In Situ Treatment Technology - Chapter 6 ppt

Nyer CONTENTS IntroductionDesign CriteriaRegulatory RequirementsMass of ContaminantsLifecycle Emission ConcentrationCiting and Utility ConsiderationsTreatment Technologies Adsorption-Bas

Kidd, Donald F & Nyer, Evan K."Air Treatment for In Situ Technologies"

In Situ Treatment Technology

Boca Raton: CRC Press LLC,2001

CHAPTER 6

Donald F Kidd and Evan K Nyer

CONTENTS

IntroductionDesign CriteriaRegulatory RequirementsMass of ContaminantsLifecycle Emission ConcentrationCiting and Utility ConsiderationsTreatment Technologies

Adsorption-Based Treatment TechnologiesOff-Site Regenerable/Disposable Gas Phase GACOn-Site Regenerable GAC

Resin Adsorption SystemsOxidation-Based TechnologiesThermal Oxidation/IncinerationCatalytic Oxidation

Biological TechnologiesScrubbers

Technology Selection SummaryEmission Control Case StudySite Description

Remedial ApproachEmission Control Design BasisAlternatives EvaluationSystem PerformanceReferences

In Chapter 1 we introduced the concept that most in situ treatment processeswere simply a switch from water to air as the carrier This chapter will look attreating the air carrier as it is brought above-ground

Above-ground vapor treatment of emissions from soil vapor extraction, air ing, and air stripping applications often represents the largest portion of the overallcost of implementing these technologies Figure 1 represents a pie chart of overallproject costs associated with a vapor extraction system which operates for 3 years

sparg-at a vapor recovery rsparg-ate of 300 cfm and a declining influent concentrsparg-ation from2,000 ppm (hydrocarbon vapors) to 5 ppm over the project lifetime It is assumedthat vapors are treated using catalytic oxidation for the first two years, and thatgranular activated carbon (GAC) is used for treatment during the last year of treat-ment Figure 1 shows that air emission control and O&M related costs might beover 50 percent of overall project costs

Due to the magnitude of air emission control costs, the design engineer mustcarefully evaluate and select the most appropriate technology Control technologyselection must consider several criteria that will be introduced in this chapter:

• Regulatory requirements

• Overall mass of VOCs to be treated

• Anticipated decline rate of VOC concentrations over project lifetime cycle design)

(life-• Citing considerations

• Utility availability

Figure 1 Pie chart of overall VES project costs.

• Organic and inorganic composition of process vapor stream

• Other project specific considerationsThe most common air emission control technologies can be classified as adsorp-tive, oxidative, or biological The adsorption based technologies include off-siteregenerable/disposable vapor phase GAC, on-site steam regenerable GAC, and on-site regenerable macroreticular resin systems Oxidative technologies include cata-lytic oxidation, and thermal oxidation Biological based systems have gained atten-tion in the last few years and have become commercially available Less commonlyutilized technologies include scrubbing, vapor compression, UV/ozone oxidation,and refrigeration The most commonly utilized technologies will be introduced inthis chapter

DESIGN CRITERIA Regulatory Requirements

Vapor emissions from site remediation activities are generally not permanentsources of discharge The short duration of the emission may exempt its permittingand control in some states Often however, in cities or states where overall air qualitystandards are not met (nonattainment areas), or in states with strict emission controlstandards, permitting and vapor treatment is required

Emission requirements are quite variable within the different states Emissionrequirements may be based on total mass emissions per hour or per day Mass basedemission criteria may be for total VOCs and/or may be compound specific Thedesign engineer must select the air emission control system based upon the mostlimiting criteria For example, VES emissions for a gasoline release contain a variety

of compounds If air emission standards require a maximum 15 pounds/day totalVOC limit, and an overall benzene limit of one lb/hr, then system design must bebased upon the limiting regulatory requirement In this example, the limiting criterion

is likely based on total VOC emissions, since benzene usually constitutes only asmall fraction of the total mass of gasoline Because of the variable composition ofgasoline and variations in constituent volatility, vapor samples collected during pilottesting are often used to determine the factors dictating emission control design.Alternately, emission control criteria may simply require that the emission ratesnot cause an exceedance in ambient air quality or other risk based criteria Thisgenerally requires that the point source of discharge be modeled using air dispersiontechniques (Bethea 1978) Several states impose both mass emission and concen-tration criteria Some states require that all emissions be treated using Best AvailableControl Technology (BACT) regardless of the magnitude of the emissions (thisrequirement calls for emission control even if the process stream already complieswith mass emission requirements) Therefore, the first thing that the design engineermust do is to acquire the local and state regulations before trying to design a vaportreatment system Based upon lifecycle design (Chapter 2), emission rates declineduring site remediation Permit preparation should account for this temporal change

It is plausible to limit the site operation (hours/day or number of extraction wells)

to stay within permitting limits (without treatment) until emission rates drop as thecleanup progresses This approach will likely increase site remediation time frameand the design engineer must conduct a cost-benefit analysis in order to justify merits

of this phased start up method

Mass of Contaminants

An estimation of the total mass of VOCs that may be recovered by the ation system is often a requirement prior to determining the appropriate treatmenttechnology This is particularly true for adsorbent based treatment systems such ascarbon or resin-based controls For example, if 1,000 pounds of gasoline are known

remedi-to be in the subsurface, and one expects that 65 percent of the mass will be recovered

by vapor extraction, 30 percent will be biodegraded, and 5 percent will not berecovered (ratios are based upon empirical projection), an estimate can be formulatedfor expected adsorbent consumption Assuming a 7 percent by weight adsorptioncapacity for GAC, approximately 9,300 lbs of GAC will be required The cost ofother technologies can also be estimated based upon the mass of VOCs and expectedflow rates

There are several simple methods to estimate the mass of VOCs in the subsurface.Once this is calculated, an estimate can be made of the amount (percent) that isexpected to be extracted for above-ground treatment Often the final estimate isbased on an average of the various estimation methods An excellent starting point

is direct knowledge of the amount of contaminants released Time is also an tant factor in that spills will weather and naturally degrade (see Chapter 7 on In Situ

impor-Bioremediation)

Second, the total mass may be estimated by using soil contaminant tions, groundwater concentrations, soil gas concentrations, and NAPL thickness atthe various locations across the site (see Equation 1 in Chapter 3) The use ofweighted average methods (concentration and expected flow from each zone) andsubdivision of the site into small quadrants (based on the available data) will yieldmore accurate mass estimates

concentra-It should be noted that soil analytical methods often underestimate the amount

of adsorbed VOCs due to significant losses during the sampling procedures (USEPA1991) Use of methanol extraction/preservation methods can often lead to soil con-taminant levels that may be one to two orders of magnitude higher than conventionalmethods It should also be noted that core samples represent a small statisticalpercentage of the sampled media, and therefore, are inherently inaccurate

Third, in instances where limited information is available, gross estimates of thetotal mass of contamination in the subsurface may be evaluated using partitioningcoefficients For example, if no soil contamination data is available, groundwaterdata, knowledge of the compound octanol-water partition coefficient, and soilorganic content can be used to estimate the amount of VOCs adsorbed to the soil(Equation 3 in Chapter 3) It should be stated that these equations assume equilibriumconditions persist in the subsurface Nonequilibrium conditions generally dominate

in the subsurface, and partitioning based calculations underestimate the adsorbedmass This is because a large portion of the mass may be restricted from being inequilibrium with the surrounding soil vapor/groundwater due to nonequilibrium typeadsorptive or mass transfer limitations (Brusseau 1991).

Lifecycle Emission Concentration

Design of cleanup strategies to accommodate the lifecycle of the project hasbeen emphasized several times in this book This is particularly true for the treatment

of vapor emissions from VES, where concentrations may drop four orders of nitude over a project lifetime (Figure 2) Emission control technology selection ismore significantly affected by concentration than volume through-put for vaporphase treatment than for liquid phase treatment For example, an air stripper willgenerally be chosen for the treatment of 100 ppm or 100 ppb of groundwatercontaminated by BTEX compounds This choice will be made for almost any flowrate On the other hand, an emission stream of 10,000 ppm vapors from a VES stack(300 cfm), is best treated by a thermal oxidizer As the concentrations drop to 1,000ppm, the vapors are best treated by catalytic oxidation At influent concentrations

mag-of 20 ppm, the optimal choice may be GAC The original design must encompassall of these criteria, not just the initial influent concentration

This dynamic need to modify treatment technologies necessitates foresight fromthe design engineer for vapor emission system design The systems must be designedand installed with sufficient flexibility to allow for future modifications For example,

at a site where VES emissions are expected to be above 300 ppm for 6 months and

Figure 2 Typical decline curve for VES emissions.

then drop off rapidly (typical small service station, limited spill situation), a catalyticoxidizer may be rented for the first six months and subsequently GAC may beinstalled at the site Many localities have recognized the benefits and needs for adynamic emission control scheme for these typically short duration emission sources.

In efforts to streamline the local regulatory approval process, often permits are issuedfor various locations, allowing one system to be placed at several sites within theagency’s jurisdiction Such permits allow for more rapid deployment of the appro-priate emission control system, while reducing the burden for detailed evaluation of

an often lengthy permit application When considering the merits of an adaptiveemission control scheme, the treatment system citing must accommodate futureneeds of the contingent system allowing adequate space and utility connections foreach component of the treatment process equipment

Typically, the engineer needs to predict the decline curve for the emissions fromthe air treatment system and subsequently prepare a cost analysis for the variousoptions at varying concentrations A typical cost analysis table is shown in Table 1.Modeling of the remedial system performance to predict the decline curve may beconducted In many instances, this modeling is not performed and empirical methods(fitting the concentration decay to an empirical logarithmic decay equation over atime period based upon past experience) are used for its prediction The use ofempirical methods is generally acceptable in the consulting industry for purposes

of air emission selection due to the costs of modeling and its inherent uncertainties.For example, it is not critical to know whether a catalytic oxidizer will run for six

or seven months before switching to GAC, what is important is the ability to planfor and switch to GAC

Table 1Cost Analysis Spreadsheet for Vapor Treatment Costs

Influent Concentration VOC/day GAC cost/day

Catalytic Oxidation cost/day

1 ppm = 3.26 mg/cu meter (benzene)

100 cfm operation for 24 hour per day GAC adsorption capacity = 15% by weight Carbon Costs = $6/lb (new plus regeneration and changeout cost) Catalytic oxidation unit rental is $3,000/month

Catalytic oxidation power consumtion is $350/month (at 500 ppm influent); assume costs are slightly higher at lower concentrations; slightly lower at higher concentrations assume costs are slightly higher at lower concentrations; slightly lower at high concentrations Cost per day of catalytic oxidizer is (3,000+350)/12= 114

Citing and Utility Considerations

There are several citing considerations that need to be evaluated prior to treatmenttechnology selection Some of these constraints are presented below:

• Availability of utilities

• Utility cost analysis

• Access issues relating to O&M

in sufficient pressure to be utilized by the treatment equipment In residential borhoods, natural gas lines may not have sufficient pressure for adequate operation

neigh-of some thermal oxidation units Even where available, high pressure gas lineconnections to the oxidizer typically require additional lead time for permitting andinstallation Electrical power must be available in the appropriate phase and voltage

to power the equipment At remote sites, where utility availability is limited, propanetanks can be utilized The design engineer needs to conduct a cost analysis in order

to choose the most appropriate power source (natural gas, electrical, propane, oil,etc.) for powering the treatment unit When available, natural gas tends to be thelowest cost option in many locations The use of propane, in addition to increasedcost per BTU (generally 1.5 to 2 times higher than natural gas), also presents otheroperational problems such as increased fouling of burner components, as well aslogistical problems created from scheduling fuel deliveries

Remedial systems are unplanned installations Sites and neighborhoods are ously developed without planning for a potential remedial system installation Thisunplanned remedial system, therefore, needs to be located to accommodate severalfactors that may sometimes be conflicting It must be located to attain permitting,meet regulatory stack height and air dispersion requirements, fit in with the naturalsetting, and be accessible for routine operation and maintenance Concurrently, thesystem must not be offensive to neighbors and have a stack height that meets localzoning laws

obvi-TREATMENT TECHNOLOGIES Adsorption-Based Treatment Technologies

Adsorption is a process by which material accumulates on the interface betweentwo phases In the case of vapor phase adsorption, the accumulation occurs at theair/solid interface The adsorbing phase is called the adsorbent and the substance

being adsorbed is termed an adsorbate It is useful to distinguish between physicaladsorption, which involves only relatively weak intramolecular bonds, and chemi-sorption, which involves essentially the formation of a chemical bond between thesorbate molecule and the surface of the adsorbent Physical adsorption requires lessheat of activation than chemisorption and tends to be more reversible (easierregeneration).

GAC is the most popular vapor phase adsorbent in the site remediation industry

A number of new synthetic resins, however, have shown increased reversibility andhave higher adsorption capacities for certain compounds

The most efficient arrangement for conducting adsorption operations is thecolumnar continuous plug flow configuration known as a fixed bed In this mode,the reactor consists of a packed bed of adsorbent through which the stream undertreatment is passed As the air stream travels through the bed, adsorption takes placeand the effluent is purified (Figure 3)

The part of the adsorption bed that displays the gradient of concentration istermed the mass transfer zone (MTZ) The amount of adsorbate within the bedchanges with time as more mass is introduced to the adsorbent bed As the saturated(spent or used) zone of the bed increases, the MTZ travels downward and eventuallyexits the bed This gives rise to the typical effluent concentration versus time profile,called the breakthrough curve (Figure 4) The reader is referenced to several text-books for adsorption theory, multicomponent effects, isotherm description, and mod-eling (Noll, Vassilios, and Hou 1992 and Faust and Aly 1987) This basic knowledge

of adsorption theory is critical to proper understanding and selection of the variousadsorbents

Figure 3 Concentration profile along an adsorbent column.

Off-Site Regenerable/Disposable Gas Phase GAC

Gas phase GAC is an excellent adsorbent for many VOCs commonly encountered

in vapor extraction, air sparging, vacuum-enhanced recovery, and conventionalgroundwater extraction The adsorption capacity of GAC is often quantified as themass of contaminant that is adsorbed per pound of GAC This nominal adsorptioncapacity is a useful guide for pure compound adsorption but can be misleading whencomplex mixtures of VOCs are treated Breakthrough, or GAC bed life, is definedwhen breakthrough occurs for the compound most difficult to adsorb In instanceswhere multicomponent mixtures are present, the adsorption capacities for eachcompound are generally lower than for pure compounds Isotherm data (mg adsor-bate removed per g adsorbent at a constant temperature) and other product specificdata are generally available for contaminants of interest from carbon vendors as well

as in the published literature Pilot testing for GAC feasibility is rarely conducted,except in instances where complex mixtures of VOCs are encountered

GAC is generally a good adsorbent for hydrocarbon origin VOCs, and somechlorinated VOCs GAC has limited adsorption capacity for ketones and generallypoor adsorption of volatile alcohols Table 2 shows typical adsorption removalefficiencies for a variety of VOCs by GAC under constant temperature and moistureconditions (as stated in the table)

GAC adsorption capacity is significantly enhanced if the vapor stream’s relativehumidity is kept low The use of water knock outs, demisting, desiccants, and airstream temperature adjustments are therefore common pretreatment steps to enhanceGAC performance Adsorption capacity may be as much as 10 times higher for alow humidity stream than for a humid air stream This is particularly true for lower

Figure 4 Breakthrough curve for a typical adsorber column.

concentrations of VOCs The humidity effects are less pronounced at higher VOCconcentrations (Figure 5) Temperature elevation after water demisting/desiccationincreases the amount of moisture that the air stream can sustain thus reducing therelative humidity It should be noted that the adsorption vessels must also be keptwarm so as to avoid water condensation on the GAC and for the air stream tomaintain the reduced relative humidity.

The use of off-site regenerable/disposable GAC for emissions control is oftenlimited to instances where the mass loading is low and therefore GAC consumption

is low As a general guideline adsorption capacities for adsorbable VOCs are in therange of 2 to 20 percent range by weight GAC costs are in the range of $3/lb (ifpurchased in canisters; $1/lb if purchased as carbon only) plus an additional $0.5

to $1/lb for regeneration/disposal GAC vessels may be purchased as canisters (200

lb size, where the vessel and GAC are replaced after consumption), larger replaceableplastic/fiberglass canisters, or as conventional steel vessels wherein only the GAC

Table 2 Adsorption Capacity of GAC for Some Common VOCs

Carbon Capacity* (lb org/100 lbs GAC)

*Note: Carbon capacities are based on Calgon BPL carbon at 70 F and

1 atm Values adjusted to reflect usage per 100 lbs GAC.

Source: Calgon Bulletin #23-77a (2/86).

Figure 5 Effects of relative humidity on TCE adsorption by GAC.

is replaced Vessels are most commonly used in series in order to allow for effluentstream sampling between vessels in order to more accurately predict breakthroughtimes (Figure 6) Single vessels may also be used in conjunction with vapor effluentdetectors that can shut down the system or switch spent vessels to stand by unusedvessels GAC vessel/blower selection must account for head losses through thecarbon system to ensure maintenance of the desired airflow.

On-Site Regenerable GAC

Liquid phase GAC is difficult to regenerate at low temperatures due to adsorption

of background metals and TOC (typically naturally occurring humic and fulvicacids) In the vapor phase, the GAC does not see the metals nor the nonvolatile TOCand is therefore amenable to on-site, low temperature regeneration Air emissioncontrol using steam regenerable GAC generally utilizes a two bed system wherebyone bed is being utilized, while the second bed is either being regenerated or in astand by mode Regeneration is accomplished by passing low-pressure steam through

Figure 6 Photo of typical GAC system in series.

the carbon vessel, which desorbs the contaminants (due to the raised temperature).The contaminated steam is subsequently cooled and separated from the nonaqueousorganics by gravity The condensed organics generally require disposal, whereas thecontaminated steam may undergo water treatment (especially if a groundwatertreatment system exists on-site) or may also require off-site disposal If the con-densed steam is treated on-site, it should be metered into the groundwater treatmentsystem since it is generally much more contaminated than the groundwater.

If regenerable systems are used for adsorption of chlorinated VOCs, the vesselsshould be lined or made of an acid-resistant material The adsorption/regenerationcycle results in formation of some hydrochloric acid (HCl) within the vessels TheHCl formation will reduce the pH of the condensed steam

Regenerable GAC units are also available with nitrogen regeneration This isparticularly useful to minimize steam disposal and eliminate problems with HClformation during steam regeneration Regenerable beds will usually have a perfor-mance lifetime, since adsorption capacity tends to diminish with continued regen-eration (typically decay to 70 percent of original capacity) System lifetime rangesare dependent on frequency of regeneration but are typically in the 3- to 7-year range.Regeneration can be, (1) manual during a site operation and maintenance visits;(2) by a timer system that starts up the boiler/regeneration prior to the time ofexpected breakthrough; or (3) using an effluent detector system (typically either aflame ionization or photoionization detector) which initiates vessel alternation andregeneration based upon breakthrough After completion of the regeneration, thevessel will usually undergo a drying cycle in order to prepare the vessel for the nextadsorption cycle Figure 7 provides a schematic of a regenerable GAC system Figure

8 is photo of a commercially available system

Figure 7 Regenerable GAC system schematic.

The regeneration capability of the GAC allows for treatment of more nated air streams than is possible with off-site regenerable GAC In VES applications,regenerable GAC systems are often thought of as best suited for high flow applica-tions with moderate VOC loadings It can be used for the treatment of halogenatedand nonhalogenated air streams During adsorption of highly oxidizable VOCs(ketones) in the presence of ozone or other oxidizing agents, the GAC bed may beprone to bed fires Fire detection and suppression systems may be considered underthese circumstances.

contami-As a general guideline, fully manual regenerable systems with a boiler may bepurchased for roughly $60,000 for a 500 cfm dual bed application System automa-tion with effluent detector will generally add approximately $40,000

Resin Adsorption Systems

The use of adsorbent resins for water treatment began in the 1960s with Rohmand Hass’ introduction of their Macroreticular Ambersorb adsorbent (Faust and Aly

Figure 8 Regenerable GAC system photo.

1987) These synthetic resins were designed to have low adsorption for backgroundTOC and metals and thereby enable the liquid phase adsorbent to better adsorb theVOCs Elimination of TOC and metal adsorption would also reduce biofouling (lessTOC for microbial growth) The liquid phase resins were also designed to be steamregenerable since metals and TOC fouling would not impact regeneration For avariety of economic and performance reasons, liquid phase adsorbent resins havegained limited acceptance New resins by both Rohm and Hass and Dow within thepast five years, however, may make liquid phase resin adsorption more economicallycompetitive.

It had been commonly prescribed that these resins be prewet prior to their use

In the 1980s, it was observed that the resins have an almost similar adsorptioncapacity in their prewet condition (Yao and Tien 1990 and Rixey and King 1989).These resins tend to adsorb VOCs in a similar fashion to GAC, but also tend todemonstrate some absorption within the resin itself (Gusler, Brown, and Cohen1993) The regeneration capability of absorbed VOCs is not currently well quantified.Despite this uncertainty, the resins appear to have a few advantages over vapor phaseGAC The resins have a higher adsorption capacity for some VOCs than GAC, areless influenced by relative humidity than GAC, and their on-site regeneration tends

to produce less acid when adsorbing chlorinated VOCs than GAC The resins,however, appear to have some catalytic ability, and therefore break down some ofthe halogenated VOCs to form the hydrochloric acid (but still less than GAC) Theresins also have less oxidative capacity and are less prone to bed fires than GAC foradsorption of readily oxidizable compounds

In particular, the resins appear to better adsorb vinyl chloride than conventionalGAC systems When present, vinyl chloride can effectively preclude GAC adsorption

as the sole emission control process due to it low adsorptive capacity and typicallylow emission requirements Chemical oxidation using potassium-permanganateimpregnated activated alumina has been found an effective alternative In this appli-cation, the activated alumina is placed at the end of the treatment train where vinylchloride is essentially the only organic compound reaching the oxidizing agent Bythis configuration, oxidation of other organic compounds is minimized, similarlyminimizing the consumption of the oxidizer

An additional advantage of the resin system over steam regenerable GAC, is theelimination of steam disposal/treatment Although prices have been declining forthese systems, they are still higher than conventional regenerable GAC systems Theaverage price of the resin is $70/lb in comparison to $1/lb for GAC (if purchased

in bulk; $3/lb in canister form) The resin’s superior performance for certain pounds and its minimal disposal and O&M costs, however, may make it more costeffective than regenerable GAC for certain installations In general, the resin adsorp-tion technology is not considered competitive for hydrocarbons, since these com-pounds can be more economically destroyed by catalytic or thermal oxidation Theresin technology is most applicable for moderately contaminated halogenated VOCvapor streams However, even these economics are changing as better catalysts aredeveloped for thermal destruction of chlorinated hydrocarbons