Miljøprojekt Nr. 1194 – Kemisk oxidation af sediment- og grundvandsforureningen på depotet ved Høfde 42 - fase 1: Projektbeskrivelse og forundersøgelser

This report presents a literature review of in situ chemical oxidation (ISCO) technologies that are being considered for remediation of organophosphorous pesticides in groundwater at the Høfde 42 Harboøre Tange (Cheminova) Site in Ringkjøbing County, Denmark (the Site). The Site is located on the western coast of Denmark, on a beach adjacent to the North Sea. GeoSyntec and Dr. Joseph Pignatello of the Connecticut Agricultural Experiment Station and Yale University have prepared this report for COWI A/S as the first phase in a project to evaluate ISCO technologies for remediation of an organophosphorus pesticide contamination in groundwater at the Site. The findings of this literature review will be used to design laboratory treatability tests to further evaluate the effectiveness of selected ISCO technologies for application at the Site.

State-of-the-science literature regarding the use of ISCO techniques for the treatment of organophosphorus pesticides in groundwater has been reviewed. Particular emphasis has been given to chemical oxidation technologies that are applied for remediation in situ, including conventional Fenton’s reagent, modified Fenton’s reagent, ozone, ozone and peroxide, permanganate, persulfate, and heat-activated persulfate. Of these technologies, the focus of this document has been on those technologies that have already been shown, through laboratory research, to effectively oxidize phosphorothionate pesticides.

1.1 Site Conditions

This literature review has focused on parathion, methyl parathion, malathion, and amino parathion, which are primary contaminants of concern (CoCs) in the source area at the Site (Ringkjøbing County, 2004). These chemicals are present as an immiscible, dense nonaqueous phase liquid (DNAPL) in the upper sand aquifer at the Site. A groundwater plume of these pesticides and their degradation products extends from the DNAPL source area towards the North Sea. The dissolved phase chemicals have also been transported into the lower sand aquifer at the Site. The conditions in this plume are highly acidic (typical pH ranges from 2.4 to 4.3). A summary table of the measured concentrations of these compounds in the DNAPL source area, as presented in an overview of the Site conditions by Ringkjøbing County, Department of Environment and Infrastructure (2004), is provided in Table 1.

It should be noted that in addition to the organophosphorus pesticides listed in Table 1, lesser concentrations of alkylated polysulfides, triethyl phosphate and mercury are also present in the DNAPL at the Site. The scope of the literature review did not include evaluating the effects of chemical oxidation upon these chemicals; however, these effects should be evaluated prior to initiating any oxidation field program at the Site. For example, mercury is a redox-sensitive metal that has the potential to be affected by an ISCO program. Geochemical modeling and/or laboratory treatability testing could be used to assess the impact of an ISCO program on the solubility and speciation of this metal.

1.2 Outline of this Document

The remainder of this report outlines the results of the literature review including:

Key properties of organophosphorus pesticides (Section 2.0);
An overview of the ISCO technologies reviewed in this document, including their general chemistry, considerations for in situ treatment, advantages and disadvantages of each technology, and design, operation and maintenance, safety, and cost considerations for field implementation (Section 3.0);
Reaction chemistry of Fenton’s reagent with organophosphorus pesticides and the implications for field implementation at the Site (Section 4.0);
Reaction chemistry of permanganate with organophosphorus pesticides and the implications for field implementation at the Site (Section 5.0);
Reaction chemistry of persulfate with organophosphorus pesticides and the implications for field implementation at the Site (Section 6.0);
Reaction chemistry of ozone and ozone + peroxide with organophosphorus pesticides and the implications for field implementation at the Site (Section 7.0);
A description of selected case studies of field applications of the candidate oxidants identified from the literature review (Section 8.0);
A summary of the results of the literature review and recommendations for bench scale and field scale testing (Section 9.0); and,
References (Section 10.0).

2 Key Properties of Organophosphorus Pesticides

The key chemical properties of organophosphorus pesticides reviewed in this memorandum are summarized in Table 2 (Chemfinder.com, 2004; Fjordbøge, 2005; Schwarzenbach et al., 2003; Montgomery, 2000; Budavari et al., 1996). Key properties of these pesticides that impact their partitioning and transport in soil and groundwater include the following:

They are generally liquid at room temperature, with a density in the range of 1.2-1.3 grams per milliliter (g/mL), making them a DNAPL when released as a pure phase in groundwater;
They have low aqueous solubility and vapor pressure;
They partition readily into organic phases. As a result, these compounds adsorb strongly to natural organic matter in soils;
They degrade to varying extents via hydrolysis, yielding water soluble compounds. However, at sites where the groundwater is relatively acidic (i.e., pH<5) such as the Høfde 42 site, this hydrolysis reaction may be very slow to non-existent (Montgomery, 2000); and
They biodegrade to varying extents via both anaerobic and aerobic pathways; however, many of the reaction products are toxic themselves (Montgomery, 2000) and, biodegradation can be significantly inhibited at the low pH levels that are typical of the Høfde 42 site.

3 Overview of Oxidants for In Situ Treatment

ISCO is an emerging technology for the treatment of hazardous waste. ISCO refers to a group of specific technologies that each use differing combinations of oxidants and delivery techniques. ISCO has been shown to destroy or degrade an extensive variety of hazardous wastes in groundwater and soil, including fuel hydrocarbons, chlorinated solvents (e.g., perchloroethene [PCE] and trichloroethene [TCE]), fuel oxygenates (e.g., methyl-tert-butyl-ether [MTBE]), and polycyclic aromatic hydrocarbons (PAHs). To date, the technology has not been as widely applied with pesticides; however there are a number of bench scale studies that suggest that ISCO may be appropriate for field application for certain classifications of pesticides as well. A schematic of a typical ISCO field application is shown in Figure 1.

Various oxidants have been used in laboratory and field applications to aggressively destroy organic chemicals, including Fenton’s reagent, permanganate, persulfate, ozone, and ozone combined with peroxide. These oxidants react to varying degrees with organic contaminants (i.e., breaking molecular bonds of and capturing electrons from the contaminant) and convert them into degradation products. Depending on the parent compound, the final reaction products may be innocuous compounds commonly found in nature such as carbon dioxide (CO₂), water and inorganic chloride.

The following subsections provide an overview of the general chemistry of each of the oxidants, advantages and disadvantages for each oxidant, considerations for in situ treatment, and the design, operations and maintenance, safety and cost issues for ISCO.

3.1 Chemistry of Oxidants

Because not all ISCO treatments are applicable for all contaminants, site contaminants and conditions must be understood in order to choose the appropriate oxidant and delivery method. The treatment effectiveness of chemical oxidants currently in use varies based on several factors, including the redox potential (E°) of the oxidant, and the reactive specificity of the oxidant toward a given type of contaminant. Permanganate (E° – 1.70), for example, has been shown to be primarily effective for oxidizing chlorinated ethenes, but not chloroethanes or fuel hydrocarbons. In contrast, Fenton’s reagent (E° – 2.76) is known to oxidize fuel hydrocarbons, PAHs, pesticides, polychlorinated biphenyls (PCBs), and most types of chlorinated solvents. The redox potentials of the primary oxidants currently in use are summarized in the Table 3.

In selecting an appropriate ISCO technology, an understanding of the geochemical conditions at a given site is essential since the applied reagents could be consumed by natural organic matter or dissolved iron rather than the contaminants, resulting in poorer than expected treatment. Groundwater geochemistry may also need to be adjusted to more optimal conditions prior to treatment (e.g., lowering of pH during application of conventional Fenton’s reagent at sites where groundwater pH is near neutral).

Specific details of the chemistry of each oxidant are described further in the following subsections.

3.1.1 Fenton’s Reagent

The Fenton reaction has been studied extensively in regard to waste treatment applications (Pignatello et al., 2006). Hydrogen peroxide is an effective oxidizing agent; however, to achieve contaminant reduction in a reasonable time, iron or iron salts are used as a catalyst (the combination is referred to as Fenton’s reagent). It generates hydroxyl radicals through a series of reactions with hydrogen peroxide catalyzed by iron ions, which undergo a redox cycle between the +II and +III oxidation states. The hydroxyl radicals (OH^•) serve as powerful, effective and nonspecific oxidizing agents. The mechanism is complex but can be summarized by the following steps:

Fe(II) + H₂O₂ → Fe(III) + OH^- + HO^• (1)

Fe(III) + H₂O₂ → Fe(II) + HO₂^• + H⁺ (2)

HO^• + R-H → H₂O + R^• (3)

where R-H is an organic compound. The optimum pH is slightly less than 3, and rates normally drop precipitously above pH 4 (Pignatello et al., 2006).

There are several variants of the Fenton reaction that mainly contribute to the regeneration of soluble Fe(II). One of the most important is the photo-assisted Fenton, or photo-Fenton reaction, shown below, which is initiated when the solution is irradiated with ultraviolet (UV) or UV/visible light.

Fe(III)-L + hv → Fe(II) + L^• (L = OH^- or organic ligand) (4)

Although photolysis is not possible for in situ treatment, studies that involve photo-assistance are included in Section 4 for their relevance to the dark reaction. Photo-assistance generally increases the rate. Moreover, it often alters the product distribution, since some of the products undergo further oxidation because they form photo-labile complexes with Fe(III).

Low pH conditions are often impractical to maintain under field conditions due to the enormous buffering capacity associated with most native soils. In another variation of the Fenton reaction, it has been demonstrated that free radical generation and contaminant oxidation can be promoted at neutral pH using a modified Fenton’s process that uses hydrogen peroxide, Fe(II), and chelating agents (e.g., citric acid or nitrilotriacetic acid) to keep iron in solution without the need for acid pH conditions (Watts et al., 1999).

Fenton’s reagent produces a strongly exothermic reaction with a very short half-life. As a result, the persistence of Fenton’s reagent in the subsurface is relatively short, and the ROI of injected Fenton’s reagents is characteristically low.

3.1.2 Permanganate

Permanganate is an oxidizing agent that has an affinity for oxidizing organic compounds, particularly those containing electron-rich carbon-carbon (C=C) double bonds, aldehyde groups or hydroxyl groups. The reaction between permanganate and chlorinated ethenes involves an electrophilic attack on the ethene’s C=C double bonds and the formation of a cyclic hypomanganate ester. Rapid hydrolysis of the cyclic ester results in the production of CO₂. There are two permanganate salts, potassium permanganate (KMnO₄) and sodium permanganate (NaMnO₄). The half reaction for MnO₄^- for pH in the range of 3.5 to 12 is:

MnO₄^- + 2H₂O + 3e^- → MnO₂(s) + 4OH^- (5)

This reaction indicates that oxidation by MnO₄^- at neutral pH is accompanied by the production of manganese dioxide (MnO₂) solid and release of hydroxide (OH^-). However, under acidic conditions (for pH less than 3.5) the dominant half reaction for MnO₄^- is:

MnO₄^- + 8H⁺ + 5e^- → Mn²⁺ + 4H₂O (6)

Thus, under acidic conditions, hydrogen is consumed to produce water and Mn (II) is yielded.

In comparison to Fenton’s reagent, permanganate is a weaker oxidant, but it has a longer reactive half-life. Consequently, it is easier to control delivery and distribution of permanganate in the subsurface, and permanganate is more amenable to recirculatory designs for source area remediation.

3.1.3 Persulfate

Persulfate (a.k.a., peroxodisulfate; S₂O₈^2-) has received attention recently as a potential oxidant for ISCO treatment of organic contaminants. Persulfates are common oxidants used in plating, organic chemical synthesis, polymerization, and metal surface cleaning. The sodium salt (Na₂S₂O₈) is highly soluble in water (730 grams per liter (g/L) at 25 degrees Celsius (°C) and 860 g/L at 50°C) and can easily form a concentrated solution for subsurface delivery. Persulfate salts dissociate in water to persulfate anions which, although strong oxidants, are kinetically slow in destroying many organic contaminants. For ISCO applications, potassium persulfate has a low solubility, and the injection of ammonium persulfate may lead to the generation of ammonia, which is a regulated CoC in groundwater. Therefore, the most common salt used is sodium persulfate. The persulfate anion is a more powerful oxidant than hydrogen peroxide. Decomposition reactions vary with persulfate concentration, pH, and oxygen, and hydrogen peroxide or peroxymonosulfate can be produced. Under dilute acid conditions, hydrolysis of the persulfate anion yields bisulfate anions and hydrogen peroxide.

At ambient temperatures (15 to 20°C), the persulfate ion can act as an oxidant:

S₂O₈^-2 + 2e^- → 2SO₄^-2E° = 2.01 V (7)

Thermal homolysis (Eq. 8) of persulfate or its reaction with a suitable reductant, such as ferrous ion (Eq. 9-10), leads to the sulfate radical, which dramatically increases the oxidative strength of persulfate. Manganese or copper can also be used as reductants for formation of the sulfate radical.

heat + S₂O₈^2- → 2 SO₄^{- •} (8)

Fe²⁺ + S₂O₈²^• → Fe³⁺ + SO₄^{- •} + SO₄²^- (9)

Fe²⁺ + SO₄^{- •} → Fe³⁺ + SO₄²^{- •} (10)

SO₄^{- •} + H₂O → SO₄²^- + OH^• (11)

3.1.4 Ozone/Ozone with Peroxide

Ozone (O₃) can be used alone or in combination with peroxide to form one of the strongest oxidants available for ISCO. Ozone can oxidize organic contaminants either by direct oxidization by ozone or generation of free radical intermediates. The hydroxyl radicals are nonselective oxidizers that rapidly attack organic contaminants and break down their carbon-to-carbon bonds. Ozone can oxidize compounds such as aromatics and chlorinated alkenes, although oxidation by hydroxyl radicals is faster than oxidation by ozone itself.

Most of the literature on the use of ozone alone or in combination with hydrogen peroxide has been in water rather than in soil treatment. Ozone is reactive with certain functional groups, but in water ozone also produces hydroxyl radicals that often account for most of the reactivity (Figure 2).

Hydrogen peroxide is frequently added to ozonated water to increase the ozone decomposition rate. By accelerating the ozone decomposition rate, the hydroxyl radical concentration is elevated, which increases the overall oxidation rate. The balance of the reactions generated is shifted to increase the contribution of indirect oxidation by the hydroxyl radicals over the direct ozone oxidation, which may be preferable for some target compounds.

3.2 Advantages and Disadvantages of Each Oxidant

The advantages and disadvantages of the various chemical oxidation methods are as follows:

3.2.1 Fenton’s Reagent

Advantages:

The oxidant materials are inexpensive and readily available.

A wide range of chemicals can be treated including chlorinated solvents, fuel hydrocarbons, coal tar, PCBs, and PAHs.

Disadvantages:

The technology is limited by the interference of subsurface impurities and carbonate since bicarbonate and organic matter will create competing reactions that hinder performance.

An extremely exothermic reaction occurs, which can create safety and handling issues. Ground heaving and surface damage is possible for poorly designed applications.

For conventional Fenton’s applications, addition of concentrated acid is required, and groundwater pH post-treatment can be quite low (<5). In this case, treatment with Fenton’s reagent can effectively sterilize the soil and limit secondary treatment choices.

As a result of the short reaction half-life, the ROI of the reagent can be relatively small, thereby requiring a higher density of injection wells relative to other ISCO options.

Associated heat and bubbling can cause significant volatilization of volatile target contaminants, thereby necessitating supplemental use of soil vapor extraction to capture vapors in some cases.

3.2.2 Permanganate

Advantages:

It is typically more stable and safer to handle than Fenton’s reagent, does not require pH adjustment and produces less heat and insoluble gas in the treatment zone.

The relatively long reaction half-life (lower reactivity) of permanganate allows for flushing of treatments throughout subsurface, improved delivery of oxidant (e.g., relative to Fenton’s reagent), and greater ability to oxidize contaminants diffusing from the aquifer matrix.

Disadvantages:

Permanganate treats a narrower range of contaminants than the other oxidants. Although it can treat chlorinated ethenes, permanganate is not effective at treating chlorinated ethanes and may have limited effectiveness against benzene, toluene, ethylbenzene, or xylenes (BTEX).

Permanganate can be expensive.

Permeability reductions can occur near DNAPL source zones due to the formation of MnO₂ precipitates (e.g., MacKinnon and Thomson, 2002; Dai and Reitsma, 2002; Lee et al., 2003) and/or rapid production of CO₂(g) (Dai and Reitsma, 2002), resulting in less effective treatment over time.

Recirculation systems are prone to fouling with MnO₂ precipitates.

Strongly oxidizing conditions are created that can persist post-treatment, which may impact the effectiveness or choice of polishing technology (if required).

Dissolved metals mobilization may occur in some aquifers, depending on the mineral content of the geological material present.

Manganese precipitated as MnO₂ may mobilize as dissolved manganese if the groundwater geochemistry becomes reducing upon termination of the treatment.

3.2.3 Persulfate

Advantages:

Unlike permanganate, persulfate does not result in the accumulation of manganese, a constituent which could become a CoC over the long-term if MnO₂ dissolves.

Sulfate is a primary end-product of ISCO with persulfate, and sulfate can serve as an electron acceptor to facilitate subsequent degradation of any co-occurring fuel hydrocarbons residual.

A wide range of chemicals including chlorinated solvents, fuel hydrocarbons, and PAHs can be treated.

Disadvantages:

Persulfate typically requires the use of heat or ferrous iron catalyst. Ferrous ions require low pH or chelating agents to remain in solution. It may be necessary to lower the pH as with peroxide systems to achieve this environment.

Fe(II) does not appear to effectively activate persulfate with chlorinated ethanes (1,1,1-trichloroethane [1,1,1-TCA], etc.) and methanes (chloroform, etc.). However, recent work with persulfate under alkaline conditions demonstrates effectiveness against these contaminants.

The catalytic effect of the iron appears to decay with time and distance from injection. This decrease could be the result of either poor transport of the dissolved Fe(II) in a soil environment or the depletion of the iron as it activates the persulfate. Chelated iron effectively increases the iron solubility and longevity of Fe(II) in the groundwater.

Low pH conditions may be generated by persulfate decomposition, which can cause dissolved metal concentrations to increase in the groundwater. Natural soil buffering capacity can help alleviate this phenomenon.

Persulfate may degrade soft metals such as copper or brass. Well construction and injection materials should be compatible with long-term persulfate exposure. Appropriate materials include stainless steel, high-density polyethylene, and polyvinyl chloride (PVC).

As with all oxidants, metals can be mobilized within the treatment zone due to a change in oxidation states and/or pH.

3.2.4 Ozone/Ozone with Peroxide

Advantages:

The gaseous nature of ozone allows for ease of delivery through the vadose zone compared with the liquid oxidants (Looney and Falta, 2000).

Disadvantages:

The short half-life of ozone substantially limits its ability to migrate through the soil, thus this oxidant is generally considered useful only for small scale or vadose zone applications.

Even with the addition of peroxide, there is still a short reaction half-life and the ROI of the reagent can be relatively small, thereby requiring a relatively high density of injection wells.

An ozone generation system requires a large capital investment.

Ozone can create an indoor air quality issue.

Ozone is highly reactive with aquifer solids and groundwater constituents, yielding a high oxidant demand.

As with all oxidants, metals can be mobilized within the treatment zone due to a change in oxidation states and/or pH.

As with Fenton’s reagent, ozonation can be expected to decrease the indigenous microorganisms within the treated site; however, microbial populations are expected to rebound at a rate that depends inversely on the duration of ozonation (Jung et al, 2005).

3.3 General Considerations for In Situ Treatment

The considerations to be made when evaluating ISCO as a remedy for a site include the properties of the oxidant, the hydrogeology, and the type and distribution of contaminants in the subsurface. Based upon these data, an evaluation of the most appropriate oxidant and application method can be made. These considerations are described in the following subsections. The specific reactions between the oxidants and the CoCs for the Site are further described in Sections 4-7.

3.3.1 Oxidant Properties

Table 4 presents key oxidant properties to be considered for ISCO. The chemical and physical properties of chemical oxidants vary widely. As such, the design considerations for ISCO remedies differ significantly between the various types of chemical oxidants. For example, ISCO using Fenton’s reagent is an unstable, exothermic treatment process, while treatment using permanganate is a stable process that does not result in heat generation. The half-life for Fenton’s reactions is on the order of seconds to minutes; while the half-life for permanganate and persulfate reactions is on the order of days to months. Similarly, ozone is highly reactive and unstable, with a short reaction half-life. ISCO remedies that use Fenton’s reagent in DNAPL source areas typically involve 3 to 5 batch injections of reagent over a 4 to 12 month period. In contrast, ISCO remedies that use permanganate or persulfate may involve batch injection or continuous recirculation of reagents throughout the treatment zone.

Applications of Fenton’s reagent in the field must recognize the potential for heat and gas generation, ground heaving (e.g., proximal to asphalt), and corrosion of materials in the subsurface. Conventional Fenton’s reagent remedies require low-pH conditions, which can be engineered via the addition of acids. Alternatively, neutral-pH Fenton’s formulations can be used that employ chelating agents that increase the solubility of Fe(II) and Fe(III) and prevent precipitation of iron oxides at circumneutral pH.

Most of the ISCO reagents can result in the mobilization of metals from aquifer solids, depending on the composition of the aquifer matrix. A primary end-product of ISCO using permanganate is the precipitation of MnO₂, which may accumulate and occupy a significant fraction of the aquifer pore space. ISCO using persulfate generates sulfate, an innocuous end product that can be subsequently used to stimulate biodegradation of any remaining organics. Similarly, Fenton’s reagent generates iron oxides (rust), which may subsequently stimulate anaerobic ferrogenic biodegradation of residual contamination.

3.3.2 Geology

ISCO has been applied in both unconsolidated and consolidated media. To effectively degrade contaminants, the oxidant must come into contact with the contaminant molecules. As with all remediation technologies that require the delivery of an amendment to the treatment zone, the more heterogeneous the media the more difficult it is to effectively distribute the amendments. Subsurface heterogeneities, preferential flowpaths, or low soil permeability can result in uneven flushing of the oxidant through the subsurface, resulting in untreated contaminants. To properly design and implement ISCO the treatment site must be adequately characterized including determining the nature and mass of contaminants (including the sorbed, dissolved and/or non-aqueous phases), having an understanding to the geology (including migration pathways for the contaminants), and an understanding of the hydrogeology of a given site.

3.3.3 Plume versus Source Areas

ISCO can be applied over a range of contaminant concentrations from source area to plume concentrations. Chemical oxidants treat contaminants in the dissolved phase; however, as degradation of the aqueous phase contaminants occur, enhanced desorption of the sorbed contaminants and dissolution of nonaqueous phase liquid (NAPL) contaminants can occur. The remedial design must account for the dissolved-phase, non-aqueous phase and sorbed mass for effective site remediation. For some ISCO technologies (e.g., Fenton’s reagent), source area applications are the most suitable use of the technology as a result of the short reaction half-life. For other ISCO technologies such as permanganate, both source and plume application are possible because of the better reactivity longevity of the oxidant. For certain plume applications, treatment using ISCO may be prohibitively expensive due to the size of the plume, or lower cost of other treatment alternatives (e.g., bioremediation). Chemical oxidation accelerates the remediation of NAPL source areas through treatment of dissolved phase contamination near the NAPL/water interface. The destruction of the dissolved-phase contaminants enhances the dissolution gradient at the NAPL/water interface, increasing the overall mass transfer of the contaminant from the NAPL to the dissolved phase, and thus depleting the NAPL at a faster rate. The maximum enhancement of NAPL removal is primarily based upon the total effective surface area over which oxidation can occur and the reaction rate. Other factors that influence NAPL removal include effective delivery of the oxidant to the contaminated media, consumption of the oxidant by other organic material in the aquifer, and the contaminant solubility.

3.3.4 Application/Installation Methods

In general, ISCO amendments are introduced to the subsurface through a number of injection wells or temporary/direct-push injection points. To achieve adequate contact between the oxidant and the contamination, an adequate fraction of the pore-volume of the target area must be filled or flushed with the oxidant. However, care must be taken not to displace the contamination with excessive amounts of oxidant injection. Once laboratory treatability testing has been conducted to ensure that ISCO is an appropriate remediation technology for a site, pilot studies can be conducted to provide the necessary information for full-scale design including the appropriate injection well/point spacing and appropriate injection flow rates for liquid or gas amendment delivery.

Oxidant concentrations need to be high enough to meet the natural oxidant demand of the aquifer, as well as the demand of any contaminant encountered during flushing. However, excessive concentrations of oxidant are not desirable due to potential impacts on secondary groundwater quality (e.g., color, pH, dissolved metals) and higher costs.

3.4 Full Scale Design, Operations and Maintenance, Safety and Cost Considerations

3.4.1 Data Requirements for Design

Location of all underground utilities;

Vertical and horizontal delineation of contaminant distributions in zone requiring treatment;

Vertical characterization of geology within treatment zone, to understand where high-flow layers exist; design needs to consider location of preferential flow paths. For Fenton’s applications, existence of preferential flow paths (e.g., in fractured bedrock) can significantly impact fate and transport of injected oxidants and off-gas;

Characterization of aquifer geochemistry, including any reduced metal species that are susceptible to oxidation due to chemical oxidants;

Hydraulic conductivity of zone requiring treatment (may require slug testing and/or aquifer pump testing);

Estimate of groundwater flow rates and travel times under operating conditions;

Soil oxidant demand test (can be screened in the field, using test kits, and/or fully studied in the laboratory);

In some cases, laboratory bench test to confirm treatability, determine site-specific treatment rates, and potential for mobilization of oxidizable metals (e.g., manganese and chromium) from aquifer solids; and

In some cases, pilot testing to confirm ROI, fate of injected oxidants, mobilization of metals, and overall treatment performance. Results from pilot testing are used to support design of a full-scale system.

3.4.2 O&M Requirements

For batch injection approaches, operation and maintenance (O&M) requirements are minimal because remedy implementation is rapid and of short duration;

For continuous injection or recirculation systems (e.g., permanganate recirculation in a NAPL source area), O&M can include replacement of in-line filters, maintenance of pumps, redevelopment of injection and extraction wells, maintenance of oxidant supply tanks, monitoring of water level in extraction and injection wells, etc.;

Performance monitoring (e.g., quarterly) to assess treatment performance, fate of oxidants, and mobilization of dissolved metals; and

For batch injection approaches, three to five injections might be required (as determined by performance monitoring) for treatment objectives to be achieved.

3.4.3 Safety

During the application of ISCO there are a number of health and safety considerations which are unique to these technologies, including the following (ITRC, 2005):

Safe storage and handling of the oxidants is essential.

If permanganate and persulfate are purchased as a solid powder, the powder must be controlled to control the potential for inhalation as it is harmful to the respiratory system.

Ozone requires careful monitoring as it is toxic to breathe and increases the flammability of many materials.

Ozone generation can require using high-voltage equipment.

Fenton’s reagent and persulfate can require the injection of multiple reagents for mixture in the subsurface. Proper injection equipment must be used to minimize the potential for above ground reactions, which can be exothermic or explosive.

If underground utilities exist there is a potential for preferential migration of oxidants and/or contaminants in the subsurface.

3.4.4 Cost

The major cost items that should be included in a cost estimate for ISCO are pre-treatment activities (laboratory and/or pilot study work), fixed cost items including injection point and sampling well installation; and, variable cost items including site supervision, chemicals, sampling and analysis for process control and O&M. For oxidants that are added in a batch approach, the major cost items are generally the installation of the injection points and the chemical reagents. For oxidants added in a recirculation approach the major cost items are generally the process equipment required for recirculation and the chemical reagents. Development of cost estimates for field work is generally developed based upon the results of laboratory testing which is used to confirm the reaction rates of CoCs and the oxidant demand of the site matrix. Costs for example Fenton’s applications are provided by U.S. Department of Energy ([USDOE]1999a), Yin and Allen (1999), and ITRC (2005).

4 Fenton’s Reagent Reactions with Organophosphorus Pesticides

4.1 Reaction Chemistry with Organophosphorus Pesticides and other Contaminants at the Cheminova Site

Among the hundreds of papers on the Fenton reaction there are relatively few dealing specifically with the Cheminova CoCs, and only one discusses treatment in soil media. Among the Cheminova CoCs, rate constants for the elementary reaction with hydroxyl radical (Eq. 3) are known only for 4-nitrophenol (k_OH = 3.8 × 10⁹ per mol per second [M^-1s^-1]), its conjugate base 4-nitrophenoxide ion (k_OH = 7.6 × 10⁹ M^-1s^-1), and dimethyl phosphate (k_OH = 1.2 × 10⁸ M^-1s^-1). (Radiation Chemistry Data Center of the Notre Dame Radiation Laboratory, http://allen.rad.nd.edu/). The former is a likely byproduct of methyl parathion and parathion oxidation, while the latter is a likely byproduct of methyl parathion oxidation.

Employing the dark Fenton reaction, Doong and Chang (1998) observed only 20% loss of malathion after 24 hours. However, in the photo-Fenton reaction, malathion was 94% reacted in 30 minutes (min) giving quantitative yield of sulfate, 35% yield of phosphate, and no loss of total organic carbon (TOC) (Huston and Pignatello, 1999). Organic products included formate, oxalate and acetate. In another photo-Fenton study (Doong and Chang, 1998), 1 gram per liter (g/L) Fe⁰ (as iron powder) gave comparable results as 50 micromoles (mM) Fe²⁺.

Methyl parathion was rapidly degraded by the photo-Fenton reaction (Pignatello and Sun, 1995). Under initial conditions listed in Table 5, methyl parathion reacted completely in 5 min, giving stoichiometric yields of sulfate and nitrate within that time. Phosphate was evolved stoichiometrically within 30 min. The transient organic intermediates identified included 4-nitrophenol, methyl paraoxon, dimethyl phosphate and oxalate. 4-Nitrophenol disappeared within 5 min. Under the same conditions, a commercial standard of uniformly-¹⁴C-labelled 4-nitrophenol evolved nitrate and ¹⁴CO₂, both quantitatively. A standard of dimethyl phosphate was degraded with a half-life of ~12 min. Importantly, maximum methyl paraoxon yield from methyl parathion was <<1%. Degradation of a commercially available analog of methyl paraoxon, ethyl paraoxon, degraded with a half-life of ~2 min, giving 4-nitrophenol as the major organic product.

Methyl parathion was treated in soil slurries under Fenton conditions using a chelated form of Fe(III) (Pignatello and Day, 1996). Since inorganic ferric ion is insoluble above pH ~3 (well below the normal pH of the soil, ~ 6), the chelating agent served to enhance the soluble concentration of Fe(III) and, thus, increase its availability in the Fenton reaction (Pignatello and Baehr, 1994; Nam et al., 2001). Two chelating agents were investigated: nitrilotriacetate (NTA) and N-(2-hydroxyethyl)iminodiacetate (HEIDA). Their performance was comparable.

Soil pre-equilibrated with methyl parathion was slurried with water (1:1 mass ratio) containing the Fenton reagents. The loss of methyl parathion, which leveled off after about 10 hours, reached as high as 88%, depending on reagent concentrations and temperature. The optimum concentration of the Fe-chelate was 0.01 moles per kilogram (mol/kg). Using this concentration of chelating agent, at 21°C, up to 80% loss of methyl parathion could be achieved with 6 mol/kg H₂O₂ added in two batches, while at the slightly elevated temperature of 35°C 88% loss could be achieved with 1 mol/kg H₂O₂, corresponding to a peroxide-to-methyl parathion molar ratio of about 130. This ratio is about an order of magnitude greater than that required to completely mineralize methyl parathion in water. The higher oxidant demand in soil is due to catalytic decomposition of H₂O₂ by soil components and scavenging of hydroxyl radicals by natural organic matter. A partial characterization of products indicated stoichiometric yields of nitrate and sulfate, extensive degradation of the ring (4-nitrophenol was detected in less than 5% yield), but only partial degradation of the organophosphorus group (dimethyl phosphate was determined in 22-36% yield). The presence of methyl paraoxon was not determined. The pH of the soil declined from 6 to 4.1 during the treatment.

4.2 Reaction Chemistry with Other Organophosphorus Pesticides

Dimethoate degradation by the photo-Fenton reaction was studied by Nikolaki et al. (2005). They observed extensive loss of TOC from solution and generation of sulfate (60% yield), phosphate (50%), and ammonium (40%) after ~3 h of UV irradiation. In addition, they detected transient levels of dimethyl phosphate, N-methylacetamide, and formic acid.

Structural formulas

In the dark Fenton reaction (50 mM Fe²⁺, 0.6 mM H₂O₂, pH 7.2) Doong and Chang (1998) observed moderate losses after 24 hours of methamidophos (39%), diazinon (29%), phorate (28%) and EPN (12%). Much faster rates and greater losses of these organophosphorus pesticides were obtained when the samples were illuminated with a medium-pressure UV lamp. The order in reactivity was phorate > methamidophos > EPN > diazinon > malathion.

p-Nitrophenol, a common byproduct of the parathions, reacts rapidly with Fenton’s reagent, and has often been used to model reaction kinetics (Khan et al., 2005; Chirchi and Ghorbel, 2002; Goi and Trapido, 2002; Kiwi et al., 1994; Lipczynska-Kochany, 1992; Kavitha and Palanivelu, 2005). Degradation rapidly proceeds to ring-opened products that depend on whether the solution is illuminated (Kavitha and Palanivelu, 2005; Pignatello and Sun, 1995). In the dark, the major products are nitrite, nitrate, oxalate, acetate and CO₂. Only about 30% of the theoretical amount of CO₂ is evolved. Under UV or solar illumination, oxalate and acetate are mineralized via photolysis of their Fe(III) complexes to provide the remaining CO₂. In both dark and illuminated reactions, nitrite is oxidized and the final product is nitrate in stoichiometric yield.

4.3 Summary of Implications for Treatability Test

Successful Fenton treatment of soils on a laboratory scale has been demonstrated for chlorinated solvents, PAHs, PCBs, pesticides, explosives, fuels and fuel components (reviewed in Pignatello et al., 2006).

Commercial-scale systems based on Fenton technologies have been explored for the treatment of groundwater and soils by in situ and ex situ approaches (USEPA 1998; USDOE, 1999a). The general flow configuration for in situ treatment consists of a mixing head that combines catalyst and hydrogen peroxide solutions from separate reservoirs in the injection well. Positive displacement of reaction solutions into the aquifer are maintained by externally applied compressed air or the back pressure from CO₂ and O₂ generated from the oxidation reactions. In-place soil mixing has been used to increase contact between oxidants and contaminants, but this approach is generally limited to applications in shallow aquifer systems. Field trials conducted at sites with chlorinated solvent DNAPL contamination have shown some success at reducing groundwater contaminant concentrations (USDOE, 1999a). The observed ‘rebound’ of contaminant concentrations is likely attributable to poor contact between oxidant and contaminant zones, as well as desorption from the aquifer matrix.

Fenton technologies applied to the cleanup of natural solids face significant obstacles: interference by soil components, the pH limitation typical of Fenton reactions, difficulties in effective dispersal of reagents, and potential alteration of the soil environment.

The amount of hydrogen peroxide needed to transform, and especially mineralize, a given concentration of contaminant in soil is often far greater (factor of 10-100) than in aqueous systems due to i) the presence of natural organic matter; ii) nonproductive catalyzed decomposition of H₂O₂ to O₂ and H₂O, or iii) the presence of inorganic reductants in soil that consume H₂O₂ (Pignatello et al., 2006).

Lowering the pH to below 4 keeps Fe(III) soluble and reduces nonproductive decomposition of H₂O₂ (Baciocchi et al., 2003). Acidification of soil is difficult due to the high buffering capacity of soil, and is potentially polluting itself. The pH of the substrata at the Site (2.4 - 4.3) may be favorable for the Fenton reaction and obviate the need for acidification. If not, the addition of ferric ion chelating agents may be considered.

Oxidant addition can initially reduce aquifer microbial populations to nil (Miller et al., 1996), but populations can rebound (Ferguson et al., 2004; Chapelle et al., 2005). The formation of large amounts of iron oxyhydroxide precipitate (an end product of the Fenton reaction) may induce shifts in microbial communities to this solid as a terminal electron acceptor from others (e.g., SO₄²^-) (Chapelle et al. 2005).

Since Fe occurs naturally, one could hope to avoid having to add it. However, addition of peroxide alone is usually insufficient to achieve degradation before it decomposes nonproductively (Pignatello and Baehr, 1994; Miller et al., 1996). While iron minerals (goethite, magnetite, hematite) can catalyze the Fenton reaction, they are much less reactive than soluble iron, especially when the pH is not lowered.

Application of Fenton’s reagent in the field for in situ decontamination of aquifers may also suffer from problems related to effective dispersal of reagents. Iron(III) sorbs strongly to mineral surfaces and organic matter depending on pH and may not travel very far from the point of injection. Aquifer plugging from precipitated iron oxyhydroxides has been an issue in some field trials. Another problem is the rapid decomposition of peroxide near the point of injection which can result in gas eruption on the surface. Addition of phosphate stabilizers is thought to help in this regard (Kakarla and Watts, 1997), but after a while phosphate may be depleted by adsorption.

Fenton’s chemistry is the only oxidant commonly used for ISCO that has been well demonstrated to oxidize the types of organophosphorus pesticides found at the Høfde 42 site. It is recommended that treatability tests be conducted with Site groundwater to evaluate the treatment performance of conventional and/or modified Fenton’s reagent. The reactivity and degradation products from the organophosphorus pesticides should be quantified to evaluate the applicability of Fenton’s reagent for field pilot testing at the Site.

5 Permanganate Reactions with Organophosphorus Pesticides

5.1 Reaction Chemistry with Organophosphorus Pesticides and other Contaminants at the Cheminova Site

Permanganate oxidizes some organic compounds and is itself reduced to MnO₂ which is environmentally benign. The products of the reaction include hydrogen ions or hydroxide ions, depending on the target compound. Permanganate is much more specific in its reactivity with organic compounds than hydroxyl radical-generating reagents. In organic synthesis permanganate is widely used to hydroxylate or cleave alkenes, convert alcohols to ketones or carboxylic acids, oxidize aldehydes to carboxylic acids, and oxidize amines to nitro compounds. A few other less common oxidations are also known. In situ treatment applications of permanganate for soil or groundwater contaminants have largely been restricted to chlorinated alkenes, which are effectively mineralized by the reagent. Nevertheless, oxidation of compounds not typically thought of as susceptible to permanganate has been observed. Brown et al. (2003) spiked soil (1.3% organic matter) with a mixture of six PAHs and observed 8 to 72% degradation depending on the PAH. Gates-Anderson et al. (2001) obtained 99% loss each of naphthalene, phenanthrene and pyrene spiked to low organic carbon soils. MTBE is oxidized slowly (Damm et al., 2002). Even RDX (hexahydro-1,3,5-trinitro-1,3,5-triazine), an explosive in which the carbon atoms are on average in a highly oxidized state, was 87% mineralized by permanganate in slurries of aquifer solids (Adam et al., 2004).

Very few published studies exist of permanganate treatment of the Site contaminants. Pugliese et al (2004) investigated the ability of permanganate solutions to remove pesticide residues on nectarines. While removal of malathion and methyl parathion from the fruit by the reagent (25 milligram per liter [mg/L] KMnO₄) was no more effective than plain water, they showed in separate experiments that 0.1 molar (M) KMnO₄ converted malathion to malaoxon and methyl parathion to methyl paraoxon in 5 min, although the yields were not reported. The oxons are regarded as toxic. Nitrophenols are reported to be oxidized by permanganate (Radhakrishnamurti and Sahu, 1976). No data is available on the other Site contaminants.

5.2 Summary of Implications for Treatability Test

Well-to-well recirculation techniques have been successful in remediation of chlorinated alkene DNAPL contamination by permanganate (Lowe et al, 2002). However, in many field applications, increased injection pressures and reduced circulation rates occur as a result of pore clogging by particulate MnO₂ and the effects of CO₂ evolution (Siegrist et al., 2002; Li and Schwartz, 2000; Conrad et al., 2002; MacKinnon and Thomson, 2002). For example, in flow tank experiments, Li and Schwartz (2000) found that the DNAPL (TCE) mass removal rate decreased dramatically as treatment proceeded due to increasing flow divergence around zones of higher DNAPL saturation as MnO₂ precipitates and CO₂ gas reduced permeability in these zones. This can cause a large amount of permanganate to leave the contaminated zone during flushing (Lee et al., 2003), and thus the potential for pollution of the aquifer by permanganate. In other cases the effects of pore plugging were not significant (Struse et al., 2002; Crimi and Siegrist, 2003). For example, in aquifer solids taken from a DNAPL-contaminated Launch Complex at Cape Canaveral Air Station in Florida, Crimi and Siegrist (2003) showed there was potential for long-term immobilization of a portion of introduced manganese and no induced loss in subsurface permeability due to deposition of manganese oxides particles. Permanganate treatment did, however, cause elevated manganese, chromium, and nickel concentrations in site ground water within the treated region.

One advantage of permanganate is that it is more persistent than peroxide or ozone in the subsurface. Permanganate is stable in aqueous solution. In soil, it can be expected to be consumed by reaction with organic matter and metal ions in low oxidation states; this would have to be determined on soil specific to the site. In soil column and batch studies, Mumford et al. (2005) showed that permanganate reacts with aquifer materials by fast (>7 g of MnO₄ per kg per day (g/kg/day)) and slow (~0.005 g/kg/day) rates, but it was unclear whether the slower stage was due to intrinsically slower reactions or diffusion rate-limitations. Only a fraction (10-40%) of the organic carbon was mineralized over a 14-week period. Unlike peroxide or ozone, permanganate is not expected to be catalytically decomposed by soil components. Another advantage of permanganate oxidation over the Fenton reaction is that it generally does not require pH adjustment. Permanganate compared favorably with Fenton’s reagent for treatment of a mixture of volatile organic compounds (TCE, PCE and 1,1,1-TCA) or PAHs (naphthalene, phenanthrene, pyrene) in soil slurries (Gates-Anderson et al., 2001); the Fenton reaction required acidification to pH 3.

Various phase-transfer catalysts may be employed to facilitate reactions with permanganate. Co-solvents such as acetic acid, acetone, or tert-butyl alcohol (TBA) can increase the solubility of certain organic compounds in water where the oxidation takes place (House, 1972; Zhai et al., 2006). Kang et al. (2004) describe the use of paraffin wax-encapsulated KMnO₄ to treat chlorinated DNAPLs; the particles preferentially accumulated at the DNAPL interface where permanganate was released as the wax dissolved in the DNAPL. The surfactant sodium dodecylsulfate increased the rate of permanganate oxidation of TCE by several fold, even at concentrations below its critical micelle concentration (Li, 2004). Cationic phase-transfer catalysts—tetraethylammonium bromide, tetrabutylammonium bromide, and pentyltriphenylphosphonium bromide—also enhanced the rate of TCE destruction in two-phase mixtures of the DNAPL and water, although the effect was more modest (Seol and Schwartz, 2000). It is essential to recognize, however, that co-solvents may interfere with mineralization (Zhai et al., 2006), or may substantially increase the oxidant demand if the co-solvent can be oxidized by the oxidant. In addition, co-solvents typically need to be added at a minimum concentration of 10% of the solution in order to increase the solubility of target contaminants (Schwarzenbach et al., 1993).

Permanganate potentially can attack the thiophosphate group to form the corresponding oxon; the amino group of amino parathions to form the corresponding nitro compound; or the aromatic ring of the parathions and their byproducts, leading to hydroxylation of the ring and possibly ring cleavage. The products resulting from initial attack at these positions may react further with permanganate to give innocuous products. The limited treatability results available (Pugliese et al., 2004) suggested that organophosphorus pesticides are oxidized by permanganate, however the final products from this oxidation reaction are not known. Further laboratory work is necessary to further evaluate this oxidant’s capability with organophosphorus pesticides. Thus, a laboratory treatability test of permanganate is recommended for the Høfde 42 project.

6 Persulfate Reactions with Organophosphorus Pesticides

6.1 Reaction Chemistry with Organophosphorus Pesticides and other Contaminants at the Cheminova Site

Persulfate has only recently been widely evaluated for in situ treatment of groundwater contaminants. As a result, no data is available on Site contaminants or related compounds. A discussion of the reactivity of persulfate towards other organic contaminants follows. The thermally-induced reaction is slow at room temperature but still perceptible for many compounds (Huang et al., 2002; Huang et al., 2005). For example, Huang et al. (2002) observed half-lives of MTBE ranging from 14.8 hours (h) at 20°C to 0.25 h at 50°C at 31 mM initial persulfate concentration, affording an Arrhenius activation energy of 103 kilojoules per mol (kJ/mol). MTBE gave products typical of hydroxyl radical advanced oxidation processes (AOPs). However, MTBE reacted dramatically slower (roughly, factor of seven) in groundwater than in phosphate buffer solution. This was attributed to radical scavenging by carbonate ion.

Huang et al. (2005) monitored the rates of degradation of 59 volatile organic compounds (VOCs) present as a mixture in persulfate solution at 1 or 5 g/L. At the higher concentration, degradation was complete or nearly so for 37 out of 59 compounds. Activation energies (measured between 20 and 40°C) for select compounds ranged from 41 kJ/mole (vinyl chloride) to 92.9 kJ/mole (o-xylene). The most reactive compounds were the alkenes and the substituted benzenes. The least reactive were the halogenated alkanes.

Liang et al. (2003) observed rapid oxidation of 60 mg/L TCE or 1,1,1- TCA above 40°C in water using a 10:1 persulfate:contaminant molar ratio. Under comparable conditions in soil slurries (1:5 soil-water mass ratio), the half lives were several times greater. Organic matter appeared to act in a dual role as competing substance (i.e., radical scavenger) and reaction promoter by providing a source of ferrous ion. Further study showed that supplemental Fe(II) accelerates degradation of TCE (Liang et al., 2004a) but that Fe(II) participates as a stoichiometric reagent (i.e., is not regenerated from the (III) state) and leads to unproductive decomposition of persulfate presumably by scavenging SO₄^{- •} as shown in Eq. 9.

Addition of sodium thiosulfate following Fe(II) addition gave opposing effects of regenerating Fe(II) from Fe(III) on the one hand and scavenging SO₄^{- •} and OH^• on the other. Addition of citrate ion improved the ability of Fe(II) to accelerate degradation (Liang et al., 2004b). Optimum results were obtained at a persulfate-citrate-Fe²⁺-TCE molar ratio of 20:2:10:1. Citrate may influence the reaction through its ability to chelate Fe(II) or Fe(III), but exactly how it participates is unclear.

6.2 Summary of Implications for Treatability Test

Persulfate readily attacks aromatic compounds and, therefore, is likely to react with parathions and p-nitrophenol. Byproducts are expected to be similar to those obtained in hydroxyl radical AOPs. Few data exist, however, to allow prediction of the reactivity of persulfate towards Sulfotepp, malathion, triethyl phosphate and diethyl phosphate. The rate constant between hydroxyl and dimethyl phosphate (k_OH = 1.2 × 10⁸ M^-1s^-1; Buxton et al., 1988) is relatively low. In any case, reactions at in situ temperature are expected to be slow. The use of radio frequency or electrical resistance heating may be considered for activation of this oxidant. However, subsurface heating to achieve ISCO using persulfate would make a field program very costly. Given the lack of available laboratory evidence for treatment of organophosphorus pesticides with this oxidant, and the likely high costs associated with a heat activated persulfate field program, laboratory testing of persulfate is not recommended for the Høfde 42 project at this time.

7 Ozone and Ozone+Hydrogen Peroxide Reactions with Organophosphorus Pesticides

7.1 Reaction Chemistry with Cheminova Site Organophosphorus Pesticides and other Contaminants

The only data available on ozonation among the Site contaminants pertains to parathion, methyl parathion, and 4-nitrophenol. Information regarding treatment conditions, rates, and end products for ozone studies identified in this literature review are summarized in Table 6.

Ozonation of parathion dissolved in 95% ethanol resulted in substantial conversion to paraoxon accompanied by formation of sulfate (Gunther et al., 1970). Paraoxon was not destroyed under the same conditions. The reaction in ethanol may be relevant to Høfde 42 because the solvation properties of ethanol may mimic those of the DNAPL existing at the Site. In water, Ku et al. (1998) obtained stoichiometric yields of nitrate and sulfate after reacting parathion with ozone for 1 h. Paraoxon was not monitored. Phosphate and carbonate (from CO₂) evolved more slowly indicating the production of organophosphorus esters and ring breakdown products as transient intermediates. The rate of parathion loss was dependent on ozone concentration and independent of pH in the range 3-9. Laplanche et al. (1984) observed ozone-mediated decomposition of parathion to phosphate and p-nitrophenol. Meijers et al. (1995) report that, at the ozone dosage required to just reach the disinfection level of drinking water, methyl parathion was only partially degraded. Spencer et al. (1980) demonstrated ozone transformation of parathion to paraoxon on soil dusts or clay particles at 30% relative humidity. The yields of paraoxon varied with ozone concentration (30 ppbv or 300 ppbv) and were lower in the absence of UV light. The reactivity of paraoxon separately was not examined.

p-Nitrophenol, a common byproduct of the parathions, reacts quite rapidly with ozone in the dark and (especially) with UV light, and is often used to model ozone reaction kinetics (Beltrán et al., 1992; Shi et al., 2005; Beltrán et al., 2005; Gimeno et al., 2005; Goi et al., 2004; Yu and Yu, 2000; Yu and Yu, 2001; Ku et al., 1998; Barberis and Howarth, 1991). Among the products detected (Shi et al., 2005; Goi et al., 2004; Yu and Yu, 2001) are: catechol, 4-nitrocatechol, o-benzoquinone, p-benzoquinone, hydroquinone, phenol, fumaric acid, maleic acid, oxalic acid, formic acid, nitrate and oxidative coupling products.

Meijers et al. (1995) studied the ozonation of several organophosphorus pesticides in drinking water at ozone dosages required to just reach the disinfection level. The extent of degradation ranged from 96% to 28% and followed the order: dimethoate > diazinon ~ methyl parathion > chlorfenvinphos >tetrachlorvinphos. The extent of degradation increased with pH (7.2-8.3), temperature (5-20°C) and O₃/DOC ratio (0.53-0.95). The products were not identified. Meijers et al. (1995) observed that the addition of hydrogen peroxide, prior to ozonation, increased the formation of hydroxyl radicals, and improved the extent of pesticide treatment. Out of 23 pesticides tested, 21 including methyl parathion were effectively degraded by AOP at an ozone dosage of 3.0 mg/L (O₃/DOC = 1.4 g/g) and H₂O₂/O₃ ratio of 0.5 to 2.5 g/g. The authors concluded that relative to treatment by ozone alone, persistent pesticides can be degraded more effectively by dosing with hydrogen peroxide followed by ozonation.

7.2 Summary of Implications for Treatability Test

Ethyl and methyl paraoxon may be produced in high yield during ozonation reactions. Since these compounds are highly toxic, their concentrations must be monitored carefully in treatability studies, and conditions recommended for field application must be chosen to minimize or eliminate their formation.

Ozone can be injected as a gas or as a solution in water (solubility, 1 g/L at 0°C). For contaminants in the saturated zone, ozone can be injected either as a gas (sparging) or in aqueous solution. For contaminants in the unsaturated zone the preferred method is gas injection. Simulations indicate that ozone can be effective in the unsaturated zone provided that efficient circulation of ozone is achieved (Shin et al., 2004). These simulations were based on a contaminant (TCE) concentration of only 0.0015 mol/kg, far below what is necessary to form a residual NAPL phase. The rate of ozone delivery in the unsaturated zone is dictated by the relative gas-to-liquid film transfer rate (including chemical reaction) divided by the gas convection rate (Sung and Huang, 2002).

Little information on ozonation in the field is available in the peer-reviewed literature. Cases in which ozonation has been used for subsurface treatment (often, coupled with soil vapor extraction or air stripping) are summarized briefly in USEPA, 1998. Among these cases is one involving semi-volatile contaminants where — like the Site contaminants ¾ gas stripping is not likely to contribute significantly to contaminant removal compared to oxidation. At this former wood treating site in Sonoma, California (USEPA, 1998), contamination by pentachlorophenol and creosote extended from shallow soils down to the water table. After one month of continuous ozonation, pentachlorophenol and PAH concentrations in the solids were reduced by 38 to 99.5%. However, it was not specified where the samples were collected with respect to the water table.

Most reports in the primary literature are based on batch or soil column experiments, sometimes combined with mathematical modeling. When ozone is pumped through an artificially contaminated soil column the degree of degradation increases with increasing ozonation time, decreasing soil particle size, decreasing contaminant concentration, decreasing moisture content, and decreasing scavenger concentration (e.g., soil organic matter or bicarbonate ion) (Zhang et al, 2005; Masten and Davies, 1997).

Ozone must be delivered quickly to the contaminant plume because of its inherent instability. Ozone self-decomposes in both air and water, with a half-life of a few days and a few tens of minutes, respectively, at 20°C. Surfaces catalyze ozone decomposition. For example, the decomposition of ozone in a dry uncontaminated silica sand column of low organic matter content (0.03%) was first order and ozone had a half-life of 1.1 h (Yu et al., 2005). The rate of decomposition decreased with increasing moisture content. Shin et al. (2004) partitioned the rate of ozone decomposition in reactors with soil but without contaminant into self-decomposition, surface-catalyzed decomposition, and consumption by soil organic matter. The latter two processes are expected to dominate under field conditions. Organic matter, in contrast to surfaces, becomes less and less effective with increasing exposure to ozone. It was calculated that between 25 and 40% of soil organic matter is available for consumption by ozone. With increasing water content the rates of both surface-catalyzed decomposition and consumption by organic matter are reduced (Choi et al., 2002; Jung et al., 2004).

The question of whether or not hydrogen peroxide should be combined with ozone is relevant. The production of hydroxyl radical from ozone is not dependent on addition of hydrogen peroxide, since hydroperoxyl radical (HO₂^•) and its conjugate anion, superoxide ion (O₂^-^•), may be generated by reaction of organoradical intermediates with oxygen. Compounds that convert OH^• to HO₂^•/O₂^-^• ¾ not all do ¾ act as promoters of the chain (see Figure 2). Hydrogen peroxide can be a chain carrier though its conjugate anion (HO₂^- ; pK_a(H₂O₂) = 11.6). However, under the acidic conditions of the Site, the concentration of HO₂^- will be exceedingly low. It therefore seems unlikely that hydrogen peroxide would improve remediation at the Site. Nevertheless, hydrogen peroxide is known to enhance reactivity of ozone under certain conditions, and the combination of ozone and hydrogen peroxide has been shown to achieve effective treatment of certain CoCs, including methyl parathion and other persistent pesticides (e.g., see USEPA, 2004; and Meijers et al. 1995).

Further laboratory work is necessary to evaluate the capability of ozone and ozone + peroxide to treat organophosphorus pesticides. Thus, a laboratory treatability test of ozone and/or ozone + peroxide is recommended for the Høfde 42 project.

8 Case Studies of ISCO Field Applications

Two case studies have been selected to illustrate the efficacy of Fenton’s reagent and ozone with peroxide for in situ treatment. The first case study is an outline of the pilot-scale application of Fenton’s reagent at the Savannah River Site in Aiken, South Carolina (USDOE, 1999a). The second case study is a summary of an ozone and peroxide pilot and full scale tests conducted at an active retail gas station in Riverside, California (Brackin et al., 2005), which would be similar in approach to an ozone application for source zone remediation in groundwater. Table 7 summarizes the site conditions and remedial activities at both of these sites. Sections 8.1 to 8.2 below include summaries of the treatment operation and lessons learned from each application.

The following references available on the World Wide Web provide descriptions of additional ISCO case studies that demonstrate the efficacy of this technology for DNAPL remediation:

Environmental Security Technology Certification Program. 1999. In Situ Chemical Oxidation, Technology Status. http://www.estcp.org/documents/techdocs/ISO_Report.pdf

Interstate Technology and Regulatory Council. 2005. Technical and Regulatory Guidance for In Situ Chemical Oxidation of Contaminated Soil and Groundwater. http://www.itrcweb.org/Documents/ISCO-2.pdf

U.S. Department of Energy. 1999a. Fenton’s Reagent. DOE/EM-0484. Office of Environmental Management, Office of Science and Technology. http://apps.em.doe.gov/OST/pubs/itsrs/itsr2161.pdf

U.S. Department of Energy. 1999b. In Situ Chemical Oxidation Using Potassium Permanganate. DOE/EM-0496. Office of Environmental Management, Office of Science and Technology. http://apps.em.doe.gov/OST/pubs/itsrs/itsr167.pdf

U.S. Environmental Protection Agency. 1998. Field Applications of In Situ Remediation Technologies: Chemical Oxidation. Office of Solid Waste and Emergency Response. EPA 542-R-98-008. http://www.clu-in.org/download/remed/chemox.pdf

Yin, Y, and H.E. Allen. 1999. In Situ Chemical Treatment. Ground-Water Remediation Technologies Analysis Center. Technology Evaluation Report No. TE-99-01. http://www.groundwatercentral.info/org/pdf/E_inchem.pdf.

8.1 Case Study #1: Fenton’s Reagent Demonstration, Savannah River Site Aiken, South Carolina

A demonstration of ISCO using Fenton’s reagent was conducted at the Savannah River Site in Aiken, South Carolina. The site geology included permeable sands with low fines alternating with clayey sand and clay units. The area targeted for the demonstration was approximately 50 feet (ft) by 50 ft. The area was adjacent to a known source of DNAPL with an estimated TCE mass of 600 pounds (lbs) within the treated area. The demonstration was conducted over a 6 day period.

Fenton’s reagent, comprised of a catalyst solution of 100 parts per million (ppm) ferrous sulfate, pH-adjusted with concentrated sulfuric acid, was introduced to the subsurface using 4 injection points which used a patented mixing and injection process. Injections were conducted in batch mode with one batch injected per day. Following 6 days of injection, the site was characterized to determine treatment efficiency.

A destruction efficiency of 94% was achieved, based upon results of soil sampling. In addition, groundwater concentrations were substantially decreased from 119.49 mg/L PCE and 21.31 mg/L TCE before treatment to 0.65 mg/L PCE and 0.07 mg/L TCE. Corresponding increases in chloride concentration confirmed that the DNAPL removal was the result of oxidation. Some metals mobilization was observed, however the concentrations remained below levels of concern.

An evaluation of the costs to implement the Fenton’s reagent technology was completed based upon the results of the demonstration, and compared to the cost per pound of DNAPL treated by pump-and-treat. The break even point for costs for Fenton’s reagent treatment versus pump-and-treat was sensitive to the depth of contamination and total mass. For the Savannah River Site, the break even point ranged from 6500 to 9500 pounds of DNAPL as depth of contamination increased from 60 to 155 ft. For sites with less than 4000 lbs DNAPL the unit cost is >$100/lb DNAPL and for sites with approximately 1000 lbs of DNAPL unit costs increase to greater than $700/lb of DNAPL.

Lessons learned during the technology application included the following (text in italics has been excerpted from USDOE [1999a]):

Design Issues:

"The efficiency of the process increases at higher contaminant concentrations and decreases as target treatment levels become more stringent.

Higher H₂O₂ concentrations provide faster reaction times, significantly greater removal of DNAPL type contaminants, but less efficient H₂O₂ use.

Highly alkaline soils may require mineral acid addition to bring the pH into the optimal range.

Organic carbon content may impact treatment because the hydroxyl radical is relatively nonselective. However, no significant effect was observed with contaminant levels of 500-2000 ppm with total organic carbon of 0.1 to 1.3 (Watts et al. 1994).

For in situ groundwater treatment, the number and pattern of injectors and monitoring wells must be designed to ensure maximum coverage of the treatment zone. Because the cost is related to depth (cost per well was approximately $70/ft) and amount of DNAPL, the number and spacing of the wells becomes critical. The heterogeneity of the subsurface at the site will also control the number and spacing of wells required.

Duration of operation is not a linear function of volume of DNAPL. Factors affecting the duration of the treatment include: permeability, heterogeneity, and geochemistry of the aquifer."

Implementation Considerations:

"When implementing ISCO using Fenton’s Reagent, general operation considerations include:

pH of the system must be between 3 and 6, for traditional Fenton’s reagent.

The rate of the reaction increases with increasing temperature (although the efficiency declines above 40 to 50°C).

For most applications the valence of the iron salts used doesn’t matter (+2 versus +3) nor does it matter whether a chloride or sulfate salt of the iron is used, although chlorine salts may generate high rates of chloride during application.

Due to oxidation of the subsurface, metals that are mobile under these conditions may be released at some sites. This should be considered during the technology selection process.

Implementation of this technology does not require permanent infrastructure, such as a permanent power source (temporary power is required), permanent water and chemical tanks, etc. Temporary power is required for operation of the system. This is much less expensive for the short duration of operation, typically less than 1 month and in many instances 1 to 2 weeks. Also required is a constant supply of water for process, as well as emergency, purposes. For remote sites where a distribution line with potable water is not available tanks for water storage are appropriate. During the demonstration, approximately 1000 gallons of water per day were used for a 6-day period.

The end products of in situ oxidation are very appealing. No waste is generated from the treatment process, and no material is brought to the surface.

At complex sites in situ oxidation using Fenton’s Reagent should be considered in tandem with other technologies. For example, if in situ bioremediation is considered as a polishing step, the pH should be held above 4.0 during the treatment operations."

Note that the above reflect the conclusions and opinions stated in USDOE (1999a), do not entirely reflect the opinions of the authors of this literature review report.

8.2 Case Study #2: Ozone with Peroxide, Active Retail Gas Station, Riverside, California

A pilot test and subsequent full scale implementation of AOP was performed at an active retail gas station in Riverside, California. The site geology included silts and sandy silts to approximately 15 feet below ground surface (bgs), fine-and coarse-grained sands to about 35 feet bgs, and weathered bedrock between 35 and 40 feet bgs at some locations. The water table was generally observed at 33 feet bgs. The primary CoCs at the site included total petroleum hydrocarbon ([TPH] 5,000 – 35,000 micrograms per liter (µg/L)), MTBE (5,000 to 200,000 µg/L), TBA (1,000 to 20,000 µg/L), and with BTEX (500 to 5,000 µg/L).

A pilot test was conducted using two sparge wells placed in the center of a source area for the injection of ozone, oxygen, hydrogen peroxide, and air into the contaminated groundwater. This combination of reagents was selected to provide the AOP process, to distribute the oxidant in the subsurface, and to stimulate aerobic bioremediation. Subsequent to the pilot test, a full-scale remediation program was conducted that used six nested sparge points. The sparging devices were installed to a total depth of 40 feet bgs, about 8-10 feet below the top of the groundwater table. Four wells were used as monitoring points during the pilot test and five wells during the full scale program to observe ROI and water quality changes.

Groundwater samples were collected to evaluate changes in water quality. Samples were taken twice a week and analyzed for pH, temperature, conductivity, oxidation-reduction potential (ORP), and dissolved oxygen (DO), as well as the CoCs at the site. During the pilot and full scale tests ozone was injected into each sparging well at a 4% concentration at a rate of 0.25 lbs/well/day. During the pilot trial, hydrogen peroxide at a concentration of 7.75% was injected into each well at a rate of 1.2 lbs/well/day. For the full-scale operation, a 10% solution was injected at a rate of 0.5 lbs/well/day. Compressed air was used to assist in moving the oxidants away from the injection well.

The pilot test was operated for a period of 22 days. Results from the pilot test confirmed a substantial reduction in target contaminants at the monitoring points, although reduction in TBA concentrations was lower than that of the other target chemicals. The pilot test also confirmed an active ROI from the oxidant injection of 12 ft. Based upon these results, the full scale program was implemented.

The full scale program operated for a period of 3 months. The oxidant injection wells were placed around the perimeter of the gas station and downgradient of the source area. After three months of operation, concentrations in all monitoring and extraction wells onsite were below laboratory detection limits for TPH, BTEX, and MTBE. Also, the concentration of TBA was reduced by 90 to 99.95 % in the monitoring wells, suggesting additional treatment required to fully remove this chemical. Other important monitoring results included a slight pH increase (0.25 units) over the operating period, a slight increase in temperature of two degrees, an increase in DO, and an increase in ORP.

For this site, a high amount of chemical oxygen demand (COD) was present, which was compensated for by using a relatively high oxidant dose rate to achieve target chemical destruction. The ROI achieved during full scale operation was at least 20 feet.

9 Summary and Recommendations

Fenton’s reagent, permanganate, persulfate, ozone and ozone with peroxide are oxidants that are commonly used for treatment of organic contaminants. Certain ISCO technologies have been demonstrated to be a successful approach for remediation of NAPL source areas at numerous sites including the two case studies reported in this document. It should be recognized, however, that ISCO only treats dissolved-phase contamination. Therefore, the rate of treatment of NAPL source areas by ISCO is limited by mass transfer (dissolution) of NAPL constituents from the nonaqueous to the aqueous phase.

To date, research reported in the literature has reported oxidation of organophosphorus pesticides with Fenton’s reagent, permanganate, and ozone. Specific conclusions from this literature review for each of the oxidants reviewed includes:

Fenton’s reagent is known to oxidize some organophosphorus pesticides (e.g., methyl parathion, malathion), and should be tested with Site groundwater to confirm treatability under Site conditions;
Ozone has been demonstrated to oxidize some organophosphorus pesticides (e.g., parathion), and should be tested with Site groundwater to confirm treatability under Site conditions. The low pH conditions at the Site are favorable toward maximizing the reactive half-life of ozone in groundwater;
The effectiveness of AOP (ozone + hydrogen peroxide) for treating methyl parathion has been demonstrated, and the combination of ozone + hydrogen peroxide is reported to yield more hydroxyl radicals than ozone alone. As such, ozone + hydrogen peroxide should be tested to confirm treatability of other Site organophosphorus pesticides under Site conditions;
The literature is inconclusive regarding the treatability of organophosphorus pesticides by permanganate. Nevertheless, permanganate remains an attractive option for ISCO, given its success with other applications in Denmark and its relative ease of deployment; and
No literature could be identified regarding treatment of organophosphorus pesticides by persulfate. In any case, the requirement for heat activation makes persulfate a relatively less attractive option for ISCO at the Site.

Based upon these recommendations, the bench treatability tests will be performed using Fenton’s reagent and ozone (or ozone + peroxide) treatments. Testing with permanganate should also be performed. The purpose of these tests will be to confirm the efficiency and extent of treatment of the principal contaminants at the Høfde 42 site by these chemical oxidants. The results from these treatability tests will be used to evaluate which oxidants would be appropriate for pilot scale testing at the Site.

10 References

References listed below include all the references that are cited in preceding sections, as well as references for additional publications that are relevant to ISCO technologies and treatment of organophosphorus pesticides by various chemical oxidants. All the references cited below were reviewed as part of the preparation of this report.

Adam, M. L., Comfort, S. D., Morley, M. C., and Snow, D. D. (2004). Remediating RDX-contaminated ground water with permanganate: Laboratory investigations for the Pantex perched aquifer. Journal of Environmental Quality 33, 2165-2173.

Baciocchi, R., Boni, M. R., and D'Aprile, L. (2003). Hydrogen peroxide lifetime as an indicator of the efficiency of 3-chlorophenol Fenton's and Fenton-like oxidation in soils. Journal of Hazardous Materials B96, 305-329.

Barberis, K., and Howarth, C. R. (1991). Reactivity studies in the ozonolysis of pollutants using a perforated spinning disk reactor to enhance mass-transfer. Ozone-Science & Engineering 13, 501-519.

Beltran, F. J., Gomezserrano, V., and Duran, A. (1992). Degradation kinetics of para-nitrophenol ozonation in Water. Water Research 26, 9-17.

Beltran, F. J., Rivas, F. J., and Gimeno, O. (2005). Comparison between photocatalytic ozonation and other oxidation processes for the removal of phenols from water. Journal of Chemical Technology and Biotechnology 80, 973-984.

Benitez, F. J., Beltranheredia, J., and Gonzalez, T. (1991). Absorption kinetics of ozone in aqueous-solutions of malathion. Ozone-Science & Engineering 13, 487-499.

Benitez, F. J., Beltranheredia, J., and Gonzalez, T. (1992). Ozonation of pesticides - variables effect on degradation - stoichiometric ratio. Anales De Quimica 88, 548-555.

Brackin, J., Hellmann, G., Gustafson, D., 2005. Aggressive remediation of adsorbed and dissolved gasoline, MTBE, and TBA contaminants using an innovative in-situ advanced chemical oxidation system. Proceedings of the 2005 NGWA Conference on MTBE and Perchlorate: Assessment, Remediation, and Public Policy.

Brown, G. S., Barton, L. L., and Thomson, B. M. (2003). Permanganate oxidation of sorbed polycyclic aromatic hydrocarbons. Waste Management 23, 737-740.

Budavari, S. et. al., 1996, The Merck Index, 12th Ed., Merck Research Laboratories.

Buxton, G. V., Greenstock, C. L., Helman, W. P., and Ross, A. B. (1988). Critical review of rate constants for reactions of hydrated electrons, hydrogen atoms and hydroxyl radicals (^.OH/^.O-) in aqueous solution. J. Phys. Chem. Ref. Data 17, 513-886.

Chapelle, F. H., Bradley, P. M., and Casey, C. C. (2005). Behavior of chlorinated ethene plume following source-area treatment with Fenton's reagent. Groundwater Monitoring and Remediation 25, 131-141.

Chemfinder.com, 2004. Chemical Database. Downloaded from: http://chemfinder.cambridgesoft.com/, 21 March 2006.

Chirchi, L., and Ghorbel, A. (2002). Use of various Fe-modified montmorillonite samples for 4-nitrophenol degradation by H₂O₂. Applied Clay Science 21, 271-276.

Conrad, S. H., Glass, R. J., and Peplinski, W. J. (2002). Bench-scale visualization of DNAPL remediation processes in analog heterogeneous aquifers: surfactant floods and in situ oxidation using permanganate. Journal of Contaminant Hydrology 58, 13-49.

Crimi, M. L., and Siegrist, R. L. (2003). Geochemical effects on metals following permanganate oxidation of DNAPLs. Ground Water 41, 458-469.

Damm, J. H., Hardacre, C., Kalin, R. M., and Walsh, K. P. (2002). Kinetics of the oxidation of methyl tert-butyl ether (MTBE) by potassium permanganate. Water Research 36, 3638-3646.

Dai, Q., and Reitsma, S. 2002. Two-dimensional experimental studies of permanganate flushing of pooled DNAPL. Third International Conference on Remediation of Chlorinated and Recalcitrant Compounds; Monterey, CA; USA; 20-23 May 2003.

Doong, R. A., and Chang, W. H. (1998a). Photodegradation of parathion in aqueous titanium dioxide and zero valent iron solutions in the presence of hydrogen peroxide. Journal of Photochemistry and Photobiology a-Chemistry 116, 221-228.

Doong, R.A., and Chang, W.H. (1998b). Photoassisted iron compound catalytic degradation of organophosphorous pesticides with hydrogen peroxide. Chemosphere 37, 2563-2572.

Environmental Security Technology Certification Program. (1999). In Situ Chemical Oxidation, Technology Status. http://www.estcp.org/documents/techdocs/ISO_Report.pdf

Ferguson, S. H., Woinarski, A. Z., Snape, I., Morris, C. E., and Revill, A. T. (2004). A field trial of in situ chemical oxidation to remediate long-term diesel contaminated Antarctic soil. Cold Regions Science and Technology 40, 47-60.

Fjordbøge, A. (2005). Evaluering af nul-valent jern til oprensning af Høfde 42 Depotet. Institut for Miljø & Ressourcer, Danmarks Tekniske Universitet. November.

Gates-Anderson, D. D., Siegrist, R. L., and Cline, S. R. (2001). Comparison of potassium permanganate and hydrogen peroxide as chemical oxidants for organically contaminated soils. Journal of Environmental Engineering-ASCE 127, 337-347.

Gimeno, O., Carbajo, M., Beltran, F. J., and Rivas, F. J. (2005). Phenol and substituted phenols AOPs remediation. Journal of Hazardous Materials 119, 99-108.

Goi, A., and Trapido, M. (2002). Hydrogen peroxide photolysis, Fenton reagent and photo-Fenton for the degradation of nitrophenols: a comparative study. Chemosphere 46, 913-922.

Goi, A., and Trapido, M. (2004). Combined methods for remediation of nitrophenol-contaminated soil. Fresenius Environmental Bulletin 13, 1191-1196.

Goi, A., and Trapido, M. (2004). Degradation of polycyclic aromatic hydrocarbons in soil: The Fenton reagent versus ozonation. Environmental Technology 25, 155-164.

Goi, A., Trapido, M., and Tuhkanen, T. (2004). A study of toxicity, biodegradability, and some by-products of ozonised nitrophenols. Advances in Environmental Research 8, 303-311.

Gunther, F. A., Ott, D. E., and Ittig, M. (1970). Oxidation of parathion to paraoxon. II. By use of ozone. Bulletin of Environmental Contamination and Toxicology 5, 87-94.

Haapea, P., and Tuhkanen, T. (2005). Aged chlorophenol contaminated soil's integrated treatment by ozonation, soil washing and biological methods. Environmental Technology 26, 811-819.

Hafiz, A. A. (2005). Metallosurfactants of Cu(II) and Fe(III) complexes as catalysts for the destruction of paraoxon. Journal of Surfactants and Detergents 8, 359-363.

House, H. O. (1972). Modern Synthetic Reactions. 2nd/Ed. W.A. Benjamin, Inc., Menlo Park, CA.

Hsu, M. I. Y., and Masten, S. J. (1997). The kinetics of the reaction of ozone with phenanthrene in unsaturated soils. Environmental Engineering Science 14, 207-218.

Huang, K. C., Couttenye, R. A., and Hoag, G. E. (2002). Kinetics of heat-assisted persulfate oxidation of methyl tert-butyl ether (MTBE). Chemosphere 49, 413-420.

Huang, K. C., Zhao, Z. Q., Hoag, G. E., Dahmani, A., and Block, P. A. (2005). Degradation of volatile organic compounds with thermally activated persulfate oxidation. Chemosphere 61, 551-560.

Huston, P. L., and Pignatello, J. J. (1999). Degradation of selected pesticide active ingredients and commercial formulations in water by the photo-assisted Fenton reaction. Water Research 33, 1238-1246.

Interstate Technology and Regulatory Council. (2005). Technical and Regulatory Guidance for In Situ Chemical Oxidation of Contaminated Soil and Groundwater. http://www.itrcweb.org/Documents/ISCO-2.pdf.

Jung, H., Ahn, Y., Choi, H., and Kim, I. S. (2005). Effects of in-situ ozonation on indigenous microorganisms in diesel contaminated soil: Survival and regrowth. Chemosphere 61, 923-932.

Jung, H., Kim, J., and Choi, H. (2004). Reaction kinetics of ozone in variably saturated porous media. Journal of Environmental Engineering-ASCE 130, 432-441.

Kakarla, P. K. C. and Watts, R.J. (1997). Depth of Fenton-like oxidation in remediation of surface soil. Journal of Environmental Engineering-ASCE 123, 11-17, 1997.

Kang, N., Hua, I., and Rao, P. S. C. (2004). Production and characterization of encapsulated potassium permanganate for sustained release as an in situ oxidant. Industrial & Engineering Chemistry Research 43, 5187-5193.

Kavitha, V., and Palanivelu, K. (2005). Degradation of nitrophenols by Fenton and photo-Fenton processes. Journal of Photochemistry and Photobiology A: Chemistry 170, 83-95.

Khan, E., Babcock, R. W., Hsu, T. M., and Lin, H. (2005). Mineralization and biodegradability enhancement of low level p-nitrophenol in water using Fenton's reagent. Journal of Environmental Engineering-ASCE 131, 327-331.

Kiwi, J., Pulgarin, C., and Peringer, P. (1994). Effect of Fenton and photo-Fenton reactions on the degradation and biodegradability of 2-nitrophenols and 4-nitrophenols in water-treatment. Applied Catalysis B-Environmental 3, 335-350.

Ku, Y., Chang, J. L., and Shen, Y. S. (1998). Decomposition of ethylparathion in aqueous solution by ozonation. Water Air and Soil Pollution 103, 341-355.

Laplanche, A, Martin, G, Tonnard, F. (1984). Ozonation Schemes of Organo Phosphorus Pesticides Application in Drinking Water Treatment
Ozone: Science and Engineering 6, 207-219.

Lee, E. S., Seol, Y., Fang, Y. C., and Schwartz, F. W. (2003). Destruction efficiencies and dynamics of reaction fronts associated with the permanganate oxidation of trichloroethylene. Environmental Science & Technology 37, 2540-2546.

Li, X.D., and Schwartz F.W. (2000). Efficiency problems related to permanganate oxidation schemes. Second International Conference on Remediation of Chlorinated and Recalcitrant Compounds; Monterey, CA; USA; 22-25 May. pp. 41-48.

Li, Z. H. (2004). Surfactant-enhanced oxidation of trichloroethylene by permanganate - proof concept. Chemosphere 54, 419-423.

Liang, C. J., Bruell, C. J., Marley, M. C., and Sperry, K. L. (2003). Thermally activated persulfate oxidation of trichloroethylene (TCE) and 1,1,1-trichloroethane (TCA) in aqueous systems and soil slurries. Soil & Sediment Contamination 12, 207-228.

Liang, C. J., Bruell, C. J., Marley, M. C., and Sperry, K. L. (2004a). Persulfate oxidation for in situ remediation of TCE. I. Activated by ferrous ion with and without a persulfate-thiosulfate redox couple. Chemosphere 55, 1213-1223.

Liang, C. J., Bruell, C. J., Marley, M. C., and Sperry, K. L. (2004b). Persulfate oxidation for in situ remediation of TCE. II. Activated by chelated ferrous ion. Chemosphere 55, 1225-1233.

Lim, H. N., Choi, H., Hwang, T. M., and Kang, J. W. (2002). Characterization of ozone decomposition in a soil slurry: kinetics and mechanism. Water Research 36, 219-229.

Lipczynska-Kochany, E. (1992). Degradation of nitrobenzene and nitrophenols by means of advanced oxidation processes in a homogeneous phase: Photolysis in the presence of hydrogen peroxide versus the Fenton reaction. Chemosphere 24, 1369-1380.

Looney, B.B, and Falta, R.W. (2000). Vadose Zone Science and Technology Solutions. Battelle Press, Columbus OH.

Lowe, K. S., Gardner, F. G., and Siegrist, R. L. (2002). Field evaluation of in situ chemical oxidation through vertical well-to-well recirculation of NaMnO4. Ground Water Monitoring and Remediation 22, 106-115.

MacKinnon, L. K., and Thomson, N. R. (2002). Laboratory-scale in situ chemical oxidation of a perchloroethylene pool using permanganate. Journal of Contaminant Hydrology 56, 49-74.

Masten, S. J., and Davies, S. H. R. (1997). Efficacy of in-situ ozonation for the remediation of PAH contaminated soils. Journal of Contaminant Hydrology 28, 327-335.

Meijers, R. T., Oderwald-Muller, E. J., Nuhn, P., and Kruithof, J. C. (1995). Degradation of pesticides by ozonation and advanced oxidation. Ozone-Science & Engineering 17, 673-686.

Montgomery, (2000). Groundwater Chemicals: Desk Reference, 3^rd Edition. Lewis Publishers, 2000.

Miller, C. M., Valentine, R. L., Roehl, M. E., and Alvarez, P. J. J. (1996). Chemical and microbiological assessment of pendimethalin-contaminated soil after treatment with Fenton's reagent. Water Research 30, 2579-2586.

Moss, R. A., and Ragunathan, K. G. (1998). Remarkable acceleration of dimethyl phosphate hydrolysis by ceric cations. Chemical Communications, 1871-1872.

Mumford, K. G., Thomson, N. R., and Allen-King, R. M. (2005). Bench-scale investigation of permanganate natural oxidant demand kinetics. Environmental Science & Technology 39, 2835-2840.

Nam, K., and Kukor, J. J. (2000). Combined ozonation and biodegradation for remediation of mixtures of polycyclic aromatic hydrocarbons in soil. Biodegradation 11, 1-9.

Nam, K., Rodriguez, W., and Kukor, J. (2001). Enhanced degradation of polycyclic aromatic hydrocarbons by biodegradation combined with a modified Fenton reaction. Chemosphere 45, 11-20.

Nikolaki, M. D., Oreopoulou, A. G., and Philippopoulos, C. J. (2005). Photo-Fenton assisted reaction of dimethoate in aqueous solutions. Journal of Environmental Science and Health, Part B, 233 - 246.

Pignatello, J. J., and Baehr, K. (1994). Ferric complexes as catalysts for "Fenton" degradation of 2,4-D and metolachlor in soil. Journal of Environmental Quality 23, 365-370.

Pignatello, J. J., and Day, M. (1996). Mineralization of methyl parathion insecticide in soil by hydrogen peroxide activated with iron(III)-NTA or -HEIDA complexes. Haz. Waste. Haz. Mater. 13, 237-244.

Pignatello, J. J., Oliveros, E., and MacKay, A. (2006). Advanced oxidation processes for organic contaminant destruction based on the Fenton reaction and related chemistry. Critical Reviews in Environmental Science and Technology 36, 1-84.

Pignatello, J. J., and Sun, Y. (1995). Complete oxidation of metolachlor and methyl parathion in water by the photoassisted Fenton reaction. Water Research 29, 1837-1844.

Pugliese, P., Molto, J. C., Damiani, P., Marin, R., Cossignani, L., and Manes, J. (2004). Gas chromatographic evaluation of pesticide residue contents in nectarines after non-toxic washing treatments. Journal of Chromatography A. 1050, 185-191.

Radhakrishnamurti, P. S., and Sahu, S. N. (1976). Kinetics and mechanism of oxidation of nitrophenols and nitroanilines by potassium-permanganate. Journal of the Indian Chemical Society 53, 1154-1155.

Schwarzenbach, R.P., Gschwend, P.M., and Imboden D.M. (1993). Environmental Organic Chemistry. John Wiley & Sons, Inc. New York.

Seol, Y., and Schwartz, F. W. (2000). Phase-transfer catalysis applied to the oxidation of nonaqueous phase trichloroethylene by potassium permanganate. Journal of Contaminant Hydrology 44, 185-201.

Shi, H. X., Xu, X. W., Xu, X. H., Wang, D. H., and Wang, Q. D. (2005). Mechanistic study of ozonation of p-nitrophenol in aqueous solution. Journal of Environmental Sciences-China 17, 926-929.

Shin, W. T., Garanzuay, X., Yiacoumi, S., Tsouris, C., Gu, B. H., and Mahinthakumar, G. K. (2004). Kinetics of soil ozonation: an experimental and numerical investigation. Journal of Contaminant Hydrology 72, 227-243.

Siegrist, R. L., Urynowicz, M. A., Crimi, M. L., and Lowe, K. S. (2002). Genesis and effects of particles produced during in situ chemical oxidation using permanganate. Journal of Environmental Engineering-ASCE 128, 1068-1079.

Spear, R. C., Lee, Y. S., Leffingwell, J. T., and Jenkins, D. (1978). Conversion of parathion to paraoxon in foliar residues - effects of dust level and ozone concentration. Journal of Agricultural and Food Chemistry 26, 434-436.

Spencer, W. F., Adams, J. D., Shoup, T. D., and Spear, R. C. (1980). Conversion of parathion to paraoxon on soil dusts and clay-minerals as affected by ozone and UV Light. Journal of Agricultural and Food Chemistry 28, 366-371.

Staehelin, J., and Hoigne, J. (1985). Decomposition of ozone in water in the presence of organic solutes acting as promoters and inhibitors of radical chain reactions. Environmental Science & Technology 19, 1206-1213.

Struse, A. M., Siegrist, R. L., Dawson, H. E., and Urynowicz, M. A. (2002). Diffusive transport of permanganate during in situ oxidation. Journal of Environmental Engineering-ASCE 128, 327-334.

Sung, M. H., and Huang, C. P. (2002). In situ removal of 2-chlorophenol from unsaturated soils by ozonation. Environmental Science & Technology 36, 2911-2918.

U.S. Department of Energy. (1999a). Fenton’s Reagent. DOE/EM-0484. Office of Environmental Management, Office of Science and Technology. http://apps.em.doe.gov/OST/pubs/itsrs/itsr2161.pdf

U.S. Department of Energy. (1999b). In Situ Chemical Oxidation Using Potassium Permanganate. DOE/EM-0496. Office of Environmental Management, Office of Science and Technology. http://apps.em.doe.gov/OST/pubs/itsrs/itsr167.pdf

U.S. Environmental Protection Agency. (1998). Field Applications of In Situ Remediation Technologies: Chemical Oxidation. Office of Solid Waste and Emergency Response. EPA 542-R-98-008. http://www.clu-in.org/download/remed/chemox.pdf

U.S. Environmental Protection Agency. (2004). AOP treats dioxane-contaminated groundwater. Technology News and Trends, Issue 11. March.

Trapido, M., and Kallas, J. (2000). Advanced oxidation processes for the degradation and detoxification of 4-nitrophenol. Environmental Technology 21, 799-808.

Wang, Y. X., Liang, Z., Yuan, X. M., and Xu, Y. S. (2005). Preparation of cellular iron using wastes and its application in dyeing wastewater treatment. Journal of Porous Materials 12, 225-232.

Watts, R. J., Bottenberg, B. C., Hess, T. F., Jensen, M. D., & Teel, A. L. (1999). Role of reductants in the enhanced desorption and transformation of chloroaliphatic compounds by modified Fenton’s reactions. Environmental Science & Technology 33, 3432-3437.

Watts et al. (1994). On Site Treatment of Contaminated Soils Using Hydrogen Peroxide. Project Report T9234-06, Washington State Transport Center, Washington State University.

Yin, Y, and H.E. Allen. (1999). In Situ Chemical Treatment. Ground-Water Remediation Technologies Analysis Center. Technology Evaluation Report No. TE-99-01. http://www.groundwatercentral.info/org/pdf/E_inchem.pdf.

Yu, C. P., and Yu, Y. H. (2000). Identifying useful real-time control parameters in ozonation process. Water Science and Technology 42, 435-440.

Yu, C. P., and Yu, Y. H. (2001). Mechanisms of the reaction of ozone with p-nitrophenol. Ozone-Science & Engineering 23, 303-312.

Yu, D. Y., Bae, W., Kang, N. G., Banks, M. K., and Choi, C. H. (2005). Characterization of gaseous ozone decomposition in soil. Soil & Sediment Contamination 14, 231-247.

Zhai, X. H., Hua, I., Rao, P. S. C., and Lee, L. S. (2006). Cosolvent-enhanced chemical oxidation of perchloroethylene by potassium permanganate. Journal of Contaminant Hydrology 82, 61-74.

Zhang, H., Ji, L., Wu, F., and Tan, J. (2005). In situ ozonation of anthracene in unsaturated porous media. Journal of Hazardous Materials 120, 143-148.

Tables

TABLE 1 SUMMARY OF ORGANOPHOSPHORUS PESTICIDE PRESENCE AT THE HØFDE 42 SITE

(Modified from Ringkjøbing County, [2004])

Chemical	Total Aqueous and Suspended Colloid Concentration in “Hot Spot” Area (mg/L)	Maximum Ground- water Concentra- tion (mg/L)	Typical DNAPL Composition
Parathion	190	20	57%
Methyl parathion	450	1.6	16%
Malathion	18	0.5	1.5%
Sulfotepp	140	2	3%
Amino-parathion	28000	5	Not listed

TABLE 2 SUMMARY OF KEY PROPERTIES OF ORGANOPHOSPHORUS PESTICIDES

Click here to see Table 2

TABLE 3. REDOX POTENTIALS OF COMMON CHEMICAL OXIDANTS

Oxidant	Formula	E° (mV)
Fenton’s reagent	H₂O₂+ Fe²⁺	2.76
Activated persulfate (via heat or metals)	S₂O₈	2.6
Ozone	O₃	2.07
Persulfate	S₂O₈	2.01
Hydrogen peroxide	H₂O₂	1.76
Permanganate	MnO₄	1.70

TABLE 4. OXIDANT PROPERTIES TO BE CONSIDERED FOR IN SITU TREATMENT WITH ISCO (modified from ITRC, 2005)

	Fenton’s reagent	Ozone	Permanganate	Persulfate
Vadose zone	Successful	Successful	Successful	Successful
treatment
Saturated zone	Successful	Site specific	Successful	Successful
Treatment		Not well transported in groundwater
Potential detrimental effects	Gas evolution, heat generation, by-products, resolubilization of metals	Gas evolution, by-products, resolubilization of metals	By-products, resolubilization of metals	By-products, resolubilization of metals
pH/alkalinity	Effective over a wide pH range, but carbonate alkalinity must be taken into consideration. Addition of chelating agents can overcome pH/alkalinity issues	Effective over a wide pH range, but carbonate alkalinity must be taken into consideration	Effective over a wide range	Effective over a wide pH range, but carbonate alkalinity must be taken into consideration
Persistence	Rapidly degraded in contact with soil/groundwater . Generally unstable with short half-life	Easily degraded in contact with soil/groundwater	The oxidant is very stable	The oxidant is very stable
Oxidant demand	Soil oxidant demand varies with soil type; contaminant oxidant demand is based on total mass and mass distribution (sorbed, dissolved and free phase)
Soil permeability and heterogeneity	Low-permeable soils and subsurface heterogeneity offer a challenge for the distribution of injected or extracted fluids

TABLE 5. SUMMARY OF CHEMINOVA CONTAMINANTS DEGRADED BY THE FENTON REACTION

Compound	Conditions	Rate data	Product (yield)	Reference
Methyl parathion	aqueous, light (1 × 10¹⁸ photons/L/s blacklamp UV), 0.1 mM MP, 1 mM Fe³⁺, 10 mM H₂O₂, pH 2.8	complete loss of MP in 5 min	SO₄²^- (100%); NO₃^- (100%); PO₄³^- (100%); transient products: 4-nitrophenol, methyl paraoxon (trace), dimethyl phosphate, oxalate	Pignatello and Sun, 1995
Methyl parathion	soil slurry, dark, MP = 7.6 mmol/kg; H₂O₂ = 0-6 mol/kg; Fe = 0-0.1 mol/kg; Fe as NTA or HEIDA chelate; initial pH ~6; temp, 10-60°C	up to 88% loss of MP depending on reagent concns. and temp.	using 0.01 mol/kg Fe-NTA, 1-2 mole/kg H₂O₂, and based on MP degraded in 3 h: SO₄²^- (106-112%); NO₃^- (100-107%); dimethyl phosphate (22-36%); p-nitrophenol (<5%)	Pignatello and Day, 1996
malathion	aqueous, dark, 50 mM Fe²⁺, 0.6 mM H₂O₂, pH 7	20% loss mal in 24 h	n.d.	Doong and Chang, 1998
malathion	aqueous, light (230 mW/cm² medium pressure Hg lamp), 10 mg/L mal, 0.6 mM H₂O₂, pH 7 a) 50 mM Fe²⁺ b) 1 g/L iron powder	both cases: 100% loss mal in 150 min, k_obs = 0.012 min^-1	n.d.	Doong and Chang, 1998
malathion	aqueous, light (1.2 × 10¹⁹ photons/L/s blacklamp UV), 0.2 mM mal, 0.5 mM Fe³⁺, 10 mM H₂O₂	94% loss mal in 0.5 h	SO₄²^- (115%); PO₄³^- (35%); formate; oxalate; acetate	Huston and Pignatello, 1999

TABLE 6. OZONE REACTION WITH CHEMINOVA CONTAMINANTS WHERE PRODUCTS HAVE BEEN IDENTIFIED

Compound	Conditions	Rate information	Product (yield)	Reference
parathion	ethanolic solution (5 min at 40 mL/min ozone)	n.d.	paraoxon (30%-“quantitative”); sulfate (34-38%)	Gunther et al., 1970
parathion	water, pH 3-9	pseudo 1^st order k = ~0.17 min^-1	nitrate (stoich.); sulfate (stoich.); phosphate; carbonate;	Ku et al., 1998
parathion	adsorbed on soil dust or clay mineral, exposed to 30 or 300 ppbv ozone, 30% relative humidity; w/wo UV	rate followed up to 140 h	paraoxon (0.9-8.5%)	Spencer et al., 1980
methyl parathion	pH 7.2-8.3; 5-20°C; O₃/DOC = 0.53-0.95	64-89 (increases with pH, T and dose)	products not identified	Meijers et al., 1995

TABLE 7. SUMMARY OF CHEMICAL OXIDATION CASE STUDIES

Parameter	Savannah River Site	Active Retail Gas Station
Technology description	Fenton’s Reagent	Ozone and Peroxide
Soil type	permeable sands with low fines alternating with clayey sand and clay units	silt and sand layer (15 ft thick), fine and course grained sands (to 35 ft depth), weathered bedrock (to 40 ft depth)
Depth to groundwater	130 ft bgs	33 ft bgs
Contaminants of concern	TCE and PCE - DNAPL composition 95% TCE and 5 % PCE	TPH (5 – 35 mg/L), MTBE (5 to 200 mg/L), TBA (1 to 20 mg/L), and BTEX (0.5 to 5 mg/L)
NAPL evidence	Observed DNAPL in bottom of wells; groundwater concentrations of 120 mg/l PCE and 21 mg/l TCE	Free product (light NAPL) observed prior to a previous MPE demonstration at the site
Volume treated	68,702 ft³	not provided
Area treated	50 × 50 ft²	not provided, each sparge point had an estimated ROI of >20 ft
Depth treated	124 to 152 ft bgs	~33 to 40 ft bgs
NAPL mass targeted	600 lbs	not provided
Remediation infrastructure	4 injection wells; 3 groundwater monitoring wells; 3 vadose zone monitors; proprietary injection process	2 injection wells during pilot test, 6 during full scale; 4 monitoring wells during pilot test, 5 during full scale above ground ozone generator and amendment dosing equipment including specialized well head construction for amendment addition
Remediation duration	6 day period of injection	Pilot: approximately 22 days Full scale; 3 months reported, activities ongoing
Remedial costs	$511,115 site preparation and operation activities, drilling, construction, operations, sampling, pre- and post- demonstration characterization, demobilization and reporting and project management.	not provided
Performance results:	94% destruction of total VOCs, 95% PCE and 88% TCE	After 3 months of full-scale operation: TPH, MTBE, BTEX 100% reduction TBA reduced by 90 to 99.95%

Figures

Figure 1. Typical in-situ chemical oxidation application

Figure 1. Typical in-situ chemical oxidation application

Figure 2. Reactions of ozone in water in the presence of reactive solutes (Staehelin and Hoigné, 1985). M represents a reactive solute, such as an organic compound.

Figure 2. Reactions of ozone in water in the presence of reactive solutes (Staehelin and Hoigné, 1985). M represents a reactive solute, such as an organic compound.

Bilag 2 Litteraturstudie

Literature Review

In Situ Chemical Oxidation of Organophosphorous Pesticides in Groundwater

Table of Contents

1 Introduction

1.1 Site Conditions

1.2 Outline of this Document

2 Key Properties of Organophosphorus Pesticides

3 Overview of Oxidants for In Situ Treatment

3.1 Chemistry of Oxidants

3.1.1 Fenton’s Reagent

3.1.2 Permanganate

3.1.3 Persulfate

3.1.4 Ozone/Ozone with Peroxide

3.2 Advantages and Disadvantages of Each Oxidant

3.2.1 Fenton’s Reagent

Advantages:

Disadvantages:

3.2.2 Permanganate

Advantages:

Disadvantages:

3.2.3 Persulfate

Advantages:

Disadvantages:

3.2.4 Ozone/Ozone with Peroxide

Advantages:

Disadvantages:

3.3 General Considerations for In Situ Treatment

3.3.1 Oxidant Properties

3.3.2 Geology

3.3.3 Plume versus Source Areas

3.3.4 Application/Installation Methods

3.4 Full Scale Design, Operations and Maintenance, Safety and Cost Considerations

3.4.1 Data Requirements for Design

3.4.2 O&M Requirements

3.4.3 Safety

3.4.4 Cost

4 Fenton’s Reagent Reactions with Organophosphorus Pesticides

4.1 Reaction Chemistry with Organophosphorus Pesticides and other Contaminants at the Cheminova Site

4.2 Reaction Chemistry with Other Organophosphorus Pesticides

4.3 Summary of Implications for Treatability Test

5 Permanganate Reactions with Organophosphorus Pesticides

5.1 Reaction Chemistry with Organophosphorus Pesticides and other Contaminants at the Cheminova Site

5.2 Summary of Implications for Treatability Test

6 Persulfate Reactions with Organophosphorus Pesticides

6.1 Reaction Chemistry with Organophosphorus Pesticides and other Contaminants at the Cheminova Site

6.2 Summary of Implications for Treatability Test

7 Ozone and Ozone+Hydrogen Peroxide Reactions with Organophosphorus Pesticides

7.1 Reaction Chemistry with Cheminova Site Organophosphorus Pesticides and other Contaminants

7.2 Summary of Implications for Treatability Test

8 Case Studies of ISCO Field Applications

8.1 Case Study #1: Fenton’s Reagent Demonstration, Savannah River Site Aiken, South Carolina

8.2 Case Study #2: Ozone with Peroxide, Active Retail Gas Station, Riverside, California

9 Summary and Recommendations

10 References

Figures