In Situ Treatment Technology - Chapter 3 pptx

Rorech CONTENTS IntroductionContaminant Partitioning in the SubsurfaceAirflow Requirements and CapabilitiesAirflow Capability Airflow RequirementsEvaluation of Conditions Where VES is Appli

Rorech, Gregory J "Vapor Extraction and Bioventing"

In Situ Treatment Technology

Boca Raton: CRC Press LLC,2001

CHAPTER 3 Vapor Extraction and Bioventing

Gregory J Rorech

CONTENTS

IntroductionContaminant Partitioning in the SubsurfaceAirflow Requirements and CapabilitiesAirflow Capability

Airflow RequirementsEvaluation of Conditions Where VES is ApplicableContaminant Properties

Vapor PressureSolubilityHenry’s LawOther Molecular PropertiesSummary

Properties of the SoilBulk Density/Soil PorositySoil Adsorption

Soil MoistureSite Surface TopographyDepth to Water TableSite HomogeneityModeling Tools for Vapor Extraction System DesignEngineering Design Model

Flow ModelsMultiphase Transport ModelsPilot Studies

Laboratory Studies

Field Pilot StudiesVapor Extraction Testing WellVapor Extraction Monitoring WellsStage 1: Pilot Test PlanningStage 2: Conducting the Pilot TestStage 3: Evaluating the DataEvaluation of Pneumatic PermeabilityExtraction Well Placement

System DesignBioventingIntroductionAdvantages of Vapor Phase BiotreatmentPerformance Criteria and Bioventing Plan ProtocolsLaboratory Testing

Field Respirometry TestingSoil Gas Permeability TestingBioventing System ConfigurationsCleanup Goals/Costs

Case StudyDesign/Operating CriteriaDesign Example

ProblemSolutionPilot Test PlanningConducting the Pilot TestEvaluating the DataReferences

INTRODUCTION

The vapor extraction and bioventing technologies induce airflow in the face using an above-ground vacuum blower/pump system Adequate air movementwithin the contaminated zones is of primary importance to the success of the VES.The induced airflow brings clean air in contact with the contaminated soil, NAPL,and soil moisture The contaminated soil gas is drawn off by the VES and the air

subsur-in the soil matrix becomes recharged with new vapor phase contamsubsur-ination as thesoil/pore water/soil gas/NAPL partitioning is re-established

Bioventing, or bioenhanced vapor extraction, is a remedial method similar tovapor extraction in that it relies upon an increase in the flow of air through thevadose zone Vapor extraction is performed to volatilize the volatile organic constit-uents in situ In bioventing, the increase in the flow of air is to provide oxygen inthe subsurface to optimize natural aerobic biodegradation, and this becomes thedominant remedial process While the design criteria for vapor extraction and bio-venting are different, once the physical system is in operation both processes occur.Compounds that are volatile move with the air, and compounds that are degradable

have an increased rate of degradation The design sections of this chapter will showhow to set up the process so that one process is favored over the other.

Figure1 shows the basic components of a VES Subsurface vapors are withdrawnthrough an extraction well that may be vertically or laterally constructed Recoveredvapors are routed to an above-ground vapor treatment unit, if required The key to

a successful design is to place the wells and equipment so that when the system is

in operation, an airflow pattern is created across the entire section of the unsaturatedzone that is contaminated The designer must also be careful that the air does notmove through a small percentage of the area due to porous geology (short circuiting)

The distribution of contaminants within the four phases (NAPL/vapor phase/soilmoisture/adsorbed to soil) can be represented by mathematical equations, and com-puter models can simulate the distribution For large projects where accurate pre-diction of system performance is required, pilot testing of the VES is often per-formed The design engineer must gather information during the pilot testingactivities or the predesign phase of the project to calibrate the simulation models.This will, in turn, aid in ascertaining achievable closure criteria to be compared withrisk assessment goals, closure timeframes, and optimal system operating parameters(vacuum, flow rates, moisture content, well location, well screening, number ofwells) Projects for which this level of sophistication is conducted may includelitigation-oriented Superfund projects or projects where the site owner desires accu-rate prediction of the lifetime of the remedial system

For small projects, such as a gasoline service stations with shallow vadosecontamination in a high permeability formation, modeling is often not conducted

If simulation models are not utilized, the design engineer must choose the requiredsubsurface airflow velocities to achieve cleanup goals using published literature

Figure 1 Basic VES.

values or other empirical design concepts At a typical service station type of VES,the engineer is often constrained by subsurface structures for well locations Due tothese constraints, the systems may not be optimally designed and are often overde-signed to overcome site access limitations.

A vacuum pump or blower is the tool that is used to create subsurface airflow(Figure 1) The vacuum created at the extraction well head is an indication of thesubsurface soil resistance to airflow If the subsurface is very porous there will bevery little vacuum at the extraction well head regardless of vacuum pump that isutilized (i.e., low vacuum capability regenerative blower or high vacuum liquid-ringpump) If there is little resistance, there will be little resultant vacuum If oneincreases the flow from a given well, a higher vacuum will be required at theextraction well head because of increased subsurface resistance created by the higherflow Vacuum is an indication of subsurface resistance to flow and is not the variablethat is critical to VES success Sufficient airflow is the critical variable The vacuumapplication is a system operational parameter that allows creation of the desiredairflow

This chapter will discuss the theory of vapor extraction, site and chemicalparameters that are used to predict its applicability, modeling of VES, pilot testing,system design criteria, and the biological enhancement that results from vapor flow.Vapor phase treatment options will be discussed in Chapter 6

CONTAMINANT PARTITIONING IN THE SUBSURFACE

Contaminants that are released to the environment will be distributed in thesubsurface in a manner consistent with their physical properties (vapor pressure,solubility, etc.) as well as related to properties of the soil (type, organic content,moisture extent, etc.) This subsurface distribution (in pore water, vapor, adsorbed

to soil, or NAPL) is termed partitioning

Transfer of the contaminants between phases (dissolved, vapor, adsorbed, andfree phase) is affected by the relative affinity of the contaminant to each phase.These affinities can be evaluated using the constituent partitioning coefficients tothe various phases The interphase partitioning coefficients can be expressed as theconcentration ratio of the constituents in each phase, and this ratio is dictated bythe equilibrium relations in the subsurface (Equations (1)-(3)) Since, to a largeextent, the interphase transfer is governed by these equilibrium partitioning relation-ships, the most effective remediation will create the subsurface conditions that willdrive the interphase transfer towards the phases that allow for the most efficient massremoval Site remediation can therefore be viewed as implementing changes orperturbations to the subsurface that will drive chemical and/or biological processestowards the site remediation goals (Sims 1990) The subsurface change that iseffected during vapor extraction is the replenishment of the subsurface soil vaporutilizing air as the carrier, and therefore, driving the contamination to the vaporphase where it is collected for above-ground treatment Figure2 is a schematic ofthe environmental compartment model showing the goals of vapor extraction

When a volatile NAPL is present on the soil, the bulk of the mass removal byVES will come from direct volatilization of the NAPL This would be similar to afan blowing past a pool of gasoline Research workers (Hoag et al 1989) have shownthat the bulk (more than 95 percent) of the NAPL can be removed within passage

of several hundred pore volumes of air through the experimental soil columns (NAPLwithin the dry soil void space) In field applications, where airflow is usually overthe NAPL layer rather than through it, VES still often recovers the bulk of the NAPLwith passage of several hundred pore volumes of air Under conditions of NAPLpresence, mass removal rates are often linearly correlated with airflow rates WhenNAPL is not present in the subsurface, airflow requirements become very different,and are often governed by nonequilibrium rate limiting conditions

The following section describes the partitioning of contaminants in the face These equations are the basis of numerical simulation models that attempt topredict the remediation process of vapor extraction Models achieve this prediction

subsur-by repeatedly evaluating the new partitioning relationships after passage of cleanair past the contaminated soil Modeling is discussed later in this chapter

Under moist soil conditions, contaminant partitioning in the vadose zone (noresidual NAPL) can be described by the following equation (1) Figure 3 is anillustration of the equation

(1)

where CT = Total quantity of chemical per unit soil volume; CA = Adsorbed chemicalconcentration; CL = Dissolved chemical concentration; CG = Vapor concentration;

pb = Soil bulk density; θL = Volumetric water content; and θG = Volumetric air content

Figure 2 Environmental compartment model for VES.

C T = p b C A+θL C L+θG C G

The equilibrium relationship between vapor concentration (CG) and the ated pore water concentration (CL) is given by Henry’s Law

associ-(2)

Henry’s Law is often thought of in relation to air stripping In air stripping, airremoves dissolved VOCs from the water stream The efficiency of this removalprocess is related to the compound’s Henry’s constant Under moist soil conditions,the extracted vapors during vapor extraction similarly remove VOCs from water.The success of this removal process is similarly related to the compound’s Henry’sconstant

Likewise, the relationship between equilibrium solution concentration andadsorbed concentration is given by

(3)

where Kd (L3/M) = the distribution coefficient expressed as Kd = focKoc, andwhere

foc = the mass fraction of organic carbon, and Koc = the organic carbon partitioningcoefficient

Equation (3) is generally considered to be valid for soils with high organiccontent (foc>0.1 percent solids) For soils with lower organic carbon content (foc <0.1percent solids), sorption to mineral grains may be dominant (Piwoni and Bannerjee

1989 and Brusseau, Jessup, and Rao 1991) Adsorption is further discussed later inthe chapter There are several good sources for organic carbon partitioning coefficient

Figure 3 Equation schematic.

C G = K H C L

C A = K d C L

(Koc) values (U.S EPA 1996 and Fetter 1994) The EPA document checked variousliterature sources and then calculated the geometric mean of what they consideredthe most reliable sources This document also provided some relationships forvarious types of compounds to estimate Koc , if Kow is known, as presented below

in Equations (4) and (5) Other relationships for these and other types of compoundsare also available in the literature with some slight variation in each source.For polyaromatic hydrocarbons, polychlorinated biphenols and phthalates

(4)For volatile organic compounds and chlorinated pesticides

(5)

Mathematical relationships can also be developed to quantify biological and/orother reaction transformations of the chemical contaminants These relationshipscan subsequently be used to develop mathematical models of the subsurface condi-tions under advective air movement conditions Mathematical models can be used

to simulate subsurface changes caused by the VES (perturbation) and allow the user

to select the most efficient perturbation that leads to the most contaminant massremoval

The above equations have defined partitioning of contaminants in the subsurfacewithout air movement Partitioning without advective air movement occurs viadiffusion The VES induces airflow (advective) past the contaminated zone; there-fore, under VES operating conditions, both diffusive and advective transport areoccurring

Under the assumptions of uniform moisture distribution across the soil andincompressible air phase, the advective-dispersive transport equation in cartesiancoordinates can be written as (Armstrong, Frind, and McClennan 1994, and Gierke,Hutzler, and McKenzie 1989)

(6)

where subscripts G, L, and A designate the gaseous, dissolved, and sorbed phases

of the contaminant CA, CL, CG, Pb., θG, and θL are as defined above The air velocitycomponent is derived from the air continuity equation and the subsequent application

of the Darcy equation to pressure The continuity equation states that the same mass

of material entering a unit volume of space must also exit that volume space withoutbiodegradability The Darcy equation relates groundwater velocity to hydraulic gra-

movement rather than groundwater The term Dij is the dispersion tensor, defined interms of longitudinal and transverse dispersivity and the diffusion coefficient Equa-

K oc= logK ow–0.094log

K oc =0.78 logK ow+0.151log

tions (1), (2), and (3) can be substituted into Equation (6) in order to represent thetransport equation in terms of the gaseous phase only to yield

(7)

(8)

where R is the retardation factor

Equations (7) and (8) therefore, are a mathematical representation of ing under diffusive and advective conditions Actual field observations, however,indicate that the equations are only valid for diffusion dominated or weakly advec-tive conditions

partition-Under strongly advective conditions that can be found while operating a VES,the above equations do not account for the asymptote that is observed in fieldapplications Tailing is the phenomenon that is often observed in VES applicationswhereby the contaminant mass removal rates are slower and the residual mass ofcontaminant adsorbed to the soil after vapor extraction is greater than what would

be predicted by the equilibrium equation, Equation (8) Equilibrium predicted tailing and nonequilibrium type tailing effects are shown schematically in Figure 4

non-Figure 4 Equilibrium asymptotic tailing effect.

As discussed in Chapter 2, the performance of a system in removing

contami-nants changes with time In most VES applications, the initial stages of the project

yield the highest mass removal The tailing effect implies that efficiency is decreasing

with system lifetime and that closure goals or system operational modifications

should account for this temporal change

Several authors (Armstrong, Frind, and McClennan 1994; Gierke, Hutzler, and

McKenzie 1989; Sleep and Syikey 1989; and Brusseau, Jessup, and Rao 1991) have

shown that the tailing effects can be represented by first order mass transfer,

non-equilibrium, physical, and/or chemical processes These nonequilibrium

modifica-tions to the equilibrium relamodifica-tionships (Equamodifica-tions (1)-(3), (6)-(8)) will not be

pre-sented in this chapter; however, the reader is referenced to the appropriate research

publications for details In brief, the nonequilibrium notions relate to the existence

of a rate limiting criteria governing the mass transfer process for VES Under fully

wetted soil conditions without NAPL, the rate limiting step may be transfer across

the air/water interphase, the soil/water interphase, dead end micropore effects, and/or

a combination of all effects Simply stated, these nonequilibrium effects slow down

the vapor extraction process once the bulk of the contamination has been removed

(Armstrong, Frind, and McClennan 1994) have conducted a sensitivity analysis on

several of these nonequilibrium rate limiting conditions

A sensitivity analysis or demonstration of the physical limitation of the VES

ability is a powerful tool in ascertaining the system’s capability and achievable

closure goals This can be utilized to negotiate reasonable closure criteria with

regulatory agencies and/or modify system operation during the project’s lifetime to

minimize expenses while maximizing mass removal VES is a very powerful mass

removal technique However, as the historic research and long-term operations have

shown, VES can not overcome the natural limitations of the geology Once most of

the mass has been removed by the VES, an alternative treatment technique may have

to finish the removal

AIRFLOW REQUIREMENTS AND CAPABILITIES

The need to understand and predict the subsurface mass transfer relationships

relates to the practical need to deliver the required airflow to achieve the remedial

goals Often the designer will only want the minimum subsurface air movement to

achieve the remedial goals, since excessive airflow results in larger, expensive off

gas treatment equipment as well as higher operating costs

In instances where NAPL is present in pockets, pools, or as a layer atop the

groundwater, mass removal will often be linearly or semi-linearly rated to the airflow

This does not imply that if NAPL is present, high airflow is required, since often

the NAPL is removed rapidly, leading to site conditions that may not require further

high airflow Airflow generation capability (airflow that can be generated based upon

subsurface soil conditions) and airflow requirement (to achieve remediation) needs

must be met in order to appropriately install a VES

Airflow Capability

Figure5 presents predicted airflow rates per unit well screen (from an extraction

well) depth for a 4-inch diameter extraction well and a wide range of soil

perme-abilities and applied vacuums (Johnson et al 1990, and Johnson et al 1991) The

graph was generated by solving the logarithmic Jacob’s equation and assuming a

40 foot radius of air collection (i.e., zone of influence) The Jacob’s equation is the

same Jacob’s equation used in groundwater applications to determine zone of

influ-ence of a pumping test In this instance, however, it is utilized for subsurface airflow

rather than groundwater flow This figure provides an excellent screening tool to

determine the necessary vacuum equipment to generate the required airflow This

will be further discussed later in the chapter

The measurement of vacuum at locations away from the vapor extraction well

implies that subsurface airflow is present at that location Subsurface vacuum is easy

to measure in the field, therefore, it is often measured in place of subsurface airflow

during VES pilot testing and application

Figure6 illustrates how quickly vacuum measurement profiles die off away from

the extraction well point This quick decay of induced vacuum readings in turn

implies that subsurface airflow quickly dies off at increasing distances from the

extraction well If the airflow was perfectly radial from the well, then the airflow

velocity would decrease at a rate proportional to the square root of the distance from

the well Several design techniques will be discussed later in the chapter that focus

on maintaining the airflow across the contaminated area of the vadose zone and not

simply in a radial flow from the extraction well This will help produce the required

airflow with a minimum of equipment

Figure 5 Airflow generation plot.

Airflow Requirements

Delivering the required airflow requirements to achieve the cleanup criteria is

the basic design goal for VES installation This basic design goal, however, remains

the most difficult to predict due to our limited understanding of subsurface

condi-tions This understanding is required to quantify the mathematical relationships used

in formulating the simulation models The most distant location from the extraction

well should receive sufficient airflow to achieve remediation The most distant

location from the extraction well is termed the zone of influence Sufficient wells

are spaced in the contaminated area to deliver the minimally acceptable airflows

across the entire site

The required airflow at the most distant location can be determined by (1)

comparison of published literature values for similar soil types and contaminants;

(2) by conducting bench scale column tests to determine this required airflow for

contaminants and soil conditions (discussed later in the chapter); or (3) by conducting

computer simulation modeling (discussed later in the chapter)

EVALUATION OF CONDITIONS WHERE VES IS APPLICABLE

Vapor extraction system efficiency is affected by parameters relating to the

contaminants to be removed and by variables relating to the site to be remediated

Contaminant properties that affect VES are vapor pressure, solubility, Henry’s Law

constant, biodegradability, and other molecular structure properties (Jury et al 1990)

Vapor pressure and solubility tend to be the most important chemical parameters

affecting VES performance Soil properties that affect VES performance include soil

Figure 6 Vacuum tailing effect.

porosity, soil adsorption, soil moisture, site topography, depth to water, and site

homogeneity Of all the soil properties, permeability tends to be the most critical

parameter relating to system success

CONTAMINANT PROPERTIES Vapor Pressure

Vapor pressure is the parameter that can be used to estimate a compound’s

tendency to volatilize and partition into the gaseous state The vapor pressure of a

compound is defined as the pressure exerted by the vapor at equilibrium with the

liquid phase of the compound in the system at a given temperature Vapor pressure

is often expressed in millimeters of mercury Table 1 provides a listing of vapor

pressures of some common environmental contaminants When chemicals exist in

pure form, the vapor pressure of the contaminant is very important to its removal

efficiency by a VES The higher the vapor pressure, the more it is amenable to vapor

extraction In turn, the lower the contaminant’s vapor pressure, the less likely it will

be volatilized and the greater reliance on bioventing (if the contaminant is

biode-gradable) for successful remediation For mixtures of compounds (i.e., gasoline),

the composition of the mixture also has a bearing on the vapor pressure according

to the following relationship

(9)

where Pi* = Equilibrium partial pressure of component i in the organic mixture; Xi =

Mole fraction of component i in the organic compound mixture; Ai = Activity

coefficient of component i in the organic compound mixture; and Pio =Vapor pressure

of component i as a pure compound

Table 1 Contaminant Properties—Vapor

Pressure Parameters

Compound

Vapor Pressure (mmHg)

=

For example, the above relationship states that the vapor pressure of benzene isrelated to the percentage (mole fraction) of component benzene in gasoline.

If NAPL is not present in the soil, vapor pressure becomes a less accuratepredictor of VES efficiency since other relationships (adsorption to soil, moisturecontent, etc.) become more important in governing system success It should benoted, however, that even under conditions of no NAPL presence, a compound must

be volatile enough to be removed by VES Sufficiently high vapor pressure cantherefore be viewed as a prerequisite for successful vapor extraction Although thedefinition of sufficiently high vapor pressure is rather subjective, one to two milli-meters of mercury (mm Hg) should be used as a guideline Compounds with lowervapor pressures will likely be more slowly removed and/or a greater reliance will

be required on in situ biological breakdown of the compounds.

Solubility

Aqueous solubility is one of the most important parameters governing the titioning, transport, fate, and therefore, the ultimate remediation of site contaminants.Solubility can be defined as the maximum amount of a constituent that will dissolve

par-in pure water at a specified temperature For organic mixtures such as gasolpar-ine,solubility is additionally a function of the mole fraction of each individual constituent

in the mixture according to Equation (10) Chapter 1 contains a table listing purewater solubilities for some common environmental contaminants

(10)

where Ci* = Equilibrium concentration of component i in the organic mixture; Xi =Mole fraction of component i in the organic compound mixture; Ai = Activitycoefficient of component i in the organic compound mixture; and Cio = Equilibriumsolute concentration of component i as a pure compound

Under most vapor extraction scenarios, the vadose soil is relatively moist (10 to

14 percent by weight) and contaminants are generally dissolved in the soil porewater Solubility is also a critical factor for the bioventing of contaminants, sincebiological degradation is enhanced and simplified if the contaminants are moreavailable for microbial uptake by being dissolved in the pore water A soil moisture

of 12 percent by weight is generally required for adequate bioventing

Henry’s Law

The interaction of solubility and vapor pressure produces a behavioral cation that renders the additive effects of solubility and vapor pressure nonlinear.This interaction has particular impact on volatilization of organics from water Thereader is referenced to basic chemistry textbooks for the derivation of Henry’s Law(Mahan 1966) Henry’s constant is functionally defined as the ratio of saturatedvapor density to chemical solubility for a given compound

modifi-C i

*

X i A i C i o

=

at the water/vapor interphase; and Csi = Compound concentration in the water phase

at the water/vapor interphase

The Henry’s Law relationship is depicted graphically in Figure7Table2 presents

a list of various compounds Henry’s Law constants Under moist soil conditions,therefore, VES efficiency is Henry’s Law dependent much like VOC removal by airstripping For example, although acetone is very volatile, it is not well remediated

by a VES due to its high water solubility Acetone tends to biodegrade readily,however, and is therefore amenable to bioventing

Much like air stripping efficiency is temperature dependent, VES efficiency has

a temperature dependence The temperature relationship is more complex for VESdue to the existence of multiple system variables such as biodegradation, adsorption,etc In general, higher temperatures in the vadose zone enhance volatilization whichimproves operation of a VES Increased temperature also enhances biodegradation,increases the rate of desorption, and weakens the adsorption binding

Other Molecular Properties

There are several other molecular properties of the contaminant that influencethe success of vapor extraction or bioventing Although these properties are not assignificant as vapor pressure, solubility, and Henry’s Law, and hence may be con-sidered secondary, these properties often may be the rate-limiting criteria to siteremediation Compound size, molecular weight, electronegativity, and polarity affect

Figure 7 Henry’s Law.

H = C sg⁄C si

its adsorption to soil particles and its travel through soil micropores Larger, bulkier(branched chains) molecules travel more slowly within soil micropores and tend toadsorb more strongly to soil surfaces Once the bulk of the more accessible (largepores, not directly adsorbed to soil) contaminant is removed, the final moleculesremoval tends to be rate limiting Polarity and electronegativity relate to a com-

Table 2 Henry’s Law Constants

Compound

Henry’s Law Constant a

1 Per Hydro Group, Inc., 1990.

2 Solubility and vapor phase pressure data from Handbook of

Environ-mental Data on Organic Chemicals, 2nd edition, by Karel Verschueren,

Van Nostrand Reinhold, New York 1983.

3 Michael C Kavanaugh and R Rhodes Trussel, “Design of aeration towers

to strip volatile contaminants from drinking water,” Journal AWWA, p.

685, December 1980.

4 Coskun Yurteri, David F Ryan, John J Callow, Mirat D Gurol, “The effect

of chemical composition of water on Henry’s law constant,” Journal

WPCE, Volume 59, Number 11, p 954, November 1987.

pound’s effective charge, and its interaction with the soil’s surface charge Thesetopics are further discussed below.

Summary

Vapor pressure and solubility are the most dominant contaminant propertiesaffecting the success of a VES Secondary contaminant properties include molecularweight, compound size and structure, and surface charge The properties of thecompounds are only part of the controlling factors for VES The compounds interactwith the soil The stronger the interaction, the less effective the VES The degree ofinteraction is dependent upon the properties of the compounds discussed above andthe properties of the soil

PROPERTIES OF THE SOIL

The typical volumetric composition of soil is depicted in Figure8

Bulk Density/Soil Porosity

Decreasing soil permeability will generally reduce the efficiency of a VES becausethe diffusive transport from the soil matrix to the soil surface and in turn to the soilgas Also, the increased path length (path to convective airflow) and decrease in crosssectional area for airflow will reduce transport A secondary influence of decreasingsoil porosity is the increase in soil surface areas available for contaminant binding

Figure 8 Typical volumetric composition of soil.

The permeability will also have an economic affect on the VES The less permeablethe soil, the higher the vacuum that is required to maintain the same airflow rate.The zone of influence will also be affected, requiring an increasing number of wells

to cover a volume of soil

Soil Adsorption

There are two methods by which the soil can adsorb the organic contaminant.The soil organic content or its mineral adsorptive surface sites are both capable ofcontaminant adsorption The adsorption of contaminants to the soil organic or min-eral clay surfaces tends to increase the immobile fraction of the contaminant(adsorbed to the soil) and to decrease the vacuum extraction system efficiency Soiladsorptive interactions become particularly important (rate limiting) under drier soilmoisture conditions

The clay soil composition is strongly correlated to many water transport andretention properties which influence the volatilization process Clay holds onto watertightly and is poorly transmissive This presence of water in the soil pores reducesthe available space for vapor transport The effects of soil moisture or pore water

on VES are further described below Clay soils also tend to have small microporesmaking transport path lengths longer The longer transport path lengths reduce theefficiency of vapor extraction

Clay mineral surfaces tend to have a net negative charge They will influencethe adsorption reaction of most compounds in some fashion Clays are an excellentadsorbent of positively charged molecules (such as metals) or very polar organiccompounds

Soil total organic carbon (TOC) matter content is strongly correlated with thebinding capacity for organic chemicals This general correlation is valid as a generalguidance despite the wide variation in the makeup of soil organic content, its state

of decomposition, and consequently its binding capacity The soil adsorption lation coefficient is likely to be highest when the soil organic content is high, butobservations of strong binding have been documented for soil TOC as low as 0.1percent A relationship describing the soil’s binding capacity to organics was defined

corre-in Equation (3) The parameter Koc in Equation (3) is defined as

Due to the preponderance of other soil organic matter surfaces and the nonpolarnature of most organic contaminants, there is usually little correlation between claycontent and VOC adsorption (Jury et al 1990)

Most organic contaminants are more easily adsorbed to the soil than they aredesorbed It therefore takes much longer and requires more energy to remove thecontaminants from the subsurface than it does to distribute them This phenomenon,known as hysterisis, tends to slow down the vapor extraction process more thanwould be predicted by simple adsorption isotherm data

Soil Moisture

Soil moisture is a very important parameter for VES success High soil moisturecontent limits air advection travel pathways by occupying void space Since move-ment of VOCs is much faster in the gas phase than in the liquid phase, it would beexpected that VOC removal by vacuum extraction would be enhanced by decreasingsoil moisture This trend is not always observed The lack of soil moisture allowsfor contaminant adsorption to soil surfaces to play a more prominent role in masstransfer as the water particles are removed from the surface If the soil adsorptivecapacities are strong, the benefits of soil dewatering (increased air travel pathways)may be partially offset by this increased soil binding capacity (Sims 1990 andThibaud, Erkey, and Akgerman 1993) The moisture content at which a decrease invapor concentration during VES operation is observed is often termed the criticalmoisture content and is often empirically defined as one monolayer of water mole-cules coating the soil surface Recent research observations hypothesize that waterparticles may act to kick out adsorbed organics, thereby enhancing VES operationunder certain conditions Figure 9 provides a schematic of this concept If soiladsorptive capacities are very weak (low TOC), it may be advantageous to conductvapor extraction under drier conditions

The notion that an optimal moisture content exists for a given contaminant, basedupon its Henry’s constant and the soil binding capacity, should allow for someprocess control and optimization of the VES performance Although theoreticallypossible, this notion is rarely applied in the industry due to limited modelingresources and incomplete understanding of the site’s partitioning coefficients Also,

it is not practical to regulate soil moisture It would be expensive to try to dry the

Figure 9 Soil moisture effect schematic.

air that is entering the vadose zone for the VES Even with injection wells it maynot be possible to completely eliminate the natural moisture of that climate Moisturecan be added in desert environments This is usually only considered for bioventingprojects.

Site Surface Topography

Site surface topography can greatly influence the success of a VES Ideally thesite should be covered by an impermeable surface such as a pavement or concrete.The covered surface serves two functions First, it minimizes the infiltration ofrainwater to the vadose soils and consequently allows some control over soil mois-ture Second, the covered surface eliminates the possibility of extraction well shortcircuiting (Figure10), where the majority of the extracted volume of air is comingfrom near the ground surface, thereby more distant locations from the extractionwells receive minimal airflow The basic design premise of any VES is to create anairflow pattern across the contaminated zone, and short circuiting to the surface mayprevent that pattern

Short circuiting may also be due to the presence of higher permeability zonessuch as utility trenches If operation of a VES is conducted in a zone that is prone

to short circuiting, a higher number of VES wells will generally be required Thiswill, in turn, lead to higher airflows and higher capital costs for air treatment devices

In order to minimize the effects of surface short circuiting, wells should not bescreened within 5 feet of the surface In instances where a surface seal is notavailable, plastic sheeting can be applied, preferably buried under one foot of cover

to enhance system performance

Figure 10 Surface seal effects schematic.

Depth to Water Table

If the VES well penetrates the water table (use of a converted monitoring wellfor vapor extraction) and vacuum is applied to the well, the water table within thewell will rise by an amount equal to the level of applied vacuum A 60 inch watercolumn vacuum at the well head will therefore result in a water table rise of 5 feet(Figure11) If there are only 5 feet of well screen above the water table, the watertable rise may clog the available well screen This scenario may be encountered in

a situation where the water table is shallow or when the well is inappropriatelydesigned Horizontal wells can be utilized in shallow water table situations therebyenlarging the available screen length and reducing well head vacuum, thus minimiz-ing water uplift

Installing the VES well with well screen above the water table will minimizewater table uplift as the vacuum at the water surface will be less than at the wellhead

As Figure6 points out, wellhead vacuum dissipates quickly away from the extractionwell As a general rule, the bottom of the VES well should be a minimum of 2 to

3 feet above the water table, if possible, to prevent the effect

Site Homogeneity

Site homogeneity is important to ensure that airflow reaches all areas requiringremediation The air carrier must flow past the contaminants if they are to beremoved Transport of the contaminants by the carrier airflow minimizes diffusion

Figure 11 Water uplift effects schematic.

requirements for mass removal This reduces the travel path length to remediationand expedites the cleanup As an example, NAPL floating on top of the water tabletakes longer to be volatilized because airflows over it rather than through it Labexperiments where air is drawn through NAPL saturated soils usually result in NAPLvolatilization that is much faster than volatilization of NAPL floating on the watertable.

Site nonhomogeneity can be partially alleviated by varying well screen designs

to maximize air movement in contaminated zones, fully opening some extractionwells in low permeability zones and closing others in high permeability zones, and

by possibly breaking the area into more than one zone based upon permeability.Wells in high permeability zones can be connected to a moderate vacuum blowerand wells screened in low permeability zones can be connected to high vacuumliquid ring type pumps (Figure12)

The presence of micro lenses of highly adsorptive, low permeability soils willoften be rate limiting during VES operation (Figure 13) These highly adsorptivelenses are not accessed by the carrier air and rely on concentration gradient drivendiffusion

At a VES site the presence of utility trenches (generally constructed of highpermeability fill material) or other high permeability airflow paths may also provideshort-circuit pathways In these instances, well screening and/or placement mayrequire adjustment to accommodate the contaminant distribution This adjustmentusually requires deeper well screens and a higher density of extraction wells

Figure 12 Advective/diffusive airflow schematic.

MODELING TOOLS FOR VAPOR EXTRACTION SYSTEM DESIGN

Computer simulation modeling is utilized to better design vapor extraction tems This enhancement is due to the simulation of complex subsurface processes.This allows the model user to vary operating parameters and observe a simulation

sys-of the result For example, the model allows for comparing subsurface flow regimeswith 10 or 20 extraction wells The modeler can therefore evaluate the benefits to

be gained by installing the additional 10 wells and decide whether the added benefit

is worth the additional installation and operational costs

For the purposes of this discussion, three classes of modeling tools for use inVES design are described This is a broad, crude model classification, but is useful

in presenting the basic requirements for VES design The following three differenttypes of models will be discussed

1 Engineering design model

2 Airflow models

3 Multiphase transport models

Engineering Design Model

Models are available that simulate vacuum losses in piping networks, vacuumequipment, well design, valving networks, and above-ground air treatment equip-ment These models are particularly useful for the design of larger systems (greaterthan 10 extraction points) where these calculations may become cumbersome Thesemodels are often based on solution of Hardy Cross type of pneumatic flow equations

Figure 13 Two-zone venting system.

Flow Models

Flow type models Airflow SVE (Waterloo Hydrogeologic Software, Waterloo,Ontario Canada), HYPER VENTILATE (U.S Environmental Protection Agency),and AIR 3D (ARCADIS Geraghty & Miller) simulate airflow Airflow SVE is also

a transport model while the others do not simulate mass transfer among the variouscompartments These models are simple to operate and allow the user to locate wells

in order to achieve a predetermined vacuum or conversely (Airflow and AIR 3D)allow the user to translate vacuum influence readings into cross sectional airflow at

a given remote location In either case the models allow for vacuum and airflowprofile distributions across the generating two or three (AIR 3D) dimensional profiles.Flow models do not account for volatilization from pore water, soil surfaces, orany equilibrium partitioning If flow models are utilized, the user must select theminimally acceptable airflow across the site based on published literature values forthe contaminant of concern within similar soils For example, if TCE is to beextracted from sandy soils, a minimally acceptable airflow across the most distantlocation from the extraction well must be selected A limited but useful database(minimally acceptable airflows) exists in the published literature for this purpose.Airflow models therefore allow for selection and optimization of well placementand well screening within the contaminated zone This selection is based on apredetermined minimal acceptable airflow

Multiphase Transport Models

Multiphase transport modeling is usually not required for the majority of simpleVES applications Vapor extraction at a typical quarter-acre sandy soil, shallow watertable, service station does not require multiphase transport modeling The installedsystems are usually small, overdesigned, and to a large extent rely on bioventing tosupplement any shortcomings of VES design, since the majority of contaminantsare biodegradable

At more complex sites, however, where any or all of the following is desired,multiphase transport modeling can be very useful

1 System layout and sizing optimization is desired due to the site’s large size (too expensive to overdesign)

2 System operating parameter optimization is desired

3 Vapor extraction feasibility requires demonstration

4 Contaminants may not be biodegradable

5 Impacts to groundwater may be significant

6 Closure time-frame prediction is required

Solution of Equation (8) with the requisite modifications to account for librium, nonlinear behavior is the basis for the multiphase transport models Severalmodels have been developed by researchers (Armstrong, Frind, and McClennan1994), but few models that can adequately describe multiphase transport are com-mercially available

nonequi-Multiphase transport modeling allows for simulating the remediation processover time based on the selected well layout The multiphase transport models aretherefore often preceded by simple airflow models that locate extraction wells Themultiphase transport models simulate the VES perturbation (inducement of airflow)

to the contaminant partitioning and predict contaminant concentrations within thevarious media at future points in time The modeler can vary the site’s soil moisture,airflow, or any other parameter and observe the predicted effects

A numerical model is therefore the best means of understanding the masstransfer between the liquid (soil pore water) and gas phase, and the degradation ofconstituents into different species MOTRANS, developed for EPA by Parker andKaluarachchi at The Virginia Polytechnic Institute and State University in Blacks-burg, Virginia is commercially available MOTRANS can be used to simulate eithertwo-phase flow of water and NAPL in a system with gas present at a constantpressure, or three-phase flow of water, NAPL, and gas at variable pressure Systemswith no NAPL present or with immobile NAPL at a residual saturation may also

be modeled by an option that enables elimination of the NAPL flow equation Thetransport module can handle up to five components that partition among water,NAPL, gas, or solid phases assuming either a local equilibrium interphase masstransfer or first-order kinetically controlled mass transfer

The flow of water and vapor and the transport of constituents in the vadose zoneare highly complex processes The equations governing these processes are stronglynonlinear, difficult to solve, and require extensive data input to characterize thephysical properties of both the media and the fluids In general, the principal limi-tation in applying modeling codes is characterization of the problem Migrations ofconstituents in the vadose zone are controlled by local heterogeneities, which may

be difficult to define

In addition, the physical properties characterizing the relative permeabilities andfluid retention characteristics are rarely collected Multiphase flow and multicom-ponent transport require specification of permeability/saturation/capillary pressurerelationships, air-water capillary retention function parameters, NAPL surface ten-sion and interfacial tension with water, NAPL viscosity, NAPL density, maximumresidual NAPL saturation, soil permeabilities and dispersivities, initial phase con-centrations, equilibrium partition coefficients, component densities, diffusion coef-ficients, decay coefficients, mass transfer coefficients, and boundary conditions.These relationships and parameters can be determined from direct measurements inlaboratory treatability tests that can accompany modeling efforts or found through

a literature search Multi-phase models can therefore potentially account for:

1 Advection

2 Dispersion

3 Sink/source mixing

4 Chemical and equilibrium partitioning

The models can potentially simulate the removal of the contaminants from thesubsurface under a variety of conditions (different flow velocity, different well screen

positions, different moisture levels, different extraction and passive well locations,different concentration profiles, etc.) The models can allow the user to optimallychoose:

1 Well screening positions

Pilot studies are generally conducted in order to gather relevant information

to design a full-scale VES Field pilot studies gather information regarding thepneumatic flow characteristics of the vadose zones and the extracted air quality.Laboratory soil column and soil cube tests allow for simulation of the vaporextraction remediation process by passing air through a small amount of soil Thisalso allows for selection of the minimally acceptable airflow velocity for removal

of the contaminants

LABORATORY STUDIES

Laboratory studies are often the best method to optimize the required airflowand moisture content, as well as other control parameters for a VES Laboratorystudies allow for manipulation of these parameters under controlled conditions andcan be done in conjunction with modeling and field pilot studies to optimize systemperformance Due to the expense associated with these activities, their implementa-tion is rare except at large sites contaminated with nonbiodegradable VOCs wherevolatilization must be optimized

Lab studies can be conducted using columns or soil cubes Figure14 presents

a schematic of the laboratory setup Soil cubes offer the benefit of providing bettersimulation of actual airflow profiles in the subsurface However, soil cubes will not

be able to simulate the macrogeological conditions of a sand lens, clay lens, or anyother major change in the geology

Researchers have increasingly concentrated on evaluating the minimally able airflow velocity for successful vapor extraction (Armstrong, Frind, and McClen-nan 1994, Gierke, Hutzler, and McKenzie 1989, Sleep and Syikey 1989 and Brus-seau, Jessup, and Rao 1991) It is anticipated that upon completion of sufficient labstudies, a matrix of contaminant type, soil type, moisture content, and minimalairflow velocity can be compiled This matrix would greatly improve the design ofVES by practitioners

accept-FIELD PILOT STUDIES

This section will describe the test wells, procedures, and equipment required toconduct a pilot or feasibility test A pilot test will allow determination of thesubsurface air permeability, the expected mass removal rates of contaminants atsystem start-up, and allow for the selection of appropriate vapor treatment equip-ment The pilot test should also determine the required vacuum and number ofextraction points to achieve the desired airflows during VES application A pilot testshould be conducted prior to installation of most VES applications Only systemswhose size does not greatly exceed (two to three times) the size of the pilot systemshould be installed without pilot testing For small systems, the cost of the pilot testcan sometimes not be justified

A typical pilot test set up is shown in Figure 15 The pilot test will includeinstallation of a minimum of one extraction well, several observation wells (orpoints), and hook up of the extraction well to the vacuum equipment Upon start up

of the vacuum pump, several field measurements from the wells and the extractionvacuum pump are taken prior to the conclusion of the test

Vapor Extraction Testing Well

The extraction well should be located near the center of the contaminated zone

in order to ensure gathering of data that may be representative of start up conditions.The vapor extraction well should be 2 to 4 inches in diameter At most locations, 2inch diameter construction is adequate; 4 inch diameter construction is advisable in

Figure 14 Lab columns/cubes.