hillel, 1980) and in a series of articles
by Dragun (1988, 1988a, 1988b). These references provide
excellent discussions of these characteristics and their influ-
ence on water movement in soils as well as contaminant fate and
transport.
SOIL CHARACTERISTICS DATA TYPES REQUIRED
FOR MODELING
The information presented here is not intended as a review of all
data types required for all models, instead it presents a sampling
of the more appropriate models used in risk assessment and
remedial design.
Uses of Vadose Zone Models for Cercla Remedial
Response Activities
Models are used in the CERCLA RI/FS process to estimate
contaminant fate and transport. These estimates of contami-
nant behavior in the environment are subsequently used for:
• Risk assessment. Risk assessment includes contaminant
release assessment, exposure assessment, and determining
risk-based clean-up levels. Each of these activities requires
estimation of the rates and extents of contaminant movement
in the vadose zone, and of transformation and degradation
processes.
• Effectiveness assessment of remedial alternatives. This
task may also require determination of the rates and extents
of contaminant movement in the vadose zone, and of rates
and extents of transformation and degradation processes.
Technology-specific data requirements are cited in the Soil
Characterization Data Type Required for Remedial Alterna-
tive Selection Section.
The types, quantities, and quality of site characterization data
required for modeling should be carefully considered during Rl/
FS scoping. Several currently available vadose zone fate and
transport models are listed in Table 3. Soil characterization data
types required for each model are included in the table. Model
documentation should be consulted for specific questions con-
cerning uses and applications.
The Superiund Exposure Assessment Manual discusses vari-
ous vadose zone models (US EPA, 1988e). This document
should be consulted to select codes that are EPA-approved.
Data Types Required for Modeling
Soil characterization data types required for modeling are in-
cluded in Tables 1 and 2. Most of the models are one- or
two-dimensional solutions to the advection-dispersion equa-
tion, applied to unsaturated flow. Each is different in the extent
to which transformation and degradation processes may be
simulated; various contaminant release scenarios are accom-
modated; heterogeneous soils and other site-specific charac-
teristics are accounted for. Each, therefore, has different data
type input requirements.
All models require physicochemical data for the contaminants of
concern. These data are available in the literature, and from
EPA databases (US EPA, 1988c,d). The amount of physico-
chemical data required is generally related to the complexity of
the model. The models that account for biodegradation of
organics, vapor phase diffusion and other processes require
more input data than the relatively simpler transport models.
Data Quality and Quantity Required for Modeling
DQOs for the modeling task should be defined during RI/FS
scoping. The output of any computer model is only as valid as
the quality of the input data and code itself. Variance may result
from the data collection methodology or analytical process, or as
a result of spatial variability in the soil characteristic being
measured.
In general, the physical and chemical properties of soils vary
spatially. This variation rarely follows well defined trends; rather
it exhibits a stochastic (i.e., random) character. However, the
stochastic character of many soil properties tends to follow
classic statistical distributions. For example, properties such as
bulk density and effective porosity of soils tend to be normally
distributed (Campbell, 1985). Saturated hydraulic conductivity,
in contrast, is often found to follow a log-normal distribution.
Characterization of a site, therefore, should be performed in
such a manner as to permit the determination of the statistical
characteristics (i.e., mean and variance) and their spatial
correlations.
(Continued on page 8)
5
-------
TABLE 1. MEASUREMENT METHODS FOR PRIMARY SOIL CHARACTERISTICS
NEEDED TO SUPPORT CERCLA DECISION-MAKING PROCESS
Measurement Technique/Method (w/Reference)
Soil Characteristic* Held Laboratory Calculation or Lookup Method
Bulk density
Soil pH
Texture
Depth to
ground water
Horizons or
stratigraphy
Hydraulic
conductivity
(saturated)
Water retention
(soil water
characteristic
curves)
Air permeability
and water content
relationships
Porosity (pore
volume)
Climate
Neutron probe (ASTM, 1985),
Gamma radiation (Blake and Hartage,
1986, Blake, 1965).
Measured in field in same manner as
in laboratory.
Collect composite sample for each soil
type. No field methods are available,
except through considerable
experience of "feeling" the soil for an
estimation of % sand, silt, and clay.
Ground-water monitoring wells or
piezometers using EPA approved
methods (EPA 1985a).
Soil pits dug with backhoe are best. If
safety and cost are a concern, soil
bores can be collected with either a
thin wall sample driver and veilmayer
tube (Brown et al., 1990).
Auger-hole and piezometer methods
(Amoozeger and Warrick, 1986) and
Guelph permeameter (Reynolds &
Elrick, 1985; Reynolds & Elrick, 1986).
Field methods require a considerable
amount of time, effort, and equipment.
For a good discussion of these methods
refer to Bruce and Luxmoore (1986).
None
Coring or excavation for lab analysis
(Blake and Hartage, 1986).
Using a glass electrode in an aqueous
slurry (ref. EPRIEN-6637) Analytical
Method - Method 9045, SW-846, EPA.
ASTM D 522-63 Method for Particle
Analysis of Soils. Sieve analysis better at
hazardous waste sites because organics
can effect hydrometer analysis
(Kluate, 1986).
Not applicable.
Not applicable.
Constant head and falling head methods
(Amoozeger and Warrick, 1986).
Obtained through wetting or drainage of
core samples through a series of known
pressure heads from low to high or high
to low, respectively (Klute, 1986).
Several methods have been used,
however, all use disturbed soil samples.
For field applications the structure of
soils are very important, For more
information refer to Corey (1986).
Gas pycnometer (Danielson and
Sutherland, 1986).
Precipitation measured using either
Sacramento gauge for accumulated value
or weighing gauge or tipping bucket gauge
for continuous measurement (Finkelstein
et al., 1983; Kite, 1979). Soil temperature
measured using thermocouple.
Not applicable.
Not applicable.
Not applicable.
Not applicable.
Not applicable.
May be possible to obtain information
from SCS soil survey for the site.
Although there are tables available that
list the values for the saturated
hydraulic conductivity, it should be
understood that the values are given for
specific soil textures that may not be the
same as those on the site.
Some look-up and estimation methods
are available, however, due to high
spatial vanabiltiy in this characteristic
they are not generally recommended
unless their use is justified.
Estimation methods tor air permeability
exist that closely resemble the estimation
methods for unsaturated hydraulic
conductivity. Example models those
developed by Brooks and Corey (1964)
and van Genuchten (1980).
Calculated from particle and bulk
densities (Danielson and Sutherland,
1986).
Data are-provided in the Climatic Atlas of
the United States or are available from
the National Climatic Data Center,
Asheville, NC Telephone (704) 259-0682.
Soil characteristics are discussed in general except where specific cases relate to different waste types (i e„ metals, hydrophobic organics or polar organics).
6
-------
TABLE 2. MEASUREMENT METHODS FOR ANCILLARY SOIL PARAMETERS
NEEDED TO SUPPORT CERCLA DECISION-MAKING PROCESS
Measurement Technique/Method (w/Reference)
Soil Characteristic* Field Laboratory Calculation or Lookup Method
Organic carbon Not applicable.
Capacity Exchange See Rhoades tor field methods.
Capacity (CEC)
Erodibility
High temperature combustion (either
wet or dry) and oxidation techniques
(Powell et al., 1989) (Powell, 1990).
(Rhoades, 1982).
Not applicable.
Water erosion
Universal Soil Loss
Equation (USLE)
or Revised USLE
(RUSLE)
Wine erosion
Vegetative cover
Soil structure
Organic carbon
partition
cooefficient (KJ
Redox couple ratios
of waste/soil system
Measurement/survey of slope (in ft
nse/ft run or %), length of field,
vegetative cover.
Air monitoring for mass of containment
Field length along prevailing wind
direction.
Visual observation and documented
using map. USDA can aid in identification
of unknown vegetation.
Classified into 10 standard kinds - see
local SCS office for assistance (Soil
Survey Staff, 1990) or Taylor and
Ashcrott (1972), p. 310.
In situ tracer tests (Freeze and Cherry,
1979).
Platium electrode used on lysimeter
sample (ASTM, 1987)
Not applicable.
Not applicable
Not applicable.
Not applicable.
(ASTM E 1195-87,1988)
Same as field
Estimated using standard equations and
graphs (Israelsen et al., 1980) field data
tor slope, field length, and cover type
required as input Soils data can be
ODtained from the local Soil Conservation
Service (SCS) office.
A modified universal soil loss equation
(USLE) (Williams, 1975) presented in
Mills et al., (1982) and US EPA (1988d)
source for equations.
The SCS wind loss equation (Israelsen
et al., 1980) must be adjusted (reduced)
to account for suspended particles of
diameter <1 O^m Cowherd et al., (1985)
for a rapid evaluation (<24 hr) of particle
emission fro a Superfund site.
See local soil survey for the site
Calculated tram K , water solubility
(Mills et al., 1985; Sims et al., 1986).
Can be calculated from concentrations of
redox pairs or 02 (Stumm and Morgan, 1981)
Liner soil/water
partition coefficient
Soil oxygen
content (aeration)
In situ tracer tests (Freeze and Cherry,
1979)
Batch experiment (Ash et al., 1973);
column tests (van Genuchten and
Wierenga, 1986)
02 by membrane electrode 02 diffusion Same as field,
rate by Pt microeledrode (Phene, 1986).
O. by field GC (Smith, 1983)
Mills et al., 1985.
Calculated from pE (Stumm and Morgan,
1981) or from 02 and soil-gas diffusion
rate.
(Continued)
-------
TABLE 2. (CONTINUED)
Measurement Technique/Method (w/Refetence)
. Soil Characteristic* Field Laboratory Calculation or Lookup Method
Soil temperature (as Thermotery (Taylor and Jackson, 1986).
it affects volatilization)
Clay mineralogy Parent material analysis.
Same as field.
Brown and Associates (1980).
Unsaturated
hydraulic
conductivity
Moisture content
Soil biota
Unsteady dranage-fiux (or instantaneous
profile) method and simplified unsteady
drainage flux method (Green et al.,
1986).The instantaneous profile method
was initially developed as a laboratory
method (Watson, 1966), however it was
adapted to the field (Hillel et al. 1972).
Constant-head borehole infiltration
(Amoozegar and Warrick, 1986).
Two types of techniques - indirect and
direct. Direct menthods, (i.e., gravimetric
sampling), considered the most accurate,
with no calibration required. However,
methods are destructive to field systems
Methods involve collecting samples,
weighing, drying and re-weighing to
determine field moisture. Indirect methods
rely on calibration (Klute, 1986).
No standard method exists (see model or
remedial technology for input or remedial
evaluation procedures).
X-ray diffraction
-------
TABLE 3. SOIL CHARACTERISTICS REQUIRED FOR VADOSE ZONE MODELS
Model Name
[References)]
Properties and Parameters
Help
(A,B)
Sesoil
(C,D)
Creams
(E,F)
PRZM
(G,H,I)
Vadoft
(H,J)
Minteq
(J)
Fowl™
(K)
Rrtz
(L)
Vlp
(M)
Chemflo
(N)
Soil bulk density
o
•
•
o
•
o
•
•
•
•
Soil pH
o
•
o
o
o
•
•
o
o
o
Soil texture
•
o
•
• •
•
o
0
•
•
o
Depth to ground water
o
•
o
o
•
o
0
o
o
o
Horizons (soil layering)
•
•
•
o
•
o
o
o
o
o
Saturated hydraulic conductivity
ft
•
•
e
ft
o
ft
ft
ft
ft
Water retention
•
•
•
ft
•
o
ft
o
o
•
Air permeability
o
•
o
o
o
o
o
o
•
o
Climate (precipitation)
•
•
•
•
o
o
ft
•
•
•
Soil porosity
•
•
•
•
ft
o
o
ft
ft
o
Soil organic content
o
•
•
•
ft
ft
0
•
•
o
Cation Exchange Capacity (CEC)
o
•
o
o
o
ft
o
o
o
o
Degradation parameters
•
•
•
•
•
o
o
•
•
ft
Soil grain size distribution
o
o
o
o
o
o
o
o
o
o
Soil redox potential
o
o
o
o
o
•
o
o
o
o
Soil/water partition coefficients
o
•
•
•
ft
ft
ft
ft
ft
ft
Soil oxygen content
o
o
o
o
o
o
o
o
ft
o
Soil temperature
0
•
o
•
ft
ft
o
ft
ft
o
Soil mineralogy
o
•
o
0
o
0
o
o
o
o
Unsaturated hydraulic conductivity
•
•
•
•
ft
o
•
o
o
•
Saturated soil moisture content
•
•
•
•
ft
0
•
•
•
•
Microorganism population
o
o
o
0
o
o
o
o
o
o
Soil respiration
o
0
0
o
o
o
o
o
o
o
Evaporation
•
•
•
ft
o
0
o
•
•
•
Air/water contaminant densities
o
o
0
0
o
0
•
•
•
o
Air/water contaminant viscosities
o
o
o
G
o
o
o
o
o
o
REFRENCES
A. Schroeder,etal., 1984. F. Devaurs and Spnnger, 1988 K Hostev;- =• v-w pa, iqoo ft Required ONot required OUsed indirectly'
B Schroeder, etal ,1984a. G Carsel ei al, 1984 L No>¦'. h to
C. Bonazountas and Wagner, 1984. H Dean etal., 1989. M Steve"- v i ' Used in ther estimation ol other required
D. Chen, Wollman, and bu, 1987. l. Deanetal!l98Sa N No!z">- 1- > \V- characteristics or the intrpretatoon of the models,
E Leonard aid Ferre'ra 198^ Prwn -c<- but not directtv entered as inout to models
-------
Application of stochastic models to hazardous waste sites has
two main advantages. First, this approach provides a rigorous
way to assess the uncertainty associated with the spatial vari-
ability of soil properties. Second, the approach produces model
predictions in terms of the likelihood of outcomes, i.e., probabil-
ity of exceeding water quality standards. The use of models at
hazardous waste sites leads to a thoughtful and objective
treatment of compliance issues and concerns.
In order to obtain accurate results with models, quality data
types must be used. The issue of quality and confidence in data
can be partially addressed by obtaining as representative data
as possible. Good quality assurance and quality control plans
must be in place for not only the acquisition of samples, but also
for the application of the models (van der Heijde, et al., 1989).
Specific soil characteristics vary both laterally and vertically in
an undisturbed soil profile. Different soil characteristics have
different variances. As an example, the sample size required to
have 95 percent probability of detecting a change of 20 percent
in the mean bulk density at a specific site was 6; however, for
saturated hydraulic conductivity the sample size would need to
be 502 (Jury, 1986). A good understanding of site soil charac-
teristics can help the investigators understand these variations.
This is especially true for most hazardous waste sites because
the soils have often been disturbed, which may cause even
greater variability.
An important aspect of site characterization data and models is
that the modeling process is dynamic, i.e., as an increasing
number of "simplifying" assumptions are needed, the complexity
of the models must increase to adequately simulate the addi-
tional processes that must be included. Such simplifying as-
sumptions might include an isotropic homogeneous medium or
the presence of only one mobile phase (Weaver, et al., 1989).
In order to decrease the number of assumptions required, there
is usually a need to increase the number of site-specific soil
characteristic data types in a model (see Table 2); thus providing
greater confidence in the values produced. For complex sites,
an iterative process of initial data collection and evaluation
leading to more data collection and evaluation until an accept-
able level of confidence in the evaluation can be reached can be
used.
Table 3 identifies selected unsaturated zone models and their
soil characteristic needs. For specific questions regarding use
and application of the model, the reader should refer to the
associated manuals. Some of these models are also reviewed
by Donigan and Rao (1986) and van der Heijde et al. (1988).
SOIL CHARACTERISTICS DATA TYPES REQUIRED
FOR REMEDIAL ALTERNATIVE SELECTION
Remedial Alternative Selection Procedure
The CERCLA process involves the identification, screening and
analysis of remedial alternatives at uncontrolled hazardous
waste sites (US EPA, 1988c). During screening and analysis,
decision values for process-limiting characteristics for a given
remedial alternative are compared to site-specific values of
those characteristics. If site-specific values are outside the
range required for effective use of a particular alternative, that
alternative is less likely to be selected. Site soil conditions are
critical process-limiting characteristics.
Process-Limiting Characteristics
Process-limiting characteristics are site- and waste-specific
data types that are critical to the effectiveness and ability to
implement remedial processes. Often, process-limiting charac-
teristics are descriptors of rate-limiting steps in the overall
remedial process. In some cases, limitations imposed by
process-limiting characteristics can be overcome by adjustment
of soil characteristics such as pH, soil moisture content, tem-
perature and others. In other cases, the level of effort required
to overcome these limitations will preclude use of a remedial
process.
Decision values for process limiting characteristics are increas-
ingly available in the literature, and may be calculated for
processes where design equations are known. Process limiting
characteristics are identified and decision values are given for
several vadose zone remedial alternatives in Table 4. For
waste/site characterization, process-limiting characteristics
may be broadly grouped in four categories:
1. Mass transport characteristics
2. Soil reaction characteristics
3. Contaminant properties
4. Engineering characteristics
Thorough soil characterization is required to determine site-
specific values for process-limiting characteristics. Most reme-
dial alternatives will have process-limiting characteristics in
more than one category.
Mass Transport Characteristics
Mass transport is the bulk flow, or advection of fluids through
soil. Mass transport characteristics are used to calculate
potential rates of movement of liquids or gases through soil and
include:
Soil texture
Unsaturated hydraulic conductivity
Dispersivity
Moisture content vs. soil moisture tension
Bulk density
Porosity
Permeability
Infiltration rate, stratigraphy and others.
Mass transport processes are often process-limiting for both in
situ and extract-and-treat vadose zone remedial alternatives
(Table 4). In situ alternatives frequently use a gas or liquid
mobile phase to move reactants or nutrients through contami-
nated soil. Alternatively, extract-and-treat processes such as
soil vapor extraction (SVE) or soil flushing use a gas or liquid
mobile phase to move contaminants to a surface treatment site.
For either type of process to be effective, mass transport rates
must be large enough to clean up a site within a reasonable time.
Soil Reaction Characteristics
Soil reaction characteristics describe contaminant-soil interac-
tions. Soil reactions include bio- and physicochemical reactions
that occur between the contaminants and the site soil. Rates of
reactions such as biodegradation, hydrolysis, sorption/desorp-
tion, precipitation/dissolution, redox reactions, acid-base
reactions, and others are process-limiting characteristics for
(Continued on page 12)
10
-------
TABLE 4. SOIL CHARACTERIZATION CHARACTERISTICS REQUIRED FOR REMEDIAL TECHNOLOGY EVALUATION,
(US EPA, 1988e,f; 1969a,b; 1990; Sims et al., 1986; Sims, 1990; Towers etal., 1989)
Technology
Process
Limiting Characteristics
Site Data
Required
Pretreatment/
materials handling
Large particles interfere
Clayey soils or hardpan
difficult to handle
Particle size
distribution
Wet soils difficult
to handle
Soil moisture content
Soil vapor
extraction
Applicable only to volatile
organics w/significant vapor
pressure >1 mm Hg
Contaminants
present
Low soil permeability inhibits
air movement
Soil permeability
Soil hydraulic conductivity
>1E-8 cm/sec required
Hydraulic
conductivity
Depth to ground water
>20 ft recommended
Depth to ground water
High moisture content
inhibits air movement
Soil moisture content
High organic matter
content inhibits
contaminant removal
Organic matter content
In situ enhanced
bioremediation
Applicable only to
specific organics
Contaminants present
Hydraulic conductivity
>1E-4 cm/sec preferred
to transport nutrients
Hydraulic conductivity
Stratification should be
minimal
Soil stratigraphy
Lower permeability layers
difficult to remediate
Soii stratigraphy
Temperature 15-45°C
required
Soil temperature
Moisture content 40-80%
of that at -1/3 bars tension
preferred
Soil moisture
characteristic curves
pH 4.5-3.5 required
Soil pH
Presence of microbes
required
Plate count
Minimum 10% air-filled
porosity required for
aeration
Porosity and soil
moisture content
Thermal treatment Applicable only to organics Contaminants present
Soil moisture content Soil moisture content
afreets handling and
heating requirements
Technology
Process
Limiting Characteristics
Site Data
Required
Thermal treatment
(continued)
Particle size affects
feeding and residuals
Particle size
distribution
pH <5 and >11 causes
corrosion
pH
Solidification/
stabilization
Not equally effective for
all contaminants
Contaminants
present
Fine particles < No. 200
mesh may interfere
Particle size
distribution
Oil and grease >10%
may interfere
Oil and grease
Chemical
extraction
(slurry reactors)
Not equally effective
for all contaminants
Particle size <0.25 in.
Contaminants
present
Particle size
distribution
pH <10
pH
Soil washing
Not equally effective
for all contaminants
Contaminants
present
Silt and clay difficult
to remove from wash
fluid
Particle
size distribution
Soil flushing
Not equally effective
for all contaminants
Contaminants
present
Required number of
pore volumes
Infiltration rate
and porosity
Glycolate
dechlonnation
Not equally effective
for all contaminants
Contaminants
present
Moisture content <20%
Moisture content
Low organic matter
content required
Organic carbon
Chemical oxidation/ Not equally effective
reduction (slurry for all contaminants
reactor)
Oxidizable organics
interfere
Contaminants
present
Organic carbon
pH <2 interferes
PH
In situ
vitrification
Maximum moisture
content of 25% by weight
Moisture
content
Particle size <4 inches
Requires soil hydraulic
conductivity <1E-5 cm/sec
Particle size
distribution
Hydraulic conductivity
-------
many remedial alternatives (Table 4). Soil reaction character-
istics include:
Kd, specific to the site soils and contaminants
Cation exchange capacity (CEC)
Eh
pH
Soil biota
Soil nutnent content
Contaminant abiotic/biological degradation rates
Soil mineralogy
Contaminant properties, described below, and others.
Soil reaction characteristics determine the effectiveness of
many remedial alternatives. For example, the ability of a soil to
attenuate metals (typically described by Kd) may determine the
effectiveness of an alternative that relies on capping
and natural attenuation to immobilize contaminants.
Soil Contaminant Properties
Contaminant properties are critical to contaminant-soil interac-
tions, contaminant mobility, and to the ability of treatment
technologies to remove, destroy or immobilize contaminants.
Important contaminant properties include:
Water solubility
Dielectric constant
Diffusion coefficient
Molecular weight
Vapor pressure
Density
Aqueous solution chemistry, and others.
Soil contaminant properties will determine the effectiveness of
many treatment techniques. For example, the aqueous solution
chemistry of metal contaminants often dictates the potential
effectiveness of stabilization/solidification alternatives.
Soil Engineering Characteristics and Properties
Engineering characteristics and properties of the soil relate both
to implementability and effectiveness of the remedial action.
Examples include the ability of the treatment method to remove,
destroy or immobilize contaminants; the costs and difficulties in
installing slurry walls and other containment options at depths
greater than 60 feet; the ability of the site to withstand vehicle
traffic (trafficabiIity); costs and difficulties in deep excavation of
contaminated soil; the ability of soil to be worked for implemen-
tation of in situ treatment technologies (tilth); and others.
Knowledge of site-specific engineering characteristics and
properties is therefore required for analysis of effectiveness and
implementability of remedial alternatives. Engineering charac-
teristics and properties include, but are not limited to:
Trafficability
Erodability
Tilth
Depth to groundwater
Thickness of saturated zone
Depth and total volume of contaminated soil
Bearing capacity, and others.
SUMMARY AND CONCLUSIONS
The goal of the CERCLA RI/FS process is to reach a ROD in a
timely manner. Soil characterization is critical to this goal. Soil
characterization provides data for RI/FS tasks including deter-
mination of the nature and extent of contamination, risk as-
sessment, and selection of remedial techniques.
This paper is intended to inform investigators of the data types
required for RI/FS tasks, so that data may be collected as
quiokly, efficiently, and cost effectively as possible. This
knowledge should improve the consistency of site evaluations,
improve the ability of OSCs and RPMs to communicate data
needs to site contractors, and aid in the overall goal of reaching
a ROD in a timely manner.
REFERENCES
American Society for Testing and Materials, 1985. Density of
soil and soil aggregate in place by nuclear methods (shallow
depth). ASTM, Philadelphia, PA.
ASTM, 1987. American Society for Testing and Materials,
Standard practice for oxidation-reduction potential of water.
ASTM D1498-76. ASTM, Philadelphia, PA.
ASTM, 1987. American Society for Testing and Materials.
Standard Test Method for Determining a Sorption Constant
(Koc) for an Organic Chemical in Soil and Sediments E1195-87.
Annual Book ASTM Standards, Vol. 11.02 p. 731.
Amoozegar, A. and A. W. Warrick, 1986. Hydraulic Conductivity
of Saturated Soils. In Klute, A. ed. Methods of Soil Analysis Part
1: Physical and Mineratogical Methods, 2nd edition. Monograph
9 (Part 1), American Society of Agronomy, inc./Soil Science
Society of America, Inc. Publisher, Madison, Wl.
Andersson, J. and A. M. Shapiro, 1983. "Stochastic Analysis of
One-Dimensional Steady-State Unsaturated Flow: A
Comparison of Monte Carlo and Perturbation Methods," Water
Resources Research, Vol. 19, No. 1, pp. 121-133.
Ash, S. G., R. Brown, and D. H. Everett, 1973. A high-precision
apparatus for the determination of adsorption at the interface
between a solid and a solution. J. Chem. Thermodynamics 5:
239-246.
Barth, D. S., B. J. Mason, T. H. Starks, and K. W. Brown, 1989.
Soil Sampling Quality Assurance User's Guide. EPA 600/8-89/
046, U.S. Environmental Protection Agency, Environmental
Monitoring Systems Laboratory, Las Vegas, NV.
Blake, G. R., 1965. Bulk Density. In Black, C. A. ed. Methods
of Soil Analysis. Part 1. Monograph 9, Part 1, Am. Soc. of
Agronomy, Madison, Wl.
Blake, G. R. and K. H. Hartge, 1986. Bulk density. In Klute, A.
ed. Methods of Soil Analysis Part 1: Physical and Mineralogical
Methods, 2nd edition. Monograph 9 (Part 1), American Society
of Agronomy, Inc./Soil Science Society of America, inc.
Publisher, Madison, Wl.
12
-------
Bonazountas, M. and J. M. Wagner, 1984. SESOIL: A
Seasonal Soil Compartment Model. Contract No. 68-01 -6271,
Draft Report from Arthur D. Little, Inc. U.S. Environmental
Protection Agency. Office of Toxic Substances, Washington,
DC.
Brady, Nyle C., 1974. The Nature and Properties of Soils,
MacMillan Publishing Co., Inc., NY.
Brooks, R. H. and A. T. Corey, 1964. "Hydraulic properties of
porous media", Hydrology Paper No. 3,27 pp. Colorado State
University, Fort Collins, CO.
Brown, D. S. and J. D. Allison, 1987. MINTEOA1 Equilibrium
Metal Speciation Model: A User's Manual. U. S. Environmental
Protection Agency, Environmental Research Laboratory,
Athens, GA.
Brown, K. W. and Associates, 1980. Hazardous waste land
treatment. Draft edition. SW-874. U.S. Environmental
Protection Agency, Cincinnati, OH.
K. W. Brown, R. P. Breckenridge, and R. C. Rope, 1990. U.S.
Fish and Wildlife Service Contaminant Monitoring Operations
Manual: Appendix J, Soil Sampling Reference Field Methods,
EGG-EST-9222, EG&G Idaho, Inc, Idaho Falls, ID.
Bruce, R. R. and R. J. Luxmoore, 1986. Water Retention- Field
Methods. In Klute, A., ed. Methods of Soil Analysis Part 1:
Physical and Mineralogical Methods, 2nd edition. Monograph 9
(Part 1), American Society of Agronomy, IncVSoil Science
Society of America, Inc. Publisher, Madison, Wl.
Campbell, G. S., 1985. Soil Physics with Basic, Elsevier, New
York, NY.
Carsel, R. F., C. N. Smith, L. A. Mulkey, J. D. Dean, and P.
Jowise, 1984. Users Manual for the Pesticide Root zone Model
(PRZM): Release 1. U. S. Environmental Protection Agency,
Environmental Research Laboratory, Athens, GA.
Chen, J.,S. Wollman.andJ. Liu, 1987. User's Guide to SESOIL
Execution in GEMS. GSC-TR8747. Prepared by General
Sciences Corporation. U.S. Environmental Protection Agency.
Office of Pesticides and Toxic Substances. Washington, DC.
Clark, I., 1982. Practical Geostatistics, Applied Science
Publishers Ltd, London, England.
Clifton, P. M., R. G. Baca, R. C. Arnett, 1985. "Stochastic
Analysis of Groundwater Traveltimes for Long-Term Repository
Performance Assessment," in the Proceedings of the Materials
Research Society Symposium-Scientific Basis for Nuclear
Waste Management, Boston, MA.
Corey, A. T„ 1986. Air Permeability. In Klute, A., ed. Methods
of Soil Analysis Part 1: Physical and Mineralogical Methods, 2nd
edition. Monograph 9 (Part 1), American Society of Agronomy,
Inc./Soil Science Society of America, Inc. Publisher, Madison,
Wl.
Cowherd, C., Mulseki, G. E., Englehart, P. J., and Gillette, D. A.,
1985. PB85-192219, Rapid assessment of exposure to
particulate emissions from surface contamination sites.
Midwest Research Institute, Kansas City, MO
Danielson, R. E. and P. L.Sutherland, 1986. Porosity. /nKlute,
A. ed. Methods of Soil Analysis Part 1: Physical and
Mineralogical Methods, 2nd edition. Monograph 9 (Part 1),
American Society of Agronomy, IncJSoil Science Society of
America, Inc. Publisher, Madison, Wl.
Davis, J. C., 1986. Statistics and Data Analysis in Geology,
Second Edition, John Wiley and Sons, New York, NY.
Dean, J. D„ P. S. Huyakom, A. S. Donigian, Jr., K. A. Voos, R.
W. Schanz, and R. F. Carsel, 1989. Risk of Unsaturated/
Saturated Transport and Transformation of Chemical
Concentrations (RUSTIC): Volume I. Theory and Code
Verification. EPA/600/3-89/048a. U.S. Environmental
Protection Agency, Environmental Research Laboratory,
Athens, GA.
Dettinger, M. D. and J. L. Wilson, 1981. "First Order Analysis
of Uncertainty in Numerical Models of Groundwater Flow, Part
1, Mathematical Development," Water Resources Research,
Vol. 16, No. 1, pp. 149-161.
Devaurs, M. and E. Springer, 1988. "Representing Soil
Moisture in Experimental Trench Cover Designs for Waste
Burial with the CREAMS Model". Hazardous Waste and
Hazardous Material. Vol. 5, No. 4, pp. 295-312.
Donigian, A. S., Jr. and P.S. C. Rao, 1986 Overview of
Terrestrial Processes and Modeling. In Hern, S. C. and S. M.
Melancon. 1986. Vadose Zone Modeling of Organic
Pollutants, Lewis Publishers, Inc., Chelsea, Ml.
Dragun, J. 1988. The Fate of Hazardous Materials in Soil
(What Every Geologist and Hydrogeologist Should Know), Part
1. HMC 1(2): 30-78.
Dragun, J. 1988a. The Fate of Hazardous Materials in Soil
(What Every Geologist and Hydrogeologist Should Know), Part
2. HMC 1(3): 40-65.
Dragun, J. 1988b. The Fate of Hazardous Materials in Soil
(What Every Geologist and Hydrogeologist Should Know), Part
3. HMC 1(5): 24-43.
Englund, E. and A. Sparks, 1988. GEO-EAS (Geostatistical
Environmental Assessment Software) User's Guide. EPA/600/
4-88/033.
Eslinger, P. W. and B. Sagar, 1988. EPASTAT: A Computer
Model for Estimating Releases at the Accessible Environment
Boundary of a High-Level Nuclear Waste Repository -
Mathematical Model and Numerical Model, SD-BWI-TA-022,
Rockwell Hanford Operations, Richland, WA.
Finkelstein, F.L., D. A. Mazzarella, T. A. Lockhart, W. J. King,
and J. H. White, 1983. Quality Assurance Handbook for Air
Pollution Measurement Systems. IV: Meteorological
Measurements, EPA-600/4-82-060, Washington, DC.
Freeze, R. A. and J. A. Cherry, 1979. Groundwater. Prentice-
Hall. Englewood Cliffs, NJ.
13
-------
Gardner, W. H. 1986. Water content. In Klute, A., ed. Methods
of Soil Analysis Part 1: Physical and Mineraiogical Methods, 2nd
edition. Monograph 9 (Part 1), American Society of Agronomy,
IncVSoil Science Society of America, inc. Publisher, Madison,
Wl.
Gee, G. W. and J. W. Bauder. Particle-size Analysis. In Klute,
A., ed. Methods of Soil Analysis Part 1: Physical and
Mineraiogical Methods, 2nd edition. Monograph 9 (Part 1),
American Society of Agronomy, Inc./Soil Science Society of
America, Inc. Publisher, Madison, Wl.
Green, R. E., L. R. Ahuja, and S. K. Chong. 1986. Hydraulic
Conductivity, Diffusivity, and Sorptivity of Unsaturated Soils:
Field Methods. In Klute, A., ed. Methods of Soil Analysis Part 1:
Physical and Mineraiogical Methods, 2nd edition. Monograph 9
(Part 1), American Society of Agronomy, Inc./Soil Science
Society of America, Inc. Publisher, Madison, Wl.
Hillel, D., 1980. Application oi Soil Physics, Academic Press,
Inc., New York, NY.
Hillel, D., 1980a. Fundamentals of Soil Physics, Academic
Press, Inc., New York, NY.
Hillel, D., V. D. Krentos.Y.Stylianou, 1972. "Procedure and Test
of an Internal Drainage Method for Measuring Soil Hydraulic
Characteristics in situ", Soil Science 114:395-400.
Hostetler, C. J., R. L. Erikson, and D. Ral, 1988. The Fossil Fuel
Combustion Waste Leaching (FOWLJ Code: Version 1. User's
Manual. EPRI EA-57420CCM. Electric Power Research
Institute. Palo Alto, CA.
Israelsen, C. E., Clyde, C. G., Fletcher, J. E„ Israelsen, E.K.,
Haws, F. W., Packer, P. E., and Farmer, E.E., Erosion Control
During Highway Construction. Manual on Principles and
Practices. Transportation Research Board, National Research
Council, Washington, DC 1980.
Jenkins, R. A., W. H. Griest, R. L. Moody, M. V. Buchanan, M.
P. Maskarinec, F. F. Dyer, C. -h. Ho, 1988. Technology
Assessment of Field Portable Instrumentation for Use at Rocky
Mountain Arsenal, ORNLVTM-10542, Oak Ridge National
Laboratory, Oak Ridge, TN.
Jury.W. A., 1986. Spatial Variability of Soil Properties. In Hern,
S. C. and S. M. Melancon. Vadose Zone Modeling of Organic
Pollutants. Lewis Publishers, Inc., Chelsea, Ml.
Kite, J. W„ 1979. Guideline for the Design, Installation, and
Operation of a Meteorological System, Radian Corporation,
Austin, TX.
Klute, A., 1986. Water Retention: Laboratory Methods, /n Klute,
A. ed. Methods of Soil Analysis Part 1: Physical and
Mineraiogical Methods, 2nd edition. Monograph 9 (Part 1),
American Society of Agronomy, Inc./Soil Science Society of
America, Inc. Publisher, Madison, Wl.
Leonard, R. A., and V. A. Ferreira, 1984. "CREAMS2 - The
Nutrient and Pesticide Models", Proceedings of the Natural
Resources Modeling Symposium, U.S. Department of
Agriculture.
Mills, W. B., D. B. Procella, M. J. Ungs, S. A. Gherini, K. V.
Summers, L Mok, G. L Rupp, G. L Bowie, and D. A. Haith, 1985.
EPA/600/6-85-002a, Water quality assessment: A screening
procedure for toxic and conventional pollutants in surface and
ground water. Part 1. Tetra Tech Inc., Lafayette, CA.
Mills, W. B„ Dean, J. D„ Porcella.D. B.,etal., 1982. Waterquality
assessment: a screening procedure for toxic and conventional
pollutants: parts 1, 2, and 3, Athens, GA: U.S. Environmental
Protection Agency. Environmental Research Laboratory. Office
of Research and Development. EPA-600/6-82/004 a.b.c.
Mualem, Y. 1986. Hydraulic Conductivity of Unsaturated Soils:
Prediction and Formulas. In Klute, A. ed. Methods of Soil Analysis
Part 1: Physical and Mineraiogical Methods, 2nd edition.
Monograph 9 (Part 1), American Society of Agronomy, lnc7Soil
Science Society of America, Inc. Publisher, Madison, Wl.
Neilson, D. R. and J. Bouma, eds, 1985. Soil Spatial Variability,
Center for Agricultural Publishing and Documentation,
Wageningen, the Netherlands.
Nofziger, D. L, K. Rajender, S. K. Nayudu, and P. Y. Su, 1989.
CHEMFLOW: One-Dimensional Water and Chemical
Movement in Unsaturated Soils. EPA/600/8-89/076. U. S.
Environmental Protection Agency. Robert S. Kerr Environmental
Research Laboratory, Ada, OK.
Nofziger, D. L. and J. R. Williams, 1988. Interactive Simulation
of the Fate of Hazardous Chemicals During Land Treatment of
Oily Wastes: RITZ User's Guide. EPA/600/8-88/001. U.S.
Environmental Protection Agency. Roberts. Kerr Environmental
Research Laboratory, Ada, OK.
Phene, C. J., 1986. Oxygen electrode measurement. In Klute,
A. ed. Methoids of Soil Analysis Part 1: Physical and Mineraiogical
Methods, 2nd edition. Monograph 9 (Part 1), American Society
of Agronomy, lnc./Soil Science Society of America, Inc.
Publisher, Madison, Wl.
Powell, R.M., Bledsoe, B. E., Johnson, R. L., and G. P. Curtis,
"Interlaboratory Methods Comparison for the Total Organic
Cartoon Analysis of Aquifer Materials", Environmental Science
and Technology, Vol. 23, pp. 1246-1249.
Powell, R.M., 1990. "Total Organic Carbon Determinations in
Natural and Contaminated Aquifer Materials, Relevance and
Measurement", Proceedings of the Fourth National Outdoor
Action Conference on Aquifer Restoration, Ground water
Monitoring and Geophysical Methods (National Water Well
Association), May 14-17,1990, Las Vegas, NV.
Reynolds, W. D. and D. E. Elrick, 1985. In situ Measurement of
Field-Saturated Hydraulic Conductivity, Sorptivity and the a-
Parameter using the Guelph Permeameter. Soil Science
140(4):292-302.
Reynolds, W. D. and D. E. Elrick, 1986. A Method for
Simultaneous in situ Measurement in the Vadose zone of Field-
Saturated Hydraulic Conductivity, Sorptivity, and the
Conductivity-Pressure Head Relationship. Ground Water
Monitoring Review 6(1):84-95.
14
-------
Rhoades, J.D., 1982. Cation Exchange Capacity. In Page, A.
L., R. H. Miller, and D. R. Keeney, eds, Methods of Soil Analysis,
Part 2, Chemical and Microbiological Properties, 2nd edition,
American Society of Agronomy Monograph 9 (Part 2), Madison,
Wl.
Roco, M. C., J. Khadilkar, and J. Zhang, 1989. 'Probabilistic
Approach for Transport of Contaminants Through Porous
Media," International Journal for Numerical Methods in Fluids,
Vol. 9, pp. 1431-1451.
Sagar, B., P. W. Eslinger, and R. G. Baca, 1986. "Probabilistic
Modeling of Radionuclide Release at the Waste Package
Subsystems Boundary of a Repository in Basalt," Nuclear
Technology, Vol. 75, pp. 338-349.
Schroeder, P.R., J. M. Morgan, T. M. Walski, and A. C. Gibson,
1984. Hydrologic Evaluation of Landfill Performance (HELP)
Model: Volume I. User's Guide for Version 1. EPA/530-SW-84-
009, U.S. Environmental Protection Agency. Municipal
Environmental Research Laboratory, Cincinnati, OH.
Schroeder, P. R., A. C. Gibson, and M. D. Smolen, 1984.
Hydrologic Evaluation of Landfill Performance (HELP) Model:
Volume II. Documentation for Version 1. EPA/530-SW-84-010.
U. S. Environmental Protection Agency. Municipal
Environmental Research Laboratory, Cincinnati, OH.
Sims,R. C. 1990. Soil RemediationTechniques at Uncontrolled
Hazardous Waste Sites: A Critical Review. Journal of the Air
and Waste Management Association, Vol. 40, No. 5, pp. 704-
732.
Sims, J. L, R. C. Sims, and J. E. Matthews, 1989. EPA/600/9-
89/073, Bioremediation of Contaminated Surface Soils, US EPA
Environmental Research Laboratory, Ada, OK.
Sims, R. C., D. Sorenson, J. Sims, J. McLean, R. Mahmood, R.
Dupont, J. Jurinak, and K. Wagner, 1986. Contaminated
Surface Soils In-Place Treatment Techniques. Pollution
Technology Review No. 132. Noyes Publications, Park Ridge,
NJ.
Smith, K. A. 1983. Gas chromatographic analysis of the soil
atmosphere. In K. A. Smith (ed.) Soil Analysis. Instrumental
techniques and related procedures. Marcel Dekker Inc. New
York, NY.
Soil Conservation Service (SCS), USDA, 1951. Soil survey
manual. U.S. Department of Agriculture Handbook 18, p. 228,
U.S. Government Printng Office, Washington, DC.
Soil Survey Staff, 1990. Keys to Soil Taxonomy. Soil
Management Support Services. SMSS Technical Monograph
#19, 4th edition. Virginia Polytechnic Institute, International
Soils, Department of Crop and Soil Environmental Science,
Blacksburg, VA.
Stevens, D. K., W. J. Grenney, Z. Yan, and R. C. Sims, 1989.
Sensitive Parameter Evaluation for a Vadose Zone Fate and
Transport Model. EPA/600.2-89/039. U. S. Environmental
Protection Agency. Robert S. Kerr Environmental Research
Laboratory, Ada, OK
Stumm, W. and J. J. Morgan, 1981. Aquatic Chemistry. 2nd
edition. Wiley-lnterscience, NY.
Taylor, S. A. and G. L'Ashcroft, 1972. Physical Edaphology.
The Physics of Irrigated and Nonirrigated Soils, W. H. Freeman
and Company, San Francisco, CA.
Taylor, S. A. and R. D. Jackson, 1986. Temperature. /nKlute,
A. ed. Methods of Soil Analysis Part 1: Physical and
Mineralogical Methods, 2nd edition. Monograph 9 (Part 1),
American Society of Agronomy, InciSoil Science Society of
America, Inc. Publisher, Madison, Wl.
Towers, D. S., M. J. Dent, and D. G. Van Arnam, 1988.
Evaluation of In Situ Technologies for VHOs Contaminated Soil.
In: Proceedings of the 6th National Conference on Hazardous
Wastes and Hazardous Materials. Sponsored by the
Hazardous Materials Control Research Institute.
US EPA, 1985a. Practical Guide for Ground-water Sampling,
EPA 600/2-85-104, Environmental Research Laboratory. Ada,
OK.
US EPA, 1985b. Compilation of Air Pollutant Emission Factors.
Volume 1. Stationary Point and Area Sources. Fourth Edition.
Office of Research and Development. Research Triangle Park,
NC.
US EPA, 1987a. Data Quality Objectives for Remedial
Response Activities, EPA/540/G-87/003 (NTIS PB88-131370),
Office of Emergency and Remedial Response and Office of
Waste Programs Enforcement, Washington, D.C. 20460.
US EPA, 1987b. Compendium of Superfund Field Operating
Methods, EPA-540 P-87:001. OSWER Directive 9355.0-14.
Office of Solid Waste and Emergency Response, U.S.
Environmental Protection Agency, Washington, DC.
US EPA, 1988a. Field Screening Methods for Hazardous Waste
Site Investigations, Proceedings from the First International
Symposium, October 11-13,1988.
US EPA, 1988b. Field Screening Methods Catalog. User's
Guide. EPA/540/2-88/005. U.S. Environmental Protection
Agency, Office of Emergency and Remedial Response,
Washington, DC.
US EPA, 1988c. Guidance for Conducting Remedial
Investigations and Feasibility Studies Under CERCLA: Interim
Final. EPA/540/G-89/004. Office of Emergency and Remedial
Response, U. S. Environmental Protection Agency,
Washington, DC.
US EPA. 1988d. Superfund Exposure Assessment Manual.
EPA-540-1-88-001. OSWER Directive 9285.5-1. Office of
Remedial Response, U.S. Environmental Protection Agency,
Washington, DC.
US EPA, 1988e. Technology Screening Guide for Treatment of
CERCLA Soils and Sludges. EPA/540/2-88/004; NTIS# PB89-
132674. U.S. Environmental Protection Agency, Washington.
DC.
15
-------
US EPA, 1988f. Cleanup of Releases from Petroleum USTs:
Selected Technologies. EPA/530/UST-88/001. U.S.
Environmental Protection Agency, Office of Underground
Storage Tanks, Washington, DC 20640.
US EPA, 1989a. Seminar on Site Characterization for
Subsurface Remediations. CERI-89-224. U.S. Environmental
Protection Agency, Office of Research and Development,
Washington, DC 20460.
US EPA, 1989b. Bioremediation of Hazardous Waste Sites
Workshop: Speaker Slide Copies and Supporting Information.
CERI-89-11. U.S. Environmental Protection Agency, Office of
Research and Development, Washington, DC 20460.
US EPA, 1990. Handbook on In Situ Treatment of Hazardous
Waste-Contaminated Soils. EPA/540/2-90/002. U.S.
Environmental Protection Agency Risk Reduction Engineering
Laboratory, Cincinnati, OH.
van der Heijde, P. K. M, A. I. El-Kadi, and S. A. Williams, 1988.
Ground Water Modeling: An Overview and Status Report. EPA/
600/2-89/028.
van der Heijde, P.K.M., W. I. M. Elderhorst, R. A. Miller, and M.
J. Trehan, 1989. The Establishment of a Groundwater
Research Data Center for Validation of Subsurface Flow and
Transport Models. EPA/600/2-89/040, July 1989.
van Genuchten (in press). Proceedings of the International
Workshop on Indirect Methods for Estimating the Hydraulic
Properties of Unsaturated Soils, Riverside, CA, October 11-13,
1989. Univ. of CA-Riverside and U.S. Department of
Agriculture.
van Genuchten, M. Th. 1980. A Closed-form equation for
predicting the hydraulic conductivity of unsaturated soils. Soil
Sci. Soc. Am. J. 44:892-898.
van Genuchten, M. and P. J. Wierenga, 1986. Solute dispersion
coefficients and retardation factors. In Klute, A. ed. Methods of
Soil Analysis Part 1: Physical and Mineralogical Methods, 2nd
edition. Monograph 9 (Part 1), American Society of Agronomy,
IncVSoil Science Society of America, Inc. Publisher, Madison,
Wl.
Watson, K. K. 1986. An Instantaneous Profile Method for
Determining the Hydraulic Conductivity of Unsaturated Porous
Media. Water Resources Research 2:709-715.
Weaver, J., C. G. Enfield, S. Yates, D. Kreamer, and D. White,
1989. Predicting Subsurface Contaminant Transport and
Transformation: Considerations for Model Selection and Field
Validation. EPA/600/2-89/045, August 1989.
Whittig, L. D. and W. R. Allardice, 1986. X-Ray Diffraction
Techniques. In A. Klute, ed. Methods of Soil Analysis Part 1:
Physical and Mineralogical Methods, 2nd edition. Monograph 9
(Part 1), American Society of Agronomy, IncTSoil Science
Society of America, Inc. Publisher, Madison, Wl.
Williams, J. R. 1975. Sediment-yield prediction with the
universal equation using runoff energy factor. In Present and
prospective technology for predicting sediment yields and
sources. ARS-S-40. U.S. Department of Agriculture.
Yates, S. R. and M. V. Yates, 1990. Geostatistics for Waste
Management: A User's Guide for the GEOPACK (Version 1.0)
GeostatisHcal Software System. EPA/600/8-90/004, January
1990.
United States Center for Environmental Research BULK RATE
Environmental Protection Information POSTAGE & FEES PAID
Agency Cincinnati, OH 45268 EPA PERMIT NO. G-35
Official Business
Penalty for Private Use $300
EPA/540/4-91/003
-------
United States Office of Office of Solid Waste EPA/540/S-93/501
Environmental Protection Research and and Emergency May 1993
Agency Development Response
EPA Engineering Issue
In Situ Bioremediation of Contaminated Unsaturated
Subsurface Soils
J.L. Sims*, R.C. Sims*, R.R. Dupont*, J.E. Matthews** and H.H. Russell**
Introduction
An emerging technology for the remediation of unsaturated
subsurface soils involves the use of microorganisms to
degrade contaminants which are present in such soils.
Understanding the processes which drive in situ
bioremediation, as well as the effectiveness and efficiency of
the utilization of these systems, are issues which have been
identified by the Regional Superfund Engineering Forum as
concerns of Superfund decision makers.
The Regional Superfund Engineering Forum is a group of
EPA professionals, representing EPA's Regional Superfund
Offices, committed to the identification and resolution of
engineering issues impacting the remediation of Superfund
sites. The Forum is supported by and advises the Superfund
Technical Support Project.
Although in situ bioremediation has been used for a number
of years in the restoration of ground water contaminated by
petroleum hydrocarbons, it has only been in recent years that
in situ systems have been directed toward contaminants in
unsaturated subsurface soils. Research has contributed
greatly to understanding the biotic, chemical, and hydrologic
parameters which contribute to or restrict the application of in
situ bioremediation and has been successful at a number of
locations in demonstrating its effectiveness at field scale.
This document is one in a series of engineering issue papers
which have been prepared in response to needs expressed
by the Engineering Forum. It is based on findings from the
research community in concert with experience gained at
sites undergoing remediation. The intent of the document is
to provide an overview of the factors involved in in situ
bioremediation, outline the types of information required in
the application of such systems, and point out the
advantages and limitations of this technology.
For further information, contact John Matthews (405) 436-
8600 or Dr. Hugh Russell (405) 436-8612.
Background
Bioremediation of contaminated surface soils using in situ
systems, prepared bed, and above-ground bioreactors, has
been previously addressed with regard to characterization,
environmental processes and variables, and field-scale
applications (Sims et al.,1989). This paper will address
processes which are currently being utilized or are in
development to treat contaminated unsaturated subsurface
soils in place.
In situ biological remediation of subsurface soils
contaminated with organic chemicals is an alternative
treatment technology that, in certain cases, can meet the
goal of achieving a permanent cleanup at hazardous waste
sites. Use of such alternatives is encouraged by the U.S.
Environmental Protection Agency (U.S. EPA) for
implementing the requirements of the Superfund
Amendments and Reauthorization Act (SARA) of 1986.
Bioremediation of subsurface soils is consistent with the
philosophical thrust of SARA, for it involves use of naturally
occurring microorganisms to degrade and/or detoxify
hazardous constituents in the soil to protect public health and
the environment. Use of in situ subsurface bioremediation
Utah State University
Robert S. Kerr Environmental Research Laboratory,
U.S. EPA
Technology Innovation Office
Office of Solid Waste and Emergency
Response, US EPA, Washington, D.C.
Walter W. Kovalick, Jr., Ph.D.
Director
Superfund Technology Support Center for
Ground Water
Robert S. Kerr Environmental
Research Laboratory
Ada, Oklahoma
-------
techniques in conjunction with chemical and physical
treatment processes, i.e., "treatment trains," is an effective
means for comprehensive site-specific remediation (Ross et
al., 1988; Sims, 1990). For instance, bioremediation maybe
utilized to lower the concentration of organic contaminants in
a soil matrix before stabilization or solidification is used as a
remedial alternative for metals.
Bioremediation has been shown effective in reducing the
overall mass of a variety of organic contaminants. Full scale
systems have been utilized to remediate soil contaminated
with both crude and refined petroleum hydrocarbons (i.e.,
diesel fuel, gasoline), creosote, and pentachlorophenol. To
date, it has not been shown effective at removing highly
structured, highly insoluble compounds such as
polychlorinated biphenyls and dioxins.
For the purposes of this document, subsurface soil refers to
unsaturated soil within the vadose zone at depths greater
than three feet below the land surface. The vadose zone
extends from the ground surface to the upper surface of the
principal water-bearing formation (Everett et al., 1982). The
vadose zone usually consists of three to six feet of topsoil
(weathered geological materials) which gradually merges with
deeper underlying earth materials such as depositional or
transported clays or sands. In this zone, water primarily
coexists with air, though saturated regions may occur.
Perched water tables may develop at interfaces of layers
(soils having different textures) of soil having less hydraulic
conductivity. Prolonged infiltration also may result in
transient saturated conditions. In some regions, the entire
vadose zone may be hundreds of feet thick and the travel
time of constituents to ground water can be hundreds or
thousands of years. Other regions may be underlain by
shallow potable aquifers that are especially susceptible to
contamination due to short transport times and reduced
potential for pollutant attenuation by soil materials and
processes.
This document addresses specific environmental processes,
factors, and data requirements for characterizing and
evaluating the application of subsurface in situ
bioremediation, and describes selected field-scale
applications of recovery and delivery systems to enhance in
situ subsurface soil bioremediation.
Overview: In Situ Subsurface Microbial Processes
and Controlling Environmental Factors
The rate and extent of biodegradation of organic chemicals
during subsurface in situ bioremediation are influenced by
several site-specific factors. These include type and activity
of microbial populations; chemical environmental factors;
bioavailability of the target chemical(s) and other substrates
required for co-metabolism, i.e., electron donor; mass
transport of moisture, nutrients, and oxygen (the terminal
electron acceptor in aerobic metabolism); toxicity; and
stratigraphy, heterogeneity, and geochemistry of the surface
or subsurface environment. A detailed discussion of the
impact of these and other factors on bioremediation can be
found in 'Transport and Fate of Contaminants in the
Subsurface" (EPA/625/4-89/019) and "Bioremediation of
Contaminated Surface Soils" (EPA/600/9-89/073).
Microbial Populations
Successful in situ bioremediation depends on the presence of
appropriate microbial populations which can be stimulated to
degrade contaminants of concern by modifying or otherwise
managing environmental conditions at a site. Results of
microbial characterization of deep subsurface materials have
indicated that: (1) microorganisms are present at populations
sufficient to change the chemistry of the environment when
stimulated; (2) the microbial communities are diverse and
carry out a wide range of chemical transformations; (3) a
majority (>95%) of the microbes are chemotrophic bacteria
that degrade organic chemicals to obtain energy; and (4)
environmental characteristics identified previously (oxygen
concentration, nutrient status, moisture content) are
important in influencing microbial activity and degradation
patterns (Fliermans and Hazen, 1990).
Microbial communities in the subsurface are diverse and
adaptable. Microbial populations at older sites are usually
acclimated to the contaminants of concern. Therefore, levels
of critical nutrients or electron acceptors, toxicity, and
adverse environmental conditions most often are the major
factors which limit the extent and rate of in situ
bioremediation.
Critical Environmental Conditions
There are several environmental conditions that affect activity
of soil microorganisms. These factors, along with individual
soil and waste characteristics, all interact to affect microbial
activity at specific contaminated sites. Many of these
conditions can be managed to enhance biodegradation of
organic constituents in subsurface soils. Optimum ranges for
the most critical of these factors are presented in Table 1.
Water content of soil is an important factor which regulates
microbial activity. Soil water serves as the transport medium
through which many nutrients and organic constituents
diffuse to the microbial cell, and through which metabolic
waste products are removed. Soil water also affects soil
aeration status, nature and amount of soluble materials,
osmotic pressure, pH of the soil solution, and unsaturated
hydraulic conductivity of the soil (Paul and Clark, 1989). The
water content of deeper subsurface soils may vary greatly.
Unsaturated soil samples have been obtained even from
cores collected below the water table in deep subsurface
environments, and the low water content was shown to
adversely affect microbial activity (Kieft et al., 1990).
Biodegradation rates often depend on the rate at which
terminal electron acceptors can be supplied. A large fraction
of the microbial population within soils are aerobes which use
oxygen as the terminal electron acceptor. Oxygen can be
easily depleted in subsurface soils where there is an oxygen
demand due to plant root respiration or due to normal
2
-------
Table 1. Critical environmental factors tor microbial activity (Sims et al., 1984; Huddleston et a!., 1986; Rochklnd and
Blackburn, 1986; Paul and Clark, 1989)
Environmental Factor
Optimum Levels
Available soil water
Oxygen
Redox potential
pH
Nutrients
25%-85% of water holding capacity. -0 01 MPa
Aerobic metabolism- Greater than 0.2 mg/l dissolved
oxygen, minimum air-filled pore space
of 10%,
Anaerobic metabolism 02 concentrations <1%
Aerobes and facultative anaerobes greater than
50 millivolts, Anaerobes less than 50 millivolts
5.5-8,5
Sufficient nitrogen, phosphorus, and other nutrients so not
limiting to microbial growth
Temperature
Suggested C.N.P ratio of 100 10.1
15°C-45°C (Mesophiles)
microbial activity throughout the depth of the unsaturated
zone. Oxygen levels tend to decrease in soils having high
clay and organic matter content. Clayey soils tend to retain
higher moisture content, which restricts oxygen diffusion,
while organic matter may increase microbial activity and
deplete available oxygen. Under these circumstances,
oxygen may be consumed faster than it can be replaced by
diffusion from the atmosphere, and the soil may become
anoxic.
Facultative anaerobic organisms (which can use oxygen or
alternative electron acceptors such as nitrate or sulfate in the
absence of oxygen) and obligate anaerobic organisms then
become the dominant populations under such conditions.
The sequence of use of various electron acceptors is
determined by the redox potential and the electron affinity of
the electron acceptors present (Zehnder and Stumm, 1988).
The potential of alternative electron acceptors has been
evaluated with nitrate at field scale for contaminants
(including benzene, toluene, and xylene) in an aquifer
environment (Hutchins et al., 1991).
Redox potential also affects metabolic processes in
subsurface microbial populations (Paul and Clark, 1989).
Redox potential provides a measurement of electron density
and is related to the oxygen status of a subsurface soil. As
oxygen is removed and a system becomes more reduced,
there is a corresponding increase in electron density,
resulting progressively in an increased negative potential.
Soil pH affects growth and activity of subsurface soil
microorganisms. Fungi are generally more tolerant of acidic
soil conditions (below pH 5) than bacteria. Solubility of
phosphorus, a critical nutrient in biological systems, is
maximized at a pH value of 6.5. A specific contaminated soil
system may require management of soil pH to achieve levels
that maximize microbial activity. Control of pH to enhance
microbial activity may also aid in the immobilization of
hazardous metals in a subsurface soil system (a pH level
greater than 6 is recommended to minimize metal transport).
Subsurface soil pH may be managed through addition of an
aqueous phase containing pH adjusting chemicals through
gravity delivery systems such as infiltration galleries or
surface irrigation systems.
Microbial metabolism and growth depends upon adequate
supplies of essential macro- and micro-nutrients. Critical
nutrients such as nitrogen and phosphorous must be present
and available to microorganisms in: (1) usable form; (2)
appropriate concentrations; and (3) proper ratios (Dragun,
1988). If wastes are high in carbon (C), and low in nitrogen
(N) and phosphorus (P), biodegradation will cease when
available N and P are depleted. Therefore, fertilization of
subsurface soils may be required as a management
technique to enhance microbial degradation.
Soil temperature affects microbial growth and metabolic
activity. Biodegradation rates decrease as temperature
drops and essentially cease at temperatures below 0° C.
While surface soils exhibit both diurnal and seasonal
variations in temperature, changes of temperature decrease
with depth. Generally, only the top 30 feet of the subsurface
profile are affected by seasonal variations in temperature;
temperature is generally constant and corresponds to the
3
-------
mean annual air temperature of the locality (Kuznetsov et al.,
1963; Matthess, 1982). In the United States, temperatures in
this zone range from 3°C to 25° C (Dunlap and McNabb,
1973). Due to the high specific heat of water, wet soils are
less subject to larger diurnal changes than dry soils (Paul and
Clark. 1989).
Bioavailability is a general term which refers to the
accessibility of contaminants by degrading populations.
There are two major components involved: (1) a physical
aspect related to phase distribution and mass transfer
limitations of the contaminant, and (2) a physiological aspect
related to the suitability of the contaminant as a substrate.
Major factors which affect bioavailability include water
solubility and sorption. Target chemicals may occur in one or
more of the four phases comprising the subsurface soil
environment: (1) soil solids, including organic matter and
inorganic sand, silt, and clay particles; (2) soil water; (3) soil
gas; and (4) often a nonaqueous phase liquid (NAPL). In
general, chemicals that distribute to the water phase (more
soluble) are more bioavailable than chemicals that either sorb
strongly to solid phases or occur in a NAPL phase. NAPLs
are generally degraded from the water:NAPL interface inward
since the aqueous phase contains nutrients, oxygen, and
moisture required for microbial life processes. The
bioavailability of a NAPL phase may be increased by
increasing the surface area to volume ratio of NAPL
elements. This increases mass transfer of nutrients,
moisture, and oxygen; and decreases toxicity by decreasing
interfacial concentrations (Symons and Sims, 1988).
Substrate chemicals in the gas phase have also been found
to be bioavailable (Dupont et al., 1991; Miller et al., 1991).
Generally, chemicals that are highly sorbed, such as high
molecular weight PAHs present in creosote, petroleum, and
manufactured town gas plant wastes, are found to be
degraded at slower rates than chemicals that are only slightly
sorbed. Since the majority of the mass of target constituents
at many contaminated sites is associated with NAPL and/or
solid phases, these represent the greatest challenge with
regard to in situ bioremediation.
Bioavailability is also a function of the biodegradability of the
target chemical, i.e., whether it acts as a substrate,
cosubstrate, or is recalcitrant. The target chemical may be
physically available (i.e., water soluble and/or not sorbed to
solids) but not useful as a metabolic substrate.
Contaminants of concern may not be the dominant organic
substrate in a system. When the target chemical cannot
serve as a substrate (source of carbon and energy) for
microorganisms, but is oxidized in the presence of a
substrate already present or added to the subsurface, the
process is referred to as cooxidation and the target chemical
is defined as the cosubstrate (Keck et al., 1989; Sims et
al.,1989). Cooxidation processes are important for the
biodegradation of high molecular weight polycyclic aromatic
hydrocarbons (PAHs), and some chlorinated solvents,
including trichloroethylene (TCE). Contaminants with
complex molecular structures or high degrees of toxicity may
not be degradable, and may persist or be recalcitrant under
aerobic conditions. Examples of recalcitrant compounds
include highly oxidized halogenated compounds such as
polychlorinated biphenyls (PCBs), pesticides such as
toxaphene, and dioxin contaminants present in wood-
preserving wastes.
The toxicity of the environment may be reduced by
decreasing the concentration of a toxic waste (e.g., creosote)
or chemical (e.g., pentachlorophenol) within one or more
subsurface phases. Concentrations of toxic chemicals in the
gas phase may be reduced through soil vacuum extraction; in
the water phase through soil flushing; in the NAPL phase
through soil flushing with water containing viscosifiers, or with
solvents or surfactants; and in the soil solid phase by
inducing partitioning of contaminants from solid to fluid
phases. All mobile phases in the subsurface have potential
for escape; therefore, containment strategies are often
necessary while the constituents within the phase are
biodegraded.
Heterogeneity of the subsurface environment limits the rate
and extent of in situ bioremediation. Restrictive layers (e.g.,
clay lenses), although more resistant to contamination, are
also more difficult to remediate due to poor permeability and
low rates of diffusion. Clay soils have larger porosities than
silty or sandy soils and therefore larger storage capacities for
contaminants, but have greater resistance to fluid flow
including aqueous, gas, and NAPL phases. Also clay layers
with poor hydraulic conductivity are less permeable to
nutrients and oxygen. In sites that have substantial clay and
silt deposits, more permeable soils will become preferential
conduits for remedial fluids, and the clay/silt deposits will
require much longer time frames for remediation. For
example, heterogeneity of the subsurface with respect to soil
layering and chemical parameters at a gas works site in the
United Kingdom presented constraints on the feasibility of
utilizing in situ bioremediation (Thomas et al„ 1991).
Enhancement of In Situ Subsurface Bioremediation
The method of enhancing in situ bioremediation efforts
depends on the four phases in which contaminants can
occur, heterogeneity of subsurface matrix, and the types of
delivery and recovery systems utilized. Removing limiting or
controlling factors and establishing favorable conditions are
the primary goals of recovery and/or delivery systems.
Enhancement may be achieved by increasing bioavailability;
reducing toxicity; increasing delivery of moisture, nutrients,
and oxygen; and/or by introducing substrates that stimulate
indigenous microbial degradative activity.
A variety of strategies may be implemented to maximize
biodegradation activity in contaminated subsurface soils. The
success of in situ bioremediation efforts is often determined
by the effectiveness of the recovery and delivery systems
used to remove major sources of contaminants and to
transport nutrients and electron acceptors to the location of
the remaining contaminants. Establishing optimum levels of
essential nutrients and electron acceptors at specific
subsurface locations is often driven by physical limitations of
4
-------
the subsurface matrix on transport of fluids (liquids or gases)
used to deliver these amendments. Overcoming these
limitations is the primary goal of a delivery system, and the
development of adequate delivery technologies continues to
be the major challenge ol in situ bioremediation. A summary
of delivery and recovery techniques commonly used to
manage subsurface remediation is provided in Table 2.
Making the Saturated Zone Unsaturated
Advantages of Unsaturated Systems
Because hydraulic conductivity is a function of soil moisture
content, changing a saturated soil into an unsaturated soil
greatly reduces the hydraulic conductivity and therefore the
downward transport of chemicals in the water phase to the
ground water. Also, because oxygen diffuses through air
10,000 times faster than through water, an unsaturated
environment may be maintained in an aerobic condition more
easily than a saturated environment in the presence of
oxygen-demanding chemicals (Table 3). Soil pore space that
contains a gas phase also allows removal of volatile contam-
inants (via soil vacuum extraction) in a direction that is away
from the ground water. Therefore, management of a site to
change the saturated zone to an unsaturated condition may
reduce potential for ground-water contamination as well as
enhanced oxygen delivery to stimulate in situ biodegradation.
Physical Containment
There are a variety of approaches to establishing and
maintaining dewatered conditions. In order to adequately
dewaterthe subsurface, it is often necessary to physically
isolate the treatment zone. Impermeable subsurface barriers
can prevent the migration of ground water by preventing
uncontaminated water from entering the contaminated site
and stopping contaminated water from leaving. Extraction
systems or drains must then be used to remove the ground
water to create an unsaturated zone.
Commonly used barriers include slurry walls, grout curtains,
and sheet piling cutoff walls to retard the flow of water under
and through a site (Devinny et al., 1990).
Ground-Water Removal
Ground-water removal can be accomplished by hydraulic
pumping and/or drainage trenches. Hydraulic pumping using
a well-point system is one such technique (Devinny et al.,
1990) using short lengths of plastic or Teflon well screen
placed in the saturated zone.
Ground water can also be removed using subsurface drains
or drainage ditches. Subsurface drains are constructed by
excavating a trench to the desired depth, partially backfilling
the trench with highly permeable sand or gravel, placing a
plastic or ceramic drain tile in the sand and gravel bed, and
completing the backfilling (Devinny et al., 1990).
Drainage ditches or surface drains are similar to subsurface
drains except that no collection pipes or tiles and backfills are
used. They may be used at sites underlain by poorly
permeable soils (Devinny et al., 1990).
Recovery and Delivery Technologies for Subsurface
Bioremediation
Recovery and delivery technologies are those that facilitate
transport of materials either out of or into the subsurface
(Murdoch et al., 1990). Recovery technologies are primarily
utilized for contaminant source reduction. High levels of
contamination present as either trapped residuals or NAPLs
can severely limit success of bioremediation attempts.
Therefore, removal of as much of this initial contaminant
mass as possible is a prerequisite to in situ bioremediation
efforts.
Specific recovery and delivery technologies for enhancing in
situ bioremediation of subsurface soils are identified in
Table 2. Each identified technology is discussed below with
regard to its applications and limitations, and current status.
Recovery Technologies
The principal recovery technologies used for subsurface
remediation depend on the ability to move fluids. Also
involved is the ability to move contaminants by altering their
solubility or sorption characteristics (Murdoch et al., 1990).
These techniques are used to move materials from the
subsurface soil environment in order to enhance in situ
bioremediation by addressing one or more limiting factors
identified in Tables 1 and 2, including: soil vacuum extraction,
soil flushing, steam stripping, and radio frequency heating.
Soil vacuum extraction
Soil vacuum extraction (SVE) (also referred to as subsurface
or forced air venting, in situ air stripping, or soil vapor
extraction) involves the removal of contaminants carried in
the soil gas phase by reduction of the vapor pressure within
the soil pores by applying a vacuum. As clean air is drawn
through the soil, the contaminants are removed. This
process is driven by concentration differences between solid,
aqueous, and NAPL phases and the clean air that is
introduced through the soil vacuum extraction process.
Vacuum extraction is most applicable to sites contaminated
with highly volatile compounds, such as those associated with
gasoline and solvents (e.g., perchloroethylene,
trichloroethylene, dichloroethylene, trichloroethane, benzene,
toluene, ethylbenzene, and xylene).
Important soil characteristics that should be measured or
estimated to determine the feasibility of vacuum extraction at
a specific site include physical factors that control the rate
and extent of air flow through contaminated soil, and
chemical factors that determine the amount of contaminant
that partitions from soil to air. These factors include: bulk
density (weight per volume), total porosity (void spaces
5
-------
Table 2. Management strategies for addressing factors limiting in situ bloremediation of subsurface soils
Limiting Factor
Management Response
Delivery or Recovery Technique
Bioavailability limited due to NAPL
Reduce NAPL mass
Gravity or forced delivery, Soil flushing,
Steam stripping, Hydraulic fracturing
Bioavailability limited by sorption or slow
mass transport through soil matrix
Reduce sorption, Increase mass transport Soil flushing, Steam stripping, Hydraulic
fracturing
Moisture
Add water or water saturated air
Gravity or forced delivery; Bioventing, Cyclic
pumping
Nutrients
Add nutrients in water or as ammonia gas Gravity or forced delivery; Bioventing, Cyclic
pumping
Oxygen/Redox
Add air
Bioventing, Hydraulic fracturing, Cyclic
pumping, Radial drilling, Kerfing
Toxicity
Remove chemicals
Soil vacuum extraction, Soil flushing, Steam
stripping
pH
Adjust soil pH
Gravity or forced delivery
Temperature
Increase temperature
Radio frequency heating. Steam stripping
Substrate Addition
Add in water or air
Gravity or forced delivery, Bioventing,
Hydraulic fracturing
Heterogeneity
Add or withdraw material in more restrictive
layers
Cyclic pumping, Hydraulic fracturing, Radial
drilling, Kerfing
Table 3. Carrier fluid oxygen supply requirements (Dupont et al., 1991)
Carner g Carrier/g 02
Water
Air Saturated 110,000
Pure 02 Saturated 22,000
500 m^L H202 (100% Utilization) 2,000
Air (20 0% 02) 13
6
-------
between soil grains) and air-filled porosity (that portion of the
total porosity filled with air); diffusivity of volatiles (amount of
volatiles which move through an area overtime); soil
moisture content (percentage of void spaces filled with
water); air phase permeability (ease with which air moves
through soils); texture; structure; mineral content; surface
area; temperature; organic carbon content; heterogeneity;
depth of air permeable zone; and depth to water table
(Metcalf & Eddy, Inc., 1991). Soils at sites where vacuum
extraction is used should be fairly homogeneous and have
high permeability, porosity, and uniform particle-size
distributions (Metcalf & Eddy, Inc., 1991). Soil vapor
transport can be severely limited in a soil with high bulk
density, high soil water or high NAPL content, low porosity,
and low permeability. In heterogeneous soils, air flows
preferentially through more permeable zones, leaving less
permeable zones untreated.
Contaminant characteristics that affect the feasibility of
vacuum extraction include the extent and degree of
contamination, vapor pressure, Henry's law constant,
aqueous solubility, diff usivity, and partition coefficients. Due
to the high solubility of many organic contaminants in NAPL
phases, the presence of NAPL in subsurface soil systems
may significantly affect the distribution of the compounds in
various phases, and their fate in SVE systems. Specific
contaminant and soil conditions that determine the feasibility
of vacuum extraction are presented in Table 4.
The efficiency of a vacuum extraction system can be
enhanced in several ways. For example, a system of air
injection wells can be installed at the perimeter of a
contaminated area (Metcalf & Eddy, Inc., 1991) which can
be connected to air blowers to force air into the soil or
remain open to the atmosphere. Use of air injection wells
can result in increased soil air flow rates and a larger area
through which clean air can move.
Pulsed pumping may be used to give contaminants time to
desorb from solid surfaces, diffuse from restricting layers,
and volatilize from residual saturation (NAPL) in the soil pore
space. Using pulsed pumping for recovery of contaminants
allows a lower volume of air with higher concentrations of
contaminants to be recovered.
If ground water is at or near the zone of soil contamination,
water table rise may occur due to reduced air pressure near
extraction wells (Metcalf & Eddy, Inc, 1991). Ground-water
Table 4. Conditions affecting feasibility of use of vacuum extraction (U.S. EPA, 1990; Metcalf & Eddy, Inc., 1991)
Condition
Favorable
Unfavorable
Contaminant:
Dominant form
Vapor phase
Solid or strongly sorbed to soil
Vapor pressure
>100 mm of mercury
<10 mm of mercury
Water solubility
<100 mg/l
>1,000 mg/l
Henry's Law Constant
>0.01 (dimensionless)
<0 01 (dimensionless)
Soil
Temperature
>20°C (usually will require external heading
of soils)
<10°C (common in northern climates)
Air conductivity
>10"4 cm/s
<10"® cm/s
Moisture content
<10% (by volume)
>10% (by volume)
Composition
Homogeneous
Heterogeneous
Surface area of soil matrix
<0.1 m2/g of soil
>1 0 m2/g of soil
Depth to ground water
>20 m
<1 m
7
-------
pumping may be used to counteract the water table rise, as
well as to expose additional contaminated soil that can be
treated by vacuum extraction.
Horizontal extraction wells (wells drilled parallel to ground
surface) have been used for deep subsurface contamination
at the U.S. Department of Energy Savannah River facility to
access larger areas of the contaminated site (Hazen,1992).
This use of horizontal wells may be a means to reduce costs
associated with deep subsurface remediation since only a
single hole may be required to access contaminated areas
instead of many vertical wells.
The performance of a vacuum extraction system is monitored
by system operational characteristics and by treatment
efficiency characteristics (Metcalf & Eddy, Inc., 1991).
System characteristics include strength of vacuum applied,
air flow rate, and contaminant concentrations and moisture
content in the vented gas. Wells are used to monitor
pressure in the contaminated area. Efficiency of treatment is
monitored by soil gas analyses, and soil core analyses to
determine residual concentration of contaminants. For more
detailed discussions of soil venting evaluation, see
"Evaluation of Soil Venting Application" (EPA/540/S-92/004).
Since soil vacuum extraction is an in situ treatment technique
that requires only addition of ambient air to the subsurface, it
can be applied with little disturbance to existing facilities and
operations (Metcalf & Eddy, Inc., 1991). SVE can be used at
sites where areas of contamination are large and deep, or
when the contamination is beneath a building. The system
can be easily modified, depending on additional analytical
and subsurface characterization data and/or changing site
conditions. Even if vacuum extraction can be implemented at
a site, most of the conditions listed in Table 4 must be met, or
the cost and time for cleanup will be prohibitive.
The use of SVE at remedial sites has been reviewed by the
U.S. EPA (1989a) and classified as a developed technology
for remedial applications. It is currently the most commonly
used in situ remedial technology (Murdoch et al., 1990). Soil
vacuum extraction may be used to reduce toxic concen-
trations of contaminants to levels which are more conducive
to bioremediation. In addition, it will also deliver oxygen to
the subsurface which is required by aerobic bacteria.
Soil flushing
In situ soil flushing is used to accelerate movement of
contaminants through unsaturated materials by solubilizing,
emulsifying, or chemically modifying the contaminants. A
treatment solution is applied to the soil and allowed to
percolate downward and interact with contaminating
chemicals. Contaminants are mobilized by the treatment
solulion and transported downward to a saturated zone
where they are captured in drains or wells and pumped to the
surface for recovery, treatment, or disposal (Murdoch et al.,
1990). In combination with bioremediation, the flushing
solution may be amended with nutrients to enhance biological
activity (Metcalf & Eddy, Inc, 1991).
Treatment solutions are delivered to the contaminated zone
by using either gravity or forced methods. Forced delivery
consists of various pumping techniques. Gravity delivery
methods include surface flooding, ponding, spraying, ditching
and subsurface infiltration beds and galleries (Amdurer et al.,
1986). Barriers, such as slurry walls, may be required to
prevent the transport of contaminants away from the site
(Metcalf & Eddy, Inc. 1991). A ground-water extraction
system must be used to capture the flushing solution and
associated contaminants. In some cases, the flushing
solution may be treated to remove the contaminants and
reused, and in others it may require disposal.
Efficiency of soil flushing is related to two processes: the
increase in hydraulic conductivity that accompanies an
increase in water content of unsaturated soil, and the
selection of treatment solutions with regard to the
composition of the contaminants and the contaminated
medium. The hydraulic conductivity of soils decreases
markedly with decreases in water content; therefore, the flow
of liquids through unsaturated soils is extremely slow and the
recovery of contaminants by conventional pumping
techniques is not possible. With soil flushing, the water
content and consequently the hydraulic conductivity of the
soil is increased (Murdoch et al., 1990). However,
heterogeneities in soil permeability may result in incomplete
removal of contaminants.
At sites where water-soluble contaminants are present, water
can be used to flush or mobilize the contaminants (Metcalf &
Eddy, Inc., 1991). Surfactants can be added to increase the
mobility of hydrophobic organic contaminants, such as oils
and petroleum. Examples of other flushing solutions include:
acidic aqueous solutions (for the removal of metals and basic
organic constituents including amines, ether, and anilines),
basic solutions, chelating agents, oxidizing agents, and
reducing agents. Toxicity of flushing solutions to soil
microorganisms should be considered when followed by
bioremediation of residual contamination. The flushing
solution may change physical and chemical properties of the
soil environment that affect bioremediation potential.
The level of treatment that will be achieved is dependent on
selection of an appropriate flushing solution, extent and time
of contact between the solution and waste constituents, soil
partition coefficients of the waste constituents, and the
hydraulic conductivity of the soil (Metcalf & Eddy, Inc., 1991).
Soil flushing is not applicable to soils with low hydraulic
conductivities (e.g., less than 1 ft/day), or for contaminants
that are strongly sorbed to the soil (e.g., PCBs, dioxin).
Soil flushing has been classified by the U.S. EPA as a
developed technology used for recovery in remedial
applications (Murdoch et al., 1990) Although the technology
has been tested at field-scale, soil flushing has not yet been
used extensively in large-scale clean-up operations. As with
SVE systems, soil flushing may be utilized with bioremed-
iation as a coupled technology. Soil flushing may initially be
utilized to lower toxic or extreme concentrations of contam-
inants to a manageable level for biological processes which
8
-------
may be utilized as a polishing step to remove those
contaminants which were not removed through the flushing
process. If biological processes are used during or after soil
flushing, the compatibility of the soil flushing solution with
subsurface bacteria must always be considered.
Delivery Techniques
The major limiting factor to the bioremediation of amenable
compounds is the delivery of required nutrients, co-oxidation
substrates, electron acceptors or other necessary enhancers
of microbial growth. Delivery techniques are used to add
required materials to the subsurface environment to enhance
in situ bioremediation by addressing one or more limiting
factors identified in Tables 1 and 2. A variety of delivery
techniques are in use or are being developed (Figures 1-3).
These include soil venting, gravity and forced hydraulic
delivery, hydraulic fracturing of low permeability zones, radial
drilling, and cyclic pumping. Of these, only gravity and forced
hydraulic delivery and venting systems are in common use at
sites. The other three approaches are still in developmental
stages.
In Use: Gravity/Forced Hydraulic Delivery and
Bioventing
Gravity and Forced Hydraulic Delivery
Irrigation technologies were among the first utilized for
enhancing in situ biodegradation. Gravity methods are used
to deliver water and amendments to the contaminated
subsurface by applying the solutions directly over the
contaminated area. Applied solutions then percolate
downward through the subsurface to contaminated zones.
Application methods consist of both surface and subsurface
spreading (Amdurer et al., 1986).
Surface application methods include flooding, ponding,
ditches, and sprinkler systems. These methods are generally
applicable to contamination at depths less than 15 feet.
Flooding is a surface application method in which the solution
is spread over the land surface in a thin sheet. Flooding is
applicable to sites that are flat or gently sloped (i.e., less than
3 percent slope), uniform, without gullies or ridges, and have
soils with high hydraulic conductivities (i.e, greater than 10"3
cm/sec; such as those found in sands, loamy sands, and
sandy loams).
Ponding can be used to increase the infiltration rate of the
applied solution above that achieved by flooding. Ponds are
constructed by excavating into the ground or by constructing
low berms The depth of the solution in the pond becomes
the driving force to increase infiltration rates. Ponding can be
used in sandy or loamy soils and in flat areas.
The ditch method of surface spreading utilizes flat-bottomed,
shallow, narrow ditches to transport the solution over the land
surface; allowing for infiltration of the solution into the ground
through both bottom and side surfaces. Gradients in the
ditches are kept small to prevent erosion as well as to allow
residence time for infiltration. Ditches may be constructed by
excavating surface materials or by building small
embankments. Ditches are used at sites where it is not
desirable to completely cover an entire area with the solution.
I Gravel Bed
V v \ v
Figure 1. Schematic of a sprinkling system used to deliver nutrients to contaminated subsurface soil.
9
-------
Extracted Air
ft
Injected Air
Extracted Air
0
Top Soil
Figure 3. Schematic of a ponding system used to deliver nutrients to contaminated subsurface soil.
10
-------
Sprinkler systems can be used to deliver solutions uniformly
and directly to the ground surface. These systems are less
susceptible to topographical constraints than flooding and
ponding. Sprinkler systems have been used successfully to
deliver nutrients and moisture to bioventing systems where
the site was contaminated to a depth of 50 feet (Dupont et
al., 1991).
Subsurface gravity delivery systems include infiltration
galleries (or trenches) and infiltration beds. These systems
are applicable to sites where there is deep contamination or
where the surface layers have low permeability. Subsurface
systems consist of excavations filled with a porous medium
(e.g., coarse sands or gravels) that distribute solutions to the
contaminated area. An infiltration gallery consists of a pit or
trench that is filled with gravel or stones. The solution fills the
pores in the gallery and is distributed to the surrounding soils
in both the vertical and horizontal directions. This system is
most applicable to sites with sandy or loamy soils. In sites
with silty soils, an infiltration gallery can be used but
application rates will be reduced. Solutions can be
introduced into the gallery by injection at locations along the
length of the gallery or through perforated distribution pipe.
Infiltration galleries can be used in sites with steep slopes
(i.e., up to 25 percent slope) and uneven terrain. Infiltration
beds are similar to galleries but are wider and contain more
than one perforated distribution pipe. Infiltration occurs
almost entirely through the bottom, with little infiltration
through sidewall surfaces. This system is applicable to soils
with sandy and loamy textures, but limited to sites where the
topography is relatively flat (i.e., with less than 5 percent
slope) and the terrain is even. Beds can saturate larger
areas than a single trench and are easier to install than a
multi-trench system.
Forced systems deliver fluids under pressure into a
contaminated area through open end or slotted pipes that
have been placed to deliver the solution to the zone requiring
treatment (Amdurer et al., 1986). These systems are
generally applicable to soils with hydraulic conductivities
greater than 10*4 cm/sec (i.e., fine sandy or coarse silty
materials) and high effective porosities (i.e., ranging from 25
to 55 percent). A maximum injection pressure must be
established to prevent hydraulic fracturing and uplift in the
subsurface, which would cause the fluid to travel upward
rather than through the contaminated area. Unlike gravity
systems, a forced delivery system is theoretically
independent of surface topography and climate.
Design considerations for gravity and forced delivery systems
are presented in Amdurer et al. (1986). Application of gravity
delivery systems in subsurface bioremediation systems has
been demonstrated in bioventing systems (Dupont et al.,
1991; Miller et al., 1991). In Russia, methane-oxidizing
bacteria grown in fermenters have been injected into lateral
core holes in a coal mine (Fliermans and Hazen, 1990). This
process has been shown to reduce methane concentration in
the air by 50-60 percent in one month, thus reducing the risk
of explosions and fire.
Soil Bioventing
Soil bioventing incorporates soil vacuum extraction processes
to deliver oxygen to the subsurface to enhance in situ
bioremediation of organic contaminants. The large amounts
of oxygen-saturated water required for bioremediation often
cannot be delivered due to hydraulic conductivity limitations.
For example, benzene and hexane, which are common
hydrocarbon contaminants, require more than 3 g 02 per g of
hydrocarbon for mineralization. Soil bioventing is applicable
to remediation of contaminants of low volatility and can also
reduce concentrations of volatile contaminants in off-gases,
thus reducing the amount of contaminants requiring off-gas
treatment.
To accomplish bioventing, soil vacuum extraction processes
are operated at lower than usual air flow rates to reduce
vapor extraction quantities and maximize vapor retention
times. Soil moisture levels necessary for biological activity
are usually higher than those recommended for optimum
vacuum extraction operations. The addition of nutrients may
also enhance bioremediation. Nutrient addition can be
accomplished by surface application, incorporation by tilling
into surface soil, and transport to deeper layers through
applied irrigation water. Increased soil temperatures have
been shown to enhance biodegradation rates in bioventing
systems (Miller et at., 1991). Possible means of increasing
soil temperature include the use of heated air, heated water,
or low-level radio-frequency heating. High temperature
should be avoided, since this can result in decreased
microbial populations and/or activity.
Soil bioventing has been demonstrated in several field
applications (Dupont et al., 1991; Hinchee et al., 1991;
Hoeppel et. al., 1991; Miller et al., 1991; van Eyk and
Vreeken, 1991; Urlings et al., 1991). At Hill Air Force Base in
Utah, a JP-4 jet fuel spill occurred in January 1985 that
resulted in the contamination of approximately 0.4 hectares
(1 acre) to a depth of approximately 50 feet with approxi-
mately 25,000 gallons of JP-4 (Dupont et al., 1991). Soil total
petroleum hydrocarbon (TPH) concentrations at the site were
as high as 15,000 mg/kg, with average TPH levels of 1,500
mg/kg. Site soil consists of mixed coarse sand and gravel
deposits with interspersed, discontinuous clay stringers to a
confined ground-water table located approximately 600 feet
below ground surface. Prior to initiating a full-scale vacuum
extraction project, the fuel tanks were excavated,
refurbished, and installed in an above-ground concrete
cradle. Excavated soil was placed in a pile and subjected to
vacuum extraction.
An SVE system consisting of 15 wells in the undisturbed soil
and 10 wells in the excavated soil pile and under the tanks
was installed to provide access to the contaminated soil and
allow flexibility in the operation of the venting system. The
system was operated in a conventional mode to maximize
the recovery of volatile components of the JP-4 through
volatilization. Venting was initiated on December 18, 1988, at
a rate of 1,270 ft3/hr (approximately 0.04 pore volumes/day),
11
-------
and gradually increased to approximately 74,000 ft3/hr
(approximately 2.5 pore volumes/day) as the hydrocarbon
levels in the vent gas decreased over time. The venting rate
during the start-up period was limited by the operating
conditions of the catalytic incinerator used to treat the
collected vent gas. This high-rate operating mode was
maintained from December 18, 1988, through September 15,
1989 with approximately 340 pore volumes (245 x 10s ft3) of
soil gas extracted from the site.
In situ respiration tests conducted during the high-rate SVE
operating period indicated that significant respiration was
occurring without nutrient or moisture addition, and that
enhancement of biodegradation might be possible under
modified site management conditions. Biodegradation was a
significant removal mechanism during the initial high-rate
venting, accounting for 15 to 25 percent of the recovered
hydrocarbon. To assess the potential for enhancing
biodegradation rates, a series of laboratory and field
biotreatability studies were conducted to evaluate moisture
and nutrient additions. The effect of SVE system operational
parameters on biodegradation rates was also evaluated by
decreasing airflow rates and increasing flow path length.
A number of in situ respiration tests were conducted during
the field studies to assess the impact of different engineering
management options on microbial activity. A total of three
tests were conducted to monitor the effect of different
management approaches.including: (1) flow rate and
operating configuration modifications , (2) moisture addition,
and (3) moisture and nutrient addition. Biodegradation
reactions were estimated based on cumulative oxygen
consumption and carbon dioxide production. All
biodegradation calculations were normalized to background
C02 and 02 concentrations so that the effects of field
management techniques could be isolated from changes in
background respiration taking place during the study.
The results of these studies indicated that moisture addition
and operational modifications significantly enhanced
biodegradation rates. Based on analyses of 02 uptake rates,
moisture addition (35% to 50% field capacity) was shown to
statistically accelerate in situ respiration at the site. However,
nutrient addition generally did not statistically increase the
degradation rates of residual JP-4 constituents. The
operational modifications (reduced air flow rate, increased
path length) significantly improved biodegradation rates. Fuel
removal due to biodegradation increased to greater than 80
percent, resulting in an additional 12,000 lb of total petroleum
hydrocarbons being degraded during the bioventing portion of
the study. Initial hydrocarbon (on a carbon equivalent basis)
removal rates of 70 lb/day were maintained at an average
rate of greater than 100 lb/day following system operating
modifications.
Soil bioventing was also investigated at Tyndall Air Force
Base in Florida to remediate sandy soils contaminated by
past jet fuel storage activities (Miller, 1990; Miller et at.,
1991). Hydrocarbon concentrations in the soil ranged from
30 to 23,000 mg/kg. The contaminated area was dewatered
prior to system installation. The impact of moisture and
nutrient addition was investigated during a 7-month period.
Moisture addition had no significant effect on biodegradation
rate in this system. Nutrient addition also did not affect
biodegradation rate, since naturally occurring nutrients were
present in adequate quantities to support the amount of
biodegradation observed Biodegradation rates were shown
to be affected by soil temperature and followed predicted
rates based on the van't Hoff-Arrhenius equation. Fifty-five
percent removal was attributed to biodegradation during the
period of study, but a series of flow rate tests showed that
biodegradation could be increased to 85 percent by
decreasing air flow rates. The optimal air flow conditions were
found to be the removal of 0.5 air flow volumes per day. The
contaminated gas phase was drawn through clean soil to
increase gas residence time within the soil. This augmented
in situ biodegradation and eliminated the need for off-gas
treatment as well as reducing exposure to off-gas.
Research: Hydraulic Fracturing, Radial Drilling
Research areas are focusing on methods to increase the
capacity of current systems to deliver increased
concentrations of required solutions to the subsurface. Two
of these systems are discussed below.
Hydraulic fracturing
Hydraulic fracturing is a technique that involves using
hydraulic pressure to induce cracking in rock or clay/silt
lenses in the vicinity of a borehole, which develops a larger
framework of interconnected pore space. The newly created
pore space is filled with solid, granular materials, which can
act as permeable channels to increase the rate and area of
delivery of fluids containing nutrients or oxygen to the
subsurface (Murdoch et al., 1990; Murdoch et al., 1991;
Davis-Hoover et al., 1991). The hydraulic fractures may be
filled with granules of slow-dissolving nutrients or oxygen-
releasing chemicals, which may provide a reservoir of these
compounds for the enhancement of bioremediation. This
technique could also potentially be used in recovery systems,
e.g., by increasing extraction of vapor phases in soils with low
permeabilities, or by forming horizontal sheet-like drains to
capture leachates in soil flushing systems.
Hydraulic fracturing has been successfully utilized in
petroleum engineering in many types of geologic materials,
ranging from granite to poorly consolidated sediments. For
remedial applications, it has been demonstrated in soft clay
soils at shallow depths, but has not yet been demonstrated in
a wide range <5f soils or at waste sites. For use in remedial
applications, hydraulic fracturing has been classified by the
U.S. EPA as an emerging technology (i.e., research on its
use is in progress) (Murdoch et al., 1990).
Radial drilling
Radial well technology consists of drilling horizontal wells
radially oulward Irom a central borehole. This enhances
access to a contaminated subsurface environment by
12
-------
increasing the volume serviced by each vertical well
(Murdoch et al., 1990). Radial wells can be placed at the
same level or on multiple levels in the same borehole. The
use of horizontal wells allows access to fracture zones that
are perpendicular to the ground surface and allows
contaminated areas to be entered laterally rather than
vertically.
Radial wells have been installed in both consolidated rock
and unconsolidated materials (Murdoch et al., 1990). In
unconsolidated formations, drilling rates range from 5 to 120
ft/min, while in very hard, homogeneous basalt, rates range
from 0.10 to 0.50 ft/min. For use in remedial applications,
radial well drilling has been classified by the U.S. EPA as an
emerging technology (i.e., research on its use is in progress)
(Murdoch et al., 1990).
Waste, Soil, and Site Information Requirements for
Evaluation and Management of In Situ
Bioremediation
Adequate site characterization including: surface and
subsurface soil characteristics, hydrogeology, and
microbiological characteristics, serve as the basis for rational
design of any subsurface soil bioremediation system. A
thorough site characterization is necessary to determine both
the three-dimensional extent of contamination as well as
engineering and management constraints which may limit the
rate and extent of remediation. Specific characterization
information regarding waste, soil, and hydrogeology is
required in order to assess the potential effectiveness of
bioremediation. Specific waste characterization information
required includes the relative aerobic biodegradability of
waste chemicals under optimum conditions. Important
hydraulic, physical, and chemical properties of soils that
affect the behavior of organic constituents in the vadose zone
are presented in Sims et al. (1989). Subsurface soil
characterization information required includes identification of
limiting soil environmental factors identified in Table 1.
Required site characterization information includes
identification of potential limiting factors with regard to relative
ease of delivery and recovery listed in Table 2.
Based upon waste, subsurface soil, and site characterization
information, appropriate containment strategies need to be
considered for the mobile contaminant phases associated
with the subsurface (Figure 1). Naturally occurring
containment may be sufficient with regard to preventing
escape of mobile phases under existing site conditions.
However, other containment strategies may need to be
considered if materials are to be added or removed from the
subsurface to stimulate microbial activity. These may include
volatiles removed in vacuum extraction, water used to add
oxygen and nutrients, or NAPLs if soil flushing is carried out.
For each chemical (or chemical class), the information
required is summarized as: (1) characteristics related to
potential leaching, e.g.. water solubility, octanol/water
partition coefficient, solid sorption coefficient; (2) volati-
lization, e.g., vapor pressure, relative volatilization index;
(3) Henry's Law Constant; (4) potential biodegradation, e.g.,
half-life, degradation rate, biodegradability index; and (5)
chemical reactivity, e.g , hydrolysis half-life, soil redox
potential (Sims et al., 1984; Sims et al., 1989).
Information from waste and site characterization studies, and
laboratory evaluations of biodegradation may be integrated
by using appropriate mathematical models to predict: (1) the
potential for bioremediation of and (2) the potential for cross-
contaminating other media (i.e., ground water under the
contaminated area, atmosphere over the site or at the site
boundaries, surface waters, etc). The models used will be
highly dependent on site characteristics and contaminants of
interest. These may range from "back-of-the-envelope"
calculations to sophisticated fate and transport computer
models.
Mass Balance Approach to In Situ Subsurface
Bioremediation
Successful subsurface bioremediation depends upon
thorough characterization and management of each
subsurface phase with regard to containment, stimulation of
microbial activity, and monitoring strategies. The chemical
mass balance approach provides a framework for evaluating,
managing, and monitoring subsurface soil bioremediation
(Sims, 1990). Mass balance helps obtain specific information
that is needed to determine fate and behavior, evaluate and
select management options for in situ bioremediation, and
monitor treatment effectiveness for specific chemicals in
specific subsurface phases. The information needed to
construct a mass balance for subsurface contamination
simultaneously addresses site characterization and
biodegradation rates.
A necessary first step in mass balance requires
characterizing each phase present in the subsurface (Figure
1) with regard to location, amount, and heterogeneity of the
subsurface environment to assess which chemicals are
associated with which phase(s). This information allows
determination of the relative bioavailability of chemicals. For
example, chemicals associated with aqueous and gas phases
are generally more bioavailable than chemicals associated
with solid and NAPL phases. In addition, chemicals
associated with aqueous and gas phases are more prone to
migration. This information also allows determination of the
need for containment by defining where contamination is
migrating under the influence of natural processes. The
problem can be defined in the context of mobility versus
biodegradation for chemicals. Is the rate of biodegradation
(either natural or enhanced) such that chemicals which are
prone to leaching or volatilization degrade before either
occurs'' Using mathematical models or other tools,
chemicals can be ranked in order of their relative tendencies
to leach, volatilize, or remain in-place under subsurface site-
specific conditions. Containment and management options
can then be selected that address specific escape and
attenuation pathways. For example, SVE may be
appropriate as a managerial tool to remove highly volatile,
biologically recalcitrant chemicals from soil before switching
13
-------
to a bioventing mode to remove less volatile, easily
biodegraded compounds. Specific waste phases may be
addressed at specific times during bioremediation. Finally,
comprehensive monitoring programs can be designed to
track specific chemicals in specific phases in the subsurface
at specific times.
After a phase is contained through natural or managed
processes, techniques to enhance microbial activity may be
applied. Monitoring strategies can then be designed to
ensure that the rate and extent of biodegradation within each
phase, as well as transfer of chemicals between phases, are
measured. Biodegradation rates of organic compounds in
soil systems are generally measured by monitoring their
disappearance in a soil through time. Rates of degradation
are often expressed as a function of the concentration of one
or more of the constituents being degraded. This is
accomplished by measuring at specific time intervals the
concentration of contaminants of interest (in the medium of
interest, i.e., soil phase, gas phase, etc.), through a properly
designed sampling and analysis plan. This sampling and
analysis plan should be statistically valid and provide
sufficient information to determine the rate of disappearance
of contaminants of interest or appropriate surrogates, such
as petroleum hydrocarbons (TPH). Care should be taken to
ensure that transfer or partitioning of contaminants from one
phase to another is not misinterpreted as biodegradation
within the source phase. Abiotic losses such as volatilization
and leaching must be defined in order to accurately
determine biodegradation rates. Identification of metabolic
transformation products is also necessary since metabolites
may be more mobile or toxic than the parent compounds. In
addition, measuring only for parent compounds and not
metabolites may tremendously overestimate extent of
biodegradation. In addition, identification of metabolites is
warranted when known daughter products are toxic.
Recommendations
There is currently a lack of information concerning some
aspects of in situ bioremediation of subsurface soils. Specific
areas where additional information is required include site
characterization with regard to effects of physical, chemical,
and hydrologic properties on microbial distribution, numbers,
and activity. Field research to obtain these types of
information is currently limited; however, this information is
required in order to estimate the feasibility of bioremediation
for subsurface contamination. Implementation of subsurface
remediation is currently limited to a significant extent by the
difficulty of establishing adequate systems for delivery and
recovery of chemicals for augmenting biological activity. As
research continues, these difficulties may be overcome as
more information becomes available concerning the
applicability of innovative technologies in the remediaiton of
contaminated soil.
References
Amdurer, M., R.T. Fellman, J. Roetzer, and C. Russ.
Systems to Accelerate In Situ Stabilization of Waste
Deposits. EPA/540/2-86/002, Hazardous Waste Engineering
Research Laboratory, U.S. Environmental Protection Agency,
Cincinnati, OH.
Balkwill, D.L., and F.J. Wobber. 1989. Deep Microbiology
Transitional Program Plan. DOE/ER-0328, Office of Energy
Research, Office of Health and Environmental Research,
U.S. Department of Energy, Washington, DC.
Brown, R.A., Norris, R.A., and Raymond, R.L. 1984. Oxygen
transport in contaminated aquifers with hydrogen peroxide.
Proceedings, Petroleum Hydrocarbons and Organic
Chemicals in Groundwater-Prevention, Detection and
Restoration Conference, Houston, TX, and National Water
Well Association, Worthington, OH.
Davis-Hoover, W.J., L.C. Murdoch, S.J. Vesper, H.R.
Pahren, O.L. Sprockel, C.L. Chang, A. Hussain, and W.A.
Ritschel 1991. Hydraulic fracturing to improve nutrient and
oxygen delivery for in situ bioreclamation. pp. 67-82. In:
In Situ Bioreclamation: Applications and Investigations for
Hydrocarbon and Contaminated Site Remediation (R.E.
Hinchee and R.F. Olfenbuttel, eds.). Butterworth-Heinemann,
Boston, MA.
Dev, H., and D. Downey. 1988. Zapping hazwastes. Civil
Engineering (August): 43-45.
Dev, H„ J.B. Condorelli, C. Rogers, and D. Downey. 1986. In
situ frequency heating process for decontamination of soil,
pp. 332-339. In: Solving Hazardous Waste Problems,
Learning from Dioxins. ACS Symposium Series 338,
American Chemical Society, New York, NY.
Devinny, J.S., L.G. Everett, J.C.S. Lu, and R.L. Stollar. 1990.
Subsurface Migration of Hazardous Wastes. Van Nostrand
Reinhold, New York, NY.
Dragun, J. 1988. The Soil Chemistry of Hazardous Materials.
Hazardous Materials Control Research Institute, Silver
Spring, MD.
Dunlap, W. J., and J. F. McNabb. 1973. Subsurface
Biological Activity in Relation to Ground Water Pollution.
EPA-660/2-73-014, Robert S. Kerr Environmental Research
Laboratory, U.S. Environmental Protection Agency,
Ada, OK.
Dupont, R.R., W.J. Doucette, and R.E. Hinchee. 1991.
Assessment of in situ bioremediation potential and the
application of bioventing at a fuel-contaminated site. pp. 262-
282. In: In Situ Bioreclamation: Applications and
Investigations for Hydrocarbon and Contaminated Site
Remediation (R.E. Hinchee and R.F. Olfenbuttel, eds.).
Butterworth-Heinemann, Boston, MA.
Everett, L.G., E.W. Hoylman, L.G. McMillion, and L.G.
Wilson. 1982. Vadose zone monitoring concepts at landfills,
impoundments, and land treatment disposal areas. In:
Management of Uncontrolled Hazardous Waste Sites.
14
-------
Hazardous Materials Control Research Institute, Silver
Spring, MD.
Fliermans, C.B., andT.C. Hazen (eds.). 1990. Microbiology
of the Deep Subsurface, Proceedings, First International
Symposium. U.S. Department of Energy and Westinghouse
Savannah River Company, Savannah, GA, January 15-19.
Ghiorse, W.C., and J.T. Wilson. 1988. Microbial ecology of
the terrestrial subsurface. Advances in Applied Microbioloqy
33:107-172.
Hazen, T.C. 1992. Full scale underground injection of air,
methane, and other gases via horizontal wells for in situ
bioremediation of chlorinated solvent contaminated ground
water and soil. Proceedings, Bioremediation Case Studies at
Federal Facilities, Oak Ridge National Laboratory, Oak
Ridge, TN, August.
Hinchee, R.E, D.C. Downey, R.R. DuPont, P.K. Aggarwal
and R.N. Miller. 1991. Enhancing biodegradation of
petroleum hydrocarbons through soil venting. Journal of
Hazardous Materials 27:315-325.
Hoeppel, R.E., R.E. Hinchee and M.F. Arthur. 1991.
Bioventing soils contaminated with petroleum hydrocarbons.
Journal of Industrial Microbiology 8:141-146.
Huddleston, R.L., C.A. Bleckmann, and J.R. Wolfe. 1986.
Land treatment biological degradation processes, pp. 41-61.
In: (Land Treatment: A Hazardous Waste Management
Alternative (R.C. Loehr and J.F. Malina, eds.). Water
Resources Symposium No. 13, Center for Research in Water
Resources, University of Texas at Austin, Austin, TX.
Hutchins, S.R., W.C. Downs, G.B. Smith, D.A. Kovacs, D.D.
Fine, R.H. Douglas, and D.J. Hendrix. 1991. Effect of nitrate
addition on biorestoration of fuel-contaminated aquifer: Field
Demonstration. Ground Water 29:571-580.
Keck, J., R.C. Sims, M. Coover, K. Park, and B. Symons.
1989. Evidence for cooxidation of polynuclear aromatic
hydrocarbons in soil. Water Research 23:1467-1476.
Kieft, T.L., L.L. Rosacker, D. Willcox, and A.J. Franklin. 1990.
Water potential and starvation stress in deep subsurface
microorganisms, pp. 4-99 - 4-112. In: Microbiology of the
Deep Subsurface, Proceedings, First International
Symposium (C.B. Fliermans and T.C. Hazen, eds). U.S.
Department of Energy and Westinghouse Savannah River
Company, Savannah, GA, January 15-19.
Kuznetsov, S.I., M.W. Ivanov, and N.N. Lyalikova. 1963.
Introduction to Geological Microbiology. (C.H. Oppenheimer,
ed.). McGraw-Hill, New York, NY.
Madsen, E.L., and J.-M. Bollag. 1989. Aerobic and anaerobic
microbial activity in deep subsurface sediments from the
Savannah River Plant. Geomicrobiology Journal 7:93-101.
Matthess, G. 1982. The Properties of Ground Water. Wiley,
New York, NY.
Metcalf & Eddy, Inc. 1991. Stabilization Technologies for
RCRA Corrective Actions. EPA/625/6-91/026, Center for
Environmental Research Information, U.S. Environmental
Protection Agency, Cincinnati, OH.
Miller, R.N. 1990. A Field Scale Investigation of Enhanced
Petroleum Hydrocarbon Biodegradation in the Vadose Zone
Combining Soil Venting as and Oxygen Source with Moisture
and Nutrient Addition Ph.D. Dissertation. Department of Civil
and Environmental Engineering, Utah State University,
Logan, UT.
Miller, R.N., C.C. Vogel, and R.E. Hinchee. 1991. A field-
scale investigation of petroleum hydrocarbon biodegradation
in the vadose zone enhanced by soil venting at Tyndall AFB,
Florida, pp. 283-302. In: In Situ Bioreclamation: Applications
and Investigations for Hydrocarbon and Contaminated Site
Remediation (R.E. Hinchee and R.F. Olfenbuttel, eds.).
Butterworth-Heinemann, Boston, MA.
Murdoch, L., B. Patterson, G. Losonsky, and W. Harrar.
1990. Technologies of Delivery and Recovery for the
Remediation of Hazardous Waste Sites. EPA/600/2-89/066,
Risk Reduction Engineering Laboratory, U.S. Environmental
Protection Agency, Cincinnati, OH.
Murdoch, L., G. Losonky, P. Cluxton, B. Patterson, I. Klich,
and B. Braswell. 1991. Feasibility of Hydraulic Fracturing of
Soil to Improve Remedial Actions. EPA/600/2-91/012, Risk
Reduction Engineering Laboratory, U.S. Environmental
Protection Agency, Cincinnati, OH.
Noble, D.G. 1963. Well points for dewatering groundwater.
Ground Water 1:21 -36.
Paul, E.A., and F.E. Clark. 1989. Soil Microbiology and
Biochemistry. Academic Press, Inc., San Diego, CA.
Rittmann, B.E., and P.L. McCarty. 1980. Model of steady-
state film biofilm kinetics. Biotechnology and Bioengineering
22:23-43.
Rochkind, M.L., and J.W. Blackburn. 1986. Microbial
Decomposition of Chlorinated Aromatic Compounds. EPA/
600/2-86/090, Hazardous Waste Engineering Research
Laboratory, U.S. Environmental Protection Agency,
Cincinnati, OH.
Ross, D., T.P. Marziarz, and A.L. Bourquin. 1988.
Bioremediation of hazardous waste sites in the USA: Case
histories, pp. 395-397. In- Superfund '88, Proc., 9th National
Conference, Hazardous Materials Control Research Institute,
Silver Spring, MD.
Sims, R.C. 1990. Soil remediation techniques at uncontrolled
hazardous waste sites: A critical review. Journal of the Air &
Waste Management Association 40:704-732.
15
-------
Sims, J.L., R.C. Sims, and J.E. Matthews. 1989.
Bioremediation of Contaminated Surface Soils. EPA/600/9-
89/073, Robert S. Kerr Environmental Research Laboratory,
U.S. Environmental Protection Agency, Ada, OK.
Sims, R.C., D.L. Sorensen, J.L. Sims, J.E. McLean, R.
Mahmood, and R.R. Dupont. 1984. Review of In Place
Treatment Techniques for Contaminated Surface Soils.
Volume 2: Background Information for In Situ Treatment.
EPA/540/2-84-003b, Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency,
Cincinnati, OH.
Stevens, D.K., W.J. Grenney, and Z. Yan. 1988. User's
Manual: Vadose Zone Interactive Processes Model.
Department of Civil and Environmental Engineering, Utah
State University, Logan, UT.
Stevens, D.K., Grenney, W.J., Z. Yan, and R.C. Sims 1989.
Sensitive Parameter Evaluation for a Vadose Zone Fate and
Transport Model EPA/600/2-89/039, Robert S. Kerr
Environmental Research Laboratory, U.S. Environmental
Protection Agency, Ada, OK.
Symons, B.D., and R.C. Sims. 1988. Assessing detoxification
of a complex hazardous waste, using the Microtox bioassay.
Archives of Environmental Contamination and Toxicology
17:497-505.
Thomas, A.O., P.M. Johnston, and J.N. Lester. 1991. The
characterization of the subsurface at former gasworks sites in
respect of in situ microbiology, chemistry, and physical
structure. Hazardous Waste & Hazardous Materials 8: 341-
365.
Urlings, L.G.C.M., F. Spuy, S. Coffa, H.B.R.J, van Vree.
1991. Soil vapour extraction of hydrocarbons: In Situ and On-
Site Biological Treatment, pp. 321-336. In: In Situ
Bioreclamation: Applications and Investigations for
Hydrocarbon and Contaminated Site Remediation (R.E.
Hinchee and R.F. Olfenbuttel, eds.). Butterworth-Heinemann,
Boston, MA.
U.S. EPA. 1988. Interactive Simulation of the Fate of
Hazardous Chemicals during Land Treatment of Oily Wastes:
RITZ User's Guide. EPA/600/8-88-001, Robert S. Kerr
Environmental Research Laboratory, U.S. Environmental
Protection Agency, Ada, OK.
U.S. EPA. 1989a. State of Technology Review: Soil Vapor
Extraction Systems. EPA/600/2-89/024, Risk Reduction
Engineering Laboratory, U.S. Environmental Protection
Agency, Cincinnati, OH.
U.S. EPA. 1989b. The Superfund Innovative Technology
Evaluation Program: Technology Profiles. EPA/540/5-89-013,
Risk Reduction Engineering Laboratory, U.S. Environmental
Protection Agency, Cincinnati, OH.
U.S. EPA. 1990. Assessing UST Corrective Action
Technologies: Site Assessment and Selection of Unsaturated
Zone Treatment Technologies. EPA/600/2-90/011, Risk
Reduction Engineering Laboratory, U.S. Environmental
Protection Agency, Cincinnati, OH.
van Eyk, J. and C. Vreeken. 1991. In Situ and on-site subsoil
and aquifer restoration at a retail gasoline station, pp. 303-
320. In: In Situ Bioreclamation: Applications and
Investigations for Hydrocarbon and Contaminated Site
Remediation (R.E. Hinchee and R.F. Olfenbuttel, eds.).
Butterworth-Heinemann, Boston, MA.
Wobber, F.J. 1989. Deep Microbiology Transitional Program
Implementation Plan. DOE/ER-0431, Office of Energy
Research, Office of Health and Environmental Research,
U.S. Department of Energy, Washington, DC.
Zehnder, A J.B., and W. Stumm. 1988. Geochemistry and
biochemistry of anaerobic habitats. In: Biology of Anaerobic
Microorganisms (A.J.B Zehnder, ed.). John Wiley and Sons,
New York, NY.
16
-------
United States Office of Office of Solid Waste EPA/540/S-92/003
Environmental Protection Research and and Emergency February 1992
Agency Development Response
&EPA Ground Water Issue
In-Situ Bioremediation of
Contaminated Ground Water
J.L. Sims*, J.M. Sufllta", and H.H. Russell0
An emerging technology for the remediation of ground water is
the use of microorganisms to degrade contaminants which are
present in aquifer materials. Understanding the processes
which drive in-situ bioremediation, as well as the effectiveness
and efficiency of the utilization of these systems, are issues
which have been identified by the Regional Superfund Ground
Water Forum as concerns of Superfund decision makers. The
Forum is a group ol ground-water scientists and engineers,
representing EPA's Regional Superfund Offices, organized to
exchange up-to-date information related to ground-water
remediation at Superfund sites.
Although in-situ bioremediation has been used for a number of
years in the restoration of ground water contaminated by
petroleum hydrocarbons, it has only been in recent years that
this technology has been directed toward other classes of
contaminants. Research has contributed greatly to
understanding the biotic, chemical, and hydrologic parameters
which contribute to or restrict the application of in-situ
bioremediation, and has been successful at a number of
locations in demonstrating its effectiveness at field scale.
This document is one in a series of Ground Water Issue
papers which have been prepared in response to needs
expressed by the Ground Water Forum, ft is based on
findings from the research community in concert with
experience gained at sites undergoing remediation. The intent
of the document is to provide an overview of the factors
involved in in-situ bioremediation, outline the types of
information required in the application of such systems, and
point out the advantages and limitations of this technology.
For further information contact Dr. Hugh Russell, RSKERL,
FTS 743-2444, commercial number (405) 332-8800.
Summary
in-situ bioremediation, where applicable, appears to be a
potential cost-effective and environmentally acceptable
remediation technology. Suflita (1989) identified
characteristics of the ideal candidate site for successful
implementation of in-situ bioremediation. These characteristics
included: (1) a homogeneous and permeable aquifer; (2) a
contaminant originating from a single source; (3) a low
ground-water gradient; (4) no free product; (5) no soil
contamination; and (6) an easily degraded, extracted, or
immobilized contaminant. Obviously, few sites meet these
characteristics. However, development of information
concerning site specific geological and microbiological
characteristics of the aquifer, combined with knowledge
concerning potential chemical, physical, and biochemical fate
of the wastes present, can be used to develop a
bioremediation strategy for a less-than-ideal site.
Introduction
In-situ bioremediation is a technology to restore aquifers
contaminated with organic compounds. Organic contaminants
found in aquifers can be dissolved in water, attached to the
aquifer material, or as freephase or residual phase liquids
referred to as NAPLs which are liquids that do not readily
dissolve in water (Palmer and Johnson, 1989c). Generally,
* Soil Scientist, Utah Water Research Laboratory,
Utah State University
b Professor, Dept of Botany and Microbiology, University
of Oklahoma
c Research Microbiologist, Robert S. Kerr Environmental
Research Laboratory
Superfund Technology Support Center for
Ground Water
Robert S. Kerr Environmental
Research Laboratory
Hocy Ada, Oklahoma
TaC^nf^Ogy IrtrKJvfcfiprtOfgo*
Office ol SoBd Waste and Emergency
: US EPA, Washlrigtoft, &C .
Waiter W. KbvaM, i)r., PtiK ..
Director
Printed on Recycled Paper
-------
NAPLs are subdivided into two classes: those that are lighter
than water (LNAPLs density <1.0), and those with a density
greater than water (DNAPLs density >1.0). LNAPLs include
hydrocarbon fuels, such as gasoline, heating oil, kerosene, jet
fuel, and aviation gas. DNAPLs include chlorinated
hydrocarbons, such as 1,1,1 -trichloroethene, carbon tetra-
chloride, chbrophenols, chlorobenzenes, tetrachloroethylene,
PCBs, and creosote.
In this discussion, a technical approach is presented to assess
the potential implementation of bioremediation at a specific
site contaminated with an organic compound. The approach
consists of (1) a site investigation to determine the transport
and fate characteristics of organic waste constituents in a
contaminated aquifer, (2) performance of treatability studies to
determine the potential for bioremediation and to define
required operating and management practices, (3) develop-
ment of a bioremediation plan based on fundamental
engineering principles, and (4) establishment of a monitoring
program to evaluate performance of the remediation effort.
The pattern of contamination from a release of contaminants
into the subsurface environment, as would occur from an
underground leaking storage tank containing NAPLs, is
complex (Figure 1) (Palmer and Johnson, 1989c; Wilson et al.
1989). As contaminants move through the unsaturated zone,
a portion is left behind, trapped by capillary forces. If the
release contains volatile contaminants, a plume of vapors
forms in the soil atmosphere in the vadose zone. If the
release contains NAPLs less dense than water (LNAPLs),
they may flow by gravity down to the water table and spread
laterally. The oily phase can exist either as a free product,
which can be recovered by pumping, or as a residual phase
after the pore spaces have been drained. Contaminants
associated with NAPLs can also partition into the aquifer's
solid phase or in the vapor phase of the unsaturated zone. If
the release contains DNAPLs, these contaminants can
penetrate to the bottom of an aquifer, forming pools in
depressions. In either case, when ground water comes into
contact with any of these phases, the soluble components are
dissolved into the water phase.
There are a number of techniques available to remediate
ground water contaminated with organic compounds including:
physical containment such as slurry walls, grout curtains, and
sheet pilings (Ehrenfield and Bass, 1984); hydrodynamic
control using pumping wells to manipulate the hydraulic
gradient or interceptor systems (Canter and Knox, 1985);
several methods of free product recovery (Lee and Ward
1986); and (4) extraction of contaminated ground water
followed by treatment at the surface (Keely, 1989; U.S. EPA,
1989b).
Alternatively, contaminated ground water can be treated in
place, without extraction using in-situ chemical treatment or
biological treatment with microorganisms (Thomas et al.,
1987c). An advantage of in-situ treatment strategies is that
treatment can take place in multiple phases.
In-situ chemical treatment techniques are similar to methods
used to treat contaminated materials after withdrawal or
excavation, but are directly applied to the materials in place
(Ehrenfield and Bass, 1984). Chemical treatment may involve
neutralizing, precipitating, oxidizing or reducing contaminants
Figure 1. Region* of contamination in • typical release from an underground storage tank (Wilson at al., 1989).
2
-------
by injecting reactive materials into a contaminated leachate
plume through injection wells, but may be limited by mass
transport and concentration dependence. For treatment of
shallow contaminated aquifers, permeable treatment beds
containing reactive materials such as activated carbon or ion
exchange resins may be constructed downgradient to
intercept and treat the contaminated plume.
Biological in-situ treatment of aquifers is usually accomplished
by stimulation of indigenous microorganisms to degrade
organic waste constituents present at a site (Thomas and
Ward, 1989). The microorganisms are stimulated by injection
of inorganic nutrients and, if required, an appropriate electron
acceptor, into aquifer materials.
Most biological in-situ treatment techniques in use today are
variations of techniques developed by researchers at Suntech
to remediate gasoline-contaminated aquifers. The Suntech
process received a patent titled Reclamation of Hydrocarbon
Contaminated Ground Waters (Raymond, 1974). The process
involves the circulation of oxygen and nutrients through a
contaminated aquifer using injection and production wells (Lee
et al., 1988). Placement of the wells is dependent on the area
of contamination and the porosity of the formation, but are
usually no more than 100 feet apart. The nutrient amendment
consists of nitrogen, phosphorus, and other inorganic salts, as
required, at concentrations ranging from 0.005 to 0.02 percent
by weight. Oxygen for use as an electron acceptor in
microbial metabolism is supplied by sparging air into the
ground water. If the growth rate of microorganisms is 0.02 g/l
per day, the process is estimated to require approximately 6
months to achieve 90 percent degradation of the hydro-
carbons present. Cleanup is expected to be most efficient for
ground waters contaminated with less than 40 ppm of
gasoline. After termination of the process, the numbers of
microbial cells are expected to return to background levels.
In addition to stimulating indigenous microbial populations to
degrade organic waste constituents, another technique, which
has not yet been fully demonstrated, is the addition of
microorganisms with specific metabolic capabilities to a
contaminated aquifer (Lee et al., 1988). Populations that are
specialized in degrading specific compounds are selected by
enrichment culturing or genetic manipulation. Enrichment
culturing involves exposure of microorganisms to increasing
concentrations of a contaminant or mixture of contaminants.
The type of organism (or group of organisms) that is selected
or acclimates to the contaminant depends on the source of the
inoculum, the conditions used for the enrichment, and the
substrate. Examples of changes that may occur during an
acclimation period include an increase in population of
contaminant degraders, a mutation that codes for new
metabolic capabilities, and the induction or derepression of
enzymes responsible for degradation of specific contaminants
(Aelion et al., 1987).
h is important to note that the inoculation of a specialized
microbial population into the environment may not produce the
desired degree of degradation for a number of reasons
(Goldstein et al., 1985; Lee et al., 1988; Suflita 1988b).
Factors that may limit the success of inoculants include
contaminant concentration, pH, temperature, salinity, and
osmotic or hydrostatic pressure. They may act alone or
collectively to inhibit the survival of the microorganisms. The
subsurface environment may also contain substances or other
organisms that are toxic or inhibitory to the growth and activity
of the inoculated organisms. In addition, adequate mixing to
ensure contact of the organism with the specific organic
constituent may be difficult to achieve at many sites.
Successful inoculation of introduced organisms into simpler,
more controllable environments (e.g., bioreactors such as
waste-water treatment plants) to accomplish degradation has
been demonstrated. However, effectiveness of inoculation into
uncontrolled and poorly accessible environments such as the
subsurface is much more difficult to achieve, demonstrate and
assess (Thomas and Ward, 1989).
Genetic manipulation of microorganisms to produce
specialized populations to degrade specific contaminants
involves the acceleration and focusing of the process of
evolution (Kilbane, 1986; Lee et al., 1988). Genetic
manipulation can be accomplished by exposure of organisms
to a mutagen, followed by enrichment culturing to isolate a
population with specialized degradative capabilities, or by the
use of DNA recombinant technology to change the genetic
structure of a microorganism. The use of genetically engi-
neered organisms in the environment is illegal without prior
government approval (Thomas and Ward, 1989). In addition,
the introduction of genetically engineered organisms into the
environment would meet the same kind of barriers to success
as organisms developed by enrichment culturing, or more.
Additional methods that have been suggested to enhance
biodegradation include: cross acclimation, which involves the
addition of a readily degradable substrate to aid in the
biodegradation of more recalcitrant molecules; and analog
enrichment, which involves the addition of a structural analog
of a specific contaminant in order to induce degradative
enzyme activity that will affect both the analog and the specific
contaminant (Suflita, 1989a).
In most contaminated aquifers, the hydrogeologic system is so
complex, in terms of site characteristics and contaminant
behavior, that a successful remediation process must rely on
the use of multiple treatment technologies (Wilson et al.,
1986). The combination of several technologies, in series or
in parallel, into a treatment process train may be necessary to
restore ground-water quality to a required level. Barriers and
hydrodynamic containment controls alone serve as only
temporary plume control measures, but can be integral parts
of withdrawal and treatment or in-situ treatment measures.
A possible treatment train might consist of: (1) source removal
by excavation and disposal; (2) free product recovery to
reduce the mass in order to decrease the amount of
contaminants requiring treatment; and (3) in-situ treatment of
remaining contamination. When applicable, biological in-situ
treatment offers the advantage of partial or complete
destruction of organic contaminants, rather than transfer or
partitioning of contaminants to different phases of the
subsurface.
in-Sltu Bioremedlatlon Technical Process
The in-situ bioremediation technical process consists of the
following activities:
1. performance of a thorough site investigation;
2. performance of treatability studies;
3
-------
3. removal of source of contamination and recovery of
free product;
4. design and implementation of the bioremediation
technology; and
5. evaluation of performance of the technology through
a monitoring program (Lee and Ward, 1986; Lee et
al., 1988).
A thorough site investigation in which biological, contaminant,
and aquifer characterization data are integrated, is essential
for the successful implementation of a bioremediation system.
Biological characterization is required to determine if a viable
population of microorganisms is present which can degrade
the contaminants of concern. An assessment of waste
characteristics provides information for determining whether
bioremediation, either alone or as part of a treatment train, is
feasible for the specific contaminants at the site. Aquifer
characteristics provide information on the suitability of the
specific environment for biodegradative processes, as well as
information required for hydraulic design and operation of the
system.
Bioremediation of an aquifer contaminated with organic
compounds can be accomplished by the biodegradation of
those contaminants and result in the complete mineralization
of constituents to carbon dioxide, water, inorganic salts, and
cell mass, in the case of aerobic metabolism; or to methane,
carbon dioxide, and cell mass, in the case of anaerobic
metabolism. However, in the natural environment, a
constituent may not be completely degraded, but transformed
to an intermediate product or products, which may be equally
or more hazardous than the parent compound. In any event,
the goal of in-situ bioremediation is detoxification of a parent
compound to a product or products that are no longer
hazardous to human health and the environment.
In 1973 a review of ground-water microbiology was published
by researchers at the U.S. EPA Robert S. Kerr Environmental
Research Laboratory (RSKERL) (Dunlap and McNabb, 1973)
that stimulated research into microbiology of the subsurface.
Previously, biological activity in the subsurface environment
below the root zone was considered unlikely and that
microbial activity in the subsurface could not be of significant
importance (Lee et al., 1986). However, as methods for
sampling unconsolidated subsurface soils and aquifer
materials without contamination from surface materials
(Dunlap et al., 1977, Wilson et al., 1983, McNabb and Mallard,
1984) as well as methods to enumerate subsurface microbial
organisms (Ghiorse and Wilson, 1988) were developed,
evidence for microbial activity in the subsurface became
convincing.
Bacteria are the predominant form of microorganisms that
have been found in the subsurface, although a few higher life
forms have been detected (Ghiorse and Wilson, 1988; Suflita,
1989a). The majority of microorganisms in pristine and
uncontaminated aquifers are oligotrophic, because organic
materials available for metabolism are likely present in low
concentrations or difficult to degrade. Organic materials that
enter uncontaminated subsurface environments are often
refractory humic substances that resist biodegradation while
moving through the unsaturated soil zone.
Many subsurface microorganisms are metabolically active and
can use a wide range of compounds as carbon and energy
sources, including xenobiotic compounds (Lee et al., 1988).
Compounds such as acetone, ethanol, isopropanol, tert-
butanol, methanol, benzene, chlorinated benzenes,
chlorinated phenols, polycyclic aromatic hydrocarbons, and
alkylbenzenes have been shown to degrade in samples of
subsurface aquifer materials.
The rate and extent of biotransformation of organic
compounds at a specific site are controlled by geochemical
and hydraulic properties of the subsurface (Wilson et al.,
1986). Populations of microorganisms increase until limited by
a metabolic requirement, such as mineral nutrients, substrates
for growth, or suitable electron acceptors. At this point, the
' rate of transformation of an organic material is controlled by
transport processes that supply the limiting factor. Since most
subsurface microorganisms are associated with the solid
phase, the limiting factor must be delivered to the microbes by
advection and diffusion through the mobile phases. Below the
water table, all transport must be through liquid phases, and
as a result, aerobic metabolism may be severely limited by the
very low solubility of oxygen in water. As oxygen becomes
limiting, aerobic respiration slows. However, other groups of
organisms become active and continue to degrade contamina-
ting organic materials. Under conditions of anoxia, anaerobic
bacteria can use organic chemicals or several inorganic
anions as alternate electron acceptors (Suflita, 1989a).
Even though microorganisms may be present in a
contaminated subsurface environment and have
demonstrated the potential to degrade contaminants in
laboratory studies, they may not be able to degrade these
contaminants without a long period of acclimation. Acclimation
results in development of the capability to accomplish
degradation.
In summary, the rate of biological activity in the subsurface
environment is generally controlled by:
1. the concentration of required nutrients in the mobile
phases;
2. the advective flow of the mobile phases or the
steepness of concentration gradients within the
phases;
3. opportunity for colonization in the subsurface by
metabolically active organisms or groups of
organisms capable of degradation of the specific
contaminants present;
4. presence, availability, and activity of appropriate
enzymes for degradation of specific contaminants
present; and,
5. toxicity exhibited by the waste or co-occurring
material(s) (Wilson et al., 1986; Suflita, 1989a).
Methods to Collect Biological Samples
Traditionally, unconsolidated soils or sediments are sampled
through a hollow-stem auger with a split-spoon core barrel or
a conventional thin-walled sample tube (Acker 1974, Scatf et
al., 1981; Wilson et al., 1989). The hollow-stem auger acts as
a temporary casing to keep the borehole open until a sample
can be acquired. A borehole is drilled down to the depth to be
sampled and a core barrel is inserted through the annular
opening in the auger and driven or pushed while rotating the
auger into the earth to collect the sample. These tools are
effective in both unsaturated and saturated cohesive
4
-------
materials, but are not as effective in unconsolidated sands as
it is difficult to keep aquifer material out of the hollow stem
auger (a phenomenon referred to as "heaving") and to keep
the sample in the core barrel while the sample is being
retrieved to the surface. In recent years there have been
many improvements in sampling the subsurface, particularly
with respect to heaving materials (Zapico et al., 1987; Leach
etal., 1988)
Just as it is important to protect the integrity of samples while
coring, it is as important to assure integrity while transferring
sample material to containers which are to be returned to the
laboratory for analysis. To prevent contamination of aquifer
material samples from introduced microorganisms and to
protect samples from the atmosphere to prevent injury of
anaerobic microorganisms, samples are extruded inside a
nitrogen-filled glove box (Figure 2). The glove box is prepared
for sample collection by filling it with the desired number of
sterile sampling jars and sterile paring devices, sealing the
box, and then purging it with nitrogen gas. A slight positive
pressure of nitrogen is maintained in the box by purging during
extrusion and collection of the samples.
Biological Characterization
A wide variety of methods are available to detect, enumerate,
and estimate biomass and metabolic activities of subsurface
microorganisms. These methods include: direct light and
epifluorescence microscopy, viable counts (e.g., plate counts,
most probable number counts, and enrichment culture
procedures), and biochemical indicators of metabolic activity
such as ATP, GTP, phospholipid, and muramic acid (Ghiorse
and Wilson 1988). Levels of microorganisms ranging from 10*
to 10r cells/g of dry aquifer material have been reported from
uncontaminated shallow aquifers (Ghiorse and Balkwill, 1985;
Lee et al., 1988). Often the distribution of microorganisms in
aquifers, as it is in soils, is sporadic and nonuniform, indicating
the presence of micro-environments conducive to growth and
activity.
Waste Characterization
The source of contamination is usually the primary object of
remedial activities (Wilson et al. 1989) as the treatment of
plume areas will not be effective if the source continues to
release contaminants. Information concerning: (1) the areal
location of the source area and contaminant plumes; (2)
amounts of contaminants in the source area; and (3) amounts
of contaminants released into the subsurface are required to
select and apply an appropriate remediation technology and to
determine cost and time requirements for completion of a
remedial action. If in-situ bioremediation is selected as the
remedial technology, information concerning the amount and
distribution of contamination is used in conjunction with
hydrogeological site characteristics to locate injection and
extraction wells and to optimize pumping rates and
concentrations of amendments, such as nutrients and
alternate electron acceptors.
The use of conventional monitoring wells can generally
accurately define the geometry of the ground-water plume
(Palmer and Johnson, 1989a; Wilson et al., 1989). However,
there are important factors that control the quality of
information collected from a network of monitoring wells,
which include the amount of well purging done prior to
sampling (Barcelona and Helfrich, 1986), method of sampling
(Stolzenburg and Nichols, 1985), and method of well
construction and installation (Keely and Boateng, 1987).
Methods for ground-water sampling are presented by Scalf et
al. (1981), Ford et al. (1984), and Barcelona et al. (1985).
Other methods used for detecting contaminant plumes in the
subsurface include geophysical techniques such as surface
resistivity and electromagnetic surveys, chemical time-series
sampling tests (Palmer and Johnson, 1989a), and vapor
Sample Head
Space Analysis
Vent
Flushing Vent
r
j- A
•
tr ®
r-'
Flow Regulator
and Indicator
Sample Tube
from Extruder
Iris Port
Figure 2. Field sampling glove box (Wilson, et al., 1989).
5
-------
monitoring wells (Devitt et al., 1988; Palmar and Johnson,
1989b).
The distribution of the source area and the extent of
contamination should also be characterized by collecting
cores of the solid aquifer materials. Precise information is
required to define the vertical extent of contamination so that
nutrients, oxygen and other amendments injected into the
aquifer contact the contaminants. Injection into a clean part of
the aquifer is a wasted effort and may give the false
impression that the region of aquifer between the injection and
recovery wells is clean (Figure 3).
Additional characteristics of waste contaminants present at a
specific site that should be considered are related to their
environmental fate and behavior in specific aquifer materials
(Armstrong, 1987; Johnson et al., 1989). These character-
istics include physical and chemical properties that determine
recalcitrance, reactivity, and mobility of contaminants at the
site. Information concerning partitioning of contaminants
between aquifer solids and water is especially important. This
information is used to evaluate the extent and rate of release
of contaminants into the ground water, their mobility, and the
quantity of electron acceptors and inorganic nutrients that
must be supplied to support in-situ bioremediation.
Aquifer Characterization
Important geological characteristics of an aquifer that should
be considered during a site investigation include the
composition and heterogeneity of aquifer material, specific
yield, hydraulic connections to other aquifers, magnitude of
water table fluctuations, ground-water flow rate and direction,
hydraulic conductivity distribution, permeability, bulk density,
and porosity (Lee et al., 1988; Palmer and Johnson, 1989a).
Hydraulic conductivity (K) is an especially important
characteristic since the aquifer must be permeable enough to
allow the transport of electron acceptors and inorganic
nutrients to the microorganisms in the zone of contamination.
Permeable aquifer systems, i.e., aquifers with K values of 10"4
cm/sec or greater, are usually considered good candidates for
in-situ bioremediation (Thomas and Ward, 1989).
Hydraulic conductivity of an aquifer can be determined by a
variety of methods (Thomas et al., 1987b, Palmer and
Johnson, 1989a). Knowledge of K values at multiple locations
is necessary because of the heterogeneity of aquifer
materials. Laboratory methods are also available for
determining hydraulic conductivity, but field-measured values
represent average properties over a larger volume and utilize
less disturbed materials (Palmer and Johnson, 1989a).
Aquifer characteristics play an extremely important role in
determining the effectiveness of in-situ bioremediation. Even
in the presence of organisms acclimated to the specific waste
constituents present in an aquifer, biodegradation of
contaminants may be limited by unfavorable aquifer
characteristics that affect microbial activity including:
1. insufficient concentrations of dissolved oxygen for
aerobic metabolism of compounds susceptible to
aerobic degradation;
2. excessive oxygen that inhibits anaerobic
biodegradation of many halogenated compounds in
the subsurface;
Injection
^e" \ Land Surface
+ 4 ¦* + 1r + ¦+ 1
Water
Table
Contaminated : j—k Djrection 0f p|0w cz^>
Interval : ^
j c=^> Wasted c=£>
i
Figure 3. The value of accurately locating the contaminated Interval (Wilson et al., 1989).
6
-------
3. lack of a suitable alternative electron acceptor, if
oxygen is unavailable or not usable;
4. insufficient inorganic nutrients, such as nitrogen,
phosphorus, and trace minerals;
5. presence of toxic metals or other toxicants; and
6. other aquifer characteristics, such as pH, buffering
capacity, salinity, osmotic or hydrostatic pressures,
radiation, sorptive capacity, and temperature
(Armstrong, 1987; Lee et al., 1988).
Treatability Study
A treatability study is designed to determine if bioremediation
is possible at a specific site, and whether there are any
biological barriers to attaining clean-up goals. Even though
the scientific literature may indicate that a specific chemical is
likely to biodegrade in the environment, a treatability study
using site specific variables should be used to confirm that
contention (Suflita, 1989a). Microcosms are generally used to
conduct treatability studies. Pritchard (1981) defined a
microcosm as "a calibrated laboratory simulation of a portion
of a natural environment in which environmental components,
in as undisturbed a condition as possible, are enclosed within
definable physical and chemical boundaries and studied under
a set of laboratory conditions." Microcosms may range from
simple batch incubation systems to large and complex flow-
through devices (Suflita, 1989a).
Results of a treatability study can also provide an estimate of
the rate and extent of remediation that can be attained if
microorganisms are not limited by the rate of supply of an
essential growth factor or by the presence of an unfavorable
environmental factor.
Treatability studies to determine inorganic nutrient and
electron acceptor requirements of subsurface microorganisms
present at a specific site should be conducted using samples
of subsurface solids as well as the ground water. Nutrient and
electron acceptor requirements that will enable indigenous
microorganisms to efficiently degrade organic contaminants
present at a specific site can be determined by incubating
contaminated subsurface materials with combinations of levels
of inorganic nutrients and electron acceptors. Studies should
be performed under conditions that simulate field
environmental conditions. Results of the studies are used to
design the bioremediation program as well as to optimize the
treatment strategy.
Design and Implementation of an In-Sltu
Bioremediation System
Before implementation of an in-situ bioremediation system, the
source of contamination in the soil and in the ground water
should be removed as much as possible. In the case of a
liquid fuel spill, source removal may consist of recovery of
LNAPL free product from the ground water. Depending on the
characteristics of the aquifer and contaminants, free product
can account for as much as 91 percent of the spilled
hydrocarbon, with the remaining hydrocarbon (accounting for
9-40 percent of the spill) sorbed to the soil or dissolved in the
ground water (Lee et al., 1986).
Physical recovery techniques, based on the fact that LNAPL
hydrocarbons are relatively insoluble in and less dense than
water, are used to remove free product from a contaminated
site. Physical recovery often accounts for only 30 to 60
percent of spilled hydrocarbon before yields decline.
Continued pumping of recovery wells may be used to contain
a spill while in-situ bioremediation is being implemented. If a
spill is comprised of DNAPLs, which may sink to the bottom of
the aquifer, physical recovery may be considerably more
difficult to achieve.
Information from the performance of site characterization and
treatability studies may be integrated with the use of
comprehensive mathematical modeling to estimate the
expected rates and extent of treatment at the field scale
(Javandel, 1984; Keely, 1987). The specific model chosen
should incorporate biological reaction rates, stoichbmetry of
waste transformation, mass-transport considerations, and
spatial variability in treatment efficiency (U.S. EPA, 1989a).
After assessment of site characterization and treatability
studies, if results indicate that in-situ bioremediation is
applicable to the site and will be an effective clean-up
technology, the information collected is used to design and
implement the system.
When in-situ bioremediation of a contaminant ground-water
plume involves using methods to enhance the process, such
as the addition of nutrients, additional oxygen sources, or
other electron acceptors, the use of hydraulic controls to
minimize migration of the plume during the in-situ treatment
process may be required (Thomas et al., 1987c; U.S. EPA,
1989a). In general, hydraulic control systems are generally
less costly and time consuming to install than physical
containment structures such as slurry walls. Well systems are
also more flexible, for pumping rates and well locations can be
altered as the system is operated over a period of time.
Pumping-injection systems can be used to: (1) create
stagnation zones at precise locations in a flow field; (2) create
gradient barriers to pollution migration; (3) control the
trajectory of a contaminant plume; and (4) intercept the
trajectory of a contaminant plume (Shafer, 1984). The choice
of a hydraulic control method depends on geological
characteristics, variability of aquifer hydraulic conductivities,
background velocities, and sustainable pumping rates (Lee et
al. 1988). Typical patterns of wells that are used to provide
hydraulic controls include; (1) a pair of injection-production
wells; (2) a line of downgradient pumping wells; (3) a pattern
of injection-production wells around the boundary of a plume;
and (4) the "double-cell" hydraulic containment system. The
"double-cell" system utilizes an inner cell and an outer
recirculation cell, with four cells along a line bisecting the
plume in the direction of flow (Wilson, 1984).
Well systems also serve as injection points for addition of the
materials used for enhancement of microbial activity and for
control of circulation through the contaminated zone. The
system usually includes injection and production wells and
equipment for the addition and mixing of the nutrients (Lee et
al., 1988). A typical system in which microbial nutrients are
mixed with ground water and circulated through the
contaminated portion of the aquifer through a series of
injection and recovery wells is illustrated in Figure 4 (Raymond
et al., 1978; Thomas and Ward, 1989).
Materials can also be introduced to the aquifer through the
use of infiltration galleries (Figure 5) (Brenoel and Brown,
7
-------
Figure 4. Typical schematic for aerobic subsurface bloremedlatfon (Thomas and Ward, 1989).
Air Compressor or Nutrient Addition
Figure 5. Use of Infiltration gallery for recirculation of water and nutrients in in-situ bloremediation (Thomas and Ward, 1989).
8
-------
1985; Thomas and Ward, 1989). Infiltration galleries allow
movement of the injection solution through the unsaturated
zone as well as the saturated zone, resulting in potential
treatment of source materials that may be trapped in the pore
spaces of the unsaturated zone.
Amendments to the aquifer are added to the contaminated
aquifer in alternating pulses. Inorganic nutrients are usually
added first through the injection system, followed by the
oxygen source. Simultaneous addition of the two may result in
excessive microbial growth close to the point of injection and
consequent plugging of the aquifer. High concentrations of
hydrogen peroxide (greater than 10%) can be used to remove
biofouling and restore the efficiency of the system.
Operations Monitoring
Both the operation and effectiveness of the system should be
monitored (Lee et al., 1988). Operational factors of impor-
tance include the delivery of inorganic nutrients and electron
acceptor, the point of the delivery within the aquifer in relation
to the contaminated portion of the plume, and the effective-
ness of containment and control of the contaminated plume.
Measurements of dissolved oxygen and nutrient levels in
ground-water samples are recommended to assess whether
or not bioremediation is being accomplished. Increases in
microbial activities in samples of aquifer materials may be
quantified relative to plume areas prior to treatment, areas
within the plume that did not receive treatment, and control
areas outside the plume. Carbon dioxide levels in ground-
water samples may also be a useful indicator of microbial
activity (Suflita, 1989b).
Measurement of contaminant levels should indicate that
concentrations of contaminants are decreasing in areas
receiving treatment and remaining relatively unchanged in
areas that are not. If degradation pathways of specific
contaminants are known, measurement of presence and
concentrations of metabolic products may be used to
determine whether or not bioremediation is occurring. Both
soil and ground-water samples should be collected and
analyzed to develop a thorough evaluation of treatment
effectiveness. The use of appropriate control samples, e.g.,
assays of untreated areas or areas outside the plume, is
highly recommended to confirm the effectiveness of the
bioremediation technology (Suflita, 1989b).
The frequency of sampling should be related to the time
expected for significant changes to occur along the most
contaminated flow path (U.S. EPA, 1989a). Important
considerations include time required for water to move from
injection wells to monitoring wells, seasonal variations in water
table elevation or hydraulic gradient, changes in the
concentration of dissolved oxygen or alternative electron
acceptor, and costs of monitoring.
Advantages and Limitations In the Use of In-SItu
Bioremediation
There are a number of advantages and disadvantages in the
use of in-situ bioremediation (Lee et al., 1988). Unlike other
aquifer remediation technologies, it can often be used to treat
contaminants that are sorbed to aquifer materials or trapped in
pore spaces. In addition to treatment of the saturated zone,
organic contaminants held in the unsaturated and capillary
zones can be treated when an infiltration gallery is used.
The time required to treat subsurface pollution using in-situ
bioremediation can often be faster than withdrawal and
treatment processes. A gasoline spill was remediated in 18
months using in-situ bioremediation, while pump-and-treat
techniques were estimated to require 100 years to reduce the
concentrations of gasoline to potable water levels (Raymond
et al., 1986). In-situ bioremediation often costs less than other
remedial options. The areal zone of treatment using
bioremediation can be larger than with other remedial
technologies because the treatment moves with the plume
and can reach areas that would otherwise be inaccessible.
There are also disadvantages to in-situ bioremediation
programs (Lee et al., 1988). Many organic compounds in the
subsurface are resistant to degradation. In-situ
bioremediation usually requires an acclimated population of
microorganisms which may not develop for recent spills or for
recalcitrant compounds. Heavy metals and toxic
concentrations of organic compounds may inhibit activity of
indigenous microorganisms. Injection wells may become
clogged from profuse microbial growth resulting from the
addition of nutrients and oxygen.
In-situ bioremediation is difficult to implement in low-
permeability aquifers that do not permit the transport of
adequate supplies of nutrients or oxygen to active microbial
populations. In addition, bioremediation projects require
continuous monitoring and maintenance for successful
treatment.
References
Aelion, C. M., C. M. Swindoll, and F. K. Pfaender. 1987.
Adaptation to and Biodegradation of Xenobiotic Compounds
by Microbial Communities from a Pristine Aquifer. Appl.
Environ. Microbiol. 532212-2217.
Acker, W. L., III. 1974. Basic Procedures for Soil Sampling and
Core Drilling. Acker Drill Co. Scranton, PA.
Armstrong, J. 1987. Some Problems in the Engineering of
Groundwater Cleanup, p. 110-120. In: N. Dee, W. F.
McTernan, and E. Kaplan (eds.) Detection, Control, and
Renovation of Contaminated Ground Water. Am. Soc. Civil
Eng., New York, NY.
Barcelona, M.J. and J.A. Helfrich. 1986. Well Construction and
Purging Effects on Ground-Water Samples. Environ. Sci.
Technol. 20:1179-1184.
Barcelona, M.J., J.P. Gibb, J.A. Helfrich, and E.E. Garske.
1985. Practical Guide for Ground-Water Sampling. EPA/600/
2-85/104, Robert S. Kerr Environmental Research Laboratory,
U.S. Environmental Protection Agency, Ada, OK.
Brenoel, M„ and R.A. Brown. 1985. Remediation of a Leaking
Underground Storage Tank with Enhanced Bioreclamation.
Proc., Fifth National Symp. on Aquifer Restoration and Ground
Water Monitoring. National Water Well Assoc., Worthington,
OH.
9
-------
Canter, L.W., and R.C. Knox. 1985. Ground Water Pollution
Control. Lewis Publishers, Chelsea, Ml.
Devitt, D.A., R.B. Evans, W.A. Jury, T.H. Starks, B. Eklund, A.
Ghalsan. 1988. Soil Gas Sensing for Detection and Mapping
of Volatile Organics. EPA/600/8-87/036, Environmental
Monitoring and Support Laboratory, U.S. Environmental
Protection Agency, Las Vegas, NV.
Dunlap, W.J. and J.F. McNabb. 1973. Subsurface Biological
Activity in Relation to Ground Water Pollution. EPA-660/2-73-
014, Robert S. Kerr Environmental Research Laboratory, U.S.
Environmental Protection Agency, Ada, OK.
Dunlap, W.J., J.F. McNabb, M.R. Scalf, and R.L. Cosby. 1977.
Sampling for Organic Chemicals and Microorganisms in the
Subsurface. EPA-600/2-77-176, Robert S. Kerr Environmental
Research Laboratory, U.S. Environmental Protection Agency,
Ada, OK.
Ehrenfield, J. and J. Bass. 1984. Evaluation of Remedial
Action Unit Operations of Hazardous Waste Disposal Sites.
Pollution Technology Review No. 110, Noyes Publications,
Park Ridge, NJ.
Ford, P.J., P.J. Turina, and D.E. Seely. 1984. Characterization
of Hazardous Waste Sites - A Methods Manual, Volume II:
Available Sampling Methods. EPA/600/4-84/076. Robert S.
Kerr Environmental Research Laboratory, U.S. Environmental
Protection Agency, Ada, OK.
Ghiorse, W.C., and D.L. Balkwill. 1985. Microbiology of
Ground Water Environments. In: G. E. Januer (ed.) Progress
in Chemical Disinfection II: Problems at the Frontier. State
Univ. of New York (SUNY) at Binghamton, Binghamton, NY.
Ghiorse, W.C., and J.T. Wilson. 1988. Microbial Ecology of the
Terrestrial Subsurface. Adv. Appl. Microbiol. 33:107-172.
Goldstein, R.M., LM. Maltory, and M. Alexander. 1985.
Reasons for Possible Failure of Inoculation to Enhance
Biodegradation. Appl. Environ. Microbiol. 50:977-983.
Javandel, I., C. Doughty, and C.F. Tsang. 1984. Groundwater
Transport: Handbook of Mathematical Models. Water
Resources Monograph No. 10, Am. Geophysical Union,
Washington, DC.
Johnson, R.L, C.D. Palmer, and W. Fish. 1989. Subsurface
Chemical Processes. In: Transport and Fate of Contaminants
in the Subsurface. EPA/625/4-89/019, Robert S. Kerr
Environmental Research Laboratory, U.S. Environmental
Protection Agency, Ada, OK.
Keely, J.F. 1987. The Use of Models in Managing Ground-
Water Protection Programs. EPA/600/8-87/003. Robert S.
Kerr Environmental Research Laboratory, U.S. Environmental
Protection Agency, Ada, OK.
Keely, J.F. and K. Boat eng. 1987. Monitoring Well Installation,
Purging, and Sampling Techniques. Part II: Case Histories.
Ground Water 25:427-439.
Kilbane, J J. 1986. Genetic Aspects of Toxic Chemical
Degradation. Microbiol. Ecol. 12:135-145.
Leach, L.E., F.P. Beck, J.T. Wilson, and D.H. Kampbell. 1988.
Aseptic Subsurface Sampling Techniques for Hollow-Stem
Auger Drilling. Proc., Second National Outdoor Action
Conference on Aquifer Restoration, Ground Water Monitoring
and Geophysical Methods 1:31 -51.
Lee, M.D. and C.H. Ward. 1986. Ground Water Restoration.
Report submitted to JACA Corporation, Fort Washington, PA.
Lee, M.D., J.M. Thomas, R.C. Borden, P.B. Bedient, J.T.
Wilson, and C. H. Ward. 1988. Biorestoration of Aquifers
Contaminated with Organic Compounds. CRC Critical Rev.
Environ. Controls 1829-89.
McNabb, J.F., and G.E. Mallard. 1984. Microbiological
Sampling in the Assessment of Groundwater Pollution. In: G.
Bitton and C.P. Gerba (eds.), Groundwater Pollution
Microbiology. John Wiley & Sons, New York, NY.
Palmer, C.D., and R.L. Johnson. 1989a. Determination of
Physical Transport Parameters. In: Transport and Fate of
Contaminants in the Subsurface. EPA/625/4-89/019, Robert
S. Kerr Environmental Research Laboratory, U.S.
Environmental Protection Agency, Ada, OK.
Palmer, C.D., and R.L. Johnson. 1989b. Physical Processes
Controlling the Transport of Contaminants in the Aqueous
Phase. In: Transport and Fate of Contaminants in the
Subsurface. EPA/625/4-89/019, Robert S. Kerr Environmental
Research Laboratory, U.S. Environmental Protection Agency,
Ada. OK.
Palmer, C.D., and R.L. Johnson. 1989c. Physical Processes
Controlling the Transport of Non-Aqueous Phase Liquids in
the Subsurface. In: Transport and Fate of Contaminants in the
Subsurface. EPA/625/4-89/019, Robert S. Kerr Environmental
Research Laboratory, U.S. Environmental Protection Agency,
Ada. OK.
Pritchard, P.H. 1981. Model Ecosystems. ln:R.A. Conway
(ed.), Environmental Risk Analysis for Chemicals. Van
Nostrand Reinhold, New York, NY.
Raymond, R.L. 1974. Reclamation of Hydrocarbon
Contaminated Ground Waters. U.S. Patent Office,
Washington, DC. Patent No. 3,846,290. Patented
November 5,1974.
Raymond, R.L., R.A. Brown, R.D. Norris, and E.T. O'Neill.
1986. Stimulation of Biooxidation Processes in Subterranean
Formations. U.S. Patent Office, Washington, DC. Patent No.
4,588,506. Patented May 13, 1986.
Raymond, R.L., V.W. Jamison. J.O. Hudson, R.E. Mitchell,
and V.E. Farmer. 1978. Field Application of Subsurface
Biodegradation of Gasoline in a Sand Formation. API Publi-
cation 4430, American Petroleum Institute, Washington, DC.
Scalf, M.R., J.F. McNabb, W.J. Dunlap, R.L. Cosby, and J.
Fryberger. 1981. Manual of Ground Water Sampling
Procedures. National Water Well Assoc., Worthington, OH.
Schafer, J.M. 1984. Determining Optimum Pumping Rates for
Creation of Hydraulic Barriers to Ground Water Pollutant
Migration. In: D.M. Nielsen (ed.), Proc., Fourth National Symp.
10
-------
on Aquifer Restoration and Ground Water Monitoring. National
Water Well Assoc., Worthington, OH.
Stolzenburg, T.R. and D.G. Nichols. 1985. Preliminary Results
on Chemical Changes in Groundwater Samples Due to
Sampling Devices. EPRI Report No. EA-4118, Electric Power
Research Institute, Palo Alto, CA.
Suflita, J.M. 1989a. Microbial Ecology and Pollutant
Biodegradation in Subsurface Ecosystems. In: Transport and
Fate of Contaminants in the Subsurface. EPA/625/4-89/019,
Robert S. Kerr Environmental Research Laboratory, U.S.
Environmental Protection Agency, Ada, OK.
Suflita, J.M. 1988b. Microbiological Principles Influencing the
Biorestoration of Aquifers. In: Transport and Fate of
Contaminants in the Subsurface. EPA/625/4-89/019, Robert
S. Kerr Environmental Research Laboratory, U.S.
Environmental Protection Agency, Ada, OK.
Thomas, J.M., and C.H. Ward. 1989. In Situ Biorestoration of
Organic Contaminants in the Subsurface. Environ. Sci.
Technol. 23:760-766.
Thomas, J.M., M.D. Lee, and C.H. Ward. 1987a. Use of
Ground Water in Assessment of Biodegradation Potential in
the Subsurface. Environ. Toxicol. Chem. 6:607-61 4.
Thomas, J.M., H J. Marlow, R.L. Raymond, and C.H. Ward.
1987b. Hydro logic Considerations for In Situ Biorestoration.
US/USSR Symposium on Fate of Pesticides and Chemicals in
the Environment, Oct. 12-16, Iowa City, IA (under review by
John Wiley & Sons, Inc., New York, NY).
Thomas, J.M., M.D. Lee, P.B. Bedient, R.C. Borden, LW.
Canter, and C.H. Ward. 1987c. Leaking Underground Storage
Tanks: Remediation with Emphasis on In Situ Biorestoration.
EPA/600/2-87/008, Robert S. Kerr Environmental Research
Laboratory, U.S. Environmental Protection Agency, Ada, OK.
U.S. Environmental Protection Agency. 1989a. Bioremediation
of Hazardous Waste Sites Workshop: Speaker Slide Copies
and Supporting Information. CERI-89-11, Center for
Environmental Research Information, U.S. Environmental
Protection Agency, Cincinnati, OH.
Wilson, J.L 1984. Double-cell Hydraulic Containment of
Pollutant Plumes. In: D. M. Nielsen (ed.), Proc., Fourth
National Symp. on Aquifer Restoration and Ground Water
Monitoring. National Water Well Assoc., Worthington, OH.
Wilson, J.T., and B.H. Wilson. 1985. Biotransformation of
Trichloroethylene in Soil. Appi. Environ. Microbiol. 49:242-243.
Wilson, J.T. and C.H. Ward. 1987. Opportunities for
Bioreclamation of Aquifers Contaminated with Petroleum
Hydrocarbons. Dev. Industrial Microbiol. (J. Industrial
Microbiol. Suppl. 1 ) 27:109-116.
Environmental Research Laboratory, U.S. Environmental
Protection Agency, Ada, OK.
Wilson, J.T., J.F. McNabb, D.L Balkwill, and W.C. Ghiorse.
1983. Enumeration and Characterization of Bacteria
Indigenous to a Shallow Water-Table Aquifer. Ground Water
21:134-142.
Wilson, J.T., L.E. Leach, M.J. Henson, and J.N. Jones. 1986.
In Situ Biorestoration as a Ground Water Remediation
Technique. Ground Water Monitoring Rev. 6:56-64.
Zapico, M.M., S. Vales, and J.A. Cherry. 1987. A Wireline
Piston Core Barrel for Sampling Cohesionless Sand and
Gravel Below the Water Table. Ground Water Monitoring Rev.
7(3):74-82.
Wilson J., L. Leach, J. Michalowski, S. Vandegrift, and R.
Callaway. 1989. In Situ Bioremediation of Spills from
Underground Storage Tanks: New Approaches for Site
Characterization, Project Design, and Evaluation of
Performance. EPA/600/2-89/042, Robert S. Kerr
11
it U.S. GOVERNMENT PRINTING OFFICE: l*»2 - «4!MX)3/407Z«
-------
BULK RATE
United States Center for Environmental Research POSTAGE & FEES PAID
Environmental Protection Information EPA
Agency Cincinnati, OH 45268 PERMIT No. G-35
Official Business
Penalty for Private Use, $300
Please make all necessary changes on the above label,
detach or copy, and return to the address In the upper
left-hand comer.
R you do not wish to recede these reports CHECK HERE Q ;
detach, or copy this cover, and return to the address In the
upper left-hand comer.
E PA/540/S-92/003
-------
United States Office of Office of Solid Waste EPA/540/S-92/004
Environmental Protection Research and and Emergency April 1992
Agency Development Response
&EPA Ground Water Issue
Evaluation of Soil Venting Application
Dominic C. DiGiulio*
Introduction
The Regional Superfund Ground-Water Forum is a group of
scientists, representing EPA's Regional Superfund Offices,
organized to exchange up-to-date information related to
ground-water remediation at Superfund sites. One of the
major issues of concern to the Forum is the transport and fate
of contaminants is soil and ground water as related to
subsurface remediation.
The ability of soil venting to inexpensively remove large
amounts of volatile organic compounds (VOCs) from
contaminated soils is well established. However, the time
required using venting to remediate soils to low contaminant
levels often required by state and federal regulators has not
been adequately investigated. Most field studies verify the
ability of a venting system to circulate air in the subsurface
and remove, at least initially, a large mass of VOCs. They do
not generally provide insight into mass transport limitations
which eventually limit performance, nor do field studies
generally evaluate methods such as enhanced biodegradation
which may optimize overall contaminant removal. Discussion
is presented to aid in evaluating the feasibility of venting
application. Methods to optimize venting application are also
discussed.
For further information contact Dominic DiGiulio (405)332-
8800 or FTS 700-743-2271 at RSKERL-Ada.
Determining Contaminant Volatility
The first step in evaluating the feasibility of venting application
at a hazardous waste site is to assess contaminant volatility. If
concentrations of VOCs in soil are relatively tow and the
magnitude of liquid hydrocarbons present in the soil is
negligible, VOCs can be assumed to exist in a three-phase
system (i.e., air, water, and soil), as illustrated in Figure 1. If
soils are sufficiently moist, relative volatility in a three-phase
system can be estimated using equation (1) which
incorporates the effects of air-water partitioning (Henry's
constant) and sorption (soil-water partition coefficient).
h
C.
1
(PaVoc^h) + e/Kh + ~
(1)
where:
C. -
C. -
1 -
K,
e
~
Vapor concentration of VOCs in gas phase(mg/
cm® air)
Total volatile organic concentration (mg/cm3 soil)
Bulk density (g/cm3)
Organic carbon-water partition coefficient
(cm3/g)
Fraction of organic carbon content (g/g)
Henry's Constant (mg/cm3air/mg/cm3water)
Volumetric moisture content (cm3/cm3)
Volumetric air content (cm3/cm3)
Caution must be exercised when using this approach since
this relationship is based on the assumption that solid phase
sorption is dominated by natural organic carbon content. This
assumption is frequently invalid in soils below the root zone
where soil organic carbon is less than 0.1%.
Equation (1) can be used to evaluate individual VOC
contaminant reduction trends and attainment of soil-based
Environmental Engineer, Roberts. Kerr Environmental
Research Laboratory
Superfund Technology Support Center for
Ground Water
Robert S. Kerr Environmental
Research Laboratory
*4.
-------
/
"\
Air
V
J
k„ ©v©
< >
Kp = Soil-water partition coefficient
Kh = Henry's Constant
Figure 1. Three phase system.
remediation standards. Vapors should be collected from
dedicated vapor probes under static (venting system not
operating) conditions. This estimate is valid only for soils in the
immediate vicinity of the probe intake. This approach
minimizes sample dilution and collection of vapor samples
under nonequilibrium conditions. It, however, necessitates
periodic cessation of venting. When the vapor concentration
for a VOC approaches a corresponding total soil
concentration, actual soil samples can be collected to confirm
remediation. This approach has several benefits over
conventional soil samples collection and analysis. At lower
VOC concentration levels, collection of static vapor samples is
likely more sensitive than soil collection and analysis due to
VOC loss in the latter procedure. Siegrist and Jenssen (1990)
demonstrated substantial VOC loss during normal soil sample
collection, storage, and analysis. Also, comparing contaminant
reduction trends strictly with soil samples is difficult due to
spatial variability in soils. No two soil samples can be collected
at the exact same location. In addition, soil gas analyses can
be accomplished more quickly and inexpensively than soil
sample collection, thus enabling more frequent evaluation of
trends. A potential disadvantage of using this approach is
inability to distinguish VOC vapors emanating from soils as
opposed to ground water. Hypothetically, soils could be
remediated to desired levels with probes still indicating
contamination above remediation standards. This concern
could be alleviated to some degree by determining the
presence of a diffusion vapor gradient from the water table
using vertically placed vapor probes.
If soils are visibly contaminated or the presence of
nonaqueous phase liquids (NAPLs) is suspected in soils
based on high contaminant, total organic carbon, or total
petroleum hydrocarbon analysis, contaminants are likely
present in a four phase system as illustrated in Figure 2.
Under these circumstances, most of the VOC mass will be
associated with the immiscible fluid and assuming that the
fluid acts as an ideal solution, volatilization will be governed by
Raoult's Law.
Pa-XaP°a (2)
where:
Pa - vapor pressure of component over
solution (mm Hg)
Figure 2. Four phase system.
Xfl - mole fraction of component in
solution
P°a - saturated vapor pressure of pure
component (mm Hg)
In a four-phase system, contaminant volatility will be governed .
by the VOC's vapor pressure and mole fraction within the
immiscible fluid. The vapor pressure of all compounds
increases substantially with an increase in temperature while
solubility in a solvent phase is much less affected by
temperature. This suggests that soil temperature should be
taken into account when evaluating VOC recovery for
contaminants located near the soil surface (seasonal
variations in soil temperature quickly dampen with depth). For
instance, if conducting a field test to evaluate potential
remediation of shallow soil contamination in the winter, one
should realize that VOC recovery could be substantially
higher during summer months, and low recovery should not
necessarily be viewed as venting system failure.
As venting proceeds, lower molecular weight organic
compounds will preferentially volatilize and degrade. This
process is commonly described as weathering and has been
examined by Johnson (1989) in laboratory experiments.
Samples of gasoline were sparged with air and the
concentration and composition of vapors were monitored. The
efficiency of vapor extraction decreased to less than 1% of its
initial value even though approximately 40% of the gasoline
remained. Theoretical and experimental work on product
weathering indicate the need to monitor temporal variation in
specific VOCs of concern in extraction and observation wells.
Evaluating Air Flow
Air permeability (kfl) in soil is a function of a soil's intrinsic
permeability (k,) and liquid content. At hazardous waste sites,
liquid present in soil pores is often a combination of soil water
and immiscible fluids. Air permeability (k ) can be estimated
by multiplying a soil's intrinsic permeability (k) by the relative
permeability (kr). 1
The dimensionless ratio k, varies from one to zero and
describes the variation in air permeability as a function of air
2
-------
saturation. Equations developed by Brooks and Corey (1964)
and Van Genuchten (1980) are useful in estimating air
permeability as a function of air saturation or liquid content.
The Brooks-Corey equation to estimate relative permeability of
a non-wetting fluid (i.e. air) is given by:
kf-(1 -S/ (1 -Se(2+X)/X) (4)
where:
Se » effective saturation
X > a pore distribution parameter
placement and screened intervals of extraction and
observation wells and applied vacuum rates during a pump
test is often based on preliminary mathematical modeling.
Evaluating Mass Transfer Limitations and
Remediation Time
The effects of mass transport limitations are usually
manifested by a substantial drop in soil vapor contaminant
concentrations as illustrated in Figure 4 or by an asymptotic
increase in total mass removal with operation time. Typically,
when venting is terminated, an increase in soil gas
The effective saturation is given by:
S„ .
(ill)
(5)
Where:
e
c
0- ¦
volumetric moisture content
total porosity
residual saturation
The pore size distribution parameter and residual water con-
tent can be estimated using soil-water characteristic curves
which relate matric potential to volumetric water content.
When initially developing an estimate of relative permeability
for a given soil texture and liquid content, values for e, 6^ S#,
and X can be obtained from the literature. Rawls et al.
(1982) summarized geometric and arithmetic means for
Brook-Corey parameters for various USDA soil texturai
classes. Figure 3 illustrates relative permeability as a function
of volumetric moisture content for clayey soils assuming e ¦
0.475,0r - 0.090, and X - 0.131.
The most effective method of measuring air permeability is by
conducting a field pneumatic pump test Using permeameters
or other laboratory measurements provide information on a
relatively small scale. Information gained from pneumatic
pump tests is vital in determining site-specific design
considerations (e.g., spacing of extraction wells). Selecting the
0.2 0.3 a4 as
Moisture (8)
Figure 3. Relative permeability vs moisture content of day.
Re-Start Yield Spike
Time
Figure 4. Concentration vs. time.
concentration is observed over time. Slow mass transfer with
respect to advective air flow is most likely caused by diffusive
release from porous aggregate structures or lenses of lesser
permeability as illustrated in Figure 5. The time required for
the remediation of heterogeneous and fractured soils depends
on the proportion of contaminated material exposed to direct
bulk airflow. It would be expected that long-term performance
of venting will be limited to a large degree by gaseous and
liquid diffusion from soil regions not exposed to direct airflow.
Regardless of possible causes, the significance of mass
transport limitations should be evaluated during venting field
tests. This can be achieved by pneumatically isolating a small
area of a site and aggressively applying vacuum extraction
until mass transport limitations are realized. Isolation can be
achieved by surrounding extraction wells with passive inlet or
air injection wells as shown in Figure 6. Quantifying the
effects of mass transport limitations on remediation time might
then be attempted by utilizing models incorporating mass
transfer rate coefficients.
The discrepancy frequently observed between mass removal
predicted from equilibrium conditions using Henry's Law
constants and that observed from laboratory column and field
studies is sometimes reconciled by the use of "effective or
lumped' soil-air partition coefficients. These parameters are
determined from laboratory column tests and are then used for
model input to determine required remediation times. While
this method does indirectly account for mass transport
3
-------
Air Advection
/k k k k k i\ ^ S A A A A A \
< Molecular Diffusion > — \ t t t t t/
xt ~ ~ ~ ~ >*.
X a—-
LowPermeabflty AA A\
Strata XT T T T T/
\ „
<$ i i i ti
..... ^ h^lecular ^itluskyi^\
Air Advection \NLJ_i_Lij/
PAMralat. IBM
Figure 5. Effect of Low Permeability Lenses.
limitations, problems may arise when one attempts to
quantitatively describe several processes with lumped
parameters. The primary concern is whether the lumped
parameter is suitable for use only under the laboratory
conditions from which it was determined, or whether it can be
transferred for modeling use in the field. Perhaps the most
direct method of accounting for mass transport limitations
would be to incorporate diffusive transfer directly into
convective-dispersive vapor transport models.
Enhanced Aerobic Blodegradatlon
With the exception of a few field research projects, soil
vacuum extraction has been applied primarily for removal of
volatile organic compounds from the vadose zone. However,
circulation of air in soils can be expected to enhance the
aerobic biodegradation of both volatile and semivolatile
organic compounds. One of the most promising uses of this
technology is in manipulating subsurface oxygen levels to
maximize in-situ biodegradation. Bioventing can reduce vapor
treatment costs and can result in the remediation of
semivolatile organic compounds which cannot be removed by
physical stripping alone.
Venting circulates air in soils at depths much greater than are
possible by tilling, and oxygen transport via the gas phase is
much more effective than injecting or flooding soils with
oxygen saturated liquid solutions.
Hinchee (1989) described the use of soil vacuum extraction at
Hill AFB, Utah for oxygenation of the subsurface and the
enhancement of biodegradation of petroleum hydrocarbons in
soils contaminated with JP-4 jet fuel. Figures 7 and 8 illustrate
subsurface oxygen profiles at the Hill site prior to and during
venting. It is evident that soil oxygen levels dramatically
increased following one week of venting. Soil vapor samples
collected from observation wells during periodic vent system
shutdown revealed rapid decreases in oxygen concentration
and corresponding C02 production suggesting that aerobic
biodegradation was occurring at the site. Laboratory
treatability studies using soils from the site demonstrated
increased carbon-dioxide evolution with increasing moisture
content when enriched with nutrients. It is worthwhile to note
that soils at Hill AFB were relatively dry at commencement of
field vacuum extraction indicating, that the addition of moisture
could perhaps stimulate aerobic biodegradation even further
under field operating conditions.
When conducting site characterization and field studies, it is
recommended that C02 and 02 levels be monitored in soil
vapor probes and extraction well offgas to allow the
assessment of basal soil respiration and the effects of site
management on subsurface biological activity. These
measurements are simple and inexpensive to conduct and
can yield a wealth of information regarding:
1. The mass of VOCs and semivolatiles which have
undergone biodegradation versus volatilization. This
information is crucial if subsurface conditions (e.g.,
moisture content) are to be manipulated to enhance
biodegradation to reduce VOC offgas treatment oosts
and maximize semivolatile removal.
2. Factors limiting biodegradation. If 02 and CO, monitoring
reveals low 02 consumption and C02 generation while
readily biodegradable compounds persist in soils, further
characterization studies could be conducted to determine
10
o
10 -
0 24ft
J 30
7.32 meten
o Vmonf Probe duller
1 40
* Pusive Inlet Well
SO
¦ Vent Well
A Borahols Sampling Location*
CO
DfOalio, 1H9
70
20
30
Disttikoe (CooO
40 50
60
70
90
'^U/-
_
5* ^
IZ/
"iw
¦ M
1*
¦T
IK
Vent
Well #7
¦R
1Q
¦ Y
Figure 6. Proposed Pilot Test Design.
Figure 7. Oxygen concentration In vadose zone before venting.
4
-------
Distance (fleet)
0 10 2030405060709090
Figure 8. Oxygen concentration In vadose zone after venting.
if biodegradation is being limited by insufficient moisture
content, toxicity (e.g. metals), or nutrients.
3. Subsurface air flow characteristics. Observation wells
which indicate persistent, low Oz levels may indicate an
insufficient supply of oxygenated air at that location
suggesting the need for air injection, higher extraction
well vacuum, additional extraction wells, or additional
soils characterization which may indicate high moisture
content or the presence of immiscible fluids impeding the
flow of air.
Location and Number of Vapor Extraction Wells
One of the primary objectives in conducting a venting field
test is to evaluate the initial placement of extraction wells to
optimize VOC removal from soil. Placement of extraction
wells and selected applied vacuum is largely an iterative
process requiring continual re-evaluation as additional data
are collected during remediation. Vacuum extraction wells
produce complex three-dimensional reduced pressure zones
in affected soils. The size and configuration of this affected
volume depends on the applied vacuum, venting geometry
(e.g., depth to water table), soil heterogeneity, and intrinsic
(e.g., permeability) and dynamic (e.g., moisture content)
properties of the soil. The lateral extent of this reduced
pressure zone (beyond which static vacuum is no longer
detected) is often termed the radius or zone of influence
(ROI). Highly permeable sandy soils typically exhibit large
zones of influence and high air flow rates whereas less
permeable soils, such as silts and clays, exhibit smaller zones
of influence and low air flows.
Measured or anticipated radii of influence are often used to
space extraction wells. For instance, if a ROI is measured at
10 feet, extraction wells are placed 20 feet apart. However,
this strategy is questionable since vacuum propagation and
air velocity decrease substantially with distance from an
extraction well. Thus, only a limited volume of soil near an
extraction well will be effectively ventilated regardless of the
ROI. Johnson (J.J., 1988) describes how the addition of 13
extraction wells within the ROI of other extraction wells
increased blower VOC concentration by 4000 ppmv and mass
removal by 40 kg/day. They concluded that the radius of
influence was not an effective parameter for locating
extraction wells and that operation costs could be reduced by
increasing the number of extraction wells as opposed to
pumping at higher rates with fewer wells.
Determining the propagation of induced vacuum requires
conducting pneumatic pump tests in which variation in static
vacuum is measured in vapor observation wells at depth and
distance from extraction wells. Locating extraction and
observation wells along transects as illustrated in Figure 4
minimizes the number of observation wells necessary to
evaluate vacuum propagation at linear distances from
extraction wells. Pressure differential can be observed at
greater distances than would otherwise be possible in other
configurations.
Propagation of vacuum in soils as a function of applied
vacuum can be determined by conducting pneumatic pump
tests with incrementally increasing flow or applied vacuum.
Vacuum is increased after steady state conditions (relatively
constant static vacuum measurements in observation wells)
exist in soils from the previously applied vacuum. A step pump
test will indicate a significant increase in static vacuum or air
velocity with increasing applied vacuum near an extraction
well. However, at distance from an extraction well, a
significant increase in static vacuum will not be observed with
an increase in applied vacuum. Pneumatic pump tests allow
determination of radial distances from extraction wells in which
air velocity is sufficient to ensure remediation.
After the initial placement of extraction wells has been
established based on the physics of air flow, an initial applied
vacuum must be selected to ensure optimal VOC removed. In
regard to mass transfer considerations, the vent rate should
be increased if a significant corresponding mass flux is
observed. Even though an increased venting rate may not
substantially increase the propagation of vacuum with
distance, air velocity will increase near the extraction well. If
most contaminants are in more permeable deposits, an
increase in applied vacuum will increase mass removal
eventually to a point of diminishing returns or until the system
is limited by diffusion. Note that this strategy is for
optimization of volatilization not biodegradation. Optimizing in-
srtu biodegradation often necessitates reducing air velocity in
soil. As a result, vapor treatment costs are minimized but
overall mass flux decreases. Thus, in-situ biodegradation of
VOCs minimizes overall costs but may extend venting
operation time.
During a field test, it is desirable to operate until mass
transport limitations are realized in order to evaluate the long
term performance of the technology. This can be achieved by
isolating small selected areas of a site by the use of passive
air inlet wells. When attempting to evaluate diffusion limited
mass removal in isolated areas, applied vacuum should
remain high and the distance between passive inlet and
extraction wells should be minimized. Too often, venting field
tests are conducted for relatively short periods of time (e.g.,
2 - 21 days) which only results in assessment of air
permeability and initial mass removal. Longer field studies
(e.g., 6 months -12 months) enable better insight into mass
5
-------
transfer limitations which eventually govern venting
effectiveness.
Screened Interval
The screened interval of extraction wells will play a significant
role in directing air flow through contaminated soils. Minimum
depths are recommended by some practitioners for venting
operation to avoid short-circuiting of air flow. However, the
application of venting need not be limited by depth to water
table since horizontal vents can be used in lieu of vertically
screened extraction wells to remediate soils with shallow
contamination. Often, it is desirable to dewater contaminated
shallow aquifer sediments for venting application. For
remediation of more permeable soils with deep contamination,
an extraction well should be screened at the maximum depth
of contamination or to the seasonal low water table,
whichever is shallowest, to direct airflow and reduce short-
circuiting. For less permeable soils, or for more continuous
vertical contamination, a higher and longer screened interval
may be useful. In stratified systems, such as in the presence
of clay layers between more permeable deposits, more than
one well will be required, each venting a distinct strata.
Screening an extraction well over two strata of significantly
different permeability will result in most air flow being directed
only in the strata of greater permeability. It is important to
screen extraction wells over the interval of highest soil
contamination to avoid extracting higher volumes of air at
lower vapor concentration.
During venting, the reduced pressure in the soil will cause an
upwelling of the water table. The change in water table
elevation can be determined from the predicted radial
pressure distribution. Johnson et al. (1988) indicated that
upwelling can be significant under typical venting conditions.
Water table rise will cause contaminated soil lying above the
water table to become saturated, resulting in decreased mass
removal rates. Ground water upwelling due to venting system
operation can be minimized with concurrent water table
dewatering.
Placement of Observation Wells
Observation wells are essential in determining whether
contaminated soils are being effectively ventilated and in the
evaluation of interactions among extraction wells. The more
homogeneous and isotropic the unsaturated medium, the
fewer the number of vapor monitoring probes required. To
adequately describe vacuum propagation during a field test,
usually at least three observation well clusters are needed
within the ROI of an extraction well. At least one of these
clusters should be placed near an extraction well because of
the logarithmic decrease in vacuum with distance. The depth
and number of vapor probes within a cluster depends on the
screened intervals of extraction wells and soil stratigraphy.
However, vertical placement of vapor probes might logically
be near the soil-water table interface, soil horizon interfaces,
and near the soil surface. As previous mentioned, the use of
air flow modeling can assist in optimizing the depth and
placement of vapor observation wells and in the interpretation
of data collected from these monitoring points.
When constructing observation wells K is desirable to minimize
vapor storage volume in the screened interval and sample
transfer line. This will minimize purging volumes and ensure a
representative vapor sample in the vicinity of each observation
well. Analysis of soil gas in an on-site field laboratory is
preferred to provide real time data for implementation of
engineering controls and process modifications. It is
recommended that steel canisters, sorbent tubes, or direct GC
injection be used in lieu of Tedlar bags when possible
because of potential VOC loss through bag leakage or
diffusion within the teflon material itself. This problem may
lead to erroneous analytical results and the potential of a false
negative indication of soil remediation at low soil gas
concentrations.
Summary/Conclusions
While the application of soil vacuum extraction is conceptually
simple, its success depends on understanding complex
subsurface physical, chemical, and biological processes
which provide insight into factors limiting venting performance.
Optimizing venting performance is critical when attempting to
meet stipulated soil-based clean-up levels required by
regulators. The first step in evaluating a venting application is
to assess contaminant volatility. Volatility is a function of a
contaminant's soil-water partition coefficient and Henry's
constant if present in a three-phase system, and a
contaminant's vapor pressure and mole fraction in an
immiscible fluid, if present in a four phase system. Volatility is
greatly decreased when soils are extremely dry. As vacuum
extraction proceeds, lower molecular weight organic
compounds preferentially volatilize and biodegrade.
Decreasing mole fractions of lighter compounds and
increasing mole fractions of heavier compounds affect
observed offgas concentrations. Understanding contaminant
volatility is necessary when attempting to utilize offgas vapor
concentrations as an indication of venting progress.
The significance of mass transport limitations should be
evaluated during venting field tests. Long term performance of
venting will most likely be limited by diffusion from soil regions
of lesser permeability which are not exposed to direct airflow.
Mass transport limitations can be assessed by isolating a
small area of a site and aggressively applying vacuum
extraction. Simplistic methods to evaluate remediation time
should be avoided. One of the most promising uses of vacuum
extraction is in manipulating subsurface oxygen levels to
enhance biodegradation. When conducting field studies, it is
recommended that CO. and Oz levels be monitored in vapor
probes to evaluate the feasibility of VOC and semivolatile
contaminant biodegradation.
Air permeability in soil is a function of a soil's intrinsic
permeability and liquid content. Relative permeability of air
can be estimated using relationships developed by Brooks
and Corey (1964) and Van Genuchten (1980). The most
effective method of measuring air permeability is by
conducting pneumatic pump tests. Information gained from
pneumatic pump tests can be used to determine site-specific
design considerations such as the spacing of extraction wells.
Measured or anticipated zones of influence are not particularly
useful in spacing extraction wells. Extraction wells should be
located to maximize air velocity in contaminated soils.
Pneumatic pump tests with increasing applied vacuum may be
useful in determining radial distances from extraction wells in
which air velocity is sufficient to ensure remediation.
6
-------
Screened intervals should be located at or below the depth of
contamination. In stratified soils, more than one well is
necessary to ventilate each strata. At least three observation
well clusters are usually necessary to observe vacuum
propagation within the radius of influence of an extraction well.
Logical vertical placement of vapor probes might be near the
soil-water table interface, soil horizon interfaces, and near the
soil surface.
References
(1) Brooks, R.H., and Corey, A.T., 1964. Hydraulic
Properties of Porous Media, Colorado State University,
Fort Collins, CO., Hydrol. Pap. No. 3, 27 pp.
(2) Hinchee, R.E., 1989. Enhanced Biodegradation through
Soil Venting, Proceedings of the Workshop on Soil
Vacuum Extraction, Robert S. Kerr Environmental
Research Laboratory, Ada, Oklahoma, April 27-28,1989.
(3) Johnson, J.J., 1988. In Situ Air Stripping: Analysis of
Data from a Project Near Benson, Arizona, Master of
Science Thesis, Colorado School of Mines, Colorado.
(3) Johnson, P.C., Kemblowski, M.W., and Colthart, J.D.,
1988. Practical Screening Models for Soil Venting
Applications, NWWA/API Conference on Petroleum
Hydrocarbons and Organic Chemicals in Groundwater,
Houston, TX, 1988.
(4) Johnson, R.L, 1989. Soil Vacuum Extraction: Laboratory
and Physical Model Studies, Proceedings of the
Workshop on Soil Vacuum Extraction, Robert S. Kerr
Environmental Research Laboratory, Ada, Oklahoma,
April 27-28,1989.
(5) Rawls, W.J., Brakensiek, D.L, and Saxton, K.E., 1982.
Estimation of Soil Water Properties, Transactions of the
ASAE, 1982, pp. 1316-1328.
(6) Siegrist, R. L., and Jenssen, P. C„ 1990. Evaluation of
Sampling Method Effects on Volatile Organic Compound
Measurements in Contaminated Soils, Environ. Sci.
Technol., Vol. 24, No. 9, p. 1387-1392.
(7) Van Genuchten, M.T., 1980. A Closed-Form Equation for
Predicting the Hydraulic Conductivity of Unsaturated
Soils, Soil Sci. Soc. Am. J., 44:982-898.
7
All.S. GOVERNMENT PRINTING OFFICE: l**I • MS-CS0/4O260
-------
United States
Environmental Protection
Agency
Center for Environmental
Research Information
Cincinnati, OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use $300
EPA/540/S-92/004
-------
United States Office of Office of Solid Waste EPA/540/4-91-002
Environmental Protection Research and and Emergency March 1991
Agency Development Response
&EPA Ground Water Issue
DENSE NONAQUEOUS PHASE LIQUIDS
Scott G. Huling* and James W. Weaver**
Background
The Regional Superfund Ground Water Forum is a group of
EPA professionals representing EPA's Regional Superfund
Offices, committed to the identification and the resolution of
ground water issues impacting the remediation of Superfund
sites. The Forum is supported by and advises the Superfund
Technical Support Project. Dense nonaqueous phase liquids is
an issue identified by the Forum as a concern of Superfund
decision-makers. For further information contact Scott G.
Huling (FTS:743-2313), Jim Weaver (FTS:743-2420), or
Randall R. Ross (FTS: 743-2355).
Introduction
Dense nonaqueous phase liquids (DNAPLs) are present at
numerous hazardous waste sites and are suspected to exist at
many more. Due to the numerous variables influencing DNAPL
transport and fate in the subsurface, and consequently, the
ensuing complexity, DNAPLs are largely undetected and vet
are likely to be a significant limiting factor in site remediation.
This issue paper is a literature evaluation focusing on DNAPLs
and provides an overview from a conceptual fate and transport
point of view of DNAPL phase distribution, monitoring, site
characterization, remediation, and modeling.
A nonaqueous phase liquid (NAPL) is a term used to describe
the physical and chemical differences between a hydrocarbon
liquid and water which result in a physical interlace between a
mixture of the two liquids. The interface is a physical dividing
surface between the bulk phases of the two liquids, but
compounds found in the NAPL are not prevented from
solubilizing into the ground water. Immiscibilhy is typically
determined based on the visual observation of a physical
interface in a water- hydrocarbon mixture. There are numerous
methods, however, which are used to quantify the physical and
chemical properties of hydrocarbon liquids (31).
Nonaqueous phase liquids have typically been divided into two
general categories, dense and light. These terms describe the
specific gravity, or the weight of the nonaqueous phase liquid
relative to water. Correspondingly, the dense nonaqueous
phase liquids have a specific gravity greater than water, and
the light nonaqueous phase liquids (LNAPL) have a specific
gravity less than water.
Several of the most common compounds associated with
DNAPLs found at Superfund sites are included in Table 1.
These compounds are a partial list of a larger list identified by a
national screening of the most prevalent compounds found at
Superfund sites (65). The general chemical categories are
halogenated/non-halogenated semi-volatiles and halogenated
volatiles. These compounds are typically found in the following
wastes and waste-producing processes: solvents, wood
preserving wastes (creosote, pentachlorophenol), coal tars,
and pesticides. The most frequently cited group of these
contaminants to date are the chlorinated solvents.
DNAPL Transport and Fate - Conceptual Approach
Fate and transport of DNAPLs in the subsurface will be
presented from a conceptual point of view. Figures have been
selected for various spill scenarios which illustrate the general
behavior of DNAPL in the subsurface. Following the
conceptual approach, detailed information will be presented
explaining the specific mechanisms, processes, and variables
which influence DNAPL fate and transport. This includes
DNAPL characteristics, subsurface media characteristics, and
saturation dependent parameters.
Unsaturated Zone
Figure 1 indicates the general scenario of a release of DNAPL
into the soil which subsequently migrates vertically under both
the forces of gravity and soil capillarity. Soil capillarity is also
responsible for the lateral migration of DNAPL. A point is
reached at which the DNAPL no longer holds together as a
continuous phase, but rather is present as isolated residual
globules. The fraction of the hydrocarbon that is retained by
capillary forces in the porous media is referred to as residual
Environmental Engineer," Research Hydrologist, U.S.
Environmental Protection Agency, Robert S. Kerr Environmental
Research Laboratory, Ada, Oklahoma.
/ \
^ "T" %
8 XsefHotooy ft
|upport
Superfund Technology Support Center for
Ground Water
Robert S. Kerr Environmental
Research Laboratory
Ada, Oklahoma
Technology lonovattort Office :
Office of Solid Waste and Emergency
Response, OS EPA, Washington, D.C.
Water W„ Kovalick, Jr.(Ph.D.
Director
Prinied on Recycled Papei
-------
Table 1. Most prevalent chemical compounds at U.S. Superfund Sites (65) with a specific gravity
greater than one.
Density Dynamic[2] Kinematic Water[4] Henry's Law Vapor[6]
Compound [1] Viscosity Viscosity[3] Solub. Constant[5] Pressure
Halogenated Semi-volatlles
1,4-Dichlorobenzene
1.2475
1.2580
1.008
8. 0 E+01
1.58 E-03
6 E-01
1,2-Dichlorobenzene
1.3060
1.3020
0.997
1.0 E+02
1.88 E-03
9.6 E-01
Aroclor 1242
1.3850
4.5 E-01
3.4 E-04
4.06 E-04
Aroclor 1260
1.4400
2.7 E-03
3.4 E-04
4.05 E-05
Aroclor 1254
1.5380
1.2 E-02
2.8 E-04
7.71 E-05
Chlordane
1.6
1.1040
0.69
5.6 E-02
2.2 E-04
1 E-05
Dieldrin
1.7500
1.86 E-01
9.7 E-06
1.78 E-07
2,3,4,6-Tetrachlorophenol
1.8390
1.0 E+03
Pentachlorophenol
1.9780
1.4 E+01
2.8 E-06
1.1 E-04
Halogenated Volatiles
Chlorobenzene
1.1060
0.7560
0.683
4.9 E+02
3.46 E-03
8.8 E+00
1,2-Dichloropropane
1.1580
0.8400
. 0.72
2.7 E+03 .
3.6 E-03
3.95 E+01
1,1-Dichloroethane
1.1750
0.3770
0.321
5.5 E+03
5.45 E-04
1.82 E+02
1,1-Dichloroethylene
1.2140
0.3300
0.27
4.0 E+02
1.49 E-03
5 E+02
1,2-Dichloroethane
1.2530
0.8400
0.67
8.69 E+03
1.1 E-03
6.37 E+01
Trans-1,2-Dichloroethylene
1.2570
0.4040
0.321
6.3 E+03
5.32 E-03
2.65 E+02
Cis-1,2-Dichloroethylene
1.2480
0.4670
0.364
3.5 E+03
7.5 E-03
2 E+02
1,1,1-Trichloroethane
1.3250
0.8580
0.647
9.5 E+02
4.08 E-03
1 E+02
Methylene Chloride
1.3250
0.4300
0.324
1.32 E+04
2.57 E-03
3.5 E+02
1,1,2-Trichtoroethane
1.4436
0.1190
0.824
4.5 E+03
1.17 E-03
1.88 E+01
Trichloroethylene
1.4620
0.5700
0.390
1.0 E+03
8.92 E-03
5.87 E+01
Chloroform
1.4850
0.5630
0.379
8.22 E+03
3.75 E-03
1.6 E+02
Carbon Tetrachloride
1.5947
0.9650
0.605
8.0 E+02
2.0 E-02
9.13 E+01
1,1,2,2-Tetrachloroethane
1.6
1.7700
1.10
2.9 E+03
5.0 E-04
4.9 E+00
Tetrachloroethylene
1.6250
0.8900
0.54
1.5 E+02
2.27 E-02
1.4 E+01
Ethylene Dibromide
2.1720
1.6760
0.79
3.4 E+03
3.18 E-04
1.1 E+01
Non-halogenated Semi-volatiies
2-Methyl Napthalene
1.0058
2.54 E+01
5.06 E-02
6.80 E-02
o-Cresol
1.0273
3.1 E+04
4.7 E-05
2.45 E-01
p-Cresol
1.0347
2.4 E+04
3.5 E-04
1.08 E-01
2,4-Dimethylphenol
1.0360
6.2 E+03
2.5 E-06
9.8 E-02
m-Cresol
1.0380
21. 0
20
2.35 E+04
3.8 E-05
1.53 E-01
Phenol
1.0576
3.87
8.4 E+04
7.8 E-07
5.293E-01
Naphthalene
1.1620
3.1 E+01
1.27 E-03
2.336E-01
Benzo(a)Anthracene
1.1740
1.4 E-02
4.5 E-06
1.16 E-09
Flourene
1.2030
1.9 E+00
7.65 E-05
6.67 E-04
Acenaphthene
1.2250
3.88 E+00
1.2 E-03
2.31 E-02
Anthracene
1.2500
7.5 E-02
3.38 E-05
1.08 E-05
Dibenz(a,h)Anthracene
1.2520
2.5 E-03
7.33 E-08
1 E-10
Fluoranthene
1.2520
2.65 E-01
6.5 E-06
E-02 E-06
Pyrene
1.2710
1.48 E-01
1.2 E-05
6.67 E-06
Chrysene
1.2740
6.0 E-03
1.05 E-06
6.3 E-09
2,4-Dinilrophenol
1.6800
6. 0 E+03
6.45 E-10
1.49 E-05
Miscellaneous
Coal Tar
1.028(7)
18.98<7>
Creosote
1.05
1.08<8>
|1] g/cc
[2] centipoise (cp), water has a dynamic viscosity of 1 cp at 20°C
[3] centistokes (cs)
[4] mg/l
[5] atm-m3/mol
[6] mm Hg
[7] 45" F (70)
[8] 15.5°C, vanes with creosote mix (62)
2
-------
Residual Saturation of
DNAPL in Soil From Spill
Infiltration and
4 "TTi ^ Leaching
Tf T,
Groundwater
Flow
After. Waterloo Centre for Groundwater Research, 1969
Residual
Saturation o(
DNAPL In
Vadoee Zone
Infiltration, Leaching and
Mobile DNAPL Vapors
Plume From DNAPL
Soil Vapor
Groundwater
Plume From DNAPL Flow
Re6ldual Saturation
Alter. Waterloo Cente tor Groundwater Reeeer*. 1989
Figure 1. The entire volume of DNAPL is exhausted by residual
saturation in the vadose zone prior to DNAPL reaching
the water table. Soluble phase compounds may be
leached from the DNAPL residual saturation and
contaminate the ground water.
saturation. In this spill scenario, the residual saturation in the
unsaturated zone exhausted the volume of DNAPL, preventing
it from reaching the water table. This figure also shows the
subsequent leaching (solubilization) of the DNAPL residual
saturation by water percolating through the unsaturated zone
(vadose zone). The leachate reaching the saturated zone
results in ground-water contamination by the soluble phase
components of the hydrocarbon. Additionally, the residual
saturation at or near the water table is also subjected to
leaching from the rise and fall of the water table (seasonal, sea
level, etc.).
Increasing information is drawing attention to the importance of
the possibility that gaseous-phase vapors from NAPL in the
unsaturated zone are responsible for contaminating the ground
water and soil (18,47). It is reported that the greater "relative
vapor density" of gaseous vapors to air will be affected by
gravity and will tend to sink. In subsurface systems where
lateral spreading is not restricted, spreading of the vapors may
occur as indicated in Figure 2. The result is that a greater
amount of soils and ground water will be exposed to the
DNAPL vapors and may result in further contamination. The
extent of contamination will depend largely on the partitioning
of the DNAPL vapor phase between the aqueous and solid
phases.
DNAPL Phase Distribution - Four Phase System
It is apparent from Figures 1 and 2 that the DNAPL may be
present in the subsurface in various physical states or what is
referred to as phases. As illustrated in Figure 3, there are four
possible phases: gaseous, solid, water, and immiscible
hydrocarbon (DNAPL) in the unsaturated zone. Contaminants
associated with the release of DNAPL can, therefore, occur in
four phases described as follows:
1. Air phase - contaminants may be present as vapors;
2. Solid phase - contaminants may adsorb or partition onto
the soil or aquifer material;
3. Water phase - contaminants may dissolve into the water
according to their solubility; and
Figure 2. Migration of DNAPL vapors from the spill area and
subsequent contamination of the soils and ground
water.
4. Immiscible phase - contaminants may be present as
dense nonaqueous phase liquids.
The four phase system is the most complex scenario because
there are four phases and the contaminant can partition
between any one or all four of these phases, as illustrated in
Figure 4. For example, TCE introduced into the subsurface as
a DNAPL may partition onto the soil phase, volatilize into the
soil gas, and solubilize into the water phase resulting in
contamination in all four phases. TCE can also partition
between the water and soil, water and air, and between the soil
and air. There are six pathways of phase distribution in the
unsaturated zone. The distribution of a contaminant between
these phases can be represented by empirical relationships
referred to as partition coefficients. The partition coefficients, or
the distribution of the DNAPL between the four phases, is
highly site-specific and highly dependent on the characteristics
of both the soil/aquifer matrix and the DNAPL. Therefore, the
distribution between phases may change with time and/or
location at the same site and during different stages of site
remediation.
Solid
Water
Air \
- DNAPL
Figure 3. A DNAPL contaminated unsaturated zone has four
physical states or phases (air, solid, water, Immiscible).
The contaminant may be present In any one, or all four
phases.
3
-------
Four Phase System
Partition Coefficients
K = Soil-water partition coefficient
Kh - Henry's Constant
K' c DNAPL-water partition coefficient
K" = DNAPL-air partition coefficient
DNAPL y
Water v == Soil
K
• lur DiGiiilo. IMO (fl)
Figure 4. Distribution of DNAPL between the four phases found
in the vadose zone.
The concept of phase distribution is critical in decision-
making. Understanding the phase distribution of a DNAPL
introduced into the subsurface provides significant insight in
determining which tools are viable options with respect to site
characterization and remediation.
DNAPL represented by residual saturation in the four phase
diagram is largely immobile under the usual subsurface
pressure conditions and can migrate further only: 1) in water
according to its solubility; or 2) in the gas phase of the
unsaturated zone (47). DNAPL components adsorbed onlo the
soil are also considered immobile. The mobile phases are,
therefore, the soluble and volatile components of the DNAPL
in the water and air, respectively.
The pore space in the unsaturated zone may be filled with one
or all three fluid phases (gaseous, aqueous, immiscible). The
presence of DNAPL as a continuous immiscible phase has the
potential to be mobile. The mobility of DNAPL in the
subsurface must be evaluated on a case by case basis. The
maximum number of potentially mobile fluid phases is three.
Simultaneous flow of the three phases (air, water, and
immiscible) is considerably more complicated than two-phase
flow (46). The mobility of three phase flow in a four-phase
system is complex, poorly understood, and is beyond the
scope of this DNAPL overview. The relative mobility of the two
phases, water and DNAPL, in a three-phase system is
presented below in the section entitled "Relative Permeability."
Generally, rock aquifers contain a myriad of cracks (fractures)
of various lengths, widths, and apertures (32). Fractured rock
systems have been described as rock blocks bounded by
discrete discontinuities comprised of fractures, joints, and
shear zones which may be open, mineral-filled, deformed, or
any combination thereof (61). The unsaturated zone overlying
these fractured rock systems also contain the myriad of
preferential pathways. DNAPL introduced into such formations
(Figure 5) follow complex pathways due to the heterogeneous
distribution of the cracks, conduits, and fractures', i.e.,
preferential pathways. Transport of DNAPL may follow non-
Darcian flow in the open fractures and/or Darcian flow in the
porous media filled fractures. Relatively small volumes of
NAPL may move deep, quickly into the rock because the
Figure 5. DNAPL spilled into fractured rock systems may
follow a complex distribution of the preferential pathways.
retention capacity offered by the dead-end fractures and the
immobile fragments and globules in the larger fractures is so
small (32). Currently, the capability to collect the detailed
information for a complete description of a contaminated
fractured rock system is regarded as neither technically
possible nor economically leasible (61).
Low permeability stratigraphic units such as high clay content
formations may also contain a heterogeneous distribution of
preferential pathways. As illustrated in Figure 6, DNAPL
transport in these preferential pathways is correspondingly
complex. Typically, it is assumed that high clay content
formations are impervious to DNAPL. However, as DNAPL
spreads out on low permeable formations it tends to seek out
zones of higher permeability. As a result, preferential pathways
allow the DNAPL to migrate further into the low permeable
formation, or through it to underlying stratigraphic units. It is
apparent from Figures 5 and 6 that the complexity of DNAPL
transport may be significant prior to reaching the water table.
Saturated Zone
The second general scenario is one in which the volume of
DNAPL is sufficient to overcome the fraction depleted by the
residual saturation in the vadose zone, as illustrated in Figure
7. Consequently, the DNAPL reaches the water table and
contaminates the ground water directly. The specific gravity of
DNAPL is greater than water, therefore, the DNAPL migrates
into the saturated zone. In this scenario, DNAPL continues the
vertical migration through the saturated zone until the volume
is eventually exhausted by the residual saturation process or
until it is intercepted by a low permeable formation where it
begins to migrate laterally.
DNAPL Phase Distribution - Three Phase System
Due to the lack of the gaseous phase, the saturated zone
containing DNAPL is considered a three-phase system
consisting of the solid, water, and immiscible hydrocarbon
(Figure 8). Contaminant distribution in the three-phase system
is less complex than the four-phase system. Again, this is
highly dependent on the characteristics of both the aquifer
4
-------
Figure 6. DNAPL spilled Into a low permeable formation may
follow a complex distribution of preferential pathways.
The volume of DNAPL is exhausted in the vadose zone
prior to reaching the water table.
matrix and the DNAPL Figure 9 indicates the three phases
and the transfer of the mass of contaminant between the
phases. In this scenario, there are only three pathways of
phase distribution in the saturated zone.
Note that when the DNAPL is represented by residual
saturation in the three-phase system, the mobile phase of the
contaminant is the water soluble components of the DNAPL
and the immobile phases are the residual saturation and the
adsorbed components of the DNAPL associated with the
aquifer material. The main mobilization mechanism of the
residual saturation is removal of soluble phase components
into the ground water. When the DNAPL is present as a
continuous immiscible phase, it too is considered one of the
mobile phases of the contaminant. While the continuous phase
DNAPL has the potential to be mobile, immobile continuous
phase DNAPL may also exist in the subsurface. Although the
saturated zone is considered a three-phase system, gaseous
vapors from DNAPL in the unsaturated zone does have the
Figure 8. A DNAPL contaminated saturated zone has three
phases (solid, water, Immiscible). The contaminant
may be present in any one, or all three phases.
potential to affect ground-water quality, as was indicated earlier
in Figure 2.
Assuming the residual saturation in the saturated zone does
not deplete the entire volume of the DNAPL, the DNAPL will
continue migrating vertically until it encounters a zone or
stratigraphic unit of lower permeability. Upon reaching the zone
of lower permeability, the DNAPL will begin to migrate laterally.
The hydraulic conductivity in the vertical direction is typically
less than in the horizontal direction. K is not uncommon to find
vertical conductivity that is one-fifth or one-tenth the horizontal
value (4). It is expected that DNAPL spilled into the subsurface
will have a significant potential to migrate laterally. If the lower
permeable boundary is "bowl shaped", the DNAPL will pond as
a reservoir (refer to Figure 10). As illustrated in Figure 11, it is
not uncommon to observe a perched DNAPL reservoir where a
discontinuous impermeable layer; i.e., silt or clay lens,
intercepts the vertical migration of DNAPL. When a sufficient
volume of DNAPL has been released and multiple
discontinuous impermeable layers exist, the DNAPL may be
present in several perched reservoirs as well as a deep
Residual
Saturation of
DNAPL in Soil
From Spill
— Infiltration and
Leaching
Groundwater
Flow
Residual
Saturation in Saturated Zone
After. Waterloo Cenve for Groundwater Research, 1M9
Figure 7. The volume of DNAPL Is sufficient to overcome the
residual saturation in the vadose zone and
consequently penetrates the water table.
Three Phase System
Water
K' = DNAPL-water partition coefficient
K = Soil-water partition coefficient
Figure 9. Distribution of DNAPL between the three phases found
in the saturated zone.
5
-------
&«oh*d
Contammviti
>>>yf
s / / / /
~777lr/ / / / ,y-y-jr
Low Pwrn—bl»
Unit
RtuduaJ
SMjrnon > DNAPL Pool
Grountfwatar
Flow
SANO
AOUFER
yTTTJJWw \fv777777777TTTTT77/
AIMr. Wantloo Cm lor GraundNtnr RMurctl. 1M9
Figure 12. Perched and deep DNAPL reservoirs.
Figure 10. Migration of DNAPL through the vadose zone to an
impermeable boundary.
reservoir (refer to Figure 12). Lateral migration continues until
either the residual saturation depletes the DNAPL or an
impermeable depression immobilizes the DNAPL in a reservoir
type scenario. Soluble-phase components of the DNAPL will
partition into the ground water from both the residual saturation
or DNAPL pools. The migration of DNAPL vertically through
the aquifer results in the release of soluble-phase components
of the DNAPL across the entire thickness of the aquifer. Note,
that ground water becomes contaminated as it flows through,
and around, the DNAPL contaminated zone.
As indicated earlier, DNAPL will migrate laterally upon
reaching a stratigraphic unit of lower permeability. Transport of
DNAPL will therefore be largely dependent on the gradient of
the stratigraphy. Occasionally, the directional gradient of an
impermeable stratigraphic unit may be different than the
direction of ground-water flow as illustrated in Figure 13a. This
may result in the migration of the continuous phase DNAPL in
a direction different from the ground-water flow. Nonhorizontal
stratigraphic units with varying hydraulic conductivity may also
convey DNAPL in a different direction than ground-water flow,
and at different rates (refer to Figure 13b). Determination of the
direction of impermeable stratigraphic units will therefore
provide useful information concerning the direction of DNAPL
transport.
Figure 11. Perched DNAPL reservoir.
Similar to the unsaturated zone, the saturated zone also
contains a complex distribution of preferential pathways from
cracks, fractures, joints, etc. DNAPL introduced into such
formations correspondingly follow the complex network of
pathways through an otherwise relatively impermeable rock
material. Other pathways which may behave as vertical
conduits for DNAPL include root holes, stratigraphic windows,
disposal wells, unsealed geotechnical boreholes, improperly
sealed hydrogeological investigation sampling holes and
monitoring wells, and old uncased/unsealed water supply wells
(72). Transport of the DNAPL may migrate very rapidly in these
open conduits or follow Darcian flow in the surrounding porous
media or porous media filled fractures. A relatively small
volume of DNAPL can move deep into a fractured system due
to the low retentive capacity of the fractured system.
Consequently, fractured clay or rock stratigraphic units, which
are often considered lower DNAPL boundary conditions, may
have preferential pathways leading to lower formations, as
depicted in Figure 14. Careful inspection of soil cores at one
Superfund site indicated that DNAPL flow mainly occurred
through preferential pathways and was not uniformly
distributed throughout the soil mass (8). Due to the complex
Figure 13a. Stratigraphic gradient different from ground water
gradient results in a different direction of flow of the
ground water and continuous phase DNAPL.
6
-------
Where Ko> Kxs
Kx = Horizontal Hydraulic Conductivity
Groundwater
Flow
Figure 13b. Non-horizontal stratigraphlc units with variable
hydraulic conductivity may convey DNAPL in a
different direction than the ground water flow
direction.
distribution of preferential pathways, characterization of the
volume distribution of the DNAPL is difficult.
Important DNAPL Transport and Fate Parameters
There are several characteristics associated with both the
subsurface media and the DNAPL which largely determine the
fate and transport of the DNAPL. A brief discussion of these
parameters is included to help identify the specific details of
DNAPL transport mechanisms. Several of the distinctive
DNAPL phenomena observed on the field-scale relates back to
phenomena at the pore-scale. Therefore, it is important to
understand the principles from the pore-scale level to develop
an understanding of field-scale observations, which is the scale
at which much of the Supeiiund work occurs. A more
complete and comprehensive review of these parameters is
available (2,36,71).
DNAPL Characteristics
Density
Fluid density is defined as the mass of fluid per unit volume,
i.e. g/cm3. Density of an immiscible hydrocarbon fluid is the
parameter which delineates LNAPL's from DNAPL's. The
property varies not only with molecular weight but also
molecular interaction and structure. In general, the density
varies with temperature and pressure (2). Equivalent methods
of expressing density are specific weight and specific gravity.
The specific weight is defined as the weight of fluid per unit
volume, i.e. lb/ft3. The specific gravity (S.G.) or the relative
density of a fluid is defined as the ratio of the weight of a given
volume of substance at a specified temperature to the weight
of the same volume of water at a given temperature (31). The
S.G. is a relative indicator which ultimately determines whether
the fluid will float (S.G.< 1.0) on, or penetrate into (S.G.>1.0)
the water table. Table 1 contains a list of compounds with a
density greater than one that are considered DNAPL's. Note,
however, that while the specific gravity of pentachlorophenol
and the non-halogenated semi-volatiles is greater than 1.00,
these compounds are a solid at room temperature and would
not be expected to be found as an immiscible phase liquid at
wood preserving sites but are commonly found as contami-
nants. Pentachlorophenol is commonly used as a wood
preservant and is typically dissolved (4-7%) in No. 2 or 3 fuel
oil.
Viscosity
The viscosity of a fluid is a measure of its resistance to flow.
Molecular cohesion is the main cause of viscosity. As the
temperature increases in a liquid, the cohesive forces
decrease and the absolute viscosity decreases. The lower the
viscosity, the more readily a fluid will penetrate a porous
media. The hydraulic conductivity of porous media is a function
of both the density and viscosity of the fluid as indicated in
equation [1]. It is apparent from this equation that fluids with
either a viscosity less than water or fluids with a density greater
than water have the potential to be more mobile in the
subsurface, than water.
Figure 14. DNAPL transport In fracture and porous media
stratigraphlc units.
K = k p g where, K = hydraulic conductivity [1]
H k = intrinsic permeability
p ¦= fluid mass density
g = gravity
H = dynamic (absolute) viscosity
Results from laboratory experiments indicated that several
chlorinated hydrocarbons which have tow viscosity (methylene
chloride, perchloroethylene, 1,1,1-TCA, TCE) will infiltrate into
soil notably faster than will water (47). The relative value of
NAPL viscosity and density, to water, indicates how fast it will
flow in porous media (100% saturated) with respect to water.
For example, several low viscosity chlorinated hydrocarbons
(TCE, tetrachloroethylene, 1,1,1-TCA, Methylene Chloride,
Chloroform, Carbon Tetrachloride, refer to Table 1) will flow
1.5-3.0 times as fast as water and higher viscosity compounds
including light heating oil, diesel fuel, jet fuel, and crude oil (i.e.
LNAPL's) will flow 2-10 times slower than water (45). Both coal
tar and creosote typically have a specific gravity greater than
one and a viscosity greater than water. It is interesting to note
7
-------
that the viscosity of NAPL may change with time (36). As fresh
crude oils lose the lighter volatile components from
evaporation, the oils become more viscous as the heavier
components compose a larger fraction of the oily mixture
resulting in an increase in viscosity.
Solubility
When an organic chemical is in physical contact with water, the
organic chemical will partition into the aqueous phase. The
equilibrium concentration of the organic chemical in the
aqueous phase is referred to as its solubility. Table 1 presents
the solubility of several of the most commonly found DNAPL's
at EPA Superfund Sites. The solubility of organic compounds
varies considerably from the infinitely miscible compounds,
including alcohols (ethanol, methanol) to extremely low
solubility compounds such as polynuclear aromatic
compounds.
Numerous variables influence the solubility of organic
compounds. The pH may affect the solubility of some organic
compounds. Organic acids may be expected to increase in
solubility with increasing pH, while organic bases may act in
the opposite way (31). For example, pentachlorophenol is an
acid which is ionized at higher pH's. In the ionized form,
pentachlorophenol would be more soluble in water (59).
Solubility in water is a function of the temperature, but the
strength and direction of this function varies. The presence of
dissolved salts or minerals in water leads to moderate
decreases in solubility (31). In a mixed solvent system,
consisting of water and one or more water-miscible
compounds, as the fraction of the cosolvent in the mixture
increases, the solubility of the organic chemical increases
exponentially (12). In general, the greater the molecular weight
and structural complexity of the organic compound, the lower
the solubility.
Organic compounds are only rarely found in ground water at
concentrations approaching their solubility limits, even when
organic liquid phases are known or suspected to be present.
The observed concentrations are usually more than a factor of
10 lower than the solubility presumably due to diffusional
limitations of dissolution and the dilution of the dissolved
organic contaminants by dispersion (74). This has also been
attributed to: reduced solubility due to the presence of other
soluble compounds, the heterogeneous distribution of DNAPL
in the subsurface, and dilution from monitoring wells with long
intake lengths (10). Detection of DNAPL components in the
subsurface below the solubility should clearly not be
interpreted as a negative indicator for the presence of DNAPL.
In a DNAPL spill scenario where the DNAPL or its vapors are
in contact with the ground water, the concentration of the
soluble phase components may range from non-detectable up
to the solubility of the compound. The rate of dissolution has
been expressed as a function of the properties of the DNAPL
components (solubility), ground water flow conditions,
differential between the actual and solubility concentration, and
the contact area between the DNAPL and the ground water
(10). The contact area is expected to be heterogeneous and
difficult to quantify. Additionally, as the time of contact
increases between the DNAPL and the water, the
concentration in the aqueous phase increases.
Vapor Pressure
The vapor pressure is that characteristic of the organic
chemical which determines how readily vapors volatilize or
evaporate from the pure phase liquid. Specifically, the partial
pressure exerted at the surface by these free molecules is
known as the vapor pressure (30). Molecular activity in a liquid
tends to free some surface molecules and this tendency
towards vaporization is mainly dependent on temperature. The
vapor pressure of DNAPL's can actually be greater than the
vapor pressure of volatile organic compounds. For example, at
20 C, the ratio of the vapor pressures of TCE and benzene is
1.4(1).
Volatility
The volatility of a compound is a measure of the transfer of the
compound from the aqueous phase to the gaseous phase. The
transfer process from the water to the atmosphere is
dependent on the chemical and physical properties of the
compound, the presence of other compounds, and the physical
properties (velocity, turbulence, depth) of the water body and
atmosphere above it. The factors that control volatilization are
the solubility, molecular weight, vapor pressure, and the nature
of the air-water interface through which it must pass (31). The
Henry's constant is a valuable parameter which can be used to
help evaluate the propensity of an organic compound to
volatilize from the water. The Henry's law constant is defined
as the vapor pressure divided by the aqueous solubility.
Therefore, the greater the Henry's law constant, the greater the
tendency to volatilize from the aqueous phase, refer to Table 1.
Interfacial Tension
The unique behavior of DNAPLs in porous media is largely
attributed to the interfacial tension which exists between
DNAPL and water, and between DNAPL and air. These
interfacial tensions, result in distinct interfaces between these
fluids at the pore-scale. When two immiscible liquids are in
contact, there is an interfacial energy which exists between the
fluids resulting in a physical interface. The interfacial energy
arises from the difference between the inward attraction of the
molecules in the interior of each phase and those at the
surface of contact (2). The greater the interfacial tension
between two immiscible liquids; the less likely emulsions will
form; emulsions will be more stable if formed, and the better
the phase separation after mixing. The magnitude of the
interfacial tension is less than the larger of the surface tension
values for the pure liquids, because the mutual attraction of
unlike molecules at the interface reduces the large imbalance
of forces (31). Interfacial tension decreases with increasing
temperature, and may be affected by pH, surfactants, and
gases in solution (36). When this force is encountered between
a liquid and a gaseous phase, the same force is called the
surface tension (66).
The displacement of water by DNAPL and the displacement of
DNAPL by water in porous media often involves a phenomena
referred to as immiscible fingering. The lower the interfacial
tension between immiscible fluids, the greater the instability of
the water:DNAPL interface and thus the greater the immiscible
fingering (27). The distribution of the fingering effects in porous
media has been reported to be a function of the density,
viscosity, surface tension (27) and the displacement velocity
8
-------
(13) of the fluids involved as well as the porous media
heterogeneity (28).
Wettability
Wettability refers to the relative affinity of the soil for the
various fluids - water, air, and the organic phase. On a solid
surface, exposed to two different fluids, the wettability can be
inferred from the contact angle (66), also referred to as the
wetting angle, refer to Figure 15. In general, if the wetting angle
is less than 90 degrees, the fluid is said to be the wetting fluid.
In this scenario, water will preferentially occupy the smaller
pores and will be found on solid surfaces (14). When the
wetting angle is near 90 degrees, neither fluid is preferentially
attracted to the solid surfaces. If the wetting angle is greater
than 90 degrees, the DNAPL is said to be the wetting fluid. The
wetting angle is an indicator used to determine whether the
porous material will be preferentially wetted by either the
hydrocarbon or the aqueous phase (71). Wettability, therefore,
describes the preferential spreading of one fluid over solid
surfaces in a two-fluid system. The wetting angle, which is a
measure of wettability, is a solid-liquid interaction and can
actually be defined in terms of interfacial tensions (71).
Several methods have been developed to measure the wetting
angle (36,71). In most natural systems, water is the wetting
fluid, and the immiscible fluid is the non-wetting fluid. Coal tar
may be the exception (i.e. contact angle greater than 90
degrees), which is mainly attributed to the presence of
surfactants (70). The wetting fluid will tend to coat the surface
of grains and occupy smaller spaces (i.e. pore throats) in
porous media, the non-wetting fluid will tend to be restricted to
the largest openings (47).
The wetting angle depends on the character of the solid
surface on which the test is conducted. The test is conducted
on flat plates composed of minerals which are believed
representative of the media, or on glass. Contact angle
measurements for crude oil indicates that the wetting angles
vary widely depending on the mineral surface (53). Soil and
aquifer material are not composed of homogeneous mineral
composition nor flat surfaces. The measured wetting angle can
only be viewed as a qualitative indicator of wetting behavior.
The reader is recommended to refer to reference No. 31 for
review of the basic principles and for various techniques to
measure the following DNAPL parameters: density, viscosity,
interfacial tension, solubility, vapor pressure, and volatility.
O
V
8
o
y «<90°
¦6 fDNAP^
Xn
WamngRuti DNAPL
Wtttng Rutd Wttar
Water
Water
Fluid Relationships
?Y5terri
Wettina Fluid
Non-Wettina Fluid
air water
water
air
air: DNAPL
DNAPL
air
water:DNAPL
water
DNAPL
air:DNAPL:water
water>organioair
(i)
(1) Wettrg (lutdordar
f Mm. Waterloo Centra for
V. Groundwatar RasMrch, 1969 J
Figure 15. Wetting angle and typical wetting fluid relationships.
Subsurface Media Characteristics
Capillary Force/Pressure
Capillary pressure is important in DNAPL transport because it
largely determines the magnitude of the residual saturation that
is left behind after a spill incident. The greater the capillary
pressure, the greater the potential for residual saturation. In
general, the capillary force increases in the following order;
sand, silt, clay. Correspondingly, the residual saturation
increases in the same order. Capillary pressure is a measure
of the tendency of a porous medium to suck in the wetting fluid
phase or to repel the nonwetting phase (2). Capillary forces are
closely related to the wettability of the porous media. The
preferential attraction of the wetting fluid to the solid surfaces
cause that fluid to be drawn into the porous media. Capillary
forces are due to both adhesion forces (the attractive force of
liquid for the solids on the walls of the channels through which
it moves) and cohesion forces (the attraction forces between
the molecules of the liquid) (32). The capillary pressure
depends on the geometry of the void space, the nature of
solids and liquids, the degree of saturation (2) and in general,
in-creases with a decrease in the wetting angle and in pore
size, and with an increase in the interfacial tension (71). All
pores have some value of capillary pressure. Before a
nonwetting fluid can enter porous media, the capillary pressure
of the largest pores (smallest capillary pressure) must be
exceeded. This minimum capillary pressure is called the entry
pressure.
In the unsaturated zone, pore space may be occupied by
water, air (vapors), or immiscible hydrocarbon. In this scenario,
capillary pressure retains the water (wetting phase) mainly in
the smaller pores where the capillary pressure is greatest. This
restricts the migration of the DNAPL (non-wetting phase)
through the larger pores unoccupied by water. Typically,
DNAPL does not displace the pore water from the smaller
pores. It is interesting to note that the migration of DNAPL
through fine material (high capillary pressure) will be impeded
upon reaching coarser material (low capillary pressure).
The capillary fringe will obstruct the entry of the DNAPL into
the saturated zone. When a sufficient volume of DNAPL has
been released and the "DNAPL pressure head" exceeds the
water capillary pressure at the capillary fringe (entry pressure),
the DNAPL will penetrate the water table. This is why DNAPL
is sometimes observed to temporarily flatten out on top of the
water table. Similarly, laboratory experiments have been
conducted in which DNAPL (tetrachloroethylene) infiltrating
through porous media was found to flow laterally and cascade
off lenses too fine to penetrate (28), (refer to Figure 11). This
was attributed to the inability of the DNAPL to overcome the
high capillary pressure associated with the lenses. Logically,
when "DNAPL pressure head" exceeds the capillary pressure,
the DNAPL will penetrate into the smaller pores. These
laboratory experiments are important because they illustrate
that small differences in the capillary characteristics of porous
media can induce significant lateral flow of non-wetting fluids.
A comprehensive investigation of capillary trapping and
multiphase flow of organic liquids in unconsolidated porous
media revealed many intricacies of this process in the vadose
and saturated zone (66). An important note is that while
capillary pressure is rarely measured at hazardous waste sites,
9
-------
the soil texture (sand, silt, clay) is usually recorded during
drilling operations and soil surveys. This information, along with
soil core analyses will help to delineate the stratigraphy and
the volume distribution of NAPL.
Pore Size Distribution/Initial Moisture Content
In natural porous media, the geometry of the pore space is
extremely irregular and complex (2). The heterogeneity of the
subsurface environment i.e. the variability of the pore size
distribution, directly affects the distribution of the capillary
pressures along the interfaces between the aqueous and
immiscible phases (50). In saturated column experiments, it
was observed that NAPL preferentially traveled through strings
of macropores, almost completely by-passing the water filled
micropores (66). In the same study, a heterogeneous
distribution of coarse and fine porous material was simulated.
Most of the incoming organic liquid preferentially traveled
through the coarse lens material.
In short term column drainage experiments, results indicated
that the particle grain size is of primary importance in
controlling the residual saturation of a gasoline hydrocarbon
(19). Fine and coarse sands (dry) were found to have 55%
and 14% residual saturation, respectively. The finer the sand,
the greater the residual saturation. During these experiments,
the residual saturation was reduced 20-30% in a medium
sand and 60% in a fine sand when the sands were initially wet.
Soil pore water held tightly by capillary forces in the small
pores will limit the NAPL to the larger pores, and thus, result in
lower residual saturation. In a similar laboratory (unsaturated)
column study, the smaller the grain size used in the
experiment, the greater the residual saturation of the NAPL
(74). The residual saturation in the saturated column
experiments was found to be greater than the unsaturated
columns and was independent of the particle size distri-
bution.
These observations follow traditional capillary force theory.
Residual saturation resulting from a ONAPL spill in the
unsaturated zone is highly dependent on the antecedent
moisture content in the porous media. When the moisture
content is low, the strong capillary forces in the smaller pores
will tenaciously draw in and hold the DNAPL. When the
moisture content is high, the capillary forces in the smaller
pores will retain the soil pore water, and DNAPL residual
saturation will mainly occur in the larger pores. Therefore,
greater residual saturation can be expected in dryer soils.
Correspondingly, NAPL will migrate further in a wetter soil,
and displacement of NAPL from small pores is expected to
be more difficult than from large pores.
Stratigraphic Gradient
DNAPL migrating vertically will likely encounter a zone or
stratigraphic unit of lower vertical permeability. A reduction in
the vertical permeability of the porous media will induce lateral
flow of the DNAPL The gradient of the lower permeable
stratigraphic unit will largely determine the direction in which
the DNAPL will flow. This is applicable to both the saturated
and unsaturated zones. As depicted in Figures 13a and 13b,
the lateral direction of DNAPL flow may be in a different
direction than ground-water flow.
Ground Water Flow Velocity
The ground water flow velocity is a dynamic stress parameter
which tends to mobilize the hydrocarbon (39). As the ground
water velocity increases, the dynamic pressure and viscous
forces increase. Mobilization of DNAPL occurs when the
viscous forces of the ground water acting on the DNAPL,
exceeds the porous media capillary forces retaining the
DNAPL.
Saturation Dependent Functions
Residual Saturation
Residual saturation is defined as the volume of hydrocarbon
trapped in the pores relative to the total volume of pores (38)
and therefore is measured as such (74). Residual saturation
has also been described as the saturation at which NAPL
becomes discontinuous and is immobilized by capillary forces
(36). The values of residual saturation vary from as low as 0.75
-1.25% for light oil in highly permeable media to as much as
20% for heavy oil (50). Residual saturation values have also
been reported to range from 10% to 50% of the total pore
space (39,74). Other researchers reported that residual
saturation values appear to be relatively insensitive to fluid
properties and very sensitive to soil properties (and
heterogeneities) (66). Laboratory studies conducted to predict
the residual saturation in soils with similar texture and grain
size distribution yielded significantly different values. It was
concluded that minor amounts of clay or silt in a soil may play
a significant role in the observed values.
In the unsaturated zone during low moisture conditions, the
DNAPL residual saturation will wet the grains in a pendular
state (a ring of liquid wrapped around the contact point of a
pair of adjacent grains). During high moisture conditions, the
wetting fluid, which is typically water, will preferentially occupy
the pendular area of adjacent grains and the hydrocarbon will
occupy other available pore space, possibly as isolated
droplets. In the saturated zone, the DNAPL residual saturation
will be present as isolated drops in the open pores (47).
Furthermore, results of laboratory experimentation indicated
that residual saturation increased with decreasing hydraulic
conductivity in both the saturated and unsaturated zones and
that the residual saturation is greatest in the saturated zone.
Laboratory experiments indicated that vadose zone residual
saturation was roughly one third of the residual saturation in
the saturated zone (66). The increase in residual saturation in
the saturated zone is due to the following: [1] the fluid density
ratio (DNAPL:air versus DNAPL:water above and below the
water table, respectively) favors greater drainage in the vadose
zone; [2] as the non-wetting fluid in most saturated media,
NAPL is trapped in the larger pores; and, [3] as the wetting
fluid in the vadose zone, NAPL tends to spread into adjacent
pores and leave a lower residual content behind, a process
that is inhibited in the saturated zone (36). Thus, the capacity
for retention of DNAPLs in the unsaturated zone is less than
the saturated zone.
Relative Permeability
Relative permeability is defined as the ratio of the permeability
of a fluid at a given saturation to its permeability at 100%
saturation. Thus it can have a value between 0 and 1 (71).
10
-------
Figure 16 illustrates a relative permeability graph for a two fluid
phase system showing the relationship between the observed
permeability of each fluid for various saturations to that of the
observed permeability if the sample were 100% saturated with
that fluid (73). The three regions of this graph are explained as
follows (71): Region I has a high saturation of DNAPL and is
considered a continuous phase while the water is a
discontinuous phase, therefore, water permeability is low.
Assuming the DNAPL is the non-wetting fluid, water would fill
the smaller capillaries and flow through small irregular pores. In
Region II, both water and DNAPL are continuous phases
although not necessarily in the same pores. Both water and
- Increasing DNAPL Saturation
Increasing Water Saturation 1
rntmm tnd WUOt. 1171
Figure 16. Relative permeability graph.
NAPL flow simultaneously. However, as saturation of either
phase increases, the relative permeability of the other phase
correspondingly decreases. Region III exhibits a high
saturation of water while the DNAPL phase is mainly
discontinuous. Water flow dominates this region and there is
little or no flow of DNAPL.
Both fluids flow through only a part of the pore space and thus
only a part of the cross section under consideration is available
for flow of each fluid. Therefore, the discharge of each fluid
must be lower corresponding to its proportion of the cross
sectional area (46).
Figure 17 is another relative permeability graph which
demonstrates several points. Small increases in DNAPL
saturation results in a significant reduction in the relative
permeability of water. However, a small increase in water
saturation does not result in a significant reduction in DNAPL
relative permeability. This figure identifies two points, S01 and
S02, where the saturation of the DNAPL and the water are
greater than 0 before there is a relative permeability for this
fluid. The two fluids hinder the movement of the other to
different degrees and both must reach a minimum saturation
before they achieve any mobility at all (47). These minimum
saturations, for the water and DNAPL, are identified as
irreducible and residual saturation, respectively.
( After Schwille, t988 ^
Figure 17. The relative permeability curves for water and a
DNAPL in a porous medium as a function of the pore
space saturation.
Site Characterization for DNAPL
Characterization of the subsurface environment at hazardous
waste sites containing DNAPL is complex and will likely be
expensive. Specific details associated with the volume and
timing of the DNAPL release are usually poor or are not
available and subsurface heterogeneity is responsible for the
complicated and unpredictable migration pathway of
subsurface DNAPL transport. As discussed previously, slight
changes in vertical permeability may induce a significant
horizontal component to DNAPL migration.
Site characterization typically involves a significant investment
in ground-water analyses. Although analysis of ground water
provides useful information on the distribution of the soluble
components of the DNAPL, the presence of other phases of
the DNAPL may go unrecognized. The investigation must,
therefore, be more detailed to obtain information concerning
the phase distribution of the DNAPL at a site. Site
characterization may require analyses on all four phases
(aqueous, gaseous, solid, immiscible) to yield the appropriate
information (refer to Table 2). In brief, data collected on the
various phases must be compiled, evaluated and used to help
identify: where the contaminant is presently located; where it
has been; what phases it occurs in; and what direction the
mobile phases may be going. A comprehensive review of site
characterization for subsurface investigations is available (68).
Development of monitoring and remediation strategies can be
focused more effectively and efficiently after a clear definition
of the phase distribution has been completed.
Ground Water
Ground water analyses for organic compounds, in conjunction
with ground water flow direction data, has repeatedly been
used to: delineate the extent of ground water contamination
from DNAPL; determine the direction of plume migration; and
11
-------
Table 2 - Phase Distribution of DNAPL in the Subsurface
MATRIX
PHASE
1.
ground water
aqueous - soluble components of DNAPL
2.
soil/aquifer
material
t
solid - adsorbed components of DNAPL
on solid phase material
3.
DNAPL
immiscible - continuous phase (mobile),
residual saturation (immobile)
4.
soil gas
gaseous - volatile components
to identify probable DNAPL source area(s). While this
approach has been used successfully to characterize the
distribution of contaminants in the subsurface, there are
limitations. For example, since DNAPL and ground water may
flow in different directions, as indicated in Figures 13a and 13b,
ground water analyses may not necessarily identify the
direction of DNAPL migration.
Ground water analyses may be useful to identify probable
DNAPL source areas, but, estimating the volume of DNAPL in
the subsurface is limited using this approach. Soluble phase
components of DNAPL are rarely found in excess of 10% of
the solubility even when organic liquids are known or
suspected to be present. The concentration of soluble DNAPL
components in the ground water is not only a function of the
amount of DNAPL present, but also the chemical and physical
characteristics of the DNAPL, the contact area and time
between the ground water and DNAPL, and numerous
transport and fate parameters (retardation, biodegradation,
dispersion, etc.). One technique has been developed using
chemical ratios in the ground water as a means of source
identification and contaminant fate prediction (18).
Soll/Aqulfer Material
Exploratory Borings
Physical and chemical analyses of soil and aquifer material
(drill cuttings, cores) from exploratory borings will provide
useful information in the delineation of the horizontal and
vertical mass distribution of DNAPL. While simple visual
examination for physical presence or absence of contamination
might seem like a worthwhile technique, it can be deceiving
and does nothing to sort out the various liquid phases and their
relationship to each other (71). A quantitative approach is
necessary to determine DNAPL distribution.
Drill cuttings or core material brought to the surface from
exploratory borings can be screened initially to help delineate
the depth at which volatile components from the various
phases of the hydrocarbon exists. The organic vapor analyzer
and the HNU are small portable instruments that can detect
certain volatile compounds in the air. These methods are used
to initially screen subsurface materials for volatile components
of DNAPL. identification of individual compounds and their
concentrations may be confirmed by other, more precise,
analyses.
Analysis of the soil or aquifer material by more accurate
means, such as gas chromatography or high pressure liquid
chromatography, will take longer but will provide more specific
information on a larger group of organic compounds, i.e.,
volatile/non-volatile, and on specific compounds. This
information is necessary to help fix the horizontal and vertical
mass distribution of the contaminant and to help delineate the
phase distribution. These analyses do not distinguish between
soluble, sorbed or free-phase hydrocarbon, however; a low
relative concentration indicates that the contaminant may
mainly be present in the gaseous or aqueous phases; and a
high relative concentration indicates the presence of sorbed
contaminant or free phase liquid either as continuous-phase or
residual saturation. A more rigorous set of analyses is required
to distinguish between the various phases.
Additional tests to identify the presence of NAPL in soil or
aquifer core sample are currently undeveloped and research in
this area is warranted. Squeezing and immiscible displacement
techniques have been used to obtain the pore water from
cores (40). Other methods of phase separation involving
vacuum or centrifugation may also be developed for this use. A
paint filter test was proposed in one Superfund DNAPL field
investigation where aquifer cores were placed in a filter/funnel
apparatus, water was added, and the filtrate was examined for
separate phases. These core analysis techniques have
potential to provide valuable field data to characterize NAPL
distribution.
Cone Penetrometer
The cone penetrometer (ASTM D3441-86)(69) has been used
for some time to supply data on the engineering properties of
soils. Recently, the application of this technology has made the
leap to the hazardous waste arena. The resistance of the
formation is measured by the cone penetrometer as it is driven
vertically into the subsurface. The resistance is interpreted as
a measure of pore pressure, and thus provides information on
the relative stratigraphic nature of the subsurface. Petroleum
and chlorinated hydrocarbon plumes can be detected most
effectively when the cone penetrometer is used in conjunction
with in-situ sensing technologies (48). Features of the cone
penetrometer include: a continuous reading of the stratigraphy/
permeability; in-situ measurement; immediate results are
available; time requirements are minimal; vertical accuracy of
stratigraphic composition is high; ground-water samples can be
collected in-situ; and the cost is relatively low.
Data from the cone penetrometer can be used to delineate
probable pathways of DNAPL transport. This is accomplished
by identifying permeability profiles in the subsurface. A zone of
low permeability underlying a more permeable stratigraphic
unit will likely impede vertical transport of the DNAPL. Where
such a scenario is found, a collection of DNAPL is probable
and further steps can be implemented to more accurately and
economically investigate and confirm such an occurrence.
This general approach has successfully been implemented at
one Superfund site (8).
DNAPL
Well Level Measurements
In an effort to delineate the horizontal and vertical extent of the
DNAPL at a spill site, it is important to determine the elevation
12
-------
%
if
\7
Measured > Actual
/a/ "¦
DNAPL Pool
/
Measured
^^^Impermeable z'
yz///////////^^^ PW^sr. ^
of DNAPL in the subsurface. Monitoring DNAPL elevation over
lime will indicate the mobility of the DNAPL. There are several
methods that can be used to determine the presence of
DNAPL in a monitoring well. One method relies on the
difference in electrical conductivity between the DNAPL and
water. A conductivity or resistivity sensor is lowered into the
well and a profile is measured. The interface of the DNAPL is
accurately determined when the difference in conductivity is
detected between the two fluids. This instrument may also be
used to delineate LNAPL. A transparent, bottom-loading bailer
can also be used to measure the thickness (and to sample) of
DNAPL in a well (36). The transparent bailer is raised to the
surface and the thickness of the DNAPL is made by visual
measurement.
Several laboratory and field studies have been performed
which investigate the anomaly between the actual and
measured LNAPL levels in ground-water wells (15,16,24,25).
The anomaly between actual and measured NAPL thickness in
the subsurface is also applicable to DNAPL, but for different
reasons. The location of the screening interval is the key to
understanding both scenarios. First, if the well screen interval
is situated entirely in the DNAPL layer, and the hydrostatic
head (water) in the well is reduced by pumping or bailing, then
to maintain hydrostatic equilibrium, the DNAPL will rise in the
well (36,44,71) (refer to Figure 18). Secondly, if the well screen
extends into the barrier layer, the DNAPL measured thickness
will exceed that in the formation by the length of the well below
the barrier surface (36) (refer to Figure 19). Both of these
scenarios will result in a greater DNAPL thickness in the well
and thus a false indication (overestimate) of the actual DNAPL
thickness will result. One of the main purposes of the
monitoring well in a DNAPL investigation is to provide
information on the thickness of the DNAPL in the aquifer.
Therefore, construction of the well screen should intercept the
ground water:DNAPL interface and the tower end of the screen
should be placed as close as possible to the impermeable
stratigraphic unit.
Figure 18. A well screened only In the DNAPL in conjunction
with lower hydrostatic head (i.e. water) in the well
may result in an overestlmatlon of DNAPL thickness.
Figure 19. A well screened into an impermeable boundary
may result in an over-estimation of the DNAPL
thickness.
DNAPL Sampling
Sampling of DNAPL from a well is necessary to perform
chemical and physical analyses on the sample. Two of the
most common methods used to retrieve a DNAPL sample from
a monitoring well are the peristaltic pump and the bailer. A
peristaltic pump can be used to collect a sample if the DNAPL
is not beyond the effective reach of the pump, which is typically
less than 25 feet. The best method to sample DNAPL is to use
a double check valve bailer. The key to sample collection is
controlled, slow lowering (and raising) of the bailer to the
bottom of the well (57). The dense phase should be collected
prior to purging activities.
Soil-Gas Surveys
A soil-gas survey refers to the analysis of the soil air phase as
a means to delineate underground contamination from volatile
organic chemicals and several techniques have been
developed (34,52). This investigative tool is mainly used as a
preliminary screening procedure to delineate the areal extent
of volatile organic compounds in the soil and ground water.
This method is quick, less expensive than drilling wells and can
provide greater plume resolution (33).
Data from a soil-gas survey is a valuable aid in the
development of a more detailed subsurface investigation
where ground water monitoring wells and exploratory borings
are strategically located for further site characterization. There
are limitations to soil-gas surveys (26,52) and data
interpretation must be performed carefully (35,49). Soil-gas
investigations have mainly been conducted to identify the
location of the organic contaminants in ground water. At the
time of this publication, the scientific literature did not contain
information specifically applicable to the delineation of DNAPL
from soil-gas survey data. However, it is surmisable that soil-
gas surveys can be used to help delineate DNAPL residual
saturation in the unsaturated zone or the location of perched
DNAPL reservoirs.
13
-------
Miscellaneous
Pumping Systems
The vertical migration of DNAPL in the saturated zone will
eventually be challenged by a low permeability stratigraphic
unit. According to the principles of capillary pressure, the lower
permeability unit will exhibit a greater capillary pressure.
Displacement of water by DNAPL requires that the hydrostatic
force from the mounding DNAPL exceed the capillary force of
the low permeability unit. The Hobson formula is used to
compute the critical height calculation to overcome the
capillary pressure under different pore size conditions (70).
In an effort to minimize further DNAPL contamination as a
result of drilling investigations, precautionary steps should be
taken. Penetration of DNAPL reservoirs in the subsurface
during drilling activities offers a conduit for the DNAPL to
migrate vertically into previously uncontaminated areas. It is
very easy to unknowingly drill through a DNAPL pool and the
bed it sits on, causing the pool to drain down the hole into a
deeper part of the aquifer or into a different aquifer (32).
Special attention to grouting and sealing details during and
after drilling operations will help prevent cross-contamination.
Precautionary efforts should also be considered when a
DNAPL reservoir is encountered during drilling operations. The
recommended approach is to cease drilling operations and
install a well screen over the DNAPL zone and cease further
drilling activities in the well. If it is necessary to drill deeper,
construction of an adjacent well is recommended. Alternatively,
if it is not necessary to screen off that interval, it is
recommended to carefully seal off the DNAPL zone prior to
drilling deeper.
Well construction material compatibility with DNAPL should be
investigated to minimize downhole material failure. A
construction material compatibility review and possible testing
will prevent the costly failure of well construction material. The
manufacturers of well construction material are likely to have
the most extensive compatibility data and information
available.
Remediation
Pumping represents an important measure to stop the mobile
DNAPL from migrating as a separate phase by creating a
hydraulic containment and by removal of DNAPL (44). Very
simply, DNAPL recovery is highly dependent on whether the
DNAPL can be located in the subsurface. The best recovery
scenario is one in which the DNAPL is continuous and has
collected as a reservoir in a shallow, impermeable subsurface
depression. Once the DNAPL has been located and recovery
wells are properly installed, pumping of pure phase DNAPL is
a possible option but depends largely on site specific
conditions which include, but are not limited to: DNAPL
thickness, viscosity, and permeability.
Many DNAPL reservoirs in the subsurface are of limited
volume and areal extent. Therefore, it can be expected that
both the level of DNAPL (saturated thickness) in the well will
decline from the prepumping position and the percentage of
DNAPL in the DNAPLrwater mixture will decrease rather
rapidly. Correspondingly, DNAPL recovery efficiency
decreases. Field results indicate that recovery wells screened
only in the DNAPL layer will maintain maximum DNAPLiwater
ratios (102). Well diameter was not found to influence long
term DNAPL recovery; however, large diameter wells allow
high volume pumping for short durations; and small diameter
wells result in lower DNAPL:water mixtures and greater
drawdown.
An enhanced DNAPL recovery scheme may be used to
improve recovery efficiency. An additional well is constructed
with a screen interval in the ground water zone located
vertically upward from the DNAPL screen intake. Ground water
is withdrawn from the upper screen which results in an
upwelling of the DNAPL (70), refer to Figure 20. The upwelling
of the DNAPL, coal tar in this case, improved the rate (twofold)
at which the coal tar was recovered resulting in a more efficient
operation. The ground water withdrawal rate must be carefully
determined; too much will result in the coal tar from rising
excessively and being either mixed (emulsions) with or
suppressed by the higher water velocity above; too low will not
Remediation of DNAPL mainly involves physical removal by
either pumping or trench-drainline systems. Removal of
DNAPL early in the remediation process will eliminate the main
source of contaminants. This step will substantially improve the
overall recovery efficiency of the various DNAPL phases
including the long term pump and treat remediation efforts for
soluble components. Remediation technologies such as
vacuum extraction, biodegradation, ground water pumping,
and soil flushing is mainly directed at the immobile DNAPL and
the various phases in which its components occur. Physical
barriers can be used in an effort to minimize further migration
of the DNAPL.
Clean-up of DNAPL can involve sizable expenditures: they are
difficult to extract and the technology for their removal is just
evolving (43). Historically, field recovery efforts usually proceed
with a poor understanding of the volume distribution of the
DNAPL. This reflects the difficulties involved in adequate site
characterization, poor documentation of the release, and the
complexity associated with the DNAPL transport in the
subsurface.
Figure 20. A DNAPL recovery system where deliberate
upwelling of the static coal-tar surface Is used to
increase the flow of product into the recovery wells.
14
-------
caused upwelling. An estimate of this upwelling can be
calculated using the simplified Ghyben-Herzberg Principle
under ideal conditions (4). Laboratory studies indicated that
dimethyl phthalate (1.19 g/cc) recovery rate was doubled or
tripled over the conventional, non-upconing, recovery scheme
(75). A similar application of this technique was used to
increase the level of DNAPL (solvents) in a sandstone bedrock
formation (11). Other enhanced DNAPL recovery techniques
were implemented utilizing both water flooding and wellbore
vacuum. Essentially, this minimized drawdown, allowing a
maximum pumping rate of the DNAPL:water mixture. Both
techniques offered significant advantages in terms of the rate
and potential degree of DNAPL removal (8).
The highly corrosive nature of some DNAPL's may increase
maintenance problems associated with the recovery system. A
design consideration during any DNAPL recovery program
should include a material compatibility review to minimize
downhole failures. This is applicable to the well construction
material and the various appurtenances of the recovery
system. Manufacturers of the construction material would
most likely have the best compatibility information available.
While most scientists agree that the residual saturation of
immiscible hydrocarbon droplets in porous media are
immobile, researchers have investigated the mobility of
residual saturation in porous media for enhanced oil recovery
and for NAPL remediation at spill sites. Specifically, this
includes a complex interplay between four forces (viscous,
gravity, capillary, buoyancy). These forces are dependent on
both the chemical and physical characteristics of the DNAPL
and porous media. The mobilization of residual saturation
mainly hinges on either increasing the ground water velocity
which increases the viscous forces between the residual
saturation and the ground water, or decreasing the interfacial
tension between the residual saturation and the ground water
which decreases the capillary forces.
The capillary number is an empirical relationship which
measures the ratio between the controlling dynamic stresses
(absolute viscosity and ground water velocity) and static
stresses (interfacial tension) of the residual saturation (39). The
former are the viscous stresses and the dynamic pressure in
the water which tend to move the oil. The latter are the
capillary stresses in the curved water/oil interfaces which tend
to hold the oil in place. As the capillary number is increased,
the mobility of the residual saturation increases. In a laboratory
column study, the capillary number had to be increased two
orders of magnitude from when motion was initiated to
complete displacement of the hydrocarbon in a sandstone core
(74). In a glass bead packed column, only one order of
magnitude increase was required. However, a higher capillary
number was required to initiate mobility. The difference in
mobility between the two columns was attributed to the pore
geometry, i.e. size, shape.
There are limitations to residual saturation mobilization. The
ground water gradient (dh/dl) necessary to obtain the critical
capillary number to initiate blob mobilization would be 0.24. To
obtain complete NAPL removal would require a gradient of 18
(3). Ground water gradients of this magnitude are unrealistic.
Another estimate of the gradient necessary to mobilize carbon
tetrachloride in a fine gravel and medium sand was 0.09 and
9.0 respectively (74). The former gradient is steep but not
unreasonable and the latter gradient is very steep and
impractical to achieve in the field. The same researchers
concluded from more recent, comprehensive studies, that the
earlier predictions were optimistic, and that the gradient
necessary to mobilize residual organic liquid is clearly
impractical (66). Another limitation is that along with residual
saturation mobilization, the NAPL blobs disperse into smaller
blobs and that the blob distribution was dependent on the
resulting capillary number (6). Recovery of the NAPL residual
saturation by pumping ground water may be more feasible
where the porous media is coarse and capillary forces are low,
i.e. coarse sands and gravel. However, even in this scenario, it
is expected that the radius of residual saturation mobilization
would be narrow.
It is held in petroleum engineering theory that the only practical
means of raising the capillary number dramatically is by
lowering the interfacial tension (39) and that this can be
achieved by using surfactants (66). Surlactants reduce the
interfacial tension between two liquids, and therefore, are
injected into the subsurface for enhanced recovery of
immiscible hydrocarbons. In laboratory experiments, surfactant
flushing solutions produced dramatic gains in flushing even
after substantial water flushing had taken place (54).
Unfortunately, surfactants can be quite expensive and cost
prohibitive in NAPL recovery operations. Surfactants are
usually polymeric in nature and a surfactant residue may be
left behind in the porous media which may not be
environmentally acceptable. Additionally, surfactants may be
alkaline and thus affect the pH of the subsurface environment.
It has been suggested that such a surfactant may inhibit
bacterial metabolism and thus preclude subsequent use of
biological technologies at the site. Significant research in this
area is currently underway which may uncover information
improving the economics and feasibility of this promising
technology.
In summary, practical considerations and recommendations
concerning the mobilization and recovery of residual saturation
include the folbwing: greater effectiveness in very coarse
porous media i.e. coarse sands and gravel; recovery wells
should be installed close to the source to minimize flow path
distance; a large volume of water will require treatment/
disposal at the surface; compounds with high interfacial
tension or viscosity will be difficult to mobilize; and implemen-
tation of linear one-dimensional sweeps through the zones of
residual saturation (74) and surfactants will optimize recovery.
Pumping the soluble components (aqueous phase) of DNAPL
from the immiscible (continuous and residual saturation), solid
(sorbed), and gaseous phases has been perhaps one of the
most effective means to date to both recover DNAPL from the
subsurface and to prevent plume migration. Recovery of
soluble components quite often has been the only remediation
means available. This is largely attributed to the inability to
locate DNAPL pools and due to low, DNAPL yielding
formations. The basic principles and theory of pump and treat
technology and the successes and failures have been
summarized in other publications (64,67) and is beyond the
scope of this publication.
Pumping solubilized DNAPL components from fractured rock
aquifers historically has been plagued with a poor recovery
efficiency. Although the rock matrix has a relatively small
intergranular porosity, it is commonly large enough to allow
dissolved contaminants from the fractures to enter the matrix
15
-------
T-r GiewdWa« SjIioi
Qiotfid Sufaoa
Lu£_
OHUPt ^^DNAPCOnrtw
r / / / /7v ////ft
ry OrouTd War Sufai
DNAPL Surf at
Qmtrd SuHaoa
/ / / 9 / / / / ? /~r-
OM Distribution
• DNAPL dartsar than ground wator,
has aocumiatod at ho base of f»
DNAPL Mounding
• Drvardown of N ovsrtying water
(aba by pumping tha watar dranins
raaufts r moundng of (he DNAPL
DNAPL Recovery
Ground W«r Sulao*
DNAPL. Sufi
'/y / // / yyy
• Pumping from bofh 9» watar and
DNAPL fra^na induces inaaasng
DNAPL low to tha DNAPL dranlna
• Saparata production of DNAPL and
pound watar reduce* above pound
aaparrtm requwamvrts
• A Row pat) of maxvnum fcrmaScn
parmaatilty to DNAPL a estebehed
at t>a baaa of re ahMum
All* Satertll.lMa
Figure 21. Trench recovery system of DNAPL utilizing the dual
drainline concept.
by diffusion and be stored there by adsorption (32). The
release of these components is expected to be a slow diffusion
dominated process. This is because little or no water flushes
through dead-end fracture segments or through the porous,
impervious rock matrix. Therefore, clean-up potential is
estimated to be less than that expected for sand and gravel
aquifers.
Trench Systems
Trench systems have also been used successfully to recover
DNAPL and are used when the reservoir is located near the
ground surface. Trench systems are also effective when the
DNAPL is of limited thickness. Recovery lines are placed
horizontally on top of the impermeable stratigraphic unit.
DNAPL flows into the collection trenches and seep into the
recovery lines. The lines usually drain to a collection sump
where the DNAPL is pumped to the surface. Similar to the
pumping system, an enhanced DNAPL recovery scheme may
be implemented using drain lines to improve recovery
efficiency. This "dual drain line system" (41) utilizes a drain line
located in the ground water vertically upward from the DNAPL
line. Ground water is withdrawn from the upper screen which
results in an upwelling of the DNAPL which is collected in the
lower line, refer to Figure 21. This increases the hydrostatic
head of the DNAPL. Excessive pumping of either single or dual
drain line systems may result in the ground water "pinching off
the flow of DNAPL to the drain line. An advantage of the dual
drain system is that the oikwater separation requirements at
the surface are reduced.
Vacuum Extraction
Soil vacuum extraction (SVE) is a remediation technology
which involves applying a vacuum to unsaturated subsurface
strata to induce air flow. Figure 22 illustrates that the volatile
contaminants present in the contaminated strata will evaporate
and the vapors are recovered at the surface and treated.
Common methods of treatment include granular activated
carbon, catalytic oxidation, and direct combustion. SVE can
effectively remove DNAPL present as residual saturation or its
soluble phase components in the unsaturated zone. In general,
vacuum extraction is expected to be more applicable for the
chlorinated solvents (PCE, TCE, DCE) than the polycyclic
aromatic compounds (wood preserving wastes, coal tars, etc.).
When DNAPL is present in perched pools (Figure 12) it is more
effective to remove the continuous phase'DNAPL prior to the
implementation of SVE. The same strategy is applicable in the
saturated zone where DNAPL removal by SVE is attempted
concomitantly with lowering the water table. Upon lowering the
water table, SVE can be used to remove the remnant volatile
wastes not previously recovered. Often, the precise location of
the DNAPL is unknown; therefore, SVE can be used to
remediate the general areas where the presence of DNAPL is
suspected. Removal of DNAPL by SVE is not expected to be
as rapid as direct removal of the pure phase compound. One
advantage of SVE however, is that the precise location of the
DNAPL need not be known.
Important parameters influencing the efficacy of SVE concern
both the DNAPL and porous media. Porous media specific
parameters include: soil permeability, porosity, organic carbon,
moisture, structure, and particle size distribution. DNAPL
specific parameters include: vapor pressure, Henry's constant,
solubility, adsorption equilibrium, density, and viscosity (20).
These parameters and their relationships must be evaluated
on a site specific basis when considering the feasibility of
vacuum extraction and a practical approach to the design,
construction, and operation of venting systems (22).
Additionally, soil gas surveys which delineate vapor
concentration as a function of depth is critical in locating the
contaminant source and designing an SVE system.
Historically, SVE has been used to remove volatile compounds
from the soil. Recently it has been observed that SVE
enhances the biodegradation of volatile and semivolatile
organic compounds in the subsurface. While SVE removes
volatile components from the subsurface, it also aids in
supplying oxygen to biological degradation processes in the
unsaturated zone. Prior to soil venting, it was believed that
biodegradation in the unsaturated zone was limited due to
inadequate concentrations of oxygen (17). In a field study
where soil venting was used to recover jet fuel, it was observed
that approximately 15% of the contaminant removal was from
the result of microbial degradation. Enhanced aerobic
biodegradation during SVE increases the cost effectiveness of
the technology due to the reduction in the required above
ground treatment.
Vacuum extraction is one form of pump and treat which occurs
in the saturated zone where the fluid is a gas mixture.
Therefore, many of the same limitations to ground water pump
and treat are also applicable to vacuum extraction. While the
application of vacuum extraction is conceptually simple, its
success depends on understanding complex subsurface
16
-------
Figure 22. Vacuum extraction of DNAPL volatile components
in the unsaturated zone. As shown here, vapors are
treated by thermal combustion or carbon adsorp-
tion and the air is discharged to the atmosphere.
chemical, physical, and biological processes which provide
insight into factors limiting its performance (9).
Biodegradation
The potential for biodegradation of immiscible hydrocarbon is
highly limited for several reasons. First, pure phase
hydrocarbon liquid is a highly hostile environment to the
survival of most microorganisms. Secondly, the basic
requirements for microbiological proliferation (nutrients,
electron acceptor, pH, moisture, osmotic potential, etc.) is
difficult if not impossible to deliver or maintain in the DNAPL A
major limitation to aerobic bioremediation of high
concentrations of hydrocarbon is the inability to deliver
sufficient oxygen. A feasible remediation approach at sites
where immiscible hydrocarbon is present is a phased
technology approach. Initial efforts should focus on pure phase
hydrocarbon recovery to minimize further migration and to
decrease the volume of NAPL requiring remediation.
Following NAPL recovery, other technologies could be phased
into the remediation effort. Bioremediation may be one such
technology that could be utilized to further reduce the mass of
contaminants at the site. NAPL recovery preceding
bioremediation will improve bioremediation feasibility by
reducing the toxicity, time, resources, and labor.
Similar to other remediation technologies, a comprehensive
feasibility study evaluating the potential effectiveness of
bioremediation is critical and must be evaluated on a site
specific basis. A comprehensive review of biodegradation of
surface soils, ground water, and subsoils of wood preserving
wastes, i.e. PAH's (29,37,51,62,63) are available. A
comprehensive review of microbial decomposition of
chlorinated aromatic compounds is also available (58).
Soil Flushing
Soil flushing utilizing surfactants is a technology that was
developed years ago as a method to enhance oil recovery in
the petroleum industry. This technology is new to the
hazardous waste arena and available information has mainly
been generated from laboratory studies. Surfactant soil
flushing can proceed on two distinctly different mechanistic
levels: enhanced dissolution of adsorbed and dissolved phase
contaminants, and displacement of free-phase nonaqueous
contaminants. These two mechanisms may occur
simultaneously during soil flushing (42).
Surfactants, alkalis, and polymers are chemicals used to
modify the pore-level physical forces responsible for
immobilizing DNAPL. In brief, surfactants and alkalis reduce
the surface tension between the DNAPL and water which
increases the mobility. Polymers are added to increase the
viscosity of the flushing fluid to minimize the fingering effects
and to maintain hydraulic control and improve flushing
efficiency. Based on successful laboratory optimization studies
where an alkali-polymer-surfactant mixture was used, field
studies were conducted on DNAPL (creosote) which resulted
in recovery of 94% of the original DNAPL (42). Laboratory
research has also been conducted which indicated that
aqueous surfactants resulted in orders of magnitude greater
removal efficiency of adsorbed and dissolved phase
contaminants than water flushing (55).
Depth to contamination, DNAPL distribution, permeability,
heterogeneities, soil/water incompatibility, permeability
reduction, and chemical retention are important factors when
considering soil flushing (42). Prior to this technology being
cost effective in the fiula, surfactant recycling will be necessary
to optimize surfactant use (55). Soil flushing is complex from a
physical and chemical point of view; is relatively untested in the
field; and will likely be challenged regulatorily. Considerable
research currently being conducted in this area may result in
the increased use of this technology to improve DNAPL
recovery in the future.
Thermal methods of soil flushing involve injecting hot water or
steam in an effort to mobilize the NAPL. The elevated
temperature increases volatilization and solubilization and
decreases viscosity and density. A cold-water cap is used to
prevent volatilization. The mobile phases of the DNAPL are
then recovered using a secondary approach, i.e. pumping,
vacuum extraction etc. This approach (Contained Recovery of
Oily Wastes) to enhance recovery of DNAPL is currently under
EPA's Superfund Innovative Technology Evaluation Program
and a pilot-scale demonstration is forthcoming (21). A
limitation in the use of thermal methods is that the DNAPL may
be converted to LNAPL due to density changes (36). The
adverse effects from this are that the DNAPL, existing as a thin
layer, becomes buoyant and mobilizes vertically resulting in a
wider dispersal of the contaminant. Other limitations involve
the high energy costs associated with the elevated water
temperature and the heat loss in the formation (36).
Physical Barriers
Physical barriers may be used to prevent the migration of
DNAPL's in the subsurface and are typically used in
conjunction with other recovery means. One feature of physical
17
-------
barriers is the hydraulic control it offers providing the
opportunity to focus remediation strategies in treatment cells.
Unfortunately, physical barriers, while satisfactory in terms of
ground water control and containment of dissolved-phase
plumes, may contain small gaps or discontinuities which could
permit escape of DNAPL (7). Chemical compatibility between
physical barriers and construction material must agree to
insure the physical integrity of the barrier. The history of the
performance of these containment technologies is poorly
documented and is mainly offered here for completeness of
review. A more complete review of these physical barriers is
available (5,56).
Sheet piling involves driving lengths of steel that connect
together into the ground to form an impermeable barrier to
lateral migration of DNAPL. Ideally, the bottom of the sheet pile
should be partially driven into an impermeable layer to
complete the seal. Slurry walls involve construction of a trench
which is backfilled with an impermeable slurry (bentonite)
mixture. Grouting is a process where an impermeable mixture
is either injected into the ground or is pumped into a series of
interconnected boreholes which together form an impermeable
boundary. Again, the main feature of these techniques is to
physically isolate the DNAPL.
In summary, site characterization and remediation options for
sites containing DNAPL are limited. Field data from site
characterization and remediation efforts are also limited. This
is largely due to the complexity of DNAPL transport and fate in
the subsurface, poorly developed techniques currently
available to observe and predict DNAPL in the subsurface, and
to the fact that this issue has not been widely recognized until
recently. Clearly, there is a growing realization within the
scientific and regulatory community that DNAPL is a significant
factor in limiting site remediation. Correspondingly, current
research efforts within the private, industrial, and public sectors
are focusing on both the fundamentals and applications
aspects of DNAPL behavior in subsurface systems.
Additionally, the number of field investigations reflecting an
increased awareness of DNAPLs, is growing.
DNAPL Modeling
A modeling overview report identified nineteen (numeric and
analytic) multiphase fbw models which are currently available
(60). Most of these models were developed for salt water
intrusion, LNAPL transport, and heat flow. Four models are
qualitatively described as immiscible flow models but do not
specifically indicate DNAPL. A more recent model has been
developed which simulates density driven, three phase flow,
that is capable of modeling DNAPL transport (23). Presently,
very little information is available on DNAPL modeling in the
scientific literature.
Multiphase flow modeling involves modeling systems where
more than one continuous fluid phase (NAPL, water, gaseous)
is present. Modeling any subsurface system requires a
conceptual understanding of the chemical, physical, and
biological processes occurring at the site. Modeling of
simultaneous flow of more than one fluid phase requires a
conceptual understanding of the fluids and the relationship
between the fluid phases. The significance of multiphase flow
over single phase fbw is the increased complexity of fluid flow
and the additional data requirements necessary for modeling.
As presented earlier, numerous variables strongly influence
DNAPL transport and fate, and consequently, the
mathematical relationship of these variables is complex.
Therefore, it follows that DNAPL modeling presents paramount
technical challenges.
Presently, it is exceedingly difficult to obtain accurate field data
which quantitatively describes DNAPL transport and fate
variables within reasonable economic constraints. DNAPL
transport is highly sensitive to subsurface heterogeneities
(8,27,28) ^hich compounds the complexity of modeling.
Heterogeneities are, by nature, difficult to identify and quantify
and models are not well equipped to accommodate the
influence of heterogeneities. Additbnally, relative permeability
and capillary pressure functions must be quantified to identify
the relationship between fluids and between the fluids and the
porous media. Unfortunately, these parameters are very
difficult to measure, particularly in three phase systems. Prior
to an investment of time and money to model a given site, a
careful evaluation of the specific objectives and the confidence
of the input and anticipated output data should be performed.
This will help illuminate the costs, benefits, and therefore, the
relative value of modeling in the Superfund decision making
process.
In summary, DNAPL modeling at Superfund sites is presently
of limited use. This is mainly due to: the fact that very little
information is available in the scientific literature to evaluate
previous work; accurate and quantitative input data is expected
to be costly; the sensitivity of DNAPL transport to subsurface
heterogeneities; and, the difficulty in defining the
heterogeneities in the field and reflecting those in a model.
However, multiphase fbw models are valuable as learning
tools.
References
1. Baehr, A.L, Selective Transport of Hydrocarbons in the
Unsaturated Zone Due to Aqueous and Vapor Phase
Partitioning, Water Resources Research. Vol. 23, No. 10,
pp. 1926-1938, 1987.
2. Bear, J., 1972, Dynamics of Fluids in Porous Media.
American Elsevier Publishing Co., New York, 763 p.
3. Bouchard, D., Contaminant Transport in the Subsurface;
Sorption Equilibrium and the Role of Nonaqueous Phase
Liquids, in, Intermedia Pollutant Transport; and Field
Measurement. (David T. Allen, Yoram Cohen and Isaac
R. Kaplan, Eds.), New York, Plenum Pub. Corp., pp. 189-
211.
4. Bower, H„ Groundwater Hvdroloav. McGraw-Hill Book
Co., 1978, 480 pp.
5. Canter, L. W. and R. C. Knox, Ground Water Pollution
Control. Lewis Publishers Inc., Chelsea, Mich., 1986, 526
PP-
6. Chatzis, I., M.S. Kuntamukkuia, and N.R. Morrow, Blob-
size Distributbn as a Function of Capillary Number in
Sandstones, Paper 13213, Presented at SPE Annual
Tech. Conference and Exhibition. Houston, TX, 1984.
18
-------
7. Cherry, J.A., S. Feenstra, B.H. Kueper and D.W.
McWhorter, "Status of In Situ Technologies for Cleanup
of Aquifers Contaminated by DNAPL's Below the Water
Table," in, International Specialty Conference on How
Clean is Clean? Cleanup Criteria for Contaminated Soil
and Groundwater. Air and Waste Management
Association, pp. 1-18, November 6-9, 1990.
8. Connor, J.A., C.J. Newell, D.K. Wilson, Assessment,
Field Testing, Conceptual Design for Managing Dense
Nonaqueous Phase Liquids (DNAPL) at a Superfund
Site, in. Proceedings of Petroleum Hydrocarbons and
Organic Chemicals in Ground Water: Prevention.
Detection, and Restoration. A Conference and
Exposition, The Westin Galleria, Houston, TX, Vol. 1, pp.
519-533, 1989.
9. DiGiulio, D.C. and J.S. Cho, Conducting Field Tests for
Evaluation of Soil Vacuum Extraction Application, in,
Proceedings of the Fourth National Outdoor Action
Conference on Aouifer Restoration. Ground Water
Monitoring, and Geophysical Methods. Las Vegas, NV,
May 14-17, 1990, pp. 587-601.
10. Feenstra, S., Evaluation of Multi-Component DNAPL
Sources by Monitoring of Dissolved-Phase
Concentrations, in, Proceedings of the Conference On
Subsurface Contamination bv Immiscible Fluids.
International Association of Hydrogeologists, Calgary,
Alberta, April 18-20, 1990.
11. Ferry, J.P. and P.J. Dougherty, Occurrence and
Recovery of a DNAPL in a Low-Yielding Bedrock Aquifer,
in, Proceedings of Ihe NWWA/API Conference on
Petroleum Hydrocarbons and Organic Chemicals in
Ground Water - Prevention. Detection and Restoration.
Nov. 12-14, Houston, TX., 1986, pp. 722-733.
12. Fu, J.K. and R.G. Luthy, Effect of Organic Solvent on
Sorption of Aromatic Solutes onto Soils, Journal of
Environmental Engineering. Vol. 112, No. 2, pp. 346-366,
1986.
13. Glass, R.J., T.S. Steenhuis, and J.Y. Parlange,
Mechanism for Finger Persistence in Homogeneous
Unsaturated Porous Media: Theory and Verification, Soil
Science. 148(1), pp. 60-70, 1989.
14. Hall, A.C., S.H. Collins, and J.C. Melrose, Stability of
Aqueous Wetting Films, Society of Petroleum
Engineering Journal. 23(2), pp. 249-258, 1983.
15. Hall, R.A., S.B. Blake, and S.C. Champlin, Jr.,
Determination of Hydrocarbon Thickness in Sediments
Using Borehole Data, in, Proceedings of the 4th National
Symposium and Exposition on Aouifer Restoration and
Ground Water Monitoring. Columbus, OH, pp. 300-304,
May 23-25, 1984.
16. Hampton, D.R., and P.D.G. Miller, Laboratory
Investigation of the Relationship Between Actual and
Apparent Product Thickness in Sands, in, Proceedings of
Petroleum Hydrocarbons and Organic Chemicals in
Ground Water: Prevention. Detection, and Restoration. A
Conference and Exposition, The Westin Galleria,
Houston, Texas, Vol. 1, pp. 157-181, November 9-11,
1988.
17. Hinchee, R.E., D.C. Downey, R.R. Dupont, P. Aggarwal,
and R.N. Miler, Enhancing Biodegradation of Petroleum
Hydrocarbon Through Soil Venting, Journal of Hazardous
Materials, (accepted) 1990.
18. Hinchee, R.E. and H.J. Reisinger, A Practical Application
of Multiphase Transport Theory to Ground Water
Contamination Problems, Ground Water Monitoring
Review, pp. 84-92, Winter, 1987.
19. Hoag, G.E. and M.C. Marley, Gasoline Residual
Saturation in Unsaturated Uniform Aquifer Materials,
Journal of Environmental Engineering. Vol. 112, No. 3,
pp. 586-604, 1989.
20. Hutzler, N.J., B.E. Murphy, and J.S. Gierke, Review of
Soil Vapor Extraction System Technology, Presented at
Soil Vapor Extraction Technology Workshop. June 28-29,
1989, Edison, New Jersey.
21. Johnson, L.A. and F.D. Guffey, "Contained Recovery of
Oily Wastes, Annual Progress Report." Western
Research Institute, Laramie, Wyoming, June, 1989.
22. Johnson, P.C., C.C. Stanley, M.W. Kemblowski, D.L.
Byers, and J.D. Colthart, A Practical Approach to the
Design, Operation, and Monitoring of In Situ Soil-Venting
Systems, Ground Water Monitoring Review, pp. 159-178,
Spring 1990.
23. Katyal, A.K., J.J Kaluarachchi, and J.C. Parker, MOFAT:
A Two-Dimensional Finite Element Program for
Multiphase Flow and Multicomponent Transport.
Program Documentation, Version 2.0, Virginia
PolyTechnic Institute and State University, 58 pp.,
August, 1990.
24. Kemblowski, M.W. and C.Y. Chiang, Analysis of the
Measured Free Product Thickness in Dynamic Aquifers,
in. Proceedings of Petroleum Hydrocarbons and Organic
Chemicals in Ground Water: Prevention. Detection, and
Restoration. A Conference and Exposition, The Westin
Galleria, Houston, Texas, Vol. 1, pp. 183-205, November
9-11, 1988.
25. Kemblowski, M.W. and C.Y. Chiang, Hydrocarbon
Thickness Fluctuations in Monitoring Wells, Ground
Water. Vol. 28, No. 2, pp. 244-252, 1990.
26. Kerfoot, H.B., Is Soil-Gas Analysis an Effective Means of
Tracking Contaminant Plumes in Ground Water? What
are the Limitations of the Technology Currently
Employed? Ground Water Monitoring Review, pp. 54-57,
Spring 1988.
27. Kueper, B.H. and E.O. Frind, An Overview of Immiscible
Fingering in Porous Media, Journal of Contaminant
Hydrology. Vol. 2, pp. 95-110, 1988.
28. Kueper, B.H., W. Abbott, and G. Farquhar, Experimental
Observations of Multiphase Flow in Heterogeneous
Porous Media, Journal of Contaminant Hvdroloov. Vol. 5,
pp. 83-95,1989.
19
-------
29. Lee, M.D., J.M. Thomas, R.C. Borden, P.B. Bedient, J.T.
Wilson, and C.H. Ward, Biorestoration of Aquifers
Contaminated with Organic Compounds, National Center
for Ground Water Research, CRC Critical Reviews in
Environmental Control. Vol. 18, Issue 1, pp. 29-89, 1988.
30. Undeburg, M.R., 1986, Civil Engineering Reference
Manual. 4th edition, Professional Publications Inc.
Belmont, CA.
31. Lyman, W.J., W.F. Reehl, and D.H. Rosenblatt,
Handbook of Chemical Property Estimation Methods.
McGraw-Hill Book Company, 1982.
32. Mackay, D.M. and J.A. Cherry, Ground-Water
Contamination: Pump and Treat Remediation,
Environmental Science & Technology. Vol. 23, No. 6, pp.
630-636, 1989.
33. Marrin, D.L. and G.M. Thompson, Gaseous Behavior of
TCE Overlying a Contaminated Aquifer, Ground Water.
Vol.25, No. 1,pp. 21-27,1987.
34. Marrin, D., Kerfoot, H, Soil-gas surveying techniques
Environmental Science & Technology. Vol. 22, No. 7, pp.
740-745, 1988.
35. Marrin, D.L., Soil-Gas Sampling and Misinterpretation,
Ground Water Monitoring Review, pp. 51-54, Spring
1988.
36. Mercer, J.W. and R.M. Cohen, A Review of Immiscible
Fluids in the Subsurface: Properties, Models,
Characterization and Remediation, Journal of
Contaminant Hvdroloav. Vol. 6, pp. 107-163,1990.
37. Mississippi Forest Products Laboratory. Proceedings of
the Bioremediation of Wood Treating Waste Forum.
Mississippi State University, March 14-15,1989.
38. Morrow, N.R., Interplay of Capillary, Viscous and
Bouyancy Forces in the Mobilization of Residual Oil, The
Journal of Canadian Petroleum. Vol. 18, No. 3, pp. 35-46,
1979.
39. Ng, K.M., H.T. Davis, and L.E. Scriven, Visualization of
Blob Mechanics in Flow Through Porous Media,
Chemical Engineering Science. Vol. 33, pp. 1009-1017,
1978.
40. Patterson, R.J., S.K. Frape, L.S. Dykes, and R.A.
McLeod, A Coring and Squeezing Technique for the
Detailed Study of Subsurface Water Chemistry,
Canadian Journal Earth Science. Vol. 15, pp. 162-169,
1978.
41. Sale, T., CH2M Hill, and Kuhn, B., Recovery of Wood-
Treating Oil from an Alluvial Aquifer Using Dual
Drainlines, in, Proceedings of Petroleum Hydrocarbons
and Organic Chemicals in Ground Water: Prevention-
Detection. and Restoration. A Conference and
Exposition, The Westin Galleria, Houston, Texas, Vol. 1,
pp. 419-442, November 9- 11,1988.
42. Sale, T„ K. Pbntek, and M. Pitts, Chemically Enhanced
In-Situ Soil Washing, in Proceedings of the Conference
on Petroleum Hydrocarbons and Organic Chemicals in
Ground Water: Prevention. Detection, and Restoration.
Houston, TX, November 15-17, 1989.
43. Schmidtke, K., E. McBean, and F. Rovers, Drawdown
Impacts in Dense Non-Aqueous Phase Liquids, in
NWWA Ground Water Monitoring Symposium. Las
Vegas, Nevada, pp. 39-51, May, 1987.
44. Schmidtke, K„ E. McBean, and F. Rovers, Evaluation of
Collection Well Parameters for DNAPL, Journal of
Environmental Engineering, accepted, August, 1990.
45. Schwille, F., Groundwater Pollution in Porous Media by
Fluids Immiscible With Water, The Science of the Total
Environment. Vol. 21, pp. 173-185, 1981.
46. Schwille, F„ Migration of Organic Fluids Immiscible with
Water in the Unsaturated Zone, in, Pollutants in Porous
Media: The Unsaturated Zone Between Soil Surface and
Groundwater. (B. Yaron, G. Dagan, J. Goldshmid, Eds.)
Springer-Verlag, New York, pp. 27-48,1984.
47. Schwille, F.. Dense Chlorinated Solvents in Porous and
Fractured Media: Model Experiments (English
Translation), Lewis Publishers, Ann Arbor, Ml 1988.
48. Seitz, W.R., In-Situ Detection of Contaminant Plumes in
Ground Water, Special Report 90-27, U.S. Army Corps of
Engineers, Cold Regions Research & Engineering
Laboratory, August 1990, 12 pp.
49. Silka, L, Simulation of Vapor Transport Through the
Unsaturated Zone - Interpretation of Soil-Gas Surveys,
Ground Water Monitoring Review, pp. 115-123, Spring
1988.
50. Sitar, N., J.R. Hunt, and K.S. Udell, Movement of
Nonaqueous Liquids in Groundwater, in, Proceedings of
a Speciality Conference. Geotechnical Practice tor Waste
Disposal '87. University of Michigan, Ann Arbor, Ml, pp.
205-223, June 15-17, 1987.
51. Sims, R., Soil Remediation Techniques at Uncontrolled
Hazardous Waste Sites. Air & Waste Management
Association. Vol. 40, No. 5, pp. 704-732, May 1990.
52. Thompson, G., and Marrin, D., Soil Gas Contaminant
Investigations: A Dynamic Approach, Ground Water
Monitoring Review, pp. 88-93, Summer, 1987.
53. Treiber, L.E., D.L. Archer, and W.W. Owens, A
Laboratory Evaluation of Wettability of Fifty Oil-Producing
Reservoirs, Society of Petroleum Engineering Journal.
12(6), 531-540.
54. Tuck, D.M., P.R. Jaffe, and D.A. Crerar, Enhancing
Recovery of Immobile Residual Non-Wetting
Hydrocarbons from the Unsaturated Zone Using
Surfactant Solutions, in, Proceedings of Petroleum
Hydrocarbons and Organic Chemicals in Ground Water:
Prevention. Detection, and Restoration. A Conference
20
-------
and Exposition, The Westin Galleria, Houston, Texas,
Vol. 1, pp. 457-478, November 9-11,1988.
55. U.S. EPA,, Treatment of Contaminated Soils with
Aqueous Surfactants, EPA/600/2-85/129, NTIS PB86-
122561, 84 pp., 1985.
56. U.S. EPA, Handbook Remedial Action at Waste Disposal
Sites, EPA/625/6-85/006, October, 1985.
57. U.S. EPA, RCRA Ground-Water Monitoring Technical
Enforcement Guidance Document (TEGD), OSWER
Directive 9950.1, 1986c.
58. U.S. EPA, Microbial Decomposition of Chlorinated
Aromatic Compounds, EPA/600/2-86/090, September
1986.
59. U.S. EPA, Characterization and Laboratory Soil
Treatability Studies for Creosote and Pentachlorophenol
Sludges and Contaminated Soil, EPA/600/2-88/055 or
NTIS Publication #PB89-109920, 138 p., 1988.
60. U.S. EPA, Ground Water Modeling: An Overview and
Status Report, EPA/600/2-89/028, December, 1988.
61. U.S. EPA, Contaminant Transport in Fractured Media:
Models for Decision Makers, EPA/600/SF-88/002,
October, 1988.
62. U.S. EPA, Characterization and Laboratory Soil
Treatability Studies for Creosote and Pentachlorophenol
Sludges and Contaminated Soil, EPA/600/2-88/055,
September 1988.
63. U.S. EPA, Bioremediation of Contaminated Surface
Soils, EPA-600/9-89/073, 23 pp.. August 1989.
64. U.S. EPA, Performance Evaluations Of Pump And Treat
Remediations, Superfund Ground Water Issue, EPA/540/
4-89/005, 19 pp.,1989.
65. U.S. EPA, Subsurface Contamination Reference Guide,
EPA/540/2- 90/011, October, 1990.
66. U.S. EPA, Laboratory Investigation of Residual Liquid
Organics from Spills, Leaks, and Disposal of Hazardous
Wastes in Groundwater, EPA/600/6-90/004, April, 1990.
67. U.S. EPA, Basics of Pump and Treat Ground Water
Remediation Technology, EPA-600/8-90/003, 31 pp.,
March 1990.
68. U.S. EPA, Site Characterizations for Subsurface
Remediations, EPA/625/ - /(in press) 1990.
69. U.S. Federal Highway Administration, Guidelines for
Cone Penetration Test: Performance and Design,
FHWA-T5-78-209 (TS 78 No. 209) February, 1977.
70. Villaume, J.F., P.C. Lowe, and D.F. Unites, Recovery of
Coal Gasification Wastes: An Innovative Approach, in,
Proceedings Third National Symposium on Aouifer
Restoration and Ground Water Monitoring. National
Water Well Association, Worthington, OH, pp. 434-445,
1983.
71. Villaume, J.F., Investigations at Sites Contaminated with
Dense, Non-Aqueous Phase Liquids (NAPLs), Ground
Water Monitoring Review. Vol. 5, No. 2, pp. 60-74,1985.
72. Waterloo Centre for Ground Water Research, University
of Waterloo Short Course, "Dense Immiscible Phase
Liquid Contaminants in Porous and Fractured Media,"
Kitchener, Ontario, Canada, Nov. 6-9, 1989.
73. Williams, D.E. and D.G. Wilder, Gasoline Pollution of a
Ground- Water Reservoir - A Case History, Ground
Water. Vol. 9, No. 6, pp. 50- 54, 1971.
74. Wilson, J.L. and S.H. Conrad, Is Physical Displacement
of Residual Hydrocarbons a Realistic Possibility in
Aquifer Restoration?, in, Proceedings of the NWWA/API
Conference on Petroleum Hydrocarbons and Organic
Chemicals in Ground Water—Prevention. Detection, and
Restoration. The Intercontinental Hotel, Houston, Texas,
pp. 274-298, November 5-7,1984.
75. Wisniewski, G.M., G.P. Lennori, J.F. Villaume, and C.L.
Young, Response of a Dense Fluid Under Pumping
Stress, in, Proceedings of the 17th Mid-Atlantic Industrial
Waste Conference. Lehigh, University, pp. 226-237,
1985.
21
~ U.S. GOVERNMENT PRINTING OFTICE- 1991 - MH-IK7/256I3
-------
United States Center for Environmental Research
Environmental Protection Information BULK RATE
Agency Cincinnati OH 45268 POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use, S300
Please make all necessary changes on the above label
detach or copy and return to the address in the upper
left-hand corner
It you oo not wish to receive these reports CHECK HERE u,
detach, or copy this cover and return to the address m the
upper left hand corner
EPA/540/4-91/002
-------
United States
Environmental Protection
Agency
Research and Development
Robert S. Kerr Environmental
Research Laboratory
Ada, OK 74820
EPA/600/S-93/004 May 1993
&ERA ENVIRONMENTAL
RESEARCH BRIEF
COMPLEX MIXTURES AND GROUNDWATER QUALITY
M.L. Brusseau*
INTRODUCTION
The occurrence of organic chemicals in soil and groundwater
has become an issue of great interest and import. Concomitantly,
research on the transport and fate of organic contaminants in
subsurface environments has expanded greatly in recent years.
Much of this research has been focused on dissolved
constituents in aqueous systems. However, the behavior of
"complex mixtures" is beginning to receive increased attention.
By complex mixture we mean any system other than the simple
system of water containing a single solute. Examples of
pertinent problems involving complex mixtures include the
transport of oxygenated gasoline in the subsurface, the
dissolution of diesel fuel and coal-tar, and the use of chemical
agents such as surfactants or solvents to enhance the removal
of contaminants by pump-and-treat remediation. A discussion
of these few selected examples will serve to highlight some of
the issues associated with complex mixtures, with a focus on
potential groundwater contamination and remediation.
COMPLEX MIXTURES AND SUBSURFACE
CONTAMINATION
Miscible Organic Liquids and Alternative Fuels
Concern about air pollution and the dependency on foreign
sources of oil has led to major programs promoting the use of
alternative fuels in the U.S.A. Currently, oxygenates, either
neat or as additives, appear to be the principal alternative fuel
Department of Soil and Water Science
University of Arizona, Tucson, Arizona
candidates (Haggin, 1989). Of the oxygenates, methanol and
ethanol are the primary miscible compounds in use (Hanson,
1991). The advent of alternative fuels has fomented increased
interest in the transport and fate of miscible organic liquids in
the subsurface. It has also increased interest in the effects of
these liquids on the transport and fate of other contaminants.
1. Transport and Fate of Miscible Organic Liquids in
the Subsurface
The sorption of miscible organic liquids by soil is generally
extremely low. Little sorption is expected for compounds such
as methanol and ethanol because of their polarity and large
(infinite) aqueous solubility. The minimal sorption of alcohols
has been widely demonstrated in the chromatography literature.
Limited data forsoil systems has also shown negligible sorption
of alcohols (cf., Garrett etal., 1986; Wood etal., 1990). Hence,
these compounds will be minimally retarded and will travel
through the subsurface at essentially the velocity of water. This
large mobility can be a useful characteristic. For example,
alcohols may be useful as an "early warning" sign of the
impending arrival of a contaminant plume emanating from a fuel
spill. In regard to the use of alcohols for in-situ soil washing, the
greater mobility means that an injected pulse of alcohol may be
able to overtake a plume of a retarded solute.
Alcohols such as methanol have been reported to be
biodegradable under both aerobic and anaerobic conditions
(cf., Colby et al., 1979; Lettinga et al., 1981; Novak et al., 1885).
However, the concentrations of alcohol at which biodegradation
occurred were less than 1 %. Large concentrations (> 10%) of
alcohol are generally considered to be toxic to most
microorganisms and therefore not biodegradable.
Printed on Recycled Paper
-------
2. Effect of Miscible Organic Liquids on the
Subsurface Environment
The addition of a miscible organic liquid, such as methanol, to
water results in a reduction of surface tension. For example,
surface tension is reduced by approximately one-half in systems
containing 5% acetone (Paluch and Rybska, 1991) or 50%
methanol (Wells, 1981). Very large reductions in interfacial
(liquid-liquid) tension are required to mobilize immiscible liquids
trapped in porous media (Puig etal., 1982). Using the surface
tension data as a guide, cosolvents will probably not produce
such large reductions in interfacial tension. Thus, the presence
of a cosolvent is not expected to produce emulsions or to
mobilize residuals of immiscible liquids.
The presence of organic liquids has been shown to cause
shrinking of clay materials and of soils consisting of large
portions of clay. For example, clay materials have been
demonstrated to shrink (in relation to status in aqueous system)
with the addition of acetone or ethanol (Green et al., 1983;
Brown and Thomas, 1987; Chen et al., 1987). This shrinkage
can result in an increase in hydraulic conductivity (Brown and
Thomas, 1987). Thus, it is possible that the presence of large
concentrations of cosolvent could cause shrinking and cracking
of subsurface domains containing large fractions of clay. This
perturbation may alter the hydraulic conductivity and, thereby,
affect fluid flow and solute transport.
The presence of organic liquids can also affect the properties of
naturally occurring organic components of the soil. It is well
known in polymer science that organic liquids can cause organic
polymers to swell. The degree of swelling is dependent upon
the properties of the solvent (polarity) and of the polymer (type,
structure). The addition of an organic liquid has been shown to
cause natural organic materials to swell (Freeman and Cheung,
1981; Lyon and Rhodes, 1991). One potential effect of the
swelling of organic matter associated with the subsurface solid
phase is a reduction in permeability due to blockage of pores.
Given the relatively small content of organic matter associated
with most subsurface materials, this effect will probably not lead
to a measurable reduction in permeability in most cases.
Another potential effect is the dissolution of components (e.g.,
humic or fulvic acids) from the solid-phase organic matter. A
great deal of research has been reported describing the effect
of dissolved organic matter on the solubility, sorption, and
transport of organic and inorganic compounds. There is a
possibility that large concentrations of cosolvents could extract
organic material from the soil, and that this dissolved organic
matter could affect the transport of contaminants. This effect
will probably be of importance for limited conditions, i.e., for
systems with high organic-carbon content soils and highly
hydrophobic compounds.
Another potential effect of the swelling of organic matter is the
enhanced release of organic compounds (contaminants) residing
in the matrix of organic matter. It is generally accepted that the
organic fraction of soil is the predominant sorbent for low-
polarity organic compounds. It is likely that sorbed organic
compounds reside in internal as well as external domains of the
organic matter. It is quite possible that high concentrations of
cosolvents could enhance the release of organic contaminants
retained within the organic phase. The swelling of the organic
matrix with the addition of a cosolvent allows greater diffusive
mass transfer and, thus, enhances the release of sorbed
compounds (Freeman and Cheung, 1982; Brusseau et al.,
1991 a). This concept is used in analytical chemistry in terms of
solvent extraction of contaminated soils. This is discussed in
more detail in the following section.
As previously mentioned, large concentrations (> 10%) of
alcohol are generally considered to be toxic to most
microorganisms. Hence, it is possible that a release of a fuel
containing large concentrations of alcohol could deleteriously
affect the subsurface biota. The potential effect of large
concentrations of alcohols on microbial communities in the
subsurface appears to have received minimal attention.
3. Effect of Miscible Organic Liquids on the Transport
and Fate of Organic Contaminants in the
Subsurface
The influence of an organic liquid (cosolvent) on the solution-
phase activity of organic compounds is dependent upon the
nature of the solute and of the solvent-cosolvent system. For
many of the systems of environmental interest, water is the
solvent, the cosolvent is less polar than water, and the solutes
are of relatively low polarity. For this case, the addition of a
cosolvent tends to increase the amount of solute that can reside
in solution under equilibrium conditions. A simple relationship
describing the influence of cosolvent on the solubility of a solute
in the mixed-solvent system is the log-linear cosolvency model
(Yalkowsky et al., 1972)
log Sm = log Sw + o fc (1)
where S is the solubility in water (w) and mixed-solvent (m), o
represents the cosolvency power of the cosolvent expressed as
the slope of the solubilization profile (i.e., log solubility versus
fc), and fc is the volume fraction of organic cosolvent.
Given that the sorption of low-polarity organic compounds by
soils, sediments, and aquifer materials ("soil") is considered to
be driven primarily by an entropic, solute-solvent interaction
process, it is expected that the presence of a cosolvent should
significantly affect sorption. A log-linear cosolvency model,
relating the equilibrium sorption constant (K ) to the volume
fraction of cosolvent, for sorption of organic solutes from binary
mixed solvents has been presented in the chromatography and
soil science literature (Dolan etal., 1979; Rao etal., 1985). This
equation is:
iog Kpm = log Kp w - ao fc (2)
where K m and Kp w are the equilibrium sorption constants
(ml g- ')' for the'mixed-solvent and aqueous systems,
respectively, and a is an empirical constant that represents
any deviation of the sorption-fc functionality from that observed
for solubilization. The latter term is generally considered to
represent solvent-sorbent interactions.
The decrease in Kp caused by addition of a cosolvent results in
a reduction in retardation (i.e., retardation factor, R, = 1 +
(p/0)Kp where p and 9 are soil bulk density and volumetric
water content, respectively). The cosolvency effect has
been demonstrated by experiment to cause a decrease in the
sorption and retardation of many organic solutes (cf., Nkedi-
Kizza etal., 1985; 1987; 1989; Fu and Luthy, 1986; Wood etal.,
2
-------
1990; Brusseau et al., 1991a). An example of this effect is
shown in Figure 1, where log Kp values obtained for sorption of
anthracene by a sandy soil are plotted versus volume fraction
of methanol (Figure 1 A). The effect of methanol on the transport
of anthracene in a column packed with the sandy soil is shown
in Figure 1B.
The discussion of the cosolvency effect presented above was
focused on low-polarity organic compounds. A number of
environmentally important compounds, however, are ionizable
acids or bases (e.g., phenols, amines). The impact of organic
cosolvents on the sorption of ionizable organic solutes has
received very little attention to date. The decreases in sorption
of ionizable solutes, present in the neutral form, obtained with
increasing fraction of cosolvent were similar to those observed
for nonionizable solutes (Fu and Luthy, 1986; Leeet al., 1991),
as might be expected. In these cases, however, the system pH
was fixed. The impact of cosolvents on sorption of ionizable
solutes in systems where pH is not controlled is of great
interest, considering the effect organic cosolvents can have on
the pH of the system and on the pKa of the solute. The pKa of
an ionizable solute changes with the composition of the solvent
because of the so-called medium effect, which results from
differences in solvent-solvent and solute-solvent interactions
(cf., Bates, 1969). The pKa value of an organic acid will increase
with increasing fraction of cosolvent (cf., Parsons and Rochester,
1975; Rubinoand Berryhill, 1986), while that of an organic base
will decrease (cf., Gowland and Schmid, 1969). Observe that
for both cases, the shift in pKa promotes formation of the neutral
species. This shift in speciation could significantly affect the
nature and magnitude of sorption.
To illustrate the impact of cosolvent on transport of ionizable
solutes, breakthrough curves obtained for pentafluorobenzoate
in water and methanol systems are compared in Figure 2. Note
that no sorption is observed for the aqueous system and that the
retardation factor is, therefore, 1. No sorption of
pentafluorobenzoate is expected since it is in the anionic form
under the experimental conditions. The fact that sorption is
essentially nonexistent for many organic acids under conditions
typical to the subsurface (pKa«pH; net negative surfaces) has
fomented the use of these organic acids as groundwater tracers.
In contrast, R is greater than 1 for the methanol system. This
change in R would negate the use of pentafluorobenzoate as a
tracer to delineate the velocity of fluid flow. The increase in
retardation with addition of an organic cosolvent has also been
observed for other acids such as dicamba, 2,4-
dichlorophenoxyacetic acid, and chlorophenols (Hassett et al.,
1981; Brusseau, 1990; Lee et al., 1993). This phenomenon
may be important at waste-disposal sites, where ionogenic
chemicals may co-exist with organic solvents.
In comparison to the amount of research devoted to the effect
of cosolvents on solubility and equilibrium sorption of organic
contaminants, there has been little work reported on the impact
of cosolvents on nonequilibrium sorption of organic solutes. A
decrease in the asymmetry of breakthrough curves with
increasing volume fraction of cosolvent was reported by Nkedi-
Kizza et al., (1987), who were investigating the transport of two
herbicides (diuron and atrazine) in columns packed with a
sandy soil. Breakthrough-curve asymmetry, which was
attributed by the authors to nonequilibrium sorption, decreased
with increasing cosolvent content suggesting that the rate of
sorption is greater in the presence of a cosolvent. The sorption
of dioxins by soils from water/methanol mixtures was observed
to be more rapid at higher methanol contents (Walters and
Guiseppi-Elie, 1988). The desorption rate constant (k2) has
been observed to increase with increasing fraction of cosolvent
(Nkedi-Kizzaetal., 1989; Shorten and Elzerman, 1990; Brusseau
et al., 1991a; Lee et al., 1991).
1.5
0.5
-0.5 -
-1.5
log Kp = -4.07 fc + 1.635
r* = 0.988
0.3 0.5
Volume Fraction Methanol, f c
0.7
60 80 100
Pore Volume
Figure 1. The influence of methanol on the sorption and transport of anthracene in a Eustis sand; A) the log-iinear relationship between
the equilibrium sorption constant (K ) and volume fraction of cosolvent (fc). B) The influence of cosolvent on the retardation
and transport of anthracene. Data from Brusseau et al.t 1991a.
3
-------
re
v + $ fc
(3)
where k2 m and k2 w are the reverse sorption-rate constants for
the mixed-solvent'and aqueous systems, respectively; 4> = aao;
and a is the slope of the linear relationship between log k2 w and
log K . The validity of this model was substantiatedf'using
experimental data. Examples of their results are presented in
Figure 3. The mechanism responsible for the cosolvency effect
on sorption kinetics was postulated to involve changes in
conformation of the organic carbon associated with the sorbent.
These conformational changes were induced by the changes in
solvent polarity that resulted from the addition of a cosolvent.
The concentrations of cosolvent required to produce substantial
enhancement in solubility and reduction in sorption are relatively
large (% level) for many solutes of interest. Thus, it has been
difficult to envision scenarios wherein cosolvency could be
important. The use of oxygenated and alternative fuels, however,
has presented cases where cosolvency could be very important.
For example, the presence of the cosolvent in alternative fuels
(e.g., 50% methanol, 50%gasoline) could enhancethe transport
of the gasoline constituents contained in the fuel, thus increasing
the potential for groundwater contamination resulting from a
spill. In any case, the effect would probably be limited to the
region near the spill (i.e., the near-field domain).
Immiscible Liquids: Multi-Component Systems,
Dissolution Kinetics, and Transport of Co-Solutes
The disposition of immiscible organic liquids in the subsurface
is of interest to environmental scientists, hydrologists,
environmental/civil engineers, and petroleum engineers. The
vast majority of research performed by these groups has
focused on the movement, entrapment, and displacement of
the liquid (cf., Marie, 1981; Schwille, 1988). This reflects
concerns associated with petroleum-reservoir engineering as
well as remediation of solvent- and petroleum- contaminated
sites. Other aspects that have begun to receive attention are
the dissolution of residual immiscible phases, including the
partitioning behavior of multi-component liquids and the rate of
mass transfer to the aqueous phase, and the effect of immiscible
liquids on the transport of co-solutes.
1. Transport, Entrapment, and Dissolution of
Immiscible Organic Liquids in the Subsurface
The movement, entrapment, and mobilization of immiscible
organic liquids in porous media has been the focus of a
tremendous research effort. Entire volumes have been published
on this subject and there is no need to reproduce this material.
Instead, the dissolution of immiscible organic liquids, a topic
that has received less attention, will be briefly discussed.
Mass transfer of a constituent between two liquids can be
represented by (Cussler, 1984):
acr/at = kr (Krcm - cr)
(4)
where Cr and Cm are the concentrations of the solute in the
residual and aqueous phases, respectively; Kr is the liquid-
liquid partition coefficient; kr is the mass-transfer constant (1 /T);
and t is time (T). The appropriate driving force for mass transfer
is the difference between the actual solute concentration in the
residual phase and that attained at equilibrium (KrCm) (Cussler,
1984). Equation 4 is based on a macroscopic approach and the
mass transfer term is a global parameter. Microscopic
approaches where mass transfer across individual interfaces is
explicitly simulated have also been developed. This latter
approach, however, is constrained by the difficulty of specifying
the nature and magnitude of the interfaces present in the
system.
Consideration of the kinetics of dissolution of residual phases
of immiscible organic liquids is a departure from the majority of
models developed for multi-phase systems, which are based on
instantaneous attainment of equilibrium between residual and
water phases. The results of several laboratory experiments
have suggested that mass transfer between immiscible liquid
and water is relatively rapid (cf., van der Waarden et al., 1971;
Fried et al., 1979; Schwille, 1988; Miller et al., 1990). Other
investigations, however, have shown that liquid-liquid transfer
can be significantly rate-limited, especially under conditions
that may be found in the field (cf., Hunt et al., 1988; Powers et
al., 1991; Brusseau, 1992a). Thus, the use of the equilibrium
assumption for mass transfer in the development of mathematical
models is still open to question. Much additional research is
needed in this area to identify the conditions under which
dissolution will be rate limited and the local equilibrium
assumption is not valid. Liquid-liquid mass transfer in
heterogeneous porous media is of special concern.
2. Partitioning of Multi-Component Liquids
While some of the most widely studied immiscible liquids are
composed of a single component (e.g., trichloroethene), many
others (e.g., gasoline, diesel fuel, coal tar) are multi-component
liquids. Knowledge of the partitioning behavior of multi-
component liquids is essential to the prediction of their impact
on groundwater quality. The partitioning of components into
4
-------
Figure 3. The effect of methanol on the reverse sorption rate coefficient (kj); figure adapted from Brusseau et at., 1991a.
water is controlled by the aqueous solubility of the component
and the composition of the liquid. A simple approach to
estimating partitioning involves an assumption of ideal behavior
in both aqueous and organic phases and the application of
Raoult's law:
Cw, = X°, Sw, (5)
where C™ is aqueous concentration (mol/l) of component i, Sw,
is aqueous solubility (mol/l) of component i, and X° is mole
fraction of component i in the organic liquid. The liquid-liquid
partition coefficient, K,, is given by:
K, = C°l/X°l Sw, (6)
where C° is concentration (mol/l) of i in the organic liquid and
where C°/X° is equivalent to the inverse of the molar volume of
the organic liquid. The Raoult's law-based approach has been
used successfully to predict aqueous-phase concentrations of
compounds (or partition coefficients) for gasoline (Cline et al.,
1991), diesel fuel (Lee et al., 1992a), and coal-tar (Lee et al.,
1992b) systems (see Figure 4). One result of this and other
work (Banerjee, 1984; Picel et al., 1988; Vadas et al., 1991) is
that it appears that many multi-component liquids can be
approximated as ideal mixtures.
3. Effect of Immiscible Liquids on Solute Transport.
The impact of immiscible liquids present as a separate phase
on the sorption and transport of organic solutes was evaluated
by Brusseau (1990). An analysis of experimental data obtained
from systems where an immiscible liquid (e.g., toluene) was the
mobile phase showed that the retardation of organic solutes
(e.g., benzene) was near unity and much lower than that which
would be obtained with water as the solvent. This enhanced
transport by mobile immiscible liquids is to be expected based
upon the relative solubilities of low-polarity organic solutes in
organic liquids and water.
The opposite effect is observed, however, when the immiscible
liquid is present as a fixed residual phase. The residual phase
serves as a sink for organic solutes, resulting in enhanced
retention and retardation. For example, the presence of a
residual phase of aviation gas was observed to increase retention
of petroleum constituents (e.g., toluene) in columns packed
with an aquifer material (Bouchard et al., 1989). The presence
of residual petroleum or PCB oils was shown to increase the
sorption of pentachlorophenol, toluene, and 2-chlorobiphenyl
(Boyd and Sun, 1990). A large increase in retardation of
naphthalene was observed when a residual phase of
tetrachloroethene was emplaced in a column packed with
5
-------
-4.5 -4 -3.5
log [S, moles/L]
Figure 4. Comparison of data obtained from multi-component
partitioning experiments to ideal behavior predicted
by use of Raoult's Law. A) Gasoline system, data from
Cllneetal., 1991; B) Diesel Fuel system, data from Lee
et al., 1992a; C) Coal Tar system, data from Lee et al.,
1992b. S Is aqueous solubility of the compound,
and are the equilibrium partition coefficients
ofthe compounds for distribution between the organic
and aqueous phases.
aquifer material (Brusseau, 1990). A mathematical model
describing the effect of immobile immiscible organic phases on
the transport of solutes was presented by Brusseau (1992a).
The model was used to predict the transport of toluene in a
column packed with an aquifer material contaminated with a
residual of aviation gas (data reported by Bouchard et al.,
1989). The simulated prediction produced with the model
provided a good description of the data (see Figure 5). Based
on these investigations, it appears possible that residual phases
of immiscible organic liquids can serve as long-term sinks and
sources for organic solutes.
When multiple contaminants are present in solution, a primary
question to be addressed is the occurrence of antagonistic or
synergistic interactions among the solutes, and between the
solutes and the solid and aqueous phases. The presence of a
cosolute at high concentrations can affect the behavior of
organic compounds in several ways, resulting in the following
three phenomena: (1) competitive sorption; (2) cooperative
sorption; and (3) cosolvency. The first and third phenomena
reduce sorption and thus enhance the transport of solutes,
whereas cooperative sorption has the opposite effect. A potential
source of these multi-contaminant solutions is the dissolution of
immiscible liquids into water residing in or entering the
subsurface, the relatively slow movement of water in the
subsurface creates the possibility of relatively high solute
concentrations (e.g., near X° Sw, limit) in the vicinity of the
immiscible liquid phases.
Competitive sorption, where sorption of a solute is reduced by
the presence of a co-solute, has been investigated by several
researchers and their results have generally shown no
competition for nonionic, low-polarity organic solutes such as
naphthalene and chlorinated benzenes (cf., Karickhoff et al.,
1979; Chiou etal., 1983). Indeed, non-competition is considered
a defining characteristic of the sorption of nonionic, low-polarity
organic solutes by a "partitioning" mechanism (Chiou et al.,
1983). However, some researchers have reported relatively
small decreases in sorption resulting from competition (Maclntyre
and deFur, 1985; Abdul and Gibson, 1986; McGinley et al.,
1989). The vast majority of studies on sorption in multi-solute
systems have used sorbents with relatively high organic-carbon
contents (i.e., greater than 0.1 %). Conversely, few studies
have been reported for systems comprised of sorbents containing
small organic-carbon contents, which are representative of
many sand aquifers. The sorption of trichloroethene and p-
xylene from single and binary solute solutions by two organic-
carbon-poor aquifer materials was examined by Lee et al.
(1988). They observed no difference in sorption between the
single and binary systems. The sorption of trichloroethene by
a sandy aquifer material in single and ternary solute systems
was observed by Brusseau and Rao (1991) to be essentially
identical.
Cooperative sorption, where sorption of nonionic, low-polarity
organic solutes is enhanced by the presence of other nonionic,
low-polarity organic solutes, has been studied by few
researchers. Brusseau (1991) investigated the effect of a
nonionic, low-polarity cosolute (tetrachloroethene) on the
sorption of three nonionic, low-polarity organic chemicals
(naphthalene, p-xylene, 1,4-dichlorobenzene) by two aquifer
materials with small organic-carbon contents (< 0.03 %). In all
cases, the sorption of the primary solute was enhanced by the
presence of high concentrations of tetrachloroethene.
Equilibrium sorption constants measured in binary-solute
systems were 1.5 to 3 times larger than those measured for the
single-solute systems. Hence, tetrachloroethene had a
synergistic (i.e., cooperative), rather than an antagonistic (i.e.,
competitive), effect on the sorption of the primary solutes. The
enhanced sorption was postulated to result from sorbed
tetrachloroethene increasing the effective organic carbon content
of the sorbent. Enhanced sorption was observed by Onken and
Traina (1991) in recently reported experiments that used
synthetic organo-clay complexes. They examined the sorption
of pyrene by CaCO, treated with humic acid to obtain an organic
carbon content of 0.003%. The sorption of pyrene in a binary
6
-------
P « 50
R = 13.5
01 = 0 1
fe = 014
03 = 0 76
J # \
km0 - 0 36
J #\
k,0 = 30
/ 4 • Tohjene
/ lrom Bouchard el ai (1989)
• ^ — Prediction from MPNEOIL 4 model
• \ ol Brusseau (1990)
2
\—X. I ! ¦ '
20
40 60
Pore Volumes
80
100
COMPLEX MIXTURES AND REMEDIATION OF
CONTAMINATED SOIL AND GROUNDWATER
"Pump-and-treat" is one of the most commonly used techniques
for attempting to remediate contaminated groundwater. In fact,
approximately 68% of Superfund Records of Decision list
pump-and-treat as the primary remediation technique (Travis
and Doty, 1990). Confidence in and popularity of pump-and-
treat is beginning to wane as its effectiveness has proven to be
questionable. In a recent analysis of 19 active or completed
pump-and-treat operations, it was concluded that, although
groundwater extraction is an effective method for containing
plumes, it is not practicable to rely solely on pump-and-treat to
achieve health-based cleanup objectives (Haley et a I., 1991). It
was recommended that methods to enhance extraction
effectiveness and efficiency be considered. In order to design
enhanced removal techniques, the factors responsible for poor
performance of pump-and-treat must be understood.
Figure 5. The effect of a residual phase of immiscible organic
liquid on the transport of toluene in an aquifer material;
figure adapted from Brusseau 1990,1992a.
solution with anthracene as the cosolute was greater than that
measured with no anthracene. The potential for competitive or
cooperative interactions associated with large concentrations
of solute dissolving from immiscible liquids requires further
study, especially for systems consisting of solids with small
organic-carbon contents.
Competitive and cooperative sorption result primarily from
solute-sorbent interactions. In contrast, cosolvency, where the
cosolute is considered a cosolvent, results from solute-solvent
interactions. Interest in cosolvency is focused on the impact of
the cosolvent on the physicochemical properties of water and
the resultant effects on solute behavior in the mixed-solvent
system. The vast majority of research on cosolvency has
involved miscible liquids, as discussed above. It might be
expected that the cosolvency effect of immiscible liquids present
at concentrations below phase separation will generally follow
the behavior of miscible cosolvents, with two major differences.
First, the immiscible liquids are generally of lesser polarity than
are the miscible solvents. By this measure, the immiscible
cosolvent should have a greater effect on the solubility and
sorption of a low-polarity organic solute. However, the lower
polarity of the immiscible cosolvent also limits the amount of
cosolvent that can reside in the aqueous phase. Thus, the
volume fraction of many immiscible cosolvents may be limited
to less than 1%. These small volume fractions may not be
sufficient to induce a significant cosolvency effect.
The cosolvency of water-immiscible liquids was investigated by
Pinal etal. (1990) and Raoetal. (1990). They found that, while
the impact of immiscible cosolvents on solubility and sorption of
hydrophobic organic solutes depended upon the polarity of the
cosolvent, the general trends were similar to those observed for
miscible cosolvents. However, for some immiscible liquids, the
presence of a miscible cosolvent was required to enhance the
solubility of the immiscible liquid to levels such thatthe immiscible
cosolvent had an appreciable cosolvency effect. Much more
research is needed on the potential cosolvency effect of solutes
dissolving from immiscible liquids.
Two phenomena relating to poor performance of pump-and-
treat systems have been observed at many sites. The first is the
so-called "tailing" phenomenon, wherein the rate of reduction in
contaminant concentration in water declines greatly after a
relatively short phase of rapid reduction. This behavior results
in an asymptotic concentration-time profile and greatly delayed
cleanup times. The second phenomenon has been popularly
termed "rebound" and is characterized by an increase in
contaminant concentration after cessation of pumping. Both of
these phenomena greatly reduce the efficacy of pump-and-
treat remediation systems.
Factors Influencing the Efficacy of Pump-and-Treat
Remediation
The tailing and rebound phenomena discussed above are
indicative of nonideal contaminant transport. The fact that
transport nonideality can have a significant impact on the
effectiveness of pump-and-treat remediation is just beginning
to be acknowledged (cf., Hall, 1988; Brusseau and Rao, 1989;
Keely, 1989; MacKay and Cherry, 1989). Of primary concern
for this technique is the removal efficiency associated with a
given pumping regime or, in other words, the amount of time
and water required to flush the aquifer to a specified contaminant-
concentration level. The concentration/time function is sensitive
to nonideal transport. In general, most nonideality factors will
increase the time and the volume of water required to effect
remediation.
Some of the major factors that can cause nonideal transport are
briefly discussed.
1. Flow in heterogeneous porous media: Aquifers are
heterogeneous in nature; hydraulic conductivity and sorption
capacity are generally the two most significant properties.
The hydrodynamics of fluid flow in heterogeneous systems
causes nonideal solute transport. For example, the
existence of low-conductivity media (e.g., silt/clay lenses)
within a sandy aquifer creates domains through which
advective flow and transport are minimal in comparison to
the surrounding sand. Contaminant associated with the
silt/clay lenses, orthe"non-advective" domain, is released
to flowing groundwater primarily by pore-water diffusion.
Increasing the flow rate can increase the state of
7
-------
disequilibrium between the advective and non-advective
domains and result in delayed removal (i.e., "tailing"). The
effects of variable velocity fields caused by hydraulic
conductivity heterogeneity can also be caused by sorption
capacity variability.
2. Sorption/desorption kinetics: Recent research has revealed
that adsorption/desorption of organic solutes by aquifer
materials can be significantly rate limited (Lee et al., 1988;
Ball and Roberts, 1991; Brusseau and Reid, 1991; Brusseau
et al., 1991b). The rate-limiting mechanism apparently
involves constrained diffusion within the sorbent matrix
(Ball and Roberts, 1991; Brusseau et al., 1991c). The
validity of the local equilibrium assumption is dependent, in
part, upon the hydrodynamic residence time of the
contaminant in the system, which is a function of, among
other factors, pore-water velocity. Increasing the velocity,
as is done in pump-and-treat, can cause or enhance
nonequilibrium conditions as a result of reduced residence
time. Nonequilibrium will produce aqueous-phase
concentration values lower than those obtained under
ideal, equilibrium conditions. Thus, tailing will occur and
removal by flushing will take longer.
3. Immiscible liquid dissolution kinetics: In many cases,
residual phases of immiscible organic liquids may exist in
portions of the contaminated subsurface. It has been
shown that very large pore-water velocities (i.e., hydraulic
gradients) are required to displace residual saturation
(Wilson and Conrad, 1984; Willhite, 1986; Hunt et al.,
1988). Hence, the primary means of removal will be
dissolution into water and volatilization into the soil
atmosphere. The immiscible liquid, therefore, serves as a
long-term source of contaminant. As discussed above, the
dissolution of immiscible liquid into water may be rate
limited and, in such cases, would be dependent upon pore-
water velocity. Increased velocity would enhance
nonequilibrium conditions and, thus, result in tailing and
delayed removal.
4. Contaminant Aging: Recent research has shown that
contaminants that have been in contact with porous media for
long times are much more resistant to desorption, extraction,
and degradation. For example, contaminated soil samples
taken from field sites exhibit solid:aqueous distribution ratios
that are much larger than those measured or estimated based
on spiking the porous media with the same contaminant (e.g.,
adding contaminant to uncontaminated sample) (Steinberg et
al., 1987; Pignatello etal., 1990; Smith etal., 1990; Scribner
et al. 1992). In addition, the desorption rate coefficients
determined for previously contaminated media collected from
the field have been shown to be much smaller than the values
obtained for spiked samples (Steinberg et al., 1987). These
field-based observations are supported by laboratory
experiments that show desorption rate coefficients to decrease
with increasing time of contact prior to desorption (Karickhoff,
1980; McCall and Agin, 1985; Coates and Elzerman, 1986;
Brusseau et al., 1991c). These phenomena are significant
not only because of the delayed removal they can cause, but
also because the aged contaminants appear to be highly
resistant to degradative processes (cf., Steinberg etal., 1987;
Scribneretal., 1992). Thus, these aged contaminant residues
may be resistant to remediation, except perhaps by use of an
enhancement technique.
5. Other Factors: Otherfactors, such as nonuniform flowpaths
and stagnation zones, can contribute to observed nonideal
phenomena such as tailing during a pump-and-treat
remediation. The effects of these factors are, however,
much more a function of well-field dynamics than
contaminant-media interactions and, as such, would not be
affected by chemical enhancements.
It is apparent from the above discussion that several factors
influencing contaminant transport can have deleterious effects
on the efficacy of pump-and-treat remediation. These effects
can create conditions where the expected, desirable result of
large decreases in remediation time is not obtained when
pumping is initiated or increased. These factors must be
considered when designing pump-and-treat remediation
systems.
Unfortunately, there has been very little quantitative analysis of
the impact of nonideal transport on aquifer flushing. An example
taken from one of the few such analyses is presented in Figure
6 (adapted from Brusseau, 1993). The data presented in the
figure were obtained from a pilot-scale aquifer flushing system
wherein a two-well injection-withdrawal couplet was used to
evaluate the effect of injecting clean water into a contaminated
aquifer (Whiffen and Bahr, 1984). These, as well as other, data
were used by Brusseau (1992b) to evaluate the ability of a multi-
factor nonideality model to predict field-scale solute transport.
The data were subsequently used to quantitatively evaluate the
effect of porous-media heterogeneity and nonequilibrium
sorption on the effectiveness of pump-and-treat (Brusseau,
1993). The predicted removal curve for the case of uniform
aquifer properties and instantaneous sorption/desorption is
shown in Figure 6. It is evident that the prediction greatly
underestimates the volume of water required to remove the
contaminant. The predicted simulation obtained for the case of
variable hydraulic conductivity and rate-limited sorption/
desorption matches the field data extremely well (see Figure 6).
A comparison of this prediction to the one obtained for ideal
conditions clearly illustrates the effect that nonideal transport
factors can have on aquifer flushing.
The predicted removal of contaminant for the case of spatially
variable hydraulic conductivity and instantaneous sorption/
desorption is also shown in Figure 6. While this prediction does
not match the early field observations, at large pore volumes
the simulated curve approaches the curve obtained by including
the combined effects of variable conductivity and rate-limited
sorption/desorption. This suggests that, while both factors
contribute to nonideal transport, spatially variable conductivity
may be the more important factor constraining the efficacy of
aquifer flushing in this system. The knowledge of which factor
or factors is the major cause of nonideal transport is essential
in the design of an effective method for enhancing the efficiency
of a pump-and-treat operation.
Chemical Enhancement of Pump-and-Treat
Remediation
Several chemical-based techniques for enhancing contaminant
removal in the subsurface are under investigation (e.g., addition
of'surfactants, cosolvents, complexing agents), and each has
advantages and disadvantages. A detailed discussion of
chemical enhancement techniques was presented by Palmer
and Fish (1992). However, several aspects relating to the
8
-------
•1 '• x
1 ' 1
J : i
1 • 1
• i
1 i
l : I
• diethylether field data
: i
¦ 1 ' 1
— Predicted simulation
\ i
(variable hydraulic conductivity)
1 ' '
- \ : 1
(rate-limited desorption)
1 ' 1
\ ; 1
Predicted simulation
1 ; 1
(variable hydraulic conductivity)
•k. 1
(instantaneous desorption)
A 1
~~ Predicted simulation
•\i
(uniform hydraulic conductivity)
' l\
(instantaneous desorption)
V
i*.
\ ' ,
\
PORE VOLUMES
Figure 6. The effect of nonideal transport on removal of organic
contaminants from aquifers by flushing. Field data
from Whiffen and Bahr, 1984; model used for
simulations from Brusseauetal. 1989. Figure adapted
from Brusseau 1993.
impact of nonideal transport phenomena on the efficacy of
chemical enhancement were not discussed.
Surfactants are currently the focus of the research effort on
chemical enhancements and, based on preliminary laboratory
data, appear to have promise for enhancing pump-and-treat
remediation in some situations. The use of dissolved organic
matter (DOM) and of cosolvents is also being investigated,
albeit at a smaller scale. Miscible cosolvents, such as methanol,
reduce the net polarity of the mixed solvent when added to
water and thereby increase the quantity of a nonionic organic
compound that can dissolve in the mixed solvent. This increase,
in turn, results in a smaller equilibrium sorption constant and
less attendant retardation. Thus, the addition of a cosolvent
can reduce the volume of water required to flush a contaminant
from porous media by altering the equilibrium phase distribution.
A similar result is obtained with surfactants and DOM, although
by different mechanisms. Hence, surfactants, DOM, and
cosolvents act to increase the aqueous-phase concentration of
organic compounds, the so-called "solubilization" effect. This
effect is of special interest for the removal of residual phases of
immiscible liquids. The other major method of removing trapped
residual phases, mobilization, will not be considered in the
present discussion.
A comparison of the relative degree to which aqueous-phase
concentration of contaminant is enhanced by the various
additives favors the surfactants. However, a comparison of this
type can be very misleading without considering such factors as
potential interactions between the additive and the porous
media. It is well known, for example, that surfactant molecules
(cf., Ducreux et al„ 1990; Kan and Tomson, 1990; Jafvert and
Heath, 1991) and DOM (cf., Dunnivantetal., 1992; Mooreetal.,
1992) can sorb to surfaces of solids, thereby reducing the
concentration of additive available for dissolving the contaminant.
In addition, surfactants and DOM may precipitate under certain
conditions. In contrast, most subsurface solids have a low
affinity for miscible solvents such as methanol. Thus, it is
possible that, whereas the "active" mass of a surfactant or DOM
may be significantly less than the total mass injected into the
subsurface, that of a solvent may be essentially the same.
A comparison of the impact of several potential chemical
additives on the apparent solubility of selected organic
compounds was developed by collecting and synthesizing data
reported in the literature (see Table 1). The effect of sorption
and precipitation of the additives was taken into account. The
results of the analyses are presented in Figure 7a-c. For all
three solutes, the nonionic surfactant (Triton), with low assumed
sorption, produced the greatest enhancement. The cosolvent
(ethanol) produced the lowest degree of enhancement for all
three solutes. The solubilization effect of ethanol increases
dramatically at cosolvent concentrations above those used in
these analyses. It is readily apparent that the relative
enhancement effect will vary by solute, and by other factors
such as the nature of the sorbent. The comparison of the
effectiveness of various additives under a range of conditions
is a topic requiring more research.
The primary criterion upon which chemical enhancement
additives are judged is their solubilization potential. The impact
of interactions between the additive and the solid phase on this
enhancement is an important factor to consider, as discussed
above. However, there are several other factors that should
also be considered when selecting an enhancement agent. In
this regard, cosolvents have several benefits that surfactants
and DOM do not.
First, the addition of a cosolvent increases the magnitude of the
desorption rate coefficient (not to be confused with an increase
in the rate of desorption), thereby reducing the time required to
attain equilibrium. This reduction in the degree of nonequilibrium
would result in reduced tailing during pumping. This, in turn,
would decrease the volume of water and the time required to
remove the contaminant by flushing. As previously discussed,
rate-limited desorption may impose a significant constraint on
the efficacy of pump-and-treat remediation. If so, the ability of
a cosolvent to reduce the degree of nonequilibrium would be a
major attribute. There is no reason to expect surfactants or
DOM to increase desorption rate coefficients.
Second, cosolvents may be able to "extract" the highly retained,
aged contaminants that have been observed in field studies
(see discussion above). There is no reason to suppose that
surfactants or DOM could act in an "extractive" manner.
Conversely, there is good reason to suppose that cosolvents
could enhance the release of aged contaminants, based on the
results of solvent extraction techniques used in the analysis of
contaminated soils (cf., Sawhney et al., 1988) and on the
results of experiments that evaluated the effect of cosolvents on
the desorption of organic compounds (Freeman and Cheung,
1981; Nkedi-Kizza et al., 1989; Brusseau et al., 1991a).
Third, cosolvents may be able to access contaminant that is
residing in low hydraulic-conductivity domains such as clay
lenses. During a pump-and-treat remediation, as discussed
above, contaminant in these domains is probably removed
primarily through diffusion. The clay particles provide a large
surface area with which a surfactant or DOM may interact and
thereby reduce its availability for enhancing contaminant
9
-------
§ ,0
CQ
ts
c
g 8
c
o
2 •
>
4)
CC
Trlchloroethene
A
Etnand
SOS
Humic acKJ
0. q Tnton, R-2
~ ~ Tnton. R«10
E
D
• \am—*
i i i
20 40 60
Additive Concentration (g/l)
100
o
O
©
120
100
60
40
Naphthalene
EthancH
SDS
Hume aod
O" O Tnton, R-2
0 g Trnon, R-10
B
O'
"IS--—
———
20 40 60 60
Additive Concentration (g/l)
100
5000
4000
o
o
©
5 2000
cc
1000
Pyrene
EthanoJ
SOS
Hutnic aod
©¦¦•--© Triton, R»2
~ q Tnton. R-10
.¦O
.O'
~
Etnanoi
1 Ltlr
0
1
20 40 60 80
Additive Concentration (g/l)
100
Figure 7. The effect of several additives on the aqueous-phase
concentration of (A) trlchloroethene, (B) naphthalene,
and (C) pyrene.
removal. In addition, the sorption of the surfactant or DOM can
enhance the retention of the organic solutes by providing an
increase in stationary organic carbon. Surfactant micelles and
larger DOM particles may possibly be excluded from the smaller
pore-size domains, which would limit accessibility. Cosolvents
such as methanol do not sorb significantly to solid surfaces and,
because of their small size, would not be excluded from any
pore domains in which contaminants would be found. In
addition, as discussed above, cosolvents have been found to
cause cracking of clayey materials. This cracking results in
larger permeabilities, which could enhance the rate of
contaminant removal from the lenses. Thus, in comparison to
surfactants and DOM, cosolvents may have a much greater
potential for enhancing the release of contaminants trapped in
fine-grained media.
Fourth, cosolvents have the potential for being used in an
integrated, chemical-biological remediation technique. For
example, methanol is the initial intermediate in the oxidation of
methane by methanotrophic bacteria. The addition of methanol
to the groundwater environment at low concentrations may
stimulate useful cometabolic transformations, causing the
destruction of otherwise refractory contaminants such as
trichloroethene. Under (locally) anaerobic conditions, cosolvent
addition may also drive reductive dehalogenation, particularly
of compounds such as tetrachloroethene (cf., DiStefano et al.,
1991; Gibson and Sewell, 1992). It is possible to envision
situations where addition of cosolvents such as methanol or
ethanol may initiate transformations that result directly or
indirectly in degradation to non-toxic products. The negative
effects of high concentrations of cosolvent on the subsurface
microbial community may initially preclude the development of
biodegradative activity. However, such activity could occur
following dilution of the cosolvent during transport.
Considering the preceding discussion, cosolvents may have
specific properties that make them useful for enhanced pump-
and-treat. However, given these same properties, it is likely
that the use of cosolvents will be limited to smaller scale
problems. The clean-up of near-field contamination problems
is probably where cosolvents can be put to best use.
CONCLUSION
Experience has shown that many soil and groundwater
contamination problems involve complex mixtures of chemicals.
As discussed in this monograph, these mixtures may affect
contaminant behavior through a variety of mechanisms. Because
many of these mechanisms are not well understood, approaches
for dealing with complex mixtures in the subsurface often
involve direct application or untested extrapolation of knowledge
derived from relatively simple aqueous systems. Not surprisingly,
the results are frequently less than satisfactory.
The primary purpose of this paper is to identify and discuss, in
a generic sense, some of the important processes which must
be considered when dealing with complex mixtures in the
subsurface, and to illustrate how these may impact groundwater
quality. From the discussion, it is apparent that complex
mixtures may play a role in groundwater reclamation as well as
degradation of groundwater quality. Equally apparent, however,
is the need for improved scientific understanding of the processes
associated with the transport of complex mixtures and of the
10
-------
Table 1. Enhanced Solubilization Data Collected From the Literature
Additive Compound Sorption of Additive*
SDS
TCE (Shiau etal., 1992)
R = variable (Jafvert and Heath, 1991
SDS
Naphthalene (Gannon et al., 1989)
R = variable "
SDS
Pyrene (Jafvert, 1991)
R = variable
Triton
TCE (West, 1992)
R = 2,10 (Kan and Tomson, 1990)
T riton
Naphthalene (Edwards et al., 1991)
R = 2,10
Triton
Pyrene (Edwards et al., 1991)
R = 2,10
Ethanol
TCE (Morris et al., 1988)
R = 1 (Wood etal., 1990)
Ethanol
Naphthalene (Morris et al., 1988)
R = 1
Ethanol
Pyrene (Morris et al., 1988)
R = 1
Humic Acid
TCE (Garbarini and Lion, 1986)
R = 2 (Dunnivant et al., 1992)
Humic Acid
Naphthalene (McCarthy and Jiminez, 1985)
R = 2
Humic Acid
Pyrene (Gauthier et al., 1987)
R = 2
SDS = sodium dodecyl sulfate; TCE = trichloroethene; Triton = triton X-100; Humic Acid = Aldrich humic acid; "Retardation
factor, R, of an additive in a hypothetical soil was estimated from data reported in the references cited in this column.
influence that chemical mixtures have on the behavior of
specific contaminants.
Disclaimer
The information in this document has been funded in part by the
U.S. Environmental Protection Agency under Cooperative
Agreement No. CR-818757. This document has been subjected
to the Agency's peer review and has been approved for
publication as an EPA document.
Quality Assurance Statement
This project did not involve physical measurements and, as
such, did not require a QA plan.
11
-------
REFERENCES
Abdul, A.S., and Gibson, T.L., Equilibrium batch experiments with
six polycyclic aromatic hydrocarbons and two aquifer materials,
Hazard. Waste Hazard. Mater., 3,125,1986.
Ball, W.P. and Roberts, P.V., Long-term sorption of halogenated
organic chemicals by aquifer materials- Part 2. intraparticle diffusion.
Environ. Sci. Technol., 25, 1237, 1991.
Banerjee, S., Solubility of organic mixtures in water, Environ. Sci.
Technol., 18, 587,1984.
Bates, R.G., Medium effects and pH in nonaqueous solvents,
chap. 2 in: Solute-Solvent Interactions, Coetzee, J.F. and Ritchie,
C.D., eds., Marcel Dekker, New York, N.Y., 1969.
Bouchard, D.C., Enfield, C.G., and Piwoni, M.D., Transport
processes involving organic chemicals, in: Reactions and
Movement of Organic Chemicals in Soils, SSSA Special Publ. No.
22, pp. 349-371, Soil Sci. Soc. Am., Madison, Wl, 1989.
Boyd, S.A., and Sun, S., Residual petroleum and
polychlorinatedbiphenyl oils as sorptive phases for organic
contaminants in soils, Environ. Sci. Technol., 24,142, 1990.
Brown, K.W. and Thomas, J.C., A mechanism by which organic
liquids increase the hydraulic conductivity of compacted clay
materials. Soil Sci. Soc. Amer. J., 51,1452,1987.
Brusseau, M. L., Mass transfer processes and field-scale transport
of organic solutes, pp. 816-840 in: Transport and Mass Exchange
Processes in Sand and Gravel Aquifers: Field and Modelling
Studies, G. Moltyaner, ed., Atom. Energy Canada, Chalk River,
Ontario, Canada, 1990.
Brusseau, M.L., Cooperative sorption of organic chemicals in
systems composed of low organic carbon aquifer materials, Environ.
Sci. Technol., 25,1747,1991.
Brusseau, M.L., Rate-limited mass transfer and transport of organic
solutes in porous media that contain immobile immiscible organic
liquid. Water Resour. Res., 28, 33, 1992a.
Brusseau, M.L., Transport of rate-limited sorbing solutes in
heterogeneous porous media: Application of a onendimensional
multifactornonideality model to field data. Water Resour. Res.,28,
2485,1992b.
Brusseau, M.L., The effect of porous-media heterogeneity and
rate-limiting desorption on pump-and-treat remediation. In:
Proceedings of National Meeting of the American Chemical
Society, Division of Environ. Chem. Vol. 33, No. 1, pp. 65-68.
American Chemical Society, Washington, DC, 1993.
Brusseau, M.L. and Rao, P.S.C, Sorption nonideality during
organic contaminant transport in porous media. CRC Critical
Reviews in Environ. Control, 19, 33,1989.
Brusseau, M.L. and Rao, P.S.C., Influence of sorbate structure on
nonequilibrium sorption of organic compounds, Environ. Sci.
Technol., 25,1501,1991.
Brusseau, M.L. and Reid, M.L., Nonequilibrium sorption of organic
chemicals by low organic-carbon aquifer materials. Chemosphere,
22, 341,1991.
Brusseau, M.L, Jessup.R.E., Rao, P.S.C., Modeling the transport
of solutes influenced by multiprocess nonequilibrium, Water Resour.
Res., 25, 1971, 1989.
Brusseau, M.L., Wood, A.L., and Rao, P.S.C., The influence of
organic cosolvents on the sorption kinetics of hydrophobic organic
chemicals. Environ. Sci. Technol., 25, 903, 1991a.
Brusseau, M.L., Larsen, T„ and Christensen, T.H., Rate-limited
sorption and nonequilibrium transport of organic chemicals in low
organic carbon aquifer materials. Water Resour. Res., 27,1137,
1991b.
Brusseau, M.L., Jessup, R.E., and Rao, P.S.C., Nonequilibrium
sorption of organic chemicals: Elucidation of rate-limiting processes.
Environ. Sci. Technol., 25, 134, 1991c.
Chen, S., Low, P.F., Cushman, J.H., and Roth, C.B., Organic
compound effects on swelling and flocculation of upton
montmorillonite, Soil Sci. Soc. Amer. J., 51,1444,1987.
Chiou, C.T., Porter, P.E., andSchmedding, D.W., Partition equilibria
of nonionic organic compounds between soil organic matter and
water, Environ. Sci. Technol., 17, 227, 1983.
Cline, P.V., Delfino, J.J., and Rao, P.S.C., Partitioning of aromatic
constituents into water from gasoline and other complex solvent
mixtures. Environ. Sci. Technol., 25, 914,1991.
Coates, J.T., and Elzerman, A.W., Desorption kinetics for selected
PCB congeners from river sediments. J. Contam. Hydrol., 1,191,
1986.
Colby, J., Dalton, H., and Whittenburg, R„ Biological and
biochemical aspects of microbial growth on C, compounds, Ann.
Rev. Microbiol.,)33,481,1979.
Cussler, E.L., Diffusion: Mass Transfer in Fluid Systems, Cambridge
Univ. Press, Cambridge, 1984.
DiStefano, T.D., Gosset, J.M., and Zinder, S.H., Reductive
dechlorination of high concentrations of tetrachloroethene to ethene
by an anaerobic enrichment culture in the absence of
methanogenesis. App. Environ. Micro., 57,2287,1991.
Dolan, J.W., Gant, J.R., and Snyder, L.R., Gradient elution in high-
performance liquid chromatography, J. Chromat., 165, 31,1979.
Ducreux, J., Bocard, C., Muntzer, P., Razakarisoa, O., and Zilliox,
L., Mobility of soluble and non-soluble hydrocarbons in
contaminated aquifer, Water Sci. Tech., 22, 27,1990.
Dunnivant, F.M., Jardine, P.M., Taylor, D.L., and McCarthy J.F.,
Transport of naturally occurring dissolved organic carbon in
laboratory columns containing aquifer material, Soil Sci. Soc.
Amer. J., 56, 437,1992.
Edwards, D.A., Luthy, R.G., and Liu, Z., Solubilization of Polycyclic
aromatic hydrocarbons in micellar nonionic surfactant solutions,
Environ. Sci. Technol., 25,127, 1991.
Freeman, D.H. and Cheung, L.W., A gel-partition model for
organic desorption from a pond sediment, Science, 214, 790,
1981.
12
-------
Fried, J.J., Muntzer, P., and Zilliox, L., Ground-water pollution by
transfer of oil hydrocarbons, Groundwater, 17, 586, 1979.
Fu, J., and Luthy, R.G., Effect of organic solvent on sorption of
aromatic solutes onto soils. J. Environ. Engin., 112, 346, 1986.
Gannon, O.K., Bibring, P., Raney, K., Ward, J.A., and Wilson, D.J.,
Soil cleanup by in-situ surfactant flushing. III. Laboratory results,
Separ. Sci. Technol., 24,1073, 1989.
Garbarini, D.R., and Lion, L.W., Influence of the nature of soil
organics on the sorption of toluene and trichloroethylene, Environ.
Sci. Technol., 20, 1263, 1986.
Garrett, P., Moreau, M., and Lowry, J.D., MTBE as a ground water
contaminant, pp. 227-238 in: Petroleum Hydrocarbons and Organic
Chemicals in Ground Water: Prevention, Detection, and
Restoration, Nat. Water Well Assoc., Dublin, OH, 1986.
Gauthier, T.D., Seitz, W.R., and Grant, C.L., Effects of structural
and compositional variations in dissolved humic material on pyrene
Koc values, Environ. Sci. Technol., 21, 243,1987.
Gibson, S.A. and Sewell, G.W., Simulation of reductive
dechlorination of tetrachloroethene in anaerobic aquifer
microcosms by addition of short-chain organic acids or alcohols,
App. Environ. Microbiol., 58, 1392, 1992.
Gowland, J.A. and Schmid, G.H., Two linear correlations of pKa vs.
solvent composition, Canad. J. Chem., 47, 2953-2958,1969.
Green, W.J., Lee, G.F., Jones, R.A., and Palit, T., Interaction of
clay soils with water and organic solvents: Implications for the
disposal of hazardous wastes, Environ. Sci. Technol., 17, 278,
1983.
Haggin, J., Alternative fuels to petroleum gain increased attention.
Chem. Engin. News, 67, 25,1989.
Haley, J.L., Hanson, B., Enfield, C., and Glass, J., Evaluating
the effectiveness of groundwater extraction systems.
Groundwater Monit. Rev., 11, 119, 1991.
Hall, C.W., Practical limits to pump-and-treat technology for
aquifer remediation, Hazard. Mater. Tech. Center Report, 7, 3,
1988.
Hanson, D., Air pollution cleanup: Pact set for reformulating
gasolines. Chem. Engin. News, 69, 4, 1991.
Hassett, J.J., Banwart. W.L., Wood, S.G., and Means, J.C.,
Sorption of -naphthol: Implications concerning the limits of
hydrophobic sorption, Soil Sci. Soc. Amer. J., 45, 38, 1981.
Hunt, J.R., Sitar, N., and Udell, K.S., Nonaqueous phase liquid
transport and cleanup 1. Analysis of mechanisms. Water
Resour. Res., 24, 1247, 1988.
Jafvert, C.T., and Heath, J.K., Sediment- and saturated-soil-
associated reactions involving an anionic surfactant
(dodecylsulfate). 1. Precipitation and micelle formation, Environ.
Sci. Technol., 25, 1031, 1991.
Jafvert, C.T., Sediment-and saturated-soil-associated reactions
involving an anionic surfactant (dodecylsulfate). 2. Partition of
PAH compounds among phases, Environ. Sci. Technol., 25,
1039, 1991.
Kan, A.T., and Tomson, M.B., Groundwater transport of
hydrophobic organic compounds in the presence of dissolved
organic matter, Environ. Toxic. Chem., 9, 253, 1990.
Karickhoff, S.W., Sorption kinetics of hydrophobic pollutants in
natural sediments, pp. 193-205 in: Contaminants and
Sediments, R.A. Baker, ed. Ann Arbor Science, Ann Arbor, Ml,
1980.
Karickhoff, S.W., Brown, D.S., and Scott, T.A., Sorption of
hydrophobic pollutants on natural sediments, Water Res., 13,
241, 1979.
Keely, J.F., Performance evaluations of pump-and-treat
remediations. U.S. E.P.A. Ground Water Issue Paper, Oct.
1989.
Lee, L.S., Rao, P.S.C., Brusseau, M.L., and Ogwada, R.A.,
Nonequilibrium sorption of organic contaminants during flow
through columns of aquifer materials, Environ. Tox. Chem., 7,
779, 1988.
Lee, L.S., Rao, P.S.C., and Brusseau, M.L., Nonequilibrium
sorption and transport of neutral and ionized chlorophenols,
Environ. Sci. Technol., 25, 722, 1991.
Lee, L.S., Hagwell, M., Delfino, J.J., and Rao, P.S.C., Partitioning
of polycyclic aromatic hydrocarbons from diesel fuel into water,
Environ. Sci. Technol., 26, 2104, 1992a.
Lee, L.S., Rao, P.S.C., and Okuda, I., Equilibrium partitioning
of polycyclic aromatic hydrocarbons from coal tar into water,
Environ. Sci. Technol., 26, 2110, 1992b.
Lee, L.S., Bellin, C.A., Pinal, R., and Rao, P.S.C., Cosolvent
effects on sorption of organic acids by soils from mixed-solvents,
Environ. Sci. Technol., 1993 (in press).
Lettinga, G., deZeeuw, W., and Ouborg, E., Anaerobic treatment
of wastes containing methanol and higher alcohols, Water
Res., 15, 171, 1981.
Lyon, W.G. and Rhodes, D.E., The swelling properties of soil
organic matter and their relation to sorption of non-ionic organic
compounds, U.S. Environmental Protection Agency Report EPA/
600/2-911/033, 1991.
Maclntyre, W.G., and deFur, P.O., The effect of hydrocarbon
mixtures on adsorption of substituted naphthalenes by clay and
sediment from water, Chemosphere, 14, 103, 1985.
MacKay, D.M. and Cherry, J.A., Groundwater contamination:
Pump-and-treat remediation. Environ. Sci. Technol., 23,630,1989.
Marie, C.M., Multiphase Flow in Porous Media, Gulf Publ. Co.,
Paris, 1981.
McCall, P.J., and Agin, G.L., Desorption kinetics of picloram as
affected by residence time in soil. Environ. Toxic. Chem., 4, 37,
1985.
13
-------
McCarthy, J.F., and Jiminez, B.D., Interactions between
polycyclic aromatic hydrocarbons and dissolved humic material:
Binding and dissociation, Environ. Sci. Technol., 19, 1072,
1985.
McGinley, P.M., Katz, L.E., and Weber, W.J. Abstract, Amer.
Chem. Soc., Div. of Environ. Chem., Wash. D.C., Vol. 29, pq.
146, 1989.
Miller, C.T., Poirier-McNeill, M.M., and Mayer, A.S., Dissolution
of trapped nonaqueous phase liquids: Mass transfer
characteristics. Water Resour. Res., 26,2783, 1990.
Moore, T.R., de Souza, W., and Koprivnjak, J.F., Controls on
the sorption of dissolved organic carbon by soils, Soil Science,
154,120, 1992.
Morris, K.R., Abramowitz, R., Pinal, R., Davis, P., and Yalkowsky,
S.H., Solubility of aromatic pollutants in mixed solvents,
Chemosphere, 17, 285, 1988.
Nkedi-Kizza, P., Rao, P.S.C., and Hornsby, A.G., The influence
of organic cosolvents on sorption of hydrophobic organic
chemicals by soils. Environ. Sci. Technol., 19, 975, 1985.
Nkedi-Kizza, P., Rao, P.S.C., and Hornsby, A.G., The influence
of organic cosolvents on leaching of hydrophobic organic
chemicals through soils, Environ. Sci. Technol., 21, 1107,
1987.
Nkedi-Kizza, P., Brusseau, M.L., Rao, P.S.C., and Hornsby,
A.G., Nonequilibrium sorption during displacement of
hydrophobic organic contaminants and 45Ca through soil
columns with aqueous and mixed solvents. Environ. Sci.
Technol., 23, 814, 1989.
Novak, J.T., Goldsmith, C.D., Benoit, R.E., and O'Brien, J.H.,
Biodegradation of methanol and tertiary butyl alcohol in
subsurface systems, pp. 71-85 in: Degradation, Retention, and
Dispersion of Pollutants in Ground Water, Inter. Assoc. Water
Pollution Control Research, Great Britain, 1985.
Onken, B.M. and Traina, S.J. Agronomy Abst., Amer. Soc.
Agron., Madison, Wl. 1991.
Palmer, C.D., and Fish, W., Chemical enhancements to pump-
and-treat remediation. Groundwater Issue Paper, EPA/540/S-
92/001, 1992.
Paluch, M. and Rybska, J., Influence of some organic substances
on the surface potential and on the surface tension at the water/
air interface, J. Coll. Inter. Sci., 145, 219, 1991.
Parsons, G.H. and Rochester, C.H., Acid ionization constants
of 4-substituted phenols in methanol+water mixtures, J. Chem.
Soc., Trans. Faraday Soc. I, 71(5), 1058-1068, 1975.
Picel, K.C., Stamoudis, V.C., and Simmons, M.S., Distribution
coefficients for chemical components of a coal-oil/water system,
Water. Res., 22, 1189, 1988.
Pignatello, J.J., Frink, C.R., Marin, P.A., and Droste, E.X.,
Field-observed ethylene dibromide in an aquifer after two
decades. J. Contam. Hydrol., 5, 195,1990.
Pinal R., Rao, P.S.C., Lee, L.S., Cline, P.V., and Yalkowsky,
S.H., Cosolvency of partially miscible organic solvents on the
solubility of hydrophobic organic chemicals, Environ. Sci.
Technol., 24, 639, 1990.
Powers, S.E., Loureiro, C.O., Abriola, L.M., and Weber, W.J.,
Theoretical study of the significance of nonequilibrium dissolution
of nonaqueous phase liquids in subsurface systems, Water
Resour. Res., 27, 463, 1991.
Puig, J.E., Scriven, L.E., Davis, H.T., and Miller, W.G., Fluid
microstructures and enhanced oil recovery, pp. 1 -27 in: Interfacial
Phenomena in Enhanced Oil Recovery, D. Wasan and A.
Payatakes, eds., Amer. Inst. Chem. Eng., New York, NY, 1982.
Rao, P.S.C., Hornsby, A.G., Kilcrease, D.P., and Nkedi-Kizza,
P., Sorption and transport of hydrophobic organic chemicals in
aqueous and mixed solvent systems: Model development and
preliminary evaluation. J. Environ. Qual., 14, 376, 1985.
Rao, P.S.C., Lee, L.S., and Pinal, R., Cosolvency and sorption
of hydrophobic organic chemicals, Environ. Sci. Technol., 24,
647, 1990.
Rubino, J.T. and Berryhill, W.S., Effects of solvent polarity on
the acid dissociation constants of benzoic acids, J. Pharm. Sci.,
75(2), 182-186, 1986.
Sawhney, B.L., Pignatello, J.J., and Steinberg, S.M.,
Determination of 1,2-dibromoethane (EDB) in field soils:
implications for volatile organic compounds, J. Environ. Qual.,
17,149,1988.
Schwille, F., Dense Chlorinated Solvents in Porous and
Fractured Media, Lewis Publishers, Chelsea, Ml, 1988.
Scribner, S.L., Benzing, T.R., Sun, S., and Boyd, S.A., Desorption
and bioavailability of aged simazine residues in soil from a
continuous corn field. J. Environ. Qual., 21,1115, 1992.
Shiau, B.J., Sabatini, D.A., Gupta, A., and Harwell, J.H.,
Enhanced solubilization and mobilization of subsurface DNAPL
contamination using edible surfactants, pp. 266-268. In:
Proceed, of the International Conf. on Subsurface Restoration,
June 21 -24, Dallas, TX, National Center Groundwater Research,
Houston, TX, 1992.
Shorten, C.V., and Elzerman, A.W., Methods for the
determination of PAH desorption kinetics in coal fines and coal
contaminated sediments. Chemosphere, 20, 137, 1990.
Smith, J.A., Chiou, C.T., Kammer, J.A., and Kile, D.E., Effect of
soil moisture on the sorption of trichloroethene vapor to vadose-
zone soil at Picatinny arsenal, New Jersey. Environ. Sci.
Technol., 24, 676, 1990.
Steinberg, S.M., Pignatello, J.J., and Sawhney, B.L., Persistence
of 1,2-dibromoethane in soils: Entrapment in intraparticle
micropores. Environ. Sci. Technol., 21,1201,1987.
Travis, C.C. and Doty, C.B., Can contaminated aquifers at
Superfund sites be remediated? Environ. Sci. Technol., 24,
1464,1990.
14
-------
Vadas, G.G., Maclntyre, W.G., and Burris, D.R., Aqueous
solubility of liquid hydrocarbon mixtures containing dissolved
solid components, Environ. Toxic. Chem., 10, 633, 1991.
van der Waarden, M., Bridie, A.L.A.M., and Groenewoud,
W.M., Transport of mineral oil components to groundwater I.
Model experiments on the transfer of hydrocarbons from a
residual oil zone to trickling water, Water Res., 5, 213, 1971.
Walters, R.W. and Guiseppi-Elie, A., Sorption of 2,3,7,8-
Tetrachlorodibenzo-p-dioxin to soils from water/methanol
mixtures, Environ. Sci. Technol., 22, 819, 1988.
Wells, M.J.M., Ph.D. Thesis, Auburn, University, Auburn, AL, 1981.
West, C.C., Surfactant-enhanced solubilization of
tetrachloroethylene and degradation products in pump and
treat remediation, Chap. 12 in: Transport and Remediation of
Subsurface Contaminants, D.A. Sabatini and R.C. Knox, eds.,
Amer. Chem. Soc., Wash. D.C., 1992.
Whiffen, R.B. and Bahr, J.M., pp. 75-81 in: Proc. of the Fourth
Natl. Symp. on Aquifer Restoration and Groundwater Monitoring,
National Water Well Assoc., Dublin, OH, 1984.
Willhite, G.P., Waterflooding, Society of Petroleum Engineers,
Richardson, TX, 1986.
Wilson, J.L., and Conrad. S.H., Is physical displacement of
residual hydrocarbons a realistic possibility in aquifer
restoration?, paper presented at Petroleum Hydrocarbons and
Organic Chemicals in Ground Water, pp. 274-298, Nat. Water
Well Assoc., Dublin, OH, 1984.
Wood, A.L., Bouchard, D.C., Brusseau, M.L., and Rao, P.S.C.,
Cosolvent effects on sorption and mobility of organic
contaminants in soils. Chemosphere, 21, 575, 1990.
Yalkowsky, S.H., Flynn, G.L., and Amidon, G.L., Solubilities of
nonelectrolytes in polar solvents. J. Pharm.Sci.,61,983,1972.
15
'U.S. Government Printing Office: 1993 — 750-071/60245
-------
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268
Official Business
Penalty for Private Use
$300
EPA/600/S-93/004
Please make all necessary changes on the below label,
detach or copy, and return to the address in the upper
left-hand corner.
If you do not wish to receive these reports CHECK HERE C:
detach, or copy this cover, and return to the address in the
upper left-hand corner.
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
-------
United States Office of Office of Solid Waste EPA/540/4-90/054
Environmental Protection Research and and Emergency January 1991
Agency Development Response
f/EPA Ground Water Issue
Reductive Dehalogenation of Organic
Contaminants in Soils and Ground Water
Judith L. Sims, Joseph M. Suflita, and Hugh H. Russell
The Regional Superfund Ground-Water Forum is a group of
ground-water scientists, representing EPA's Regional Superfund
Offices, organized to exchange up-to-date Information related
to ground-water remediation of superfund sites. One of the
major issues of concern to the Forum is the transport and fate
of contaminants In soil andgroundwaterasrelatedtosubsurf ace
remedatlon. Process which influence the behavior of contaminants
In the subsurface must be considered both in evaluating the
potential for movement as well as in designing remediation
activities at hazardous waste sites. Such factors not only tend to
regulate the mobility of contaminants, but also their form and
stability. Reductive dehalogenation is a process which may
prove to be of paramount importance In dealing with a particularly
persistent class of contaminants often found in soil and ground
water at superfund sites. This paper summarizes concepts
associated with reductive dehalogenation and describes
applications and limitations to Its use as a remediation technology.
Forfurther information contact Dr. Hugh Russell, FTS 743-2444
at RSKERL-Ada.
Abstract
Although the processes involved in dechlorination of many of
these organic compounds are well understood in the fields of
chemistry and microbiology, technological applications of these
processes to environmental remediation are relatively new-
partlcularly at pilot or field scale. It is well established, however,
that there are several mechanisms which result In dehalogenation
of some classes of organic contaminants, often rendering them
less offensive environmentally. These include: stimulation of
metabolic sequences through Introduction of electron donor
and acceptor combinations; addition of nutrients to meet the
needs of dehalogenating microorganisms; possible use of
engineered micro-organisms; and use of enzyme systems
capable of catalyzing reductive dehalogenation.
The current state of research and development In the area of
reductive dehalogenation is discussed along with possible
technological applications of relevant processes and mechanisms
to the remediation of soil and ground water contaminated with
chlorinated organics. In addition, an overview of research
needs is suggested which might be of interest for development
of in situ systems to reduce the mass of halogenated organic
contaminants in soil and ground water.
Introduction and large scale production of synthetic halogenated
organic chemicals overthe last 50 years has resulted in a group
of contaminants which tend to persist in the environment and
resist both biotic and abiotic degradation. The low solubility of
these types of contaminants, along with their toxicity and tendency
to accumulate in food chains, make them particularly relevant
targets for remediation activities.
Introduction
Large scale production of synthetic halogenated organic
compounds, which are often resistant to both biotic and abiotic
degradation, has occurred only In the last few decades (Hutzinger
and Verkamp 1981). However, naturally occurring halogenated
organic compounds have existed in marine systemsforperhaps
/
u
r&lON
Printed on Recycled Paper
echnical
bupport
t
o
| roject .
Superfund Technology Support Center for Ground Water
Robert S. Kerr Environmental
Research Laboratory
Ada, OK
-------
millions of years. These compounds, including aliphatic and
aromatic compounds containing chlorine, bromine, or iodine,
are produced by macroalgae and invertebrates. The presence
of these natural compounds, at potentially high concentrations,
may have resulted in populations of bacteria that are eflective
dehalogenators (King 1988). Microorganisms exposed to
halogenated compounds in soil and ground water may also
have developed enzymatic capabilities similarto those in marine
environments. Enzyme systems that have evolved to degrade
nonchlorinated compounds may also be specific enough to
degrade those that are chlorinated. (Tiedje and Stevens 1987).
Many halogenated organic compounds are not very soluble and
tend to be highly lipophilic, therefore having the potential to
bioaccumulate Insomefood chains. These chemical properties,
along with their toxicity and resistance to degradation, present
the potential for adverse health effects and ecosystem
perturbations upon exposure (Rochkind et al. 1986).
Recent research findings indicate that anaerobic processes that
remove halogens from these compounds produce dehalogenated
compounds that are generally less toxic, less likely to
bioaccumulate, and moresusceptibletofurthermlcroblal attack,
especially by aerobic microorganisms utilizing oxidative
biodegradative processes. Both aromatic and nonaromatic
organic compounds are subject to these dehalogenation
processes. Technological applications of these processes for
remediation of contaminated soils and ground waters is of a
relatively new concept.
Recent research also has shown that anaerobic dehalogenation
reactions specifically involving reductive processes can effectively
degrade a wide variety of halogenated contaminants In soil and
ground water (Vogel et al. 1987, Kuhn and Suflita 1989a).
Organic compounds generally represent reduced forms of
carbon, making degradation by oxidation energetically favorable.
However, halogenated organic compounds are relatively oxidized
by the presence of halogen substltuents, which are highly
electronegative and thus more susceptible to reduction. A
compound with more halogen substituents is therefore more
oxidized and more susceptible to reduction. Thus, with increased
halogenation, reduction becomes more likely than does oxidation
(Vogel etal. 1987).
An organic compound is considered to be reduced if a reaction
leads to an increase in its hydrogen content or a decrease in its
oxygen content; however, many reduction reactions (e.g., the
vicinal reduction process) do not involve changes in the hydrogen
or oxygen content of a compound. Oxidation and reduction
reactions are more precisely def ined in terms of electron transfers.
An organic chemical Is said to be reduced if it undergoes a net
gain of electrons as the result of a chemical reaction (electron
acceptor), and is said to be oxidized if it undergoes a net loss of
electrons (electron donor). Under environmental conditions,
oxygen commonly acts as the electron acceptor when present.
When oxygen is not present (anoxic conditions), microoiganisms
can use organic chemicals or inorganic anions as alternate
electron acceptors under metabolic conditions referred to as
fermentative, denitrifying, sulfate-reducing or methanogenic.
Generally, organic compounds present at a contaminated site
represent potential electron donors to support microbial
metabolism. However, halogenated compounds can act as
electron acceptors, and thus become reduced in the reductive
dehalogenation process. Specifically, dehalogenation by reduction
Is the replacement of a halogen such as chloride, bromide,
fluoride, or iodide on an organic molecule by a hydrogen atom.
Vicinal reduction occurs when two halogens are released while
two electrons are Incorporated Into the compound.
An organic chemical would be expected to be reduced if the
electrode potential of the specific soli or ground-water system,
in which the chemical is present, Is less than that of the organic
chemical (Dragun 1988). The electrode potential is described
by the oxidation-reduction (redox) status of the system, referlng
to potential for the transfer of electrons to a reducible material.
The electron (e) participates in chemical reactions In soil and
ground water similar to the hydrogen Ion (H+) in that electrons
are donated from a reduced compound to an oxldidlzed. Redox
potential (Eh) is usually reported In volts and Is measured using
a reference electrode in combination with a metallic electrode,
such as platinum, which Is sensitive and reversible to oxidation-
reduction conditions.
The redox potential of a soil system is complex. The oxidation
state of each soil constituent, such as organic compounds,
humus, Iron, manganese, and sulfur, contributes to the measured
redox potential. The contribution of each constituent in a
system varies with such factors as soil water content, oxygen
activity, and pH. Weil-oxidized soils have redox potentials of 0.4
to 0.8 V, while extremely reduced soils may have potentials of
-0.1 to -0.5 V (Dragun 1988).
The potential for anaerobic biological processes to reductively
dehalogenate organic compounds may be important in the
bioremediation of soils and aquifers contaminated with these
compounds. These environments often become anaerobic due
to depletion of oxygen by the microbial degradation of more
easily degradable organic matter. When compounds can be
degraded under anaerobic conditions, the cost associated with
the maintenance of an aerobic environment by providing air,
ozone, or hydrogen peroxide would be eliminated (Sufltta et al.
1988).
While anaerobic biological mediated reductive dehalogenation
mechanisms were demonstrated as early as 1983 (Allan, 1955),
the utilization of this process as a remedial alternative to reduce
the overall mass of halogenated organic compounds from soil
and ground water Is a new concept and still subject to field
demonstrations.
For this reason research is currently underway to better define
the basic mechanisms of reductive dehalogenation reactions.
Such approaches may include: (1) stimulation of desirable
metabolic sequences in soil and ground water through the
intentional introduction of suitable electron donor and acceptor
combinations (Suflita et al. 1988); (2) addition of adequate
nutrients to meet the nutritional requirements of dehalogenating
microorganisms (Palmer et al. 1989); (3) use of engineered
2
-------
microorganisms (Palmer et al. 1989); (3) use of engineered
microorganisms with optimum dehalogenating activity (Palmer
et al. 1989); and (4) addition of cell-free enzymes capable of
catalyzing reductive dehalogenation reactions (DeWeerd and
Sufiita 1989).
Dehalogenation Mechanisms
Anaerobic reductive dehalogenation is only one of the mechanisms
available to remove halogens from organic compounds. Other
anaerobic dehalogenation processes are identified In Figure 1
(Kuhn and Sufiita 1989a). The reactions are classified according
to the type of compound undergoing dehalogenation, l.e„
aromatic or nonaromatic.
Dehalogenation of Aromatic Compounds
Two mechanisms of dehalogenation for aromatic compounds
under anaerobic conditions have been observed; reduction and
hydrolysis. Reductive mechanisms are recognized as the
predominant pathway for removal of halogens from homocyclic
aromatic rings under anaerobic conditions, while hydrolytic
dehalogenation (including both chemically and enzymatically
mediated reactions) is the preferred mechanlsmfor heterocyclic
aromatic compounds (Sufiita et al. 1982; Kuhn and Sufiita
1989a). However, Adrian and Sufiita (1989) have recently
demonstrated reductive denomination of the herbicide bromacil
under methanogenlc conditions. This is the first report of
reductive dehalogenation of a heterocyclic aromatic compound.
Reductive Dehalogenation of Aromatic Compounds
Many classes of halogenated aromatic compounds have been
shown to be degraded by reductive dehalogenation processes
(Table 1). Evidence for the involvement of microorganisms In
aryl or aromatic reductive dehalogenation reactions Include: (1)
the specificity of the reductive reaction; (2) characteristic lag
periods required before significant dehalogenation is observed;
(3) the absence of activity in autoclaved controls; and (4) the
isolation of aryl dehalogenating bacteria.
Anaerobic Dehalogenation Mechanisms
Reduction
Hydrolysis
wsmimmmmm
2e-+H+ X"
OH" X"
V ^
Reduction
(Hydrogenoiysis)
2e- + H+
llll^c-x
v, ^
iiii)c-h
Hydrolysis
(Substitution)
llll)c-x
OH"
II^C—OH
Vicinal Reduction
(Dihalo-Elimination)
X X
I I
2e-
-i-i- ^^
V.C =C^
Dehydrohalogenation _ C _ C _
I I
H
HX
=c^
Figure 1. Examples of anaerobic dehalogenation mechanisms for aromatic and nonaromatic pesticides (Kuhn and Sufiita, 1989a)
3
-------
Reductive dehalogenation Is rare in well-aerated environments.
Methanogenlc conditions, In which the typical redox potential is
-0.3 V, the preferred electron acceptor is carbon dioxide, and
the product Is methane (Dragun 1988), appear to be optimal for
this type of biotransformation. Genthner et al. (1989), have
recently investigated dehalogenation of monochlorophenols
and monochlorobenzoates under four anaerobic enrichment
conditions: methanogenic, nitrate-reducing, sulfate-reducing,
and bromoethane sulfonic acid (BESA)-amended. BESA is a
potent inhibitor of methanogenesis and was used to promote
reductive dechlorination as a terminal electron process.
Aquatic sediments used as inocula were collected from a salinity
gradient that Included both freshwater and estuarine environments
and varying degrees of exposure to Industrial effluents.
Degradation was observed least often in enrichments with
nitrate or sulfate, and most often when amended with 1 mM
BESA. In contrast to 1mM BESA, 10mM BESA prevented or
delayed the degradation of several of the chloroaromatic
compounds, suggesting inhibition of methanogenesis or inhibition
of reductive dechlorination by BESA. Other sulfur oxyanions
also have been shown to inhibit anaerobic dehalogenation
reactions where sulfate is present as an inorganic contaminant
(DeWeerd et al. 1986, Gibson and Suflita 1986, Suflita et al.
1988, Kuhn and Suflita 1989b). Additional research is being
conducted in environments where sulfate occurs naturally. King
(1988) showed that sulfate-reducing bacteria did not participate
in dehalogenation of 2,4-dibromophenol (DBP), a naturally
occurring halogenated organic compound in some marine
sediments, but did appear to degrade phenol, a metabolic
product of DBP dehalogenation.
The reductive dehalogenation of chlorinated compounds, as
shown in Table 1, Is characterized by their specificity for
compounds within a particular chemical class, for example
benzoates, phenols, or phenoxyacetates (Suflita et al. 1982,
Gibson and Suflita 1986, Suflita and Miller 1985, Kuhn and
Suflita 1989a). Recently, however, research has shown that
cross-acclimation between compound classes can occur. Struijs
and Rogers (1989) demonstrated the reductive dehalogenation
of dichloroanilines by anaerobic microorganisms In pond sediments
acclimated to dehalogenate dichlorophenols. Since both hydroxyl
and amino groups have a tendency to donate electrons, the
authors hypothesized that organisms that were capable of
dechlorinating dichlorophenols, which have been shown to be
relatively non-persistent in the environment, could possibly
dechlorinate the more persistant dichloroanilines. The
monochloroanlllnes produced by dechlorination of the
dichloroanilines were stable under anaerobic conditions, but
have been shown previously to be readily degraded under
aerobic conditions (Zeyerand Kearney 1982, Zeyeret al. 1985).
The specificity of dehalogenation also Is dependent on the
position of halogens on the aromatic ring within a class of
compounds. For example, chlorinated benzoates are generally
more readily dehalogenated at the meta position, followed by
the ortho and para positions (Suflita etal. 1982,Genthneretal.
1989). Hydroxy, alkoxy, and nitrogen-substituted aromatic
Class of Halogenated Aromatic Compounds
Carboxylated Benzenes
Oxygen-Substituted Benzenes
Nitrogen-Substituted Benzenes
Cyano-Subsfituted Benzenes
Methylene-substituted Benzenes
Chlorinated Benzenes
Polychlorinated biphenyls
Examples of Specific Compounds
Amiben
Dicamba
2,3,6-trichlorobenzoate
Pentachlorophenol
Chlorinated phenoxyacetates (e.g.,
2,4-D, 2,4,5-T
Halogenated diphenyl ether herbicides
(e.g., chloronitrofen)
3,4-DihaJogenated aromatic compounds
(diuron, DCPU, Imuran, DCIPC,
propanil)
Pentachloronitrobenzene
2,4,5,6-tetrachloroisophthalonitrile
(TPN)
Benthiocarb
Hexachlorobenzene
Araclors (commercial PCB products)
Table 1. Classes of halogenated aromatic compounds demonstrated to be susceptible to degradation by reductive dehalogenation processes
(Kuhn and Suflita 1989a).
4
-------
compounds generally are dehalogenated laster at ortho and
para halogens (Kuhn and Sufllta 1989a, 1989b), however,
Genthner et al. (1989) recently have shown that the order of
degradabllity of monochlorophenols was meta > ortho > para..
Mlkesell and Boyd (1986) have shown that three groups of
acclimated microorganisms can act in concert to completely
dehalogenate pentachlorophenol (PCP) to form phenol, a substrate
that was labile under the methanogenic conditions of their
experiments. Each type of microorganism, acclimated to one of
three monochlorophenol isomers, transformed PCP by removal
of halogens from the same relative ring positions at which they
dehalogenated the monochlorophenol substrates. The 2-
chlorophenol adapted cells dehalogenated PCP at the two ortho
positions as well as Irom the para position. Similarly, 4-
chlorophenol adapted cells cleaved the para chlorine of PCP in
addition to the two ortho substituents. In contrast, the 3-
chlorophenol adapted cells exclusively dehalogenated the meta
position.
Other studies of PCP degradation have shown accumulation of
tri- and tetrachlorophenol intermediates, which indicates that
higher halogenated phenols tend to be more readily dehalogenated
than their lesser halogenated congeners. Similarly, dehalogenation
of chlorinated anilines shows shorter lag periods and faster
dehalogenation rates with multi-halogenated compounds
compared to di- and monohaiogenated anilines. Dehalogenation
of aromatic amines occurs predominately at the ortho and para
positions as has been demonstrated with the dechlorination of
anilines (Kuhn and Suflita 1989b), though removal of meta
halogens from this group of compounds has also been
demonstrated.
Reductive dehalogenation may require the induction of enzymes
responsible for dehalogenation. DeWeerd and Suflita (1990)
have demonstrated reductive dehalogenation of 3-chlorobenzoate
using cell-free extracts of an anaerobic bacterium. The extracts
exhibited the same substrate specificity as whole cells. Rapid
dehalogenation activity was found only in extracts of cells
cultured in the presence of the halogenated molecule, indicating
that the enzymes responstole required induction. Dehalogenation
was Inhibited by sulfite, thiosulfate, and sulfide. Dehalogenation
activity was associated with the membrane fraction and required
a low potential electron donor. These results suggest that a
specific enzyme is made by the cells for dehalogenation of
selected halogenated substrates. Research into the use of
enzymes as a potential amendment to enhance bioremediation
should be encouraged.
Further evidence that reductive dehalogenation may depend
upon the induction of enzymes has been presented by Linkf ield
et al. (1989). Acclimation periods prior to detectable dehalogenation
of halogenated benzoates in anaerobic lake sediments ranged
from 3 weeks to 6 months. These periods were reproducible
over time and among sampling sites and characteristic of the
specific benzoate compound tested. The lengthy acclimation
period appeared to represent an induction phase in which little
or no aryl dehalogenation was observed. This was followed by
an exponential increase in activity typical of an enrichment
response. Extremely low activities during the early days of
acclimation, coupled with the fact that dehalogenation yields no
carbon to support microbial growth, suggests that slow continuous
growth from time of the first exposure of the chemical was not
responsible for the acclimation period. The characteristic
acclimation period for each chemical also argues against nutritional
deficiency, Inhibitory environmental conditions, orpredatlon by
protozoa or other microbial grazers as the cause of the acclimation
period. The reproducibility of the findings with time and space
and among replicates argues against genetic changes as the
explanation.
The removal of chloride or bromide from an aromatic molecule
proceeds easier when the ring also is substituted with electron
destabilizing groups, such ascarboxy, hydroxy, orcyano groups
(Kuhn and Suflita 1989a). Other chemical groups attached to
the aromatic ring by nitrogen or oxygen bonds may have the
same effect on the reductive dehalogenation reaction. However,
recent research has shown that even highly chlorinated, poorly
water soluble aromatic hydrocarbons that do not contain polar
functional groups can also undergo reductive dehalogenation.
Hexachlorobenzene (HCB) has generally been considered
recalcitrant to microbial attack, particularly In the absence of
oxygen (Bouwer and McCarty 1984, Kuhn et al. 1985); however,
HCB was shown to degrade to tri- and dichlorobenzenes by
Fathepure et al. (1988). Brown et al. (1987) performed standard
thermochemical calculations of free-energy changes associated
with the oxidation of organic compounds (In this case, glucose)
coupled with the reduction of chlorobenzene compounds. The
reactions involving HCB and monochlorobenzene offered more
energy to anaerobic bacteria than the reduction of compounds
available naturally in anaerobic environments, such as sulfate
and carbon dioxide (Table 2). Also, more energy could be
obtained from the dehalogenation of hexachlorobenzene to
benzene than the dehalogenation of monochlorobenzene,
Indicating that dehalogenation reactions are more likely to occur
with aromatic compounds containing manychloro groups since
they are more highly oxidized and more electronegative than
those containing fewer chloro groups.
Polychlorinated biphenyls (PCBs), commonly thought to be
resistant to blodegradative processes, have also been shown to
be susceptible to degradation by reductive dehalogenation
(Brown et al. 1987, Quensen et al. 1988). Brown et al. (1987)
suggest that dehalogenated products formed were less toxic
than the original PCB congeners and may possible be more
susceptible to oxidative biodegradation by aerobic bacteria.
Hydrolytic Dehalogenation of Aromatic Compounds
Hydrolytic dehalogenation represents a substitution reaction in
which a hydroxyl group replaces a halogen on an organic
molecule (Figure 1). In general, the anaerobic hydrolytic
removal of halogen substituents from homocyclic aromatic
compounds is rare (Kuhn and Sufllta 1989a), but has been
observed under aerobic conditions. Also, the enzymes involved
have been shown to be active in reduced media, and some were
inhibited by oxygen (Marks et al. 1984, Thiele et al. 1988). This
transformation has been observed in anaerobic soil for a single
5
-------
Oxidant
Reduced Product
AG
(kcal/mol)
Molecular oxygen (O,)
Hexachlorobenzene (C,CI,)
Monochlorobenzene (C,H,CI)
Sulfate (S04l)
Carbon dioxide (CO,)
Water (H,0)
Benzene (C,H,)
Benzene (C,Ht)
Reduced Sulfur
Methane (CH4)
-676.10
-410.16
-369.50
-131.78
-95.63
Table 2. Standard free-energy changes (or the oxidation of glucose to CO, and H,0 using various oxidants (Brown et al. 1987).
herbicide, flamprop-methyl; however no anaerobic bacteria
were isolated with the ability to catalyze this type of dehalogenation.
A hydrolytic defluorination product of the herbicide was identified
in anaerobic soil incubation studies (Roberts and Standen
1978).
Heterocyclic chloroaromatic compounds, such as chlorinated
triazine herbicides, tend to react more readily with hydroxy,
amino, or sulfhydryl groups than their homocyclic chemical
counterparts. Hydrolytic dehalogenation is, therelore, the
preferred mechanism for removing halogens from hetero-cyclic
aromatic compounds under anaerobic conditions (Adrian and
Suflita 1989).
The hydrolysis of triazine herbicides to form dehalogenated and
less phytotoxic products has been known for many years (Paris
and Lewis 1973). However, there has been controversy overthe
involvement of microorganisms in this process. Reactions with
reactive soil surfaces, such as clays and organic matter, appear
to be significant with regard to the rate of hydrolysis (Kuhn and
Suflita 1989a). Dechlorination ol s -triazlnes has been shown to
be catalyzed by microorganisms. This was demonstrated by
Cook and Huetter (1984,1986). The organisms studied were
aerobic, but biotransformation of the herbicides did not require
molecular 02 and was functional under anaerobic conditions.
Dehalogenation of Nonaromatic Compounds
Dehalogenation of nonaromatic compounds, particularly
halogenated aliphatic chemicals, is generally better understood
than aryl dehalogenation reactions. The reductive processes of
hydrolysis and dehydrohalogenation have been identified as
anaerobic dehalogenation mechanisms (Figure 1).
In general, biologically mediated anaerobic dehalogenation of
nonaromatic compounds tends to be f asterthan dehalogenation
of aromatic compounds, does not require long adaptation times,
and does not exhibit a high degree of substrate specificity.
Some of these reactions also are not too sensitive to the
presence of oxygen and have been observed in aerobic incubation
systems. The greater variety of reaction mechanisms potentially
available to metabolize nonaromatic halogenated compounds
in general results in rendering these compounds more susceptible
to biodegradation than the haloaromatic compounds (Vogel et
al. 1987, Kuhn and Suflita 1989a).
Dehalogenation has been demonstrated with many bacterial
species representing diverse genera Mesophiiic and thermophilic
methanogenic bacteria as well as some thermophilic clostridial
species may catalyze dehalogenation of some aliphatic
compounds. For exanpte, metabolism of hexachlorocyclohexanes
by thermophilic Clostridia was reported by Sethunathan (1973).
Dehalogenation reactions are also sometimes heat resistant,
suggesting that some reactions may not be enzymatically
mediated, and therefore not dependent on Intact microorganisms
or microbial consortia. The dehalogenation of nonaromatic
compounds can be catalyzed by transition metal complexes
with or without the Involvement of enzymes (Kuhn and Suflita
1989a).
Reductive and Vicinal Dehalogenation of Nonaromatic
Compounds
If a nonaromatic carbon atom in a synthetic molecule Is highly
halogenated, dehalogenation is more easily accomplished by
reductive, vicinal reductive or elimination reactions (Vogel et al.
1987). Compounds that have been demonstrated to be degraded
by reduction or vicinal reduction mechanisms are listed in Table
3.
Reductive and vicinal dehalogenation reactions are dependent
on the redox potential of the electron donor and acceptor. To be
thermodynamically feasible, the Eh of the electron accepting
reactant (dehalogenation) must be higher than that of the
electron donating reactant. This requirement can limit the
number of available electron donors for dehalogenation of some
compounds (Castro et al. 1985, Vogel et al. 1987, Kuhn and
Suflita 1989a). For example, free ferrous iron (Fe(ll)) has a
redox potential of+0.77 V; but most of the halogenated alkanes
and alkenes with lower redox potentials will not react with this
transition metal. However, when Fe(ll) is in a complexed form,
such as a porphyrin or as ferredoxin, the redox potential is
6
-------
Class of Halogenated Nonaromatic Compound
Examples of Specific Compounds
Aliphatic Compounds
Alicyciic Compounds
Tetrachloromethane (carbon
tetrachloride)
Trichloromethane (chloroform)
Dichloromethane (methylene chloride)
Chloromethane (methyl chloride)
Bromomethane (methyl bromide)
Trichloronitromethane (chloropicrin)
Hexachloroe thane
Tetrachloroethene (perchloroethyiene)
1,1,1 -Trichloroethane
Trlchlorethene (trlchtoroethylene)
1,2-Dichloroethane (ethylene
dichloride, EDC)
1,2-dibromoethane (ethylene
dibromide, EDB)
1,2-dibromo-3-chloropropane
(DBCP)
1,1,1 -trichloro-2,2-bis
(p-chlorophenyl)ethane (DDT) (aliphatic
portion)
Toxaphene
Mirex
Heptachlor
Hexahalocyclohexanes
Lindane and Its isomers
Table 3. Classes of halogenated nonaromatlc compounds demonstrated to be susceptible to degradation by reductive dehalogenation
processes (Kuhn and Sufllta 1989a).
dramatically lowered, and the reaction Is possible (Kuhn and
Sufllta 1989a). As examples, Fe(ll)deuteroporphin IX and
cytochrome P-450 have redox potentials of 0.00 V and -0.17 V,
respectively.
Active transition metal complexes, which Include complexes of
iron (Fe), cobalt (Co), nickel (Ni), and perhaps chromium (Cr)
and zinc (Zn), have redox potentials less than zero and can be
as low as -0.8 V for the cobalt complexed vitamin 6,2. The low
redox potentlalsoftheseelectrondonorsallowforthelrreduction
to be coupled with dehalogenation of many nonaromatic
compounds having redox potentials which range from 0 to 1.2
V (Vogel et al. 1987).
Highly halogenated aliphatic compounds have higher reduction
potentials than their lesser halogenated analogues; therefore,
more energy is released by their dehalogenation, Indicating a
greater driving force for these reactions. In general, reductive
dehalogenation of tetra- and tri-halogenated carbon atoms is
easier than di- or monohalogenated congeners (Vogel et al.
1987).
In natural environments, Fe(ll) porphyrins (e.g., cytochromes),
Co complexes (e.g., vitamin B,2), and NI complexes (e.g., F-
430) are likely to be dominant in the reductive dehalogenation
process. Dead cells can release these stable transition metal
complexes which are then more available for participation In the
dehalogenation process. Such complexes are also active in
living cells, as was demonstrated with Pseudomonas putida by
Castro etal. (1985). Pseudomonas putida contains Fe(ll)potphyrin
bound to the cytochrome P-450 complex, but movement of
halogenated compounds across the bacterial membrane and
diffusion to the active Iron center can limit the rate of dehalogenation.
Another potential reductant available for dehalogenation of
haloaliphatic compounds in natural environments is the flavin/
flavoprotein complex, which has been shown to mediate many
of the known reductive reactions of xenoblotic compounds in
laboratory studies (Esaac and Matsumura 1980). To date, no
studies have clearly demonstrated the environmental significance
of this reductant. Relative to other dehalogenation reaction
mechanisms, dehalogenation by vicinal reduction appears to be
more tolerant of oxidized conditions and may even be independent
of transition metals or metallo-organic complexes (Kuhn and
Suflita 1989a).
Dehydrohalogenation of Nonaromatic Compounds
Dehydrohalogenation is an elimination reaction in which two
groups are lost from adjacent carbon atoms so that a double
bond is formed, resulting in the release of a halogen and a
proton (HX) and the formation of an alkene (Figure 1). The rate
of dehalogenation is higher when additional chloride ions are
bonded to the carbon atom that loses Its chloride ion substituent
7
-------
(Vogel et al. 1987). Bromine atoms rather than chlorine atoms
are generally more readily eliminated by this reaction. Elimination
reactions can proceed spontaneously (1,1,1 -trichloroethane;
1,2-dibromoethane) or can be catalyzed by microbial enzymes
such as the dechlorinase enzyme which Is responsible for the
conversion of DDT to DDE--a dechlorination reaction involving
the aliphatic portion of the DDT molecule (Kuhn and Sufllta
1989a).
Hydrolytic Dehalogenation
Hydrolysis, a substitution reaction In which one substituent on a
molecule Is replaced by another, has been demonstrated with
many aliphatic compounds. Hydrolysis is favored for carbon
atoms with only one or two halogens; however, hydrolytic
dehalogenation has been shown with higher chlorinated
compounds, such as 1,1,1 -trichloroethane. This transformation
can be chemically or biologically catalyzed by methanogenic
mixed cultures and by a number of aerobic bacterial isolates.
Bromine loss tends to be favored compared to the corresponding
chlorinated compounds (Kuhn and Suflita 1989a).
Applications And Limitations of Reductive
Dehalogenation of Organic Halogenated Pollutants
The degradation of trichloroethylene (TCE), as shown in Figure
2, may be used to illustrate the potential effectiveness of the
reductive dehalogenation process to remove common pollutants
from the environment, as well as to present some of the cautions
that should be observed (Dragun 1988). TCE Is an industrial
solvent used extensively for degreasing metal as well as in dry-
cleaning operations, organic synthesis, refrigerants, and fumlgants.
Most septic tank cleaning fluids contain TCE (Craun 1984).
Also Illustrated in Figure 2 are possible degradatlve pathways of
tetrachloroethylene (PCE) and 1,1,1-trichloroethane (1,1,1-TCA).
PCE is a solvent widely used in dry cleaning and degreasing
operations; 1,1,1-TCA Is used extensively as an Industrial
cleaner and degreaser of metals, spot remover, adhesive, and
vapor pressure depressant (Craun 1984). These compounds
have relatively high water solubility (e.g., 1000 mg/l for TCE)
and are highly mobile in soils and aquifer materials and often are
found in ground waters. Since they are suspected carcinogens
(Infante and Tsongas 1982), they represent a threat to human
health.
The degradative pathway for TCE (Dragun, 1988) can be
described as follows:
(1) TCE can undergo reductive dehalogenation, i.e., the
removal of one chloride atom (CI) and the addition of
one hydrogen (H) atom. Three possible reaction
products can be formed: 1,1-dichloroethylene (1,1-
DCE), cis-1,2-dichloroethylene (c-1,2-DCE), and/or
trans-1,2-dichloroethylene (t-1,2-DCE).
Figure 2. Transformation pathways for various chlorinated volatile hydrocarbons in soil systems (Drugun 1988).
-------
(2) 1,1-DCE can undergo reductive dehalogenatlon to
form vinyl chloride, or its carbon-carbon double bond
can be reduced to form 1,1 -dlchloroethane (1,1 -DCA).
(3) The two dlchloroethylene compounds, c-1,2-DCE and
M.2-DCE can undergo reductive dehalogenation to
form vinyl chloride. Their carbon-carbon double bonds
can be reduced to form 1,2-dichloroethane (1,2-DCA).
(4) 1,1 -DCA and 1,2-DCA can undergo dehydrohalogenatlon
to form vinyl chloride. These two chemicals can also
undergo reductive dehalogenatlon to form chloroethane.
The degradation pathway of a single compound, TCE, can lead
to the production of six chlorinated volatile hydrocarbons. The
degradation of PCE can lead to the production of seven chlorinated
volatile hydrocarbons, while the degradation of 1,1,1-TCA can
lead to the production of four chlorinated hydrocarbons. Two of
the metabolic products formed, vinyl chloride and 1,1-DCA,
have been classified as a carcinogen and a probable carcinogen,
respectively (Vogel et al. 1987). The dlchloroethylene products,
c-1,2-DCE and t-1,2-DCE, and vinyl chloride are also regulated
under the 1986 Safe Drinking Water Amendments (Freedman
and Gossett 1989). Vinyl chloride is the most persistent of the
compounds under anaerobic conditions, but can be rapidly
degraded under aerobic conditions (Hartsmans et al. 1985,
Fogel et al. 1986).
Management of a bioremediation system to accomplish treatment
of these compounds in a manner to protect human health and
the environment should incorporate considerations of detoxification
as well as disappearance of the parent compounds. Disappearance
is not synonymous, however, with mineralization to Inorganic
salts, carbon dioxide, and water. Partial degradation of organic
substrates can result in the production of metabolic products
that generate their own environmental and health consequences.
Such contaminants may be of more toxlcologlcal concern than
the parent compounds (Suflita et al. 1988).
Fathepure and Boyd (1988) recently suggested that In situ
dechlorination of PCE to TCE could be enhanced by stimulating
methanogenesis. They found that reductive dechlorination of
PCE occurred only during methanogenesis, and no dechlorination
was seen when methane production ceased. There was a clear
dependence of the extent of PCE dechlorination on the amount
of methanogenic substrate (methanol) consumed. Methanogenic
bacteria are present in a diversity of environmental habitats.
Including those where chloroethylenes are commonly found as
contaminants (e.g., soils, ground waters, and aquifers near
landfills).
Abionemediation systemforchlorlnated ethylenes and ethanes
could consist of maintenance of an anaerobic environment,
followed by aeration to complete the degradation process after
anaerobic degradative processes have reduced the parent
compounds to acceptable levels. Recent research, however,
by Freedman and Gossett (1989) has shown that PCE and TCE
can be degraded to ethylene, a non-chlorinated environmentally
acceptable biotransformation product, under anaerobic
methanogenic conditions if an adequate supply of electron
donors was supplied to a mixed anaerobic enrichment culture.
Methanol was the most effective electron donor, although
hydrogen, formate, acetate, and glucose also served.
Ethylene is sparingly soluble In water and has not been associated
with any long-term toxlcologlcal problems (Autian 1980). It is
also a naturally occurring plant hormone. Since complete
conversion of VC to ethylene was not observed in the study, the
authors suggested that further research Is required to determine
the concentration of electron donors required to complete the
conversion.
A major operational cost of this method of enhanced anaerobic
bioremediation will be the supply of electron donors. Alternatively,
means to channel more of the donors Into the reductive
dechlorination process and less into methane production should
be investigated.
As proposed by Fathepure et al. (1988), a similar potential for
the use of an anaerobic environment followed by an aerobic
environment, for mineralization and detoxification of halogenated
organic pollutants, is illustrated by the degradation of
hexachlorobenzene (HCB) (Figure 3). HCB is a fungicide used
as a seed coating for cereal crops. Two pathways of dechlorination
were proposed: (1) a major pathway In which 1,3,5-
trlchlorobenzene (1,3,5-TCB) is formed via pentachlorobenzene
and 1,2,3,5-tetrachlorobenzene (1,2,3,5-TTCB); and (2) a minor
pathway in which dichlorobenzenes are formed via 1,2,4,5-
TTCB and 1,2,4-TCB.
The authors presented explanations for the existence of two
pathways. One is that there were two populations, each using a
different pathway. The other is that the products reflect the
distribution of reactive ring intermediates in which a chlorine,
between two other chlorines, was lost most readily and
dechlorination ceased when there are no adjacent chlorines as
with 1,3,5-TCB.
Reductive dechlorination appeared to occur In a stepwise
fashion until lower chlorinated compounds accumulated. Most
of the added HCB accumulated as 1,3,5-TCB, which remained
unchanged. Although metabolic products Identified In this study
were not further utilized by the anaerobic sludge populations
used to elucidate the metabolic pathways, It is likely that they
would be degraded by aerobic organisms (Reineke and Knackmuss
1984, deBont et al. 1986, Schraa et al. 1986, Spain and Nishino
1987) or by facultative anaerobes possessing dechlorinatlng
activity (Tsuchiya and Yamaha 1984).
The U.S. Environmental Protection Agency is presently sponsoring
research to develop engineered microorganisms capable of
anaerobic reductive dehalogenation of organic halogenated
compounds (Palmer et al. 1989). Desutfomonile tiedjei (DeWeerd
et al, 1990), formerly known as DCB-1, is the first obligate
anaerobe known to accomplish reductive dehalogenation. Results
using this organism indicated that no plasmld genes responsible
for dehalogenating activity could be detected. Therefore, in
9
-------
order to clone the gene or genes responsible for the activity, a
genomic library of the bacterial chromosome Is being constructed
to Isolate the dehalogenase gene. The Isolation of the gene
would be greatly facilitated by the isolation and characterization
of the requisite dehalogenase.
Summary
Bioremedlatlon of soils and ground waters contaminated with
organic pollutants involves management of the contaminated
system to control and enhance biodegradation of the pollutants
present (Sims et al. 1989, Thomas and Ward 1989). Reductive
dehalogenation appears to be a potentially powerful process for
achieving bioremediation of a site contaminated with organic
halogenated pollutants, If mechanisms and pathways of
degradation are known and can be managed to achieve removal
of the compounds of interest as well as potentially toxic metabolic
degradation products.
References
Adrian, N. R., and J.M. Suflita. 1989. Reductive dehalogenation
of a nitrogen heterocyclic herbicide in anoxic aquifer slurries.
Appl. Environ. Microbiol. 56:292-294.
Allan, J. 1955. Loss of biological efficiency of cattle-dipping
wash containing benzene hexachlorlde. Nature (London)
175:1131-1132.
Autian, J. 1980. Plastics, p. 531 -556. In: J. Doull, C.D. Klaassen,
and M.O. Amdur (eds.) Casarett and Douirs Toxicology. Macmillan
Publishing Co., Inc. New York, NY.
Bouwer, E.J., and P.L. McCarty. 1984. Utilization rates of trace
halogenated organic compounds In acetate-grown blofllms.
Biotechnol. Bioeng. 27:1564-1571.
Brown, J.F., Jr., R.E. Wagner, H. Feng, D.L. Bedard, M.J.
Brennan, J.C. Carnahan, and R.J. May. 1987. Environmental
dechlorination of PCBs. Environ. Toxicol. Chem. 6:579-593.
Castro, C.E., R. S. Wade, and N.O. Belser. 1985.
Biodehalogenatlon: Reactions of cytochrome P-450 with
polyhalomethanes. Biochemistry 24204-210.
Cook, A. M., and R. Huetter. 1984. Deethylsimazine: Bacterial
dechlorination, deamination, and complete degradation. J. Agric.
Food Chem. 32:581-585.
Cook, A. M., and R. Huetter. 1986. Ring dechlorination of
deethylsimazine by hydrolases from Rhodocoocus coralllnus.
FEMS Microbiol Letters 34:335-338.
Craun, G.F. 1984. Health aspects of groundwater pollution, pp.
135-179. In: G. Bitton and C. Gerba (eds.). Groundwater
Pollution Microbiology. John Wiley & Sons, New York, NY.
deBont, J.A.M., M.J.A.W. Vorage, S. Hart mans, and W.J. J. van
den Tweel. 1986. Microbial degradation of 1,3-dichlorobenzene.
Appl. Environ. Microbiol. 52:677-680.
DeWeerd, K.A., J.M. Suflita, T. Linkfield, J.M. Tiedje, and P.H.
Pritchard. 1986. The relationship between reductive dechlorination
and other aryl substituent removal reactions catalyzed by
anaerobes. FEMS Microbiol. Ecoi. 38:331-340.
10
-------
DeWeerd, K.A., and J.M. Suflita. 1990. Anaerobic aryl
dehalogenation of halobenzolates by cell extracts of "DesutomonUe
tledjef'. Appl. Environ. Microbiol. 56: in press (out In the October
Issue).
DeWeerd, K.A., L. Mandeico, R.S. Tanner, C.R. Woese, and
J.M. Suflita. 1990. Desulfomonlle tledjel gen. nov. and sp. nov.,
a novel anaeroic, dehalogenating sulfate-reducing bacterium.
Arch. Microbiol. 154:23-30.
Dragun, J. 1988. The Soli Chemistry of Hazardous Materials.
Hazardous Materials Control Research Institute, Silver Spring,
MD.
Esaac, E.G., and F. Matsumura. 1980. Metabolism of insecticides
by reductive systems. Pharmac. Ther. 9:1-26.
Fathepure, B.Z., and S.A. Boyd. 1988. Dependence of
tetrachloroethylene dechlorination on methanogenic substrate
consumption by Methanosarclna sp. strain DCM. Appl. Environ.
Microbiol. 54:2976-2980.
Fathepure, B.Z., J.M. Tiedje, and S.A. Boyd. 1988. Reductive
dechlorination of hexachlorobenzene to trt- and dichlorobenzenes
In anaerobic sewage sludge. Appl. Environ. Microbiol. 54:327-
330.
Fogel, M.M., A.R. Taddeo, and S. Fogel. 1986. Biodegradation
of chlorinated ethenes by a methane-utilizing mixed culture.
Appl. Environ. Microbiol. 51:720-724.
Freedman, D.L., and J.M. Gossett. 1989. Biological reductive
dechlorination of tetrachloroethylene and trichloroethylene to
ethylene under methanogenic conditions. Appl. Environ. Microbiol.
55:2144-2151.
Genthner, B.R.S., W.A. Price, II, and H. P. Prltchard. 1989.
Anaerobic degradation of chloroaromatlc compounds In aquatic
sediments under a variety of enrichment conditions. Appl.
Environ. Microbiol. 55:1466-1471.
Gibson, S.A., and J.M. Suflita. 1986. Extrapolation of
biodegradation results to groundwater aquifers: Reductive
dehalogenation of aromatic compounds. Appl. Environ. Microbiol.
52:681-688.
Hartsmans, S., J.A.M. de Bont, J. Tramper, and K.Ch.M.A.
Luyben. 1985. Bacterial degradation of vinyl chloride. Biotechnol.
Letters 7:383-388.
Hutzinger, O., and W. Verkamp. 1981. Xenobiotic chemicals
with pollution potential, pp. 3-46. In: T. Leisinger, A. M. Cook, R.
Hutter, and J. Nuesch (eds.). Microbial Degradation of Xenobiotic
and Recalcitrant Compounds. Academic Press, London, G.B.
Infante, P.F., and T.A. Tsongas. 1982. Mutagenic and oncogenic
effects of chloromethanes, chloroethanes, and halogenated
analogs of vinyl chloride. Environ. Sci. Res. 25:301-327.
King, G.M. 1988. Dehalogenation in marine sediments containing
natural sources of halophenols. Appl. Environ. Microbiol. 543079-
3085.
Kuhn, E.P., P.J. Colberg, J.L. Schnoor, O. Wanner, A.J.B.
Zehnder, and R.P. Schwarzenbach. 1985. Microbial transformation
of substituted benzenes during infiltration of river water to
ground water.laboratory column studies. Environ. Scl. Technol.
19:961-968.
Kuhn, E. P., and J. M. Suflita. 1989a. Dehalogenation of
pesticides by anaerobic microorganisms in soils and groundwater
- a review, pp. 111-180. In: B. L. Sawhney and K. Brown (eds.)
Reactions and Movement of Organic Chemicals in Soils. Soli
Scl. Soc. America Special Publication No. 22. Soil Sci. Soc.
America, Inc. Madison, Wl.
Kuhn, E.P., and J.M. Suflita. 1989b. The sequential reductive
dehalogenation of chloroanllines by microorganisms from a
methanogenic aquifer. Environ. Sci. Technol. 23:848-852.
Unkfleld.T.G., J.M. Suflita, and J.M. Tiedje. 1989. Characterization
of the acclimation period prior to the anaerobic biodegradation
of haloaromatic compounds. Appl. Environ. Microbiol. 55:2773-
2778.
Marks, T.S., A.R.W. Smith, and A.V. Quirk. 1984. Degradation
of 4-chlorobenzoic acid by an Arthrobacter sp. Appl. Environ.
Microbiol. 48:1020-1025.
Mikesell, M.D., and S.A. Boyd. 1986. Complete reductive
dechlorination and mineralization of pentachlorophenol by
anaerobic microorganisms. Appl. Environ. Microbiol. 52:861-
865.
Palmer, D.T., T. G. Unkfield, J.B. Robinson, B.R.S. Genthner,
and G. E. Pierce. 1989. Determination and enhancement of
anaerobic dehalogenation: Degradation of chlorinated organics
In aqueous systems. EPA/600/S2-88/054, U.S. Environmental
Protection Agency, Cincinnati, OH.
Paris, D.F., and D.L. Lewis. 1973. Chemical and microbial
degradation of ten selected pesticides In aquatic systems.
Residue Rev. 45:95-124.
Quensen, J.F. Ill, J.M. Tiedje, and S.A. Boyd. 1988. Reductive
dechlorination of polychiorinated biphenyis by anaerobic
microorganisms from sediments. Science 242:752-754.
Reineke, W., and H.J. Knackmuss. 1984. Microbial metabolism
of haloaromatics: isolation and properties of a chlorobenzene-
degrading bacterium. Appl. Environ. Microbiol. 47:395-402.
Roberts, T.R., and M.E. Standen. 1978. Degradation of the
herbicide flamprop-methyl in soil under anaerobic conditions.
Pestic. Biochem. Physiol. 9:322-333.
11
-------
Rochklnd, M.L, J.W. Blackburn, andG.S. Sayler. 1986. Microbial
decomposition of chlorinated aromatic compounds. EPA/600/2-
86/090, U.S. Environmental Protection Agency, Cincinnati, OH.
Schraa, G., M.L. Boone, M.M. Jetten, A. R. W. van Neerven,
P.J. Colberg, and A.J.B. Zehnder. 1986. Degradation of 1,4-
dichlorobenzene by Alcaligenes sp. strain A175. Appl. Environ.
Microbiol. 52:1374-1381.
Sethunathan, N. 1973. Microbial degradation of insecticides in
flooded soil and in anaerobic cultures. Residue Rev. 47:143-
165.
Sims, J.L., R.C. Sims, and J.E. Matthews. 1989. Bloremediation
of contaminated soils. EPA/600/9-89/073, Robert S. Kerr
Environmental Research Laboratory, U.S. Environmental
Protection Agency, Ada, OK.
Spain, J.C., and S.F. Nishino. 1987. Degradation of 1,4-
dichiorobenzene by Pseudomonas sp. Appl. Environ. Microbiol
53:1010-1019.
Struijs, J. and J.E. Rogers. 1989. Reductive dehaiogenation of
dlchloroanlllnes by anaerobic microorganisms in fresh and
dichlorophenol-acclimated pond sediment. Appl. Environ.
Microbiol. 55:2527-2531.
Suflita, J. M., A. Horowitz, D. R. Shelton, and J. M. Tiedje. 1982.
Dehaiogenation: A novel pathway for anaerobic biodegradation
of haloaromatic compounds. Science 218:1115-1117.
Suflita, J.M., and G.D. Miller. 1985. Microbial metabolism of
chlorophenolic compounds in groundwater aquifers. Environ.
Toxicol. Chem. 4:751-758.
Suflita, J.M., S.A. Gibson, and R.E. Beeman. 1988. Anaerobic
biotransformation of pollutant chemicals In aquifers. J. Ind.
Microbiol. 3:179-194.
Thomas, J. M„ and C. H. Ward. 1989. in situ biorestoration of
organic contaminants in the subsurface. Environ. Scl. Technol.
23:760-766.
Thiele, J., R. Muller, and F. Llngens. 1988. Enzymatic
dehaiogenation of chlorinated nitroaromatic compounds. Appl.
Environ. Microbiol. 54:1199-1202.
Tiedje, J.M., and T. O. Stevens. 1987. The ecology of an
anaerobic dechlorinating consortium, pp. 3-14. In: G.S. Omenn
(ed.). Environmental Biotechnology: Reducing Risks from
Environmental Chemicals through Biotechnology. Plenum Press,
New York, NY.
Tsuchlya,T.,andT. Yamaha. 1984. Reductive dechlorination of
1,2,4-trichlorobenzene by Staphylococcus epidermis isolated
from intestinal contents of rats. Agric. Biol. Chem. 48:1545-
1550.
Vogel, T. M., C. S. Criddle, and P. L. McCarty. 1987.
Transformations of halogenated aliphatic compounds. Environ.
Sci. Technol. 21:722-736.
Zeyer, J., and P.C. Kearney. 1982. Microbial degradation of
para-chloroanillne as sole carbon and nitrogen source. Pestic.
Blochem. Physiol. 17215-223.
Zeyer. J., A. Wasserfallen, and K.N. Tlmmls. 1985. Microbial
mineralization of ring-substituted anilines through an ortho-
cleavage pathway. Appl. Environ. Microbiol. 50:447-453.
AlJ. GOVERNMENT HUNTING OFFICE: mi • S4S-M7/MS37
)n rted Stales Center for Environmental Research BULK RATE
invironmental Protection Information POSTAGE & FEES PAID
sgency Cincinnati. OH 45268 EPA PERMIT NO. G-35
:*fficial Business
enalty for Private Use $300
: PA/540/4-90/054
-------
A One-Day Seminar On
Bioremediation Applications
July 27, 1993
Session IV
Treatability Studies
Presented By
Hugh Russell, Ph.D.
-------
LABORATORY TREATABILITY STUDIES
QUESTIONS
1. What is the biotic rate?
a. Replicates- sufficient to determine that time points
are real and some confidence can be placed on numbers obtained.
Minimum of three at each time point. Sampling and analytical
variability may cause errors in calculation of rates or percent
degradation.
LABORATORY TREATABILITY STUDIES
QUESTIONS
b. Controls- sufficient to determine all biotic losses
within a system are measured. Simple method- control same as
active only treated with a biocide to inhibit microbial population.
If no biotic loss occurs, then rate of loss of compound is equal to
total abiotic losses through volatilization, sorption etc.
-------
LABORATORY TREATABILITY STUDIES
QUESTIONS
c. Points- at least five required to determine rate.
Last point should be below level to which soil, groundwater etc. is
to be treated. Extrapolation of data ignores the plateau which
biological systems may reach. Caveat you may not wish to try and
reach clean-up goals in the lab, merely determine if bioremediation
is possible.
LABORATORY TREATABILITY STUDIES
QUESTIONS
d. Analysis- must be of the compound in which you are
interested. TPH may be utilized as surrogate for naphthalene, but
at start and end naphthalene concentrations should be determined.
If other than standard methods used, then limits of detection,
variability, etc. of analysis must be included.
-------
LABORATORY TREATABILITY STUDIES
QUESTIONS
2. Can Clean-up Goals be met?- This question is best
answered by a field pilot scale demonstration. Lab studies are
best suited to determine if a barrier exists to biological growth,
not if clean-up goals can be reached on a field scale. These
studies are conducted under ideal conditions which can never be
performed on a field scale.
3. What additives will increase the biological rate?- Can be
assessed by proper controls and protocols.
LABORATORY TREATABILITY STUDIES
QUESTIONS
Control-
Active 1-
Active 2-
Active 3-
NO amendments, inhibited
Nutrients (nitrogen and phosphorous)
Manure
Nutrients + Manure
-------
LABORATORY TREATABILITY STUDIES
QUESTIONS
4. Intermediates- In some cases daughter or intermediate
products and their ability to escape (volatilize, leach) are not
known. It is possible through proper studies to answer these
questions.
-------
A One-Day Seminar On
Bioremediation Applications
July 27, 1993
Session V
Regulatory Review
Presented By
David E. Giamporcaro, Esq.
-------
Federal RCRA Regulations Relevant to
Bioremediation
¦ Treatability Sample Exclusion Rule (40 CFR 261.4 (e)-(f)
'Allows for treatability studies under RCRA-
¦ Research Development & Demonstration Permit (40 CFR 270.65)
•Allows the issuance of a RCRA permit for pilot scale study
pertaining to an innovative or experimental technology-
¦ Subpart X Miscellaneous Units (40 CFR 264.600)
-Allows the issuance of a RCRA permit for a miscellaneous
unit-
¦ RCRA Permit Modification Rule: Temporary Authorization
(40 CFR 270.42(e))
-Allows the permitting agency to grant a facility temporary
authorization to perform certain activities far up to 180 days
(corrective action, cleanup St closure activities)-
-------
FEDERAL REGULATIONS AND GUIDANCE RELEVANT TO
BIOREMEDIATION
CITATION
REGULATION
DESCRIPTION
GUIDANCE I
40 CFR 261.4(e)-(f)
Treatability Study Exemption
Allows for treatability studies
; under RCRA
Conducting Treatability Studies
Under RCRA (7/92, OSWER
Directive 9380.3-09FS, NTIS
PB92-963-501)
40 CFR 270.65
Research Development and
Demonstration Permits
i Allows the issuance of a RCRA
; p«rmit for a pilot scale study
pertaining to an innovative or
experimental technology
Guidance Manual for Research
Development and
Demonstration Permits Under
40 CFR Section 270.65 (7/86,
EPA/530-SW-86-008, OSWfiR
Directive 9527.00-1 A)
40 CFR 264.600
Subpart X Miscellaneous Units
Altews the issuance of a RCRA
permit for a miscellaneous unit
ll
40 CFR 270.42(e)
RCRA Permit Modification Rules
Temporary Authorization
Allows the permitting agency to
grants facility a temporary
authorization to perform certain
activities (e.g., cleanups,
i corrective action and closure
activities) for up to 180 days
Modifying RCRA Permits (9/89, I
E P A/530-S W-89-050)
!
¦
-------
FEDERAL REGULATIONS AND GUIDANCE RELEVANT TO
BIOREMEDIATION (continued)
CITATION
REGULATION
f DliCRIPTION
GUIDANCE |
40 CFR 268.40
Land Disposal Restrictions
(LDR) Subpart D • Treatment
Standards
Sets forth RGRA hazardous
waste treatment standards
Land Disposal Restrictions |9
Summary of Requirements IS
(2/91, OSWER Directive jl
9934.0-1 A) j I
40 CFR 268.45
(August 18,1992;
57FR37279)
Treatment Standards for
Hazardous Debris
Dts<5M88§e Dlologloal destruction
of nazanfcus constituents from
debrts surface
40 CFR 268.44(h)
Variance from an LOR
Treatment Standard
Allows for a site-specific
treatability variance to be Issued
as « nonrulemaklng procedure
Regional Guide: Issuing i I
Site-Specific Treatability j 1
Variances for Contaminated I
Soils and Debris from LDRs nj
(1/92, OSWER Directive I
9380.3-08FS)
40 CFR 260, 264.552
etal. (February 16,
1993; 58FR8658)
Corrective Action Management
Unit (CAMU)
Encourages treatment, Including
use of Innovative treatment
(specifically bloremediatlon),
Instead of containment
Environmental Fact Sheet:
EPA Issues Final Rules for
Corrective Action Management |
Units and Temporary Units |
(1/93, EPA/530-F-93-001) 1
-------
OSWER: PROPOSED REGULATORY SCHEME FOR BIOREMEDIATION OF SOILS
Concentration
_ 10X Universal
— Universal
Compliance _ A ~
Estimates 74%
82%
86%
84%
92%
Key
- Meets BDAT
- Meets BDAT if 90% reduction achieved
SOIL BDAT DEVELOPMENT
-------
EPA'S TSCA SECTION 5
BIOTECHNOLOGY
PROGRAM
Chemical Control Division
Office of Pollution Prevention & Toxics
U.S. EPA
-------
TOXIC SUBSTANCES
CONTROL ACT
(TSCA)
May regulate chemical substances
which present unreasonable risks
of injury to health or the environment
"Chemical Substance" Defined
Any organic or inorganic substance of a
particular molecular identity, including substances:
(1) occurring in whole or in part from a
chemical reaction,
OR
(2) occurring in nature
Microorganisms are included in this definition
-------
NEW MICROORGANISMS
SUBJECT TO TSCA SECTION 5
ACCORDING TO THE 1986
POLICY STATEMENT
• Intergeneric - contain genetic material
from source organisms
in different genera
• Not listed on TSCA Inventory
Exclusions
« Naturally occurring microorganisms
• Genetically modified microorganisms other
than intergeneric
• Intergeneric microorganisms resulting
only from addition of well-characterized,
non-coding regulatory regions
-------
DETERMINING TSCA SECTION 5
OBLIGATIONS UNDER THE 1986
POLICY STATEMENT
No
TSCA Jurisdiction ~ OUT
Yes
v
No
New Microorganism OUT
(Intergeneric)
Yes
r
R&D
Yes i No
i y
Contained PMN Required
Yes | No
i ^
R&D Exemption Voluntary PMN
Section 720.36 or Section 720.36
-------
MICROORGANISMS
SUBJECT TO TSCA
Environmental, industrial, or consumer
uses that are not specifically excluded.
Exclusions:
• pesticides (but not pesticide intermediates)
• foods, food additives, drugs, cosmetics &
their intermediates
• tobacco & tobacco products
• nuclear materials
-------
POTENTIAL USES OF
MICROORGANISMS
SUBJECT TO TSCA
• Specialty chemical production
• Nitrogen fixation
• Bioremediation
• Biosensors
• Biomass conversion
• Mineral recovery
-------
DETERMINING TSCA SECTION 5
OBLIGATIONS UNDER THE 1986
POLICY STATEMENT
TSCA Jurisdiction
No
OUT
Yes
New Microorganism
(Intergeneric)
No
OUT
Yes
R&D
Yes | No
r |
Contained PMN Required
Yes
| No
f
1
R&D Exemption
Voluntary PMN
Section 720.36
or Section 720.36
-------
DETERMINING TSCA SECTION 5
OBLIGATIONS UNDER THE 1986
POLICY STATEMENT
TSCA Jurisdiction
No
OUT
Yes
New Microorganism
(Intergeneric)
No
OUT
Yes
Yes
R&D
Contained
No
1
f
PMN Required
Yes
No
r
1
r
R&D Exemption
Voluntary PMN
Section 720.36
or Section 720.36
-------
DETERMINING WHETHER
A MICROORGANISM IS
CONTAINED UNDER THE
1986 POLICY STATEMENT
A microorganism will be considered
environmentally contained if:
(1) The microorganism is used in a laboratory
that complies with NIH RAC guidelines.
OR
(2) The microorganism is used in a contained
greenhouse, fermentor, or other contained
structure.
-------
"Contained greenhouse, fermentor,
or other contained structure" means
a building or structure that has a
roof and walls
It should also have the following:
(1) A ventilation system to minimize microbial
release to the outdoors.
(2) A system for sterilizing water runoff
and wastes.
(3) A system for restricting insects, if these
are plausible routes for dissemination
of microorganisms.
(4) Experimenters should control pests,
sterilize soil or other material containing
microorganisms before disposal or reuse.
(5) Access limited only to those persons who
must have access for research purposes.
-------
BIOREACTOR R&D USES ELIGIBLE
FOR THE SMALL QUANTITIES EXEMPTION
FROM PMN REPORTING UNDER TSCA
Organization Microorganism Modification Chemical Degraded
Envirogen Escherichia coli K-12
SBP Pseudomonas cepacia
Pseudomonas mendoclna TCE
degradation genes
Tn5 mutant TCE
constitutive degradation
Armstrong Labs
Air Force
Pseudomonas cepacia
Tn5 mutant
constitutive degradation
TCE
-------
INTERGENERIC
MICROORGANISMS USED
IN R&D FIELD TESTS
Recipient
Microorganism
Modification
#PMNs
# Field
Tests
Rhizobium meliloti
N-fixation, Antibiotic -
resistance markers
18
4
Bradvrhizobium
japonicum
N-fixation, Antibiotic
resistance markers
6
2
Pseudomonas
aureofaciens
lacZY marker genes
from Escherichia coli
1
2
Voluntary PMNs were received for small-scale field
tests of the above microorganisms.
TSCA section 5 (e) consent orders were negotiated
for all PMNs to restrict uses to the specific field tests.
Additional testing of the same microorganisms
required modifications to the consent orders.
Different combinations of microorganisms were used
for the various field tests.
EPA coordinated with Regional and State contacts
during consent order negotiations.
-------
INTERGENERIC
MICROORGANISMS USED AS
INTERMEDIATES IN SPECIALTY
CHEMICAL PRODUCTION
Recipient
Microorganism
Modification
Use
AsDeraillus orvzae
Lipase
Detergent
Bacillus spp.
Subtilisin
Detergent
Bacillus falcalophilic^
Protease
CBI*
Bacillus licheniformis
Hydrolase
CBI
it a
Alpha amylase
Detergent
tl 6i
Alpha amylase
Detergent
Bacillus subtiiis
Protease
CBI
c< <(
Alpha amylase
Produce ethanol
for gasoline
u a
Lipase
Detergent
Escherichia coli
Insulin-like
Cell culture
growth hormone
media
* CBI = Confidential Business Information
PMNs were received for the above contained system fermentation uses.
No regulatory action was taken on these microorganisms subsequent to
their review.
-------
A One-Day Seminar On
Bioremediation Applications
July 27, 1993
Session VI
Biological Treatment of
Contaminated Water
Presented By
David K. Stevens, Ph.D.
-------
Bioprocess Workshop»
Contaminated Water Treatment
David K. Stevens
Utah State University
Utah Water Research Laboratory
Logan, UT 84322-4110
D
-------
Introduction
B—MBttttaattatMimmg—MnMnn««miiiMiiiiiiiiiiiiinniiiiiiiiiinninBatt
• Biological vs. P/C Treatment
• Tank vs. In-situ Treatment
• Approach to Session
Description
• Native organisms from aquifer used to
populate reactors
• Systems may be enhanced by 'special'
organisms
• Opportunities may exist for 'engineered'
organisms since treatment is confined
i
-------
Ecology
• Bacteria
o do most of the work in biological treatment
reactors
o many types may be present or the colonies
could be narrowly populated
Microorganisms
¦¦¦¦¦¦¦¦¦¦¦¦¦wwiwwiwiwwwwwwiiwiwiiwwwiwwwwwwiwiwwwiiiiiiiiiHMiiiiiiiiiiiiiiiHiiiiiiiiitniiMtftffitiiiiHttttitttrittrftirrHmiiiWH
Ecology
• Fungi, Protozoa, Nematodes, worms, etc.
oprimary role - maintain ecological balance
o seldom contribute to contaminant
degradation
o may be important for some xenobiotic
chemicals - e.g. white rot fungus and TNT
2
-------
Process Definition
• Suspended growth
• Fixed-film
• Sequencing batch reactors
Suspended. Growth
• organisms suspended in the liquid
• 'see' measured concentrations of the
substrates (food) and xenobiotics
(contaminants)
• rates are low and depend on the
concentration and residence time of the
organisms in the system
3
-------
Processes - Sketch
Suspended, Growth
Q, Cj
C„ V
Q, Ct
Processes - Sketch
iwtrntninmmmniimnwwwwH
Suspended Growth
Activate Sludge or Anaerobic Contact
Effluent
>-
Biological Solids Recycle
Discharge
Sludge Disposal
4
-------
Fixed-growth
• Microorganisms attached to solid
support
• High concentrations of biomass
• Considered 'high-rate' processes
• Mass transfer limits removal
Processes
Fixed-growth Media
• plastic support - many styles
o sheets
o random
o nylon fibers
• rock - trickling filter (old fashioned)
• sand - fluidized beds sgl
• activated carbon - fixed and fluidized ™
^fll 9 V
5
-------
Processes — Schematic
smmtitimstttu
Fixed-film Reactor
Fixed-film
Reactor
9 9 9 9 9
19 9 9 0
9 9 9 9 9
>9 9 9 9
9 9 9 9_9^
• 999
9 9_9
i 9
9 9
> 9 9 _
9 9 9 9
>9 9 9 9
9 9 9 9 9
19 9 9 9
9 9 9 9 9
19 9 9 9
9 9 9 9 9
I 9 9_9_9
A /V
J
Biofilm
Biofilm Support
Processes -- Characteristics
Fixed-film Reactor
• Biomass concentration — 10,000 to
50,000 mg/L
• Specific surface area — 100 -1000 m2/m3
• Degradation rates up to an order of |
magnitude higher than suspended 1
growth -- Very short HRTs j§|
^ IH B 91 9 H
6
-------
Sequencing Batch Reactors (SBRs)
• Fill and draw system
• Very effective for contaminants that are
difficult to degrade in continuous growth
system
o biomass adapts well to conditions
o relatively easy to control growth phase of
organisms
Processes - Schematic
SBRs
7
-------
Metabolism
¦""»
Substrates
• hydrocarbons
o alkanes, alkenes
o aromatics - BTEX, < 3 ringPAHs, others
• alcohols
• carbohydrates
• yes, even PCP
• many others
Metabolism
Xenobiotics
• Xenobiotic or 'pure' contaminant -
Cometabolism
o 3+ ring PAH
E
o VC/DCE/TCE/PCE \
°pcB ||
tflflff 3BI M 2891 WSM
91 1^1 m 19 !¦
8
-------
Xe nob io tics
• Non-specific enzymes
o white rot fungus - lignin peroxidase
o methanotrophs - methane mono-oxygenase
• Specific enzymes
o E. coli — mercuric reductase
Philosophy
• 'Empirical' approaches
• Mechanisic models
• Role of statistics and hypothesis testing
iiiIIMIIIIIIIIIIIIIIIIIIIIIIIII
9
-------
SBR
Reactors operating in parallel
Water is mixed with biomass present in
the reactor during fill cycle
Liquor is mixed by aeration (aerobic)
or gas recirculaton (anaerobic)
Processes - Uses
SBR i
• Conventional WW treatment ]
a
|
• Degradation of TCE by methanotrophs j
• Used for GW remediation at Love
Canal
• Others - nitrogen control, municipal J|§
wastewater J IB IB
- - - p H H H |
10
-------
SBR
High rate vis a vis continuous mixed
suspended growth systems
Can be more effective in treatment of
'specialized' contaminants - co-
metabolism
Controls can be easily automated using
programmable controllers and
Process Modeling
Purpose
• System design and control
o prediction of process performance
o testing influence of design variables
o improvement and optimization of
operations
11
-------
Process Modeling
Purpose
• Improved understanding
o Aid in correlating experimental and process
results
o Testing understanding of process
mechanisms
o Finding 'unexplained'phenomena B
n m m m in
Approach
• Mass and energy balances
o Flow and transport of contaminants
o Degradation and phase transfer
• Nature of process
o Single phase
o Multi-phase
o Steady or un-steady state
12
-------
Process Modeling
Kinetics
o Growth rates - usually use Monod or
Michaelis-Menten kinetics for growth
o Substrate removal generally regarded as
proportional to growth
O Inhibition can cloud the issue
o Approach known as unstructured modeling
Kinetics
• Many types of inhibition
o Substrate
o Product
o Indifferent - competitive, uncompetitive, etc.
o 'Toxicity' - bactericides
• Classical models often work well
13
-------
Process Modeling - Co-metabolism
Kinetics
• Need a multi-stage approach
1) growth of organisms
2) production of degradation enzymes
3) destruction of contaminant
4) fate of breakdown products
• Approach known as structured modeling
Contaminant transport
• Forms the basis of process models
• Fundamentals of mass and energy
balances
• Processes
oAdvection (flow)
o Dispersion (diffusion + mixing)
o Mass transfer across phase boundaries
14
-------
Process Modeling
Contaminant transport
• Process models similar to natural systems
• Easier in many ways because the process
is controlled
• Considerable work done on bioprocess
modeling in 1970's and 1980's
Process Modeling - Basis of models
Contaminant transport
• Reactor models - continuous
o plug flow - advection only (no axial mixing)
o CSTR - mixing only (no axial gradients)
o in-between - advection plus axial mixing |
¦¦BIB
15
-------
Process Modeling
Contaminant transport
• Reactor models - batch
o true batch
o sequencing batch
Contaminant transport
• Complete models concerned with
o electron donors (food)
o electron acceptors (02 NO/, others)
o target contaminants
o biomass
o buffering system (alkalinity; acidity, pH)
16
-------
Contaminant transport
• Few models commercially available
• Consultants
o in-house approaches may include models
o design models often empirical
o can be proprietary although many empirical
approaches have been published
Process Modeling
Contaminant transport 1
• Research
o models often used in correlating data and
designing experiments
o often have large data requirements j
o not commonly used in design
!Bflf mm WKM WM wbk
!¦ ¦ ™ ^
1
-------
mineralization by a Picmbmomis sp.: Effects of second substrates." Appl. Envir.
Microbiology, 53(11), 2517-2623.
Shamat, N. A., and Maicr, W. I. (1980). "Kinetics of biodegradatinn of chlorinated
organics." J. Water Pollution Control Federation, 52(8), 2158-2166.
Sok6T, W. (1988). "Uptake rate of phenol by Pseudomonas putida grown in unsteady
Mate." Biotech, ffioengrg., 32(9), 1097-1103.
Sleber, J., and Wierich, P. (1986). "Properties of hydroxyethane diphosphonate af-
fecting its environmental fate; Dcgradability, sludge adsorption, mobility in soils
and hioconcentralion." Chemosphere, 15(7), 929-945.
Stenstrom, M. K., Cardinal, L., and Libra, J. (1989). "Treatment of hazardous sub-
stances in wastewater treatment plants." Envir. Progress, 8(2), 107-112.
Stevens, O. K. (1988). "Interaction of mass transfer and inhibition in biofilmx." J.
Envir. Engrg., ASCE, 114(6), 1352-1358.
Suidan, M. T., ct nl. (1988). "Anaerobic hiodegradation of phenol: Inhibition ki-
netics and system stability." J. Envir. Engrg., ASCE, 114(6), 1359-1376.
Tacgcr, K., Knackmuss, H.-J., and Schmidt, E. (1988). "Biodegradability of mix-
tures of chloro- and methylsubstituted aromatics: Simultaneous degradation of 3-
chlorobenzoate and 3-methylbenzoate." Appl. Microbiology Biotech., 28(6), 603-
60R.
Templeton, L. L., and Grady, C. P. L., Jr. (1988). "Effect of culture history on the
determination of biodegradation kinetics by batch and fed-batch techniques." J.
Water Pollution Control Federation, 60(5), 651-658.
Tcipp, E., Crawford, R. L., and Hanson, R. S. (1988). "Influence of readily me-
taboliznble carbon on pentachlorophenol-degrading Flavobacterium sp." Appl. En-
vir. Microbiology, 54(10), 2452-2459.
Volskay, V. T., Jr. (1988). "Respiration inhibition kinetics assay: A microbial tox-
icity assay," thesis presented to Clemson University, at Clemson, S C., in partial
fulfillment of Ihe requirements for the degree of Master of Science.
Wang, X. (1988). "The development of a nonlinear parameter estimation technique
for use with hiodegradation data," report presented to Clemson University, at
Clemson, S.C., in partial fulfillment of Ihe requirements for the degree of Master
of Engineering.
Weightman, A. J., and Slater, J. H. (1988). "The problem of xenobiotics and re-
calcitrance." Microorganisms In action: Concept and applications in microbial
Ecology, J. M. Lynch and J. E. Ilobbie, cds., Blackwell Scientific Publications,
Oxford. England 322-347.
Williams, D. R. (1972). "Kinetics of heterogeneous substrate removal by natural
microbial populations in continuous culture," thesis presented to Purdue Univer-
sity, at West Lafnycttc, Ind.. in partinl fulfillment of the requirements for the
degree of Master of Science.
Appendix II. Notation
The following symbols are used in this paper:
K, = inhibition coefficient, ML'*;
Ks = half-saturation coefficient, ML"';
m — exponent depicting effect of substrate or inhibitor on Ks\
n = exponent depicting effect of substrate or inhibitor on jL;
q - specific substrate-removal rate, 7"';
S, = inhibitor concentration, ML'y;
5* = critical inhibitor concentration above which growth stops. AfL"';
Sx = concentration of substrate being biodegraded, ML'S;
Sj = critical substrate concentration above which growth stops, ML ;
K = true growth yield, dimcnsionlcss;
H. = specific growth rate of microorganisms, 7"-1; and
|1 = maximum specific growth rate of microorganisms, T~
826
Simple Solutions for Steady-State
Biofilm Reactors
By Padma S. Golla,1 Associate Member, ASCE,
and Thomas J. Ovcrcamp'
Abstract: Simple, analytical models for use in preliminary calculations are de-
veloped In approximate Ihe steady-Mute concentration and biofilm thickness in ¦
biofilm reactor. The model* consider the case in which ¦ single substrate limits
growth these models a'siime that flux of substrate to Ihe biofilm has a first oriler
dependence on Ihe hulk substrate concentration. This assumption restricts Ihe ute
of Ihe models to low initial substrate concentrations. One or the models assumes
plug flow, and Ihe other incorporates longitudinal dispersion of substrate. The models
allow the substrate concentration to approach Ihe minimum rale-limiting concen-
tration and Ihe biofilm thickness to approach rero as predicted by the steady-state
theory. Predictions of these models are compared lo those of a steady-slate nu-
merical model developed hy Rittmann. For an example taken from the literature,
the solution of Ihe model with longitudinal dispersion is in good agreement with
Ihe numerical solution
Introduction
A biofilm is an aggregation of bacteria and their exopolymcrs that is at-
tached to a solid surface. Environmental engineering applications of biofilms
include wastewater applications such as trickling filters, rotating biological
contactors, and fluidized bed reactors as well as new applications such as
in situ biological treatment of contaminated ground waters and denitrification
of drinking waters.
When the influent conditions to a fixed-film system are constant, the bio-
film attains steady state. The steady-state biofilm model assumes that growth
is balanced by loss due to decay and shear. At steady state, the thickness
of the biofilm is determined by the flux of substrate to the biofilm and the
growth nnd decay rates of the bacteria. Due to decay and shear losses, a
minimum substrate concentration, S„lH, is required to sustain a steady-state
biofilm (Rittman and McCarty 1980a).
This paper develops simple, analytical models that can be used for pre-
liminary design of biofilm reactors. The models arc based on the assump-
tions that n single substrate limits growth and that the flux of substrate to
the biofilm has a first-order dependence on the bulk substrate concentration.
The latter assumption restricts the use of the models to low concentrations.
Steady-state solutions are given for substrate concentration and biofilm thick-
ness as a function of distance along the reactor for both plug flow and flow
with longitudinal dispersion. These analytical solutions are compared to a
numerical solution developed by Rittman ct al. (1986).
Theory
Six major processes govern a steady-state biofilm (Rittman and McCarty
1980a). The first process is the diffusion of substrate from the bulk liquid
'Ashbrook-Simon-Harllcy, P.O. Box 16327, Houston, TX 77222.
'Prof., Envir. Systems Engrg., Clemson Univ., Clemson, SC 29634-0919.
Note. Discussion o|>en until March 1, 1991. To extend ihe closing date one month,
a written request must be filed with the ASCE Manager of Journals. The manuscript
for this paper was submitted for review and possible publication on June 19, 1989.
This paper is part of the Journal of Environmental Engineering, Vol. 116, No. 5.
September/October, 1990. ©ASCE. ISSN O733-9372/9O/O0O5-O829/S1.00 + $.15
per page. Paper No. 25095.
829
-------
U> .lie hiofilm's surface. 'I'his can he viewed as occurring across a stagnant
liquid layer of depth L (length L). The steady-state to the biofilm is given
by
"(V)
(I)
in which S„ = the bulk substrate concentration (M,L~J); S, = the intcrfacial
substrate concentration (Af,L~J); J = the substrate flux to the biofilm
L = the depth of the diffusion layer in the liquid (Ly, and D
= the molecular diffusivity of the substrate in the liquid {L T~').
The second process is substrate utilization, and the third is diffusion of
substrate within the biofilm. For steady-state conditions and Monod kinetics,
the substrate concentration within the biofilm is given by
a's, 4X,S,
Of—y = -LJ~L (2)
' at1 k, + sf
in which Df = the molecular diffusivity of substrate in the biofilm (L'T-1);
4 ~ the maximum specific substrate utilization rate Xf = the
active biomass density Sf - the substrate concentration in the bio-
film K, = the half-velocity coefficient (A/,L~3); and r = the distance
normal to the biofilm (L).
The boundary conditions for Eq. 2 arc
95,
— = 0 at 2 = 0 (3a)
fir
and
Sf = S, at t = Lf (3b)
where L, = the biofilm thickness (Z.).
The fourth process is growth, and the fifth is loss of biomass due to decay
and shear. By equating growth of the biofilm to losses due to decay and
shear, and the steady-state thickness is given by
YJ
L, = (4)
' b'X,
in which Y = the true growth yield and b' = the loss coefficient
due to decay and shear (7""1) as given by Rittmann and McCarty (1980a)
nnd Rittmnnn (I982;i).
In a very long reactor, the substrate falls to a minimum concentration:
- «».i (5)
in which Smm — the concentration required to support a single layer of cells
at steady-state conditions.
The sixth process is the transport of substrate down the reactor. This is
described by the steady-state, advection-dispersion equation in the bulk liq-
uid:
830
in which v = the superficial velocity of the bulk liquid (L7"'); x *=¦ the
distance along the reactor (L); = the hydrodynamic dispersion coefficient
and a = the specific surface area of the reactor (/,"'). The boundary
conditions chosen by Rittmann (1982b) are
vS/a = vSh — £>» — at x — 0 (7a)
fix
and
as„
— = 0 at x = L, (7b)
9x
in which = the substrate concentration fed to the reactor (M,L~*); and
L, = the reactor's length (L). The inlet boundary condition is the constant
flux boundary condition that assumes that the flux due to advection and dis-
persion equals the flux of substrate fed to the reactor. When Dh is small,
the inlet substrate concentration is nearly equal to the concentration fed to
the reactor, 5M. When Dk is large, the inlet substrate concentration is less
than 5m. The outlet boundary condition assumes that there is no change in
the substrate concentration upon leaving the reactor (Danckwerts 1953).
Steady-State Biofilm Flux
The substrate flux into a steady-state biofilm is determined from the so-
lution of Eqs. I and 2 subject to the boundary conditions given in Eqs. 3a
and 3b. A variety of techniques including numerical methods, analytical so-
lutions for special cases, or approximate solutions have been developed to
estimate the flux.
Rittmann and McCarty (1978) developed a procedure based on an ana-
lytical solution for substrate flux into a deep biofilm, which is defined as
one in which the substrate concentration goes to zero at z — 0. Based on
the work of Williamson and Chung (1975), they showed for Monod kinetics
that the flux into a steady-state, deep biofilm is
<•<
In this expression, the flux is a function of the intcrfacial substrate concen-
tration, S,. By combining Eqs. I and 8, they calculated the steady-state flux
to a deep biofilm as a function of the bulk concentration, Sb. By examining
the behavior of the solutions, Rittmann and McCarty (1978) proposed that
the flux could be given by nth- or variable-order kinetics:
J = CJl (9)
where C„ = the reaction coefficient (Af|""L'"",7~'); and n = the reaction
order. They observed that the reaction order varied from first order at low
concentrations to one-half order at high concentrations. By fitting equations
831
-------
to the solution for a tlcep biofilm, the reaction order was given by the expres-
sion
n = 0.75 - 0.25 tanh (0.477A) (10)
The parameter A = the logarithm of an adjusted, dimensionless concentra-
tion:
A = In (5?) - In 2 + - 1.8(1(1 (I + 2L*Df)\ + 0.353 (II)
L 2.303 J
in which St = Sh/K,\ D* = D,/D\ L* = L/hf \and 8, = \2K,Df/(qX,)]u\
The terms 5*. Of, and L* are the dimensionless substrate concentration,
biofilm dilTusivity, and diffusion layer depth, respectively.
When n is unity, the reaction coefficient is given by
2D,
C' ~ (2'% + 2LD*) ('2)
In later work, Rittmann and McCarty (1980a) extended this approach to
shallow biofilm in which the substrate concentration does not go to zero at
z = 0. They proposed a correction factor for the substrate flux to shallow,
steady-state biofilms by fitting an expression to numerical solutions. The
steady-state flux can be computed for shallow biofilms using an iterative
procedure based on Eq. 9 and their shallow biofilm correction factor. More
recently, Sflez and Rittmann (1988) developed a revised shallow biofilm cor-
rection factor to be used with the analytical solution for deep bionims (Eq.
10). With either shallow biofilm correction factor, the substrate flux to the
biofilm goes (o zero as 5» approaches 5mi„.
Numerical Solution for Biofilm Reactors
Rittmann ct al. (1986) developed a computer program, RESTST, for mod-
cling a steady-state biofilm reactor. RESTST computes the final concentra-
tion, substrate flux, and biofilm thickness by solving the advection-disper-
sion equation, Eq. 6, with the substrate flux to the biofilm given by the
variable-order model, Eq. 9, and the shallow biofilm correction of Rittmann
and McCarty (l9K0a). The solution method is an iterative procedure subject
to the constraint that Sh would not fall lower than
Analytical Model
The solution presented in this paper is based on a linearized form of the
variable-order model of Rittmann and McCarty (1978) instead of the more
recent work of Sdcz and Rittmann (1988) because the newer solution uses
Eq. 8 to calculate the deep biofilm flux rather than a first-order form, Eq.
9 with n = I, as is needed for this method. In addition, solutions are com-
pared to the predictions of RESTST which is based on the earlier work.
So that the analytical model is consistent with the steady-state biofilm
model, it is assumed that the bulk substrate can be divided into reactive
substrate, (S» - •^mln) i and unrcactivc substrate, 5miB:
S*U) = (SM - Sn-aVU) + Sml„ (13)
832
in which .S"(.r) = the reduced normalized substrate concentration defined as:
approaches 5mi„. the predicted flux to the biofilm
approaches zero. Although this decrease is somewhat similar to that pre-
dicted using a shallow biofilm correction factor, this method should not be
viewed as a rigorous alternative to a shallow biofilm correction factor ap-
proach.
For low substrate concentration, n is approximately unity. If n = I, the
steady-state, advection-dispersion equation can be written in terms of 5* as
dS' d7S'
0 = -v — + D„ —r - aC|5' (15)
dx dx
which is linear in S'.
For the case of no longitudinal dispersion or plug flow, Eq. 15 simplifies
to a first-order differential equation. For the initial condition that S' = 0 at
x = 0, the solution for the reduced normalized substrate concentration is
S"(.r) = exp ^ C' j (16)
The solution for dispersion with first-order decay for a reactor of finite
length has been given by Danckwerts (1953) and others. The reduced nor-
malized substrate concentration is given by
Tc + I (c - I)1 1
exp (-cPe) I
L 2 2(c + I) J
+ IrrlJ "p (tI^ " HI
= = : = (17)
Tc + t tf - i1
where
c=f|+^ V"
V V1 /
and Pe = the I'eclet number based on the overall length of the reactor:
vL,
P"K <19>
Results
An example is presented to compare these analytical models to solutions
of RESTST. The example is based on the parameters of experiment BCI of
Rittmann and McCarty (1980b), who showed that their numerical solution,
which was a piedecessor of RESTST, agreed well with experimental data
for both substrate concentration and biofilm thickness. Since the order of
833
-------
FKJ. 1. Steady-Slate Solution* ¦¦ Function of Distance along Blofllm Reactor
lor: (a) Subatrate Concentration; and (b) Blofllm Thlckneea
the reaction, n, was found to be 0.9S. the assumption that n is unity and
that substrate flux is proportional to the bulk substrate concentration should
be approximately correct. For this case, Pe = 32.5, which implies that both
advcclion and dispersion are important.
Fig. I gives the steady-state results for concentration and biofilm thickness
obtained from the numerical model and the two analytical models. The square
symbols arc the values calculated using RESTST. The solid line is the so-
lution incorporating longitudinal dispersion, Gq. 17, and the dashed line is
Tor the plug flow solution, Eq. 16. Fig. 1(a) shows that the concentration
profile is almost identical for the predictions of RESTST and of the analyt-
ical model incorporating longitudinal dispersion. The concentration predicted
by the plug flow model is higher in the inlet region than that given by the
other solutions. Higher concentrations result in proportionately higher flux
to the biofilm. Since higher flux results in greater substrate utilization, the
concentration for the plug flow solution at greater distances down the reactor
is less than for those models incorporating longitudinal dispersion. Fig. 1(b)
shows the steady-state biofilm thickness. This shows that the thickness pre-
dicted by the analytical model with dispersion and that predicted by RESTST
arc almost identical. On the other hand, the thickness predicted by the plug
flow model is thicker at the entrance and thinner at distances further down
the reactor.
Due to the low initial substrate concentration in this example, the substrate
concentration falls to 5mh) and the biofilm thickness goes to zero approxi-
mately half way down the reactor. Therefore, the deep biofilm assumption
834
cannot be valid throughout the reactor, and a shallow biofilm correction fac-
tor is needed to predict the flux. RESTST uses a shallow biofilm correction
factor to lower the substrate flux. In the simple model, the use of (S» - 5„,„)
in place of in Eq. 9 also lowers the substrate flux. Although the two
methods give similar results for this example, the use of the reactive sub-
strate to compute the flux is not a rigorous, alternative approach to modeling
the flux to a shallow biofilm. The method proposed in this paper docs pro-
vide a simple model for use in preliminary calculations. Final calculations
should be made with a model such as RESTST, an updated model based on
the work of Sdez and Rittmann (1988), or a complete numerical solution of
the equations.
Summary and Conclusions
Simple, analytical models were developed to predict the steady-state con-
centration and biofilm thickness in a biofilm reactor that is limited in growth
by a single substrate. Due to the assumption or a first-order relationship
between bulk substrate and the flux of substrate to the biofilm, the models
are for low substrate concentrations. One model assumes plug flow, and the
other assumes that there is longitudinal dispersion along the reactor.
The analytical models were compared to the results predicted by the steady-
state numerical model of Rittmann et al. (1986) for a case given by Rittmann
and McCarly (1980b). The predictions of the model with longitudinal dis-
persion showed good agreement with the results of the numerical model. In
comparison, the plug flow model overpredictcd the concentration and bio-
film thickness at the inlet of the reactor and underpredictcd these variables
further downstream. This was due to the neglect of dispersion in the plug
flow model that was significant for this case. For cases with lower Peclct
numbers, such as may occur in modeling biofilm growth for in situ reme-
diation of aquifers, the differences between the models will be larger.
Although limited to low initial substrate concentrations, these models pro-
vide simple tools to estimate the substrate concentration and biofilm thick-
ness along a biofilm reactor such as described by Rittmann and McCarty
(1980b). As with any simplified engineering model, final calculations should
be made with a model that includes a shallow biofilm corrections factor and
a more general expression relating substrate flux to the biofilm and substrate
concentration.
Appendix I. References
Danckwerts, P. V. (1953). "Continuous flow systems: Distribution of residence times."
Chemical Engrg. Scl., 2(1), 1-13.
Rittmann, B. E. (1982a). "The effect of shear stress on biofilm loss rate." Biotech-
nology and Bioengineering, 24(2), 501-506.
Rillmann, B. E. (1982b). "Comparative performance of biofilm reactor types." Bio-
technology and Bioengineering. 24(6), 1341-1370.
Rillmann, B. E., Kampmeier, D. T., and Chang, H. T. (1986). "Trends of biofilm-
modeling research wilh computers," presented at the 5th Conf. on Envir. Edu-
cation, Association of Environmental Engineering Professors, Michigan Techno-
logical Univ., Jul.
Rittmann, B. E., and McCarty, P. L. (1978). "Variable-order model of bacterial-
film kinetics." J. Envir. Engrg. Div., ASCE, 104(5), 889-899.
835
-------
Kiltmann, H. li., anj McCarly. J\ L. (IVMia). Mooei oi fticauy-Miiic-niutiini ni-
neties." Biotechnology and Bioengineering, 22(11), 2359-2373,
Ritlmann, B. E., and McCarty, P. L. (1980b). "Evaluation of steady-slate-biofilm
kinetics." Biotechnology and Bioengineering, 22(11), 2343-2357.
Stfcz, P. B., and Ritlmann, B. E. (1988). "Improved pseudoanalytical solution for
steady-state biofilm kinetics." Biotechnology and Bioengineering, 32(3), 379-385.
Williamson, K. J., and Chung, T. H. (1975). "Dual limitations of substrate utili-
zation kinetics within bacterial films." Presented at the 49th Annual Meeting,
American Institute of Chemical Engineers, Mar.
Appendix II. Notation
The following symbols art used in this paper:
a = specific surface area of the reactor, L~';
b' = loss coefficient of biomass due to decay and shear, 7"';
C„ = variable-order reaction coefficient for reaction order n,
T-'i
c = velocity adjustment coefficient;
D = molecular diffusivity in bulk liquid,
D, = molecular diffusivity in biofilm, L'T"1;
D* = dimensionless diffusivity = D/D,\
Dh = hydrodynamic dispersivity, L*T~';
J = substrate flux, M,L~7T~K,
K, = half-velocity coefficient, M,L'S\
L = depth of effective diffusion layer, L;
L* = dimensionless diffusion layer depth = L/bf;
L, = biofilm thickness, L;
Lj = dimensionless biofilm thickness ¦» Lf/bf\
Lr = length of reactor, L;
n ~ reaction order;
$ = maximum specific substrate utilization rate, W.M,"1 T~';
Sh - rate-limiting bulk liquid substrate concentration,
Sm = input substrate concentration,
5* = dimensionless bulk liquid substrate concentration - Sb/K,\
Sf - rate-limiting substrate concentration in biofilm,
Smin — minimum rate-limiting substrate concentration in bulk liquid, M,L'S\
S, — rate-limiting substrate concentration at the interface, M,L~
S' — reduced normalized substrate concentration;
v =* superficial flow velocity, LT~';
X( = active biomass density,
x = length along reactor, L\
Y = true growth yield,
z = distance normal to the biofilm, L;
hf = characteristic biofilm depth = [2K,Df/{QX/))tn\ and
A = adjusted substrate concentration.
836
Modeling Inactivation of Giardia Lamblia
By Robert M. Clark,' Member, ASCE
Abstract: Under the auspice* of the Safe Drinking Witer Act (SDWA) the U.S.
liPA his promulgated the Surface Water Treatment Rule (SWTR) requiring public
water systems using surface water to provide minimum disinfection to control Giardia
lamblia, enteric viruses, and bacteria. The C-l concept (concentration of disin-
fectant in mg/L times time in minutes) Is used to establish the appropriate criteria
for a surface system to achieve at least 99.9% Inactivation of Giardia lamblia and
99.99% inactivation for viruses. In the SWTR, an empirical equation was devel-
oped based on water tempeiature, pH, concentration of chlorine, and inactivation
level to predict required disinfection criteria (C-l values). This paper describes the
development of an equation based on Chick-Watson kinetics that provides equiv-
alent information but Is theoretically more consistent.
Introduction
The Safe Drinking Water Act (SDWA) as amended in 1986 requires the
Environmental Protection Agency (EPA) to promulgate primaiy-drinking-watcr
regulations that specify criteria under which filtration is required, require
disinfection as a treatment technique for all public water systems, and es-
tablish maximum contaminant levels (MCLs) or treatment requirements for
control of Giardia lamblia, viruses, Legionella, heterotrophic plate count
bacteria, and turbidity. Treatment-technique requirements were promulgated
by EPA to fulfill the SDWA requirement for systems using surface water
and ground water under the direct influence of surface water ("National"
1989). Regulations specifying disinfection requirements for systems using
ground-water sources not under the direct influence of surface water will be
proposed and promulgated at a later date.
Under the requirements of the Surface Water Treatment Rule (SWTR) all
community arid noncommunity public water systems using surface water or
ground water under the direct influence of surface water are required to pro-
vide minimum disinfection to control Giardia lamblia. enteric viruses, and
bacteria ("National" 1989). In addition, unless the source water is well pro-
tected and meets certain water-quality criteria (total or fecal coliform and
turbidity limits), treatment must also include filtration. The treatment pro-
vided, in any i case, is required to achieve at least 99.9% removal and/or
inactivation of Giardia lamblia cysts and at least 99.99% removal and/or
inactivation of viruses (i.e., virus of fecal origin and infectious to humans).
Unfiltcrcd systems arc required^ to demonstrate that disinfection alone achieves
the minimum performance requirements by monitoring disinfectant resid-
uals), disinfectant contact time(s), pH (if chlorine is used), and water tem-
perature. These data must be applied to determine if their C-l value [the
product of disinfectant concentration (mg/L) and disinfectant contact time
(minutes)) equals or exceeds the C • I values for Giardia lamblia specified
'Dir., Drinking Water Res. Lab., Risk Reduction Engrg. Lab., U.S. Envir. Pro-
tection Agency, 26 Martin Luther King Dr., Cincinnati, OH 45268.
Note. Discussion open until March 1, 1991. To extend the closing date one month,
a written request must be filed with the ASCE Manager of Journals. The manuscript
for this paper was submitted for review and possible publication on August 14, 1989.
This paper is part of the Journal of Environmental Engineering, Vol. 116, No. 5,
September/October, 1990. ©ASCE. ISSN 0733-9372/90/0005-0837/$1.00 + $.15
per page. Paper Mo. 25088.
037
-------
I'm i»\ \"i :i n.. \ pi> m\ imm/
m: t iniroimiciil S\sienis rneincerini:. (.'oHeec ol l.ngiiiceiiii-j. I'lenison Uiuversii\. (.'lenison. SC 2%U-lWIU.
'USA '
W Gi'JtiR
HU\AG. Swiv< federal Institute lor Water Resources & Water Pollution Coniiol. fll-MMM) Duhcndorf. Switzerland
G V R Makais
Department ol C nil F.nyiiieering. Uimersily of C'ape Town. Rondebosch. 0 P . Republic of South Africa
T MaISLO
Department of Urban r.nginccnng. University of Tokyo. 7-1-1 iliwigo. liunkyo. Tokyo II' Japan
ir,ai!tcnu:ital ntovieline hioloek.il wastewater iicalmciit. nitrification. deniiritkalion
IVTKOIKCTION
.!j;h-
rr.s can investigate the performance of a number of
v'.tiuial sysiems under a variety of conditions. They
particularly useful for those who are working
r:h s>stems in which carbon oxidation, nitrification
denunfication arc accomplished with a sine!:
•j:k because the competing and parallel reactions
: such s>stems arc so complicated that it is difficult
.'stimatc intuitively their responses to changes in
•hfrri configuration or load. Unfortunately, in spue
".at benefits to be gained from the use of models.
:r,\ engineers have not yet incorporated them into
~r routine practice This is due in part to the fact
IlII report is available as the / -t H I'RC St irnlilu and
'ichniiul Report \n /. which can be obtained fron".
UWPRC. I Queen Anne's Gale. London S\V 111 'JBT.
L k
that models have often been perceived as research
tools required sophisticated computer systems. With
the widespread introduction of microcomputers,
however, it is now possible for every design office to
use process simulation models on a regular basis.
Recognizing the benefits to be derived from the
routine use of mathematical modeling, the IAWPRC
in 1983 formed a task group to promote the devel-
opment anc facilitate the application of practical
models to the design and operation of biological
wastewater treatment. The goal was to review exist-
ing models and to reach a consensus concerning the
simplest one have the capability of realistic predic-
tions of the performance of single sludge systems
carrying out carbon oxidation, nitrification and
dcniiriticaiuMi. The model was to be presented in a
way that made clear the processes incorporated into
it and the procedures for its use. The group com-
pleted its task and submitted its final report in 1986.
That report will be published in total in the new
IAWPRC series entitled l.A H PRC Scientific ami
-------
:>un
M Henze et al
Tjhle I Process kinetics jnJ sioichionv
Component -• i
Process I
12 3 4
S, Ss X, xs
7
XP
8
So
1 Aerobic growth
of heteroirophs
2 Anoxic growth
of heteroirophs
3 Aerobic growth
of autotrophs
4 "Decay" of
heteroirophs
5 "Decay" of
autotrophs
6 Ammonification of
soluble organic
nitrogen
7 "Hydrolysis" of
entrapped organics
X "Hydrolysis" of
enir.tppcd organic
nitrogen
Observed conversion
r.ncs (ML 'T ')
Stoichiometric
parameters
Heterotrophic yield YH
Autotrophic yield Y4
Fraction of biomass
yielding particulate
products fF
Mass N mass COD in
biomass irt
Mass N mass COD in
prtnluLts from biomass
I - )>
» -J
£?£
°o
U —'
£ 5
M I
i 5
tr £
i — ft -1
• -i,
-i
o
h
¦3 c
rt C
60 O
c 5
H 2
2 1
I
C
¦= 5
"i i
•6 —
2 •£;
CL 2
- 2
5 2
CL c
4 57
£
|5
u #¦»
c. —
£ S
X >¦>
3 -J
"3 "3
!
Z 1
c 7.
V c
?s
-= £
o s
n j.
A
0 —
1 c
~3 £.
J
JL
3
< £.
< £.
-J C
' - ».
lib >
— i*
-X
7 1
Technical Reports. The purpose of this paper is to
present the major points of the report to the broad
IAWPRC membership Readers desiring more detail
are urged to consult the full report.
THK MOI»KI.
The essential features of the model selected by the
l.isk group arc presented in matrix format (Peterson.
I%5) in Table I. The information in that table allows
construction of the rate equations describing the fate
of each component for inclusion in mass balance
equations appropriate for the particular wastewater
licalinenl system being modeled. All components in
llic model arc listed by symbol across the top of the
table while their definitions are given acioss the
lx>liom. 'I lie nomenclature employed is consistent
with that recommended by IAWPRC (Cirau <•/ til..
IW2). All fundamental processes which are im-
portant in single-sludge systems are listed down the
exliemc lei I column wlule the tale expressions chosen
to lepiesent them are listed on the extreme right 1 lie
parameters in those rale expressions .ire defined ip :
lower right corner. The bod> of the matrix com1--
the stoichiometric coclliucnts, which are defined "
the lower left corner. If a particular process h.is -
impact upon a given component, then the box forr?:
by the intersection ol the process row and conipor-
column is blank
The model contains certain features th.it should >
recognized. First, the concentrations ol .ill orc.r.
components, whether soluble (.*>) or parln.ul.ite il
are given in units of chemical oxygen demand IC'OD
This simplifies the stoichiometric coefficient* t.
reduces the number of conversion lactors require.:
also allows calculation nl the oxygen ici|'iirt-i!"-nl '
a simple ("Ol) balance (Gaudy and (iaud\. 11"!
One consequence of this stand.ml is lli.it the o'tii.-
tralioii'. of hiom.iss .mil other parln.iil.Mes n"
tributing to the mixed liquoi volatile suspended sul.J>
must be multiplied bv an appiopnatc factor to fv
them into mass units The coiKvriii.itioni ol
mitogen spev ics arc given as nilint'.cii < •>nsn|iirnl!)
two conversion ladois (2 S(> aiul -l ^"7) must he mcC
-------
Miukl lor singlc-sltul^e wumewjier Ircaliiiciu
lv.ui .•ti.l.itmh utiitlii iiiimi. tiihl itriiiliilKijMon
107
i:
11
I'm^cm inlr. /',(Ml 1 I 11
" >.
'i» - V,r
tin ~ tr'ir
j y„
U ;s« )., ""(a,, ~ sJ(a,„ t.y.K^.+ .Vn.,)'1'4*"
, "'(as11 i"ssll)Ull".\i1),r"
14
I
14 ? >
I
u
^11 \ H II
^^ n
A XJV"'
\\)
( -v.. \ i( / A"..,, V .vN11 1
\A0 n I .S,, / \ ^,w/ {¦ 5y/ m
Kinetic p.irjnictcrs
Hilcrnirophiw growth and dciuy
/iH. Av A'0 Kho, />w
^ Autotrophic growth and decay
£ t1 i •
3
<- Correction tacior tor anoxic
^ Z i: p "E growih of heterotrophs nf
v 3^* — 5^ 5
" r £ r Ammonificaiion. Jk„
~ = 3 = - Hvdrolw* A,. AT»
- — - ^ - ?
,y ¦€ ^ ^ r i Co.ictlion Uuor tor anonu.
c 7 ¦£ z
¦£ 7 t 7
7 *¦
ti)druK*i*
he matrix to convert nitrate nitrogen and ammo-
nitrogen concentrations. respectively, to an equiv-
ii COD basis for calculating the utilization rates
ixygcn or nitrate Finally, no inorganic particulate
irriali are included in the model because they do
. enter inio reactions or contribute to the COD.
or contribution to the total mixed liquor sus-
ided solids concentration in any reactor can be
"puled b\ multiplying ineir concentrations in the
J stream by the ratio of the concentration of
iiculate inert organic matter (X,) in the reactor to
concentration in ihe feed stream.
The advantage of the matrix formal in Table 1 is
;i ii allows the reader to see at one time the impact
all potential conversion processes on all possible
npnnents To conserve space, reference to a given
aiponcnt undergoing a particular reaction will
•netimes be referred to by a shorthand notation,
R. where C refers to component / in column / and
means reaction j in row j Construction of the
vr\ed conversion rate for a component is accom-
ihtd b\ moving down the column i corresponding
to that component and summing the product of each
process rale p times the appropriate stoichiometric
coefficient, v,
', = 1 vv
/
For example, the observed conversion rate for readily
biodegradable substrate (component 2) is
r: — v:ip. + v^/i. 4-(2)
or
-------
508
M Henze ei al
The observed reaction rates can be substituted into
mass balance equations depicting the fates of the
components within appropriate system boundaries to
formulate a mode) for any particular biological treat-
ment system.
In selecting the components and process rate equa-
tions in Table 1 the task group focused on those that
were essential to a realistic simulation, yet were
simple enough lo facilitalc the model >i>lii!ion Al-
though such rate expressions ma} not depict perfect!}
the actual occurring within a system they can be used
satisfactorily as long as they mimic well the outcome
of those events. All papers influencing the thinking of
the task group cannot be cited here Nevertheless,
attention should be called to the work of Dold ci ul
(1980), Van Haandcl ct al. (1981) and Dold and
Marais (1986) because of their extensive bibli-
ographies and their synthesis of the information
contained therein. The latter paper is particularly
important because it tested an earlier version of the
model proposed by the task group (Grady et al..
1986) and expanded it to its present form In doing
so it presented excellent verification of the model
presented here
Examination of the process rate expressions in
Table I reveals that the task group employed the
concept of switching functions >.o turn them on and
off as environmental conditions arc changed This
was particularly necessary lor process that depend
upon the type of electron acceptor I oxygen or nitrate)
present For example, the bacteria which are re-
sponsible for nitrification are capable of growth onlv
under aerobic conditions (oxygen present) and their
rale of growth will fall lo zero ,is the dissolved oxviten
(DO) concentration approaches zero, regardless of
the concentration of their energy yielding substrate
This can he modeled b} including a DO "switch" in
the process rale equations The form adopted by the
task group was
oxygen switching function = — (4)
The selection of a small value lor A',, means lhat I he
value of the switching function is near unity loi
moderate DO concentrations but decreases to zero as
the DO concentration approaches zero The fact that
the function is mathematically continuous helps to
eliminate problems of numerical instability which can
occur during simulations will) models which include
ialc ri|iialions llial aic switched mi and oil discon-
linuoii.sly. Similai, processes wlm.li occiii only when
DO is absent may l>c linneil on In a suucliinc
function of I lie form
K„
- - (5)
I
Because of Ihc long solids iclcnlniii limes (Sll Is)
and low specific growth rales iiicoipoialol inlo I lie
design of most hioln|.,.ieal wnslcwalci lieatiiii'iil sys-
tems. differences in effluent soluble biodegradafc:
substrate concentration among different systcr
configurations generall} are small Conversely. tor;
differences in activated sludge concentration jr..
electron and acceptor requirement are common c,Jr
thcrmorc. good design practice requires thai ad:
quate electron acceptor be supplied in response t>
both real-time and space-time (location) depend"
changes in demand and that final settlers and sice;,
return s\ stems be capable of handling all a;uicip.'.::v.
concentrations of solids. This suggesis thai mode
depicting substrate removal are important more iV
thcr impact upon activated sludge concentration jp.l
electron acceptor requirements than for their abi!.:'
to predict effluent substrate concentration C-v
sequent!;.-, primary consideration was given b} tr.
task group to prediction of activated sludge eonc;-
trations during selection of process stoichiometry ir.
to estimation of electron acceptor requirements o.-
ing development of the process rate expressicr-
Nevertheless, it is apparent that the values of su::;-
mg constants like Kn v.ill influence those pretl!:::j"
even though the forms of equations i-1) and 15'
chosen more for their mathematical con< e:::cnc-
for conformance lo arv fundamental r.ite l.iuv
sequently. care should he taken in The selecfon ot -¦
values for switching constants to ensure ihat ree:
predictions arc not biased
The organic matter in a v.av.ewater rr.ay be .w~
divided into :: number of categories (McK'nne;. x
Ooten. I'ii>9: D.'U! ct a; . 1 l»St') The nrvi ii'tpc"
subdivision is based on biodegMdabiiitv
Nonbiodegradable organic matter is hinlo.".
iik M and |-.'s-;-s il'.roueh in acliv.Med kludge v.-
iiiu hanged in Ioim luo IrjLtioiis uii ide"
depending >n their physical stale soluble ard rv.- .
ul.ite Inert voluble organic matter '.S.i 'eju'v
¦Askni al llie v.mie concciitialion i 1:.li n eiilerv !•¦•
suspended organic matter (.V,l become-. enmesh;.' ¦
ihe ,;cliv;:t-.d sludge, and is removed from the ¦
through sludge wastage Because the waste siix.:
How rate is smaller than the sworn inllow ruts 4
mass balance requires the concentration of V, in
svsteni 10 be higher than in iIh. inl'm m
Hiodegiailable organic mailer iiiav be divided
two fractions, readily biodegradable :.SJ and vie- •
biod.-'radahle ( A ,,) The readily biodcgr.id.ihlc mj.
rial, which consists of simple molecules, is treated
il 11 were soluble whereas the slowlv biodegradaH:
miU'ii.il t iinsiviiiii' oftiiniplex molecules, is friMi.-
;is il it wcie p.11 Ik ulale, although ,imn; im.iv iiulrc-l '¦
soluble
Kc.ul'lv luodcgi.idable material is lOiiMdered :i» S:
I lie only sulisliati- loi pnvvili ol lieli-r ¦ >1 ¦ • ipliii. S
mass (A „ „) which can lake il 111 |soe column 1. f<»
I (C.K,) and column low T It'.K ) in Table l)rtf
conveii il inlo new biom.iss iiiuler eiliier ucmU
((',!(,) «>r anoxic (oxygen abscul. nitnle ptcvmi
(( ,lt.) conditions I lie elections associated xvilb t(*
cx|ien> hi in e of oiicigv lor t ell sviulieviv an1 lurj-
-------
MikIi-I lor tingle-sludge wuslewjlci treatment
il lo I In- cxogi'iicous election acceptors {oxygen
,) in 1111raIc (( ,K .>|
>wl) l>|l •« Icf'l .1(1.11 >k" Ml list 1,11C IS ClltlSldclCll III l)L'
•\v-il from suspension insiaul.iiicotisly by en-
ncni ill the liiollm. llowcvci. oiuc llicie. il iiiiisI
.led tipe used bv I lie helcrolrophic Ih.mii.iss lor growth,
re.n noils in vol vci I in this conversion are called
niKsis" in i lie model. although in icality 11 icy
ikeK to be more complex. Il is assumed tli.it
ilysis involves no energy utilization and thus
i- in) utilization of electron acceptor associated
il. ic I here is no stoichiometric cocllicicnt at
i >1 (aK
e Ji. eradation ol slowly biodegradable substrate
\ important to realistic modeling of aclivalcd
e s> stems because it is primarily responsible for
itamment of realistic space-time and real-time
•.denl electron acceptor profiles Careful exam-
in ol all available literature revealed that \ery
experimental work has been conducted
i. on the kinetiLs and mechanisms of dc-
¦ Hin of particulate organic material. Ncvcr-
s that literature revealed that certain features
Vi.ji.ired in oic.ei lor the overall system models
e ie.disti>. electron acceptor prulilcs. One was
V r..te was Itrsl oruer witli respect to the active
'trophic biom.i.vs present Another was llut the
ipc.ired u> saturate as the amount of entrapped
-tc became large in proportion to the hiomass.
:i;ioii. the rate of hydrolvsis is usually consid-
lower than the speeiric rale of utilization of
biodegradable substraie so that il becomes the
rniline lactor in the growth of biomass when
lowlv biodcgiadahle siibs:raie is present in the
¦> a reactor Furthermore, because of the need
7>mc synthesis li was reasoned that the rate
be dependent on the concentration of the
•n acceptor present and would be lower under
conditions than under jerobic ones. Finally.
is apparently completely stopped during
anaerobic (neither oxygen nor nitrate are
tl period* for organisms ol" the type found in
ed sludge (Van Flaandcl et al.. i^Sl). Exam-
¦ of row ~ in Table I shows that all of these
¦i were incorporated into the rate expression.
.toirophiL biomass is generated by growth on
biodegradable substrate under either aerobic
or anoxic (C.R-) conditions but its generation
meJ to stop under anaerobic conditions Bio-
s lost by decay (CjR4). which incorporates a
.umber of mechanisms including endogenous
•lisni. death, nredation and lysis The most
mi technique for modeling decay under aerobic
on> is to incorporate all of the mechanisms
singie rate expression winch is first order with
to the concentration of active biomass and to
h unit of biomass COD lost result in the
ion of an equivalent amount of oxygen (Grady
Vf)
and I .mi, I 'JN(i). I his appiiiacli, huwcvci, causes
problems when other electron acceptors arc consul-
cicd. 11> ,i vi ml iliose piobleins. I lie iippi uacli adopted
was basically that of Dold vt ol (I'JHO) mid il is
depicted in row 4 of lablc I. There it can lie seen that
a In si order rale ctpicssion is iclaincd. The laic
cocllicicnt. however, is dill'cicnt in both concept and
magnitude fiom the usual decay cocllicicni. in this
ease, decay acts lo convert biomass (C,R4) lo a
combination ol slowly biodcgiadable substance
(CjR,) and paniculate products (C,K4) which are
inert lo further biological attack. The latter are
similar in concept to the endogenous mass of McKin-
ney and Oolen (l%9) and act lo reduce the viability
of the suspended solids in .i hioicacloi (Weilille anil
Jenkins, llJ71) No loss ol C'OI) is involved in this
split and no electron acceptoi is utilized. l urthcr-
inoic. decay continues at a constant rate regardless of
the type of electron acceptor present. The use of
electron acceptor normally associated with decay
occurs as a result of cell growth on the readily
biodegradable substrate which arises from hydrolysis
of the slowly biodegradable substrate released by
decay. If conditions are aerobic, oxygen will be used
(C\R,l. If conditions are anoxic, nitrate will be used
(C.,R;) If neillici oxygen nor nilrale available, no
conversion occurs and the slowly biodegradable sub-
strate accumulates. Only when aerobic or anoxic
conditions are resumed will it be converted and used.
A portion of the biomass lost by decay is reconverted
into new biomass via the use of the readily bio-
degradable substrate resulting from the slowly bio-
degradable substrate released Because of this recon-
version, the rale cocllicicnt must be higher to give the
s.ini,- net loss of biomass as in the conventional
method of modeling decay
Nitrogenous matter in a wastewater, like carbo-
naceous matter, can he divided into two categories,
nonbiodegradable and biodegradable, each with fur-
ther subdivisions. With respect to the non-
biodegradable fraction, the particulate portion is thai
associated with the nonbiodegradable particulate
COD, the soluble portion usually is negligibly small
and is not incorporated into the model. The bio-
degradable nitrogenous matter may be subdivided
into ammonia (both free and saline) (SSH). soluble
organic nitrogen (5^-,) and particulate organic nitro-
gen. (ANC). Particulate organic nitrogen is hydrolyzcd
(C|:R,) to soluble organic nitrogen (C|,R,,)in parallel
¦with hydrolysis of slowly biodegradable organic
matter. The soluble organic nitrogen is acted on by
heterotrophic bacteria (CnRft) and converted lo am-
monia (C.uR,,). The ammonia serves as the nitrogen
supply for synthesis of heterotrophic (C,UR,, CI0R_.)
and autotrophic (Cu,R3) biomass and as the energy
supply (ClyR,) for growth of autotrophic nitrifying
bacteria (C„R,). For simplicity, the autotrophic con-
version of ammonia (C'mR,) lo nitrate (C„R,) nitro-
gen is considered to be a single-slep process which
requires oxygen (C,R,). The nilrale formed may serve
-------
510
M Henze et al.
as terminal electron acceptor for heterotrophic bacte-
ria under anoxic conditions (C,R:), yielding nitrogen
gas. Cell decay of either autotrophic or heterotrophic
biomass leads to release of particulate organic nitro-
gen (Ci:R4, Ci:R3) which can re-enter the cycle.
Autotrophic biomass is formed (C6Rj) by growth
at the expense of ammonia nitrogen (C|0R3) under
aerobic conditions. The decay of autotrophs (QR,)
is handled in exactly the same manner as the decay
of heterotrophs. The justification for this is the
likelihood that the decay observed in enrichment
cultures of autotrophic bacteria is actually due to
prcdation and lysis, with subsequent growth of
adventitious heterotrophic bacteria on the Ivsis
products.
The final constituent included in the model is total
alkalinity (column 13). Although its inclusion is not
essential, it is desirable because it provides informa-
tion whereby undue changes in pi I c:in be predicted
and avoided All relictions involving addition or
abstraction of protons will cause changes in alka-
linity. but the ones of primar> importance here
involve nitrogen (Sccarcc et al.. 1980. Downing a aI.
1964) as shown in Tabic I From equilibrium chem-
istry of the carbonate system, h total alkalinity falls
below about sDg in 1 as Ca< <).. Ilien lite pii becomes
unstable and can tall to values well below h (WRC.
I9S4). Low p!I decreases the nitrification rate and
causes other problems such as corrosive and ag-
gressive effluents. bulking, etc Inclusion of the
proper input term in a mass balance equation for
alkalinity permits a user to evaluate whether the
process configuration under consideration allows
sufficient recovery of aikalmiiv durinj: dcniirilieaiion
to maintain the pi I in the pioper i.ingc regardless of
the proton release during nitrification. If not. then
appropriate chemicals, such as lime, must be added
to maintain the proper pi I
CIIARAfTERI/.ATION OK WWIEWATEH AM)
ESTIMATION OF PARAMETER VAI.fE.S
The model contains I.? components and 19 par-
imeicrs. Fortunately, eight of the parameters show
it tic variation from waste to waste ami niav he
:onsidered to be constants. Thc> are listed in Table
!. The others must be evaluated on a ease by case
T.ihlc 2 Parameters .un! ilMMilcrtsito wlmh may Ikv .iNMimcil
ytnhol *ivUiu*
x Vicltl l«»r nnlt»tiImihimss
4 Dcc.iy cuclliucnl li»i .iiiUtlrophu- hintn.r s
I t.hHimv «>{'hiooutsN trailing to p.ntuul.iU* pii*liKis
„ «tt tllllt^VH |H*I llhlNi i>l ( (III Ml Imoiii.i..
r M.»sn ««l |«ci iiui .*• «*l « « U > ttt |*i«tint « • l««mt
hmninss
„ „ Owtfcn full viinuitiiMi \«h*I1i« irnl l»»i lu'iri«»ti.»phh
Nili.ile lu)l vtUiinli«»n vti'lli* init lot J.'iitinlMiip
hrlCM»ln»|*hn Imoiiihi^
, 4 OftygcM liullhi>iiton incllau'nt lur .lulotmpluc
hiomim
basis. All of the components except X, ma> a?
in the influent to the activated sludge svstetr;.
must be evaluated in concert with the sto!chio-r
parameters In the following, the subscript I
constituent represents its concentration in :he
reactor feed.
Characterization of wastewater and estimation ¦/
chiometric coefficients
Because of the operational definitions of t
rrodcl constituents they must be determined e.xr
mentallv in a wav which consider;: «
assumptions in the model. Evaluation of K-:r
wastewater characteristics and ine model coelfa.'.
will be expedited if completely mixed activated iij.
reactors are operated al steadv state in an jcrc
mode at a number of SRTs
The lotal COD in the influent wastewater is
up of lour constituents
Influent COD = 5,, - A\ - 5 - )
Todetermine the concentration of'ne-1 ¦¦ 1
organic mailer, remove an alkiaot o: :ae .
conienis I'rcm one of the completely —ivj.: .
bemj! operated at an SRT m excess .>1 l>.
.te;.ite it in a hatch icacioi Remove samples
icall; a:k! analv/e :hem for soluble COD 1u.e .-
stable residual value is .V,;
Uefore the concentration ¦ readiiv bi.i|Je-':-u-_
subslrate. \s . can be obtained, the lie'.eviu
v:e!d. must be known \ sample w
should he settled, tillered to remove the ?Ttk, .
material anil seeded liithllv witli ,iu ik.ii..1
from one ol the completely mixed .e.uv'- \
should be removed periodical!;, ar.d hot1-, v.e .'
COD antl ihe total COD determined . -
lioplm. weld i.m be detcimitkJ iiom (lie .
Cell COD - total COD - o!uh.'e COD
A cell COD
" " iw^iihle CTJI )
II ihisisiloiiesever.il limes, an approximate !' ..i
niav be determined Anv eirors in ilns estn:-..ne ¦
be compensated for in the determination ni nr
parameters or influent concentrations
The value of ,S\, can be estnnalcd bv ine.isurin:
change m oxygen niili/.ilion rale in a mh;L' .•
pletelv mixed icactoi opciateil al .in Sl< I mar .V
under a daily cyclic square wave leeding pattern '
with Iced. 12 h without feed) (1 kama i f :il. l,Kc 1
shown in I i^ I, 11kie is a i.ipitl diop in -
cialeil onlv with (lie readiiv hioJegi.id.ihlc m.i'KU
and can be used to lind its coiieenlialii'ii
(AOUKlff l
S" " (.'(I '
-------
Model Tin iiii(jlc-tlml|jc wuilewulcr irculnieiil
III
i .
| bMi »
I 1©»d
r»top
fa«il
'1, 11 n • .
T 1a i
('.1 ' ¦
1 n
3 uj- j,j
' I
f'tvj
*2
Time (h)
20
Kv-sisvnsc ol .» uMnpk'tdv tni\ai .iiltv.ilcil slmkc
i ' i IJ h s(.|ijjic w.ivc response.is used lo determine
tL.t.iiaimii w| readily hunlcgijdahlc substrate I rot»
Lk.inu (¦/ jI (I
i =¦ i h.sn^L- in o\\gcn utilization rate following
ifed termination iML "T ')
- reactor volume (LJ)
- iced (low. rate prior to termination (L'T ')
ili-h-i imiit'ii !h.' loiiccnii11 • ms in the
ul liie total tOl). rejiIiIv biodegradable
r>! the inert soluble COD it is onl> nccessa^s
v.'. iiv' either the COD of ihe inert suspended
mallei (A(i! or the COD of the slowly
ic-bie substrate i.Vsl) because the other Can
mined b\ difference using equation (6). It is
i."wed lh.il A',, be evaluated as a parameter
:l' the model to data showing the effect of
sad::-.: production in Ihe steady-state com-
nwd .icm .:ied sludge .e.iei.'rs If ilie ion-
anion ophn. biomass is neglected. the
. . the model are \,,. A bu, k„ and
it.ori (6i :.ia> be u>ed to eliminate A',, The
-n.i> be assumed (Tabic I) ar.d the value
: be evaluated indcpeiidcnll>. as described
SRI val\:."i in e*<-ess of 5 day.. .*>, and A
Ei.gibl;. snuli allowing S.,. kn and Kt to be
Lkaraa a uL. 1^86). Thus, the value of Xn
. unknown It can be evaluaicd by using u
isionjl search routine which chooses Xn to
'.he error sun; of squares when predicted
faction rales arc compared lo measured
function of SRT This fitting acts to tune
to the particular wastewater under study
T..sj:e.\ lor am error made in the estimation
2 b„ Once A',, is known, Afj, can be
ir.im equation (6) f-'or variable strength
car. generally be assumed lhal the various
,i. constant propmlion lo one anotliei.
actuated sludge modeling it is assumed
incentrauon of biomass in the influe.U is
ompareii lo lhe aniouni formed within the
ul appioach is taken here, primarily be-
research is needed regarding the impact
of biomass in the inllucnt. No procedure is recom-
mended for mnisunng the inlUienl concentrations. If
llicrc were a desire lo include litem in Ihc moilel.
iippioprinlc microbiological method* would have to
be employed.
Examination of Table 1 reveals that the model
includes the soluble concentrations of oxygen, nitrate
plus nitrite nitrogen, and alkalinity. The concen-
trations of all of tlic.se constituents in the feed may be
measured by appropriate chemical tests.
Oxidi/able nitrogen may be present in the feed in
five forms: ammonia (.TNm); soluble, inert organic
nitrogen (.VN„); particulate, inert organic nitrogen
(.VN„); readily biodegradable organic nitrogen (.S'N,,,);
and slowly biodegradable organic nitrogen (.V*,,,).
The concentration ol ammonia iti the Iced may be
determined by appropriate analysis of a filtered sam-
ple. The concentration of 5N„ may be determined by
performing Kjcldahl nitrogen tests on aliquots of the
samples used lo determine the soluble, inert COD.
The Kjeldahl test may also be used to determine the
total concentration of soluble organic nitrogen in the
feed. Subtraction of ¦^Nll from that value approxi-
mates 5V
If .Ss.,,, and .Vm.,, arc assumed to be
piopoi Honed in the same way as the readily bio-
dcgiadablc and slowly biodegradahle COD in the
feed, then \\u, may be determined once S\t)\ is
known'
A^NDi + S.
-V,, +
(10)
There is no need to determine A's„ since it does not
appear in the model.
f-AluiiittU'ii Hi kith lli fttiriuih'h-i \
Tile purpose of the half-saturation coefficients for
oxygen (K0,, and K0 ,) and for nitrate (A'v)) is to
se. ve a» switching functions Consequently, the actual
values used are not cntical as long as they are of the
appropriate order ol magnitude and are small in
comparison lo operating conditions.
The most critical parameter for characterizing the
growth of the autotrophic biomass if the maxi-
mum specific growth rate. The recommended pro-
cedure for it> determination u to measure /j4 during
a dynamic te.st on one of the completely mixed
reactors, provided it is barely nitrifying and has a
high dissolved oxygen concentration. The sludge
wastage rate should bo decreased to make the SRT
greater than that required to achieve a high degree of
nitrification and the concentration of nitrate nitrogen
should be measured over time as it increase"; through
growth of additional Minifying bacteria If Ihe natural
logarithm of the nitrate nilrogen concentration is
plotted vs time, its slope will be ;j1 (),-/>!, where 0r
is the new SRT and h't is the traditional decay rale
coellicieni for Ilie iiitnliers (Hall, 1^74). Unlike the
situation lor heterotrophic cell mass, the traditional
decay rate constant for autotrophic bacteria, b'„ is
-------
512
M Henze et al.
numerically equivalent to the specific decay rate
coefficient in this model. bA. Its value may be assumed
(Tabic 2). Since 6X is known, /i, may be calculated.
The half saturation coefficient for the nitrifying
bacteria, tfNH, can be determined by the infinite
dilution procedure of Williamson and McCarty
(1975) which provides information on the re-
lationship between the specific nitrification rate and
the pseudo-steady state ammonia concentration.
That information can be analyzed to provide a value
for Ksh
To determine the heterotrophic decay coefficient.
bH, sludge is removed from a completely mixed
reactor and put into a batch reactor where the oxygen
uptake rate can be measured many times over a
period of several days (Ekama et al.. 1986). The slope
of a plot of the natural logarithm of the oxygen
uptake rate vs time is the traditional heterotrophic
decay coefficient, b'n. The model decay coefficient b„
can be calculated with.
bH
1 - Y„t I -/,)
(U)
The parameter is a correction laclor winch
adjusts for cither the change in f'i„ associated with
anoxic conditions, or for the fact that only a portion
of the biomass can denitrify (Baichclor. 1982). »/,, is
a correction factor which adjusts for the observation
that hydrolysis of slowly biodegradable organic
matter occur"; more slowly under anoxic conditions
than under aerobic condition1. I In- lists lo iiicasiuc
and arc performed at the Mime time by evalu-
ating oxygen and nitrate consumption rates in two
lulch icaclors which aie equivalent in every respect
except for the terminal electron acceptor (oxygen in
one and nitrate m (lie oilier). Immcdulely .tiler
bringing biomass into contact with wasicwalcr in a
batch reactor the activity in (he reactor will be
dominated by growth of the hcicrotrophs on the
readily biodegradable substrate whereas later activity
will be predominantly due to use of substrate arising
from hydrolysis of the slowly biodegradable substrate
(Pfcama a til.. I')XM. If ()!/R_ reprcsen:-: file or.yj'ei!
uptake rate during the lirst period in the aerobic
reactor and NUR, represents the nilrate utilization
rate in the anoxic reactor, then.
n, = -
(2.86)(NUR,)
OUR,
(12)
Likewise, if OUR,, represents the oxygen uptake
during I he second period and NlJRt the correspond-
ing ml i a Ic uptake i.ile, I lien
C.S6)(NIIKJ
Ml =» .
OUR,
(I M
The parameters describing biomass giowth, /i„ and
A\, aie villiict111 lo evaluate meiiuitely, I>iii tli.it is not
critical because (he model is not very sensitive to (heir
values. The main luiiction ol /i„ is (o allow (he
maximum oxygen uptake rate to be predicted. T>
suggests that measures of u„ should be based upt
oxygen uptake measurements rather than cell ero^
or substrate removal Cech ei al (1985) and Chudot
et al. (1985) have described a respirometnc proced'j
for measuring both /!„ and A\.
An important factor which has only rccenth be.
recognized is that biomass grown in different react
configurations exhibit different values of and t
even though the reactors are operated at (he sap
SRT, loading, etc. (Cech et al.. 1985: Dold j-
Marais. 1986). This suggests that care must be ljs.
in the collection and interpretation of kinetic dj'
Preliminary evidence suggests that it would be jair;
able to estimate and A.'s using biomass from ¦'
completely mixed reactor receiving a daily c>^.
square wave input of food that was used for de'e
mining the concentration of readily hiodegraJj1-
substrate in the feed.
The final parameters to he evaluated are k.. ,L
maximum specific hydrolysis rale, and A\. the u._
saturation coefficient for hydrolysis of slo«!>
> The best >v.i. to estimate '•
A", ii lining the model response to the oxygen ::r'-
pattern in Fig 1 (Dold and Marais. l9Sni Sr.*:-.'
other parameters have been selected, the
knowiis for curve-Ill .ire the two hvdrohM* ?
aiueters. .uul the technique has been lound lo Ive"
sensitive to their values.
As seen in the preceding discussion, evaluate"
the p.ii.iineleis and vli.naclei i.Mtion .a the u ¦.
water must proceed in a pariicul.ii order TV1-,
summarizes the sequence in winch tlun-js r.\. \
done
TW'ICU PVRVMf TfH KVM.KS
Table 4 presents typical ranees for values
.stiiicliMiiiicli k .mil kinetic p.ii.iinclers ,i! hculr.i!,-
and 20 C" lor domestic wastewater While the rr.. J
is iclativvh insensitive to some of these par.irr.e:.-
(scc Table 2) allowing assumed values lo he use5
is quite sensitive to others l urlhermore. since v::
parameter values are strongly dependent on spe. \
factors in the wastewater and on cimronrr.-'.
conditions, (lie desir.ihilitv of experimentally dr* •
minim; llic-ni cannot he overenipii.tsizcd
sci|ueiitl\. 1 In- value-, listed in lahle 4 .ire preserv.
simply lo i'iu1 I In- ic.ulei an .ippieei.ilion lor t.'r
range likely to lie eneounteicd. In rcvicuinii llicm t
should Iv recognized that wlule some have hccnurf
in oilier models, llicicby piovidun: a reasonable tu.'1
ol k iiowli dye I nun wliu Ii to • 11 ;iw, niliei x (e g t(.
'/.) aie relatively new so tli.it the r.uigcips
aie tentative al l>est. As the model is used and nun
-------
MimIcI inr aingtc-aluilgc wanewuier ircalineiK 313
l.tl'ii* I t'.ti.tmeici* aiist v.l\.ii4\ lc»i>Uv>» svUu.lt tmtM I* evalutiiril .u»*| httouuation itccdcd
'iyiiiln'1 NlllHC limlni
,\sm S«>tnl»U' miuitc uuiogcu uiiuruliitdiMi
nt u.ttirw.iirr
S,ikCMii.ili«>n in
u.isiru .iicr
.Ss/, Soluble meri organic mirogrn
viMiccutr.ilKMi in w.hitu tin
Ss/I| Si»lnhl» Imulrgi.itl.iltlr oi^.uiK ,SN/l
imioprn k«tikriur.ilii>ii in u:isU-u.nrt
N icUl lot ItririohopltK l>i»nn.iN>
Su ('uiKcnirjiiun ol readily bnxlegr.idjhlc YH
( ()0 >n ur.i%1cw.ticr
H, M.milium spculiL g j11 wt)i r.iic fur h4
auiolrophK hiomxo.
ASII \inni«Mii,i lull vMiirtiiou n»ollutrni
11 »j uuU'irnplui. hiom.tss
hH iHx.iy %.*»cllkiu>i tw hcicri.lmpbu. >'#i./r
bioin.i>s ^
Iikti \u\|vnJcil oigauu m.iiU'i h„. SX(.S/t
ixiiii million in w.iMi'u.iin
\ v, Mo»ly biodegradable oigjiiiL mailer .V(t. i'M, «S;l
conccnirjlion in wastewater
Sloulv hioJciir.iJjblc orgjnic nilrogcn .S\(.
voiucuimuou m wavicw.iu'i
r; 1 .»iu\hi ft,, titiiict .mo\u
il'll Jit It MIS
n, Correction lauor for hydrops under
anoxic conditions.
n>, \1.iMinum vpiMJu j:rt'\vi!i r.nc )lM, Imuiii 1
A I i.ill ..il .ii ui.»ii ll.ocnt iwi )'J(, I M, .V, .Ss,. tr
heterotrophic hionus%
Maximum >pcuhc h>drol>sia rale
A, Hali-»aiuraiion wxMhN.'.em lor hydroUsi*
id biodegradable buKsiraic
ics arc performed. our know ledge of expected
mcier values will expand
Tnir.j1 cn\iror.mcnlal conditions. pi I and tcin-
:urc jre ol'priman importance. The effects of pH
'.ii'itiejiuin have neen studied extensively jnd
u.iative rcJ.:(u>n-.hi[>•> have been proposed in the
-".Lire Heterotrophic activitv is jl.so influenced bv
pH. but to a lesser extent so that few relationships arc
available. Nevertheless, the importance of pH should
he recognized Uulh niirilication and deniinticalion
involve changes which can alter the pH if the alka-
linity of the wastewater is not adequate. Con-
sequently. the model was structured in a way which
allows alkalinity changes to be calculated, thereby
Tabic J Tvpical parameter values ji ncjtral pH and 20 C fur domnlic wjslcwalcr
I'niii
Svmbol Sioichiomeiric pjfjmctcn. Value
>•<
g cell COD formed (g N oxidized) 1
0 07-0 2*
>'»
g cell COD formed (gCOD oxidized) 1
0 46-0 M
f.
duner.^onlcvs
4l
-------
514
M. Henze t CONSTK VlNTS
Modeling of a system as complex as a single-sludge
wastewater treatment svsiem icquires ih.'t certain
simplifications and assumptions be made to nuke the
model tractable To prevent their violation it is
important that thev be dialed explicitly.
. I sMi>>ipuon\ tuiil ivsirn iioii\ i/wm hiliil n nil llw •nm'i I
(1) The system operates .it constant icmpci ature
and the parameter values are appropriate lor the
particular temperature assumed.
(2) The pH is constant and near neutrality.
(3) The nature of the organic matter is constant
although its concentration may vary.
(4) Sufficient nutrients arc present to allow- bal-
anced growth of the biomass
(5) The coil eel ion I'.iclois I'oi denili ilic.itiou. i|, ami
t]h. arc fixed and constant for a given wastewater. It
is possible that their values are influenced by system
configuration but this is not considered.
(6) The coefficients for nitrification do nut change
as other waste constitutcnts arc removed.
(7) The heterotrophic biomass is homogeneous and
docs not undergo changes in species diversity with
time
(N/T!'iC entrapment of paniculate oigamc mailer in
the biomass is instantaneous.
(lJ) Hydrolysis of organic matter anil organic nitro-
gen are coupled and occur simultaneously with equal
rates.
(10) The type of electron acceptor present docs not
aireet the loss of active biomass by decay.
Cnii\iiiiiiits on iippliailimi «/ the imitlcl
Constraints arc necessary to ensure tlut the simu-
l.iiion u'Mills have pi.uiiiul utility.
(1) The net growth rale or SRT of the biomass
must be within the range that allows a sludge willi
yooil scllliiiKcliariii'lciislics lo develop. Typically, the
SRT should lie between 3 anil 30 ilays.
(2) Hie reactor ouiliguiation sliouUt Ik* rtuch thai
(lie activated sludge concent ration entering the sclilcr
is between 750 and 7500gm"' in order to achiev
proper sludge settling
(3) The unaerated fraction of the reactor volur
should not exceed 50% because larger fractions ma
cause deterioration of sludge settling characteristic;
(4) The mixing intensity associated with ui;r
l.Hior. of the lime-dependent response >l a
tvpical diurnal patterns in tlow and concerur-:.
Although considerable space in ihe task .roup
was devoted to strategics for dc1 eloping —
codes, it is impossible lo presem all ol i!v.. ir1 -
Hon here. Howevci.it is important that -.everjl-
be made.
I list, compiled '..im'uasvs mk'i .'•> H\s';
Pascal are sufficient lo implement '. pou.- :-.il .
Second, even complex flow schemes can be ~iv.
as simple tanks-in-serics s\sie"\. there". -¦•"Ki
the computer code ami its use i or es.uupli: I .
illustrates several common processes and tlv.Mr t
analogs By careful consideration of the impv -
system characteristics the number of svstems
could be simulated is almost limitless TlurJ. ,o- -
eralilc computation lime can he saved hv pr.'T
consideration of the step si/cs used during minv i
iniegration For example, changes in DO cMir
Irations occur with time constants on the or.!.- ¦
seconds wlicicas the tunc const,nils assimated
other soluble constituents are in the order of mir.'.
and those associated with particulate constituent- "
in the onlri ol hour. I lir. nn-.ius tli.it nun--'¦ -
intcfialions can lie nested l>v n .inr ilill' icul '-ir;' :
I'oi dillcrcnt constituents lourlh. the linal settles
l>c consiileied lo he a sep.ii.ilion point Whilr t
lueaii-. I lull tin- uiod.-l will ii.il ^-ivi a tine iluu-i
siniulalion of an uclivaled sludge system in »'*!
large Iraiislers of solids occui between the .icnut
basins ami ihe si-iller. it will still give reihotu^
liine-ilcpenilenl ies|)onses ol svslcnis wlneli itrtOfff
alcd with underloaded el.uiliers Im.illv. bv fl>
plilvini: Ihe pioeess late opii-ssions lo .ippropntf
-------
Mmlel ltw ^iiigk'-iliulgi' Hmifwiiici iiimiiiiciii
I ,. var> in their
ik..iain anj speed, all are easy to use and are
;iuative of ".hat can be accomplished with the
< (>S< I t Sl(>\
odd. such as the one presented here, can have
he-.-'ticiai effects upon the practice of cmlron-
engineering Hv allowing cvploration through
:ior. of a broad range ol'system configurations,
and operational strategies, it greatly expands
.neer-; evpcr.encc base and increases his intu-
xision-making ability. Once calibrated lo a
lar wastewater, a model allows the engineer to
i large number of potential designs and elim-
Ose which are inetiicient from either a process
sconomic perspective. Furthermore, once a
has been selected a model allows the units
he N\stem to he sized so as to minimi/c system
naiiy. alter a plant has been constructed, a
;ar r>c used to investigate alternative oper-
strategics lo minimize the impact of new
the hope r»f the task group that the model
d Herein u:ll he used by practicing environ-
¦ngineers. thereby bringing them the benefits
listed above, runliernioie, n is lu>|>cd ihnt this model
will cause hoili prachtioneis and researchers lo ask
critical i|iicsiin ol .i kindle selection theory H'ul AVv
1«>. 191 IVC,
I)n!t! I' I. .mil M.irais (i v It (I'lSfi) |{v.ilu:ilion of the
ecner.il activated sliuli:e model proposed hv I lie IAWPRC
l.isk ^loup II,//, i S, i Uilutol IN. (it S9
I >i >kl P 1 . I k.im.i (i A .iiul M.n.ns (i v K (198(1) A
general model h»r the .it.1 is.1 Lcit simile process /'rut; It at.
rci liniil 12. -IT 77
Downing A I . r.uniei II A anil knowlcs Ci (19(4)
Niirilicalion in I lie .itlnalcd sludge piiH.es-, J I'rui ln\l
.S,'iwizc I'uril 64, I 111 I sx
i k.i m.i (i. A , Dokl I' I .mil M.ii,ns (i v It. (l9X(i)
Procedures lor dclcrmimng inlhieiil COD fniclions and
the m.i Milium spccilie gioulli rale ol lieierotrophs in
aetnalcd sludge systems IIij/it.Vj Av/imi/ IK, 91-114
Gaudv \ 1 Jr jnd (j.iudy I". T (1971) liiiyhmutil Concept',
lor Dcuyn ami Operation ot tlu Aelu iitcd Sludge Proi ro
t S rinironmenul Protetlion Ageiuy Water Pollution
Research Series. Report No I709U. TQJ. '19 71.
Cir.ul> I". P I Jr anil 1 im II C (19X01 llmlnviuil
It imnutlfr Iri iiinii nt. Iiiturv ami Applu uliun\ Marcel
Dekker. New York
G;^J> C' P L Jr. Cju;er \\ . Men/e M.. Maruis G v R
ard M.itsuo T i|9,Sfii \ 'iindel lor sin^le-sluilpi. w.isle-
u.iler l:e,ilme!i; -,;.sk-in - M ur .S it Li/itml IX, 47 61
Cirau P . Sill I,mi I' \1 . I len/e M . I'Im.ileli S . Ciradv C P
1. Jr. (.iuier \\ and koiler J (I9S2) Recommended
noiaiion lor u^e ir. the ocicnpuon ol biological waste-
water treatment processes Ha;. Ri^. 16. 1501-1505
Hall I R 11974) Sonic studies on nitrification in the
aeliv.ileil sIiuIlv prit.es- Hot P.tllul Conlrnl 73,
,V.s 547.
Mk.K:nncy R E and Ooien R J (19h9) Concepts of
complete mixing activated sludge Tram I9ih Sarin
Envnv Conf I mi kaiiMiv pp 32 59
Petersen I I (|9h5| Chemical Hiactiun Aitahus
Prentice-Hall. Englewood ClilTs. N.J
Seearee S N . Benninger R W . Wcbcr A S. and
Sherrard J H (19S0) Predictions of alkalinity changes in
trie activated sludge process J Hat Poiim Control Fed
52. 399-405
Van Haandel A. C.. Ekama G A and Marais G v R.
(I9SI) The activated sludge process—III Sinple sludge
denilrilie.ition lli/f AY.v 15. MVS 11*2
.iler Research Commission (I9S4) riu iir\ . and
Oli -atum of \uirtiiit Rimtiral Actuated Sludge Pro-
hjm\ Water Research Commission, PO Box 824, Pre-
toria OtlOI. South Africa.
W'.sldic C L. and Jenkins D (1971) The viability and
activity of activated sludge Hot Mr\ 5.621 WO
Wiilianisi.ii k J .iiul Met .iris P L (I97S) Rapid mea-
surement oI'Moiioi! half-velocity cocllicicnts for haclcrial
kinetics /Itoti-ihunt Him-nvni; 17, 915 924
-------
IIAI.VOKSON
92. Smith, W. S., ami ln,\, S., Food Research, 12, 400-2 (1947)
V.J. I'l'iir.nsoN, C. S., ami I'isciii r, P., A'. 1'. Sidle Agr. Exp. Slas. Tech. Bui!
273 (1911)
91 IIImfi l.li, II.. J. Bod., 54, 51.) 17(1917)
9.i. I li- ATM, II., Austral,on Font Manuf., 16 (1947)
9fi. Osir
-------
Mi )NOI)
.ill the <<¦ lis li ivi' 11i\i -= xi 2\
If r is I In- nniiilii r nf ili\i>imis per iniil lime, we li;ive ;it lime t2:
.v, - .v, 2r(,» «¦>
mi lining logarithms In the base 2.
l"Ki'2 - ll>K2'"|
' 111
uliete r is t lie mean division laic in tlir linn: interval (j —/|. In
defining it we li.ive considerI the iuciease in cell concentration.
\\ In n tlie nveiage si/r nf tin- cells does not change in the time
i n t < i va I < nnsideicd, I Ik- imi'-.i-i: in "bat leiial density" is propor-
t i•'11:11 to the inciease in i ell com enti ation. Whether growth is
estimated in terms of une hi the oilier vari;ilile, the growth rate
is the same1.
However, as established in pai Iirnl.ir liy the classical studies
nf lleiiiici (3), the a\i i'.i);r >i/e of the cells may vary considerably
f i < >in one phase to .niollni nf a giowlh cjcle. It follows that the
tuo vaiiahles, cell < on< enliation and bacterial density, are not
e 1111 i \. 11 e n (. Much ciinfic-ion has Ihcii cieated because this ini-
I •> >i 1.1111 disi iin I inn has In i n fi e<|Uent ly overlooked. Act ually, one
ni | lie oiher vaii.ible may be moie significant, depending on the
t \ pe of pi i ible i ii 11 »n si i let e 11. Iii most of t lie experimental problems
nf b.Hleiial c I ic 111 is 11 v, metabolism, and niiti ition, the significant
v.ii ialile is baiteii.il density. Cell ohm enliation is essential only
in pinbleins when1 division is ai III.illy ctincei lied, or wlieie a
knowledge of llie clei i n n I ,i i \ i ddi pnsi I it hi nf the populations is
impm taut (iniit.il ion, si lei I inn, etc.).
1 I lie iie-c of Inn li.i-r in | il.i ¦ r of Inj; l>.isr- 10 simplifies all I lie r.ilcul.itions
i mini rlnl u illi lli i.it'¦- ll i1-i"-|i( ri.ill) cumcnirnt for tlie ginpliic.'il rcjirc-
-t ii l .i I ii xi of Krimlli rill Ms. If |np of I lie li.irlrii.il ilrnsily (Ior, ¦= .V.122 lo^io) is
|'!"llril .iK-iiiiM liinr, ;i n in* n.i'r nf nni iniil ill i irilin.'i Irs col irspontls In one (livi-
i ¦ .11 (nr ilniil iliii|:) I lie 11111111 >' i nf division-. (Ii.it lit \ r- mi ni rod (luring ail)' lime
in i»r v .¦ I i= i\ ci) liy tlx1 ililfui nir of I In- •>¦ ilin.i Iri of llie rnrrc-ponilint; point1!.
II i - ili"-ir.ilili' lli 11 I Ii i-^ | ii ,m I i' i I ii ii i|i | | id mi if i.i'iK r.ili/ctl.
GROWTH OF I1AC I f .KIAI. CULTUIUCS
3 73
Although the two variables are not equivalent, it is convenient
to express growth rates in the same units (i.e., number of doublings
per hour) in both cases. When cell concentrations have been
estimated, it is equivalent to the true division rate. When bacterial
density is considered, it expresses the number of doublings of
bacterial density per unit time, or the division rate of cells postu-
lated to be of constant average size. In all that follows, unless
specified, we shall consider growth and growth rates in terms of
bacterial density.
These definitions involve the implicit assumption that in a
growing culture all the bacteria are viable, i.e., capable of division
or at least that only an insignificant fraction of the cells arc not
capable of giving rise to a clone. This appears to be a fairly good
assumption, provided homogeneous populations only arc con-
sidered. It lias been challenged however (Wilson (4)] on the basis
of comparisons of total and viable counts. Hut the cultures ex-
amined by Wilson were probably not homogeneous (sec p. 378),
and the value of the viable count in determining the "absolute"
number of cells which should be considered viable under the
conditions of the culture is necessarily doubtful (see p. 378).
Direct observations by Kelly & Ralin (5) contradict these findings
and justify the assumption. [See also Lemon (42) and Topley &
Wilson (43).]
Growth I'lIASKS
In the growth of a bacterial culture, a succession of phases,
characterized by variations of the growth rate, may be conveni-
ently distinguished. This is a classical conception, but the different
phases have not always been defined in the same way. The follow-
ing definitions illustrated in Fig. 1 will be adopted here:
1. lag phase: giowtli rate null;
2. acceleration phase: growth rale increases;
3. exponential phase: growth rate constant;
4. retardation phase: growth rate decreases;
5. stationary phase: growth rate null;
6. phase of decline: growth rate negative.
This is a generalized and rather composite picture of the growth
of a bacterial culture. Actually, any one or several of these phases
may be absent. Under suitable conditions, the lag and acceleration
-------
74
MONOIJ
li.tses m;iy often l)i: snppicssed (sec p. 388). The retardation
h.ne is frequently so slnu t ,15 to l>e imperceptible. The same is
III. 1I Mi jscs of r i ¦ w 111. I,"\vn « ur w. Inj; li.icln i.il density. Upper rtir\ r:
.i i.i(ioneiituil ^rou-lh rale: growth rale (liiriiit; the exponential
i.i-h' (R). It is gi\en hy the evpiession
1(11'.. t. — log.t.
A
h ~ 11
1 "( t«(i* s t tn r (n t.i |r," tM' <(t< f( I I 'M I.
GROWTH OF BACTERIAL CULTURES
375
when /j— h is any time interval within the exponential phase.
Lag limcand growth lag.—The lag is often defined as the duration
of the lap phase proper. This definition is unsatisfactory for two
reasons: (a) it docs not take into account the duration of the
acceleration phase; (6) due to the shape of the growth curve, it is
difficult to determine the end of the lag phase with any precision.
As proposed by I.odgc & Hinshelwood (6), a convenient lag
constant, which we shall call lag time (?'/) may be defined as the
difference between the observed time (I,) when the culture reaches
a certain density (xr) chosen within the exponential phase, and the
"ideal" time at which the same density would have been reached
(/,) had the exponential growth rate prevailed from the start, i.e.,
had the culture grown without any lag 7'i = or
logs*,. - log2r0
" '¦ J ¦
The constant thus defined is significant only when cultures
having the same exponential growth rate arc compared. A more
general definition of a lag constant should be based on physiological
rather than on absolute times. For this purpose, another constant
which may be called growth lag (L) can be defined as
L = Tt R.
L is the difference in number of divisions between observed and
ideal growth during the exponential phase. Ti and L values are
conveniently determined graphically (Fig. 2).
ON TECHNIQUES
Estimation of Growth
Hearing these definitions in mind, a few general remarks may
be made about the techniques employed for the estimation of
bacterial density and cell concentrations.
Bacterial density.—For the estimation of bacterial density,
the basic method is, by definition, the determination of the dry
weights. However, as it is much too cumbersome (and accurate
only if relatively large amounts of cells can be used) it is employed
mainly as a check of other indirect methods.
Various indirect chemical methods have been used. Nitrogen
detci miualioiis are generally found to check satisfactorily with
-------
.176
MONO!)
dry weights. When iiillines .lie grown on media containing an
;iiitf11 r111i11;11 salt as sole souice of nil rogen, estimations of (lie de-
iii'.iso of fiee a111¦ 11<>iii.i in the mt¦< 1 iiiin appear If) give adequate
I esults (7). I'.si i ma I ions of metabolic ad i\ it y (oxygen con sum pi ion,
arid production) may lie convcnienI (S), lint llieir use is obviousl\
/
Fir. 2. --I..11; linn- .iml l.i(^. Soliil linr --- nliscrx cil ^roulli. Duller I
luir -" !¦ li-.il unmili" (uiilmiil l.ij;). 7i~ l.ii; limp. L~ growth l.ij;. (Soe to\l
|. '75 )
mtv limited. ("en 11 ifui; il 1 e« I ¦ 11 it j lies have licen found of v.ilue
The most w idelv used met hrids, 113 f.u , ai e based on del01 mi 11.1-
I ions of tr.111 sin i II ed 01 si .11 li 1 ed I in lit (Actually, the in I rod net ion
.11 < 11111 1 ini iples, .ne in use. 'I he leadings of these inslin-
iiu-ii(s aie oflen ipioted without refeieiuc to direct estimations a.;
aihilraiy units of tiubidilv, the wind being used in an undefined
GROWTH OF BACTERIAL CULTURES
377
sense, or as "galvanometer deflections" which is worse. This
practice introduces no little confusion and indeterminacy in the
interpretation ami comparison of data. It should l)c avoided.
Whatever instruments are used, the readings should be checked
against bacterial density or cell concentration determinations, and
the checks should be performed not only on different dilutions of
a bacterial suspension, but at various times during the growth of
a control culture. Only thus will the effects of variations of size
of the cells be controlled. Without such controls it is impossible
to decide whether the readings can be interpreted in terms of
bacterial density or cell concentration, or both, or neither.
Actually, the instruments best fitted for the purpose appear to
be those which give readings in terms of optical density (log h/I).
With cultures well dispersed, it is generally found that optical
density remains proportional to bacterial density throughout the
positive phases of growth of the cultures (11). When this require-
ment is fulfilled, optical density determinations provide an ade-
quate and extremely convenient method of estimating bacterial
density.
It is often convenient to express optical density measurements
in terms of cell concentrations. For this purpose, the two estima-
tions should be compared during the exponential phase. The data,
expressed as cell concentrations, may then be considered as refer-
ring to "standard cells," equal in size to the real bacteria observed
during the exponential phase, larger than bacteria in the stationary
phase and probably smaller than those in the acceleration phase.
Cell concentration.—Cell concentration determinations are per-
formed either by direct counts (total counts) or by indirect (viable)
counts. The value of the first method depends very much on
technical details which cannot be discussed here. Its interpretation
depends on the properties of the strains (and media) and is
unequivocal only with organisms which do not tend to remain
associated in chains or clumps. Total counts arc evidently mean-
ingless when there is even a slight tendency to clumping.
The same remarks apply to the indirect, so called viable, counts
made by plating out suitable dilutions of the culture on solid media.
The method has an additional difficulty, as it gives only the num-
ber of cells capable of giving rise to a colony on agar under condi-
tions widely different from those prevailing in the culture. Many
organisms, such as pncuinococci (12), arc extremely sensitive to
-------
MON'OI)
sudden changes in the composition of llic medium. The mere
absence of a cat lion source will induce "flash lysis" of Bacillus
*ub!ilis (13). Such effects may be, in part at least, responsible for
I he discrepancies often found between total and viable counts.
In spite of these difficulties viable counts retain the undisputed
piivilcge of being by far the most sensitive method ami of alone
pei mining ililTeieiili.il
-------
tsu
MO\'0l>
low im>1111 to eliminate inlerfeience fiiim other potential limiting
f.ti tors (|• 11 changes, :ic( ninula t ion of metabolic products, etc.).
Wiiliin the limits thus -l'im>-• I, the lelalion between G ami the
mili.11 ronceiili alion (O "f ill'" null ifnt is, as a very crucial Mile,
found to lie the si1111>11-.-1 po^ililc, namely, linV.ir anil lo confoim
i>>t Ik- equal ion :
<; - kc.
I 11is relation i111111 i<¦ s (lint tlir amount of limiting; nutrient used
up in the formation of a unit quantity of cell substance is inde-
pendent of tIn- (-oiu t-iilr.ition of the nnliient. It implies also tliat
j r > >\\ I li slops only when I In: limiting nutrient is completely ex-
hausted, or, in oilier winds, tli.it then- is no tlirrsliold concenlia
11ori below w 11it'll grow ill is impossible (11).
Neither of these conclusions can In; considered strictly true of
course, and tlic lineai iclalion ennnot lie taken for granted >i
i. Milt it does seem lo lie a general approximation, and oven a
arkably accurate one in main- cases (I'ig. 3 and Table I). Wlieie
z
I
gso
<
H
o
h
25
t^.
i
r»o
t —» . . I—
100 ISO
ITU .
|t(] MANNl TOL PER cm
200
l-ir,. .V— Tril.ll Kiuwlli <>f /¦' roll in s\ nlliclic medium wilh organic siinici'
11 i I ¦ > I) ,1-, I ii 11 i I in ^ l.i' l"i 1 li ilin.iti .irliili.M >' iinils. One iinil is ci|iiiv.ilint
li. U ft fi%. ill y u c i^lil | ii I ml til)
it holds, the estimation <>f G affords a simple and direct measine
nt I he on 111 \ ield I A") on I lie limil ini; nut i ient, or 1
<;
c
- K
;iin<1111 df hat leiial substance formed
annum! of limiting nutrient utilized
GROWTH OF HACTRUIAL CULTURES 381
When the proportion of the dry weight representing substance
derived from the limiting nutrient is known, it is a measure of
the fiaction assimilated. If G is expressed as "standard" cell
concentration, 1/A' represents the amount of limiting nutrient
used up in the formation of a "standard" cell. Thus, when de-
termined under proper conditions, G is a constant of perfectly clear
and fundamental significance; it is a measure of the efficiency of
assiniilatory processes.
Taiii.i: I
ToIjI giowtli of purple bacteria with acetate as
limiting factor (after Vail Nicl (9)|
Acetate
(mg./iiil.)
0.5
1.0
2.0
3.0
Total growth
(ihk/iiiI.)
0.18
0.36
0.70
1.12
K
0.36
0.36
0.35
0.37
Fxtensive data on G and K values are available only with
respect to the organic source (9, 11). Little is known of K values
in the case of inorganic sources. Owing to the development of
microbiological assay methods, abundant data are available on
the quantitative relations between growth of many bacteria and
concentration of a variety of growth factors. lJut the major part
of these data do not bear any known relation to G or any other
definable growth constant, which is most unfortunate. It does
seem at least probable that in many instances, the measurement
of total growth, under conditions insuring homogeneity and
limitation of growth by a single factor, could with advantage re-
place estimations of "turbidity at 16 hours," or "galvanometer
deflections at 24 hours." It can be predicted with confidence that
in most cases linear relations would be found [see e.g. (44)], per-
mitting the estimation of K, and on which simpler and more
reproducible methods of assay could be based. Furthermore, an
intelligible and very valuable body of quantitative data on nutri-
tional requirements of bacteria would thus become accumulated.
The remarkable degree of stability and reproducibility of K
values, for a given strain and a given compound under similar
-------
.382 MONO!)
ruiuli(ions, should be emphasized. Contrary to the oilier growth
constants, it scorns to lie very little affected by hereditary vari-
ability (45):
In genoial, of the three main growth constants, total growth is
the easiest to measure with accuracy and the most stable, lis
intei pretaliou is simple and sti aiglrt foi ward, provided certain
experimental tctpiiiemeiils are met. These are remarkable proper-
ties, which could, it seems, be put to much wider use than has
hitherto been done, especially with the focussing of attention on
problems of assiinilatoi y and synthetic metabolism.
I'.xi'onlni iar. Growth Katk
The exponential phase as a sternly state: rate determining ste[>s.~ -
'I he rate of growth of a bacterial culture represents the over all
velocity of the series of reactions by virtue of which cell substance
is synthesized. Most, if not all, of these reactions are enzymatic,
llio majority probably are reversible, al least potentially. The rate
of each, considered alone, depends on the concentrations of the
roactants (metabolites) and on the amount of the catalyst
(enzyme).
During the exponential phase, the growth rate is constant. It
is reasonable to consider that a steady state is established, where
the relative eonecnliatioiis of all the rhetabolites and all the
enzymes are constant It is in fact the only phase of the growth
cm le when the properties of the cells may be considered constant
and can be dcsciibcd by a numeric value, the exponential growth
i,lie, corresponding to the ovci-all velocity of the steady slate
svstein.
It h as often been assumed th.it the over-all rate of a system of
linked reactions may be governed by the slowest, or master,
reaction. That Ihis comcplion should be used, if at all, with
extienic caiition, has also been emphasized (17, 18). On theoretical
grounds, it can be shown that the over-all rate of a system of
several consecutive ie\ersible enzvniatic reactions depends on the
rate and cquihbi imn constant of each. The reasons for this are
obvious, and wc need riot go into the mathematics of the problem.
A master reaction could take control only if its rate were very
iniu h slower than that of all theothei reactions. Where hundreds,
pei liaps thousands of reactions linked iu a network rather than
.is a chain aie ( on< einei 1, as iit the growthfof bacterial cells, such a
GROWTH OF UACTtiRIAt. CULTURES
383
situation is very improbable and, in general, the maximum growth
rate should be expected to be controlled by a large number of
different rate-determining steps. This makes it clear why ex-
ponential growth rate measurements constitute a general and
sensitive physiologic test which can be used for the study of a wide
variety of cfTccts, while, on the other hand, quantitative inter-
pretations arc subject to severe limitations. liven where the condi-
tion or agent studied may reasonably be assumed to act primarily
on a single rate determining step, the over-all effect (i.e., the growth
rate) will generally remain an unknown function of the primary
effect.
Although very improbable, it is of course not impossible that
the exponential growth rate could in certain specific cases actually
ce
o'-5
X
1.0
las
1—i—J—f
i A, i—ir
M*I0 GLUCOSE
Fro. 4.—Grow lli rale of E. coli in synthetic medium as a function of glucose
concentration. Solid line is draw n to equation (2) with Rk " 1.35 divisions per hour,
and Ci =0.22 M X 10_' (I l). Temperature 37° C.
he determined by a single master reaction. But such a situation
could hardly be assumed to prevail, in any one case, without direct
experimental evidence. Some recent attempts at making use of the
master reaction concept in the interpretation of bacterial growth
rales arc quite unconvincing in that respect (19).
Rate-concentration relations.—Notwithstanding these difficul-
ties, relatively simple empirical laws arc found to express conven-
iently the relation between exponential growth rate and concentra-
tion of an essential nutrient. Examples are provided in Figs. 4 and
S. Several mathematically different formulations could be made to
fit the data. Hut it is both convenient and logical to adopt a
hyperbolic equation:
R = rk [2]
C, + C
-------
SI
MONOI >
imilar to nil adsoiption i so 111 <¦ 1111 or to (lie Michaelis equation.
ii tin- aliovi.' «'(|ii:iti1 ,1/ liibrii ii/nw'i in IIiiIkis' imiliimi, as n fimrliini of
|Hi, u cnliiili'iii. Sutnl line iIi.iwii lo cc|H.iIi'>u (2) willi Wk—0 0(7 anil
( , - M/IS (2(1)
I
.i i;.iiii< unnces are i rnnp.itrd under other n ise identical conditions.
I In IV is no doubt lli.it it is tel,ited to (lie activity of (lie specific
n/\ nie s\s|ems iii\-«»l\<¦
-------
386 MONOI)
knowledge, tliis has been cleaily observed onl>' once (25), actually
during tlie residual growth of n st reptomycin-rec uiring B. sitbtilis
in a medium <-on(.lining no streptomycin (Tig. 6). The interpreta-
tion is obvious, albeit surprising. Growth must be limited by one
enzyme or s\ stem of enzymes, the activity of which is constant. In
other words, in the absence of stieptomycin, one rate-determining
enzyme ceases to be formed, so that by being outgrown by the
MIMU7ES
300
I k;. 6 Uc«iiln.il growth of a siH'plnin) ui> i<*«|iiiriti^-siin of Bacillus subtilis in
1 hi' <• of si11-1»Y«iiiin t in. (•! mw Ih is linr.iij for uvcr •! !ir. (2.S).
other enzjmrs, it eventually achieves tine, mastery and sets the
system at its iikii constant pace, disregarding the most funda-
mt ntal law of growth. 1
Similar svsleins of a steady state. The growth lag (L)
may be considered .i measure of the physiological distance between
the initial and the steady slate. I )epending on the specific condi-
tions and pioperlie-. of the organism, one or several or a large
number of reactions may deleimine the rale of this building up
GROWTH OF IJACTMRIAL CULTURES
387
process. Furthermore each rate-determining reaction may be
affected in either or both of two ways: (a) change in the amount
and activity of the catalyst; (b) change in the concentration of the
react a nts (metabolites).
When the phenomenon is associated with the previous ageing
of the cells of the inoculum, the chances are that it involves at
once a large number of reactions, and specific interpretations arc
impossible. Furthermore an apparent lag may be caused if a large
fraction of the incoulated cells are not viable (18). When, however,
the lag can be shown to be controlled primarily by only one re-
action, or system of reactions, the measurement of lag times
becomes a useful tool for the study of this reaction. This may often
be achieved by a careful preconditioning of the inoculated cells,
and appropriate choice of media [see e.g. (26)]. In point of fact
this technique amounts to artificially creating conditions where
one or a few rate limiting steps become true master reactions, at
least during the early stages of the lag.
Theoretically, the lagging of a reaction may be due either to
insufficient supply of a metabolite or to the state of inactivity of
the enzyme. In the first case, the technique may be used for the
study of certain essential metabolites synthesized by the cell itself
during growth, and consequently difficult to detect and identify
otherwise. Few examples of this sort are available besides the
glutamine effects studied by Mclbvain et al. (27) and the detection
of metabolites able to replace carbon dioxide (26), but it is probable
that the method could be developed.
In the second case, the technique may be useful in the study
of enzyme activation or formation. The magnesium effects de-
scribed by Lodge & llinshclwood (28) and the sulfhydryl effects
dcsciibcd by Morel (29) should probably be attributed to the
reactivation of certain enzymes or group of enzymes. However,
lag effects arc especially interesting in connection with the study
of enzymatic adaptation.
Lag and enzymatic adaptation.— Enzymatic adaptation is
defined as the formation of a specific enzyme under the influence
of its substrate (30). If cells arc transferred into a medium con-
taining, as sole source of an essential nutrient, a compound which
was not present in the previous medium, growth will be impossible
unless and until an enzyme system capable of handling the new
substrate is developed. If other potential factors of lag are elimi-
-------
MONOD
11.1 Icil, tlir t lilri inin.i linn of |,n> I imi-s beet imt-s a means tif si n < I \ ing
tIn' :nI.*|*1 i\'t- |iinprrlits nf tin* cn/\me svslein involved (Fig. 7).
Tin' technique lias pioved cspet ially usi'fill for (lie stink of
adaptive «-ii z\' im-s a 11 ackiug oigauic compounds serving as suit'
in nanic source (II, 31). Tlir work of Pollock (32) shows lliat il
i an also In- applit'il in Ilu: cast.- of adaptive systems specific for
icrlain liytlto'j.cn acceptors (nilralc anil tellatliionatc). A further
development of llif t<-e l111i(11u- is suggested liy tlie wurk of Sl.iuier
a ....... ^ j.. 3 ^ k t~
r, HOURS
lie;. 7.- Gmw ill nf I'- f"li in "viilli| iii.i in t;i i iici I oil araliinocr
mi ilium, Icm[M'-r.ilurr .17" < iura ill on jjliuosi' pmccrils without any lap. l.a^
time ('/'j) mi liiM! i>- ,i11|H .»\ini,iI i\ i-Iv - S hours {t<>). '
(3. -1 i 111 • * iiH asini'iiiiiils ma) lie especially useful in
111, • deletlion and pi eliminai y idenlilication of atlaptive efTects,
|.i!I tltiv could not, of i muse, n-placc limit: tlirecl methods of
r -1 i ma I ini; cn/\ ni.il ic .u I i vi I ics.
A luoailer .1 ppi • i.k 11 In I In: pinMcm of relations between lag
and en/\malic adaptation should also lie lonsidered. As empha-
sized 11 \ I lin-lieluniiil (IS), 11 if Ian and acceleration phases repie-
M-iit es^entialU' a pi i ices*. of eqiiilibi al inn, lilt! functioning of a
iiynlalnry inci haiiisin, l>> viilue of which a ceitain eii/.\iiie
|, a 1. 111 < ¦' inside I hr till-, is a 11 a i 111'< I. That such a mechanism must
i \isl is i > I > \ ii his, hime in its absence, the cells coultl not sin vivo
i \ en slight v.nialions of I lie external environment. However, the
ii., I hi c (,r ilii' posl ula I id niei hanisms is slill completely obscure,
the kinetic speculations of 11 inshelu ood, although intcicsting as
cmpiiical fm in ula I ions of the problem, do not throw any lighl on
GROWTH OF DACTI'KIAL CULTURES
389
the nature of the basic mechanisms involved in the regulation of
enzyme formation by the cells.
The most promising hypothesis for the time being appears to
be that this regulation is insured through the same mechanism
as the formation of adaptive enzymes, which implies the assump-
tion that all the enzymes in a cell are more or less adaptive. The
competitive efTects observed in enzymatic adaptation (11, 35, 36)
agree with the view that the regulation may be the result of a
continuous process of selection of mutually interacting enzymes
or enzyme-forming systems (30, 37). The kinetics of bacterial
growth and, in particular, the lag and acceleration phases certainly
constitute the best available material for the study of this funda-
mental problem.
Division lag.—The largest discrepancies between increase in
bacterial density and increase in cell concentration are generally
observed during the lag and acceleration phases. This phenomenon
lias been the subject of much confused discussion (38). Actually,
it has been demonstrated by Uershey (39, 40) that a definite lag
in cell concentration may occur even when there is no detectable
lag in bacterial density. This must mean that cell division mecha-
nisms may be partially inhibited under conditions which do not
afTecl (lie growth rate and general metabolism of the cell. A num-
ber of interesting observations by 1 linshelwood et al. (18) point
to the same conclusion. Further studies on the phenomenon are
desirable, as they should throw some light on the factors of cell
division in bacteria.
Tllli INT1LHPHKTATION OF C'OMPMIX GROWTH CYCLES
Afulliple exponential phases.—In many cases, the growth
cycle docs not conform to the conventional scheme represented
in Fig. 1. The interpretation of these complex growth cycles will
be briefly discussed here.
One of llie most frequently encountered exceptions is the pres-
ence of several successive exponential phases, characterized by
different values of Ii and separated by angular transition points.
This should in general be interpreted as indicating the addition
or removal of one or more rate-determining steps in the steady
state system. This type of effect may result from a change in the
composition of the medium, for instance from the exhaustion of a
compound partially covering an essential nutritional requirement
-------
390
MONOI)
(34), or from t!ic transitory accumulation of a metabolite, which
will eventually serve as a secondary nutritional source (41).
Interpretations are more ilelicate, anil more interesting, when
the cause is a change in the composition of the cells themselves.
-Sik h effects are fic
-------
MONOD
determination:, of "Uu l)inical Kinetics of the Bacterial Cell, 284 pp.
(Clarendon Press, Oxford, 1916)
19. Johnson, F. II., and Lewis, I., J. Cellular Comp. Physiol., 28, 47 (1946)
20. SellAF.Fr.R, W., Ann. inst. Pasteur, 74, 458-63 (1948)
21. McIi.wain, II., Biol. Revs., 19, 135 (1914)
22. McIi.wain, II., Advances in Enzymol., 7, 409-60 (1947)
23. Wyss, O., Proc. Soc. Exptl. Biol. Med., 48, 122 (1941)
24. Koiin, II. I., and IIakris, J. S., J. Pharmacol. Exptl. Therap., 73, 343 (1941)
25. ScilAt.FFFR, P., Compt. rend., 228, 277-79 (1949)
26. Lwoff, A., and Monod, J., Ann. inst. Pasteur, 73, 323 (1947)
27. McIi.wain, II., Fii.uks, P., Gladstone, G. P., and Knight, B. C. J. G., Bio-
chem. J., 33, 223 (1939)
28. Lonci;, R. M., and IIinsiielwood, C. N., J. Chem. Soc., 1692-97 (1939)
29. Moriu., M., Ann. inst. Pasteur, 67, 449 (1911)
30. Monod, J., Growth, 11, 223-89 (1947)
31. Monod, J., Ann. inst. J'asteitr, 69, 179 (1913)
32. Poi.i.ock, M, R,, and WAlNWKlcair, S. D., Brit. J. Etptl. Path., 29, 223—10
(1948)
33. Stanier, R. Y„ J. Bad., 54, 339 (1917)
34. Cohen, S. S., J. Biol. Chem., 177, 607-19 (1949)
35. Monod, J., Ann. inst. Pasteur, 71, 37 (1945)
36. SriKGHi.MAN, S., and Dunn, R., J. Gen. Physiol., 31, 153-73 (1947)
37. Sruxi'LMAM, S., Cold Spring Ilarbor Symposia Quant. Biol., 11, 256 -77 (1946)
38. Winsi.ow, C. E., and Wai.kI'.k, II. II., Bad. Revs., 3, 147-86 (1939)
39. llr.RSHEY, A. D., J. Bad., 37, 290 (1939)
-------
<9-1
MONOD
ID. llKRSiirv, A. D., Proc. Soc. Exptl. Biol. iftd., 38, 127-28 (1938)
11. Lwuff, A., Cold Spring Harbor Symposia Quant, Biol., It; 139-55 (1946)
12. I.emon, C. G., J. Hyg .. 33, 495 (1937)
1 o['iiy, W. W. C., and Wilson, G. S., Principles of Dacteriology and Itnrnu•
ntly, 3rd E
-------
Dynamic Model of Nitrification
in Fluidized Bed
By David K. Stevens,1 Associate Member, ASCG, P. Mac Berthouex,1
Member, ASCE, and Thomas W. Chapman3
Abstract: Fixed-film nitrification was studied in a pilot-scale fluidized bed treating
municipal secondary effluent. A mechanistic mathematical model incorporating re-
action stoichiometry, diffusion, multisubstrate kinetics with product inhibition,
fluidization. and a reactor model developed from the observed residence-time dis-
tribution, was developed to predict the steady-state and short-term dynamic per-
formance of the reactor. The model equations were solved using orthogonal col-
location with trial functions tailored to the spherical-shell biofilm geometry, and
a scmi-implicil third-order Kungc-Kuttu integration technique. Tlte slcudy-slate model
closely fit measured concentration profiles using the maximum specific growth
rates for Nilrosomonas and Nitrobacter as adjustable parameter*. The dynamic model
predicted observed responses to step and impulse changes In ammonium and nitrite
concentrations to within 3 mg/L without further parameter adjustment. The fluid-
ization model predicted the observed fluidized bed height within I cm. External
mass transfer resistance wjs small for the conditions employed und was not in-
cluded in the model.
Introduction
The number of wastewater ireatment plants requiring nitrification is in-
creasing with the greater reliance on whole-effluent toxicity testing for set-
ting discharge standards. Economical alternatives to conventional treatment
that minimize space requirements are needed in order to incorporate nitri-
fication into existing treatment process streams. One proven alternative is
the fixed-film fluidized-bed nitrification process, which offers the stability
of a high-biomass system at a short hydraulic residence time, and is efficient
over wide ranges of influent characteristics and flow. To design and operate
these systems effectively, however, a more thorough understanding of the
factors governing fluidized-bed nitrification is needed. Further, an under-
standing of .system dynamics is essential since Hows, temperatures, and in-
fluent concentrations arc all time-varying.
Mueller ct al. (I9K0) first applied (lie principles of diffusion with reaction
to fixed-film nitrification in a study, of rotating biological contactors, using
miiTohiiil kinetic equations with tip to (Incc coupled Monod terms. Liquid
und biolilm mass transfer resistances were included lor lour rcucting species.
For simplicity, they ignored nitrite oxidation and did not consider the effect
of pH on reaction rates or inhibition of ammonium oxidation by nitrite. Her-
niMiiowie/ and Ciam /an/.yk (I'HW) used a Monod kinclics-hnscd tmslcady-
state model to ptedici long-tcim dynamic bcliuvior of tlic l'luldi/cd-bcd ni-
'Asst. Prof., Dept. of Civ. and Envir. Engrg.. UMC 4110, Utah State Univ.,
Loean. UT 84322-41 10.
M'lof . Dcpt nf Civ ami P.nvii , Univ. of Wisconsin-Madison, Madison,
Wl 5J7UO.
'Prof., Dept. of Chem. Engrg., Univ. of Wisconsin-Madison, Madison, Wl.
Note. Discussion open until March 1, 1990. To extend the closing date one month,
a written icqucst must he Hied with the ASCE Manager of Journals. The manuscript
lor (Ins paper w.is submitted lor icvicw and possible publication on Mnrcli 9, I9H9.
This paper is part of the Journal of Environmental Engineering, Vol. 1 15, No. 5,
October, 1989. ©ASCE, ISSN 0733-9372/89/0005-0910/S1.00 + $.15 per page.
Paper No. 23916.
910
trification process. Their model included biofilm accumulation and a time-
variable microbial activity that adjusted to substrate levels. They used single
substrate limitation and ignored the effects of pH and product inhibition.
More recently, Siegrist and Gujer (1987) included the effects of pH in a
model of ammonium and nitrite oxidation in a laboratory trickling filter to
predict the short-term effect of depressed pH on ammonium uptake rates.
Although these models include many important features of the fixed-film
nitrification process, including mass transfer and multisubstrate limited ki-
netics, they are limited in that they include only one of the two nitrification
reaction steps, or ignore the effect of pH on reaction rates, or exclude the
inhibition of NH^ oxidation by NOf, or ignore system dynamics. These
models are unable to simulate short-term dynamic behavior, which is im-
portant in operation and control of processes where the pH is variuble und
the short-term buildup of NO^ can cause breakthrough of NH4* into the re-
ceiving water.
The objective of this study was to develop a fluidized-bed nitrification
model based on the principles of multisubstrate Monod kinetics coupled with
mass transfer resistance that incorporates the following features: (I) Reaction
stoichiometry; (2) microbial growth and substrate removal, including N02"
inhibition of Nilrosomonas and the effects of temperature and pH on nitri-
fying organism growth rates; (3) mass transfer resistance; (4) a reactor model
based on observed residence time distributions; and (5) fluidization, for steady
and unsteady state. The model was tested by comparison with steady slate
and dynamic fluidized-bed pilot-scale plant data.
Nitrification Model
Reaction Stoichiometry
The model stoichiometry for the two-step nitrification reaction was derived
using, the procedure of Christensen and McCarty (1975), who used ther-
modynamic principles to predict the relative amounts of each chemical spe-
cies used for energy production and cell synthesis. Using their notation, the
basic stoichiometric equation is
K - Kj-f.K. -f.R. (I)
where /? = the overall stoichiometric cquution; und /?„. /?,. and /?., = the
half rcuctions for the reduction of oxygen, cell synthesis foi inorgunic sub-
strates, and oxidation of NH4+ or NOj", respectively. The factors/, and/, arc
the fractions of the electron-donor energy content associated with energy
production and cell synthesis, respectively, and /, + /",= I. The factor f, is
a function of cell ugc und dccuy rule und is icpicsentcd l>y
1 + 0.20,6
~ i + e.b (2)
where O, = the cell age (7*); b = the decay rate (1/7"); and a, = the growth-
yield coefficient (electron equivalent of cells produced divided by electron
equivalent of substrate consumed), calculated by the method of Christensen
and McCarty (1975).
Eqs. 1 and 2 were applied for the two nitrification-reaction steps to give
the overall stoichiometric relations. The stoichiometric coefficients for each
911
-------
TABLE 1. Stoichiometric Coefficients for Reactions 1 and 2
Species
Reaction 1
Reaction 2
(1)
(2)
(3)
nh;
1
/a/10
o,
/../[< 2/3) + (/„/5)]
/ij/2
NOj"
1/U + (/„/10)]
1
HCOJ
[(4/3) -/„ - (19/20)/„]/[(l/6)
fj 10
+ (/„/ 20)]
HjCo;
((4/3) -/„ - (6/5)/„]/[(l/6)
2/5/.,
+ (/.,/20)]
no;
0
1
H,0
|(/„/2) + (13/20)/,. - (l/3)]/[(10/3)
/., + 13/10/,, - 1
+ /.,]
CsH,NOj biomass
/n/[(10/3) +/„]
/.i/10
residing species were identified as v,where I indicates the species (1 =
NH;, 2 = 02, 3 = NOj, 4 = HCOj, 5 = H2CO?, 6 = N03~, 7 = H20,
8 = biomass, 9 = C02, and 10 = H + ) and m indicates the reaction step (m
= 1 or 2 for oxidation of NH4* or N07, respectively). Defining the combined
lonn ll;COf = COj(«) + ll:CO.,, the equilibrium relation C02(a<7) + H20
= H2COT was used to eliminate H+ and C02(aq) from these equations and
to put the carbonate system and pH in terms of the more convenient
HCO," and HjCO? forms. The resulting stoichiometric equations are, for step
1 (hi = 1, a, = 0.109)
v, ,NH^ + v2,02 + i»4i|HCOj- = vg.|C5H7NOj
+ Vj,,N02" + v3,,H2CO? + v7|H20 (3)
and for step 2 (m = 2, a, = 0.095)
vJ.jNOr + v2iJ02 + v3i2H2CO* + v4i2HCOj
+ v,.jNH; = v8.2C3H7NO: + v,.2NOj" + v7.jH20 (4)
The pH was then calculated from the definition of the dissociation constant
(H2CO*)
PH - pK. - log, UHco.(J
Table 1 summarizes the coefficients for Eqs. 3 and 4.
(5)
Microbial Growth
The growth rates lor nitrifying bacteria arc usually assumed to follow the
Monod kinetic model with cell decay (Stratten and McCarty 1967; Downing
and Hopwood 1964; Knowlcs ct al. 1965). Stenstrom and Poduska (1980)
showed that low dissolved-oxygen concentration may also limit the growth
rate of nitrifying bacteria. Nitrification rates may also be inhibited by high
concentrations of ammonium or nitrite. Kccnan ct al. (1979) noted that am-
monium did not inhibit its own oxidation at concentrations typical of do-
mestic wastewaters, but Boon and Laudelout (1962) found that N02 con-
centrations above 260 mg/L-N inhibilcd the oxidation of nitrite by Nitrobacter
912
winogradski. Anthoniscn et al. (1976) found that each bacterium was more
sensitive to the concentration of the other's substrate than to its own. Indeed,
Engel and Alexander (1958) observed that Nitrosomonas was inhibited by
levels of 34 mg/L NOJ-N, but was not inhibited by 640 mg/L NH,-N.
Since NOj is a product of NhC oxidation, the product inhibition modi-
fication of the Monod model (Zeffren and Hall 1973) was used in this study
to predict the growth rate of Nitrosomonas, assuming that nitrite inhibits the
oxidation of ammonium and that other inhibition effects are negligible. This
model was multiplied by an oxygen-switching function to account for 02
limitation (Stenstrom and Poduska 1980): the rate expression is given in Fig.
1, row 1, column 9. In Fig. 1, the model equations for microbial growth,
substrate removal, mass transport, and reactor hydraulics are presented in
the matrix form of Petersen (1965), as suggested by Henze et al. (1987),
using the notation adopted by the International Association of Water Pol-
lution Research and Control (IAWPRC) (Grau et al. 1982). Here S, is the
local concentration of the species / (M/L3), M-m^.i 's 'he maximum specific
growth rate for Reaction 1 in Table 1 (M volatile solids/A/ volatile solids-
T), a, is the pH dependence factor for reaction 1, K„ and K,2 are the half-
saturation coefficients for species NH« and 02 (M/L1), and K, is the inhi-
bition coefficient (M/L3). The growth rate for Nitrobacter. assumed to be
limited only by the concentrations of N02 and Oj, is given in Fig. 1, row
2, column 9. Here a2, and K,5 and K,t are defined analogously to
those for the growth of Nitrosomonas.
The reaction-rate functions pi(S) and p2(S) in Fig. 1, column 9, were in-
troduced to simplify equations developed for substrate removal. These func-
tions reflect the effect of the local concentrations of the major substrates,
NH<, 02, NOi", HCOj", and H2CO*, on the growth rates of each organism
in the biofilm. They are functions only of local substrate concentrations.
Substrate Removal
The rate of consumption or production of species / in reaction m is pro-
portional to the growth rate of the organism involved in that reaction. For
the model developed in this study, the overall reaction-rate expression was
derived by adding terms in rows 1 and 2 of Fig. 1:
Pi(S) Pa(S) ,
ru, (6)
Yia r,.t
where Y,,m = the organism yield coefficient for species / in reaction m. This
equation was used to predict the consumption of ammonium (/ = 1), pro-
duction and consumption of nitrite (I = 3), the consumption of bicarbonate
(/ = 4), and the production and consumption of H2COJ (/ = 5).
A term for endogenous respiration, where mg cells decay per milligram
of 02 consumed, was added to Eq. 6 to model the consumption rate of dis-
solved oxygen (/ = 2):
Pi(S) P2(S) (X, + X2)b
fv2 — (7)
I'm Yu
Temperature Dependence or Kinetic Parameters
The maximum specific growth rates, saturation coefficients, and endog-
enous decay rates of nitrifying bacteria arc kinetic parameters that arc tcm-
913
-------
¦i >
+
¦f
¥
in
i
o
>/
cX
+
+
J
E
*
0
rs
X
» I
o o
l/l
y o
X)
?.Z
7i>?
in
o
o
o
I:
o
o
o
w
4>
o
o
z
T|,«
o tr
z
C?
9 •,
5 «
1/1 Q.
a a o
n ti
c t.
Q <0 Q
o
co"o i/t"
ii n ii
co"*ii/T
N
*5
or s
) " (»)
where T is in degrees C; Te = the reference temperature; (nm„,K„b)0 = the
value of the parameter at the reference temperature; and k = an empirical
constant. Values used for the kinetic parameters at the reference temperature
and k are shown in Table 2. The values in Table 2 were taken from Stan-
kewich and Gyger (1976), who summarized the results of several authors
reporting these values. No temperature dependence for Kj2 and K,t has been
reported and they were assumed to be temperature independent and equal to
0.5 mg/L.
pll Dependence of Growth Rates
Nitrification rates are a function of pH over the range commonly found
in nnturul waters (Huang and Hopson 1974; Srinath ct nl. 1976). Below pH
6, nitrification rates are very low. The rates increase to a muximum ut ap-
proximately pH 8.6 for Nitrosomonas and pH 8.0 for Nitrobacter. The de-
pendence of the growth rates on pH was modeled by multiplying the max-
imum giowili rules (x..,.,,, and for cucli step by pll correction factors.
ot| and oij. The pH correction factor for Nitrosomonas. a,, was determined
from an empirical approximation of the data of Huang and Hopson (1974),
who studied the dependence of growth rate on pH for Nitrosomonas in at-
tached growth systems:
a, = 0.22 + 0.39(pH - 6) - 0.014(pH - 6)J + 0.012(pH - 6)3 (9)
The data of Srinath ct a). (1976) provided an approximation for Nitrobacter:
a2 = -l.64(pH - 6) + 4,94(pH - 6)1 - 3.63(pH - 6)3
+ 0.87(pH - 6)4 (10)
These approximating functions were used to predict the pH dependence of
pH values ranging from 6 to 9 and 6 to 8 for a, and otj, respectively.
Mass Transfer and Bioparticle Model
Mass transfer limitations resulting from the biological film in the fluid-
ized-bed process were incorporated into the model. Various researchers have
915
-------
found these limitations to be important (Mulcahy and Lamotta 1978; Mul-
cahy et al. 1980; Nutt 1980; Stephenson and Murphy 1980).
External Mass Transfer
External mass transfer is the transport of chemical species to the surface
of a solid particle through a stagnant liquid film surrounding the particle.
The flux, N,i, of a species I through the stagnant film is assumed to be
proportional to the concentration difference between the bulk solution and
the particle surface (Bird ct al. I960):
V„ = USlb - S„) (11)
where S,b and Su = the bulk-solution and particle-surface concentrations. The
overall mass transfer coefficient, kt, incorporates diffusion and convection
mass transfer piocesses. Estimating k, liom first principles is not possible,
and empirical correlations have been proposed for different reactor condi-
tions (e.g., Toumic ct al. 1979). The Gupta and Thodos (1962) correlation
lor mass transfer between spheres and liquids in fluidized beds was used to
picdicl k, I'm Reynolds numbers, R, greater than I:
U,Sc~2n ( 0.863 \
k, = — 0.01 + — (12)
e V R0 58 - 0.483/
wlieic R - '//.J»)/1L; ihc Sclimiill number, St' ^/(/^p); IJ, s die su-
perficial fluid velocity (1/7); dp = the particle diameter (L); p = the ltuid
density (A//L3); p. = the fluid viscosity (M/LT), D, = the diffusivity of the
species in the fluid (L?/T)\ and € = the porosity of the expanded bed. At
steady state, the flux due to external muss transfer (Eq. I I) is equal to the
reaction rate (assumed to follow the Monod model):
" (13)
where af = the particle surface area per unit biofilm volume
The effectiveness factor for external mass transfer is defined as the ratio
of the reaction rate that would be observed under the influence of external
mass transfer resistance to that which would occur for zero external resis-
tance (i.e., S„ = Si„):
S„(K, + Slb)
ti, = (14)
+ S„)
Solving Eqs. 13 and 14 for Sh:
S„ = ^-(K* + Da - 1) + V(K* + D„ - 1)J + 4K*] (15)
where D„ = (p.™,,„Xm)/{afYKmk,Slb) is the Damkohler. number, the ratio of
the maximum reaction rate to the maximum mass transfer rate; and K* =
K,/S,b. Using Eqs. 14 and 15, t|, was plotted as a function of Dc for various
values of K* as shown in Fig. 2. If Da > 1 (ti, < 1), then external mass
trrn fer is important and should be included in the model.
Using input parameters in Tabic 3, the value of ti, from Eqs. 14 and 15
916
917
-------
fABLE 3. Parameter Values for Evaluation of External Mass Transfer Resistance
Parameter
Value
Units
(1)
(2)
(3)
V,
22.2
cm/min
d.
0.09
cm
P
1.000
mg/cm5
(a
600
mg/cm-min
e
0 5
mg/cm-min
em'/min
D,
0 0009
*,
0.34
cm/min
0.00038
1 /min
100
mg/cm1
°i
66
1/cm
y,,
0 18
mg X/mg NHJ-N
K,
0.5
mg/L
S.
667
R
3.33
s n'otted versus the bulk ammonium concentration in the inset of Fig. 2.
The conditions in Table 3 represent unfavorable external mass transfer con-
litions for ammonium (low velocity, fast reaction, and a slowly diffusing
pccics). When 6'„, > 2 mg/L, the effectiveness factor is nearly one and the
ffects of external mass transfer can be neglected. This is in agreement with
he findings of Mulcahy et al. (1980), who did a similar analysis for flu-
dized-bed denitrification.
Fluidized-bed reaciois are operated under conditions that will exhibit ex-
srnal mass transfer resistances much lower than the case shown in the inset
if Fig. 2. Throughout most of the reactor, the ammonium concentrations
re above the limiting values of 2-3 mg/L. It is generally not necessary
:i reduce the ammonium concentration below these levels for pollution-con-
rol purposes. Typicul superficial velocities ure at least twice us high as given
i the table, reaction rates are normally lower, and K, values are typically
ligher. Thus, even under extreme conditions the effects of external mass
ransfer resistance are negligible.
nternal Mass Transfer
For this model, the bioparticlc was assumed to be spherical with radius
I,, consisting of an inert, nonporous sphere of radius Rp surrounded by a
nrous biol'ilm shell of uniform thickness, 8, and density, X. Assuming a
onstant effective diffusivity, D,t (L}/T), the unsteady-state continuity equa-
on for species / in the biofilm, and the boundary and initial conditions, are
iven in Fig. 1, row 6 (Bird et al. 1960). In this equation, r is the radial
istance from the center of the particle (L), pm(S) is the reaction rate term
W/l?T), Su is the concentration of I at the particle surface (A//L3), and t
: time (7"), Since the two steps of the nitrification reaction are functions of
ve species that generally have different biofilm diffusivities, particle-con-
nuity equations were written for each species that potentially controls the
ite of each reaction.
emperaturc Dependence of Effective Diffusivity
The diffusivities of NH4+, NOj, N03~, and HCOJ in biofilms have been
918
TABLE 4. Parameter Values for Temperature Dependence of Diffusivity (Lerman
1979)
Species
D.t. (em'/s)
v, (em'/mol)
Equation
(1)
(2)
(3)
(4)
nh;
9.8 X 10"'
—
17
o,
—
25.6
18
Nor
5.5 x 10~4
—
16
hco;
5.9 x 10"'
—
16
H,CO*
—
34.0
18
Nor
9.8 x 10"'
—
16
measured by a variety of authors (e.g., Williamson and McCarty 1976) to
be 80-100% of those in water. These diffusivities in water, and those of
nonelectrolytes 02 and H2CO*, are functions of temperature. For the model
presented here, the diffusivities of electrolytes were estimated from corre-
lations proposed by Lerman (1979) over the range of 0-25° C:
D,t = D,To{ 1 + 0.047") for anions (16)
D.t = D,r„(l + 0.0487") for cations (17)
The diffusivities of nonelectrolytes (02 and HjCO*) in water were estimated
using (he correlation of Wilke and Chang (1955):
„ (T + 273.15)
D.t = 5.6 x 10-' ^ (18)
where D,T is in square centimeters per second; p. = the viscosity in poise;
and Vb = the standard molal volume of the species (cubic centimeters per
mole). The parameter values for each of these equations are given in Table
4. Since the diffusivities in the biofilm were assumed in this study to equal
those in water, Eqs. 16-18 are assumed to describe the temperature depen-
dence of D,i in biofilm.
Reactor Model
The reactor model developed in this study included transport due to bulk
fluid flow, turbulent mixing, and mass transfer processes at phase bound-
aries. Homogeneous reactions in the fluid and heterogeneous reactions at
phase boundaries were neglected.
An important aspect of any process model is the proper characterization
of the hydraulic residence time distribution (RTD). so that a suitable reactor
model can be identified. The reactor used in this study to verify the proposed
model was characterized as having a partially mixed section followed by a
plug-flow section (Stevens et al. 1986). This configuration was used because
the reactor had a relatively high turbulence level near the inlet, caused by
high local velocities as the flow passed through the perforated flow-distri-
bution plate into the bottom of the reactor. Stevens (1983) found that the
overall hydraulic regime could be simulated using a model of continuous
stirred-tank reactors (CSTRs)-in-series (Levenspiel 1972) with unequal vol-
ume reactors. This model fit the observed residence-time data well and was
easily adapted to dynamic simulation.
919
-------
The CSTRs-in-scrics model assumed that (lie reactor can be viewed as nt
discrete CSTRs such that the output from the (i - l)th reactor is the input
to the ith stage. The material balance equation for species / in the ith CSTR
is given in Fig. 1, row 8, in which V, is the total volume of stage i (L3), c,
is the porosity, Su and Su., are the concentrations of I (M/L3) in stages i
and i - 1, respectively, a„ is the specific biofilm surface area (1 /L), Nj(i/
is the surface flux of I (M/VT), and D,.,(0 is the forcing function for / (A//
1}T) describing the time-varying inputs to the reactor. The appropriate num-
ber of CSTR stages, n, was determined from the variance, a2, and mean
(retention time), r, of the RTD by n, = tj/ct2 (Lcvcnspiel 1972).
Fluidization Model
The design and analysis of the fluidized-bed biological-treatment process
requires knowledge of the steady-state and dynamic-expansion properties of
the guiwlh-support medium. The fluidization behavior of biofilm-covercd
sand particles is described in detail in Stevens and Berthouex (1985). The
model equations used to describe that behavior are summarized here.
Particle Size and Density
The biofilm thickness was estimated from measurements of the mass of
volatile solids per unit mass of support medium:
r X I/'
P,
- d
2
Lu/+"
(22)
where 5 = the biofilm thickness (Z-); d = the clean-particle diameter (L); p,
= the density of the clean particle (M/L3); Xd = the dry density of the bio-
film (M/L*)\ and p = the measured mass of volatile solids per unit mass of
clean media (M/M). The quantities d, p„ and (3 were measured by standard
techniques. The value of X,, was found either from the literature (e.g., Shieh
ct al. 1979) or estimated from the wet biofilm density, p„. (M/L1) (Process
Design Manual 1979), and measured moisture content, P, by Xd = pv(l -
P). Once the film thickness was found, the overall particle size, dp = d +
2fi, and the density, p,,, was calculated from
<23)
SU'iuly-Slntc Kliildl/.iillon Model
The model used was that of Fair and Hatch (1933), which balances the
buoyant weight of the particles against the drag force due to flow past the
particle. For a particle of diameter d,\
zj£!±)' (24)
p gp e, ViM,/
where e, = (lie expanded bed poiosity for particle size J,; k = Kozcny's
constant (= 4); U, = the superficial velocity (L/T)\ p. = the viscosity (M/
LT)\ g = the gravitational constant (L?/T)\ and i|/ = the shape factor. The
derivation of Eq. 24 is found in Stevens (1983). This equation was solved
Heratively for £, and the conti ibutions from the various size fractions were
920
summed to give the total expanded bed height, L:
L = L.(l - «.)I,-^- (25)
1 -
where L0 = the bed height corresponding to a reference porosity (L); e„ =
the reference porosity; and/, = the weight fraction of particles with diameter
d/. This model was found to be suitable for superficial velocities up to 1
cm/s (2 ft/min) for sand particle diameters between 0.02 and 0.06 cm.
Dynamic Fluidization Model
The dynamic response of fluidized-bed porosity to a step increase in flow
was described mechanistically by Slis et al. (1959). Their nonlinear model
was linearized by Fan et al. (1963) and integrated to give the fluidized bed
depth as a function of time:
L = L„ exp
(i)+ 4' - (i'
(26)
where L = the bed height at time t(L)\ Lu - the bed height before the step
change; L. = the steady-state bed height after the step change (L); and 0/
= the time constant (T), defined at initial conditions by
0 = —— (27)
' U.MI ~ i„)
Here U„ and = flow velocity (L/T) and expanded porosity before the
step change and n is defined by Fan et al. (1963).
Slis et al. (1959) showed that for step-down tests, the rate of decrease in
the bed height equals the change in the superficial velocity:
L - L„ - (U„ - UJt, 0< t < L" _ L" (28)
L„ - U
L = U, t> — (29)
U „ - U„
Eqs. 26-29 were used in this study to simulate the dynamic fluidization
behavior of the reactor.
Computational Methods
The bioparticle-continuity equation is a parabolic partial differential equa-
tion that is a boundary-value problem in the space dimension, r, and an
initial-value problem in time. The method of orthogonal collocation (Vil-
ladsen and Stewart 1967) was employed to reduce the spatial derivatives to
linear combinations of the concentrations at ne selected locations (collocation
points) in the biofilm using polynomial trial functions developed by Stevens
ct al. (1987) for the spherical-shell geometry. The method is equivalent to
a high-order finite difference scheme using only a small number of discre-
tization points. For the unsteady-state solution, this equation was reduced to
a system of ordinary differential equations in time. For the steady-state case
the model reduces to a system of nonlinear algebraic equations.
921
-------
The orthogonal collocation foi initiation of the model summarized in Fig.
1 is a system of 6h,(1 + nc) simultaneous equations, where nc = the number
Df collocation points used and n, = the number of ideal CSTRs used to
approximate the fluidized bed. These equations arc nonlinear ordinary dif-
ferential equations for the dynamic model. For the steady-state case the equa-
tions are algebraic. For a typical simulation, nc = 2 and n, = 9, so the total
number of simultaneous equations was 9x6x3= 162. A further com-
plication for the unsteady-state case was that the system of equations was
quite stiff (stiffness ratio > 1,000) and ordinary explicit integration routines
were inadequate.
The computation time was reduced by assuming no wastewater recycle
(this was the case for the experiments done to verify the model), and no
backmixing between adjacent CSTRs. This allowed the steady-state model
to be formulated as n, sets of 6(1 + nc) equations, which could be solved
sequentially rather than solving the 6/1,(1 + ne) equations simultaneously.
The unsteady-state equations were integrated using the semi-implicit third-
arder Runge-Kutta method described by Michelson (1976). This method is
very effective for stiff systems in which the total number of simultaneous
equations is moderate. Its step-size adjustment feature makes it economical,
iven when long time periods are simulated. The steady-state equations were
lolvcd using the Newton-Raphson method. Usually three to four iterations
were required for adequate convergence of the material-balance equations
(relative error = 10"").
Description of Pilot Plant
A pilot fluidized-bed nitrification plant was fabricated at the Nine Springs
nctivaled-sludge wastewater-lieaiinciii plant in Madison, Wisconsin, lor studies
lone in I'JKO anil 1 'JK2 The icactor was a 3.66 m tall, clcur I'lcxiglas cyl-
inder, of 10.16 cm inner diameter. The inlet (bottom) of the column was
fitted with a perforated PIcxiglas distribution plate 0.953 cm thick, with 25
•vculy spaced holes (I. I<> em in diameter.
The lluidi/ed-licd medium was silica sand with effective pailicle si/es of
J.48 mm and 0.55 mm and uniformity coefficients of 1.32 and 1.33 for the
1980 and I9R2 experiments, respectively. An unfluidizcd depth of 162 cm
af sand was placed on a 7.5 cm deep base of pea gravel. Influent to the
|)ilot plant was unchlorinatcd secondary effluent from the activated-sludge
>lant. The mlhient (undiluted by leeyele) ammonium concentration tunged
from 6 to 22 mg/L (as N). The How was metered into a mixing tank where
it was blended with recycled fluidized-bed effluent. The mixture was then
>xygcii!i(cd with high-pinity oxygen at pressures of 104-207 kPa (15-30
jsig) in a piopnetary oxygen-danslei device provided by Doir-Oliver Corp.
ind pumped into the fluidized bed This oxygenator was designed to produce
Jissolved-oxygcn (DO) concentrations up to 145 mg/L at a gauge pressure
sf 207 kPa (30 psig) and liquid flow of 9.5 L/min. The oxygen feed rates
varied from 50 to 150 mg/min.
Results
Fig. 3 shows steady-state conditions at a hydraulic flux of 0.37 m3/m3-
nin. The expanded bed depth was 2.96 m. The bed volatile-solids concen-
922
Distance from Reoctor Inlet (cm)
FIG. 3. Steady-State Modeling Results; NH;, NO,", NOj, and pH
FIG. 4. Unatoady-Stale Modeling Result*; 8top Change In Influent NH,
tration was approximately 13,000 mg/L with an estimated average biofilm
thickness of 50 (xm (Eq. 22). Some adjustment of ut 20" C was re-
quired to fit the model to the ammonium profile. The value used was 0.0079
hr~'. No further adjustments were required. The model gave good predic-
tions of pH, ammonium, nitrite, and nitrate.
Fig. 4 shows the measured and predicted results for a step change from
13 to 22 mg/L (as N) in influent ammonium concentration. The parameter
values used were those from the aforementioned literature, except, again,
for p.mu,i at 20° C, which was adjusted to fit the steady-state ammonium
profile (data not shown). The value used was 0.0067 hr"'. The flow rate
was constant at 2 L/min, which gave an empty-bed detention time of 7.8
923
-------
Time (min)
FIG. 5. Unsteady-State Modeling Results; Impulse Forcing of NH« and NOj
lin The influent DO concentration w;is approximately 55 mg/L. The DO
vel in the diluent from the bed dropped Irom 2.5 mg/L to 0.5 mg/L (data
at shown) as the ammonium concentration increased. Other experiments
Stevens 1983; Stevens et al. 1982) showed that complete nitrification could
2 obtained if the DO could be maintained at 1-2 mg/L.
Fig. 5 shows the measured and predicted response to injecting a pulse of
85 mg NH<-N and 185 mg NO2-N into the reactor inlet. The parameter
alues were the same as those used in the steady-state profiles shown in Fig.
. The model overestimated the peak ammonium concentration by about 3
ig/L. One reason for this may be that the model does not include the con-
ibution from suspended biomass, which would increase the NH< conver-
on and reduce the predicted peak concentration. The model provided a
.500
1
280
24U
200
160
• Onto
M".lnl ,.l I ll.ll. I> ( I'l \ \)
2 3 4
I Kiw l,\.U (L/muii)
FIG. 6. Steady-State Modeling Results for Bed Fluldization
924
220
180 ' '
0 20 40 60 80 100 120
Time (seconds)
FIG. 7. Unsteady-State Modeling Results for Bed Fluldization; Step-Up and Step-
Down Responses
better fit to the nitrite response: the result was similar to the fit of the am-
monium data, with an improved match of the peak heights. The nitrate re-
sponse did not fit as well (data not shown): the model predicted a sharper
rise in concentration than was observed.
Fig. 6 shows the model calculations and the results 'for steady-state flui-
dization experiments done in 1980 and 1982, for which the average biofilm
thicknesses were 62 and 68 p.m, respectively. The average prediction errors
for each year were 5 cm and 2 cm. The small errors result in deviations in
hydraulic-residence times in the fluidized bed of less than 2%. Examples of
the data plotted with the model predictions for unsteady-state fluidization are
given in Fig. 7. Both models fit the data well with average errors of less
than I cm.
Experimental results have been published in more detail in Stevens and
Bcrtlioucx (1982, 1985) where a simpler model and compututionul methods
were described only briefly. Here, in this paper, the writers give more at-
tention to the model, nnd only these few experimental results nrc presented.
They are sufficient to indicute that the nitrification process wus cflicicnt und
that the model was adequate to describe the process' dynamic behavior.
Summary and Conclusions
A mechanistic model was developed to describe the behavior of the flu-
idized-bed nitrification process under steady- and unsteady-state conditions.
The model predicts the fluidized bed height, the concentrations of NH<,
NOj", NOJ, Oj, HCOj, H2COj, and the pH as functions of time, position
in the reactor, and position in the biofilm under varying conditions of flow,
temperature, media volume, and influent concentrations of the chemical spe-
cies. The Monod-bascd kinetic model includes the effects of product inhi-
bition and pH on NH« and NOj oxidation inside the biofilm, but does not
predict biofilm thickness or biomass; measurements of these variables were
925
-------
used m tl 10 model Nitnlication Moichiomctry, kinetics, and diffusion of
rcactants and products were modeled using values from the literature.
Experimental data were given to demonstrate the ability of the model to
predict the fluidized hed height and concentrations in the reactor for selected
chemical species. Comparisons of (lie duta with the picdictcd curves show
that the model closely predicted steady-state concentration profiles of NH<,
NOi", NOJ, and pH in the reactor. Only a few parameters were used to fit
thr. model. Predictions of unsteady-state behavior were quite good for step
c iges in influent concentrations, but less quantitative for impulse forcings,
although the general shape and location of the model curve was satisfactory.
The orthogonal collocation method combined with the semi-implicit third-
order Runge-Kutta and Newton-Raphson methods was efficient in solving
the model equations.
The fluidized-bed process gave excellent nitrification performance at de-
tention times that were a small fraction of what would be needed in an ac-
tiva._d-sludge nitrification process. The process was able to accommodate
sudden changes (steps and impulses) in flow rate and ammonium loading.
Acknowledgments
The writers would like to thank the Madison Metropolitan Sanitary District
for their support of this research. We also thank Dorr-Oliver Corporation
for the loan of a countcrcurrcnt oxygen-transfer column. The use of the fa-
cilities of (lie Wisconsin Instructional Timesharing System is gratefully ac-
knowledged. This work was performed while David K. Stevens was a grad-
uate research assistant in Civil and Environmental Engineering at the University
of Wisconsin-Madison,
Appendix I. References
Anthonison, A. C., ct ;il. (1976). "Inhibition of nitrification by unionized ammonia
i unionized nitrous acid." J. Water Poll. Cont. Fed., 48(5), 835-852.
Bird, R. D , Stewart, W. E., and Lightfoot, E. N. (I960). Transport phenomena,
John Wiley and Sons, Inc., New York, N.Y.
Boon, B., and Laudelout, H. (1962). "Kinetics of nitrite oxidation by Nitrobacter
winogradsky." Biochem. J., 85(3), 440-447.
C',ir...tensen, D. R., and McCarty, P. L. (1975). "Multiprocess biological treatment
model." J. Water Poll. Cont. Fed., 47(11), 2652-2664.
Downing, A. L., and Hopwood, A. P. (1964). "Some observations on the kinetics
of nitrifying activated sludge plants." Schwciz. Zeitsch f Hydrol., Zurich, Swit-
zerland, 26(2), 271-277.
Engel, M. S,, and Alexander, M, (1958). "Growth and metabolism of Nitrosomonas
europea." J. Bad., 76(2), 217-222.
Fair, G., and Hatch, L (1933). "Fundamental factors governing the streamline flow
of water through sand " J. Am Water Works Assoc., 25(11), 1551-1565.
Fai., L. T., Schmitz, J. A., and Miller, E. N. (1963). "Dynamics of liquid-solid
fluidized bed expansion." J. Am. Inst. Chem. Engr., 9(2), 149-153.
Gran, P., cl al. (1982). "Recommended notation for use in the description of bio-
logical wastewater treatment processes." Water Res., 16(11), 1501-1505.
Gupta, A., and Thodos, G. (1962). "Mass and heat transfer in the flow of fluids
through fixed and fluidized beds of spherical particles." J. Am. Inst. Chem. Engr.,
8(5), 608-611.
Hen/.e, M., ct al. (1987). "A general model for single-sludge wastewater treatment
systems." Water Res.. 21(5), 505-515.
I lerinanowicz, S. W., and Ganezaiczyk, J. J. (1984). "Dynamics of nitrification in
926
a biological fluidized bed reactor—I. A mathematical model." Water Sci. Teili.,
17(2-3), 351-366.
Huang, C. S., and Hopson, N. E. (1974). "Temperature and pH effects on the bi-
ological nitrification process," presented at the Annual Winter Meeting, New York
Winer Pollution Control Association, Albany, N.Y.
Keenun, J. D., Stcincr, R. L., and Fungoroli, A. A. (1979). "Substrate inhibition
of nitrification." J. Envir. Sci. and Health, A14(5), 377-397.
Knowles, G., Downing, A. L., and Barrett, M. J. (1965). "Determination of kinetic
constants for nitrifying bacteria in mixed culture, with the aid of an electronic
computer." J. General Microbiology, 38(2), 263-278.
Lerman, A. (1979). Ceochemicalprocesses: water and sediment environments. John
Wiley and Sons, Inc., New York, N.Y.
Levenspeil, O. A, (1972). Chemical reaction engineering. John Wiley and Sons,
Inc., New York, N.Y.
Micheison, M. (1976). "An efficient general purpose method for the integration of
stiff ordinary differential equations." J. Am. Inst, of Chem. Engr., 22(3), 594-
597.
Mueller, J. A., Paquin, P., and Famularo, J., (1980). "Nitrification in rotating bi-
ological contactors." J. Water Poll. Cont. Fed., 52(4), 688-710.
Mulcahy, L. T., and LaMotta, E. J. (1978). "Mathematical model of the fluidized
bed biofilm reactor." Report No. Env. E. 59-78-2, Dept. of Civ. Engrg., Univ.
of Massachusetts, Amherst, Mass.
Mulcahy, L. T., Sliich, W. K.. and LaMotta, E. J. (1980). "Simplified mathematical
models of a fluidized bed biofilm reactor," presented at the 73rd Annual Institute
of Chemical Engineering Meeting, Chicago, 111., Nov.
Nutt, S. G. (1980). Pilot scale assessment of the biological fluidized bed process for
municipal wastewater treatment. Report prepared for the Central Mortgage and
Housing Corp. by Dearborn Envlr. Consulting Services, Missi.ssuugu, Oiituno,
Canada.
Petersen, E. (1965). Chemical reaction analysis. Prentice-Hall, Inc., Englewood Cliffs,
N.J.
Process design manual for sludge treatment and disposal. (1979). U.S. linviron-
mental Protection Agency Technology Transfer Series, Washington, D.C.. 429-
435.,
Shieh, W, K,, Sutton, P. M., and Kos, P. (1979), "Oxitron system fluidized bed
wastewater treatment process: predicting reactor biomass concentration," presented
at 52nd Annual Conf., Water Pollution, Control Federation. Houston. Tex., Oct.
Sicgrist, H., and Gujcr, W. (1987). "Demonstration of mass transfer and pH eftects
in a nitrifying biofilm." Water Res., 21(12), 1481-1487.
Slis, P. I., Willemse, T. H., and Kramers, H. (1959). "The response of the level
of a liquid fluidized bed to a sudden change in the fluidizing velocity." Appl. Sci.
Res., 8(A), 209-218.
Srinath, E. G,, Loehr, R. C., and Prakasam, T. B. S. (1976). "Nitrifying organism
concentration and activity." J. Envir. Engrg. Div., ASCE, 102(2), 449-463.
Stankewich, M. J., and Gyger, R. F. (1978). "Nitrification in oxygen-activated sludge
systems." The use of high-purity oxygen in the activated sludge process. Vol. II,
R. R. McWhirter, ed., CRC Press, West Palm Beach, Fla.
Stenstrom, M. K., and Poduska, P. A. (1980). "The effect of dissolved oxygen
concentration on nitrification." Water Res., 14(6), 643-649.
Stephenson, J. P., and Murphy, K. L. (1981). "Kinetics of biological fluidized bed
wastewater denitrification." Wastewater Tech. Ctr., Envir. Protection Service, En-
vir. Canada, Burlington, Ontario, Canada.
Stevens, D. K. (1983). "Performance and modeling of nitrification in the biological
fluidized bed," thesis presented to the University of Wisconsin at Madison, Wise.,
in partial fulfillment of the requirements for the degree of Doctor of Philosophy.
Stevens, D. K., Berthouex, P. M., and Chapman, T. W. (1982). "Dynamics and
simulation of a biological fluidized bed reactor." Proc., 1st Int. Conf. on Fixed
Film Biological Treatment, Kings Island, Ohio. Sponsored by the Univ. of Pitts-
burgh. 3, 1247-1287.
927
-------
Stevens, 13 K., and Bcithoucx, P M (1985) "Fluidization of biomass covered sand
particles." Proc of the Envir. Engrg. Specialty Conf., ASCE, Boston, Mass., 1,
246-253.
Stevens, D. K.. Berthouex, P. M., and Chapman, T. W. (1986). "The effect of
tracer diffusion in biofilm on residence time distributions." Water Re.v.. 20(1)
369-.175.
Stevens, D. K., Berthouex, P. M., and Chapman, T, W. (1987). "Calculation of
effectiveness factors in spherical shells." J. Envir. Engrg., ASCE, 113(5), 1149—
Straiten. I- I: , and McCaity, P. L (1967) "Prediction of nitrification effects on
the dissolved oxygen balance ol streams." Envir, Sci. Tech., 1(5), 405-410.
Tournic, P., Lagueric, C., and Coudrec, J. P. (1979). "Correlations for mass transfer
between nuidi7.ee! spheres and a liquid." Chcm. Engrg. Sci., 34(1), 1247-1255.
Villadsen, J. V., and Stewart, W. E. (1967). "Solution of boundury value problems
by orthogonal collocation." Chcm. Engrg. Sci.. 22(11), 1483-1501.
Wilke, C. R., and Chang, P. (1955). "Correlation of diffusion coefficients in dilute
solutions." J. Am. Insi. Chcm. Engr., 1(2), 264-270.
Williamson, K., and McCarly, 1' I.. (1976). "Verification studies of the biofilm
model lor bacterial substrate utilization." J. Water Poll. Cont. Fed., 48(2), 281 —
296.
Zellren, L., and Hall, P. L. (1973). The study of enzyme mechanisms. John Wiley
and Sons, Inc., New York. N Y
Appf.n->ix I!. Notation
The following symbols me used in this paper:
=
energy yield coefficient;
-
specific surface area in reactor section i, L~K,
h
—
cell decay Kite. T~
i)„
1 ).IMlk()lllci lllllllhcl , (|J.lllll,i„)A'„1)/(
-------
A One-Day Seminar On
Bioremediation Applications
July 27, 1993
Session VII
Case Studies
Presented By
Hugh Russell, Ph.D.
-------
United States
Environmental Protection
Agency
Office of
Research and
Development
Office of Solid Waste
and Emergency
Response
EPA/540/S-92/003
February 1992
&EPA Ground Water Issue
In-Situ Bioremediation of
Contaminated Ground Water
J.L. Sims', J.M. Suflita", and H.H. Russell0
An emerging technology for the remediation of ground water is
the use of microorganisms to degrade contaminants which are
present in aquifer materials. Understanding the processes
which drive in-situ bioremediation, as well as the effectiveness
and efficiency of the utilization of these systems, are issues
which have been identified by the Regional Superfund Ground
Water Forum as concerns of Superfund decision makers. The
Forum is a group of ground-water scientists and engineers,
representing EPA's Regional Superfund Offices, organized to
exchange up-to-date information related to ground-water
remediation at Superfund sites.
Although in-situ bioremediation has been used for a number of
years in the restoration of ground water contaminated by
petroleum hydrocarbons, it has only been in recent years that
this technology has been directed toward other classes of
contaminants. Research has contributed greatly to
understanding the biotic, chemical, and hydrologic parameters
which contribute to or restrict the application of in-situ
bioremediation, and has been successful at a number of
locations in demonstrating its effectiveness at field scale.
This document is one in a series of Ground Water Issue
papers which have been prepared in response to needs
expressed by the Ground Water Forum, h is based on
findings from the research community in concert with
experience gained at sites undergoing remediation. The intent
of the document is to provide an overview of the factors
involved in in-situ bioremediation, outline the types of
information required in the application of such systems, and
point out the advantages and limitations of this technology.
For further information contact Dr. Hugh Russell, RSKERL,
FTS 743-2444, commercial number (405) 332-8800.
Summary
In-situ bioremediation, where applicable, appears to be a
potential cost-effective and environmentally acceptable
remediation technology. Suflita (1989) identified
characteristics of the ideal candidate site for successful
implementation of in-situ bioremediation. These characteristics
included: (1) a homogeneous and permeable aquifer; (2) a
contaminant originating from a single source; (3) a tow
ground-water gradient; (4) no free product; (5) no soil
contamination; and (6) an easily degraded, extracted, or
immobilized contaminant. Obviously, few sites meet these
characteristics. However, development of information
concerning site specific geological and microbiological
characteristics of the aquifer, combined with knowledge
concerning potential chemical, physical, and biochemical fate
of the wastes present, can be used to develop a
bioremediation strategy for a less-than-ideal site.
Introduction
In-situ bioremediation is a technology to restore aquifers
contaminated with organic compounds. Organic contaminants
found in aquifers can be dissolved in water, attached to the
aquifer material, or as freephase or residual phase liquids
referred to as NAPLs which are liquids that do not readily
dissolve in water (Palmer and Johnson, 1989c). Generally,
* Soil Sdentlat, Utah Water Research Laboratory,
Utah State University
b Pro feasor, DepL of Botany and Microbiology, University
of Oklahoma
• Research Microbiologist, Robert S. Kerr Environmental
Research Laboratory
.J-**"'*
Superfund Technology Support Center for
Ground Water
Robert S. Kerr Environmental
Research Laboratory
Ada, Oklahoma
Teotadofl? Iiwawtfiia* Offc# ¦¦¦¦ —i'<
fPhJD. J •••
Printed on Recycled Paper
-------
NAPLs are subdivided into two classes: those that are lighter
than water (LNAPLs density <1.0), and those with a density
greater than water (DNAPLs density >1.0). LNAPLs include
hydrocarbon fuels, such as gasoline, heating oil, kerosene, jet
fuel, and aviation gas. DNAPLs include chlorinated
hydrocarbons, such as 1,1,1-trichloroethene, carbon tetra-
chloride, chlorophenols, chlorobenzenes, tetrachloroethylene,
PCBs, and creosote.
In this discussion, a technical approach Is presented to assess
the potential implementation of bioremediation at a specific
she contaminated with an organic compound. The approach
consists of (1) a site investigation to determine the transport
and fate characteristics of organic waste constituents in a
contaminated aquifer, (2) performance of treatability studies to
determine the potential for bioremediation and to define
required operating and management practices, (3) develop*
ment of a bioremediation plan based on fundamental
engineering principles, and (4) establishment of a monitoring
program to evaluate performance of the remediation effort.
The pattern of contamination from a release of contaminants
into the subsurface environment, as would occur from an
underground leaking storage tank containing NAPLs, is
complex (Figure 1) (Palmer and Johnson, 1989c; Wilson et al.
1989). As contaminants move through the unsaturated zone,
a portion is left behind, trapped by capillary forces. If the
release contains volatile contaminants, a plume of vapors
forms in the soil atmosphere in the vadose zone. If the
release contains NAPLs less dense than water (LNAPLs),
they may flow by gravity down to the water table and spread
laterally. The oily phase can exist either as a free product,
which can be recovered by pumping, or as a residual phase
after the pore spaces have been drained. Contaminants
associated with NAPLs can also partition into the aquifer's
solid phase or in the vapor phase of the unsaturated zone. If
the release contains DNAPLs, these contaminants can
penetrate to the bottom of an aquifer, forming pools in
depressions. In either case, when ground water oomes into
contact with any of these phases, the soluble components are
dissolved into the water phase.
There are a number of techniques available to remediate
ground water contaminated with organic compounds including:
physical containment such as slurry walls, grout curtains, and
sheet pilings (Ehrenfield and Bass, 1984); hydrodynamic
control using pumping wells to manipulate the hydraulic
gradient or interceptor systems (Canter and Knox, 1985);
several methods of free product recovery (Lee and Ward
1986); and (4) extraction of contaminated ground water
followed by treatment at the surface (Keeiy, 1989; U.S. EPA,
1989b).
Alternatively, contaminated ground water can be treated in
place, without extraction using in-situ chemical treatment or
biological treatment with microorganisms (Thomas et al„
1987c). An advantage of in-situ treatment strategies is that
treatment can take place in multiple phases.
In-situ chemical treatment techniques are similar to methods
used to treat contaminated materials after withdrawal or
excavation, but are directly applied to the materials in place
(Ehrenfield and Bass, 1984). Chemical treatment may involve
neutralizing, precipitating, oxidizing or reducing contaminants
Figure 1. Regions of contamination In a typical release from an underground storage tank (Wilaon et aL, 1989).
2
-------
by injecting reactive materials into a contaminated leachate
plume through injection wells, but may be limited by mass
transport and concentration dependence. For treatment of
shallow contaminated aquifers, permeable treatment beds
containing reactive materials such as activated carbon or ion
exchange resins may be constructed downgradient to
intercept and treat the contaminated plume.
Biological in-situ treatment of aquifers is usually accomplished
by stimulation of indigenous microorganisms to degrade
organic waste constituents present at a site (Thomas and
Ward, 1989). The microorganisms are stimulated by injection
of inorganic nutrients and, if required, an appropriate electron
acceptor, into aquifer materials.
Most biological in-situ treatment techniques in use today are
variations of techniques developed by researchers at Suntech
to remediate gasoline-contaminated aquifers. The Suntech
process received a patent titled Reclamation of Hydrocarbon
Contaminated Ground Waters (Raymond, 1974). The process
involves the circulation of oxygen and nutrients through a
contaminated aquifer using injection and production wells (Lee
et al„ 1988). Placement of the wells is dependent on the area
of contamination and the porosity of the formation, but are
usually no more than 100 feet apart. The nutrient amendment
consists of nitrogen, phosphorus, and other inorganic salts, as
required, at concentrations ranging from 0.005 to 0.02 percent
by weight. Oxygen for use as an electron aoceptor in
microbial metabolism is supplied by sparging air into the
ground water. K the growth rate of microorganisms is 0.02 g/l
per day, the process is estimated to require approximately 6
months to achieve 90 percent degradation of the hydro-
carbons present. Cleanup is expected to be most efficient for
ground waters contaminated with less than 40 ppm of
gasoline. After termination of the process, the numbers of
microbial cells are expected to return to background levels.
In addition to stimulating indigenous microbial populations to
degrade organic waste constituents, another technique, which
has not yet been fully demonstrated, is the addition of
microorganisms with specific metabolic capabilities to a
contaminated aquifer (Lee et al., 1988). Populations that are
specialized in degrading specific compounds are selected by
enrichment cutturing or genetic manipulation. Enrichment
culturing involves exposure of microorganisms to increasing
concentrations of a contaminant or mixture of contaminants.
The type of organism (or group of organisms) that is selected
or acclimates to the contaminant depends on the source of the
inoculum, the conditions used for the enrichment, and the
substrate. Examples of changes that may occur during an
acclimation period include an increase in population of
contaminant degraders, a mutation that codes for new
metabolic capabilities, and the induction or derepression of
enzymes responsible for degradation of specific contaminants
(Aelion et al., 1987).
It is important to note that the inoculation of a specialized
microbial population into the environment may not produce the
desired degree of degradation for a number of reasons
(Goldstein et aL, 1985; Lee et al., 1988; Suflita 1988b).
Factors that may limit the success of inoculants include
contaminant concentration, pH, temperature, salinity, and
osmotic or hydrostatic pressure. They may act alone or
collectively to inhibit the survival of the microorganisms. The
subsurface environment may also contain substances or other
organisms that are toxic or inhibitory to the growth and activity
of the inoculated organisms. In addition, adequate mixing to
ensure contact of the organism with the specific organic
constituent may be difficult to achieve at many sites.
Successful inoculation of introduced organisms into simpler,
more controllable environments (e.g., bioreactors such as
waste-water treatment plants) to accomplish degradation has
been demonstrated. However, effectiveness of inoculation into
uncontrolled and poorly accessible environments such as the
subsurface Is much more difficult to achieve, demonstrate and
assess (Thomas and Ward, 1989).
Genetic manipulation of microorganisms to produce
specialized populations to degrade specific contaminants
involves the acceleration and focusing of the process of
evolution (Kilbane, 1986; Lee et al., 1988). Genetic
manipulation can be accomplished by exposure of organisms
to a mutagen, followed by enrichment cutturing to isolate a
population with specialized degradative capabilities, or by the
use of DNA recombinant technology to change the genetic
structure of a microorganism. The use of genetically engi-
neered organisms in the environment is illegal without prior
government approval (Thomas and Ward, 1989). In addition,
the introduction of genetically engineered organisms into the
environment would meet the same kind of barriers to success
as organisms developed by enrichment culturing, or more.
Additional methods that have been suggested to enhance
biodegradation include: cross acclimation, which involves the
addition of a readily degradable substrate to aid in the
biodegradation of more recalcitrant molecules; and analog
enrichment, which involves the addition of a structural analog
of a specific contaminant in order to induce degradative
enzyme activity that will affect both the analog and the specific
contaminant (Suflita, 1989a).
In most contaminated aquifers, the hydrogeologic system is so
complex, in terms of site characteristics and contaminant
behavior, that a successful remediation process must rely on
the use of multiple treatment technologies (Wilson et al.,
1986). The combination of several technologies, in series or
in parallel, into a treatment process train may be necessary to
restore ground-water quality to a required level. Barriers and
hydrodynamic containment controls alone serve as only
temporary plume control measures, but can be integral parts
of withdrawal and treatment or in-situ treatment measures.
A possible treatment train might consist of: (1) source removal
by excavation and disposal; (2) free product recovery to
reduce the mass in order to decrease the amount of
contaminants requiring treatment; and (3) in-situ treatment of
remaining contamination. When applicable, biological in-situ
treatment offers the advantage of partial or complete
destruction of organic contaminants, rather than transfer or
partitioning of contaminants to different phases of the
subsurface.
In-Sltu Bloremedlatlon Technical Process
The in-situ bioremediation technical process consists of the
following activities:
1. performance of a thorough site investigation;
2. performance of treatability studies;
3
-------
3. removal of source of contamination and recovery of
free product;
4. design and implementation of the bioremediation
technology; and
5. evaluation of performance of the technology through
a monitoring program (Lee and Ward, 1986; Lee et
al., 1988).
A thorough site investigation in which biological, contaminant,
and aquifer characterization data are integrated, is essential
for the successful implementation of a bioremediation system.
Biological characterization is required to determine if a viable
population of microorganisms is present which can degrade
the contaminants of concern. An assessment of waste
characteristics provides information for determining whether
bioremediation, either alone or as part of a treatment train, is
feasible for the specific contaminants at the site. Aquifer
characteristics provide information on the suitability of the
specific environment for biodegradative processes, as well as
information required for hydraulic design and operation of the
system.
Bioremediation of an aquifer contaminated with organic
compounds can be accomplished by the biodegradation of
those contaminants and result in the complete mineralization
of constituents to carbon dioxide, water, inorganic salts, and
cell mass, in the case of aerobic metabolism; or to methane,
carbon dioxide, and cell mass, in the case of anaerobic
metabolism. However, in the natural environment, a
constituent may not be completely degraded, but transformed
to an intermediate product or products, which may be equally
or more hazardous than the parent compound. In any event,
the goal of in-situ bioremediation is detoxification of a parent
compound to a product or products that are no longer
hazardous to human health and the environment.
In 1973 a review of ground-water microbiology was published
by researchers at the U.S. EPA Robert S. Kerr Environmental
Research Laboratory (RSKERL) (Dunlap and McNabb, 1973)
that stimulated research into microbiology of the subsurface.
Previously, biological activity in the subsurface environment
below the root zone was considered unlikely and that
microbial activity in the subsurface could not be of significant
importance (Lee et al., 1986). However, as methods for
sampling unconsolidated subsurface soils and aquifer
materials without contamination from surface materials
(Dunlap et al., 1977, Wilson et al., 1983, McNabb and Mallard,
1984) as well as methods to enumerate subsurface microbial
organisms (Ghiorse and Wilson, 1988) were developed,
evidence for microbial activity in the subsurface became
convincing.
Bacteria are the predominant form of microorganisms that
have been found in the subsurface, although a few higher life
forms have been detected (Ghiorse and Wilson, 1988; Suflita,
1989a). The majority of microorganisms in pristine and
uncontaminated aquifers are oligotrophia because organic
materials available for metabolism are likely present in low
concentrations or difficult to degrade. Organic materials that
enter uncontaminated subsurface environments are often
refractory humic substances that resist biodegradation while
moving through the unsaturated soil zone.
Many subsurface microorganisms are metabolically active and
can use a wide range of compounds as carbon and energy
sources, including xenobiotic compounds (Lee et al., 1988).
Compounds such as acetone, ethanol, isopropanol, tert-
butanol, methanol, benzene, chlorinated benzenes,
chlorinated phenols, polycyclic aromatic hydrocarbons, and
alkylbenzenes have been shown to degrade in samples of
subsurface aquifer materials.
The rate and extent of biotransformation of organic
compounds at a specific site are controlled by geochemical
and hydraulic properties of the subsurface (Wilson et al.,
1986). Populations of microorganisms increase until limited by
a metabolic requirement, such as mineral nutrients, substrates
for growth, or suitable electron acceptors. At this point, the
rate of transformation of an organic material is controlled by
transport processes that supply the limiting factor. Since most
subsurface microorganisms are associated with the solid
phase, the limiting factor must be delivered to the microbes by
advection and diffusion through the mobile phases. Below the
water table, all transport must be through liquid phases, and
as a result, aerobic metabolism may be severely limited by the
very low solubility of oxygen in water. As oxygen becomes
limiting, aerobic respiration slows. However, other groups of
organisms become active and continue to degrade contamina-
ting organic materials. Under conditions of anoxia, anaerobic
bacteria can use organic chemicals or several inorganic
anions as alternate electron acceptors (Suflita, 1989a).
Even though microorganisms may be present in a
contaminated subsurface environment and have
demonstrated the potential to degrade contaminants in
laboratory studies, they may not be able to degrade these
contaminants without a long period of acclimation. Acclimation
results in development of the capability to accomplish
degradation.
In summary, the rate of biological activity in the subsurface
environment is generally controlled by:
1. the concentration of required nutrients in the mobile
phases;
2. the advective flow of the mobile phases or the
steepness of concentration gradients within the
phases;
3. opportunity for colonization in the subsurface by
metabolically active organisms or groups of
organisms capable of degradation of the specific
contaminants present;
4. presence, availability, and activity of appropriate
enzymes for degradation of specific contaminants
present; and,
5. toxicity exhibited by the waste or co-occurring
material(s) (Wilson et al„ 1986; Suflita, 1989a).
Methods to Collect Biological Samples
Traditionally, unconsolidated soils or sediments are sampled
through a hollow-stem auger with a split-spoon core barrel or
a conventional thin-walled sample tube (Acker 1974, ScaH et
al., 1981; Wilson et al., 1989). The hollow-stem auger acts as
a temporary casing to keep the borehole open until a sample
can be acquired. A borehole is drilled down to the depth to be
sampled and a core barrel is inserted through the annular
opening in the auger and driven or pushed while rotating the
auger into the earth to collect the sample. These tools are
effective in both unsaturated and saturated cohesive
4
-------
materials, but are not as effective in unconsolidated sands as
it is difficult to keep aquifer material out of the hollow stem
auger (a phenomenon referred to as "heaving") and to keep
the sample in the core barrel while the sample Is being
retrieved to the surface. In recent years there have been
many improvements in sampling the subsurface, particularly
with respect to heaving materials (Zapico et al., 1987; Leach
etal., 1988)
Just as it i9 important to protect the integrity of samples while
coring, it is as important to assure integrity while transferring
sample material to containers which are to be returned to the
laboratory for analysis. To prevent contamination of aquifer
material samples from introduced microorganisms and to
protect samples from the atmosphere to prevent injury of
anaerobic microorganisms, samples are extruded inside a
nitrogen-filled glove box (Figure 2). The glove box is prepared
for sample collection by filling it with the desired number of
sterile sampling jars and sterile paring devices, sealing the
box, and then purging it with nitrogen gas. A slight positive
pressure of nitrogen is maintained in the box by purging during
extrusion and collection of the samples.
Biological Characterization
A wide variety of methods are available to detect, enumerate,
and estimate biomass and metabolic activities of subsurface
microorganisms. These methods include: direct light and
epifiuorescence microscopy, viable counts (e.g., plate counts,
most probable number counts, and enrichment culture
procedures), and biochemical indicators of metabolic activity
such as ATP, GTP, phospholipid, and muramic acid (Ghiorse
and Wilson 1988). Levels of microorganisms ranging from 10*
to 107 cells/g of dry aquifer material have been reported from
uncontaminated shallow aquifers (Ghiorse and Balkwill, 1985;
Lee et al., 1988). Often the distribution of microorganisms in
aquifers, as it is in soils, is sporadic and nonuniform, indicating
the presence of micro-environments conducive to growth and
activity.
Waste Characterization
The source of contamination is usually the primary object of
remedial activities (Wilson et al. 1989) as the treatment of
plume areas will not be effective if the source continues to
release contaminants. Information concerning: (1) the areal
location of the source area and contaminant plumes; (2)
amounts of contaminants in the source area; and (3) amounts
of contaminants released into the subsurface are required to
select and apply an appropriate remediation technology and to
determine cost and time requirements for completion of a
remedial action. If in-situ bioremediation is selected as the
remedial technology, information concerning the amount and
distribution of contamination is used in conjunction with
hydrogeological site characteristics to locate injection and
extraction wells and to optimize pumping rates and
concentrations of amendments, such as nutrients and
alternate electron acceptors.
The use of conventional monitoring wells can generally
accurately define the geometry of the ground-water plume
(Palmer and Johnson, 1989a; Wilson et al., 1989). However,
there are important factors that control the quality of
information collected from a network of monitoring wells,
which include the amount of well purging done prior to
sampling (Barcelona and Helfrich, 1986), method of sampling
(Stolzenburg and Nichols, 1985), and method of well
construction and installation (Keely and Boateng, 1987).
Methods for ground-water sampling are presented by ScaK et
al. (1981), Ford et al. (1984), and Barcelona et al. (1985).
Other methods used for detecting contaminant plumes in the
subsurface include geophysical techniques such as surface
resistivity and electromagnetic surveys, chemical time-series
sampling tests (Palmer and Johnson, 1989a), and vapor
Sample Head
Space Analysis
Vent
Flushing Vent
^*0
w
Flow Regulator
and Indicator
Sample Tube
from Extruder
Iris Port
Figure 2. FMd sampling glove box (Wilson, et sL, 1989).
5
-------
monitoring wells (Devitt et al., 1988; Palmer and Johnson,
1989b).
The distribution of the source area and the extent of
contamination should also be characterized by collecting
cores of the solid aquifer materials. Precise information is
required to define the vertical extent of contamination so that
nutrients, oxygen and other amendments injected into the
aquifer contact the contaminants. Injection into a clean part of
the aquifer is a wasted effort and may give the false
impression that the region of aquifer between the injection and
recovery wells is clean (Figure 3).
Additional characteristics of waste contaminants present at a
specific site that should be considered are related to their
environmental fate and behavior in specific aquifer materials
(Armstrong, 1987; Johnson et al., 1989). These character-,
istics include physical and chemical properties that determine
recalcitrance, reactivity, and mobility of contaminants at the
site. Information concerning partitioning of contaminants
between aquifer solids and water is especially important. This
information is used to evaluate the extent and rate of release
of contaminants into the ground water, their mobility, and the
quantity of electron acceptors and inorganic nutrients that
must be supplied to support in-situ bioremediation.
Aquifer Characterization
Important geological characteristics of an aquifer that should
be considered during a site investigation include the
composition and heterogeneity oHtquifer material, specific
yield, hydraulic connections to other aquifers, magnitude of
water table fluctuations, ground-water flow rate and direction,
hydraulic conductivity distribution, permeability, bulk density,
and porosity (Lee et al., 1988; PaJmer and Johnson, 1989a).
Hydraulic conductivity (K) is an especially important
characteristic since the aquifer must be permeable enough to
allow the transport of electron acceptors and inorganic
nutrients to the microorganisms in the zone of contamination.
Permeable aquifer systems, i.e., aquifers with K values of 10"*
cm/sec or greater, are usually considered good candidates for
in-situ bioremediation (Thomas and Ward, 1989).
Hydraulic conductivity of an aquifer can be determined by a
variety of methods (Thomas et aL, 1987b, Palmer and
Johnson, 1989a). Knowledge of K values at multiple locations
is necessary because of the heterogeneity of aquifer
materials. Laboratory methods are also available for
determining hydraulic conductivity, but field-measured values
represent average properties over a larger volume and utilize
less disturbed materials (Palmer and Johnson, 1989a).
Aquifer characteristics play an extremely important role in
determining the effectiveness of in-situ bioremediation. Even
in the presence of organisms acclimated to the specific waste
constituents present in an aquifer, biodegradation of
contaminants may be limited by unfavorable aquifer
characteristics that affect microbial activity including:
1. insufficient concentrations of dissolved oxygen for
aerobic metabolism of compounds susceptible to
aerobic degradation;
2. excessive oxygen that inhibits anaerobic
biodegradation of many halogenated compounds in
the subsurface;
Injection
Well
Land Surface
Water
Table
¦
Contaminated
Interval
Direction of Flow czj)> (z£>
c=[> Wasted c=J>
Figure 3. The value of accurately locating the contaminated Interval (WUson M aL, 1889).
6
-------
3. lack of a suitable alternative electron acceptor, if
oxygen is unavailable or not usable;
4. insufficient inorganic nutrients, such as nitrogen,
phosphorus, and trace minerals;
5. presence of toxic metals or other toxicants; and
6. other aquifer characteristics, such as pH, buffering
capacity, salinity, osmotic or hydrostatic pressures,
radiation, sorptive capacity, and temperature
(Armstrong, 1987; Lee et al., 1988).
Treatability Study
A treatability study is designed to determine if bioremediation
is possible at a specific site, and whether there are any
biological barriers to attaining clean-up goals. Even though
the scientific literature may indicate that a specific chemical is
likely to biodegrade in the environment, a treatability study .
using site specific variables should be used to confirm that
contention (Suflita, 1989a). Microcosms are generally used to
conduct treatability studies. Pritchard (1981) defined a
microcosm as "a calibrated laboratory simulation of a portion
of a natural environment in which environmental components,
in as undisturbed a condition as possible, are enclosed within
definable physical and chemical boundaries and studied under
a set of laboratory conditions." Microcosms may range from
simple batch incubation systems to large and complex flow-
through devices (Suflita, 1989a).
Results of a treatability study can also provide an estimate of
the rate and extent of remediation that can be attained if
microorganisms are not limited by the rate of supply of an
essential growth factor or by the presence of an unfavorable
environmental factor.
Treatability studies to determine inorganic nutrient and
electron acceptor requirements of subsurface microorganisms
present at a specific site should be conducted using samples
of subsurface solids as well as the ground water. Nutrient and
electron acceptor requirements that will enable indigenous
microorganisms to efficiently degrade organic contaminants
present at a specific site can be determined by incubating
contaminated subsurface materials with combinations of levels
of inorganic nutrients and electron acceptors. Studies should
be performed under conditions that simulate field
environmental conditions. Results of the studies are used to
design the bioremediation program as well as to optimize the
treatment strategy.
Design and Implementation of an In-SItu
Bioremediation System
Before implementation of an in-sJtu bioremediation system, the
source of contamination in the soil and in the ground water
should be removed as much as possible. In the case of a
liquid fuel spill, source removal may consist of recovery of
LNAPL free product from the ground water. Depending on the
characteristics of the aquifer and contaminants, free product
can aooount for as much as 91 percent of the spilled
hydrocarbon, with the remaining hydrocarbon (accounting for
9-40 percent of the spill) sorbed to the soil or dissolved in the
ground water (Lee et a!., 1986).
Physical recovery techniques, based on the fact that LNAPL
hydrocarbons are relatively insoluble in and less dense than
water, are used to remove free product from a contaminated
site. Physical recovery often accounts for only 30 to 60
percent of spilled hydrocarbon before yields decline.
Continued pumping of recovery wells may be used to contain
a spill while in-situ bioremediation is being implemented. K a
spill is comprised of DNAPLs, which may sink to the bottom of
the aquifer, physical recovery may be considerably more
difficult to achieve.
Information from the performance of she characterization and
treatability studies may be integrated with the use of
comprehensive mathematical modeling to estimate the
expected rates and extent of treatment at the field scale
(Javandel, 1984; Keely, 1987). The specific model chosen
should incorporate biological reaction rates, stoichiometry of
waste transformation, mass-transport considerations, and
spatial variability in treatment efficiency (U.S. EPA, 1989a).
After assessment of site characterization and treatability
studies, if results indicate that in-situ bioremediation is
applicable to the site and will be an effective clean-up
technology, the information collected is used to design and
implement the system.
When in-situ bioremediation of a contaminant ground-water
plume involves using methods to enhance the process, such
as the addition of nutrients, additional oxygen sources, or
other electron acceptors, the use of hydraulic controls to
minimize migration of the plume during the in-situ treatment
process may be required (Thomas et al., 1987c; U.S. EPA,
1989a). In general, hydraulic control systems are generally
less costly and time consuming to install than physical
containment structures such as slurry walls. Well systems are
also more flexible, for pumping rates and well locations can be
altered as the system is operated over a period of time.
Pumping-injection systems can be used to: (1) create
stagnation zones at precise locations in a flow field; (2) create
gradient barriers to pollution migration; (3) control the
trajectory of a contaminant plume; and (4) intercept the
trajectoiy of a contaminant plume (Shafer, 1984). The choice
of a hydraulic control method depends on geological
characteristics, variability of aquifer hydraulic conductivities,
background velocities, and sustainable pumping rates (Lee et
al. 1988). Typical patterns of wells that are used to provide
hydraulic controls include: (1) a pair of injection-production
wells; (2) a line of downgradient pumping wells; (3) a pattern
of injection-production wells around the boundary of a plume;
and (4) the "double-cell' hydraulic containment siystem. The
"double-cell" system utilizes an inner cell and an outer
recirculation cell, with four cells along a line bisecting the
plume in the direction of flow (Wilson, 1984).
Well systems also serve as injection points for addition of the
materials used for enhancement of microbial activity and for
control of circulation through the contaminated zone. The
system usually includes injection and production wells and
equipment for the addition and mixing of the nutrients (Lee et
al., 1988). A typical system in which microbial nutrients are
mixed with ground water and circulated through the
contaminated portion of the aquifer through a series of
injection and recovery wells is illustrated in Figure 4 (Raymond
et al., 1978; Thomas and Ward, 1989).
Materials can also be introduced to the aquifer through the
use of infiltration galleries (Figure 5) (Brenoel and Brown,
7
-------
To Sewer or
Recirculate
IIK-T^r
^nmnraeenr > ¦ ¦ C
-tx-
Alr Compressor
Nutrient
Addition
Tank
Coarse Sand
¦ Production Well
Water Table
7*"
1
Spilled Materials4
Clay
Water Supply
Injection Well
D-"
Sparger
Figure 4. Typical schematic for aerobic subsurface btoramedlatlon (Thomas and Wsrd, 1989).
Air Compressor or Nutrient Addition
Figure 6. Use of infiltration gallery for recirculation of water and nutrients In in-situ biors mediation (Thomas and Ward, 1989).
8
-------
1985; Thomas and Ward, 1989). Infiltration galleries allow
movement of the injection solution through the unsaturated
zone as well as the saturated zone, resulting in potential
treatment of source materials that may be trapped in the pore
spaces of the unsaturated zone.
Amendments to the aquifer are added to the contaminated
aquifer in alternating pulses. Inorganic nutrients are usually
added first through the injection system, followed by the
oxygen source. Simultaneous addition of the two may result in
excessive microbial growth close to the point of injection and
consequent plugging of the aquifer. High concentrations of
hydrogen peroxide (greater than 10%) can be used to remove
biofouling and restore the efficiency of the system.
Operations Monitoring
Both the operation and effectiveness of the system should be
monitored (Lee et al., 1988). Operational factors of impor-
tance include the delivery of inorganic nutrients and electron
acceptor, the point of the delivery within the aquifer in relation
to the contaminated portion of the plume, and the effective-
ness of containment and control of the contaminated plume.
Measurements of dissolved oxygen and nutrient levels in
ground-water samples are recommended to assess whether
or not bioremediation is being accomplished. Increases in
microbial activities in samples of aquifer materials may be
quantified relative to plume areas prior to treatment, areas
within the plume that did not receive treatment, and control
areas outside the plume. Carbon dioxide levels in ground-
water samples may also be a useful indicator of microbial
activity (Suflita, 1989b).
Measurement of contaminant levels should indicate that
concentrations of contaminants are decreasing in areas
receiving treatment and remaining relatively unchanged in
areas that are not. If degradation pathways of specific
contaminants are known, measurement of presence and
concentrations of metabolic products may be used to
determine whether or not bioremediation is occurring. Both
soil and ground-water samples should be collected and
analyzed to develop a thorough evaluation of treatment
effectiveness. The use of appropriate control samples, e.g.,
assays of untreated areas or areas outside the plume, is
highly recommended to confirm the effectiveness of the
bioremediation technology (Suflita, 1989b).
The frequency of sampling should be related to the time
expected for significant changes to occur along the most
contaminated flow path (U.S. EPA, 1989a). Important
considerations include time required for water to move from
injection wells to monitoring wells, seasonal variations in water
table elevation or hydraulic gradient changes in the
concentration of dissolved oxygen or alternative electron
acceptor, and costs of monitoring.
Advantages and Limitations In the Use of tn-Sltu
Bioremediation
There are a number of advantages and disadvantages in the
use of in-situ bioremediation (Lee et a!., 1988). Unlike other
aquifer remediation technologies, it can often be used to treat
contaminants that are sorbed to aquifer materials or trapped in
pore spaces. In addition to treatment of the saturated zone,
organic contaminants held in the unsaturated and capillary
zones can be treated when an infiltration gallery is used.
The time required to treat subsurface pollution using in-situ
bioremediation can often be faster than withdrawal and
treatment processes. A gasoline spill was remediated in 18
months using in-situ bioremediation, while pump-and-treat
techniques were estimated to require 100 years to reduce the
concentrations of gasoline to potable water levels (Raymond
et al., 1986). In-situ bioremediation often costs less than other
remedial options. The areal zone of treatment using
bioremediation can be larger than with other remedial
technologies because the treatment moves with the plume
and can reach areas that would otherwise be inaccessible.
There are also disadvantages to in-situ bioremediation
programs (Lee et al„ 1988). Many organic compounds in the
subsurface are resistant to degradation. In-situ
bioremediation usually requires an acclimated population of
microorganisms which may not develop for recent spills or for
recalcitrant compounds. Heavy metals and toxic
concentrations of organic compounds may inhibit activity of
indigenous microorganisms. Injection wells may become
clogged from profuse microbial growth resulting from the
addition of nutrients and oxygen.
In-situ bioremediation is difficult to implement in low-
permeability aquifers that do not permit the transport of
adequate supplies of nutrients or oxygen to active microbial
populations. In addition, bioremediation projects require
continuous monitoring and maintenance for successful
treatment
References
Aelion, C. M., C. M. Swindoll, and F. K. Pfaender. 1987.
Adaptation to and Biodegradation of Xenobiotic Compounds
by Microbial Communities from a Pristine Aquifer. Appl.
Environ. Microbiol. 532212-2217.
Acker, W. L, III. 1974. Basic Procedures for Soil Sampling and
Core Drilling. Acker Drill Co. Scranton, PA.
Armstrong, J. 1987. Some Problems in the Engineering of
Groundwater Cleanup, p. 110-120. In: N. Dee, W. F.
McTernan, and E. Kaplan (eds.) Detection, Control, and
Renovation of Contaminated Ground Water. Am. Soc. Civil
Eng., New York, NY.
Barcelona, M J. and J.A. Helfrich. 1986. Well Construction and
Purging Effects on Ground-Water Samples. Environ. Sci.
Technol. 20:1179-1184.
Barcelona, MJ., J.P. Gibb. J.A. Helfrich, and E.E. Garske.
1985. Practical Guide for Ground-Water Sampling. EPA/600/
2-85/104, Robert S. Kerr Environmental Research Laboratory,
U.S. Environmental Protection Agency, Ada, OK.
Brenoel, M., and RA Brown. 1985. Remediation of a Leaking
Underground Storage Tank with Enhanced Bioreclamation.
Proc., Fifth National Symp. on Aquifer Restoration and Ground
Water Monitoring. National Water Well Assoc., Worthington,
OH.
9
-------
Canter, LW., and R.C. Knox. 1985. Ground Water Pollution
Control. Lewis Publishers, Chelsea, Ml.
Devitt, D.A., R.B. Evans, W.A. Jury, T.H. Starks, B. Eklund, A.
Ghalsan. 1988. Soil Gas Sensing for Detection and Mapping
of Volatile Organics. EPA/600/8-87/036, Environmental
Monitoring and Support Laboratory, U.S. Environmental
Protection Agency, Las Vegas, NV.
Dunlap, WJ. and J.F. McNabb. 1973. Subsurface Biological
Activity in Relation to Ground Water Pollution. EPA-660/2-73-
014, Robert S. Kerr Environmental Research Laboratory, U.S.
Environmental Protection Agency, Ada, OK.
Dunlap, WJ., J.F. McNabb, M.R. ScaH, and R.L. Cosby. 1977.
Sampling for Organic Chemicals and Microorganisms in the
Subsurface. EPA-600/2-77-176, Robert S. Kerr Environmental
Research Laboratory, U.S. Environmental Protection Agency,
Ada, OK.
Ehrenfield, J. and J. Bass. 1984. Evaluation of Remedial
Action Unit Operations of Hazardous Waste Disposal Sites.
Pollution Technology Review No. 110, Noyes Publications,
Park Ridge, NJ.
Ford, PJ„ P.J. Turina, and D.E. Seely. 1984. Characterization
of Hazardous Waste Sites • A Methods Manual, Volume II:
Available Sampling Methods. EPA/600/4-84/076. Robert S.
Kerr Environmental Research Laboratory, U.S. Environmental
Protection Agency, Ada, OK.
4
Ghiorse, W.C., and D.L Balkwill. 1985. Microbiology of
Ground Water Environments. In: G. E. Januer (ed.) Progress
in Chemical Disinfection II: Problems at the Frontier. State
Univ. of New York (SUNY) at Binghamton, Binghamton, NY.
Ghiorse, W.C., and J.T. Wilson. 1988. Microbial Ecology of the
Terrestrial Subsurface. Adv. AppL MicrobioL 33:107-172.
Goldstein, R.M., LM. Mallory, and M. Alexander. 1985.
Reasons for Possible Failure of Inoculation to Enhance
Biodegradation. Appl. Environ. Microbiol. 50:977-983.
Javandel, I., C. Doughty, and C.F. Tsang. 1984. Groundwater
Transport: Handbook of Mathematical Models. Water
Resources Monograph No. 10, Am. Geophysical Union,
Washington, DC.
Johnson, R.L, C.D. Palmer, and W. Fish. 1989. Subsurface
Chemical Processes. In: Transport and Fate of Contaminants
in the Subsurface. EPA/625/4-89/019, Robert S. Kerr
Environmental Research Laboratory, U.S. Environmental
Protection Agency, Ada, OK.
Keely, J.F. 1987. The Use of Models in Managing Ground-
Water Protection Programs. EPA/600/8-87/003. Robert S.
Kerr Environmental Research Laboratory, U.S. Environmental
Protection Agency, Ada, OK.
Keely, J.F. and K. Boateng. 1987. Monitoring Well Installation,
Purging, and Sampling Techniques. Part II: Case Histories.
Ground Water 25:427-439.
Kilbane, J J. 1986. Genetic Aspects of Toxic Chemical
Degradation. Microbiol. Ecol. 12:135-145.
Leach, LE., F.P. Beck, J.T. Wilson, and D.H. Kampbell. 1988.
Aseptic Subsurface Sampling Techniques for Hollow-Stem
Auger Drilling. Proc., Second National Outdoor Action
Conference on Aquifer Restoration, Ground Water Monitoring
and Geophysical Methods 131-51.
Lee, M.D. and C.H. Ward. 1986. Ground Water Restoration.
Report submitted to JACA Corporation, Fort Washington, PA.
Lee, M.D., J.M. Thomas, R.C. Borden, P.B. Bedient, J.T.
Wilson, and C. H. Ward. 1988. Biorestoration of Aquifers
Contaminated with Organic Compounds. CRC Critical Rev.
Environ. Controls 1829-89.
McNabb, J.F., and G.E. Mallard. 1984. Microbiological
Sampling in the Assessment of Groundwater Pollution. In: G.
Bitton and C.P. Gerba (eds.), Groundwater Pollution
Microbiology. John Wiley & Sons, New York, NY.
Palmer, C.D., and R.L Johnson. 1989a. Determination of
Physical Transport Parameters. In: Transport and Fate of
Contaminants in the Subsurface. EPA/625/4-89/019, Robert
S. Kerr Environmental Research Laboratory, U.S.
Environmental Protection Agency, Ada, OIC
Palmer, C.D., and R.L Johnson. 1989b. Physical Processes
Controlling the Transport of Contaminants in the Aqueous
Phase. In: Transport and Fate of Contaminants in the
Subsurface. EPA/625/4-89/019, Robert S. Kerr Environmental
Research Laboratory, U.S. Environmental Protection Agency,
Ada, OK.
Palmer, C.D., and R.L Johnson. 1989c. Physical Processes
Controlling the Transport of Non-Aqueous Phase Liquids in
the Subsurface. In: Transport and Fate of Contaminants in the
Subsurface. EPA/625/4-89/019, Robert S. Kerr Environmental
Research Laboratory, U.S. Environmental Protection Agency,
Ada. OK.
Pritchard, P.H. 1981. Model Ecosystems. In: R.A. Conway
(ed.), Environmental Risk Analysis for Chemicals. Van
Nostrand Reinhold, New York, NY.
Raymond, R.L 1974. Reclamation of Hydrocarbon
Contaminated Ground Waters. U.S. Patent Office,
Washington, DC. Patent No. 3,846,290. Patented
November 5,1974.
Raymond, R.L, RA Brown, R.D. Norris, and E.T. O'Neill.
1986. Stimulation of Biooxidation Processes in Subterranean
Formations. U.S. Patent Office, Washington, DC. Patent No.
4,588,506. Patented May 13,1986.
Raymond, R.L, V.W. Jamison, J.O. Hudson, R.E. Mitchell,
and V.E. Farmer. 1978. Field Application of Subsurface
Biodegradation of Gasoline in a Sand Formation. API Publi-
cation 4430, American Petroleum Institute, Washington, DC.
Scalf, M.R., J.F. McNabb, WJ. Dunlap, R.L. Cosby, and J.
Fryberger. 1981. Manual of Ground Water Sampling
Procedures. National Water Well Assoc., Worthington, OH.
Schafer, J.M. 1984. Determining Optimum Pumping Rates for
Creation of Hydraulic Barriers to Ground Water Pollutant
Migration. In: D.M. Nielsen (ed.), Proc., Fourth National Symp.
10
-------
on Aquifer Restoration and Ground Water Monitoring. National
Water Well Assoc., Worthington, OH.
Stolzenburg, T.R. and D.G. Nichols. 1985. Preliminary Results
on Chemical Changes in Groundwater Samples Due to
Sampling Devices. EPRI Report No. EA-4118, Electric Power
Research Institute, Palo Alto. CA.
Suflita, J.M. 1989a. Microbial Ecology and Pollutant
Biodegradation in Subsurface Ecosystems. In: Transport and
Fate of Contaminants in the Subsurface. EPA/625/4-89/019,
Robert S. Kerr Environmental Research Laboratory, U.S.
Environmental Protection Agency, Ada, OK.
Suflita, J.M. 1988b. Microbiological Principles Influencing the
Biorestoration of Aquifers. In: Transport and Fate of
Contaminants in the Subsurface. EPA/625/4-89/019, Robert
S. Kerr Environmental Research Laboratory, U.S.
Environmental Protection Agency, Ada, OK.
Thomas, J.M., and C.H. Ward. 1989. In Situ Biorestoration of
Organic Contaminants in the Subsurface. Environ. ScL
Technol. 23:760-766.
Thomas, J.M., M.D. Lee, and C.H. Ward. 1987a. Use of
Ground Water in Assessment of Biodegradation Potential in
the Subsurface. Environ. Toxicol. Chem. 6:607-61 4.
Thomas, J.M., H J. Marlow, R.L Raymond, and C.H. Ward.
1987b. Hydrologic Considerations for In Situ Biorestoration.
US/USSR Symposium on Fate of Pesticides and Chemicals in
the Environment, Oct 12-16, bwa City, 1A (under review by
John Wiley & Sons, Inc., New York, NY).
Thomas, J.M., M.D. Lee, P.B. Bedient, R.C. Borden, LW.
Canter, and C.H. Ward. 1987a Leaking Underground Storage
TanHs: Remediation with Emphasis on In Shu Biorestoration.
EPA/600/2-87/008, Robert S. Kerr Environmental Research
Laboratory, U.S. Environmental Protection Agency, Ada, OK.
U.S. Environmental Protection Agency. 1989a. Bioremediation
of Hazardous Waste Sites Workshop: Speaker Slide Copies
and Supporting Information. CERI-89-11, Center for
Environmental Research Information, U.S. Environmental
Protection Agency, Cincinnati, OH.
Wilson, J.L 1984. Double-cell Hydraulic Containment of
Pollutant Plumes. In: D. M. Nielsen (ed.), Proc., Fourth
National Symp. on Aquifer Restoration and Ground Water
Monitoring. National Water Well Assoc., Worthington, OH.
Wilson, J.T., and B.H. Wilson. 1985. Biotransformation of
Trichloroethylene in Soil. Appl. Environ. Microbiol. 49:242-243.
Wilson, J.T. and C.H. Ward. 1987. Opportunities for
B'oreclamation of Aquifers Contaminated with Petroleum
Hydrocarbons. Dev. Industrial Microbiol. (J. Industrial
Microbiol Suppl. 1 ) 27:109-116.
Environmental Research Laboratory, U.S. Environmental
Protection Agency, Ada, OK.
Wilson, J.T., J.F. McNabb, D.L Balkwill, and W.C. Ghiorse.
1983. Enumeration and Characterization of Bacteria
Indigenous to a Shallow Water-Table Aquifer. Ground Water
21:134-142.
Wilson, J.T., LE. Leach, MJ. Hanson, and J.N. Jones. 1986.
In Situ Biorestoration as a Ground Water Remediation
Technique. Ground Water Monitoring Rev. 656-64.
Zapico, M.M., S. Vales, and J.A. Cherry. 1987. A Wireline
Piston Core Barrel for Sampling Cohesionless Sand and
Gravel Below the Water Table. Ground Water Monitoring Rev.
7(3):74-82.
Wilson J., L Leach, J. Michalowski, S. Vandegrift, and R.
Callaway. 1989. In Situ Bioremediation of Spills from
Underground Storage Tanks: New Approaches for Site
Characterization, Project Design, and Evaluation of
Performance. EPA/600/2-89/042, Robert S. Kerr
11
GOVERNMENT PRINTING OFFICE: ml . M*-OCO/4©TW
-------
United States
Environmental Protection
Agency
Office of
Research and
Development
Office of Solid Waste
and Emergency
Response
EPA/540/S-93./501
May 1993
EPA Engineering Issue
In Situ Bioremediation of Contaminated Unsaturated
Subsurface Soils
J.L. Sims*, R.C. Sims*, R.R. Dupont*, J.E. Matthews** and H.H. Russell**
Introduction
An emerging technology for the remediation of unsaturated
subsurface soils involves the use of microorganisms to
degrade contaminants which are present in such soils.
Understanding the processes which drive in situ
bioremediation, as well as the effectiveness and efficiency of
the utilization of these systems, are issues which have been
identified by the Regional Superfund Engineering Forum as
concerns of Superfund decision makers.
The Regional Superfund Engineering Forum is a group of
EPA professionals, representing EPA's Regional Superfund
Offices, committed to the identification and resolution of
engineering issues impacting the remediation of Superfund
sites. The Forum is supported by and advises the Superfund
Technical Support Project.
Although in situ bioremediation has been used for a number
of years in the restoration of ground water contaminated by
petroleum hydrocarbons, h has only been in recent years that
in situ systems have been directed toward contaminants in
unsaturated subsurface soils. Research has contributed
greatly to understanding the biotic, chemical, and hydrologic
parameters which contribute to or restrict the application of in
situ bioremediation and has been successful at a number of
locations in demonstrating its effectiveness at field scale.
This document is one in a series of engineering issue papers
which have been prepared in response to needs expressed
by the Engineering Forum. It is based on findings from the
research community in concert with experience gained at
sites undergoing remediation. The intent of the document is
to provide an overview of the factors involved in in situ
bioremediation, outline the types of information required in
the application of such systems, and point out the
advantages and limitations of this technology.
For further information, contact John Matthews (405) 436-
8600 or Dr. Hugh Russell (405) 436-8612.
Background
Bioremediation of contaminated surface soils using in situ
systems, prepared bed, and above-ground bioreactors, has
been previously addressed with regard to characterization,
environmental processes and variables, and field-scale
applications (Sims et al.,1989). This paper will address
processes which are currently being utilized or are in
development to treat contaminated unsaturated subsurface
soils in place.
In situ biological remediation of subsurface soils
contaminated with organic chemicals is an alternative
treatment technology that, in certain cases, can meet the
goal of achieving a permanent cleanup at hazardous waste
sites. Use of such alternatives is encouraged by the U.S.
Environmental Protection Agency (U.S. EPA) for
implementing the requirements of the Superfund
Amendments and Reauthorization Act (SARA) of 1986.
Bioremediation of subsurface soils is consistent with the
philosophical thrust of SARA, for it involves use of naturally
occurring microorganisms to degrade and/or detoxify
hazardous constituents in the soil to protect public health and
the environment. Use of in situ subsurface bioremediation
Utah State University
" Robert S. Kerr Environmental Research Laboratory,
U.S. EPA
Technology Innovation Oftle®
Office of SoBd Waste Emergency
Response, US EPA, Washington, O.C.::
Walter W. Kovaltck, PhD.
- Director:" 'J.'•
Superfund Technology Support Center for
Ground Water
Robert S. Kerr Environmental
Research Laboratory
Ada, Oklahoma
-------
JUSTIFICATION: Information in this report is essential in the use of mathematical
models for remediation decision-mking at Superfund sites.
1. CONTACT (program name, mail code. name. & phone number)
Marion R. Scalf RSKERL-Ada 405-41fi-RSftf)
2. OFFICE CONTROL NO.
222488
3. DATE
May 18, 1993
4. FORM, PUBLICATION, OR ISSUANCE NO., AND TITLE OR DESCRIPTION
Engineering Issue Paper: In Situ Bioremediation of Contaminated Unsaturated
Subsurface Soils
5. QUANTITY (Unit* of finished product)
4,000
6. IS OVERTIME AUTHORIZED TO l—l I 1
MEET DESIRED DELIVERY DATE? | | Yet I X I
NO
7. DESIRED DELIVERY DATE
COMPOSITION AND PROOF
8. NUMBER OF PAGES SUBMITTED
MANUSCRIPT
9. TYPE
10. FACE
11. SIZE
12. ACTUAL IMAGE SIZE
13. PROOF
13a. GALLEY
CL ~
No Sett Hold Days
No
13b. PAGE
13c. SEND PROOF TO
No. Sets Hold Days
PRESSWORK AND BINDERY
14. NUMBER OF PIECES SUBMITTED
b. CAMERA COPY
d. ILLUSTRATIONS
15. SIZE (Inches)
Trimmed Page
b. NEGATIVES
(1) HALFTONE
c. OVERLAYS
(2) UNECUT
(3) OTHER
16. RUN (Check one)
U Head to Left
O One Side ID Heed to Right
O Hoed to Heed
O Head to Foot
17. Form* Mutt
Register
_D_
18. TEXT PAPER (Grade. Color, and Weight) 19. COLOR INK
20. Margins After
Trim
(pica* or inches)
Right
Top
Bottom
21 COVER STOCK
O Self Q
22. COLOR INK
Separate (Specify)
23. PERFORATE/SCORE Parallel to Top/Left.
OTHER:
. in. from Top/Left
24. PUNCH
~ Top/Left
B3-Ring Binder
Acco Fastener
Q Other (Specify)
26. ADDRESSING AND MAILING
26. O Side Stitch O Corner Stitch O Sew ED Assemble Only
0 Saddle Stitch O Pastefold O Perfect Bind O Band in Sets
Mailing Keys
~
Bulk Mailing
Addressing Only
27. Use separate sheet if needed for additional specifications or remarks
b. Quantity (Copies) c. OTHER MAILING (Attach labels or listing)
28. FUNDS ARE AVAILABLE (Commitment Clerk)
31. APPROPRIATION NUMBER
29. ALLOTMENT NO.
30. RETURN NEGATIVES, PLATES, COPY TO:
32.DCNNO.
33. RESTRICTIONS ON QUANTITY (Check one only)
O Internal Use Only O Reprint O External Distribution
_OjWrrtt«n_«££rovah«_on^«JrorTUhe_holder_oJ_an^j»[j£ight«hTiat»n»|_re2u^rtK>ned_
34. DELIVER TO:
35. APPROVED BY
v~-jPrQjecrt Officer
a. QUANTITY
ALL
b. AGENCY/DIVISION ROOM
U.S.EPA Publication-
BLDG.
TIM
Distribution Center
DDD, Bldg. 5, Dock 63
11U27 Kenwood Road
Cincinnati, OH 45242
EAEB
36.1 concur in the publication of the attached material and certify that it complies with Agency Order No. 2200.4A
a. SIGNATURE
b.WBB&JOT
Laboratory Director
RSKERL, Ada, OK
C. DATE
37. If this material is to be forwarded to the Office of External Affairs, indicate which of the following apply:
U Has policy implications, as per attached explanation 1 1 I
Periodical as defined by OMB Circular A-3. or other item required to be reported to OMB.
38. APPROVED BY
a. FOR THE OFFICE OF EXTERNAL AFFAIRS (Signature)
b. DATE
EPA Form 2340-1 (4-84)
PUBLICATION REVIEW RECORD AND PRINTING REQUEST
-------
techniques in conjunction with chemical and physical
treatment processes, i.e., "treatment trains," is an effective
means for comprehensive site-specific remediation (Ross et
al., 1988; Sims, 1990). For instance, bioremediation may be
utilized to lower the concentration of organic contaminants in
a soil matrix before stabilization or solidification is used as a
remedial alternative for metals.
Bioremediation has been shown effective in reducing the
overall mass of a variety of organic contaminants. Full scale
systems have been utilized to remediate soil contaminated
with both crude and refined petroleum hydrocarbons (i.e.,
diesel fuel, gasoline), creosote, and pentachlorophenoi. To
date, K has not been shown effective at removing highly
structured, highly insoluble compounds such as
polychiorinated biphenyls and dioxins.
For the purposes of this document, subsurface soil refers to
unsaturated soil within the vadose zone at depths greater
than three feet below the land surface. The vadose zone
extends from the ground surface to the upper surface of the
principal water-bearing formation (Everett et al., 1982). The
vadose zone usually consists of three to six feet of topsoil
(weathered geological materials) which gradually merges with
deeper underlying earth materials such as depositional or
transported clays or sands. In this zone, water primarily
coexists with air, though saturated regions may occur.
Perched water tables may develop at interfaces of layers
(soils having different textures) of soil having less hydraulic
conductivity. Prolonged infiltration also may result in
transient saturated conditions. In some regions, the entire
vadose zone may be hundreds of feet thick and the travel
time of constituents to ground water can be hundreds or
thousands of years. Other regions may be underlain by
shallow potable aquifers that are especially susceptible to
contamination due to short transport times and reduced
potential for pollutant attenuation by soil materials and
processes.
This document addresses specific environmental processes,
factors, and data requirements for characterizing and
evaluating the application of subsurface in situ
bioremediation, and describes selected field-scale
applications of recovery and delivery systems to enhance in
situ subsurface soil bioremediation.
Overview: In Situ Subsurface Microbial Processes
and Controlling Environmental Factors
The rate and extent of biodegradation of organic chemicals
during subsurface in situ bioremediation are influenced by
several site-specific factors. These include type and activity
of microbial populations; chemical environmental factors;
bioavailability of the target chemical(s) and other substrates
required for co-metabolism, i.e., electron donor, mass
transport of moisture, nutrients, and oxygen (the terminal
electron acceptor in aerobic metabolism); toxicity; and
stratigraphy, heterogeneity, and geochemistry of the surface
or subsurface environment. A detailed discussion of the
impact of these and other factors on bioremediation can be
found in Transport and Fate of Contaminants in the
Subsurface" (EPA/625/4-89/019) and "Bioremediation of
Contaminated Surface Soils" (EPA/600/9-89/073).
Microbial Populations
Successful in situ bioremediation depends on the presence of
appropriate microbial populations which can be stimulated to
degrade contaminants of concern by modifying or otherwise
managing environmental conditions at a site. Results of
microbial characterization of deep subsurface materials have
indicated that: (1) microorganisms are present at populations
sufficient to change the chemistry of the environment when
stimulated; (2) the microbial communities are diverse and
carry out a wide range of chemical transformations; (3) a
majority (>95%) of the microbes are chemotrophic bacteria
that degrade organic chemicals to obtain energy; and (4)
environmental characteristics identified previously (oxygen
concentration, nutrient status, moisture content) are
important in influencing microbial activity and degradation
patterns (Fliermans and Hazen, 1990).
Microbial communities in the subsurface are diverse and
adaptable. Microbial populations at older sites are usually
acclimated to the contaminants of concern. Therefore, levels
of critical nutrients or electron acceptors, toxicity, and
adverse environmental conditions most often are the major
factors which limit the extent and rate of in situ
bioremediation.
Critical Environmental Conditions
There are several environmental conditions that affect activity
of soil microorganisms. These factors, along with individual
soil and waste characteristics, all interact to affect microbial
activity at specific contaminated sites. Many of these
conditions can be managed to enhance biodegradation of
organic constituents in subsurface soils. Optimum ranges for
the most critical of these factors are presented in Table 1.
Water content of soil is an important factor which regulates
microbial activity. Soil water serves as the transport medium
through which many nutrients and organic constituents
diffuse to the microbial cell, and through which metabolic
waste products are removed. Soil water also affects soil
aeration status, nature and amount of soluble materials,
osmotic pressure, pH of the soil solution, and unsaturated
hydraulic conductivity of the soil (Paul and Clark, 1989). The
water content of deeper subsurface soils may vary greatly.
Unsaturated soil samples have been obtained even from
cores collected below the water table in deep subsurface
environments, and the low water content was shown to
adversely affect microbial activity (Kieft et al., 1990).
Biodegradation rates often depend on the rate at which
terminal electron acceptors can be supplied. A large fraction
of the microbial population within soils are aerobes which use
oxygen as the terminal electron acceptor. Oxygen can be
easily depleted in subsurface soils where there is an oxygen
demand due to plant root respiration or due to normal
2
-------
Table 1. Critical environmental (actors (or microbial activity (Sims et al., 1984; Huddleston et aJ., 1986; Rochklnd and
Blackburn. 1986; Paul and Clark. 1989)
Environmental Factor
Optimum Levels
Available soil water
Oxygen
Redox potential
PH
Nutrients
25%-85% of water holding capacity; -0.01 MPa
Aerobic metabolism: Greater than 0.2 mg/l dissolved
oxygen, minimum air-filled pore space
0(10%;
Anaerobic metabolism: 02 concentrations <1%
Aerobes and facultative anaerobes, greater than
50 millivolts; Anaerobes: less than 50 millivolts
5.5-8 5
Sufficient nitrogen, phosphorus, and other nutrients so not
limiting to microbial growth
Temperature
Suggested C:N:P ratio of 100:10:1
15°C-45°C (Mesophiles)
microbial activity throughout the depth of the unsaturated
zone. Oxygen levels tend to decrease in soils having high
clay and organic matter content. Clayey soils tend to retain
higher moisture content, which restricts oxygen diffusion,
while organic matter may increase microbial activity and
deplete available oxygen. Under these circumstances,
oxygen may be consumed faster than it can be replaced by
diffusion from the atmosphere, and the soil may become
anoxic.
Facultative anaerobic organisms (which can use oxygen or
alternative electron acceptors such as nitrate or sulfate in the
absence of oxygen) and obligate anaerobic organisms then
become the dominant populations under such conditions.
The sequence of use of various electron acceptors is
determined by the redox potential and the electron affinity of
the electron acceptors present (Zehnder and Stumm, 1988).
The potential of alternative electron acceptors has been
evaluated with nitrate at field scale for contaminants
(including benzene, toluene, and xylene) in an aquifer
environment (Hutchins et al., 1991).
Redox potential also affects metabolic processes in
subsurface microbial populations (Paul and Clark, 1989).
Redox potential provides a measurement of electron density
and is related to the oxygen status of a subsurface soil. As
oxygen is removed and a system becomes more reduced,
there is a corresponding increase in electron density,
resulting progressively in an increased negative potential.
Soil pH affects growth and activity of subsurface soil
microorganisms. Fungi are generally more tolerant of acidic
soil conditions (below pH 5) than bacteria. Solubility of
phosphorus, a critical nutrient in biological systems, is
maximized at a pH value of 6.5. A specific contaminated soil
system may require management of soil pH to achieve levels
that maximize microbial activity. Control of pH to enhance
microbial activity may also aid in the immobilization of
hazardous metals in a subsurface soil system (a pH level
greater than 6 is recommended to minimize metal transport).
Subsurface soil pH may be managed through addition of an
aqueous phase containing pH adjusting chemicals through
gravity delivery systems such as infiltration galleries or
surface irrigation systems.
Microbial metabolism and growth depends upon adequate
supplies of essential macro- and micro-nutrients. Critical
nutrients such as nitrogen and phosphorous must be present
and available to microorganisms in: (1) usable form; (2)
appropriate concentrations; and (3) proper ratios (Dragun,
1988). If wastes are high in carbon (C), and low in nitrogen
(N) and phosphorus (P), biodegradation will cease when
available N and P are depleted. Therefore, fertilization of
subsurface soils may be required as a management
technique to enhance microbial degradation.
Soil temperature affects microbial growth and metabolic
activity. Biodegradation rates decrease as temperature
drops and essentially cease at temperatures below 0° C.
While surface soils exhibit both diurnal and seasonal
variations in temperature, changes of temperature decrease
with depth. Generally, only the top 30 feet of the subsurface
profile are affected by seasonal variations in temperature;
temperature is generally constant and corresponds to the
3
-------
mean annual air temperature of the locality (Kuznetsov et al.,
1963; Matthess, 1982). In the United States, temperatures in
this zone range from 3°C to 25° C (Dunlap and McNabb,
1973). Due to the high specific heat of water, wet soils are
less subject to larger diumal changes than dry soils (Paul and
Clark, 1989).
Bioavailability is a general term which refers to the
accessibility of contaminants by degrading populations.
There are two major components involved: (1) a physical
aspect related to phase distribution and mass transfer
limitations of the contaminant, and (2) a physiological aspect
related to the suitability of the contaminant as a substrate.
Major lactors which affect bioavailability include water
solubility and sorption. Target chemicals may occur in one or
more of the four phases comprising the subsurface soil
environment: (1) soil solids, including organic matter and
inorganic sand, silt, and clay particles; (2) soil water, (3) soil
gas; and (4) often a nonaqueous phase liquid (NAPL). In
general, chemicals that distribute to the water phase (more
soluble) are more bioavailable than chemicals that either sorb
strongly to solid phases or occur in a NAPL phase. NAPLs
are generally degraded from the water:NAPL interface inward
since the aqueous phase contains nutrients, oxygen, and
moisture required for microbial life processes. The
bioavailability of a NAPL phase may be increased by
increasing the surface area to volume ratio of NAPL
elements. This increases mass transfer of nutrients,
moisture, and oxygen; and decreases toxicity by decreasing
interfacial concentrations (Symons and Sims, 1988).
Substrate chemicals in the gas phase have also been found
to be bioavailable (Dupont et al., 1991; Miller et al., 1991).
Generally, chemicals that are highly sorbed, such as high
molecular weight PAHs present in creosote, petroleum, and
manufactured town gas plant wastes, are found to be
degraded at slower rates than chemicals that are only slightly
sorbed. Since the majority of the mass of target constituents
at many contaminated sites is associated with NAPL and/or
solid phases, these represent the greatest challenge with
regard to in situ bioremediation.
Bioavailability is also a function of the biodegradability of the
target chemical, i.e., whether it acts as a substrate,
cosubstrate, or is recalcitrant. The target chemical may be
physically available (i.e., water soluble and/or not sorbed to
solids) but not useful as a metabolic substrate.
Contaminants of concern may not be the dominant organic
substrate in a system. When the target chemical cannot
serve as a substrate (source of carbon and energy) tor
microorganisms, but is oxidized in the presence of a
substrate already present or added to the subsurface, the
process is referred to as cooxidation and the target chemical
is defined as the cosubstrate (Keck et al., 1989; Sims et
al.,1989). Cooxidation processes are important for the
biodegradation of high molecular weight polycyclic aromatic
hydrocarbons (PAHs), and some chlorinated solvents,
including trichloroethylene (TCE). Contaminants with
complex molecular structures or high degrees of toxicity may
not be degradable, and may persist or be recalcitrant under
aerobic conditions. Examples of recalcitrant compounds
include highly oxidized halogenated compounds such as
polychlorinated biphenyls (PCBs), pesticides such as
toxaphene, and dioxin contaminants present in wood-
preserving wastes.
The toxicity of the environment may be reduced by
decreasing the concentration of a toxic waste (e.g., creosote)
or chemical (e.g., pentachlorophenol) within one or more
subsurface phases. Concentrations of toxic chemicals in the
gas phase may be reduced through soil vacuum extraction; in
the water phase through soil flushing; in the NAPL phase
through soil flushing with water containing viscosifiers, or with
solvents or surfactants; and in the soil solid phase by
inducing partitioning of contaminants from solid to fluid
phases. All mobile phases in the subsurface have potential
for escape; therefore, containment strategies are often
necessary while the constituents within the phase are
biodegraded.
Heterogeneity of the subsurface environment limits the rate
and extent of in situ bioremediation. Restrictive layers (e.g.,
clay lenses), although more resistant to contamination, are
also more difficult to remediate due to poor permeability and
low rates of diffusion. Clay soils have larger porosities than
silty or sandy soils and therefore larger storage capacities for
contaminants, but have greater resistance to fluid flow
including aqueous, gas, and NAPL phases. Also clay layers
with poor hydraulic conductivity are less permeable to
nutrients and oxygen. In sites that have substantial clay and
silt deposits, more permeable soils will become preferential
conduits for remedial fluids, and the clay/silt deposits will
require much longer time frames for remediation. For
example, heterogeneity of the subsurface with respect to soil
layering and chemical parameters at a gas works site in the
United Kingdom presented constraints on the feasibility of
utilizing in situ bioremediation (Thomas et al., 1991).
Enhancement of In Situ Subsurface Bioremediation
The method of enhancing in situ bioremediation efforts
depends on the four phases in which contaminants can
occur, heterogeneity of subsurface matrix, and the types of
delivery and recovery systems utilized. Removing limiting or
controlling factors and establishing favorable conditions are
the primary goals of recovery and/or delivery systems.
Enhancement may be achieved by increasing bioavailability;
reducing toxicity; increasing delivery of moisture, nutrients,
and oxygen; and/or by introducing substrates that stimulate
indigenous microbial degradative activity.
A variety of strategies may be implemented to maximize
biodegradation activity in contaminated subsurface soils. The
success of in situ bioremediation efforts is often determined
by the effectiveness of the recovery and delivery systems
used to remove major sources of contaminants and to
transport nutrients and electron acceptors to the location of
the remaining contaminants. Establishing optimum levels of
essential nutrients and electron acceptors at specific
subsurface locations is often driven by physical limitations of
4
-------
the subsurface matrix on transport of fluids (liquids or gases)
used to deliver these amendments. Overcoming these
limitations is the primary goal of a delivery system, and the
development of adequate delivery technologies continues to
be the major challenge of in situ bioremediation. A summary
of delivery and recovery techniques commonly used to
manage subsurface remediation is provided in Table 2.
Making the Saturated Zone Unsaturated
Advantages of Unsaturated Systems
Because hydraulic conductivity is a function of soil moisture
content, changing a saturated soil into an unsaturated soil
greatly reduces the hydraulic conductivity and therefore the
downward transport of chemicals in the water phase to the
ground water. Also, because oxygen diffuses through air
10,000 times faster than through water, an unsaturated
environment may be maintained in an aerobic condition more
easily than a saturated environment in the presence of
oxygen-demanding chemicals (Table 3). Soil pore space that
contains a gas phase also allows removal of volatile contam-
inants (via soil vacuum extraction) in a direction that is away
from the ground water. Therefore, management of a site to
change the saturated zone to an unsaturated condition may
reduce potential for ground-water contamination as well as
enhanced oxygen delivery to stimulate in situ biodegradation.
Physical Containment
There are a variety of approaches to establishing and
maintaining dewatered conditions. In order to adequately
dewater the subsurface, it is often necessary to physically
isolate the treatment zone. Impermeable subsurface barriers
can prevent the migration of ground water by preventing
uncontaminated water from entering the contaminated site
and stopping contaminated water from leaving. Extraction
systems or drains must then be used to remove the ground
water to create an unsaturated zone.
Commonly used barriers include slurry walls, grout curtains,
and sheet piling cutoff walls to retard the flow of water under
and through a site (Devinny et al., 1990).
Ground-Water Removal
Ground-water removal can be accomplished by hydraulic
pumping and/or drainage trenches. Hydraulic pumping using
a well-point system is one such technique (Devinny et al.,
1990) using short lengths of plastic or Teflon well screen
placed in the saturated zone.
Ground water can also be removed using subsurface drains
or drainage ditches. Subsurface drains are constructed by
excavating a trench to the desired depth, partially backfilling
the trench with highly permeable sand or gravel, placing a
plastic or ceramic drain tile in the sand and gravel bed, and
completing the backfilling (Devinny et al., 1990).
Drainage ditches or surface drains are similar to subsurface
drains except that no collection pipes or tiles and backfills are
used. They may be used at sites underlain by poorly
permeable soils (Devinny et al., 1990).
Recovery and Delivery Technologies for Subsurface
Bioremediation
Recovery and delivery technologies are those that facilitate
transport of materials either out of or into the subsurface
(Murdoch et al., 1990). Recovery technologies are primarily
utilized for contaminant source reduction. High levels of
contamination present as either trapped residuals or NAPLs
can severely limit success of bioremediation attempts.
Therefore, removal of as much of this initial contaminant
mass as possible is a prerequisite to in situ bioremediation
efforts.
Specific recovery and delivery technologies for enhancing in
situ bioremediation of subsurface soils are identified in
Table 2. Each identified technology is discussed below with
regard to its applications and limitations, and current status.
Recovery Technologies
The principal recovery technologies used for subsurface
remediation depend on the ability to move fluids. Also
involved is the ability to move contaminants by altering their
solubility or sorption characteristics (Murdoch et al., 1990).
These techniques are used to move materials from the
subsurface soil environment in order to enhance in situ
bioremediation by addressing one or more limiting factors
identified in Tables 1 and 2, including: soil vacuum extraction,
soil flushing, steam stripping, and radio frequency heating.
Soil vacuum extraction
Soil vacuum extraction (SVE) (also referred to as subsurface
or forced air venting, in situ air stripping, or soil vapor
extraction) involves the removal of contaminants carried in
the soil gas phase by reduction of the vapor pressure within
the soil pores by applying a vacuum. As clean air is drawn
through the soil, the contaminants are removed. This
process is driven by concentration differences between solid,
aqueous, and NAPL phases and the clean air that is
introduced through the soil vacuum extraction process.
Vacuum extraction is most applicable to sites contaminated
with highly volatile compounds, such as those associated with
gasoline and solvents (e.g., perchloroethylene,
trichloroethylene, dichloroethylene, trichloroethane, benzene,
toluene, ethylbenzene, and xylene).
Important soil characteristics that should be measured or
estimated to determine the feasibility of vacuum extraction at
a specific site include physical factors that control the rate
and extent of air flow through contaminated soil, and
chemical factors that determine the amount of contaminant
that partitions from soil to air. These factors include: bulk
density (weight per volume); total porosity (void spaces
5
-------
Tabic 2. Management strategies for addressing factors limiting In situ bloremedlation of subsurface soils
Limiting Factor
Management Response
Delivery or Recovery Technique
Bioavailability limited due to NAPL
Reduce NAPL mass
Gravity or forced delivery; Soil flushing,
Steam stripping, Hydraulic fracturing
Bioavailability limited by sorption or slow
mass transport through soil matrix
Reduce sorption, Increase mass transport
Soil flushing, Steam stripping, Hydraulic
fracturing
Moisture
Add water or water saturated air
Gravity or forced delivery; Bioventing, Cyclic
pumping
Nutrients
Add nutrients in water or as ammonia gas
Gravity or foroed delivery; Bioventing. Cyclic
pumping
Oxygen/Redox
Add air
Bioventing, Hydraulic fracturing, Cyclic
pumping, Radial drilling, Kerfing
Toxicity
Remove chemicals
Soil vacuum extraction, Soil flushing, Steam
stripping
PH
Adjust soil pH
Gravity or forced delivery
Temperature
Increase temperature
Radio frequency heating, Steam stripping
Substrate Addition
Add In water or air
Gravity or forced delivery; Bioventing,
Hydraulic fracturing
Heterogeneity
Add or withdraw material in more restrictive
layers
Cyclic pumping, Hydraulic fracturing, Racial
drilling, Kerfing
Table 3. Carrier fluid oxygen supply requirements (Dupont et a!., 1991)
Carrier g Carrier/g 02
Water
Air Saturated 110,000
Pure 02 Saturated 22,000
500 mg/L H202 (100% Utilization) 2.000
Air (20.0% Oj) 13
6
-------
between soil grains) and air-filled porosity (that portion of the
total porosity filled with air); diffusivity of volatiles (amount of
volatiles which move through an area over time); soil
moisture content (percentage of void spaces filled with
water); air phase permeability (ease with which air moves
through soils); texture; structure; mineral content; surface
area; temperature; organic carbon content; heterogeneity;
depth of air permeable zone; and depth to water table
(Metcalf & Eddy, Inc., 1991). Soils at sites where vacuum
extraction is used should be fairly homogeneous and have
high permeability, porosity, and uniform particle-size
distributions (Metcalf & Eddy, Inc., 1991). Soil vapor
transport can be severely limited in a soil with high bulk
density, high soil water or high NAPL content, low porosity,
and low permeability. In heterogeneous soils, air flows
preferentially through more permeable zones, leaving less
permeable zones untreated.
Contaminant characteristics that affect the feasibility of
vacuum extraction include the extent and degree of
contamination, vapor pressure, Henry's law constant,
aqueous solubility, diffusivity, and partition coefficients. Due
to the high solubility of many organic contaminants in NAPL
phases, the presence of NAPL in subsurface soil systems
may significantly affect the distribution of the compounds in
various phases, and their fate in SVE systems. Specific
contaminant and soil conditions that determine the feasibility
of vacuum extraction are presented in Table 4.
The efficiency of a vacuum extraction system can be
enhanced in several ways. For example, a system of air
injection wells can be installed at the perimeter of a
contaminated area (Metcalf & Eddy, Inc., 1991) which can
be connected to air blowers to force air into the soil or
remain open to the atmosphere. Use of air injection wells
can result in increased soil air flow rates and a larger area
through which clean air can move.
Pulsed pumping may be used to give contaminants time to
desorb from solid surfaces, diffuse from restricting layers,
and volatilize from residual saturation (NAPL) in the soil pore
space. Using pulsed pumping for recovery of contaminants
allows a lower volume of air with higher concentrations of
contaminants to be recovered.
If ground water is at or near the zone of soil contamination,
water table rise may occur due to reduced air pressure near
extraction wells (Metcalf & Eddy, Inc, 1991). Ground-water
Table 4. Condition* affecting feasibility of use of vacuum extraction (U.S. EPA, 1990; Metcalf & Eddy, Inc., 1991)
Condition
Favorable
Unfavorable
Contaminant:
Dominant form
Vapor phase
Solid or strongly sorbed to soil
Vapor pressure
>100 mm of mercury
<10 mm of mercury
Water solubility
<100mg/l
>1,000 mgd
Henry's Law Constant
>0.01 (timensionless)
<0.01 (dimensionless)
Soil:
Temperature
>20°C (usually will require external heading
of soils)
<10°C (common in northern climates)
Air conductivity
>10"4 cm/s
<10"® cm/s
Moisture content
<10% (by volume)
>10% (by volume)
Composition
Homogeneous
Heterogeneous
Surface area of soil matrix
<0.1 m2/g of soil
>1.0m2/gof soil
Depth to ground water
>20 m
<1 m
7
-------
pumping may be used to counteract the water table rise, as
well as to expose additional contaminated soil that can be
treated by vacuum extraction.
Horizontal extraction wells (wells drilled parallel to ground
surface) have been used for deep subsurface contamination
at the U.S. Department of Energy Savannah River facility to
access larger areas of the contaminated site (Hazen.1992).
This use of horizontal wells may be a means to reduce costs
associated with deep subsurface remediation since only a
single hole may be required to access contaminated areas
instead of many vertical wells.
The performance of a vacuum extraction system is monitored
by system operational characteristics and by treatment
efficiency characteristics (Metcalf & Eddy, Inc., 1991).
System characteristics include strength of vacuum applied,
air flow rate, and contaminant concentrations and moisture
content in the vented gas. Wells are used to monitor
pressure in the contaminated area. Efficiency of treatment is
monitored by soil gas analyses, and soil core analyses to
determine residual concentration of contaminants. For more
detailed discussions of soil venting evaluation, see
"Evaluation of Soil Venting Application* (EPA/540/S-92/004).
Since soil vacuum extraction is an in situ treatment technique
that requires only addition of ambient air to the subsurface, K
can be applied with little disturbance to existing facilities and
operations (Metcalf & Eddy, Inc., 1991). SVE can be used at
sites where areas of contamination are large and deep, or
when the contamination is beneath a building. The system
can be easily modified, depending on additional analytical
and subsurface characterization data and/or changing site
conditions. Even if vacuum extraction can be implemented at
a site, most of the conditions listed in Table 4 must be met, or
the cost and time for cleanup will be prohibitive.
The use of SVE at remedial sites has been reviewed by the
U.S. EPA (1989a) and classified as a developed technology
for remedial applications. It is currently the most commonly
used in situ remedial technology (Murdoch et al., 1990). Soil
vacuum extraction may be used to reduce toxic concen-
trations of contaminants to levels which are more conducive
to bioremediation. In addition, it will also deliver oxygen to
the subsurface which is required by aerobic bacteria.
Soil flushing
In situ soil flushing is used to accelerate movement of
contaminants through unsaturated materials by solubilizing,
emulsifying, or chemically modifying the contaminants. A
treatment solution is applied to the soil and allowed to
percolate downward and interact with contaminating
chemicals. Contaminants are mobilized by the treatment
solution and transported downward to a saturated zone
where they are captured in drains or wells and pumped to the
surface for recovery, treatment, or disposal (Murdoch et al.,
1990). In combination with bioremediation, the flushing
solution may be amended with nutrients to enhance biological
activity (Metcalf & Eddy, Inc, 1991).
Treatment solutions are delivered to the contaminated zone
by using either gravity or forced methods. Forced delivery
consists of various pumping techniques. Gravity delivery
methods include surface flooding, ponding, spraying, ditching
and subsurface infiltration beds and galleries (Amdurer et al.,
1986). Barriers, such as slurry walls, may be required to
prevent the transport of contaminants away from the site
(Metcalf & Eddy, Inc. 1991). A ground-water extraction
system must be used to capture the flushing solution and
associated contaminants. In some cases, the flushing
solution may be treated to remove the contaminants and
reused, and in others it may require disposal.
Efficiency of soil flushing is related to two processes: the
increase in hydraulic conductivity that accompanies an
increase in water content of unsaturated soil, and the
selection of treatment solutions with regard to the
composition of the contaminants and the contaminated
medium. The hydraulic conductivity of soils decreases
markedly with decreases in water content; therefore, the flow
of liquids through unsaturated soils is extremely slow and the
recovery of contaminants by conventional pumping
techniques is not possible. With soil flushing, the water
content and consequently the hydraulic conductivity of the
soil is increased (Murdoch et al., 1990). However,
heterogeneities in soil permeability may result in incomplete
removal of contaminants.
At sites where water-soluble contaminants are present, water
can be used to flush or mobilize the contaminants (MetcaH &
Eddy, Inc., 1991). Surfactants can be added to increase the
mobility of hydrophobic organic contaminants, such as oils
and petroleum. Examples of other flushing solutions include:
acidic aqueous solutions (for the removal of metals and basic
organic constituents including amines, ether, and anilines),
basic solutions, chelating agents, oxidizing agents, and
reducing agents. Toxicity of flushing solutions to soil
microorganisms should be considered when followed by
bioremediation of residual contamination. The flushing
solution may change physical and chemical properties of the
soil environment that affect bioremediation potential.
The level of treatment that will be achieved is dependent on
selection of an appropriate flushing solution, extent and time
of contact between the solution and waste constituents, soil
partition coefficients of the waste constituents, and the
hydraulic conductivity of the soil (Metcalf & Eddy, Inc., 1991).
Soil flushing is not applicable to soils with low hydraulic
conductivities (e.g., less than 1 ft/day), or for contaminants
that are strongly sorbed to the soil (e.g., PCBs, dioxin).
Soil flushing has been classified by the U.S. EPA as a
developed technology used for recovery in remedial
applications (Murdoch et al., 1990). Although the technology
has been tested at field-scale, soil flushing has not yet been
used extensively in large-scale clean-up operations. As with
SVE systems, soil flushing may be utilized with bioremed-
iation as a coupled technology. Soil flushing may initially be
utilized to lower toxic or extreme concentrations of contam-
inants to a manageable level for biological processes which
8
-------
may be utilized as a polishing step to remove those
contaminants which were not removed through the flushing
process. If biological processes are used during or after soil
flushing, the compatibility of the soil flushing solution with
subsurface bacteria must always be considered.
Delivery Techniques
The major limiting factor to the bioremediation of amenable
compounds is the delivery of required nutrients, co-oxidation
substrates, electron acceptors or other necessary enhancers
of microbial growth. Delivery techniques are used to add
required materials to the subsurface environment to enhance
in situ bioremediation by addressing one or more limiting
factors identified in Tables 1 and 2. A variety of delivery
techniques are in use or are being developed (Figures 1-3).
These include soil venting, gravity and forced hydraulic
delivery, hydraulic fracturing of low permeability zones, radial
drilling, and cyclic pumping. Of these, only gravity and forced
hydraulic delivery and venting systems are in common use at
sites. The other three approaches are still in developmental
stages.
In Use: Gravity/Forced Hydraulic Delivery and
Bioventing
Gravity and Forced Hydraulic Delivery
Irrigation technologies were among the first utilized for
enhancing in situ biodegradation. Gravity methods are used
to deliver water and amendments to the contaminated
subsurface by applying the solutions directly over the
contaminated area. Applied solutions then percolate
downward through the subsurface to contaminated zones.
Application methods consist of both surface and subsurface
spreading (Amdurer et al., 1986).
Surface application methods include flooding, ponding,
ditches, and sprinkler systems. These methods are generally
applicable to contamination at depths less than 15 feet.
Flooding is a surface application method in which the solution
is spread over the land surface in a thin sheet. Flooding is
applicable to sites that are flat or gently sloped (i.e., less than
3 percent slope), uniform, without gullies or ridges, and have
soils with high hydraulic conductivities (i.e, greater than 10'3
cm/sec; such as those found in sands, loamy sands, and
sandy loams).
Ponding can be used to increase the infiltration rate of the
applied solution above that achieved by flooding. Ponds are
constructed by excavating into the ground or by constructing
low berms. The depth of the solution in the pond becomes
the driving force to increase infiltration rates. Ponding can be
used in sandy or loamy soils and in fiat areas.
The ditch method of surface spreading utilizes flat-bottomed,
shallow, narrow ditches to transport the solution over the land
surface; allowing for infiltration of the solution into the ground
through both bottom and side surfaces. Gradients in the
ditches are kept small to prevent erosion as well as to allow
residence time for infiltration. Ditches may be constructed by
excavating surface materials or by building small
embankments. Ditches are used at sites where it is not
desirable to completely cover an entire area with the solution.
\*r Gravel Bed
v V \ V
Figure 1. Schematic of a sprinkling system used to deliver nutrients to contaminated subsurface soil.
9
-------
Extractad Air Injactad Air Extracted Air
0 0 0
Figure 2. Schematic of a bloventing system designed to deliver air to contaminated subsurface soil.
Figure 3. Schematic of a ponding system used to deliver nutrients to contaminated subsurface soil.
10
-------
Sprinkler systems can be used to deliver solutions uniformly
and directly to the ground surface. These systems are less
susceptible to topographical constraints than flooding and
ponding. Sprinkler systems have been used successfully to
deliver nutrients and moisture to bioventing systems where
the site was contaminated to a depth of 50 feet (Dupont et
al., 1991).
Subsurface gravity delivery systems include infiltration
galleries (or trenches) and infiltration beds. These systems
are applicable to sites where there is deep contamination or
where the surface layers have low permeability. Subsurface
systems consist of excavations filled with a porous medium
(e.g., coarse sands or gravels) that distribute solutions to the
contaminated area. An infiltration gallery consists of a pit or
trench that is filled with gravel or stones. The solution fills the
pores in the gallery and is distributed to the surrounding soils
in both the vertical and horizontal directions. This system is
most applicable to sites with sandy or loamy soils. In sites
with silty soils, an infiltration gallery can be used but
application rates will be reduced. Solutions can be
introduced into the gallery by injection at locations along the
length of the gallery or through perforated distribution pipe.
Infiltration galleries can be used in sites with steep slopes
(i.e., up to 25 percent slope) and uneven terrain. Infiltration
beds are similar to galleries but are wider and contain more
than one perforated distribution pipe. Infiltration occurs
almost entirely through the bottom, with little infiltration
through sidewall surfaces. This system is applicable to soils
with sandy and loamy textures, but limited to sites where the
topography is relatively flat (i.e., with less than 5 percent
slope) and the terrain is even. Beds can saturate larger
areas than a single trench and are easier to install than a
multi-trench system.
Forced systems deliver fluids under pressure into a
contaminated area through open end or slotted pipes that
have been placed to deliver the solution to the zone requiring
treatment (Amdurer et al., 1986). These systems are
generally applicable to soils with hydraulic conductivities
greater than 10"4 cnVsec (i.e., fine sandy or coarse silty
materials) and high effective porosities (i.e., ranging from 25
to 55 percent). A maximum injection pressure must be
established to prevent hydraulic fracturing and uplift in the
subsurface, which would cause the fluid to travel upward
rather than through the contaminated area. Unlike gravity
systems, a forced delivery system is theoretically
independent of surface topography and climate.
Design considerations for gravity and forced delivery systems
are presented in Amdurer et al. (1986). Application of gravity
delivery systems in subsurface bioremediation systems has
been demonstrated in bioventing systems (Dupont et al.,
1991; Miller et al., 1991). In Russia, methane-oxidizing
bacteria grown in fermenters have been injected into lateral
core holes in a coal mine (Fliermans and Hazen, 1990). This
process has been shown to reduce methane concentration in
the air by 50-60 percent in one month, thus reducing the risk
of explosions and fire.
Soil Bioventing
Soil bioventing incorporates soil vacuum extraction processes
to deliver oxygen to the subsurface to enhance in situ
bioremediation of organic contaminants. The large amounts
of oxygen-saturated water required for bioremediation often
cannot be delivered due to hydraulic conductivity limitations.
For example, benzene and hexane, which are common
hydrocarbon contaminants, require more than 3 g 02 per g of
hydrocarbon for mineralization. Soil bioventing is applicable
to remediation of contaminants of low volatility and can also
reduce concentrations of volatile contaminants in off-gases,
thus reducing the amount of contaminants requiring off-gas
treatment.
To accomplish bioventing, soil vacuum extraction processes
are operated at lower than usual air flow rates to reduce
vapor extraction quantities and maximize vapor retention
times. Soil moisture levels necessary for biological activity
are usually higher than those recommended for optimum
vacuum extraction operations. The addition of nutrients may
also enhance bioremediation. Nutrient addition can be
accomplished by surface application, incorporation by tilling
into surface soil, and transport to deeper layers through
applied irrigation water. Increased soil temperatures have
been shown to enhance biodegradation rates in bioventing
systems (Miller et al., 1991). Possible means of increasing
soil temperature include the use of heated air, heated water,
or low-level radio-frequency heating. High temperature
should be avoided, since this can result in decreased
microbial populations anchor activity.
Soil bioventing has been demonstrated in several field
applications (Dupont et al., 1991; Hinchee et al., 1991;
Hoeppel et. al., 1991; Miller et al., 1991; van Eyk and
Vreeken, 1991; Urlings et al., 1991). At Hill Air Force Base in
Utah, a JP-4 jet fuel spill occurred in January 1985 that
resulted in the contamination of approximately 0.4 hectares
(1 acre) to a depth of approximately 50 feet with approxi-
mately 25,000 gallons of JP-4 (Dupont et al., 1991). Soil total
petroleum hydrocarbon (TPH) concentrations at the site were
as high as 15,000 mg/kg, with average TPH levels of 1,500
mg/kg. Site soil consists of mixed coarse sand and gravel
deposits with interspersed, discontinuous clay stringers to a
confined ground-water table located approximately 600 feet
below ground surface. Prior to initiating a full-scale vacuum
extraction project, the fuel tanks were excavated,
refurbished, and installed in an above-ground concrete
cradle. Excavated soil was placed in a pile and subjected to
vacuum extraction.
An SVE system consisting of 15 wells in the undisturbed soil
and 10 wells in the excavated soil pile and under the tanks
was installed to provide access to the contaminated soil and
allow flexibility in the operation of the venting system. The
system was operated in a conventional mode to maximize
the recovery of volatile components of the JP-4 through
volatilization. Venting was initiated on December 18,1988, at
a rate of 1,270 ft3/hr (approximately 0.04 pore volumes/day),
11
-------
and gradually increased to approximately 74,000 ft3/hr
(approximately 2.5 pore volumes/day) as the hydrocarbon
levels in the vent gas decreased over time. The venting rate
during the start-up period was limited by the operating
conditions of the catalytic incinerator used to treat the
collected vent gas. This high-rate operating mode was
maintained from December 18,1988, through September 15,
1989 with approximately 340 pore volumes (245 x 106 ft3) of
soil gas extracted from the site.
In situ respiration tests conducted during the high-rate SVE
operating period indicated that significant respiration was
occurring without nutrient or moisture addition, and that
enhancement of biodegradation might be possible under
modified site management conditions. Biodegradation was a
significant removal mechanism during the initial high-rate
venting, accounting for 15 to 25 percent of the recovered
hydrocarbon. To assess the potential for enhancing
biodegradation rates, a series of laboratory and field
biotreatabilKy studies were conducted to evaluate moisture
and nutrient additions. The effect of SVE system operational
parameters on biodegradation rates was also evaluated by
decreasing airflow rates and increasing flow path length.
A number of in situ respiration tests were conducted during
the field studies to assess the impact of different engineering
management options on microbial activity. A total of three
tests were conducted to monitor the effect of different
management aipproaches.including: (1) flow rate and
operating configuration modifications , (2) moisture addition,
and (3) moisture and nutrient addition. Biodegradation
reactions were estimated based on cumulative oxygen
consumption and carbon dioxide production. All
biodegradation calculations were normalized to background
C02 and 02 concentrations so that the effects of field
management techniques could be isolated from changes in
background respiration taking place during the study.
The results of these studies indicated that moisture addition
and operational modifications significantly enhanced
biodegradation rates. Based on analyses of Oz uptake rates,
moisture addition (35% to 50% field capacity) was shown to
statistically accelerate in situ respiration at the site. However,
nutrient addition generally did not statistically increase the
degradation rates of residual JP-4 constituents. The
operational modifications (reduced air flow rate, increased
path length) significantly improved biodegradation rates. Fuel
removal due to biodegradation increased to greater than 80
percent, resulting in an additional 12,000 lb of total petroleum
hydrocarbons being degraded during the bioventing portion of
the study. Initial hydrocarbon (on a carbon equivalent basis)
removal rates of 70 lb/day were maintained at an average
rate of greater than 100 lb/day following system operating
modifications.
Soil bioventing was also investigated at Tyndall Air Force
Base in Florida to remediate sandy soils contaminated by
past jet fuel storage activities (Miller, 1990; Miller et al.,
1991). Hydrocarbon concentrations in the soil ranged from
30 to 23,000 mg/kg. The contaminated area was dewatered
prior to system installation. The impact of moisture and
nutrient addition was investigated during a 7-month period.
Moisture addition had no significant effect on biodegradation
rate in this system. Nutrient addition also did not affect
biodegradation rate, since naturally occurring nutrients were
present in adequate quantities to support the amount of
biodegradation observed. Biodegradation rates were shown
to be affected by soil temperature and followed predicted
rates based on the van't Hoff-Arrhenius equation. Fifty-five
percent removal was attributed to biodegradation during the
period of study, but a series of flow rate tests showed that
biodegradation could be increased to 85 percent by
decreasing air flow rates. The optimal air flow conditions were
found to be the removal of 0.5 air flow volumes per day. The
contaminated gas phase was drawn through clean soil to
increase gas residence time within the soil. This augmented
in situ biodegradation and eliminated the need for off-gas
treatment as well as reducing exposure to off-gas.
Research: Hydraulic Fracturing, Radial Drilling
Research areas are focusing on methods to increase the
capacity of current systems to deliver increased
concentrations of required solutions to the subsurface. Two
of these systems are discussed below.
Hydraulic fracturing
Hydraulic fracturing is a technique that involves using
hydraulic pressure to induce cracking in rock or clay/silt
lenses in the vicinity of a borehole, which develops a larger
framework of interconnected pore space. The newly created
pore space is filled with solid, granular materials, which can
act as permeable channels to increase the rate and area of
delivery of fluids containing nutrients or oxygen to the
subsurface (Murdoch et al., 1990; Murdoch et al., 1991;
Davis-Hoover et al., 1991). The hydraulic fractures may be
filled with granules of slow-dissolving nutrients or oxygen-
releasing chemicals, which may provide a reservoir of these
compounds for the enhancement of bioremediation. This
technique could also potentially be used in recovery systems,
e.g., by increasing extraction of vapor phases in soils with low
permeabilities, or by forming horizontal sheet-like drains to
capture leachates in soil flushing systems.
Hydraulic fracturing has been successfully utilized in
petroleum engineering in many types of geologic materials,
ranging from granite to poorly consolidated sediments. For
remedial applications, it has been demonstrated in soft clay
soils at shallow depths, but has not yet been demonstrated in
a wide range of soils or at waste sites. For use in remedial
applications, hydraulic fracturing has been classified by the
U.S. EPA as an emerging technology (i.e., research on its
use is in progress) (Murdoch et al., 1990).
Radial drilling
Radial well technology consists of drilling horizontal wells
radially outward from a central borehole. This enhances
access to a contaminated subsurface environment by
12
-------
increasing the volume serviced by each vertical well
(Murdoch et al., 1990). Radial wells can be placed at the
same level or on multiple levels in the same borehole. The
use of horizontal welis allows access to fracture zones that
are perpendicular to the ground surface and allows
contaminated areas to be entered laterally rather than
vertically.
Radial wells have been installed in both consolidated rock
and unconsolidated materials (Murdoch et al., 1990). In
unconsolidated formations, drilling rates range from 5 to 120
ft/min, while in very hard, homogeneous basalt, rates range
from 0.10 to 0.50 ft/min. For use in remedial applications,
radial well drilling has been classified by the U.S. EPA as an
emerging technology (i.e., research on its use is in progress)
(Murdoch et al., 1990).
Waste, Soil, and Site Information Requirements for
Evaluation and Management of In Situ
Bioremediation
Adequate site characterization including: surface and
subsurface soil characteristics, hydrogeology, and
microbiological characteristics, serve as the basis for rational
design of any subsurface soil bioremediation system. A
thorough site characterization is necessary to determine both
the three-dimensional extent of contamination as well as
engineering and management constraints which may limit the
rate and extent of remediation. Specific characterization
information regarding waste, soil, and hydrogeology is
required in order to assess the potential effectiveness of
bioremediation. Specific waste characterization information
required includes the relative aerobic biodegradability of
waste chemicals under optimum conditions. Important
hydraulic, physical, and chemical properties of soils that
affect the behavior of organic constituents in the vadose zone
are presented in Sims et al. (1989). Subsurface soil
characterization information required includes identification of
limiting soil environmental factors identified in Table 1.
Required site characterization information includes
identification of potential limiting factors with regard to relative
ease of delivery and recovery listed in Table 2.
Based upon waste, subsurface soil, and site characterization
information, appropriate containment strategies need to be
considered for the mobile contaminant phases associated
with the subsurface (Figure 1). Naturally occurring
containment may be sufficient with regard to preventing
escape of mobile phases under existing site conditions.
However, other containment strategies may need to be
considered if materials are to be added or removed from the
subsurface to stimulate microbial activity. These may include
volatiles removed in vacuum extraction, water used to add
oxygen and nutrients, or NAPLs if soil flushing is carried out.
For each chemical (or chemical class), the information
required is summarized as: (1) characteristics related to
potential leaching, e.g., water solubility, octanol/water
partition coefficient, solid sorption coefficient; (2) volati-
lization, e.g., vapor pressure, relative volatilization index;
(3) Henry's Law Constant; (4) potential biodegradation, e.g.,
half-life, degradation rate, biodegradability index; and (5)
chemical reactivity, e.g., hydrolysis half-life, soil redox
potential (Sims et al., 1984; Sims et al., 1989).
Information from waste and site characterization studies, and
laboratory evaluations of biodegradation may be integrated
by using appropriate mathematical models to predict: (1) the
potential for bioremediation of and (2) the potential for cross-
contaminating other media (i.e., ground water under the
contaminated area, atmosphere over the site or at the site
boundaries, surface waters, etc). The models used will be
highly dependent on site characteristics and contaminants of
interest. These may range from "back-of-the-envelope"
calculations to sophisticated fate and transport computer
models.
Mass Balance Approach to In Situ Subsurface
Bioremediation
Successful subsurface bioremediation depends upon
thorough characterization and management of each
subsurface phase with regard to containment, stimulation of
microbial activity, and monitoring strategies. The chemical
mass balance approach provides a framework for evaluating,
managing, and monitoring subsurface soil bioremediation
(Sims, 1990). Mass balance helps obtain specific information
that is needed to determine fate and behavior, evaluate and
select management options for in situ bioremediation, and
monitor treatment effectiveness for specific chemicals in
specific subsurface phases. The information needed to
construct a mass balance for subsurface contamination
simultaneously addresses site characterization and
biodegradation rates.
A necessary first step in mass balance requires
characterizing each phase present in the subsurface (Figure
1) with regard to location, amount, and heterogeneity of the
subsurface environment to assess which chemicals are
associated with which phase(s). This information allows
determination of the relative bioavailability of chemicals. For
example, chemicals associated with aqueous and gas phases
are generally more bioavailable than chemicals associated
with solid and NAPL phases. In addition, chemicals
associated with aqueous and gas phases are more prone to
migration. This information also allows determination of the
need for containment by defining where contamination is
migrating under the influence of natural processes. The
problem can be defined in the context of mobility versus
biodegradation for chemicals. Is the rate of biodegradation
(either natural or enhanced) such that chemicals which are
prone to leaching or volatilization degrade before either
occurs? Using mathematical models or other tools,
chemicals can be ranked in order of their relative tendencies
to leach, volatilize, or remain in-place under subsurface site-
specific conditions. Containment and management options
can then be selected that address specific escape and
attenuation pathways. For example, SVE may be
appropriate as a managerial tool to remove highly volatile,
biologically recalcitrant chemicals from soil before switching
13
-------
to a bioventing mode to remove less volatile, easily
biodegraded compounds. Specific waste phases may be
addressed at specific times during bioremediation. Finally,
comprehensive monitoring programs can be designed to
track specific chemicals in specific phases in the subsurface
at specific times.
After a phase is contained through natural or managed
processes, techniques to enhance microbial activity may be
applied. Monitoring strategies can then be designed to
ensure that the rate and extent of biodegradation within each
phase, as well as transfer of chemicals between phases, are
measured. Biodegradation rates of organic compounds in
soil systems are generally measured by monitoring their
disappearance in a soil through time. Rates of degradation
are often expressed as a function of the concentration of one
or more of the constituents being degraded. This is
accomplished by measuring at specific time intervals the
concentration of contaminants at interest (in the medium of
interest, i.e., soil phase, gas phase, etc.), through a properly
designed sampling and analysis plan. This sampling and
analysis plan should be statistically valid and provide
sufficient information to determine the rate of disappearance
of contaminants of interest or appropriate surrogates, such
as petroleum hydrocarbons (TPH). Care should be taken to
ensure that transfer or partitioning of contaminants from one
phase to another is not misinterpreted as biodegradation
within the source phase. Abiotic losses such as volatilization
and leaching must be defined in order to accurately
determine biodegradation rates. Identification of metabolic
transformation products is also necessary since metabolites
may be more mobile or toxic than the parent compounds. In
addition, measuring only for parent compounds and not
metabolites may tremendously overestimate extent of
biodegradation. In addition, identification of metabolites is
warranted when known daughter products are toxic.
Recommendations
There is currently a lack of information concerning some
aspects of in situ bioremediation of subsurface soils. Specific
areas where additional information is required include site
characterization with regard to effects of physical, chemical,
and hydrologic properties on microbial distribution, numbers,
and activity. Field research to obtain these types of
information is currently limited; however, this information is
required in order to estimate the feasibility of bioremediation
for subsurface contamination. Implementation of subsurface
remediation is currently limited to a significant extent by the
difficulty of establishing adequate systems for delivery and
recovery of chemicals for augmenting biological activity. As
research continues, these difficulties may be overcome as
more information becomes available concerning the
applicability of innovative technologies in the remediaiton of
contaminated soil.
References
Amdurer, M., R.T. Fellman, J. Roetzer, and C. Russ.
Systems to Accelerate In Situ Stabilization of Waste
Deposits. EPA/540/2-86/002, Hazardous Waste Engineering
Research Laboratory, U.S. Environmental Protection Agency,
Cincinnati, OH.
Balkwill, D.L, and F.J. Wobber. 1989. Deep Microbiology
Transitional Program Plan. DOE/ER-0328, Office of Energy
Research, Office of Health and Environmental Research,
U.S. Department of Energy, Washington, DC.
Brown, R.A., Norris, R.A., and Raymond, R.L. 1984. Oxygen
transport in contaminated aquifers with hydrogen peroxide.
Proceedings, Petroleum Hydrocarbons and Organic
Chemicals in Groundwater-Prevention, Detection and
Restoration Conference, Houston, TX, and National Water
Well Association, Worthington, OH.
Davis-Hoover, W.J., LC. Murdoch, S.J. Vesper, H.R.
Pahren, O.L Sprockel, C.L. Chang, A. Hussain, and W.A.
Ritschel. 1991. Hydraulic fracturing to improve nutrient and
oxygen delivery for in situ bioreclamation. pp. 67-82. In:
In Situ Bioreclamation: Applications and Investigations for
Hydrocarbon and Contaminated Site Remediation (R.E.
Hinchee and R.F. Olfenbuttel, eds.). Butterworth-Heinemann,
Boston, MA.
Dev, H., and D. Downey. 1988. Zapping hazwastes. Civil
Engineering (August): 43-45.
Dev, H„ J.B. Condorelli, C. Rogers, and D. Downey. 1986. In
situ frequency heating process for decontamination of soil,
pp. 332-339. In: Solving Hazardous Waste Problems,
Learning from Dioxins. ACS Symposium Series 338,
American Chemical Society, New York, NY.
Devinny, J.S., LQ. Everett, J.C.S. Lu, and R.L. Stollar.1990.
Subsurface Migration of Hazardous Wastes. Van Nostrand
Reinhold, New York, NY.
Dragun, J. 1988. The Soil Chemistry of Hazardous Materials.
Hazardous Materials Control Research Institute, Silver
Spring, MD.
Dunlap, W. J., and J. F. McNabb. 1973. Subsurface
Biological Activity in Relation to Ground Water Pollution.
EPA-660/2-73-014, Robert S. Kerr Environmental Research
Laboratory, U.S. Environmental Protection Agency,
Ada. OK.
Dupont, R.R., W.J. Doucette, and R.E. Hinchee. 1991.
Assessment of in situ bioremediation potential and the
application of bioventing at a fuel-contaminated site. pp. 262-
282. In: In Situ Bioreclamation: Applications and
Investigations for Hydrocarbon and Contaminated Site
Remediation (R.E. Hinchee and R.F. Olfenbuttel, eds.).
Butterworth-Heinemann, Boston, MA.
Everett, L.G., E.W. Hoylman, L.G. McMillion, and LG.
Wilson. 1982. Vadose zone monitoring concepts at landfills,
impoundments, and land treatment disposal areas. In:
Management of Uncontrolled Hazardous Waste Sites.
-------
Hazardous Materials Control Research Institute, Silver
Spring, MD.
Fliermans, C.B., andT.C. Hazen (eds.). 1990. Microbiology
of the Deep Subsurface, Proceedings, First International
Symposium. U.S. Department of Energy and Westinghouse
Savannah River Company, Savannah, GA, January 15-19.
Ghiorse, W.C., and J.T. Wilson. 1988. Microbial ecology of
the terrestrial subsurface. Advances in Applied Microbiology
33:107-172.
Hazen, T.C. 1992. Full scale underground injection of air,
methane, and other gases via horizontal wells for in situ
bioremediation of chlorinated solvent contaminated ground
water and soil. Proceedings, Bioremediation Case Studies at
Federal Facilities, Oak Ridge National Laboratory, Oak
Ridge, TN, August.
Hinchee, R.E, D.C. Downey, R.R. DuPont, P.K. Aggarwal
and R.N. Miller. 1991. Enhancing biodegradation of
petroleum hydrocarbons through soil venting. Journal of
Hazardous Materials 27:315-325.
Hoeppel, R.E., R.E. Hinchee and M.F. Arthur. 1991.
Bioventing soils contaminated with petroleum hydrocarbons.
Journal of Industrial Microbiology 8:141-146.
Huddleston, R.L., C.A. Bleckmann, and J.R. Wolfe. 1986.
Land treatment biological degradation processes, pp. 41-61.
In: (Land Treatment: A Hazardous Waste Management
Alternative (R.C. Loehr and J.F. Malina, eds.). Water
Resources Symposium No. 13, Center for Research in Water
Resources, University of Texas at Austin, Austin, TX.
Hutchins, S.R., W.C. Downs, G.B. Smith, D.A. Kovacs, D.D.
Fine, R.H. Douglas, and D.J. Hendrix. 1991. Effect of nitrate
addition on biorestoration of fuel-contaminated aquifer. Field
Demonstration. Ground Water 29:571-580.
Keck, J., R.C. Sims, M. Coover, K. Park, and B. Symons.
1989. Evidence for cooxidation of polynuclear aromatic
hydrocarbons in soil. Water Research 23:1467-1476.
Kieft, T.L., L.L Rosacker, D. Willcox, and A.J. Franklin. 1990.
Water potential and starvation stress in deep subsurface
microorganisms, pp. 4-99 - 4-112. In: Microbiology of the
Deep Subsurface, Proceedings, First International
Symposium (C.B. Fliermans and T.C. Hazen, eds). U.S.
Department of Energy and Westinghouse Savannah River
Company, Savannah, GA, January 15-19.
Kuznetsov, S.I., M.W. Ivanov, and N.N. Lyalikova. 1963.
Introduction to Geological Microbiology. (C.H. Oppenheimer,
ed.). McGraw-Hill, New York, NY.
Madsen, E.L., and J.-M. Bollag. 1989. Aerobic and anaerobic
microbial activity in deep subsurface sediments from the
Savannah River Plant. Geomicrobiology Journal 7:93-101.
Matthess, G. 1982. The Properties of Ground Water. Wiley,
New York, NY.
Metcalf & Eddy, Inc. 1991. Stabilization Technologies for
RCRA Corrective Actions. EPA/625/6-91/026, Center for
Environmental Research Information, U.S. Environmental
Protection Agency, Cincinnati, OH.
Miller, R.N. 1990. A Field Scale Investigation of Enhanced
Petroleum Hydrocarbon Biodegradation in the Vadose Zone
Combining Soil Venting as and Oxygen Source with Moisture
and Nutrient Addition. Ph.D. Dissertation. Department of Civil
and Environmental Engineering, Utah State University,
Logan, UT.
Miller, R.N., C.C. Vogel, and R.E. Hinchee. 1991. A field-
scale investigation of petroleum hydrocarbon biodegradation
in the vadose zone enhanced by soil venting at Tyndall AFB,
Florida, pp. 283-302. in: In Situ Bioreclamation: Applications
and Investigations for Hydrocarbon and Contaminated Site
Remediation (R.E. Hinchee and R.F. OHenbuttel, eds.).
Butterworth-Heinemann, Boston, MA.
Murdoch, L., B. Patterson, G. Losonsky, and W. Harrar.
1990. Technologies of Delivery and Recovety for the
Remediation of Hazardous Waste Sites. EPA/600/2-89/066,
Risk Reduction Engineering Laboratory, U.S. Environmental
Protection Agency, Cincinnati, OH.
Murdoch, L., G. Losonky, P. Cluxton, B. Patterson, I. Klich,
and B. Braswell. 1991. Feasibility of Hydraulic Fracturing of
Soil to Improve Remedial Actions. EPA/600/2-91/012, Risk
Reduction Engineering Laboratory, U.S. Environmental
Protection Agency, Cincinnati, OH.
Noble, D.G. 1963. Well points for dewatering groundwater.
Ground Water 1:21-36.
Paul, E.A., and F.E. Clark. 1989. Soil Microbiology and
Biochemistry. Academic Press, Inc., San Diego, CA.
Rittmann, B.E., and P.L. McCarty. 1980. Model of steady-
state film biofilm kinetics. Biotechnology and Bioengineering
22:23-43.
Rochkind, M.L., and J.W. Blackburn. 1986. Microbial
Decomposition of Chlorinated Aromatic Compounds. EPA/
600/2-86/090, Hazardous Waste Engineering Research
Laboratory, U.S. Environmental Protection Agency,
Cincinnati, OH.
Ross, D., T.P. Marziarz, and A.L. Bourquin. 1988.
Bioremediation of hazardous waste sites in the USA: Case
histories, pp. 395-397. In: Superfund '88. Proc., 9th National
Conference, Hazardous Materials Control Research Institute,
Silver Spring, MD.
Sims, R.C. 1990. Soil remediation techniques at uncontrolled
hazardous waste sites: A critical review. Journal of the Air &
Waste Management Association 40:704-732.
15
-------
Sims, J.L, R.C. Sims, and J.E. Matthews. 1989.
Bioremediation of Contaminated Surface Soils. EPA/600/9-
89/073, Robert S. Kerr Environmental Research Laboratory,
U.S. Environmental Protection Agency, Ada, OK.
Sims, R.C., O.L Sorensen, J.L Sims, J.E. McLean, R.
Mahmood, and R.R. Dupont. 1984. Review of In Place
Treatment Techniques for Contaminated Surface Soils.
Volume 2: Background Information for In Situ Treatment.
EPA/540/2-84-003b, Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency,
Cincinnati, OH.
Stevens, D.K., W.J. Grenney, and Z. Yan. 1988. User's
Manual: Vadose Zone Interactive Processes Model.
Department of Civil and Environmental Engineering, Utah
State University, Logan, UT.
Stevens, D.K., Grenney, W.J., Z. Yan, and R.C. Sims. 1989.
Sensitive Parameter Evaluation for a Vadose Zone Fate and
Transport Model. EPA/600/2-89/039, Robert S. Kerr
Environmental Research Laboratory, U.S. Environmental
Protection Agency, Ada, OK.
Symons, B.D., and R.C. Sims. 1988. Assessing detoxification
of a complex hazardous waste, using the Microtox bioassay.
Archives of Environmental Contamination and Toxicology
17:497-505.
Thomas, A.O., P.M. Johnston, and J.N. Lester. 1991. The
characterization of the subsurface at former gasworks sites in
respect of in situ microbiology, chemistry, and physical
structure. Hazardous Waste & Hazardous Materials 8: 341-
365.
Urtings, L.G.C.M., F. Spuy, S. Coffa, H.B.R.J. van Vree.
1991. Soil vapour extraction of hydrocarbons: In Situ and On-
Site Biological Treatment, pp. 321-336. In: In Situ
Bioreclamation: Applications and Investigations for
Hydrocarbon and Contaminated Site Remediation (R.E.
Hinchee and R.F. Olfenbuttel, eds.). Butterworth-Heinemann,
Boston, MA.
U.S. EPA. 1988. Interactive Simulation of the Fate of
Hazardous Chemicals during Land Treatment of Oily Wastes:
RITZ User's Guide. EPA/600/8-88-001, Robert S. Kerr
Environmental Research Laboratory, U.S. Environmental
Protection Agency, Ada, OK.
U.S. EPA. 1989a. State of Technology Review: Soil Vapor
Extraction Systems. EPA/600/2-89/024, Risk Reduction
Engineering Laboratory, U.S. Environmental Protection
Agency, Cincinnati, OH.
U.S. EPA. 1989b. The Superfund Innovative Technology
Evaluation Program: Technology Profiles. EPA/540/5-89-013,
Risk Reduction Engineering Laboratory, U.S. Environmental
Protection Agency, Cincinnati, OH.
U.S. EPA. 1990. Assessing UST Corrective Action
Technologies: Site Assessment and Selection of Unsaturated
Zone Treatment Technologies. EPA/600/2-90/011, Risk
Reduction Engineering Laboratory, U.S. Environmental
Protection Agency, Cincinnati, OH.
van Eyk, J. and C. Vreeken. 1991. In Situ and on-site subsoil
and aquifer restoration at a retail gasoline station, pp. 303-
320. In: In Situ Bioreclamation: Applications and
Investigations for Hydrocarbon and Contaminated Site
Remediation (R.E. Hinchee and R.F. Olfenbuttel, eds.).
Butterworth-Heinemann, Boston, MA.
Wobber, F.J. 1989. Deep Microbiology Transitional Program
Implementation Plan. DOE/ER-0431, Office of Energy
Research, Office of Health and Environmental Research,
U.S. Department of Energy, Washington, DC.
Zehnder, A.J.B., and W. Stumm. 1988. Geochemistry and
biochemistry of anaerobic habitats. In: Biology of Anaerobic
Microorganisms (A.J.B. Zehnder, ed.). John Wiley and Sons,
New York, NY.
16
-------
A One-Day Seminar On
Bioremediation Applications
July 27, 1993
Session VIII
Case Studies on Bioventing
Presented By
Gregory D. Sayles, Ph.D.
-------
Fundamentals of Bioventing Applied to Fuel
Contaminated Sites
Dr. R. Ryan Dupont
UWRL, UMC-8200, Utah State University, Logan, UT 84322-8200
Bioventing entails the use of soil vapor extraction (SVE) systems for the
transport of oxygen to the subsurface, where indigenous organisms are stimulated
to aerobically metabolize fuel components. Bioventing systems are designed and
configured to optimize oxygen transfer and oxygen utilization efficiency, and are
operated at much lower flow rates and with configurations much different than
those of conventional SVE systems. Bioventing system applications and design
are contrasted to those of conventional SVE systems, and the two key elements
of bioventing system design evaluation, i.e., in situ microbial activity and air
permeability determinations, are highlighted in this paper. The application of
bioventing to vadose zone bioremediation was reviewed with particular emphasis
on its advantages over aqueous based bioremediation systems in terms of its
superior oxygen transfer efficiency. Finally, the application of bioventing and
bioventing design concepts are illustrated through a case study of JP-4 jet fuel
contaminated soil remediation at Hill AFB, Utah.
INTRODUCTION
Conventional soil vacuum extraction (SVE) systems are de-
signed to optimize system performance to yield a maximum
recovery rate of volatiles from contaminated soil. Performance
may deteriorate over time, however, due to occluded residual
saturation and enrichment of residual contamination in the
less volatile waste components.
Bioventing has been successfully applied and documented
for the remediation of residual hydrocarbons remaining in soil
following high rate SVE. Bioventing entails the use of SVE
systems for the transport of oxygen to the subsurface, where
indigenous organisms are stimulated to aerobically metabolize
fuel components. Bioventing systems are designed and con-
figured to optimize oxygen transfer and oxygen utilization
efficiency, and are operated at much lower flow rates and with
configurations much different than those of conventional SVE
systems.
The bioventing processes is described in this paper, along
with details of a recommended design approach for full-scale
systems using field in situ respiration and air permeability
measurements. A case study of the performance of a bioventing
system at a JP-4 jet fuel site is briefly summarized, and the
implications of results from this study having a bearing on
future system design are highlighted.
BIOLOGICAL REMEDIATION OF CONTAMI-
NATED SOILS
The biodegradation of organic compounds in soil environ-
ments has been extensively described in the technical literature,
and details o!" metabolic pathways and microbial populations
responsible for compound biotransformation have been sum-
marized in a large number of textbooks and reviews on soil
microbial ecology [/, 2, J], For direct biodegradation of haz-
ardous organics to be successful, four conditions must be sat-
isfied. First, the contaminants of interest must be able to serve
as a carbon and energy source for the indigenous microbial
population, i.e.. it must be able to serve as an electron donor.
Secondly, an appropriate electron acceptor must be available
so that energy can be extracted from these electron donors at
environmentally significant rates. Thirdly, macro- and mi-
cronutrients essential for the production of cellular material
must be available in the appropriate ratio for microbial growth
to proceed unhindered. (A C:N:P mass ratio typical recom-
mended for soil bioremediation applications is 100:10:1). Fi-
nally, environmental conditions within the contaminated soil/
water environment must not be inhibitory to the indigenous
microflora. Soil environmental conditions of concern to ensure
effective bioremediation include: soil water at 50 to 80 "It of
soil field capacity*; 1/3 bar; soil pH from 5.5 to 8.5; soil
Environmental Proaress (Vol. 12. No. 1)
Fphrnarv 1QQ"> /IE
-------
Table 1 Carrier Fluid Oxygen Supply Requirements
Carrier Solution
Water
Air Saturated
Pure Oxygen Saturated
500mg/L H:0: (lOCfa Utilization)
Air (20.9
-------
Table 2 General Design and Application Considerations Appropriate tor Conventional Versus Bioventing
SVE Systems
Parameter
Conventional SVE
Bioventing
Compound Type
Volatile @ Room Temperature
Biodegradable
Vapor Pressure
> 100 mm Hg
—
He (dimensionless)
>0.01
—
Aqueous Solubility
< 100 mg/L
—
Soil Concentration
> 1 mg/kg.
< 1%
Depth to Ground Water
>20 ft
—
Air Phase Permeability
> 1 x 10"1 cm/s
Subsurface Conditions
Little or No Stratification
NAPL Phase
Little or None
Biodegradable
Vent Well Placement
Within Contamination
Outside Contamination
Operating Mode
Maximum Soil Gas
Maximum Retention Time
Exchange Rate
& Aerobic Conditions
Operating Flow Rates
46 to 700+ actual L/s
4.6 to 23 actual L/s
(100 to 1,500+ acfm)
(10 to 50 acfm)
Pore Volumes/d
1 to 15
0.1 to 0.5
Optimal Soil Moisture
= 25<7o Field Capacity
= 75% Field Capacity
Nutrient Requirement
—
C:N:P= 100:10:1
Soil Gas 0: Levels
—
> 2 vol%
Toxicants
—
Little or None
be accomplished fairly easily, and have been used to optimize
contaminant biodegradation at field sites when other variables,
i.e., toxicity, do not limit microbial activity [16, 17],
Oxygen transfer to the subsurface via SVE systems is gen-
erally more rapid than oxygen uptake rates observed under
field conditions [16, 17], This results in the oxygenation of
soil gas to near ambient levels if vent system blowers are op-
erated on a continuous basis. To minimize system operating
costs, and more importantly to reduce or even perhaps elim-
inate off-gas treatment requirements entirely, cyclic, or "surge"
pumping of vent systems in bioventing operations is recom-
mended. Surge pumping in a bioventing mode entails operating
the blower system until soil gas oxygen levels reach near am-
bient conditions throughout the site being remediated. The
system would then be shut off for some period of time during
which soil gas oxygen concentrations would be routinely mon-
itored until they reach a level which inhibits aerobic microbial
activity. Once this limiting soil gas concentration is reached,
the vent system would be restarted, and the on-off cycle would
continue once again. Based on a Henry's Law constant for
oxygen, this oxygen limitation would be expected to occur at
a soil gas concentration of approximately 2.0 voWo, corre-
sponding to soil water oxygen concentrations of approximately
1 mg/L. An inhibition of soil respiration has been reported at
the 2.0 vol% soil oxygen level in venting systems treating JP-
4 contaminated soils, [16] and in vented soil piles contaminated
with PCP waste \19] suggesting that this value represents a
good operating number for field scale applications.
Based on observed field respiration data from various JP-
4 jet fuel contaminated sites [20] and bioventing of PCP con-
taminated soil piles [19] field oxygen uptake rates of 0.1 to
0.6 vol^o/h (2.7 to 16 g 03/m3 soil-d @ air filled porosity =
40 vol®o) can be expected. These rates can be nearly an order
of magnitude lower as remediation progresses to near de min-
imus soil hydrocarbon levels (Dupont el at., 1991), allowing
typical bioventing systems to be operated on schedules of 8 h
on, 16 h off at the initation of remediation, to 8 h on, 7 d off
near the end of the field effort, while still maintaining aerobic
conditions within the contaminated soil during nonventing pe-
riods.
Table 2 presents a summary of general design, operational
and application considerations appropriate for conventional
SVE systems versus those utilized in a bioventing operating
mode.
Bioventing System Design
The two major design considerations for bioventing systems
are first, whether the contaminants of concern are biodegrad-
able under prevailing site conditions, i.e., whether inhibition
or toxicity is evident at the site, and secondly, whether the
required terminal electron acceptor, i.e., oxygen, can be ef-
fectively transported within the soil to encourage aerobic con-
taminant biodegradation. The first question can be answered
using soil gas composition and in situ respiration measure-
ments, while the second question is answered from in situ air
permeability measurements.
Soil Bioactivity Determinations
To determine the potential for in situ biodegradation of
vadose zone contaminants via bioventing, existing soil micro-
bial activity should be quantified during site assessment in-
vestigations. This can be readily accomplished through the
analysis of soil gas 0: and CO: composition prior to venting
activity at the site. 0= and CO: concentrations can be measured
along with volatile oreamcs during standard soil gas surveys
using a variety of measurement techniques. The author has
successfully used both »et chemical (Fyrite® oxygen/carbon
dioxide analyzer; Bacharach Instrument. Pittsburgh, PA) and
electronic (Gastechtor Model 32520X; Gastech Inc., Newark,
CA) methods for field soil gas 0: and CO: determinations.
While both gases can be easily measured, 02 concentrations
are considered a better indicator of respiration in soil systems
because there are no abiotic sinks for oxygen in these envi-
ronments. Carbon dioxide is produced through anaerobic as
well as aerobic microbial activity (2] and can also be affected
by assimilation or dissolution of carbonate rock.
The key to the evaluation of soil bioactivity using these
methods is the determination of the extent of oxygen depletion
and carbon dioxide enrichment in soil gas at a site with respect
to background, uncontaminated soil levels. It cannot be
overemphasized that these determinations must be based on a
comparison to uncontaminated soil conditions, as only levels
of O; depletion and CO: enrichment in excess of background
are indicative of increased microbial activity compared to nor-
mal basal respiration levels seen in uncontaminated soils at the
site.
Environmental Progress (Vol. 12, No. 1)
February, 1993 47
-------
If soil gas organic vapor and soil core data show contami-
nation, but microbial respiration has not yielded O, uptake
and C02 production rates above background soil levels, con-
ditions within the contaminated soil have resulted in soil mi-
crobial toxicity or severe inhibition, or significant nutrient or
moisture limitations exist at the site. Unless soil moisture is
the cause of this limitation, bioremediauon has limited appli-
cation at the site, and alternative remediation schemes should
be considered.
If soil contamination exists and microbial activity above
background levels is evident from soil gas measurements, quan-
tification of maximum respiration rates under field conditions
can be carried out utilizing in situ respiration measurement
techniques described by Hinchee et al. [20, 2l\. This method
entails the oxygenation of contaminated and uncontaminated
background subsurface soil around a soil gas probe via air
injection for a 16 to 24 hr period, followed by the measurement
of 0; uptake and CO; production at the soil gas probe over
time. The collected soil gas data are analyzed using either a
zero or first order reaction rate model to generate either zero
or first order respiration rate values (vol'fo/hr or 1/h. respec-
tively) from the slope of these linear regression relationships.
The background soil values are used to correct contaminated
soil values for basal soil respiration taking place at the site.
An inert gas tracer can be injected during soil aeration so that
respiration rate measurements can also be corrected for dif-
fusion of 0: away from, or C0; diffusion to the sampling
probe during respiration rate determinations.
Using these respiration data, in situ contaminant biodegra-
dation rates can be estimated assuming the 3:1 O::hydrocarbon
stoichiometry presented in Equation 1. In addition, required
oxygen transfer rates can be estimated, and the feasibility of
in situ bioventing, and the estimated time for remediation
under prevailing site conditions can be assessed.
In Situ Air Permeability Determinations
Once bioactivity at the site has been verified, the rate of
transfer of the electron acceptor to the contaminated soil re-
mains to be determined. This can be readily accomplished by
obtaining in situ air permeability measurements at several lo-
cations throughout the site. The approach that has become the
recommended standard for in situ soil air permeability meas-
urements was described in 1990 by Johnson et al. \22] and is
based on Darcy's Law and equations for steady-state radial
flow at a vent well. The method entails the use of a single vent
well with soil vapor probes placed radially and vertically away
from it to monitor soil gas pressure or vacuum throughout the
field site when air is extracted or injected at a constant rate at
the well head. A schematic of the instrumentation necessary
for a typical in situ permeability field study is shown in Figure
3 [22].
The governing equation for such a system assuming one-
V*»: ?]0W!ttt%r
* Pimm Sunptof Probes
FIGURE 3. Schematic of an In situ permeability field
study. From Johnson et aL [22]
dimensional radial flow from the extraction well is shown in
equation 2 \22]:
P' =-
UmV
-0.5772-In
rtfi
+ ln(0
(2)
where P' = "gauge" pressure (g/cm-s:) measured at the vapor
probes some radial distance r (cm) from the vent well at time
/(s), m = vent well screen interval (cm), k = soil gas permeability
(cm*), n = air viscosity (1.8 x 10"4 g/cm-s @ 18'C),«= soil air
filled porosity (decimal ^o), Q = volumetric air flow rate at the
vent well (em'/s), and Pim = atmospheric pressure (1
atm = 1.013 x 10" g/cm-s;).
Soil gas pressure or vacuum data collected over time at
various vapor probe locations following initiation of vent well
pumping allow the determination of in situ soil gas permeability
and its variability throughout the site. Vapor probe readings
are plotted as a function of the natural log of time, generating
a straight line with a slope equal to equation 3 [221:
Slope = -
4""U
(3)
Rearrangement of this equation allows the determination of
k directlv as:
* =
Qm
4Slopeirm
(4)
This approach to data reduction will not be possible if the
assumption of radial flow is not maintained at the field site.
Radial flow will not occur if a significant vertical air velocity
component exists due to shallow contamination and subse-
quently a small well screen interval (<10 ft), and if the soil is
coarse grained. Under these conditions, pressure/vacuum
measured in the vapor sampling points will reach constant
values very quickly, requiring that the data be reduced using
equations S and 6 [22]:
Qn In
for vacuum wells k = •
©
for extraction wells k = -
-["fe)']
(5)
(6)
where fl„ = the radius of the vent well (cm), H= the depth to
the top of the well screen (cm), /?, = the minimum radius of
vent well influence under steady-state flow conditions, and
Pw = the absolute pressure at the well head (g/cm-s2). Rt can
be estimated from inspection of field data, or by extrapolating
the relationship of vapor probe vacuum/pressure versus log(r)
to a 0 vacuum/pressure value.
Integration of Field Data
Bioventing system design can now be carried out by esti-
mating the equivalent daily oxygen demand and vent air flow
rate as determined from in situ respiration measurements, and
48 February, 1993
Environmental Progress (Vol. 12, No. 1)
-------
Oft-StU
Analytical Trader
Soil Pile
FIGURE 4. Conceptual diagram of the Hill AFB, Utah,
field venting site. From Dupont et al. [16]
by estimating pump operating conditions at these required air
flow rates based on field determined in situ air permeability
measurements. The feasibility of pulse pumping and vacuum
pump/blower scheduling can be assessed based on required
versus maximum oxygen transfer rates possible under a given
set of Held and pump/blower operational constraints. An ex-
ample of such calculations are presented below in the case
study for the bioventing system operated at Hill AFB, Utah.
Case Study—JP-4 Contaminated Site, Hill AFB, Utah
Site Description
The site at Hill AFB, Utah, was the location of a JP-4 jet
fuel spill that occurred in January 1985, after the failure of
an automatic shut-off valve. Failure of the valve resulted in
the release of approximately 102,000 L (27,000 gal) of JP-4,
some 7,600 L (2,000 gal) of which were recovered as free
product. The balance of the released fuel migrated away from
the tank and contaminated an area around it of approximately
0.4 hectares (1 ac) to a depth of approximately 15 m (50 ft).
The soil at the site consists of mixed coarse sand and gravel
deposits with interspersed, discontinuous clay stringers to a
:onfined ground water table at approximately 180 m (600 feet)
below ground surface (bgs). JP-4 contamination resulted in
ioil total pertroleum hydrocarbon (TPH) concentrations at the
iite as high as 15,000 mg/kg, with average TPH levels of
>1,000 mg/kg. Prior to initiation of the full scale venting
;vstem, the fuel tanks were excavated, refurbished, and in-
called in a concrete cradle above ground.
Acoreunuti extent of sods
>10,000 mg/kg JP-4.
immeoiat^'y foUomnQ ow spd
ttofea, bwn and GtmH I96S)
/!
N
Met en
IGURE 5. Site map showing vent well and pressure
lonitoring point locations at the Hill AFB, Utah, site.
From Hinchee et al. [18]
System Configuration and High Rate Venting Opera-
tions
A one-well pilot scale vent test was conducted by Oak Ridge
National Laboratory (ORNL) to evaluate in situ air permea-
bility at the site. This study resulted in the design by ORNL
of a venting system consisting of 15 vertical wells and 10 lateral
wells in the excavated soil pile and under the tanks (Figure 4).
The vertical wells were placed at 12 m (40 ft) intervals to a
depth of 15 m (50 ft) bgs and were slotted over an interval
from 3 to 15 m (10 to 50 ft) bgs. Twenty-one pressure mon-
itoring points (PMP) were installed at various depths through-
out the site to provide point measurements of subsurface
pressure and soil gas conditions (Figure 5), and a background
well was placed approximately 210 m (700 ft) north of the site
in the same geological unit and at the same depth as the vent
wells to provide a control for basal soil respiration levels during
the study.
Prior to initiation of bioventing studies at the Hill site, the
SVE system was operated under a conventional mode to max-
mize the recovery of volatile components of the JP-4 through
volatilization. Venting was initiated at a rate of 36 mVh (26
acfm, approximately 0.04 pore volumes/d), and gradually in-
creased to approximately 2,100 m3h/ (1,500 acfm, approxi-
mately 2.5 pore volumes/d) as the hydrocarbon levels in the
vent gas decreased over time. Vent gas was collected through
Wells V5 to VI1 (Figure 5), where the bulk of the soil con-
tamination was located. The venting rate during the start-up
period was limited by the operating conditions of the catalytic
incinerator used to treat the collected vent gas. This high-rate
operating mode continued from December 18, 1988, through
September 15, 1989, during which time approximately
7,000,000 m" (300,000.000 acf, approximately 340 pore vol-
umes) of soil gas and 62,600 kg (138,000 lb) TPH were extracted
from the site due to volatilization and in situ biodegradation
of the JP-4.
Three in situ respiration tests were conducted during the
high-rate operating period [/£] to assess microbial activity at
the site during conventional soil venting. These tests were con-
ducted at cumulative extracted air volumes of 970 m1 (42,000
acf), 13,000 m3 (540.000 acf), and 1,000,000 m3 (45,000,000
acf), and showed first order oxygen uptake rates ranging from
0.85/d to nondetectable in the pressure monitoring points
throughout the site. Comparison of results from specific mon-
itoring points over time also indicated the incremental removal
of residual hydrocarbons as respiration rates declined through-
out the treatment period. It was concluded that significant
respiration was occurring during conventional SVE without
nutrient or moisture addition, and that enhancement of biod-
egradation could be possible under modified site management
conditions. This became of increasing interest as the residual
soil TPH levels had not reached the regulatory action level of
30 mg/kg dry wt. soil, and the conventional SVE system hy-
drocarbon recovery rate decreased significantly over time due
to non-volatile residual contaminants accumulating in the soil
over time.
Modified Bioventing System Operating Conditions
Based on results of vapor probe and vent gas measurements
taken during the high-rate venting period, it- was found that
at high extraction rates, i.e., 2,100 mVh (1,500 acfm), the
entire contaminated zone was aerated to near atmospheric O?
levels. In addition, due to the extraction of vapors from the
areas of maximum contamination at the interior of the site,
hydrocarbon levels above the allowable discharge limit of 50
ppmv were found in the vent gas. To maximize biodegradation
and minimize volatilization, operating flow rates were reduced
to the lowest rates possible utilizing the existing venting system,
i.e., 490 to 970 mVh (350 to 700 acfm), and vent gas was drawn
nvironmental Progress (Vol. 12, No. 1)
February, 1993 49
-------
90
HO +
TO--
<50 +
'• odfCTBOilinn
1 ,^n Uami
Enhanced
Biodecrodfiion
Activiu«i B*cm
20 ••
lOf
Indication ei P»re*nio4'Toi*l
R#eo*#r»o JP 4 F>iod«r*M*d Lsme
0^ ijat Mcaturtmvnij >n m* Vrni
;> utT Cut
Hiph R«u Ertrmttjon
Low Kji# fcurartj
03 89 Ofl.M
Mertih of Op«ratioA
FIGURE 6. Percent recovered JP-4 attributed to bio-
degradation reactions at the Hill AFB, Utah, field soil
venting site based on oxygen depletion measured in the
SVE system vent gas.
from wells on the periphery of the site (Wells V12 to V15 in
Figure 5) to maximize the flow path and retention time of
vapors in the contaminated zone. Figure 6 presents the results
of this operating mode change in terms of the percent of total
JP-4 recovery that could be attributed to biodegradation, ex-
pressed on an oxygen consumption basis, during both the con-
ventional high-rate and modified bioventing phases of the
study. Biodegradation accounted for 15 to 20% of the re-
covered JP-4 even during high-rate venting. This rate was
drastically altered in September 1989, when JP-4 volatilization
was reduced from 90 to 180 kg/d (200 to 400 lb/d) to less than
9 kg/d (20 lb/d) by making the stated changes to the system
flow rate and extraction configuration. These changes allowed
direct discharge of vent gas without expensive off-gas catalytic
incineration treatment, and had no detrimental effect on biod-
egradation reactions. The hydrocarbon biodegradation rates
of 32 kg/d^70 lb/d) observed during high rate extraction were
maintained at an average rate of greater than 45 kg/d (100
lb/d) following system operating modifications.
In Situ Permeability Determinations
In situ permeability measurements were once again made
during the bioventing phase of this project using the vent well
and vapor probe configuration utilized during bioventing op-
erations. Results presented in Figure 7 were collected from the
indicated vent well and pressure monitoring points while ex-
tracting vapor from Vent Well 13 at an operating flow rate of
212 actual L/s (450 acfm).
Using the approach by Johnson et al., [22] a linear regression
through these data yielded Slope values as input to equation
3 as shown in Table 3. These data yielded a mean k value of
223 ±73 darcys, indicative of the clean sands and gravels pres-
ent at the site.
•
-------
Pressure Monitonnq Pom! Q
Time ih)
FIGURE 8. A sample first order regression analysis of
oxygen uptake rate data obtained during in situ respi-
ration measurements.
CO; and 0; concentrations normalized to the background well
st each sampling interval, i.e., ln(Cvent iround »ru)i:rrc i»
versus time. Each regression line was tested for the significance
of its slope, i.e., the probability of the slope not equalling zero
being >0.05. In addition, an evaluation of overlapping 95
percent confidence intervals of each regression slope was used
to test for significant differences among treatments.
Figure 8 shows results of a typical oxygen uptake rate de-
termination obtained during these in situ respiration studies.
0: uptake was found to be more consistent and more sensitive
than CO: production rates in detecting effects of treatments
on microbial activity at the sue. This was particularly true for
the moisture addition cases, where the interaction of CO: with
the added water greatly affected observed CO: production
rates.
Mean and maximum 0: uptake data are presented in Table
4 as a function of engineering management treatment along
with equivalent oxygen demand values expressed in kg/d. As
indicated in Table 4, the addition of moisture to the field site
yielded a significant increase in oxygen uptake rate not ob-
served with nutrient addition. Statistical results presented else-
where [16\ based on an analysis of overlapping 95 percent
confidence intervals of the slopes of significant regression re-
lationships for the three treatment cases during the bioventing
study indicated that in no case did nutrients significantly in-
crease respiration rates above statistically significant levels fol-
lowing moisture addition alone. C02 production rates were
not found to be significant at any monitoring point or vent
well following moisture addition indicating the sensitivity of
the C0: production measurement to changes in environmental
conditions that affect CO; distribution in the subsurface.
The 02 demand data presented in Table 4 were calculated
assuming the following: total contaminated soil vol-
Table 4 In Situ 02 Uptake Rate Data and Equivalent 02
Demand Requirements Collected from the Hill AFB,
Utah, Bioventing Field Site
Mean O-.
Mean O;
Max. 02
Max. 02
Management
Uptake
Demand
Uptake
Demand
Treatment
Rate (l/d)
(kg/d)
Rate (l/d)
(kg/d)
Low Rate
0.016
24.4
0.026
40.1
Venting
Moisture
0.030
45.3
0.168
261
Addition
Nutrient
0.016
24.4
0.060
87.1
& Moisture
Addition
Table 5 Potential 02 Transfer Rate Data and Predicted
Well Operating Vacuum at the Hill AFB, Utah, Bioventing
Field Site
Vent Well
Potential 0;
Operating
Transfer Rate
Vacuum
Q[L/s (acfm)|
(kg/d)
|g/cm-s: (in H.O)]
4.7 (10)
120
503 (0.20)
11.8 (25)
300
1,710 (0.69)
23.6 (50)
600
4,110 (1.6)
47.2 (100)
1,200
9,560 (3.8)
70.8 (150)
1,800
15.500 (6.2)
94.4 (200)
2.400
21,700 (8.7)
118 (250)
3,000
28,200 (11.3)
142 (300)
3,600
34,800 (14.0)
165 (350)
4,200
41,500 (16.6)
189 (400)
4,800
48,200 (19.4)
212 (450)
5,400
55,100 (22.1)
236 (500)
6,000
62,000 (24.9)
354 (750)
9,000
97.000 (38.9)
472 (1,000)
12,000
132,000 (53.0)
ume = 61,670 m' (2.178,000 ft3), air filled pore space = 0.4,
atmospheric O; content = 21 %, 0: density = 1.43 g/L. An anal-
ysis of the potential 0; transfer rate into the Hill site at various
flow rates is summarized in Table 5 using the assumptions
listed above for soil conditions; the mean in situ air permea-
bility value of 223 darcys determined for the site; and equation
5 with R, = 4,298 cm (141 ft) at 212 L/s (450 acfm) as deter-
mined from the extrapolation of vacuum versus radial distance
data from Table 3 to a 0 vacuum value (it was assumed that
R, varies linearly with flow rate in this coarse grained material),
Rw = 7.6 cm (3 in), H = 305 cm (10 ft), and Pum= 1.013x 10*
g/cm-s; (1 atm).
Upon comparison of Table 4 02 uptake estimates with Table
5 potential 0; transfer rates, it becomes apparent that only
very low flows, on the order of 12 L/s (25 acfm), are needed
to transfer the maximum uptake rate expected under optimized
engineering management conditions. At these low flow rates,
however, the radius of influence of the extraction wells is very
small, =244 cm (8 ft) at 12 L/s (25 acfm), limiting the effec-
tiveness of single wells to remediate large contaminated areas.
At the Hill AFB sue, the extent of contamination was roughly
6,400 cm (209 ft) square, precluding the use of this low flow
rate. What could have been implemented, however, was the
use of the multiple wells operated at higher flow rates in a
sequential fashion for short periods of time to supply the
oxygen need for microbial metabolism, while limiting the vol-
ume of contaminated air extracted from the soil. This was
done to some extent by reducing the vacuum flow rate to 212
actual L/s (450 acfm). However, rather than operating on a
continuous basis, all of the daily oxygen demand could have
been supplied to the site in little over an hour at this flow rate
(261 kg maximum demand from Table 4/5,500 kg/d supplied
from Table 5 = 0.048 d = 1.2 h to supply the demand). Because
of a radius of influence of =4,300 cm (140 ft) at 212 actual
L/s (450 acfm), the Hill system could have been operated with
two wells centered on opposite boundaries of the contaminated
area, running sequentially for approximately 0.75 h/d, for a
total operating time of 1.5 h/d. This scenario would have
supplied the entire contaminated zone with oxygen, and would
have produced the minimum volume of extracted soil gas for
treatment and/or disposal.
TPH Removal Performance
As indicated above, 65% contaminant recovery was accom-
plished by conventional high rate SVE during the first nine
Environmental Progress (Vol. 12, No. 1)
February 10QT si
-------
months of the study as 62,600 kg (138,000 lb) of TPH were
extracted from the site due to volatilization and m situ bio-
degradation. An additional 33,200 kg (73,100 lb) of JP-4 were
removed during bioventing between September 16, 1989, and
November 14, 1990, resulting in a total mass removal from
the site of 95,800 kg (211,100 lb) of JP-4 over the 23 month
operating period. Of this total, 53,650 kg (118,200 lb) of JP-
4 was attributed to volatilization as indicated by vent gas hy-
drocarbon concentrations, while the recovery of 42,150 kg
(92.900 lb) of JP-4 was attributed to biodegradation as meas-
ured from vent gas oxygen deficit determinations. This cor-
responds to a 56 to 44°7o volatilization to degradation ratio
during the entire operating period.
Final soil TPH concentrations were measured in more then
200 soil cores collected from within the area of contamination
to confirm the hydrocarbon recovery data determined from
vent gas measurements. These vent gas results were substan-
tiated as the average residual TPH soil concentration was less
than 5 mg/kg dry wt. soil at the end of the bioventing period.
This soil hydrocarbon level represents greater than 99.5970 over-
all contaminant removal and a residual TPH concentration
below those required for closure of the site.
SUMMARY AND CONCLUSIONS
The application of bioventing to vadose zone bioremediation
has been reviewed, and its advantages over aqueous based
bioremediation systems in terms of its superior oxygen transfer
ability has been highlighted. Bioventing system applications
and design were contrasted to those of conventional SVE sys-
tems, and the two key elements of bioventing system design
evaluation, i.e., in situ microbial activity and air permeability
determinations, were highlighted. Finally, the application of
bioventing and bioventing design concepts were illustrated
through a case study of JP-4 jet fuel contaminated soil re-
mediation at Hill AFB, Utah. Based on this review of bio-
venting fundamentals, and of the performance of a field-scale
bioventing system, the following conclusions can be made:
1. SVE systems can be utilized as highly efficient oxygen
transfer systems for vadose zone oxygenation. Vent system
oxygen transfer rates have been shown to be much higher
than in situ oxygen uptake rates at a number of field JP-
4 contaminated bioventing sites, providing an opportunity
for optimizing treatment through SVE system operational
modifications and venting rate controls.
2. Conventional SVE systems do differ significantly
from bioventing systems in their design orientation. Air
extraction rates are maximized for contaminant recovery
in SVE systems, while bioventing systems attempt to max-
imize vapor retention within the soil to encourage microbial
degradation of contaminant vapors.
3. Methods to reduce vapor extraction rates to maxi-
mize vapor retention times in the soil are compatible with
enhancing biodegradation reactions. These procedures re-
sult in minimizing volatilization, potentially eliminate the
need for vent gas treatment, maximize the utilization of
oxygen in situ, and provide a framework for the develop-
ment of truly optimized in situ biological treatment systems.
At the Hill AFB site, reduced flow rates and maximized
flow path distances allowed the direct discharge of vent gas
without off-gas treatment, while still being below the reg-
ulatory limit of 50 ppmv TPH.
4. For bioventing systems to be successful, contami-
nants of interest must be biodegradable under field con-
ditions at rates that can be effectively exploited. Methods
described by Hinchee et at. \20, 21] for in situ respiration
rate determinations should be utilized to quantify the pres-
ence and rate of bioactivity prior to field scale system design.
5. Methods presented by Johnson et at. [22] allow the
determination of in situ air permeability from the collection
of simple field data. These methods allow the quantitation
of air permeability and its variability under actual field
conditions, and are recommended for SVE and bioventing
field system design.
6. Conclusive evidence was provided to indicate signif-
icant biological activity at the Hill AFB. Utah, field site.
Without enhancement, a total of 15 to 20 percent of the
recovered J P-4 could be attributed to biodegradation. With
enhancemeni this proportion increased to greater than 80
percent, resulting in 33,200 kg (73,100 lb) of TPH being
biodegraded during the 14 month bioventing portion of the
study.
7. Nutrient addition at field JP-4 bioventing sites has
consistently been shown to be ineffective in stimulating
microbial respiration rates, suggesting that nutrient avail-
ability is not rate limiting under field conditions at these
sites. However, moisture addition to the 30 to 50°7o field
capacity level appears to be essential in order to optimize
microbial activity within a bioventing treatment system.
8. Operation of bioventing systems for the remediation
of JP-4 jet fuel contaminated sites appears to be optimal
for biodegradation, i.e., maximum biodegradation/mini-
mum volatilization, at 0.25 to 0.5 pore volume/d. At this
operating flow rate, soil gas retention time is sufficient to
yield 80 to 85^o hydrocarbon recovery as respiration prod-
uct gas (CO:), while minimizing TPH recovery in the form
of VOC emissions.
9. In situ field respiration studies indicated that 0: up-
take raie measurements were better indicators of biological
activity at the site than were C03 production rate deter-
minations. CO; measurement sensitivity was susceptible to
varying soil environmental conditions, notably soil water
content. Soil gas CO: measurements did not consistently
detect respiration changes during the study.
10. Quantification of in situ respiration rates and ox-
ygen transfer potential indicated that daily oxygen demand
was being satisfied in slightly over 1 hr at the lowest rate
at which the Hill AFB bioventing system could be operated,
i.e., 212 actual L/s (450 acfm). Oxygen demand could have
been satisfied at flow rates much lower than this value, but
concerns over limited radii of influence at low extraction
rates suggest operating at higher flow rates for short time
periods during remediation. Optimal bioventing system de-
sign for the Hill AFB site was suggested to be two vent
wells operating at 212 actual L/s (4S0 acfm) for 0.75 hr/d
each. This results in sufficient oxygen transfer and ensures
coverage of the entire area of contamination, while signif-
icantly reducing the volume of extracted air that must be
handled prior to discharge.
LITERATURE CITED
1. Alexander, M., Introduction to Soil Microbiology, John
Wiley and Sons, Inc., New York, NY, pp. 467 (1977).
2. Atlas, R. M.. "Microbial Degradation of Petroleum Hy-
drocarbons: an Environmental Perspective," Micro. Rev.,
45(1): 185-209 (1981).
3. Dragun, J., "Microbial Degradation of Petroleum Prod-
ucts in Soils," in Soils Contaminated by Petroleum-En-
vironmental and Public Health Effects, E. J. Calabrese
and P. T. Kostecki, Ed. John Wiley and Sons, Inc., New
York, NY, pp. 289-300 (1988).
4. Wetzel, R. S., C. M. Darst, et al.. In Situ Biological
Treatment Test at Kelly Air Force Base, Volume II—Field
Test Results and Cost Model, Final Report TR-85-52,
Headquarters Air Force Engineering Services Center, Tyn-
dall Air Force Base, FL, 1987.
5. Downey, D. C., R. E. Hinchee, et al., "Combined Bio-
logical and Physical Treatment of a Jet Fuel-Contaminated
52 February, 1993
Environmental Progress (Vol. 12, No. 1)
-------
Aquifier," Proceedings of the Petroleum Hydrocarbons
and Organic Chemicals in Ground Water: Prevention De-
tection and Restoration, Dublin, OH, pp. 627-645 (1988).
6. Hinchee, R. E., and D. C. Downey, "The Role of Hy-
drogen Peroxide in Enhanced Bioreclamation," Proceed-
ings of the Petroleum Hydrocarbons and Organic
Chemicals in Ground Water: Prevention Detection and
Restoration, Dublin, OH, pp. 715-722 (1988).
7. Lee, \1. D., J. M. Thomas, .et al., "Biorestoration of
Aquifers Contaminated with Organic Compounds," CRC
Crir. Rev. Env. Control 18(1): 29-89 (1988).
8. Hinchee, R. E., D. C. Downey, et al.. Enhanced Biore-
clamation of Jet Fuels—A Full-Scale Test at Eglin AFB,
FL, Final Report ESL-TR-88-78, Headquarters Air Force
Engineering Services Center, Tyndall Air Force Base. FL,
(1988).
9. Wilson. J. T., and C. H. Ward. "Opportunities tor Bio-
remediation of Aquifers Contaminated with Petroleum
Hydrocarbons," J. Ind. Micro., 27: 109-116 (1988).
10. Bennedsen, M. B., "Vacuum VOCs from Soil," Poll.
Eng., 19(2): 66-68 (1987).
11. Riser, E., Technology Review—In Situ/On-Sne Bio-
degradation of Refined Oils and Fuels, N68305-6317-7115,
Naval Civil Engineering Laboratory, (1988).
12. Ely, D. L., and D. A. Heffner, Process for In Situ Bio-
degradanon of Hydrocarbon Contaminated Soil, Patent
No. 4.765.902, U.S. Patent Office (1988).
13. Ostendorf, D. W., and D. H. Kampbell. "Vertical Profiles
and Near Surface Traps for Field Measurement of Volatile
Pollution in the Subsurface Environment," Proceedings
of the New Field Techniques for Quantifying the Physical
and Chemical Properties of Heterogeneous Aquifers, Dal-
las, TX, (1988).
14. Stapps. J. J. M., International Evaluation of In Situ Bio-
restoration of Contaminated Soil and Groundwater,
738708006, National Institute of Public Health and En-
vironmental Protection (RIVM), (1988).
15. Hinchee, R. E., D. C. Downey, et al.. Enhanced Bio-
degradation Through Soil Venting, SSPT 88-427, USAF,
HQ AFESC/RD, (1989).
16. Dupont, R. R., W. J. Doucette, et al., "Assessment of
In Situ Bioremediation Potential and the Application of
Bioventing at a Fuels-Contaminated Site," in In Situ Bio-
reclamation: Applications and Investigations for Hydro-
carbon and Contaminated Sue Remdiation, R. E. Hinchee
and R. F. Olfenbuttel, Ed. Butterworth-Heinemann, Bos-
ton, Massachusetts, pp. 262-282 (1991).
17. Miller, R. N.. C. C. Vogel, et al., "A Field-Scale Inves-
tigation of Petroleum Biodegradation in the Vadose Zone
Enhanced by Soil Venting at Tyndall AFB, Florida," in
In Situ Bioreclamation: Applications and Investigations
for Hvdrocaroon and Contaminated Site Remediation.
R. E. Hinchee and R. F. Olfenbuttel, Ed. Butterworth-
Heinemann. Boston, pp. 283-302 (1991).
18. Hinchee. R. E.. D. C. Downey, et al., "Enhancing Bio-
degradation of Petroleum Hydrocarbons Through Soil
Venting," J. Ha:„ Mat., 27: 315-325 (1991).
19. McGinnis. D.. R. R. Dupont, et al., "Determination of
Respiration Rates in Soil Piles to Evaluate Aeration Ef-
ficiency and Biological Activity," Proceedings of the 85th
Annual Meeting and Exhibition of the Air and Waste
Management Association, Kansas City, Missouri, Reprint
#92-13.05. pp. II (1992).
20. Hinchee, R. E., Ong, S. K. et al., "A Field Treatability
Test for Bioventing," Proceedings of the 84th Annual
Meeting and Exhibition of the Air and Waste Management
Association. Vancouver, British Columbia, Reprint #91-
19.4, pp. 13 (1991).
21. Hinchee. R. E.. S. K. Ong, et al., Test Plan and Technical
Protocol for a Field Treatability Test for Bioventing, 80
pp, Prepared for the U.S. Air Force Center of Excellence.
Brooks Air Force Base, Texas, (1992).
22. Johnson, P. C., Kemblowski. M. W., et al., "Quaniitati\e
Analysis for the Cleanup of Hydrocarbon-Contaminated
Soils by In-Situ Soil Venting," Ground Water 28(3): May-
June (1990).
Environmental Progress (Vol. 12. No. 1)
February 1993 ^
-------
REPORT
TEST PLAN AND
TECHNICAL PROTOCOL
FOR A FIELD
TREATABILITY TEST
FOR BIO VENTING
For the
U.S. Air Force
Center for Environmental Excellence
OBattelle
. . . Putting Technology To Work
May 1992
-------
Revision 2
TEST PLAN AND TECHNICAL PROTOCOL
FOR
A FIELD TREATABILITY TEST FOR BIOVENTING
by
R. E. Hinchee and S. K. Ong
Battelle
Columbus, Ohio
R. N. Miller
U.S. Air Force
Center for Environmental Excellence
Brooks Air Force Base, Texas
D. C. Downey and R. Frandt
Engineering-Sciences, Inc.
Denver, Colorado
for
U.S. Air Force
Center for Environmental Excellence
Brooks Air Force Base, Texas
May 1992
-------
This report is a work prepared for the United States Government by
Battelle. In no event shall either the United States Government or
Battelle have any responsibility or liability for any consequences of
any use, misuse, inability to use, or reliance upon the information
contained herein, nor does either warrant or otherwise represent in
any way the accuracy, adequacy, efficacy, or applicability of the
contents hereof.
-------
TABLE OF CONTENTS
Page
1.0 TEST OBJECTIVES I
1.1 Conduct Air Permeability and In Situ Respiration Tests 1
1.2 Conduct Bioventing Test 1
1.3 Use of Existing Wells and Monitoring Points 1
2.0 INTRODUCTION TO BIOVENTING AND FIELD TREATABILITY TESTS 2
2.1 Bioventing Background 2
2.1.1 Conventional Enhanced Biodegradation 2
2.1.2 Bioventing 3
2.1.3 Applications 5
2.1.4 Hill AFB Site 11
2.1.5 Tyndall AFB Site 14
2.2 Soil Gas Permeability and Radius of Influence 16
2.3 In Situ Respiration Testing 18
3.0 IN SITU RESPIRATION/AIR PERMEABILITY TEST PREPARATION 24
3.1 Site Characterization Review 24
3.2 Development of Site-Specific Test Plan 24
3.3 Application for Required Permits 26
4.0 TEST WELLS AND EQUIPMENT 27
4.1 Vent Wells 27
4.2 Soil Gas Monitoring Points 28
4.2.1 Location of Monitoring Points 30
4.2.2 Depth of Monitoring Points 30
4.2.3 Construction of Monitoring Points 31
4.2.4 Thermocouples 34
4.3 Background Well 34
4.4 Blower System 35
4.5 Field Instrumentation and Measurements 38
4.5.1 Oxygen and Carbon Dioxide 38
4.5.2 Hydrocarbon Concentration 38
4.5.3 Helium Monitoring 40
4.5.4 Temperature Monitoring 40
4.5.5 Pressure/Vacuum Monitoring 40
4.5.6 Airflow 40
4.5.6.1 Airflow Measurement - Air Permeability Test 41
4.5.6.2 Airflow Measurement — Respiration Test 41
4.5.6.3 Airflow Measurement — Bioventing Test 41
-------
5.0 TEST PROCEDURES 42
5.1 Location of Optimum Test Volume 42
5.1.1 Soil Gas Survey (for contamination < 20 ft) 42
5.1.2 Exploratory Boring in Deep Soils 43
5.2 Drilling and Installation of the Vent Well 45
5.3 Drilling and Installation of Monitoring Points 45
5.4 Background Well Installation 45
5.5 Collection of Soil Samples 45
5.6 Soil Gas Permeability Test Procedures 47
5.6.1 System Check 47
5.6.2 Soil Gas Permeability Test 47
5.6.3 Post-Permeability Test Soil Gas Monitoring 49
5.7 In Situ Respiration Test 49
5.7.1 Test Implementation 49
5.7.2 Data Interpretation 50
5.7.2.1 Oxygen Utilization 50
5.7.2.2 Helium Monitoring 54
5.8 Bioventing Test 54
5.8.1 Criteria for Conducting the Bioventing Test 57
5.8.1.1 Air Permeability/Radius of Influence 57
5.8.1.2 Biodegradation Rate 57
5.8.1.3 Regulatory Approval 58
5.8.1.4 U.S. Air Force Approval 58
5.8.2 Air Injection vs. Extraction Considerations 58
5.8.3 Blower System Installation 58
5.8.4 Blower Operation and Maintenance 59
5.8.5 Long-Term Monitoring 59
6.0 SCHEDULE 60
7.0 REPORTING 62
7.1 Test Plan 63
7.2 Monthly Reports 63
7.3 Verbal Communication 63
7.4 Site Repoits 63
8.0 RECORD OF DATA AND QUALITY ASSURANCE 64
9.0 REFERENCES 68
APPENDIX RECOMMENDED ESTIMATION METHODS FOR AIR PERMEABILITY 71
-------
TABLE OF CONTENTS
(Continued) Page
LIST OF TABLES
Table 2-1. Soil Gas Permeability Values 18
Table 2-2. Summary of Reported In Situ Respiration and Bioventing
Rate Data 22
Table 4-1. Recommended Spacing for Monitoring Points 31
Table 4-2. Monitoring Points for Example Site #2 at Millersworth AFB 34
Table 5-1. Parameters to Be Measured for the In Situ Respiration Tests 51
Table 5-2. Sample Data Set for Two In Situ Respiration Tests 52
Table A-l. Air Permeability Data Set 75
Table A-2. Field Test Data for Soil Determination of Soil Permeability
at a Gasoline-Contaminated Site 78
LIST OF FIGURES
Figure 2-1. Conceptual Layout of Bioventing Process with Air Injection Only 7
Figure 2-2. Conceptual Layout of Bioventing Process with Air Withdrawn
from Clean Soil 8
Figure 2-3. Conceptual Layout of Bioventing Process with Soil Gas Reinjection 9
Figure 2-4. Conceptual Layout of Bioventing Process with Air Injection into
Contaminated Soil. Coupled with Dewatering and Nutrient Application 10
Figure 2-5. Cumulative Hydrocarbon Removal from the Hill AFB Building 914
Soil Venting Site 12
Figure 2-6. Results of Soil Analysis at Hill AFB Before and After Venting . . .: 13
Figure 2-7. Results of Soil Analysis from Plot V2 at Tyndall AFB Before and
After Venting 15
Figure 2-8. Cumulative Percent Hydrocarbon Removal at Tyndall AFB
for Sites V1 and V2 17
Figure 2-9. Gas Injection/Soil Gas Sampling Monitoring Point Used by Hinchee et al.
(1991) in Their In Situ Respiration Studies 20
Figure 2-10. Average Oxygen Utilization Rates Measured at Four Test Sites 21
Figure 3-1. Flow Chan for Conducting Bioventing Treatability TesL 25
Figure 4-1. Typical Injection/Vacuum Venting Well Construction 29
Figure 4-2. Typical Monitoring Point Construction Detail 33
Figure 4-3. Soil Gas Permeability Instrumentation Diagram for Soil Gas Extraction 36
Figure 4-4. Soil Gas Permeability Blower System Instrumentation Diagram
for Air Injection 37
Figure 4-5. Schematic Setup for Calibration of Soil Gas Instruments:
(a) CO^ 02. and Total Hydrocarbon Analyzers
(b) Helium Detector 39
iii
-------
TABLE OF CONTENTS
(Continued) Page
Figure 5-1. Schematic Diagram of Soil Gas Sampling Using
the Stainless Steel Soil Gas Probe 44
Figure 5-2. In Situ Respiration Test Results for Two Bioventing Test Sites:
Fallon NAS. Nevada (Monitoring Point A2) and
Kenai. Alaska (Monitoring Point Kl) 53
Figure 5-3. In Situ Respiration Test Results for Monitoring Point SI. Tinker AFB.
Oklahoma. 55
Figure 5-4. In Situ Respiration Test Results for Monitoring Point K3. Kenai. Alaska .... 56
Figure 8-1. Typical Record Sheet for In Situ Respiration Test 65
Figure 8-2. Typical Record Sheet for Air Permeability Test 66
Figure 8-3. Typical Record Sheet for Long-Term Bioventing Test 67
Figure A-l. Vacuum vs. In Time, Test 2. Bioventing Pilot Test, Site 22-A20,
Beale Are, California 76
Figure A-2. Results of a Field Test to Determine Soil Permeability to Airflow, k.
September 16. 1991 79
-------
Revision 2
Page: 1
May 14. 1992
TEST PLAN AND TECHNICAL PROTOCOL
FOR
A FIELD TREATABILITY TEST FOR BIO VENTING
1.0 TEST OBJECTIVES
This test plan and technical protocol describes the methods for conducting a field
treatability test for the bioventing technology. The purpose of these field test methods is to
measure the soil gas permeability and microbial activity at a contaminated site and to evaluate
the potential application of the bioventing technology to remediate the contaminated site. The
specific test objectives are stated below.
1.1 Conduct Air Permeability and In Situ Respiration Tests
At every site, the air permeability of the soil and the air vent (well) radius of
influence will be determined. This will require air to be withdrawn or injected for approxi-
mately 8 hours at vent wells located in contaminated soils. Pressure changes will be
monitored in an array of monitoring points. Immediately following this test, an in situ
respiration test will be conducted. Air will be injected into selected monitoring points to
aerate the soils. The in situ oxygen utilization and carbon dioxide production rates will be
measured.
1.2 Conduct Bioventing Test
Using the data from the soil air permeability and in situ respiration tests, an air
injection/withdrawal rate will be determined for use in the bioventing test. A blower will be
selected, installed, and operated for 6 to 12 months, and periodic measurements of the soil
gas composition will be made, to evaluate the long-term effectiveness of bioventing.
1.3 Use of Existing Wells and Monitoring Points
The U.S. Air Force has already installed monitoring points or other wells at many
sites that will be suitable for use in this study. In keeping with the objective of developing a
cost-effective program for site remediation, every effort will be made to use existing wells
and minimize drilling costs.
-------
Revision 2
Page: 2
May 14, 1992
2.0 INTRODUCTION TO BIOVENTING AND FIELD TREATABILITY TESTS
Bioventing is the process of aerating subsurface soils to stimulate in situ
biological activity and promote bioremediation. Although it is related to the process of soil
venting (aka soil vacuum extraction, soil gas extraction, and in situ soil stripping), their
primary objectives are differenL Soil venting is designed and operated to maximize the
volatilization of low-molecular-weight compounds, with some biodegradation occurring. In
contrast, bioventing is designed to maximize biodegradation of aerobically biodegradable
compounds, regardless of their molecular weight, with some volatilization occurring. The
major difference between these technologies is that the objective of soil venting is volatiliza-
tion, and the objective of bioventing is biodegradation. Although both technologies involve
venting of air through the subsurface, the differences in objectives result in different design
and operation of the remedial systems.
2.1 Bioventing Background
Petroleum distillate hydrocarbons such as JP-4 jet fuel are generally biodegrad-
able if the naturally occurring microorganisms that acclimate to the fuels as a carbon source
are provided an adequate supply of oxygen and basic nutrients (Atlas, 1986). Natural
biodegradation does occur, and at many sites microorganisms may eventually mineralize most
of the fuel contamination. However, the process is dependent on natural oxygen diffusion
rates (Ostendorf and Kambell, 1989). As a result natural biodegradation is frequently too
slow to prevent the spread of contamination and sites may require remediation to protect
sensitive aquifers. Acceleration or enhancement of the natural biodegradation process may
prove to be the most cost-effective remediation for hydrocarbon-contaminated sites.
Understanding the distribution of contaminants is important to any in situ
remediation process. Much of the hydrocarbon residue at a fuel-contaminated site is found in
the unsaturated zone soils, in the capillary fringe, and immediately below the water table.
Seasonal water table fluctuations typically spread residues in the area immediately above and
below the water table. Any successful bioremediation effort must treat these areas. Biovent-
ing provides oxygen to unsaturated zone soils and can be extended below the water table
when integrated with a dewatering system.
2.1.1 Conventional Enhanced Biodegradation
The practice of enhanced biodegradation for treating soluble fuel components in
groundwater has increased over the past two decades (Lee et al., 1988), with less emphasis
given to enhancing biodegradation in the unsaturated zone. Currently, conventional enhanced
bioreclamation processes use water to carry oxygen or an alternative electron acceptor to the
contaminated zone. This is common whether the contamination is present in the groundwater
or in the unsaturated zone.
-------
Revision 2
Page: 3
May 14, 1992
A recent field experiment at a jet fuel-contaminated site used infiltration galleries
and spray irrigation to introduce oxygen (as hydrogen peroxide), nitrogen, and phosphorus to
unsaturated, sandy soils. The experiment was unsuccessful because the rapid decomposition
of hydrogen peroxide resulted in poor oxygen distribution (Hinchee et al., 1989).
Other attempts have been made using pure oxygen or hydrogen peroxide as
oxygen sources, and recently nitrate has been added as an alternative to oxygen. Although
results indicate better hydrogen peroxide stability than achieved by Hinchee et al. (1989), it
was concluded that most of the hydrogen peroxide decomposed rapidly (Huling et al., 1990).
Some degradation of aromatic hydrocarbons appears to have occurred; however, no change in
total hydrocarbon contamination levels was detected in the soils (Ward, 1988).
In most cases where water is used as the oxygen carrier, the solubility of oxygen
is the limiting factor for biodegradation. If pure oxygen is used and 40 mg/1 of dissolved
oxygen is achieved, approximately 80,000 lb of water must be delivered to the formation to
degrade 1 lb of hydrocarbon. If 500 mg/1 of hydrogen peroxide is successfully delivered,
then approximately 13,000 lb of water must be used to degrade the same amount of hydrocar-
bon. As a result, even if hydrogen peroxide can be successfully used, substantial volumes of
water must be pumped through the contaminated formation to deliver sufficient oxygen.
2.1.2 Bioventing
A system engineered to increase the microbial biodegradation of fuel hydrocar-
bons in the unsaturated zone using forced air as the oxygen source may be a cost-effective
alternative to conventional systems. This process provides oxygen to indigenous soil
microorganisms promoting aerobic metabolism of fuel hydrocarbons in unsaturated soils.
Depending on airflow rates, some volatile compounds may be simultaneously stripped from
contaminated soils.
When air is used as an oxygen source, 13 lb of air must be delivered to provide
the minimum oxygen required to degrade 1 lb of hydrocarbon, compared to the more than
13,000 lb of water with 500 mg/1 of hydrogen peroxide that must be delivered by conven-
tional water phase-enhanced bioreclamation processes. An additional advantage of using a
gas phase process is that gases have greater diffusivity than liquids. At many sites, geological
heterogeneities cause fluid that is pumped through the formation to be channeled into the
more permeable pathways (e.g., in an alluvial soil with interbedded sand and clay, all of the
fluid flow initially takes place in the sand). As a result, oxygen must be delivered to the less
permeable clay lenses through diffusion. In a gaseous system (as found in unsaturated soils),
this diffusion can be expected to take place at rates several orders of magnitude greater than
rates in a liquid system (as is found in saturated soils). Although it is not realistic to expect
diffusion to aid significantly in water-based bioreclamation, diffusion of oxygen in a gas
phase system may be a significant mechanism for oxygen delivery to less permeable zones.
-------
Revision 2
Page: 4
May 14, 1992
To the authors' knowledge, the first documented evidence of unsaturated zone
biodegradation resulting from forced aeration was reported by the Texas Research Institute,
Inc., in a study for the American Petroleum Institute. A large-scale model experiment was
conducted to test the effectiveness of a surfactant treatment to enhance the recovery of spilled
gasoline. The experiment accounted for only 8 gal of the 65 gal originally spilled and raised
questions about the fate of the gasoline. Subsequently, a column study was conducted to
determine a diffusion coefficient for soil venting. This column study evolved into a
biodegradation study in which it was concluded that as much as 38% of the fuel hydrocarbon
was biologically mineralized. Researchers concluded that venting would not only remove
gasoline by physical means, but also could enhance microbial activity and promote biodegra-
dation of the gasoline (Texas Research Institute, 1980; 1984).
To the authors' knowledge, the first actual field-scale bioventing experiments
were conducted by van Eyk for Shell Oil. In 1982 at van Eyk's direction, Delft Geotechnics
in The Netherlands initiated a scries of experiments to investigate the effectiveness of
bioventing for treating hydrocarbon-contaminated soils. These studies are reported in a series
of papers (Anonymous, 1986; Staatsuitgeverij, 1986; van Eyk and Vreeken, 1988, 1989a and
1989b).
Wilson and Ward (1986) suggested that using air as a carrier for oxygen could be
1,000 times more efficient than using water, especially in deep, hard-to-flood unsaturated
zones. They made the connection between soil venting and biodegradation by observing that
"soil venting uses the same principle to remove volatile components of the hydrocarbon." In
a general overview of the soil venting process, Bennedsen et al. (1987) concluded that soil
venting provides large quantities of oxygen to the unsaturated zone, possibly stimulating
aerobic degradation. They suggested that water and nutrients would also be required for
significant degradation and encouraged additional investigation into this area.
Biodegradation enhanced by soil venting has been observed at several field sites.
Investigators claim that at a soil venting site for remediation of gasoline-contaminated soil
significant biodegradation occurred (measured by a temperature rise) when air was supplied.
Investigators pumped pulses of air through a pile of excavated soil and observed a consistent
rise in temperature, which they attributed to biodegradation. They claimed that the pile was
cleaned up during the summer primarily by biodegradation (Conner, 1988). However, they
did not control for natural volatilization from the aboveground pile, and not enough data were
published to critically review their biodegradation claim.
Researchers at Traverse City, Michigan, observed a decrease in the toluene
concentration in unsaturated zone soil gas, which they measured as an indicator of fuel
contamination in the unsaturated zone. They assumed that advection had not occurred and
attributed the toluene loss to biodegradation. The investigators concluded that because
toluene concentrations decayed near the oxygenated ground surface, soil venting is an
attractive remediation alternative for biodegrading light volatile hydrocarbon spills (Ostendorf
and Kambell, 1989).
-------
Revision 2
Page: 5
May 14, 1992
The U.S. Air Force initiated its research and development (R&D) program in
bioventing in 1988 with a study at Hill Air Force Base (AFB) in Utah. During this study it
became apparent that bioventing had great potential for remediating JP-4 fuel-contaminated
soils. It was also apparent that additional research would be needed before the technology
could be routinely applied in the field. The work was initially supported by the U.S. Air
Force Civil Engineering Support Agency (AFCESA), previously known as the Air Force
Engineering and Services Center. Subsequently, they were joined in R&D support of the
technology by the U.S. Air Force Center for Environmental Excellence (AFCEE) and later by
Hill and Eielson AFBs. Following the Hill AFB study, a more controlled bioventing study
was completed at Tyndall AFB in Florida.
The Air Force currently supports a number of field programs to further test and
demonstrate the technology. After completion of the initial site testing at Hill AFB, a low-
intensity bioreclamation research program at another site was initiated in late 1989. At
Eielson AFB near Fairbanks, Alaska, a field demonstration of bioventing in a subarctic
environment was initiated in the summer of 1991. This study includes a soil heating
experiment to attempt to increase biodegradation rates.
The U.S. EPA Risk Reduction Engineering Laboratory (RREL) has become
interested in the Air Force's program, and has jointly funded and technically supported the
work at both Hill and Eielson AFBs. Additionally, the AFCESA is supporting a well-
documented bioventing demonstration at a cold weather site with field work scheduled to
begin in the summer of 1992.
2.1.3 Applications
The use of an air-based oxygen supply for enhancing biodegradation relies on
airflow through hydrocarbon-contaminated soils at rates and configurations that will
(1) ensure adequate oxygenation.for aerobic biodegradation, and (2) minimize or eliminate the
production of a hydrocarbon-contaminated off-gas. The addition of nutrients and moisture
may be desirable to increase biodegradation rates; however, field research to date does not
indicate the need for these additions (Dupont et al., 1991; Miller et al., 1991). If found
necessary, nutrient and moisture addition could take any of a variety of configurations.
Dewatering may at times be necessary, depending on the distribution of contaminants relative
to the water table. A key feature of bioventing is the use of narrowly screened soil gas
monitoring points to sample gas in short vertical sections of the soil. These points are
required to monitor local oxygen concentrations, because oxygen levels in the vent well are
not representative of local conditions.
A conventional soil venting system could be installed to draw air from a vent
well in the area of greatest contamination. This configuration would allow straightforward
monitoring of the off-gases. However, its disadvantage is that hydrocarbon off-gas concentra-
tion would probably be maximized, and could require permitting and treatment Furthermore,
all of the capillary fringe contamination may not be treated.
-------
Revision 2
Page: 6
May 14, 1992
Figure 2-1 is a schematic representation of a bioventing system that involves air
injccuon only. Although this is the lowest cost configuration, careful consideration must be
given to the fate of injected air. The objective is for most, if not all, of the hydrocarbons to
be degraded, and for C02 to be emitted at some distance from the injection point. If a
building or subsurface structure were to exist within the radius of influence of the well,
hydrocarbon vapors might be forced into that structure. Thus, protection of subsurface
structures may be required.
Figure 2-2 is an illustration of a configuration in which air is injected (the
injection may also be by passive well) into the contaminated zone and withdrawn from clean
soils. This configuration allows the more volatile hydrocarbons to degrade prior to being
withdrawn, thereby eliminating contaminated off-gases. This configuration typically does not
require air emission permitting (site-specific exceptions may apply).
Figure 2-3 illustrates a configuration that may alleviate the threat to subsurface
structures while achieving the same basic effect as air injection alone. In this configuration,
soil gas is extracted near the structure of concern and reinjected at a safe distance. If
necessary, makeup air can be added before injection.
Figure 2-4 illustrates a conventional soil venting configuration at sites where
hydrocarbon emissions to the atmosphere are not a problem. This may be the prefen-ed
configuration. Dewatering, nutrient, and moisture additions are also illustrated. Dewatering
will allow more effective treatment of deeper soils. The optimal configuration for any given
site will, of course, depend on site-specific conditions and remedial objectives.
The significant features of this technology include the following:
• Optimizing airflow to reduce volatilization while maintain-
ing aerobic conditions for biodegradation
• Monitoring local soil gas conditions to assure aerobic condi-
tions, not just monitoring vent gas composition
• Adding moisture and nutrients as required to increase bio-
degradation rates although, as stated earlier, it appears from
field studies that this may not be necessary at many if not
most sites
• Manipulating the water table (dewatering) as required for
air/contaminant contact.
-------
Revision 2
Page: 7
May 14, 1992
Cutoff Well to Prevent LoW pa|Q
Migration to Basement Air Injection
(if necessary)
Soil Gas
Monitoring
Biodegradation
of Vapors
Figure 2-1. Conceptual Layout of Biovenling Process
will) Air Injection Only.
-------
Revision 2
Page: 8
May 14, 1992
Air Injection
(Optional)
Figure 2-2. Concepfual Layout of Bioventing Process
with Air Withdrawn from Clean Soil.
KA/U*yb 03
-------
Revision 2
Page: 9
May 14, 1992
Monitoring
Figure 2-3. Conceptual Layout of Biovenliiig Process
with Soil (ias Keinjcction.
KA/ > ¦(»wc W »>
-------
Nutrient Application
Revision 2
Page: 10
May 14, 1992
Figure 2-4. Conceptual Layout of Bioventing Process
with Air Injection into Contaminated Soil,
Coupled with Dewatering and Nutrient Application.
-------
Revision 2
Page: 11
May 15, 1992
2.1.4 Hill AFB Site
A spill of approximately 25,000 gal of JP-4 jet fuel occurred when an automatic
overflow device failed at Hill AFB in Ogden, Utah. Contamination was limited to the upper
65 ft of a delta outwash of the Weber River. This surficial formation extends from the
surface to a depth of approximately 65 ft and is composed of mixed sand and gravel with
occasional clay stringers. Depth to regional groundwater is approximately 600 ft; however,
water may occasionally be found in discontinuous perched zones. Soil moisture averaged less
than 6% in the contaminated soils.
The collected soil samples had JP-4 fuel concentrations up to 20,000 mg/kg, with
an average concentration of approximately 400 mg/kg (Oak Ridge National Laboratory,
1989). Contaminants were unevenly distributed to depths of 65 ft. Vent wells were drilled to
approximately 65 ft below the ground surface and were screened from 10 to 60 ft below the
surface. A background vent was installed in an uncontaminated location in the same
geological formation approximately 700 ft north of the site.
Venting was initiated in December 1988 by air extraction at a rate of -25 cfm.
The off-gas was treated by catalytic incineration, and it was initially necessary to dilute the
highly concentrated gas to remain below explosive limits and within the incinerator's
hydrocarbon operating limits. The venting rate was gradually increased to -1,500 cfm as
hydrocarbon concentration levels dropped. During the period between December 1988 and
November 1990, more than 3.5 x 10 ft3 of soil gas were extracted from the site. In
November 1989, ventilation rates were reduced to between -300 and 600 cfm to provide
aeration for bioremediation while reducing off-gas generation. This change allowed removal
of the catalytic incinerator, saving -$6,000 per month.
During extraction, oxygen and hydrocarbon concentrations in the off-eas were
measured. To quantify the extent of biodegradation at the site, the oxygen was converted to
an equivalent basis. This was based on the stoichiometric oxygen requirement for hexane
mineralization. JP-4 hydrocarbon concentrations were determined based on direct readings of
a total hydrocarbon analyzer calibrated to hexane. Based on these calculations, the mass of
the JP-4 fuel as carbon removed was -115,000 lb volatilized and 93,000 lb biodegraded.
Figures 2-5 and 2-6 illustrate these results.
Hinchee and Arthur (1991) conducted bench-scale studies using soils from this
site and found that, in the laboratory, both moisture and nutrients became limiting after
aerobic conditions were achieved. This led to the addition of first moisture and then nutrients
in the field. The results of these field additions are shown in Figure 2-5. Moisture addition
clearly stimulated biodegradation; nutrient addition did not.
-------
Revision 2
Page: 12
May 14. 1992
Date
Figure 2-5. Cumulative Hydrocarbon Removal from (he Hill AFB
Building 914 Soil Venting Site.
50
E
20 QC
a.
10
-------
Revision 2
Page: 13
May 14, 1992
Hill AFB Building 914 Soil Samples
Depth
(feet)
L|
-------
Revision 2
Page: 14
May 14, 1992
The failure to observe an effect of nutrient addition could be explained by a
number of factors, including:
• The nutrients failed to move in the soils; this is a problem
particularly for ammonia and phosphorus (see Aggarwal et
al.. 1991).
Remediation of the site was entering its final phase, and the
nutrient addition may have been too late to result in an
observed change.
• Nutrients simply may have not been limiting.
2.1.5 Tyndall AFB Site
As a follow-up to the Hill AFB research, a more controlled study was designed at
Tyndall AFB. The experimental area in this study was located at a site where past JP-4 fuel
storage had resulted in contaminated soils. The nature and volume of fuel spilled or leaked
were unknown. The site soils are a fine- to medium-grained quartz sand. The depth to
groundwater is 2 to 4 ft
Four test cells were constructed to allow control of gas flow, water flow, and
nutrient addition. Test cells VI and V2 were installed in the hydrocarbon-contaminated zone;
the other two were installed in uncontaminated soils. Initial site characterization indicated the
mean soil hydrocarbon levels were 5,100 and 7,700 mg of hexane-equivalent/kg in treatment
plots VI and V2, respectively. The contaminated area was dewatered, and hydraulic control
was maintained to keep the depth to water at -5.25 ft This exposed more of the contaminat-
ed soil to aeration. During normal operation, airflow rates were maintained at approximately
one air-filled void volume per day.
Biodegradation and volatilization rates were much higher at the Tyndall AFB site
than those observed at Hill AFB; these higher rates were likely due to higher average levels
of contamination, wanner temperatures, and the presence of moisture. After 200 days of
aeration, an average hydrocarbon reduction of -2,900 mg/kg was observed. This represents a
reduction in total hydrocarbons of approximately 40%.
The study was terminated because the process monitoring objectives had been
met; biodegradation was still vigorous. Although the total petroleum hydrocarbons had been
reduced by only 40%, the low-molecular-weight aromatics — benzene, toluene, ethylbenzene,
and xylenes (BTEX) - were reduced by more than 90% (see Figure 2-7). It appears that the
bioventing process more rapidly removes the BTEX compounds than the other JP-4 fuel
constituents.
-------
Revision 2
Page: 15
May 14, 1992
300
Figure 2-7. Results of Soil Analysis from Plot V2 at Tyndall AFB Before and After
Venting. Each bar represents the average of 21 or more soil samples.
-------
Revision 2
Page: 16
May 15, 1992
Another important observation of this study is the effect of temperature on the
biodegradation rate. Miller (1990) found that the van Hoff-ArThenius equation provided an
excellent model of temperature effects. In the Tyndall AFB study, soil temperature varied by
only -7°C, yet biodegradation rates were approximately twice as high at 25°C than at 18°C.
In the Tyndall AFB study, the effects of moisture and nutrients were observed in
a field test. Two side-by-side plots received identical treatment, except that one (V2)
received both moisture and nutrients from the outset of the study while the other plot (VI)
received neither for 8 weeks, then moisture only for 14 weeks, followed by both moisture and
nutrients for 7 weeks. As illustrated in Figure 2-8, no significant effect of moisture or
nutrients was observed. The lack of moisture effect contrasts with the Hill AFB findings, but
is most likely the result of contrasting climatic and hydrogeologic conditions. Hill AFB is
located on a high-elevation desert with a very deep water table. Tyndall AFB is located in a
moist subtropical environment, and at the site studied, the water table was maintained at a
depth of approximately 5.25 ft
The nutrient findings suppon field observations at Hill AFB that the addition of
nutrients does not stimulate biodegradation. Based on acetylene reduction studies, Miller
(1990) speculates that adequate nitrogen was present due to nitrogen fixation. Both the Hill
and Tyndall AFB sites were contaminated for several years before the bioventing studies, and
both sites were anaerobic. It is possible that nitrogen fixation, which is maximized under
these conditions, provided the required nutrients. In any case, these findings show that
nutrient addition is not always required.
In the Tyndall study, a careful evaluation of the relationship between air flow
rates and biodegradation and volatilization was made. It was found that extracting air at the
optimal rate for biodegradation resulted in 90% removal by biodegradation and 10% removal
by volatilization. It was also found that passing the 10% volatilized through clean soil
resulted in complete biodegradation.
2.2 Soil Gas Permeability and Radius of Influence
An estimate of the soil's permeability to fluid flow (k) and the radius of influence
(Rj) of venting wells are both important elements of a full-scale bioventing design. On-site
testing provides the most accurate estimate of the soil gas permeability, k. On-site testing can
also be used to determine the radius of influence that can be achieved, for a given well
configuration and its flow rate and air pressure. These data are used to design full-scale
systems, specifically to space venting wells, to size blower equipment, and to ensure that the
entire site receives a supply of oxygen-rich air to sustain in situ biodegradation.
Soil gas permeability, or intrinsic permeability, can be defined as a soil's capacity
for fluid flow, and varies according to grain size, soil uniformity, porosity, and moisture
content The value of k is a physical property of the soil: k does not change with different
extraction/injection rates or different pressure levels.
-------
Revision 2
Page: 17
May 14. 1992
Venting Time (Days)
Figure 2-8. Cumulative Percent Hydrocarbon Removal at Tyndall AFII
for Sites VI and V2.
KA/OVb 01
-------
Revision 2
Page: 18
May 15. 1992
Soil gas permeability is generally expressed in the units cm2 or darcy (1 darcy =
1 x 10"8 cm2). Like hydraulic conductivity, soil gas permeability may vary by more than an
order of magnitude on the same site due to soil variability. Table 2-1 illustrates the range of
typical k values to be expected with different soil types.
TABLE 2-1. Soil Gas Permeability Values
Soil Type
k in Darcy
Coarse Sand
100-1000
Medium Sand
1-100
Fine Sand
0.1-1.0
Silts/Clays
<0.1
Source: Johnson et al. (1990)
The radius of influence is defined as the maximum distance from the air
extraction or injection well where measurable vacuum or pressure (soil gas movement) oc-
curs. Rj is a function of soil properties, but is also dependent on the configuration of the
venting well and extraction or injection flow rates, and is altered by soil stratification. On
sites with shallow contamination, the radius of influence can also be increased by im-
permeable surface barriers such as asphalt or concrete. These paved surfaces may or may not
act as vapor barriers. Without a tight seal to the native soil surface, the pavement will not
significantly impact soil gas flow.
Several field methods have been developed for determining soil gas permeability
(see review by Sellers and Fan, 1991). The most favored field test method is probably the
modified field drawdown method developed by Paul Johnson and associates at the Shell
Development Company. This method involves the injection or extraction of air at a constant
rate from a single venting well while measuring the pressure/vacuum changes over time at
several monitoring points in the soil away from the venting well. A detailed description of
the method, including equations to compute k, is presented in the Appendix.
2.3 In Situ Respiration Testing
As part of the Air Force's bioventing R&D program, a test was identified to
provide rapid field measurement of in situ biodegradation rates so that a full-scale bioventing
system can be designed. This section describes such a test as developed by Hinchee et al.
(1991b). This respiration test has been used at numerous sites throughout the United States.
-------
Revision 2
Page: 19
May 15, 1992
The in situ respiration test described in this protocol (Sections 4.0 and 5.0) is essentially the
same with minor modifications.
The in situ respiration test consists of placing narrowly screened soil gas
monitoring points into the unsaturated zone fuel-contaminated and uncontaminated soils and
venting these soils with air containing an inert tracer gas for a given period of time. The
apparatus for the respiration test is illustrated in Figure 2-9. In a typical experiment, two
monitoring point locations — the test location and a background control location — were used.
A cluster of three to four probes were usually placed in the contaminated soil of the test
location. A 1 to 3% concentration of inert gas was added to the air, which was injected for
about 24 hours. The air provided oxygen to the soil, while inert gas measurements provided .
data on the diffusion of 02 from the ground surface and the surrounding soil and assured that
the soil gas sampling system did not leak. The background control location was placed in an
uncontaminated site with air injection to monitor natural background respiration.
Measurements of C02 and 02 concentrations in the soil gas were taken before
any air and inert gas injection. After air and inert gas injection were turned off, C02 and 02
and inert gas concentrations were monitored over time. Before a reading was taken, the
probe was purged for a few minutes until the C02 and 02 readings were constant. Initial
readings were taken every 2 hours and then progressively over 4- to 8-hour intervals. The
experiment was usually terminated when the 02 concentration of the soil gas was -5%.
The monitoring points in contaminated soil at each site showed a significant
decline in 02 over a 40- to 80-hour monitoring period. Figure 2-10 illustrates the average
results from four sites, along with the corresponding 02 utilization rates in terms of percent of
02 consumed per hour. In general, little or no 02 utilization was measured in the uncontami-
nated background well. Inorganic uptake of 02 was assumed to be negligible, as seen by the
low available iron present in the soil. Aerating the soil for 24 hours was assumed to be
sufficient to oxidize any ferrous ions. Table 2-2 provides a summary of in situ respiration
rates and reported bioventing data.
The biodegradation rates measured by the in situ respiration test appear to be
representative of those for a full-scale bioventing system. Miller (1990) conducted a 9-month
bioventing pilot project at Tyndall AFB at the same time Hinchee et al. (1991b) were
conducting their in situ respiration test. The 02 utilization rates (Miller, 1990) measured from
nearby active treatment areas were virtually identical to those measured in the in situ
respiration test
C02 production proved to be a less useful measure of biodegradation than 02
disappearance. TTie biodegradation rate in milligrams of hexane-equivalent/kilograms of soil
per day based on C02 appearance is usually less than can be accounted for by the 02
disappearance. The Tyndall AFB site was an exception. That site had low-alkalinity soils
and low-pH quartz sands, and C02 production actually resulted in a slightly higher estimate
of biodegradation (Miller, 1990).
-------
Revision 2
Page: 20
May 14, 19D2
Pressure
Gage
3-Way Valving
i ^— Gas Sampling
Qr Port
Rotometer
2J6 or Mora Feet
f
0.5 to 2 Feet
_L
I
Ground Surface
^ Small Diameter
Probe
Screen
Rotometer
Regulator
Inert Gas
JL
Figure 2-9. (las Inject ion/Soil (ins Sampling Monitoring I'oiut Used by
llinchee el al. (1991) in Their Iii Situ Respiration Studies.
-------
Revision 2
Page: 21
May 14, 1992
25
20 -
15 H
c
fl>
10 H
5 H
Patuxent River MAS, MD
k - 0.13%/hr
IVndall AFB, FL
k - 0.43%/hr
Eielson AFB, AK
k - 0.22%/hr
60
r
0
T
20
I
40
Time (hours)
80
Figure 2-10. Average Oxygen Utilization Rales Measured
al Four Test Sites.
-------
Page: 22
May 14, 1992
TABLE 2-2. Summary of Reported In Situ Respiration and Bioventing Rate Data.
Site
Scale of Application
Contaminant
In Situ Respiration
Rates (% Oj/hr)
Estimated
Biodegradalion
Rates
Reference
Hill AFB, Utah
Full-scale, 2 years
JP-4 jet fuel
up to 0.S2
Up lo 10 mg/(kg
day)(,b)
Hinchee el al., 1991a
Tyndall AFB.
Florida
Field pilot, 1 year and
in situ respiration lest
JP-4 jet fuel
0.1 - 1.0
2-20 mg/(kg day)
Miller. 1990 and
Hinchee ct al., 1991b
The Netherlands
Undefined
Undefined
0.1 - 0.26
2-5 mg/(kg day)b
Urlings el al., 1990
The Netherlands
Field pilot, 1 year
Diesel
0.42
8 mg/(kg day)
van Eyk and Vreeken,
1989b
Undefined
Full scale
Gasoline and
diesel
—
50 kg/(well day)1
Ely and Heffner, 1988
Undefined
Full scale
Diesel
—
100 kg/(well day)c
Ely and Heffner, 1988
Undefined
Full scale
Fuel oil
—
60 kg/(well day)c
Ely and Heffner, 1988
Paluxent River
NAS, Maryland
In situ respiration test
JP-S jet fuel
0.16
3 mg/(kg day)
Hinchee el al., 1991b
Fallon NAS, Nevada
In silu respiration test
JP-S jet fuel
0.26
5 mg/(kg day)
Hinchee el al., 1991b
Eiclson AFB.
Alaska
In silu respiration lest
JP-4 jet fuel
0.05 - 0.5
1-10 mg/(kg day)
Hinchee el al., 1991b
Kenai, Alaska
In silu respiration test
Crude
Petroleum
I.I
21 mg/(kg day)
Hinchee and Ong, 1991
Tinker AFB,
Oklahoma
In silu respiration test
JP-4 and mixed
fuels
0.14 - 0.94
2.7 - 18 mg/(kg day)
Hinchee and Smith. 1991
* Rales reported by Hinchee el al., (1991) were first order willi respect to oxygen; for comparative purposes, these have been converted to zero order with
respect to hydrocarbons al an assumed oxygen concentration of 10%.
h Rates were reported as oxygen consumption rales; these have been converted to hydrocarbon degradation rales assuming a 3:1 oxygcn-io-hydrocarbon ratio.
' Units arc in kilograms of hydrocarbon degraded per 30 standard cubic feel |>cr minulc (scfui) extraction venl well per day.
-------
Revision 2
Page: 23
May 14, 1992
In the case of the higher pH and higher alkalinity soils at Fallon NAS and
Eielson AFB, little or no gaseous C02 production was measured (Hinchee et al., 1991b).
This could be due to the formation of carbonates from the gaseous evolution of C02 produced
by biodegradation at these sites. A similar problem was encountered by van Eyk and
Vreeken (1988) in their attempt to use C02 evolution to quantify biodegradation associated
with soil venting.
-------
Revision 2
Page: 24
May 14, 1992
3.0 IN SITU RESPIRATION/AIR PERMEABILITY TEST PREPARATION
The necessary preparation, procedures, and specific tasks to conduct the in situ
respiration/air permeability test are presented in the following subsections. Figure 3-1 shows
a generalized flow chart of the process.
3.1 Site Characterization Review
To initiate site characterization, the project officer will inform the contractor of
the Air Force facilities and specific sites where these tests will be conducted. The project
officer will also provide a contact person at each Air Force facility (hereafter called base
point-of-contact, or base POC). The. project officer and/or the base POC will supply any
relevant documents (site characterization reports, underground uulity drawings, remedial
investigation/feasibility studies, etc.) pertaining to the contaminated area.
A tentative test site will be selected after reviewing all preliminary documents
and consulting with the project officer and the base POC. Final approval of the test area will
be obtained from the project officer.
3.2 Development of Site-Specific Test Plan
All involved parties for a given site will be provided with a site-specific test plan.
The site-specific test plan will consist of this generic test plan with a site-specific cover letter.
The following information will typically be provided in the cover letter:
• A map showing the chosen test location, and if possible,
tentative vent well and monitoring point locations
Construction details for tentative vent well and monitoring
points
Details of any required permits and actions taken to obtain
the permits
Estimated field start date
Any anticipated deviations from the generic test plan
Site-specific support required from the base
Site-specific health and safety requirements, if required.
-------
Revision 2
Page: 25
May 14. 1992
Review existing
Documents,
Preliminary Evaluation
Inputs from
Project Officer,
Base POC
Selection of Tentative
Test Area
4
Inputs from
Project Officer,
Base POC
Develop and Submit
Site Specific
Test Plan
4
Inform Base POC
Project Officer,
and Regulators as
appropriate
Obtain Necessary
Permits/Approval
4
SITE WORK After Approval
Conduct Initial Site Survey
and/or Soil Gas Survey
(Shallow Sites)
Conflra Test
Area
Inform Project
Officer and Base
POC
Conduct In S1tu Respiration/
Air Permeability Tests
4
Institute Long Term Bioventing
Test
Hove to
Next Site
Figure 3-1. Flow Chart for Conducting Bioventing Treatability Test.
-------
Revision 2
Page: 26
May 14, 1992
The site-specific test plan will be submitted to the project officer, base POC. and
any necessary regulatory agencies for approval. The test plan will normally be submitted to
outside regulatory agencies by either the project officer or the base POC. Unless specifically
directed otherwise by the project officer, the contractor will not directly contact regulatory
agencies or submit plans to them. No site work will be initiated without the necessary
approval.
3.3 Application for Required Permits
/
As soon as a candidate site is identified by the Air Force project officer,
applications must be submitted for the required permits. Obtaining permits frequently is the
greatest holdup in accomplishing this type of field work. It is likely that no state or local
permits will be required, but this must be determined early. Types of permits that may be
required include:
• Drilling and/or well installation permits for the vent well
and/or monitoring points
• Air Emission Permit for the vent well if air is extracted.
• Site Investigation Permit or Approval. This usually will not
be necessary; however, some regulatory jurisdictions may
require permitting. This test should not normally be consid-
ered a CERCLA treatability test.
No direct contact will be made by the contractor with regulatory agencies without
project officer and base POC approval. In many cases the project officer or base POC will
handle regulatory contacts,, if they are necessary.
The contractor will coordinate with the base POC to obtain access and necessary
clearance to conduct the tests at the candidate test area. The contractor will arrange with the
base for the utilities — electricity and water — needed to execute the tests. If electricity is not
available, the contractor will provide power from portable generators. The contractor will
coordinate with the base POC to obtain any necessary security clearances or badges.
As early as possible, the contractor will supply the base POC with a list of all
personnel to be used on base, including name, social security number, place and date of birth,
and expected arrival date. The contractor will also request that the base POC initiate the
process of obtaining a digging permit.
-------
Revision 2
Page: 27
May 14. 1992
4.0 TEST WELLS AND EQUIPMENT
This section describes the test wells and equipment that are required to conduct
the field treatability tests. It must be recognized that site-specific flexibility will be required,
and thus, details will vary. Local and/or state regulatory agencies and at times individual Air
Force bases will have specific requirements that differ from specifications in this test plan.
All testing must comply with regulations, and must be acceptable to the host base.
Field notes will be maintained describing all vent well and monitoring point
construction. Deviations from standard design will be noted in the final report.
4.1 Vent Wells
A vent well and blower system will be established to provide airflow through the
subsurface, creating a pressure/vacuum gradient for air permeability testing and increasing
subsurface oxygen levels for in situ respiration testing. This 2- to 4-in. vent well will be
placed with the screened section in contaminated soil and will be located near the center of
the fuel spill. The siting and construction of the venting well will follow these general
criteria:
1. The vent well will be sited as near to the center of the spill
area as possible. This location will ensure that data gath-
ered from the test will be as representative as possible of
contaminated soil conditions. On many small sites, the vent
well used during the treatability test can be converted into
the primary vent well for extended testing.
2. The diameter of the vent well may vary between 2 and 4 in.
and will depend on the ease of drilling and the area and
depth of the contaminated volume. On most sites a 2-in.-
diameter vent will provide adequate airflow for air perme-
ability/radius of influence testing. For sites with contamina-
tion extending below 30 ft, a 3- or 4-in. vent well is recom-
mended. The cost of a larger well is a minor component of
the total drilling cost because a drill rig will be required to
drill to this depth, regardless of well diameter. Groundwater
monitoring points screened several ft above the existing
water table can also be converted to vent wells. This option
is appropriate for air injection systems but will be less
successful for air extraction systems because the applied
vacuum will cause a rise in the water table which could
rapidly submerge the screened interval.
-------
Revision 2
Page: 28
May 14, 1992
3. The vent well will normally be constructed of schedule 40
polyvinyl chloride (PVC), and will be screened with a slot
size that maximizes airflow through the soil. The screened
interval will extend through as much of the contaminated
profile as possible, with the bottom of the screen corre-
sponding to the top of the capillary fringe. For shallow
sites with groundwater less than 20 ft deep, the vent well
will be screened over the bottom half of the unsaturated
zone. For deeper wells, care must be taken in determining
the depth of the top of the screen. A deeper screen is
normally better. If the top of the screen is close to the
ground surface, much of the airflow may follow the shortest
path from near the top of the screen to the ground surface.
4. Hollow-stem augering is the recommended drilling method;
however, a solid-stem auger is also acceptable in more
cohesive soils. Whenever possible, the diameter of the
annular space will be at least two times greater than the
vent well outside diameter. The annular space correspond-
ing to the screened interval will be filled with silica sand or
equivalent In shallow softer soils, hand-augering may be
feasible. The annular space above the screened interval will
be sealed with wet bentonite and grout to prevent short-
circuiting of air to or from the surface. Figure 4-1 shows a
typical vent well.
4.2 Soil Gas Monitoring Points
Soil gas monitoring points will be used for pressure and soil gas measurements
and will be installed at a minimum of three locations, and at each location to at least three
depths. The total number will vary, with up to six monitoring point locations, and six or
more depths, depending on site conditions.
To the extent possible, the monitoring points will be located in contaminated soils
with >1,000 mg/kg of total petroleum hydrocarbon. These soils will have a strong odor and
will feel oily to the touch. It may not be possible to locate all monitoring points in contami-
nated soil, especially the points furthest from the vent well. If this is the case, it is important
to ensure that the point closest to the vent well be located in contaminated soil, and if
possible, the intermediate point be placed in contaminated soils. If no monitoring points are
located in contaminated soil, no meaningful in situ respiration test can be conducted. If the
initial oxygen levels in the soil gas are not low, i.e., below 2 to 5%, and the soil gas
hydrocarbon levels are not high, say above 10,000 ppm for relatively fresh JP-4 fuel, the
monitoring point may not be suitable for an in situ respiration test.
-------
Revision 2
Page: 29
May 14, 1992
2•4" Dla. SCH 40 PVC
Header Sloped to
Blower
To Blower
Bentonlte/Cement Grout
to Surface
2-4" Ola. SCH 40 PVC Casing
BentonKe Seal
(2' Minimum)
2*4" Dla. SCH 40 PVC Screen
Silica Sand
Undisturbed Soil
End Cap
Not to Scale
Figure 4-1. Typical Injection/Vacuum Venting Well Construction.
-------
Revision 2
Page: 30
May 14, 1992
Higher oxygen concentrations would indicate that the microbial activity is not
oxygen-limited or that there is sufficient exchange of air with the atmosphere to keep the soil
gas well-aerated. In either case, bioventing will not increase biodegradation rates. At some
sites, where less contaminated soils and low 02 concentrations are encountered, bioventing
may still be feasible. If these conditions are found, care must be taken to place the monitor-
ing points in the most contaminated soil possible.
4.2.1 Location of Monitoring Points
A minimum of 3 monitoring points is recommended; ideally these will be in a
straight line and at the intervals recommended in Table 4-1. In an unobstructed heteroge-
neous site, 3 monitoring points at these spacings are appropriate. Additional monitoring point
locations may be necessary for a variety of site-specific reasons including, but not limited to,
spatial heterogeneities, obstructions, or the desire to monitor a specific location. Additional
discussion related to monitoring point placement is found in Section 5.0, Test Procedures.
4.2.2 Depth of Monitoring Points
In general, each monitoring point will be screened to at least 3 depths. The
deepest screen will be placed either at or near the bottom of contamination if a water table is
not encountered, or a minimum of 2 to 3 ft above the water table if it is encountered.
Consideration will be given to potential seasonal water table fluctuations and soil type in
finalizing the depth. In a more permeable soil the monitoring point can be screened closer to
the water table. In a less permeable soil it must be screened further above the water table.
The shallowest screen will normally be 3 to 5 ft below land surface. The intermediate screen
will be placed at a reasonable interval at a depth corresponding to the center to upper lA of
the depth of the vent well screen.
As an example, in .a sandy soil with groundwater at 30 ft and a vent well
screened from 17.5 to 27.5 ft below land surface, reasonable screened depths for the
monitoring points would be 28 ft, 22.5 ft, and 3 ft For sites with vent wells deeper than 30
ft, more depths may be screened, depending on stratigraphy.
It will be necessary in some cases to add additional screened depths to ensure a
well-oiled soil is encountered, to monitor differing stratigraphic intervals, or to adequately
monitor deeper sites with broadly screened vent wells. If air injection is being considered in
the bioventing test, a monitoring point must be located between the vent well and any
buildings that may be at risk to assure that they are well beyond the radius of influence.
-------
Revision 2
Page: 31
May 14. 1992
TABLE 4-1. Recommended Spacing for Monitoring Points
Soil Type
Depth to Top of
Vent Well Screen
(ft)'1'
Spacing
Interval (ft)(2)
Coarse Sand
5
5-10-20
10
10-20-40
>15
20-30-60
Medium Sand
5
10-20-30
10
15-25-40
>15
20-40-60
Fine Sand
5
10-20-40
10
15-30-60
>15
20-40-80
Silts
5
10-20-40
10
15-30-60
>15
20-40-80
Clays
5
10-20-30
10
10-20-40
>15
15-30-60
(1) Assuming 10 ft of vent well screen, if more screen is
used, the >15-ft spacing will be used.
(2) Note that monitoring point intervals are based on a vent-
ing flow rate range of 1 cfm/ft screened interval for clays
to 3 cfm/ft screened interval for coarse sands.
4.2.3 Construction of Monitoring Points
Most state and local regulatory agencies do not regulate unsaturated zone soil gas
monitoring point construction. Nevertheless, prior to construction it is necessary to check
with regulators to assure compliance with any regulations that may exist.
-------
Revision 2
Page: 32
May 15. 1992
Monitoring point construction will vary depending on the depth of drilling and
the drilling technique. Basically, the monitoring points will consist of a small-diameter l/i-in.
tube to the specified depth with a screen approximately 6 in. long and lA to 1 in. in diameter.
In shallow hand-augered installations, rigid tubing (i.e., Schedule 80 lA" PVQ terminating in
the center of a gravel or sand pack may be adequate. The gravel or sand pack will normally
extend for an interval of 1 to 2 ft with the screen centered. In low-permeability soils, a larger
gravel pack may be desirable. In wet soils a longer gravel pack with the screen near the top
may be desirable. A bentonite seal at least 2 ft thick is normally required above and below
the gravel pack. Figure 4-2 shows a typical installation.
For relatively shallow installations in more permeable soils, a hand-driven
system, such as that of KVA Associates, may be used. In such a system, a sacrificial drive
point with Tygon™, Teflon™, or other appropriate tubing is driven to the desired depth.
Then, the steel outer tubing is retrieved, leaving the drive point and the inner flexible tubing
in place. Because this type of installation allows little or no sand pack or seal placement, it
should be used only in relatively permeable soils where sample collection will not be a
problem or in soils that will "self heal" to prevent short-circuiting. Surface completion of the
hand-driven points should be the same as for those installed in borings.
Tubes will be used to collect soil gas for C02 and 02 analysis in the 0.25%
range, and for JP-4 hydrocarbons in the 100 ppm range or higher. The tubing material must
have sufficient strength and be nonreactive. Sorption and gas interaction with the tubing
materials have not been significant problems for this application. If a monitoring point will
be used to monitor specific organics in the low ppm or ppb range, teflon or stainless steel
may be necessary. However, this will not normally be the case.
All tubing from each monitoring point will be finished with quick-connect
couplings and will be labeled twice. Each screened depth will be labeled as follows:
[Code for Site] — [Code for Monitoring Point] — [Depth to Center of Screened Interval].
Tabic 4-2 lists the labels used for example site #2 at Millersworth AFB. In M2,
the M is for Millersworth AFB, and the 2 is for site #2 at Millersworth. The tubing will be
labeled with a firmly attached metal tag or directly by engraving or in waterproof ink.
Instead of a metal tag, a metal plate may be placed at the bottom of the monitoring point
compartment with holes drilled for each tube. The metal plate will then be engraved,
identifying each tube where it passes through the plate. If this method is used, the tube itself
must still be labeled with ink or by engraving. The label will be placed close to the ground
so that, if the tube is damaged, the label is likely to survive.
The top of each monitoring point will be labeled to be visible from above. This
will be done either by writing in the concrete or with spray paint.
-------
Revision 2
Page: 33
May 14. 1992
Water Tight Cast
Iron Well Box
Figure 4-2. Typical Monitoring Point Construction Detail.
(Dimensions will vary for specific installations.)
-------
Revision 2
Page: 34
May 14. 1992
TABLE 4-2. Monitoring Points for Example Site #2
at Millersworth AFB
M2-A-3
(3 ft deep)
Monitoring Point A,
Closest to the vent well
M2-A-15
(15 ft deep)
M2-A-25
(25 ft deep)
M2-B-3
(3 ft deep)
Monitoring Point B,
Intermediate from vent
well
M2-B-15
(15 ft deep)
M2-B-27
(27 ft deep)
M2-C-3
(3 ft deep)
Monitoring Point C,
Farthest from vent well
M2-C-14
(14 ft deep)
M2-C-23
(23 ft deep)
The monitoring points will be finished by placement in a watertight cast iron well
box. The well box will be placed either aboveground in a concrete pad or at grade, also in
concrete. The box will be drained to prevent water accumulation.
4.2.4 Thermocouples
Two thermocouples will be installed at each site. They will be installed at the
monitoring point closest to the vent well and, as shown in Figure 4-2, at the depth of the
shallowest and deepest screen. Thermocouples used are either J or K type. The thermocou-
ple-wires will be labeled using the same system as for the tubings, except that a two-letter
word, TC, will be added to the identification label (e.g., M2-TCA-3, for the thermocouple
installed at the second Millersworth AFB site monitoring point A at the 3-ft depth).
4.3 Background Well
In addition to the vent well and the monitoring points installed in contaminated
soils, a background well will be installed in uncontaminated soil to monitor the background
respiration of natural organic matter. Soil gas in uncontaminated soil generally has 02 levels
between 15 and 20% and C02 levels between 1 and 5%. The background well will be
similar in construction to the vent well (Figure 4-1), except that the length of the screen will
be approximately 5 ft.
To the extent possible, the screen of the background well will be located at a
depth similar to that of the monitoring points and in the same stratigraphic formation. For
-------
Revision 2
Page: 35
May 14. 1992
sites deeper than 20 ft, the screen portion of the background well will be placed at 20 to 25
feet. For depths less than 20 ft. the screen portion of the background well will be placed
between 5 and 15 ft.
4.4 Blower System
The type and size of blower used on a test site will be determined based upon the
soil type, depth and area of contamination, and available power. In an attempt to reduce the
number of blower units in the pilot test inventory and to standardize piping and instru-
mentation, two typical blowers are specified:
Blower One
Application:
Contaminated interval in sandy soils and mixed sandy/silt and
sandy/clay soils.
Typical Specifications:
— Explosion-proof regenerative blower
— 20 to 90 scfm at 20" to 100" H20, respectively
— 3-HP explosion-proof motor
— Single-phase 230-V power source
Blower Two
Application:
Predominantly silt and clay soils.
Typical Specifications:
— Explosion-proof pneumatic blower
— 50 scfm at 130" H20.
— 5-HP explosion-proof motor
— Single-phase 230-V power source.
Each blower will be fitted with mounting brackets and pipe fittings to make it
compatible with the basic blower systems shown in Figures 4-3 and 4-4. Explosion-proof
blowers and motors are required when soil gas extraction is used. Explosion-proof equipment
may be required for air injection systems as well.
The blower system will be instrumented to monitor blower performance and to
provide test data such as the vent well pressure (Pw) and the gas stream flow rate (Q) .
adjusted for air density. Using these data and pressure data from each soil gas monitoring
point, k and Rj can be estimated.
-------
Sample Point
To
Atmosphere
Revision 2
Page: 36
May 14. 1992
From Vent Well
Air Velocity Pitot Tube
(or Similar Indicator)
Pressure Indicator
Temperature Indicator
Flow Control Valve
Vacuum Relief Valve
Optional; at a minimum well head
pressure, extraction gas flow rate must
be measured, and accommodation for
gas sampling made
Figure 4-3. Soil («us Permeability Blower System Instrumentation Diagram
for Soil («as Extraction.
KA/a^/voi
-------
Revision 2
Page: 37
May 14, .1992
Bom
Atmosphere
PRV* /tT\*
ili
o-KstTS
* . Btorer ^
Filler*
©
Air Velocity Pitot Tube
(or Similar Indicator)
Pressure Indicator
(V) Temperature Indicator
FCV Flow Control Valve
PRV Pressure Relief Valve
* Optional
-------
Revision 2
Page: 38
May 14, 1992
4.5 Field Instrumentation and Measurements
€
Sections 4.5.1 through 4.5.6 discuss the equipment used for measurements.
Figures supplement the text.
4.5.1 Oxygen and Carbon Dioxide
Gaseous concentrations of C02 and 02 will be analyzed using a GasTech model
32520X COj/Oj analyzer or equivalent. The battery charge level will be checked to ensure
proper operation. The air filters will be checked and, if necessary, cleaned or replaced before
the experiment is started. The instrument will be turned on and equilibrated for at least 30
minutes before conducting calibration or obtaining measurements. The sampling pump of the
instrument will be checked to ensure that it is functioning. Low flow of the sampling pump
can indicate that the battery level is low or that some fines are trapped in the pump or tubing.
Meters will be calibrated each day prior to use against purchased C02 and 02
calibration standards. These standards will be selected to be in the concentration range of the
soil gas to be sampled. The C02 calibration will be performed against atmospheric C02
(0.05%) and a 5% standard. The O2 will be calibrated using atmospheric 02 (20.9%) and
against a 5% and 0% standard. Standard gases will be purchased from a specialty gas
supplier. To calibrate the instrument with standard gases, a Tedlar™ bag (capacity -1 1) is
filled with the standard gas, and the valve on the bag is closed. The inlet nozzle of the
instrument is connected to the Tedlar™ bag, and the valve on the bag is opened (see Figure
4-5). The instrument is then calibrated against the standard gas according to the manufac-
turer's instructions. Next, the inlet nozzle of the instrument is disconnected from the
Tedlar™ bag and the valve on the bag is shut off. The instrument will be rechecked against
atmospheric concentration. If recalibration is required, the above steps will be repeated.
4.5.2 Hydrocarbon Concentration
Petroleum hydrocarbon concentrations will be analyzed using a GasTech Trace-
Techtor™ hydrocarbon analyzer (or equivalent) with range settings of 100 ppm, 1,000 ppm,
and 10,000 ppm. The analyzer will be calibrated against two hexane calibration gases (500
ppm and 4,400 ppm). The Trace-Techtor™ has a dilution fitting that can be used to calibrate
the instrument in the low-concentration range.
Calibration of the GasTech Trace-Techtor™ is similar to the GasTech Model
32402X, except that a mylar bag is used instead of a Tedlar™ bag. The 02 concentration
must be above 10% for the Trace-Techtor™ analyzer to be accurate. When the 02 drops
below 10%, a dilution fitting must be added to provide adequate oxygen for analysis.
Hydrocarbon concentrations can also be determined with a flame ionization
detector (FID), which can detect low (below 100 ppm) concentrations. A photoionization
detector (PID) is not acceptable.
-------
Revision 2
Page: 39
May 14, 1992
(a)
Instrument
Sampling
Hose
1/4" TUbing
Gas Analyzers
(COj, 02, and
Total Hydrocarbons)
1/4" Male
Quick Connect
Tedlar Bag
Filled With
Standard Gas
Sampling
Hose
Figure 4-5. Schematic Setup for Calibration of Soil Gas Instruments.
(a) C02, 02, and Total Hydrocarbon Analyzers.
(b) Helium Detector.
-------
Revision 2
Page: 40
May 14, 1992
4.5.3 Helium Monitoring
Helium in the soil gas will be measured with a Marks Helium Detector Model
9821 or equivalent with a minimum sensitivity of 100 ppm (0.01%). Calibration of the
helium detector follows the same basic procedure described for oxygen calibration, except
that the setup for calibration is different (see Figure 4-5). Helium standards used are 100
ppm (0.01%), 5,000 ppm (0.5%), and 10,000 ppm (1%).
4.5.4 Temperature Monitoring
In situ soil temperature will be monitored using Omega Type J or K thermo-
couples (or equivalent). The thermocouples will be connected to an Omega OM-400 Thermo-
couple Thermometer (or equivalent). Each thermocouple will be calibrated against ice water
and boiling water by the contractor before field installation.
4.5.5 Pressure/Vacuum Monitoring
Changes in soil gas pressure during the air permeability test will be measured at
monitoring points using Magnehelic™ or equivalent gauges. Tygon™ or equivalent tubing
will be used to connect the pressure/vacuum gauge to the quick-disconnect on the top of each
monitoring point. Similar gauges will be positioned before and after the blower unit to
measure pressure at the blower and at the head of the venting well. Pressure gauges are
available in a variety of pressure ranges, and the same gauge can be used to measure either
positive or negative (vacuum) pressure by simply switching inlet ports. Gauges are sealed
and calibrated at the factory and will be rezeroed before each test. The following pressure
ranges (in inches H20) will typically be available for this field test:
0-1". 0-5", 0-10", 0-20", 0-50", 0-100", and 0-200"
Air pressure during injection for the in situ respiration test will be measured with
a pressure gauge with a minimum range of 0 to 30 psig.
4.5.6 Airflow
Airflow measurements will be taken for both the air permeability test and the
respiration test These measurements are described in Sections 4.5.6.1 and 4.5.6.2.
4.5.6.1 Airflow Measurement — Air Permeability Test
During the air permeability test an accurate estimate of flow (Q) entering or
exiting the vent well is required to determine k and Rj. Several airflow measuring devices
are acceptable for this test procedure.
-------
Revision 2
Page: 41
May 14, 1992
Pitot tubes or orifice plates combined with an inclined manometer or differential
pressure gauge are acceptable for measuring flow velocities of 1,000 ft/min or greater (-20
scfm in a 2-in. pipe). For lower flow rates, a large rotometer will provide a more accurate
measurement. If an inclined manometer is used, the manometer must be rezeroed before and
after the test to account for thermal expansion/contraction of the water. Devices to measure
static and dynamic pressure must also be installed in straight pipe sections according to
manufacturer's specifications. All flow rates will be corrected to standard temperature and
ambient pressure (altitude) conditions.
4.5.6.2 Airflow Measurement — Respiration Test
Prior to initiating respiration tests at individual monitoring points, air will be
pumped into each monitoring point using a small air compressor as described in Section 5.7.
Airflow rates of 1 to 1.5 cfm will be used, and flow will be measured using a Cole-Palmer
Variable Area Flowmeter No. N03291-4 (or equivalent). Helium will be introduced into the
injected air at a 1% concentration. A helium flow rate of approximately 0.01 to 0.015 cfm
(0.6 to 1.0 cfh) will be required to achieve this concentration. A Cole-Palmer Model
L-03291-00 flowmeter or equivalent will be used to measure the flow rate of the helium feed
stream.
4.5.6.3 Airflow Measurement — Bioventing Test
Airflow measurements during the bioventing tests may be made as described for
the air permeability test (Section 4.5.6.1). If a single vent well and blower are used and
100% of the flow to the blower comes from the extraction well, the air flow measurement
may not be necessary. If a blower with a known pump curve is used and intake and exhaust
pressures are monitored, flow rate can be estimated from the pump curve.
-------
Revision 2
Page: 42
May 14, 1992
5.0 TEST PROCEDURES
5.1 Location of Optimum Test Area
A soil gas survey will be conducted to locate an optimum site for the vent well
and the soil gas monitoring points. Ideally, the vent well and monitoring points will be
located in well-oiled soils where the 02 is depleted and the C02 levels are elevated (see
discussion in 4.2). If at least three monitoring point screens are not located in the most
contaminated soils, then the in situ respiration test may not provide adequate information on
the biodegradation rates for the site.
5.1.1 Soil Gas Survey (for contamination < 20 ft)
A soil gas survey will be conducted prior to locating the vent well and monitor-
ing points at sites with relatively shallow groundwater where soils are penetrable to a depth of
within 5 ft of the water table using hand-driven gas probes. The survey will not be a
complete site soil gas survey to fully delineate contamination.
Accessibility to the site will be confirmed, along with possible restrictions that
may hamper the tests. Existing groundwater and soil gas monitoring wells near the test area
will be identified. Groundwater will be checked for free floating product, and soil gas from
any existing monitoring points or wells will be analyzed for 02, C02, and total hydrocarbons
before proceeding with the soil gas survey. To assist in the soil gas survey, a simple
sampling grid will be established using existing monitoring wells or prominent landmarks for
identification.
Soil gas sampling will be conducted using small-diameter (~4fc-inch OD) stainless
steel probes (KVA Associates or equivalent) with a slotted well point assembly. The
maximum depth for hand-driven probes will typically be 10 to 15 ft, depending on soil
texture. In some dense silts or clays, penetration of the soil gas probe will be less, while in
some unconsolidated sands, deeper penetration may be possible. At a given location on the
grid, a probe will be driven (manually or with a power hammer) to a depth determined by
preliminary review of the site contamination documents. Soil gas at this depth will be
analyzed for O2, C02, and total hydrocarbons. The probe will then be driven deeper, and the
soil gas will be measured. For a typical site with a depth to groundwater of 9 ft, soil gas will
be measured at depths of 2.5 ft, 5 ft, and 7.5 ft
The main criterion for selecting a suitable test site is that the microbial activity
should be oxygen-limited. Under such conditions, the 02 level will be low (usually 0 to 2%).
CO2 will be high (typically 5 to 20%, depending on soil type), and hydrocarbon content will
be high (> 10,000 ppm for most fresh JP-4 sites).
An uncontaminated site also will be located to be used as an experimental control
to monitor background respiration of natural organic matter and inorganic sources of CO:.
-------
Revision 2
Page: 43
May 14. 1992
Typical 02 and C02 levels at an uncontaminated site are 15 to 20% and 1 to 5%, respective-
ly. The hydrocarbon content in the soil gas of a contaminated site is generally below 100
ppm.
Prior to sampling, soil gas probes will be purged with a sample pump. To
determine adequate purging time, soil gas concentrations will be monitored until the
concentrations stabilize. This will not always be possible, particularly when shallow soil gas
samples are being collected, as atmospheric air may be drawn into the probe and produce
false readings. When shallow soil gas samples are collected, air withdrawal will be kept to a
minimum. Figure 5-1 shows a typical setup for monitoring soil gas.
5.1.2 Exploratory Boring in Deep Soils
On sites where contamination extends to depths greater than 20 ft. exploratory
borings will be used to ensure that the vent well and monitoring points are located in fuel-
contaminated soils. Exploratory borings that encounter significant fuel contamination will
then be completed and used as vent wells or monitoring points.
A hollow-stem auger will be used to advance the boring, and drill cuttings will be
visually checked and analyzed with a GasTech Trace-Techtor™ (or equivalent) hydrocarbon
analyzer, an equivalent explosimeter, or a FID, to determine the relative fuel contamination of
each 2- to 3-ft interval. Drill cuttings will be inspected at each contaminated interval selected
for monitoring point installations.
As the boring advances beyond 20 ft, a split-spoon sampling device will be
recommended for sampling at 5-ft intervals. Split-spoon samples will be visually checked for
fuel contamination and screened for volatile emissions by passing a hydrocarbon analyzer
slowly qver the open split spoon.
The purpose of this simple monitoring technique will be to provide air monitoring
for worker health and safety, to rapidly locate the interval of highest contamination, and to
attempt to locate the maximum depth of contamination at each site. A geologic driller's log
will be kept to identify changes in lithology, depths of apparent fuel contamination, and
sample locations. Exploratory borings will also be required to locate a clean area for
installing the background monitoring point. Careful inspection of drill cuttings and volatile
hydrocarbon monitoring will be required to ensure that soils in the control area are free of
fuel hydrocarbons.
-------
Figure 5-1. Schematic Diagram or Soil (>as Sampling
Using the Stainless Steel Soil (Jas Probe.
-------
Revision 2
Page: 45
May 14, 1992
5.2 Drilling and Installation of the Vent Well
Based on a review of available site characterization data, a preliminary location
will be proposed for the vent well. Following the soil gas survey and/or exploratory boring, a
final vent well location will be determined. If soils were proved to be sufficiently contami-
nated, the exploratory boring will be completed as the vent well. Soil samples will be
collected at a minimum interval of 5 ft in the vent well boring following the procedures
outlined in Section 5.5. Siting and construction of the vent well will follow the criteria
provided in Section 4.1.
5.3 Drilling and Installation of Monitoring Points
Based on the location of the vent well and available site characterization data, the
monitoring points will be located at points where sufficient data for the air permeability tests
can be obtained and, at the same time, they can be used for the in situ respiration test. Table
4-1 will be used as a guide to locate the monitoring points in relation to the location of the
vent well. The location of the monitoring points will also take into consideration the long-
term bioventing test that will be conducted after the in situ respiration test. The monitoring
points will generally be located in a contaminated area. Screens for the monitoring points
will have the same slot sizes as those for the vent well (see discussion in Section 4.2).
When possible, the monitoring points will be placed in hand-augered borings or
in borings augered with a small portable drill. At deeper sites, it will be necessary to hire a
driller for both the monitoring points and the vent well. When a drill rig is used, a hollow-
stem auger will most likely be used. A smaller ID auger will be used, as required, for the
vent well installation. Also as required, a solid auger will be used in shallow or cohesive
soils.
5.4 Background Well Installation
A background well will be installed in an uncontaminated location to obtain soil
gas measurements of O2 and C02 concentrations to monitor background respiration. The well
will be constructed in a manner similar to the vent well, except that it will normally be 1 in.
in diameter with a screen length of 5 ft. At sites deeper than 20 ft, the screened portion of
the background well will be placed at 20 to 25 ft, so long as it is screened in the same
geological formation as the vent well. Normally, deeper screening will be required only if
necessary to intercept the vented formation.
5.5 Collection of Soil Samples
A minimum of three to four soil samples will be collected from each site and
analyzed for physical/chemical characteristics, including nutrient concentration. At least one
representative sample of each contaminated soil type will be collected. It is important that
samples for nutrient analyses be collected from a contaminated zone; otherwise, if fixation
-------
Revision 2
Page: 46
May 14, 1992
has already occurred, the nitrogen concentration may not be representative. Soil samples will
be collected from the exploratory boring or from the borings for the vent well or monitoring
points. Soil samples will be collected from cuttings if the borings are shallow, by hand from
a hand-augered hole, or with a split-spoon sampler. Enough soil will be collected to fill a
500-ml polyethylene or glass container. The container will be sealed with a teflon-lined cap
and then placed in a cooler for shipment. Special procedures for preserving the sample will
not be required, as only inorganics and the physical properties of the soil will be analyzed.
Each soil sample will be labeled to identify the site, boring location and depth, and time of
collection. Soil samples may also be collected for total petroleum hydrocarbon (TPH)
analysis and for benzene, toluene, ethylbenzene, and xylene (BTEX) analysis. Samples to be
used for TPH, BTEX, or any other volatility analysis must be collected, bundled, stored, and
shipped in a manner that will prevent volatilization losses. The methods for this sampling are
described in other sources.
Chain-of-custody forms will accompany each shipment to the laboratory. The
soil samples will be analyzed for at least the following parameters:
pH
• total kjeldahl nitrogen (TKN)
• total phosphorus
• alkalinity
• particle size analysis
• total iron
• moisture content.
In addition to the chain-of-custody forms, each sample will be logged into the
project record book along with a complete description of where and how it was collected.
Each sample will be labeled with an identification code corresponding to its sampling
location. The code will follow the system described for labeling the monitoring points in
Section 4.2.3 as follows:
[Code for Site] — [Code for Location] — [Depth]
Location codes will include the abbreviations VW for vent well, MP for
monitoring point, BG for background well, or EB for an exploratory boring or other boring
not completed as a vent well, monitoring point, or background well. For the example site #2
at Millersworth AFB the following codes might be used:
• M2—VW—12 for a sample from site #2 at Millersworth AFB
from a depth of 12 ft from the vent well boring
• M2—MPC—28 for a sample from a depth of 28 ft from the
monitoring point C boring
-------
Revision 2
Page: 47
May 14, 1992
M2—BG—4 for a sample from a depth of 4 ft from the back-
ground boring
• M2—EB2—20 for a sample from a depth of 20 ft from the
second exploratory boring, which was subsequently grouted
and not completed as a well or monitoring point.
5.6 Soil Gas Permeability Test Procedures
This section describes the field procedures that will be used to gather data to
determine k and to estimate Rj, The Appendix provides an example data set and calculations
for the radius of influence using the dynamic and steady-state solution methods.
Prior to initiating the soil gas permeability test, the site will be examined for any
wells (or other structures) that will not be used in the test but may serve as vertical conduits
for gas flow. These will be sealed to prevent short-circuiting and to ensure the validity of the
soil gas permeability test.
5.6.1 System Check
Before proceeding with this test, soil gas samples will be collected from the vent
well, the background well, and all monitoring points, and analyzed for 02, C02, and volatile
hydrocarbons. After the blower system has been connected to the vent well and the power
has been hooked up, a brief system check will be performed to ensure proper operation of the
blower and the pressure and airflow gauges, and to measure an initial pressure response at
each monitoring point. This test is essential to ensure that the proper range of Magnehelic™
gauges are available for each monitoring point at the onset of the soil gas permeability test
Generally, a 10- to 15-minute period of air extraction or injection will be sufficient to predict
the magnitude of the pressure response, and the ability of the blower to influence the test
volume.
5.6.2 Soil Gas Permeability Test
After the system check, and when all monitoring point pressures have returned to
zero, the soil gas permeability test will begin. Two people will be required during the initial
hour of this test One person will be responsible for reading the Magnehelic™ gauges, and
the other person will be responsible for recording pressure (P0 vs. time on the example data
sheet (see Appendix Table A-2). This will improve the consistency in reading the gauges and
will reduce confusion. Typically, the following test sequence will be followed:
1. Connect the Magnehelic™ gauges to the top of each moni-
toring point with the stopcock opened. Return the gauges to
zero.
-------
Revision 2
Page: 48
May 14, 1992
2. Tum the blower unit on. and record the starting time to the
nearest second.
3. At 1-minute intervals, record the pressure at each monitor-
ing point beginning at t = 60 s.
4. After 10 minutes, extend the interval to 2 minutes. Return
to the blower unit and record the pressure reading at the
well head, the temperature readings, and the flow rate from
the vent well.
5. After 20 minutes, measure P' at each monitoring point in 3-
minute intervals. Continue to record all blower data at 3-
minute intervals during the first hour of the test.
6. Continue to record monitoring point pressure data at 3-
minute intervals until the 3-minute change in P' is less than
0.1 in. of H20. At this time, a 5- to 20-minute interval can
be used. Review data to ensure accurate data were collected
during the first 20 minutes. If the quality of these data is in
question, tum off the blower, allow all monitoring points to
return to zero pressure, and restart the test
7. Begin to measure pressure at any groundwater monitoring
points that have been converted to monitoring points.
Record all readings, including zero readings and the time of
the measurement Record all blower data at 30-minute
intervals.
8. Once the interval of pressure data collection has increased,
collect soil gas samples from monitoring points and the
blower exhaust (if extraction system), and analyze for 02,
CO2, and hydrocarbons. Continue to gather pressure data
for 4 to 8 hours. The test will normally be continued until
the outermost monitoring point with a pressure reading does
not increase by more than 10% over a 1-hour interval.
9. Calculate the values of k and Rj with the data from the
completed test; use of the Hyperventilate™ computer pro-
gram is recommended. The Appendix shows sample calcu-
lation methods for determining k and R,.
-------
Revision 2
Page: 49
May 14, 1992
5.6.3 Post-Permeability Test Soil Gas Monitoring
Immediately after completion of the permeability test, soil gas samples will be
collected from the vent well, the background well, and all monitoring points, and analyzed for
02. C02, and hydrocarbons. If the 02 concentration in the vent well has increased by 5% or
more, 02 and C02 will be monitored in the vent well in a manner similar to that described
for the monitoring points in the in situ respiration test. (Initial monitoring may be less
frequent.) The monitoring will provide additional in situ respiration data for the site.
5.7 In Situ Respiration Test
The in situ respiration test will be conducted using four screened intervals of the
monitoring points and a background well. The results from this test will determine if in situ
microbial activity is occurring and if it is 02-limited.
5.7.1 Test Implementation
Air with 1 to 2% helium will be injected into the monitoring points and back-
ground well. Following injection, the change of 02, C02, total hydrocarbon, and helium in
the soil gas will be measured over time. Helium will be used as an inen tracer gas to assess
the extent of diffusion of soil gases within the aerated zone. If the background well is
screened over an interval of greater than 10 ft, the required air injection rate may be too high
to allow helium injection. The background monitoring point will be used to monitor natural
degradation of organic matter in the soil. A schematic of the apparatus to be used in the in
situ respiration test is presented in Figure 2-9.
The 02, CO2, and total hydrocarbon levels will be measured at the monitoring
points before air injection. Normally, air will be injected into the ground for at least 20 hours
at rates ranging from 1.0 io 1:7 cfm (60 to 100 cfh). Blowers to be used will be diaphragm
compressors Model 4Z024 from Grainger (or equivalent) with a nominal capacity of 1.7 cfm
(100 cfh) at 10 psi. The helium used as a tracer will be 99% or greater purity, which is
available from most welding supply stores. The flow rate of helium will be adjusted to 0.6 to
1.0 cfh to obtain about 1% in the final air mixture which will be injected into the contaminat-
ed area. Helium in the soil gas will be measured with a Marks Helium Detector Model 9821
(or equivalent) with a.minimum sensitivity of 0.01%.
After air and helium injection is completed, the soil gas will be measured for 02,
C02, helium, and total hydrocarbon. Soil gas will be extracted from the contaminated area
with a soil gas sampling pump system similar to that shown in Figure 5-1. Typically,
measurement of the soil gas will be conducted at 2, 4, 6, and 8 hours and then every 4 to 12
hours, depending on the rate at which the oxygen is utilized. If oxygen uptake is rapid, more
frequent monitoring will be required. If it is slower, less frequent readings will be acceptable.
-------
Revision 2
Page: 50
May 14, 1992
At shallow monitoring points, there is a risk of pulling in atmospheric air in the
process of purging and sampling. Excessive purging and sampling may result in erroneous
readings. There is no benefit in over sampling, and when sampling shallow points, care will
be taken to minimize the volume of air extraction. In these cases, a low-flow extraction
pump of about 0.03 to 0.07 cfm (2.0 to 4.0 cfh) will be used. Field judgment will be
required at each site in determining the sampling frequency. Table 5-1 provides a summary
of the various' parameters which will be measured and their frequency.
The in situ respiration test will be terminated when the oxygen level is about 5%,
or after 5 days of sampling. The temperature of the soil before air injection and after the in
situ respiration test will be recorded.
5.7.2 Data Interpretation
Data from the in situ respiration and air permeability tests will be summarized,
and their Oj utilization rates, air permeability, and Rj will be computed. Further details on
data interpretation are presented in Sections 5.7.2.1 and 5.7.2.2.
5.7.2.1 Oxygen Utilization
Oxygen utilization rates will be determined from the data obtained during the
bioventing tests. The rates will be calculated as the percent change in Oj over time. Table
5-2 contains the two sets of sample data which are illustrated in Figure 5-2. The 02
utilization rate is determined as the slope of the 02% vs. time line. A zero-order respiration
rate as seen in the Fallon NAS data is typical of most sites; however, a fairly rapid change in
oxygen levels may be seen as in the data from Kenai, Alaska. In the later, the oxygen
utilization rate was obtained from the initial linear portion of the respiration curve.
To estimate biodegradation rates of hydrocarbon from the oxygen utilization rates,
a stoichiometric relationship for the oxidation of the hydrocarbon will be used. Hexane will
be used as the representative hydrocarbon, and the stoichiometric relationship used to
determine degradation rates will be:
C6H14 + 9.5 02 -> 6C02 + 7H20
Based on the utilization rates (change of oxygen [%] per day), the biodegradation
rate in terms of mg of hexane-equivalent per kg of soil per day will be estimated using the
following equation.
Kb = - Kq A D0 C/100 (1)
where:
Kb = biodegradation rate (mg/kg day)
K0 = oxygen utilization rate (percent per day)
-------
Revision 2
Page: 51
May 14, 1992
TABLE 5-1. Parameters to be Measured for the In Situ Respiration Tests
Parameter/Media
Suggested Method
Suggested Frequency
Instrument
Sensitivity
(Accuracy)
Carbon dioxide/soil gas
Infrared adsorption method, GasTech
Model 32S20X (0 to 5% and 0 to
25% carbon dioxide)
Initial soil gas sample before pumping
air. immediately after pump shut off,
every 2 hours for Ihe first 8 hours,
and then every 8 lo 10 hours
±0.2%
Oxygen/soil gas
Electrochemical cell method, GasTech
Model 32520X (0 to 21% oxygen)
Same as above
±0.5%
Total hydrocarbons (THC)/soil gas
GasTech hydrocarbon detector or
similar field instrumentation
Initial soil gas sample before pumping
air. then same as above if practical
±1 ppni
Helium
Marks Helium Detector Model 9821
or equivalent
Same as for carbon dioxide
±0.01%
Pressure
Pressure gauge (0 to 30 psia)
During air injection
0.5 psia
Row rate/air
Flowmeter
Reading taken during air injection
±5 cfh
-------
Revision 2
Page: 52
May 14, 1992
A = volume of air/kg of soil (1/kg)
D0 = density of oxygen gas (mg/1)
C = mass ratio of hydrocarbon to oxygen required for mineralization.
Using several assumptions, values for A, D0, and C can be calculated and
substituted into equation 1. Assumptions used for these calculations are:
• Porosity of 0.3 (the air-filled porosity, which can range from 0.0
to 0.6 depending on the site soils and varies with moisture
content in any given soil)
• Soil bulk density of 1,440 kg/m?
TABLE 5-2. Sample Data Set for Two In Situ Respiration Tests
Fallon NAS, Nevada
(Test Well A2)
Kenai, Alaska
(Test Well Kl)
Time
(Hours)
02(%)
C02(%)
Time
(Hours)
02(%)
C02(%)
Helium
-23.5
0.05
20.4
-22.0
3.0
17.5
—
0
20.9
0.05
0
20.9
0.05
¦ 1.8
2.5
20.3
0.08
7.0
11.0
2.7
1.4
5.25
19.8
0.10
12.25
4.8
4.6
1.4
8.75
18.7
0.13
19.50
3.5
6.0
1.3
13.25
18.1
0.16
26.25
1.8
6.5
1.0
22.75
15.3
0.14
46.00
2.0
7.0
0.9
• 27.0
15.2
0.22
32.5
13.8
0.14
37.0
12.9
0.23
46.0
11.2
0.22
49.5
10.6
0.16
-------
Revision 2
Page: 53
May 14, 1992
Time (Hours)
Figure 5-2. In Silu Respiration Test Kesults for Two Uiovenling Test Sites:
Fallon NAS, Nevada (Monitoring Point A2) and
Kenai, Alaska (Monitoring Point Kl).
M/O^/5 Oj
-------
Revision 2
Page: 54
May 14, 1992
D0 oxygen density of 1,330 mg/1 (varies with temperature,
altitude, and atmospheric pressure)
• C, hydrocarbon-to-oxygen ratio of 1/3.5 from the above equation
for hexane.
Based on the above assumed porosity and bulk density, the term A, volume of
air/mg of soil, becomes 300/1,440 = 0.21. The resulting equation is:
KB = - (Ko)(0.21)(1330)(l/3.5)/100 = 0.8 K,, (2)
This conversion factor, 0.8, was used by Hinchee et al. (1991b) in their calcula-
tions of biodegradation rates of hydrocarbons. Another way to estimate biodegradation rates
is based on C02 generation rates, but as discussed in Section 2.3, this is less reliable than
using 02 utilization rates.
5.7.2.2 Helium Monitoring
Figures 5-3 and 5-4 show typical helium data for two test wells. The helium
concentration at monitoring point SI (Figure 5-3) at Tinker AFB started at 1.5% and after 108
hours had dropped to 1.1%, i.e., a fractional loss of -0.25. In contrast, for Kenai K3 (Figure
5-4), the change in helium was rapid (a fractional drop of about 0.& in 7 hours), indicating
that there was possible short-circuiting at this monitoring point. This suggested that the data
from this monitoring point were unreliable, and so the data were not used in calculating
degradation rates.
As a rough estimate, diffusion of gas molecules is inversely proportional to the
square root of the molecular weight of the gas. Based on the molecular weights of 4 and 32
g mol for helium and oxygen, respectively, helium diffuses about 2.8 times faster than
oxygen. This translates into a fractional oxygen loss of -0.095 for SI of Tinker AFB, a
minimal loss. The data from this monitoring point were used in the calculation rates. As a
guide, data from tests where fractional helium loss is 0.4 or less over 100 hours, or an
equivalent fractional oxygen loss of 0.15, are acceptable.
5.8 Bioventing Test
The bioventing test is the third and final part of the field treatability study and
will consist of a longer term (6 months or more) air injection or withdrawal procedure. A
blower will be installed immediately following completion of the air permeability and in situ
respiration tests, and will be started before the field crew leaves the site. At some sites where
regulatory approval is pending, the bioventing blower will be installed and started at a later
date.
-------
Revision 2
Page: 55
May 14, 1992
r- 5.0
50 60 70
Time (Hours)
1—'—i—'—r
80 90 100 110 120
~ 0.0
Figure 5-3. In Situ Respiration Test Results for
Monitoring Point SI, Tinker AFB, Oklahoma.
-------
Revision 2
Page: 56
May 14, 1992
Time (Hours)
Figure 5-4. In Situ Respiration Test Results for
Monitoring Point K3, Kenai, Alaska.
- a*
-------
Revision 2
Page: 57
May 14, 1992
5.8.1 Criteria for Conducting the Bioventing Test
The contractor will plan on conducting the bioventing test at each site; however,
at some sites the bioventing test may not be appropriate (e.g., where no bioremediation is
stimulated). Upon completion of the soil gas permeability and the in situ respiration tests, the
data will be analyzed and a decision will be made as to whether the bioventing test is to be
implemented. This decision will be confirmed before the field crew leaves the site.
5.8.1.1 Air Permeability and Radius of Influence
The technology of soil venting has not advanced far enough to provide firm
quantitative criteria for deteimining the applicability of venting based solely on values of k or
Rj. In general, k must be sufficiently high to allow movement of oxygen in a reasonable time
frame (1 or 2 days) from either the vent well, in the case of injection, or the atmosphere or
uncontaminated soils, in the case of extraction. If such a flow rate cannot be achieved, 02
cannot be supplied at a rate to match its demand.
The estimated radius of influence (Rx) is actually an estimate of the radius in
which measurable soil gas pressures are affected and does not always equate to gas flow. In
highly permeable gravel, for example, significant gas flow can occur well beyond the
measurable radius of influence. On the other hand, in a low-permeability clay a small
pressure gradient may not result in significant gas flow. In this study, the assumption will be
made that the Rj does equate to the area of significant gas flow; however, care must be taken
in applying this assumption. During air permeability testing, an increase in 02 concentration
within the monitoring points is often an additional indicator of Rj.
In general, if the Rj is greater than the depth of the vent well, the site is probably
suitable for bioventing. If the Rj is less than the vent well depth, the question of practicality
arises. To scale up a bioventing project at such a site may require more closely spaced vent
wells than is either economically feasible or physically possible. The decision to proceed
with bioventing will be site-specific and somewhat subjective.
5.8.1.2 Biodegradation Rate
The decision to proceed with the bioventing will be based on the results of the
degradation rate calculations. From previous studies, the oxygen utilization rates that can be
expected from sites contaminated with jet fuel are between 0.05 to 1.0% 02/hour. If rates
within this range are obtained and are significantly greater than background, there is sufficient
evidence to assume that some microbial activity is occurring and that the addition of 02 in
these contaminated areas will enhance biodegradation. If soil gas 02 levels are above 2 to
5% prior to any air injection, or if oxygen utilization rates are not greater than background,
venting will most probably not stimulate biodegradation and consideration will be given to
terminate the bioventing effort.
-------
Revision 2
Page: 58
May 14. 1992
5.8.1.3 Regulatory Approval
Regulatory approval requirements will be defined, and if necessary, approvals will
be obtained prior to initiating the bioventing test procedures. If approval is pending, a blower
will be installed for startup at a later date. This will reduce costs by eliminating the need for
a second visit.
5.8.1.4 U.S. Air Force Approval
Both the project officer and the base POC will be notified either verbally or in
writing of the plans for initiating the bioventing test, and their approval will be required
before the test is initiated. Verbal approval will be documented by the contractor.
5.8.2 Air Injection vs. Extraction Considerations
Air injection will be used as the method of choice to provide oxygen for the
initial and extended pilot tests. Air injection does not result in a direct discharge of volatile
organics to the atmosphere and is less expensive to operate and maintain than extraction
systems. Air injection systems produce no condensate, no liquid wastes, and no contaminated
air stream, and they usually do not require air permitting. Under some circumstances the use
of soil gas extraction systems will need to be incorporated into the air injection system
design. For example, whenever the radius of pressure influence (> 0.1" H20) of a vent well
is close to basements or occupied surface structures, an air extraction system will be used to
reduce the risk of moving gases into these areas. This precaution will prevent the accumula-
tion of explosive or toxic vapors in these structures.
When necessary, soil gas will be extracted away from these structures and then
reinjected in a unsaturated zone well on the opposite side of the extraction well. If necessary,
makeup air will be added prior to reinjection to maintain oxygen levels sufficient for
biodegradation (see Figure 2-3). This configuration will also have the advantage of producing
no direct discharge of volatile organics to the atmosphere, as the volatiles will be returned to
the contaminated zone for treatment by the soil's active biomass.
i
5.8.3 Blower System Installation
On sites where initial pilot testing is successful, and the criteria in Section 5.8.1
are met, a blower system will be installed for the extended bioventing test The blower will
be configured and instrumented as shown in Figure 4-3 or 4-4. This instrumentation will
ensure that important flow rate, temperature, and pressure data can be collected by base
personnel during extended testing. The blower will be sized to provide a soil gas flow that is
sufficient to influence all monitoring points within the contaminated zone and to provide
oxygen at a rate that exceeds the highest oxygen utilization rate measured during initial
testing.
-------
Revision 2
Page: 59
May 14, 1992
Whenever possible, the blower will be sized to use the existing power source at
or near the site. All electrical connections and disconnect devices will conform to local and
base electrical codes. An explosion-proof blower and motor will be required for all extraction
systems and in all fuel storage areas where explosion-proof equipment is mandatory. After
coordination with base officials, the blower will be sited and placed in a secure and unobtru-
sive place. The blower will be placed in a small, portable protective shelter that is painted to
conform to base color schemes. This enclosure will seldom exceed a 3-ft x 4-ft footprint and
a height of 4 ft. The enclosure will protect the motor and blower from the weather and must
be adequately ventilated to prevent the motor from overheating during summer months.
If necessary in high-traffic areas, piping from the vent well to the blower will be
buried several inches below the surface to prevent damage. The blower system, monitoring
points, and piping will be installed so as to minimize interference with existing site activities.
5.8.4 Blower Operation and Maintenance
If the site is selected for extended testing, base personnel will be required to
perform a simple weekly system check to ensure that the blower is operating within its
intended flow rate, pressure, and temperature range. This check must be coordinated with the
base POC. Prior to departing the site, the contractor will provide a 4-hour on-site briefing for
base personnel who will be responsible for blower system checks. The principle of operation
will be explained, and a simple checklist and logbook will be provided for blower data.
Bioventing systems are very simple, with minimal mechanical and electrical parts. Minor
maintenance such as replacing filters or gauges, or draining condensate from knockout
chambers, will be performed by base personnel, but they will not be expected to perform
complicated repairs or analyze gas samples. Replacement filters and gauges will be provided
and shipped to the base by the contractor. Serious problems such as motor or blower failures
will be corrected by the contractor.
5.8.5 Long-Term Monitoring
Most bioventing systems will require 2 or 3 years of operation to significantly
reduce soil hydrocarbon levels. The progress of this system will be monitored by conducting
semiannual respiration tests in the vent well and in each monitoring point, and by regularly
measuring the 02, C02, and hydrocarbon concentrations in the extracted soil gas and
comparing them to background levels. If air injection is used, the blower can be temporarily
reversed and the extracted soil gas monitored for O2, CO2, and hydrocarbons. Soil gas
monitoring will be performed by specialized Air Force or contractor personnel on a quarterly
basis. Semiannual respiration tests will be performed by the Air Force or by contractor
personnel. At least twice each year, the progress of the bioventing test will be reported to the
base POC.
-------
Revision 2
Page: 60
May 14, 1992
6.0 SCHEDULE
The expected schedule for the on-site air permeability, in situ respiration, and
bioventing tests is dependent on the depth to groundwater, as follows:
Case I — (Shallow Groundwater, -20 ft or less) Day After Initiation
— Review available data and develop plan 0-5(a)
— Air Force review 8-12
— Soil gas survey 13-15
— Install vent weliymonitoring points 16-18
— Soil permeability test 19
— In situ respiration test 20-24
— Install blower and start up bioventing system 24-26
Case II — (Deep Groundwater, -20 ft or more)
— Review available data and develop plan 0-5(a)
— Air Force review 8-12
— Exploratory borings 13-15
— Install vent well/monitoring points 16-19
— Soil permeability test 20
— In situ respiration test 21-25
— Install blower and start up bioventing systemb,c 26-27
Case I and II — Bioventing Test Month After Initiation
Determine regulatory requirements^ (if any) 0
Install and stan(c) blower 1
Conduct on-site testing Every 6 months
It will be necessary to begin the process of permitting and contracting
with drillers as soon as possible after contract award, and this must be
nearly complete by day 0.
Regulatory requirements will need to be investigated and any required
permitting or approvals initiated as soon as possible after a site is
identified as a potential candidate. It is assumed in this schedule that
any required permits or approvals will have been obtained prior to
starting.
The blower will be started only after any required regulatory approvals
are received, and with the concurrence of the base POC and project
officer.
-------
Revision 2
Page: 61
May 14. 1992
These schedules are based on the assumptions that (1) no special problems will
be encountered; (2) the sites will be easily accessible; and (3) useable vent well and monitor-
ing point locations will be quickly identified. Any problems or deviations will result in a
longer time frame. Deeper drilling requirements will extend the testing schedule.
-------
Revision 2
Page: 62
May 14, 1992
7.0 REPORTING
The section describes the reports to be generated. For consistency, the following
units will be used:
— English measurements for length, volume, flow, pressure, and mass,
specifically:
feet and inches for length
• gallons and ft3 for volume
• cfh and cfm for flow
• psig for pressure
lb for mass
— Metric units for concentrations, rates, and temperature, specifically:
• mg/1 for aqueous concentrations
• mg/kg for soil concentrations
• mg/(kg day) for hydrocarbon degradatioh
• °C for temperature
— Gaseous concentrations and O2 utilization rates as follows:
• ppm for hydrocarbons (parts per million, i.e., pl/1, by volume)
• percent (%) for O2, C02, and He (percent by volume,
i.e., 1 x 100%/1)
• %/hr for 02 utilization
To avoid confusion when discussing gases, the term percent (%) will refer only to
concentration. Relative changes will be expressed as fractions. For example, if the 02
concentration changes from 20% to 15%, the change will be referred to as a 5% reduction or
a fractional reduction of 0.25, not a 25% reduction.
-------
Revision 2
Page: 63
May 14, 1992
7.1 Test Plan
A Test Plan for each site will be prepared and submitted to the project officer and
the base POC for approval. The Test Plan will consist of this generic Test Plan which
provides the scope and planned activities, and a cover letter describing site-specific applica-
tions. The Test Plan will be submitted to the project officer and base POC as early as
possible before the start of the on-site test.
7.2 Monthly Reports
The contractor will provide a written monthly progress report to the project
officer outlining the work accomplished for the month, the problems encountered, approaches
to overcome the problems, and anticipated progress for the following month. Included in this
report will be the monthly expenditure and the accumulated expenditure to date.
7.3 Verbal Communication
The contractor will be in communication with the project officer and the base
POC and will report on field activities and associated problems. Oral reports will be made
either to the project officer or base POC, upon demand and at least weekly to the project
officer.
7.4 Site Reports
The contractor will provide a letter report (normally less than 15 pages) for each
site describing the results of the soil gas permeability and in situ respiration tests as well as a
description of the bioventing test initiated. This report will normally be submitted to the
project officer, base POC, and others as directed by the project officer 60 days after comple-
tion of the treatability test
-------
Revision 2
Page: 64
May 14, 1992
8.0 RECORD OF DATA AND QUALITY ASSURANCE
A project record book will be maintained during the field tests to record events
pertaining to site activities, including sampling, changes in process conditions (flow, tempera-
ture, and pressure), equipment failure, location of the test wells, calibration, and data for the
respiration/air permeability tests and long-term bioventing test. The record book will be
reviewed by the contractor's project manager. The project officer may review the record
book upon request. Typical record sheets for the respiration and air permeability tests are
shown in Figure 8-1 and 8-2, respectively. Figure 8-3 shows a typical record sheet for the
long-term bioventing test.
Quality assurance will be implemented throughout the project through quality
planning, quality control and quality assessment. This will include daily calibration of field
analytical instrument with purchased calibration standards prior to use. Field blanks will
consist of ambient air drawn through the entire sampling train set-up in an uncontaminated
area of the field site. Quality assurance activities include a review of all field activities and
procedures by the project manager to ensure compliance with this protocol and quality
guidelines. Monthly reports to the project officer will include any significant quality
assurance problems and recommended solutions.
-------
SITE
DATE
Figure 8-1. Typical Record Sheet for In Situ Respiration Test.
MONITORING POINTS
02 METER NO.
Page: 65
May 14, 1992
C02 METER NO.
LOCATION
SAMPLER(S).
HYDROCARBON METER NO.
SHUT DOWN DATE
TIME
11
co2%
o,%
Total
Hydrocarbon
Helium
Comments
Dale/
Time
CO,%
o,%
Tola!
Hydrocarbon
Helium
Comments
-------
Figure 8-2. Typical Record Sheet Tor Air Permeability Test.
Revision L
Page: 66
May 14, 1992
SITE
(DATE
SAMPLER(S)
TYPE OF TEST
TEST DATE _
TIME
Pressure/Vacuum ("HjO)
Distance from
Vent Wdl (It)
Distance from
Vont Wet
Tim*
tn(t)
MP1
MP2
UPS
MP4
Tim*
MO
MPS
UP6
MP7
MPS
-------
SITE
DATE
LOCATION _
SAMPLER(S)
11
CO,*
o,%
ToUl
Hydrocarbon
Ak Row
Rate (chn)
Comments
Data/
Time
COj%
o,%
Total
Hydrocarbon
Air Flow
Rata (chn)
Comments
Revision 2
Figure 8-3. Typical Record Sheel for Long-Term Bioventing Test. May 14 1992
MONITORING POINTS
02 METER NO. C02 METER NO.
HYDROCARBON METER NO.
SHUT DOWN DATE TIME
-------
Revision 2
Page: 68
May 14, 1992
9.0 REFERENCES
Aggarwal, P.K., J.L. Means, and R.E. Hinchee. 1991. "Formulation of Nutrient Solutions for
In Situ Bioremediation." In: R.E. Hinchee and R.F. Olfenbuttel, Eds., In Situ Bioreclamation,
Butterworth-Heinemann, Stoneham, Massachusetts, pp. 51-66.
Anonymous. 1986. "In Situ Reclamation of Petroleum Contaminated Sub-soil by
Subsurface Venting and Enhanced Biodegradation." Research Disclosure, No. 26233, 92-93.
Atlas, R.M. 1986. "Microbial Degradation of Petroleum Hydrocarbons: An Environmental
Perspective," Microbiol. Rev. 45: 180-209.
Bennedsen, M.B., J.P. Scott, and J.D. Hartley. 1987. "Use of Vapor Extraction Systems for
In Situ Removal of Volatile Organic Compounds from Soil." In: Proceedings of National
Conference on Hazardous Wastes and Hazardous Materials, Washington, DC, 92-95.
Conner, J.S. 1988. "Case Study of Soil Venting," Poll. Eng. 7: 74-78.
Dupont, R.R., W. Doucette, and R.E. Hinchee. 1991. "Assessment of In Situ Bioremediation
Potential and the Application of Bioventing at a Fuel-Contaminated Site." In: In Situ and
On-Site Bioreclamation, R.E. Hinchee and R.F. Olfenbuttel, Eds., Butterworth-Heineman,
Stoneham, MA, pp. 262-282.
Ely, D.L., and D.A. Heffner. 1988. Process for In-Situ Biodegradation of Hydrocarbon
Contaminated Soil, U.S. Patent Number 4,765,902.
van Eyk et al. See the alphabetical listing for "V."
Hinchee, R.E., and S.K. Ong. 1991. Unpublished Data, Battelle, Columbus, Ohio.
Hinchee, R.E., and G. Smith. 1991. Unpublished Data, Battelle, Columbus, Ohio.
Hinchee, R.E., D.C. Downey, R.R. Dupont, P. Aggarwal, and R.N. Miller. 1991a. "Enhanc-
ing Biodegradation of Petroleum Hydrocarbon through Soil Venting." J. Hazardous Materials
27: 315-325.
Hinchee, R.E., S.K. Ong, and R. Hoeppel. 1991b. "A Field Treatability Test for Biovent-
ing." Proceedings of the Air and Waste Management Association, Paper 91-19.4, Vancouver,
British Columbia.
Hinchee, R.E., and M. Arthur. 1991. "Bench-Scale Studies of the Soil Aeration Process for
Bioremediation of Petroleum Hydrocarbons." J. Appl. Biochem. Biotech. 28/29: 901-906.
-------
Revision 2
Page: 69
May 14, 1992
Hinchee, R.E., D.C. Downey, J.K. Slaughter, and M. Westray. 1989. Enhanced Biorestor-
ation of Jet Fuels: A Full Scale Test at Eglin Air Force Base, Florida. Air Force Engineer-
ing and Services Center Report ESL/TR/88-78.
Huling, S.G., B.E. Bledose, and M.V. White. 1990. Enhanced Biodegradation Utilizing
Hydrogen Peroxide as a Supplemental Source of Oxygen: A Laboratory and Field Study.
EPA/600-290-006, 48 pp.
Johnson, P.C.. M.W. Kemblowski, and J.D. Colthart. 1990. "Quantitative Analysis for the
Cleanup of Hydrocarbon-Contaminated Soils by In-Situ Soil Venting." Ground Water 28(3),
May-June.
Lee, M.D., et al. 1988. "Biorestoration of Aquifers Contaminated With.Organic Com-
pounds," CRC Critical Reviews in Env. Control 18: 29-89.
Miller, R.N., R.E. Hinchee, and C. Vogel. 1991. "A Field-Scale Investigation of Petroleum
Hydrocarbon Biodegradation in the Vadose Zone Enhanced by Soil Venting at Tyndall AFB,
Florida." in: In Situ and On-Site Bioreclamation, R.E. Hinchee and R. F. Olfenbuttel, Eds.,
Vol. 1., Stoneham, MA: Butterworth Publishers.
Miller, R.N. 1990. A Field Scale Investigation of Enhanced Petroleum Hydrocarbon
Biodegradation in the Vadose Zone Combining Soil Venting as an Oxygen Source with
Moisture and Nutrient Additions, Ph.D. Dissertation. Utah State University, Logan, Utah.
Oak Ridge National Laboratory. 1989. Soil Characteristics: Data Summary, Hill Air Force
Base Building 914 Fuel Spill Soil Venting Project, an unpublished report to the U.S. Air
Force.
Ostendorf, D.W., and D.tf. Kambell. 1989. "Vertical Profiles and Near Surface Traps for
Field Measurement of Volatile Pollution in the Subsurface Environment." In: Proceedings of
NWWA Conference on New Techniques for Quantifying the Physical and Chemical Properties
of Heterogeneous Aquifers, Dallas, TX.
t
Sellers, K„ and C.Y. Fan. 1991. "Soil Vapor Extraction: Air Permeability Testing and
Estimation Methods." In: Proceedings of the 17th RREL Hazardous Waste Research
Symposium, EPA/600/9-91/002, April.
Staatsuitgeverij. 1986. Proceedings of a Workshop, 20-21 March, 1986. Bodembescher-
mingsreeeks No. 9; Biotechnologische Bodemsanering, pp. 31-33. Rapportnr. 851105002,
ISBN 90-12-054133, Ordernr. 250-154-59; Staatsuitgeverij Den Haag: The Netherlands.
Texas Research Institute. 1980. Laboratory Scale Gasoline Spill and Venting Experiment.
American Petroleum Institute, Interim Report No. 7743-5:JST.
-------
Revision 2
Page: 70
May 14, 1992
Texas Research Institute. 1984. Forced Venting to Remove Gasoline Vapor from a Large-
Scale Model Aquifer, American Petroleum Institute. Final Report No. 82101-F:TAV.
Urlings, L.G.C.M., H.B.R.J. van Vree, and W. van der Galien. 1990. "Application of
Biotechnology in Soil Remediation." Envirotech Vienna, pp. 238-251.
van Eyk, J., and C. Vreeken. 1988. "Venting-Mediated Removal of Petrol from Subsurface
Soil Strata as a Result of Stimulated Evaporation and Enhanced Biodegradation." Med. Fac.
Landbouww. Riiksuniv. Gent 53(4b): 1873-1884.
van Eyk, J., and C. Vreeken. 1989a. "Model of Petroleum Mineralization Response to Soil
Aeration to Aid in Site-Specific, In Situ Biological Remediation." In: Groundwater Contami-
nation: Use of Models in Decision-Making, Proceedings of an International Conference on
Groundwater Contamination. Jousma et al., Eds., Kluwer Boston/London, pp. 365-371.
van Eyk, J., and C. Vreeken. 1989b. "Venting-Mediated Removal of Diesel Oil from
Subsurface Soil Strata as a Result of Stimulated Evaporation and Enhanced Biodegradation."
In: Hazardous Waste and Contaminated Sites, Envirotech Vienna, Vol. 2, Session 3, ISBN
389432-009-5, Westarp Wiss., Essen, pp .475-485.
Ward, C.H. 1988. "A Quantitative Demonstration of the Raymond Process for In-Situ
Biorestoration of Contaminated Aquifers." Proceedings of NWWAIAPI Conference on Petro-
leum Hydrocarbons and Organic Chemicals in Groundwater, pp. 723-746.
Wilson, J.T., and C.H. Ward. 1986. "Opportunities for Biorcmediation of Aquifers Contami-
nated with Petroleum Hydrocarbons." J. Ind. Microbiol. 27: 109-116.
-------
appendix
RECOMMENDED ESTIMATION METHODS FOR AIR PERMEABILITY
The U.S. Environmental Protection Agency's Risk Reduction Engineering Laboratory recently
reviewed several field, laboratory, and empirical methods for determining soil gas permeability (k) and
for their appropriateness in determining the feasibility of soil vapor extraction (Sellers and Fan, 1991).
The conclusion of this literature review was a strong endorsement for a modified field drawdown
method (Johnson et al.. L990).
The field drawdown method is based on Darcy's Law and equations for steady-state radial flow to
or from a vent well. A full mathematical development of this method and supporting calculations are
provided by Johnson et al. (1990). A computer program known as Hyperventilate™ has been
produced by Johnson for storing field data and computing k and R,. This program will be used to
speed the calculation and data presentation process. The two solution methods for k are presented
below. The first solution is based on carefully measuring the dynamic response of the soil to a
constant injection or extraction rate. The second solution for k is based on steady-state conditions and
the measurement or estimation of Rj at steady state. The limitations and recommended application of
each method are presented below. Whenever possible, field data will be collected to support both
solution methods, because one or both of the solution methods may be appropriate, depending on site-
specific conditions.
Dynamic Method
This test method requires that air be extracted or injected at a constant rate from a single venting
well, while measuring the pressure changes at several soil gas monitoring points throughout the
contaminated soil volume. The equation:
Q [-0.5772 -,ln (r2 ep) + ln(t)]
P' = (
4k m(k/p) 4k Patm
is used to describe the dynamic changes in soil gas pressure/vacuum where:
P.
-
"gauge" pressure measured at distance r from the vent well at time t(g/cm-s2)
m
stratum thickness, generally the vent well screened interval (cm)
r
=
radial distance from monitoring point to vent well (cm)
k
=
soil gas permeability (cm2)
P
=
viscosity of air (1.8 x 10"4 g/cm-s at 18°C)
e
=
soil's air-filled void volume (dimensionless)
t
=
time from the start of the test (s)
Q
=
volumetric flow rate from th$ vent well (cm3/s)
Patm
=
ambient pressure (at sea levd 1.013 x 106 g/cm-s:)
-------
Revision 2
Page: 72
May 14, 1992
Equation (1) predicts that the dynamic range of P'-vs.-ln(t) is a straight line with a slope of A
where:
A =
4jcm (k/p)
solving
Qy
k =
4Ajcm
The Hyperventilate™ model is based on the dynamic method and a determination of the slope. A.
This method of determining k requires accurate field measurements of Q at the vent well and P's-vs.-
time at each monitoring point. It is most appropriately applied at sites with less permeable soils where
changes in V occur over a longer time period (10 minutes or more to monitoring point steady state).
This method can be accurate for fine sandy soils where the screened interval extends to depths of over
10 ft and when monitoring points are screened at depths of 10 ft or greater. It is less accurate for sites
where a high water table or shallow contamination limits the total depth of the vent well screen and
monitoring points to less than 10 ft. In shallow and coarse-grained soils, vacuum or pressure levels
reach steady state too rapidly to accurately plot P'-vs.-ln(t). Venting systems on shallow sandy sites
are subject to higher vertical airflow which is not as accurately described by this one-dimensional
radial flow equation.
Steady State-Method
This method for determining k can be used in situations where the dynamic method is inappropri-
ate. This method is based on the steady-state solution to equation (1).
k _ Qp ln(Rw/R,) (2)
Hr Pw [ 1 - (Patm/Pw)2]
Note: Equation (2) applies only to vent wells operating under a vacuum. If air is being injected
into the vent well the equation is modified as shown below:
k _ Qfi ln(Rw/Rj) (3)
Htc Patm [1 - (Pw/Patm)2]
where Q, m. p. and Patm have been previously defined, and
Rw = the radius of the venting well (cm)
-------
Revision 2
Page: 73
May 14, 1992
H = depth of screen (cm)
R| = the maximum radius of venting influence at steady state (cm)
Pw = the absolute pressure at the venting well (g/cm-s2)
The value of Rj can be determined by actually measuring the outer limit of vacuum/pressure
influence under steady-state conditions, or by plotting the vacuum/pressure at each monitoring point
vs. the log of its radial distance from the vent well and extrapolating the straight line to zero vacuum
or pressure. An example of this solution method is included in Calculation Data Set Two below.
Sample Calculations
Data Set One
Table A-l and Figure A-l present the results of an air permeability test conducted at Beale AFB.
CA. The soils on this site were silty with a contaminated interval (and vent well screen interval)
extending from 10 to 40 feet below ground surface. Note that the plot of P'-vs.-ln(time) is a relatively
straight line during the initial 10 minutes, In (10) = 2.3. making these data good candidates for the
dynamic solution method. Data from the initial 10 minutes of this test were entered into the Hyper-
Ventilate™ computer model to calculate a range of k values. An example of the input and output data
for this model is provided in windows AP7 and AP8.
Hyp«rV«ntjlat*6 1991
Air Permeability Test - Data. Analysis (cont.)
The permeability, k, can then be calculated by one of tvo methods:
fpv The first is applicable vben both Q (flovrate) and m (veil screen interval) are
koovn accurately. The calculated slope A is used:
t_ Qm>
4A«o
© The second approach is used vhenever Q or m are not knovn vith confidence.
In this case, both the slope, A, and intercept, B, are used:
k-4^-esp[ 0.5772 + 1-)]
4P*a A
a?7 i
-------
Revision 2
Page: 74
May 14, 1992
Air Permeability Test - Data Analysis (cont.)
1*1 20 I (ft)
(mm) (In H20)
Enter radial
(T) distances of
monitoring points
40 l(ft)
(mm) (in H20)
10
(ft)
Enter measured —«
times and gauge
vacuums
Enter (optional):
a) flovrate
' (SCFM)
51
b) screened interval
thickness
(ft)
30
C clear ^
.5
0.1
O
.5
.4
K>
1
0.21
1
1.4
1.5
0.62
1.5
2.8
ijjjji
ii&i
2
1.00
i'i'il
iiijij
2
3.6
2.5
1.2S
ill!!
2.5
4
liiiji
3
1.41
Ill:
ipS;
3
4.4
3.5
1.60
3.5
5
4
1.8
iiiii
4
5.3
ill!
4.5
1.98
ii
4.5
5.6
:!${(
5
2.12
o
5
5.8
o
(min)
(in H20)
.5
1.5
1
4.5
1.5
7.5
|l!l|
2
9
iiijli!
2.5
10
•ilii:
«h
3
10.7
i >:
3.5
11.2
J i i
4
11.8
4.5
12
4 i
;t \
5
12.4
o
C clear ^
G
>C«lCUl&tft<
j:
k«
14.2021
darcy (A) k-
6.75944
darcy(A) k-
4.00444
k-
84.6266
daicy(B) k-
34.6443
darcy (B) k-
15.9240
dairy (A)
daicv(B)
a:
Return
Explanation & Statistics 1 AP8 1
Air Permeability Test - Data Analysis (cont.)
__ Enter radial
distances of
monitoring points
Enter measured —>
times and gauge
vttcuuzns
(3^ Enter (optional):
a) flovrate
i* I 40 kfQ
(min) (inK20)
1 51 I (SCFM)
b) screened interval
thlctaiess
W
30
5.5
2.2S g
6
2.37 |
6.5
2.48 jj
7
2.55 |
7.5
2.63
8.5
2.82 f]
9.5
2.92 1
o
i" I 20 \C«lcuIat»o
k-
14.2021
darcy(A) k«
6.75944
daicy (A) k-
4.00444
k»
84.6266
dairy (B) k-
34.6443
daxcy (B) k-
15.9240
dairy (A)
darcy(B)
~1 Return Explanation & Statistics | AP8
HyparV«ntilat«C 1991
-------
TABLE A-l. Air Permeability Data Sel
[ Stcatl/-S>nlc lltnv Hale 31SCTM
Test
Time In
niapsetl 'lime
(miii) (min)
Vacuum (inches of water) *1
M»nilofiiig Point* ( MPs)
Ml* 4 Ml'5 Ml* 6 MP 7
0.0
—
0.5
-
1.0
0.00
1.5
0.41
2.0
0.69
2.5
0.92
3.0
1.10
3.5
1.25
4.0
1.39
4.5
1.50
5.0
1.61
5.5
1.70
6.0 -
1.79
6.5
1.87
7.0
1.9 J
7.5
2.01
8.5
2.14
9.5
2.25
0.5
2.35
4.0
2.64
19.0
2.94
!4.0
3.18
19.0
3.37
14.0
3.53
.19.0
3.66
44.0
3.78
0.3
08
0.4
0.7
2.2
1.7
0.0(1
0.10
0.21
0.62
1.00
1.25
1.41
1.60
1.80
1.98
2.12
2.25
2.37
2.48
2.55
2.63
2.82
2.92
2.96
3.00
3.05
3.10
3.37
3.40
3.40
27.5-2«>.5 18-20 13-15 14-16 38-40
40
30-32 38-40
Kcvisioii I
Page: 75
May 14, 1992
MP 8 MP9
0.00
0.
0.40
1.50
1.40
4.50
2.80
7.50
3.60
9.1X1
4.00
10.00
4.40
10.70
5.00
11.20
5.30
11.80
5.60
12.00
5.80
12.40
6.00
12.50
6.10
12/.0
6.20
12.60
6.30
12.70
6.40
12.70
6.50
12.40
6.50
12.50
6.50
12.50
6.50
12.40
6.40
11.90
6.20
11. (X)
6.00
10.40
5.80
9.90
20 10 <— Distance from VE - 2
3R—40 38-40 < S'rrccn hitoval «leptl«
-------
VACUUM
¦ (in—lljO)
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
0.0
0.5
000
£033 B0E
1.0
1.5 2.0 2.5
In TIMli (minutes)
Monitoring Point 9
Monitoring Point 8
Monitoring Point 7
3.0
3.5
4.0
Revision 2
I'age: 76
May 14, 1W2
Figure A-1. Vacuum vs. In Time,
Test 2, llioventing Pilot Test,
Site 22-A20, Iteale AKIl, California
-------
Revision 2
Page: 77
May 14, 1992
Computer window AP7 provides a summary of two mathematical solutions for air permeability (k)
using the dynamic method. Window AP8 is the example data entry and solution sheet. The
calculated range of k values for this test is shown at the bottom of window AP8. Permeability values
of 4 to 14 darcy are based on Equation 1 in window API and provide the most accurate estimate,
because both the extraction rate (Q) and the screened interval (m) were known for this test. The more
conservative range of 4 to 14 darcy will be used for full-scale design. These air permeability values
are approximately one order of magnitude higher than would be expected for silty soils. The presence
of 10 to 15% sand (by weight) in this soil has increased the average permeability at this site.
Data Set Two
Table A-2 and Figure A-2 are the results from a test conducted in a silty loam with a contaminated
interval of only 5.2 ft and a screened interval from 2.7 to 5.2 ft below ground surface. Note that the
almost immediate steady state reached at this site does not produce the P'-vs.-ln(time) plot required for
the dynamic solution method. In this case the steady-state solution offers the only approximation of k
and R].
k Qm ln(Rw/Rj)
Hti Ew [ 1 - (Patm/Pw)2)
For this test;
Q = 1.4 x 104 cmJ/s
H = 2 ft (61 cm)
p = 1.8 x lO"4 g/cm-s
Pw = 80"H2O vacuum x 3.61 x 10'2 psia = 2.88 psia
mH20
Pw absolute = 14.7 psia - 2.88 psia = 11.82 psia
11.82 psia x 6.9 x lOVcm-s2 = 8.16 x lOVcm-s2
psia
Patm = 1.01 x lOfycm-s2
Rw = 1 ia = 2.54 cm
R, = -15 ft (457 cm) based on all monitoring points reported in Table A-2
-------
Revision 2
Page: 78
May 14, 1992
.TABLE A-2. Field Test Data for Suil Determination of Soil Permeability
at a Gasoline-Contaminated Site
Time
(min)
Air
Flow
(cfm)
Vacuum (inches of water) measured at various monitoring points
Unit
Well
F
E
G
D
11
c
I
B
A
0.0
0
0
2
0.10
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.5
30
109
80
1.90
0.90
0.25
0.15
0.00
0.00
0.00
0.00
0.00
1.5
30
109
80
1.90
0.90
0.30
0.20
0.05
0.00
0.00
0.00
0.00
5.0
30
109
80
1.90
0.90
0.30
0.20
0.05
0.00
0.00
0.00
0.00
10.0
30
109
80
1.90
0.95
0.30
0.20
0.05
0.00
0.00
0.00
0.00
15.0
30
109
80
1.90
0.95
0.30
0.20
0.05
0.00
0.00
0.00
0.00
20.0
30
109
80
1.90
0.95
0.30
0.20
0.05
0.00
0.00
0.00
0.00
Distance
from well
(ft)
3
6
9
12
15
18
21
24
27
Rw
II
0
= 2.54 cm
= 60.96 cm
= 14,158 cinYsec
p = 1.8 x 104 g/cm-s
Palm = 8.14 x 10s Dyncs/cm2
-------
ICaNVTEST wkj
Revision 2
Page: 79
May 14, 1992
I
Figure A-2. Results of a Field Test
to Determine Soil Permeability
to Airflow, k, September 16, 1991
-------
Revision 2
Page: 80
May 14, 1992
k = (1.4 x I04cm3/s)( 1.8 x 10"'g/cm-s)ln(2.54/457)
(61 cm)(3.14)(8.16 x 105g/cm-s)(l - [ 1.01/0.816]2)
k = 1.6 x 10'7 cm2 or 0.16 darcv. which is typical for silty soils.
References
Johnson, P.C., M.W. Kemblowski. and J.D. Colthart. 1990. "Quantitative Analysis for the Cleanup of
Hydrocarbon-Contaminated Soils by In-Situ Soil Venting." Ground Water 28(3), May-June.
Sellers. K.. and C.Y. Fan. 1991. "Soil Vapor Extraction: Air Permeability Testing and Estimation
Methods." In: Proceedings of the 17th RREL Hazardous Waste Research Symposium.
EPA/600/991/002. April.
-------
REPORT
FINAL REPORT
BIOVENTING AND SOIL
VENTING (VACUUM
EXTRACTION)
TREATABILITY TESTS
FOR GREENWOOD
CHEMICAL SUPERFUND
SITE, ALBEMARLE
COUNTY, VIRGINIA
To
U.S. EPA Environmental Protection Agency
Risk Reduction Engineering Laboratory
Cincinnati, Ohio
llBaiteise
. . . Putting Technology To Wbrk
December 3, 1991
-------
-------
CONCENTRATIONS of oxygen, carbon dioxioe, total hydrocarbons and
ACETONE IN MONITORING WELLS 24 HOURS AFTER INSTALLATION
well
Depth
(feet)
$
%
THC
(ppm)
Acetone
(ppm)
C5
5
2.5
12.5
>10,000
--
C6
5
0.5
16
>10,000
1S09
15
0.8
7.5
>10,000
2553
CCIO
5
-
no readings
taken
-
15
#-
no readings
taken
—*
U1
4
10.5
9
>10,000
--
11
2.5
19
>10,000
2420
19
2.5
16
>10,000
2309
-------
Well C6 - 5 feet deep
Time (hrs)
In situ Respiration Test Results for Well C6 (5 feet)
-------
Well U1-11 Feet Deep
0 20 40 60 80 100
Time (hrs)
In situ Respiration Test Results for Well U1 (11 feet)
-------
Well U1 -19 feet deep
t
Oxygen • B/Ground li ft.
Oxygen • Ul
4
Carbon Dioxide • B/Ground IS ft ^
Carbon Dioxide • L'l
f^n p i ,¦ , , p
20 40 60 80 100
Time (hrs)
In situ Respiration Test Results for Well Ul (19 feet)
-------
Air Permeability Test Data for Well CC12
In (t (sees))
Air Permeability Test Oata for Well CC12
-------
Air Permeability Test Data for Well CC13
In (t(sees))
Air Permeability Test Data for Well CC13
-------
Air Permeability Test Data for Well CC10
In (t (sees))
Air Permeability Test Data for Well CC10
-------
Radius of Influence Plot
In (r (ft))
Plot to Compute the Radius of Influence
------- |