JOURNAL OF

Contaminant Hydrology

E L S E V I E R Journal of Contaminant Hydrology 28 (1997) 327-335

Efficacy of in-situ ozonation for the remediation of PAH contaminated soils

Susan J. Masten *, Simon H.R. Davies Department of Cit, il and Em, ironmental Engineering, Michigan State Unit,ersi O' East Lansing,

M148823, USA

Abstract

Polycyclic aromatic hydrocarbons (PAHs) are of environmental concern because many PAHs are either carcinogens or potential carcinogens. Petroleum products are a major source of PAHs. The occurrence of PAH contamination is widespread and novel treatment technologies for the remediation of contaminated soils are necessary.

Ozone has been found to be extremely useful for the degradation of PAHs in soils. For these compounds, the reaction with molecular ozone appears to be the more important degradation pathway. Greater than 95% removal of phenanthrene was achieved with an ozonation time of 2.3 h at an ozone flux of 250 mg h - 1. After 4.0 h of treatment at an ozone flux of 600 mg h 1, 91% of the pyrene was removed. We have also found that the more hydrophobic PAHs (e.g. chrysene) react more slowly than would be expected on the basis of their reactivity with ozone, suggesting that partitioning of the contaminant into soil organic matter may reduce the reactivity of the compound. Even so, after 4 h of exposure to ozone, the chrysene concentration in a contaminated Metea soil was reduced from 100 to 5 0 m g k g - t .

Ozone has been found to be readily transported through columns packed with a number of geological materials, including Ottawa sand, Metea soil, Borden aquifer material and Wurtsmith aquifer material. All of these geological materials exerted a limited (finite) ozone demand, i.e. the rate of ozone degradation in soil columns is very slow after the ozone demand is met. Moisture content was found to increase the ozone demand, most likely owing to the dissolution of gaseous ozone into the pore water. As once the initial ozone demand is met, little degradation of ozone is observed, it should be possible to achieve ozone penetration to a considerable distance away from the injection well, suggesting that in-situ ozonation is a feasible means of treating uncontaminated

* Corresponding author. Tel.: + 1 517 353-8539: fax: + 1 517 355-0250; e-mail: masten@cee.msu.edu

0169-7722/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PII S0 1 69 -7722 (97 ) 0001 9-3

328 S.J. Masten, S.H.R. Dacies / Journal of Contaminant Hydrology 28 (1997) 327-335

unsaturated soils. This is substantiated by two field studies where in-situ ozonation was apparently successful at remediating the sites. © 1997 Elsevier Science B.V.

Keywords: Ozonation; PAHs; In-situ soil remediation; Phenanthrene: Pyrene: Chrysene

I. Introduction

Polycyclic aromatic hydrocarbons (PAHs) are of environmental concern because many of these compounds are either carcinogens or potential carcinogens. The occur- rence of PAH contamination in soils is widespread, as gasoline or fuel oil spills are common; in fact it is estimated that 90-95% of all leaking underground storage tanks contain petroleum products (US EPA, 1995). Many thousands of sites exist; Miller (1994) has estimated that the Air Force alone needs to remediate more than 2000 petroleum-contaminated sites. Where a fuel spill has occurred, soils in the vadose zone will be contaminated. Bioventing and conventional soil vapor extraction (soil venting) techniques have been used to remediate petroleum contamination, but the removal efficiencies for PAHs are typically very low (R. Miller, personal communication, 1994). Since most of the PAHs are semi- or non-volatile, conventional soil venting is relatively ineffective. The higher molecular weight PAHs are resistant to bioremediation and bioremediation studies suggest that microorganisms can only degrade PAHs that are dissolved in the aqueous-phase and that PAHs sorbed onto or partitioned into a solid phase are not readily degraded by microorganisms (Luthy et al., 1994). Thus, prospects for the in-situ bioremediation of PAHs in soils are poor, unless means are found to solubilize these contaminants.

The use of solvents or surfactants to enhance the solubility of the hydrophobic components found in petroleum products has been suggested. However, high concentra- tions of the solvent or surfactant are generally required to achieve good results (Luthy et al., 1994). Also, the slow dissolution kinetics necessitate long contact-times and effective procedures to treat the extracted PAHs are often difficult to develop (Luthy et al., 1994). In the field, the heterogeneity of the deposits may cause the solvent (or surfactant solution) to bypass some of the contaminated zones. Thus, in the field, dissolution rates may be much slower than that observed in laboratory column or batch studies (Luthy et al., 1994).

2. In-situ ozonation

In-situ chemical oxidation processes may be used to overcome the limitations imposed by the low aqueous solubility of PAHs on their rates of biodegradation and/or dissolution. Chemical oxidation of these compounds is likely to enhance both their biodegradability and aqueous solubility, since the oxidation products would be more polar than the parent compounds (Gilbert, 1987). An advantage of chemical oxidation processes is that they are not subject to some of the rate limitations imposed on

S.J. Masten, S.H.R. Davies~Journal of Contaminant Hydrology 28 (1997) 327-335 329

biodegradation processes, since the chemical oxidants may diffuse in the aqueous and /or gaseous phase to the reaction site.

The use of gaseous ozone for in-situ chemical oxidation has significant advantages over aqueous-based systems for the treatment of unsaturated soils. As the diffusivity of ozone gas is much greater than that for aqueous species, the mass of ozone that one can contact with the target contaminants is much greater than that which is possible with dissolved oxidants. Additionally, as volatilization of the target chemical is also not required, it can overcome mass transfer limitations associated with soil venting. In-situ ozonation may also be used to treat the region near the water table by using sparging. In-situ ozonation would likely be more rapid than biodegradation or soil venting processes, thereby reducing the remediation time and, possibly, treatment costs.

Conceptually, in-situ ozonation is similar to soil venting. Vertical or horizontal injection wells can be used to inject ozone into the subsurface. Fig. 1 illustrates how a vertical injection well could be used to introduce ozone into unsaturated soils. Extraction wells may, if necessary, be used to control the direction of ozone flow in the subsurface. Horizontal wells could also be used to inject ozone. Nelson and Brown (1994) have claimed that this technique is more effective for the introduction of ozone into unsaturated soils than is the use of vertical wells. Horizontal wells are also advantageous when using ozone in combination with sparging technologies. However, while horizon- tal wells may have some advantages, it is not clear that these advantages warrant the additional cost necessary to install such wells.

In-situ ozonation has not been widely used in the field. However, to our knowledge, two groups have used in-situ ozonation for soil remediation. One group is a consortium of two West Coast companies (Steffen Robertson and Kirsten Consulting Engineers and Scientists and McCulloch Environmental Equipment Sales). This group has conducted three field-scale soil remediation efforts using ozone. In all cases, the contaminants of interest have been chlorinated solvents. The largest of these remediation efforts was at a site in Carson City, NV. This site had been contaminated with tetrachlorethylene (PCE). It was estimated that it would take 5 years to clean up this site using conventional pump-and-treat methods. The owner, a property developer, wanted to accelerate this

Ozone Generator ~ ' ~

"1

I j , i

Injection well I

I . . J

Exhaust blower

Extraction well

Air or oxygen containing ozone is drawn through the contaminated zone

Fig. 1. Schematic of in-situ ozone treatment system.

330 S.J. Masten, S.H.R. Davies / Journal ()f Contaminant Hydrology 28 (1997) 327-335

process, so that the site could be developed. Thirteen sparge wells and three horizontal vapor extraction wells were installed. After initial sparging and venting with air, ozone was used in both the sparging and soil vapor extraction wells. Samples taken after 4.5 months of operation indicated that the site had been remediated and the site is now closed (McCulloch, W., personal communication). At a PCE contaminated site in Colorado Springs, groundwater clean-up was achieved in 2.5 months (McCulloch, W., personal communication). The site was closed at the end of the third month. At a smaller site in Cape Cod, MA groundwater clean-up was achieved in less than 2 weeks (McCulloch, W., personal communication).

Nelson and Brown (1994) have described the use of ozone sparging to remove trichloroethylene (TCE) and dichloroethylene (DCE) from a site in Kansas. Prior to ozonation, air sparging was used to remove the bulk of the contaminants. Groundwater samples collected from monitoring wells ranging from 5 to 20 ft from the injection point indicated that in this region TCE concentrations were reduced from 54 to 100%. The lowest level of TCE removal (14%) was seen at a sampling location 5 ft from the ozone injection point, suggesting that preferential flow paths may have been established near to the injection well. DCE removal efficiencies ranged from 94 to 100% (with the exception of the nearby sampling point, where the removal was 30%).

To assess the applicability of in-situ ozonation, we have investigated the effect of soil organic matter, soil moisture and soil texture on the ozone demand exerted by the soil, along with the reactivity of selected PAHs with ozone in the soil and the byproducts formed from the reaction of ozone with PAHs.

3. Transport of ozone in unsaturated soils

One of the major concerns with the use of in-situ ozonation has been the ability to transport ozone any considerable distance in unsaturated soils. This stems from the fact that numerous researchers have shown that it is difficult to transport ozone or hydrogen peroxide in saturated soils, this observation should not be translated to the transport of gaseous ozone in unsaturated soils. The concentrations of ozone in the gas phase are orders of magnitude higher than that obtainable in aqueous solutions. Ozone is more stable in the gas phase than in water. In water, OH catalyses the autodecomposition of ozone (Langlais et al., 1991). Also, higher flow velocities can be achieved in the vadose zone than is possible in aquifers.

We have investigated the transport of ozone in several geological materials (Day and Masten, 1992; Day, 1994; Hsu, 1995; Cole et al., 1996). Ozonation was conducted by passing humidified gaseous ozone through 10 cm long soil columns. Details of the procedures used in these experiments are described by Cole et al. (1996), Hsu (1995) and Day (1994). As shown in Fig. 2, Day (1994) found that ozone is readily transported through columns packed with Ottawa sand and there is a rapid initial breakthrough of ozone (within 1.5 pore volumes). However, complete breakthrough is not achieved until nearly 5 pore volumes have been passed through the column. Work has also been conducted with a Metea subsoil. The ozone demand exerted by the Metea subsoil is greater than that of the Ottawa sand; breakthrough in a 10 cm column is observed in

S.J. Masten, S.H.R. Davies/Journal of Contaminant t-(vdrology 28 (1997) 327-335 331

1.2

1o

0.8

0.6 b

0.4

0.2

0.0

0 500 1000 1500

Time (sec) Fig. 2. The effect of soil moisture content on the breakthrough curves for the ozonation of Ottawa sand. C is

the effluent ozone concentration and C O is the influent ozone concentration (O air-dried sand, • 12%

moisture content, • 37% moisture content). Ozone flow rate: 120mlmin - j . Column length: 10 cm. Column

diameter: 5.5 cm I.D.

approximately 600 pore volumes. For Borden sand, 90% ozone breakthrough is achieved in a 30 cm column in approximately 300 pore volumes (Day, 1994). Similar results were also obtained with Wurtsmith soil (Cole et al., 1996). As shown in Fig. 3, the ozone demand of the Wurtsmith soil was completely satisfied within 30 min (approximately 60 pore volumes).

All the geological materials we have studied exert a limited (finite) ozone demand, i.e. the rate of ozone degradation in soil columns is very slow after the ozone demand is

1.2

1.0

0.8 ~W--

0.6 c3

0.4

0.2

0.0 I - - - - - 7 - - [ I - -

1000 2000 3000 4000 5000

Time (sec) Fig. 3. Breakthrough curve for the ozonation of Wurtsmith soil. C is the effluent ozone concentration and C~ is the influent ozone concentration. Moisture content: 6.8%. Inlet ozone concentration: 3.5%. Ozone flow rate:

50mlmin J. Column length: 10 cm. Column diameter: 5.5 cm I.D.

332

Table 1

S.J. Masten, S.H.R. Dauies / Journal o[ Contaminant l~vdroh)gy 28 (1997) 327-335

Geological material Energy cost/ton soil

kWh a SUS b

Ottawa sand Wurtsmith soil, Oscoda, MI Metea subsoil, East Lansing, MI Borden sand, Borden AFB, Ontario

< 0.22 < 0.013 < 4.3 < 0.26 31 1.85 44 2.64

aBased on an energy cost for 03 generation of 10 kWhlb -~ . bBased on a cost for electricity of $ 0.06 per kWh.

met (Day, 1994). Ottawa sand exerts little ozone demand (less than 0.04mg 0 3 per gramme sand). For the Metea subsoil the ozone demand is approximately 1.4 mg 03 per gramme soil and the ozone demand of Borden sand is approximately 2mg 03 per gramme sand (Day, 1994). For Wurtsmith soil, the ozone demand is approximately 0.215 mg 0 3 per gramme soil at a moisture content of 6.8%. At a moisture content of 3.2% the ozone demand is approximately 0.022 mg 03 per gramme soil. As such, once the initial ozone demand is met, little degradation of ozone is observed. Therefore, it should be possible, in the soils we have studied, to achieve ozone penetration to a considerable distance away from the injection well.

The costs associated with in-situ ozonation have been of significant concern to the engineering community. The two major costs would be associated with the purchase of the ozonation equipment and with operating and maintenance (O & M). The predominant O&M expenses would be associated with the generation of ozone. The energy costs to generate ozone to meet the ozone demand of an uncontaminated geological material are as mentioned in Table 1.

These costs are small compared to typical soil remediation costs (over $100 per ton). The ozone injection wells ought to be simple and relatively inexpensive to construct. We expect that the principal cost associated with the remediation effort would the capital (or rental) costs associated with the procurement of the ozone generation equipment and the operating cost for the ozone generation equipment. Such costs are difficult to estimate, as they are very dependent on the size of the ozone generation equipment to be used. For water treatment, the overall cost of ozone generation for a 200 lb 03 per day plant (excluding the cost of the contactor) was estimated in 1991 to be $US 1000 per day (i.e. $5.00 per lb) (Langlais et al., 1991). While comparison of the cost of ozone generation at a water treatment plant and the cost of ozone generation for subsurface remediation is at best, very rough, our preliminary results suggest that ozone venting costs are very likely to be competitive with the costs of other soil remediation techniques (typically over $100 per ton).

The effect of moisture content on the degradation of ozone in soils was investigated (Day, 1994). It was found that increasing the moisture content in Ottawa sand resulted in an increase in the ozone-demand exerted by the soil and an increase in the time required to achieve a constant ozone concentration in the effluent stream from the column. Air-dried Ottawa sand had an ozone demand of 0.036 mg g ] soil, while the ozone demand for the moist soils (12 and 37% moisture contents) were 0.048 and 0.0062 mg g-

s.J. Masten, S.H.R. Davies~Journal of Contaminant Hydrology 28 (1997) 327-335 333

soil), respectively. It is thought that this increase is due to the dissolution of gaseous ozone into the soil pore water and the subsequent self-decomposition of ozone once it is dissolved in the water.

4. Ozone reactions in aqueous solutions and in soil slurries

Ozone reacts directly with aromatic compounds via a t,3-dipolar cyclic addition of ozone across the double bond to yield an unstable intermediate, known as a trioxalane. These intermediates rapidly decompose to form catechols, phenols and carboxylic adds (Bailey, 1982). Phenol will be further oxidized by ozone to form a number of organic acids and aldehydes (Jarret and Bermond, 1983). The PAHs are reactive with both ozone and OH radicals (formed from the self-decomposition of ozone). As such, the remedia- tion of soils contaminated with the PAHs has considerable promise.

Much work has been accomplished in our laboratory investigating the aqueous decomposition products formed from the ozonation of pyrene (Yao, 1997). Numerous products are produced, but the intermediates formed are mostly biphenyls substituted with carboxylic acid and aldehyde functional groups. Toxicity testing using gap junction intercellular communication as a biological endpoint has revealed that the products formed upon the ozonation of pyrene at dosages of at least 4.5 tool of ozone per mole of pyrene are not epigenetic toxicants to rat liver epithelial cells (Upham et al., 1994, 1995).

5. In-situ ozonation of PAH contaminated soils

Polycyclic aromatic hydrocarbons (PAHs) are very reactive with ozone and as such, ozone is very effective for the removal of PAHs from soils. We have studied the degradation of pyrene, naphthalene, chrysene and phenanthrene in soils (Yao and Masten, 1992; Day, 1994). For a soil containing 100 mg kg- 1 pyrene, 81% of removal of pyrene could be achieved using an ozone dosage of approximately 500 mg kg-~ of soil (Day, 1994). In this experiment the rate of disappearance of pyrene was not measured, however, as the ozone was completely degraded within 15 min it is clear that the rate of disappearance of pyrene was fast. We have found that in soil, very hydrophobic PAHs (e.g. chrysene) react more slowly than would expected on the basis of their reactivity with ozone in water (Yao and Masten, 1992). Even so, after 4 h exposure to ozone, the chrysene concentration in contaminated Metea soil 1 was reduced from 100 to 50mgkg -~. In this soil, greater than 90% removal of pyrene and phenanthrene could be achieved with an ozonation time of only 1 h (Yao and Masten, 1992).

In Ottawa sand, it was observed that in dry soil, approximately 65% of the naphthalene was removed after 37 h of venting with air; this resulted in a residual

i The Metea soil contained 78.9% sand, 12.5% silt and 8.6% clay. The organic matter content of the soil was 0.4 _+ 0.1%.

334 S.J. Masten, S.H.R. Dat,ies / Journal of Contaminant Hydrology 28 (1997) 327-335

naphthalene concentration of 23.2 mg kg-~ soil. Air stripping (for 23 h) followed by ozone treatment (for 3.2 h) resulted in a naphthalene residual of 0.65 mg 1- ~ (approxi- mately 99.7% removal). Similar results were observed in moist soil when ozone venting was applied: the residual naphthalene concentration obtained was approximately two orders of magnitude lower than that obtained when using air venting alone (Hsu et al., 1993).

The effect of moisture content on the degradation of pyrene in soils was also investigated in cycling batch reactor systems (Day, 1994). It was found that the rate of ozone degradation increased with increasing moisture content in pyrene contaminated Metea soils. This resulted in a decrease in the efficiency of pyrene removal with increasing moisture content. For example, in air-dried Metea soil 23.6mg of ozone resulted in a pyrene removal efficiency of 63%; while in the same soil (moisture content of 2.9%), the same mass of ozone resulted in a treatment efficiency of 30%. This decrease in efficiency is thought to be due to the fact that, in these systems, the reaction was ozone-limited and the competing reactions (dissolution of ozone into soil pore water and ozone self-decomposition) consume ozone that would have otherwise reacted with pyrene. On the contrary, in systems where ozone was not limited (experiments that were conducted in soil columns with gaseous ozone being continually supplied), it was found that the presence of water increased the level of naphthalene removal as compared to that obtained in dry Ottawa sand. This was thought to be due to the water, a wetting fluid, displacing the trapped organic phase (naphthalene dissolved in toluene) from small pores, making it more available to the ozone for reaction.

6. Conclusions

Ozone has been found to be extremely effective for the degradation of PAHs in unsaturated soils. It has also been shown that ozone can be readily transported through columns packed with a number of geological materials, including Ottawa sand, Metea soil, Borden aquifer material and Wurtsmith aquifer material. All of these geological materials exerted a limited (finite) ozone demand, i.e. the rate of ozone degradation in soil columns is very slow after the ozone demand is met. As once the initial ozone demand is met, little degradation of ozone is observed, it should be possible to achieve ozone penetration to a considerable distance away from the injection well, indicating the feasibility of using ozone in situ for the treatment of uncontaminated unsaturated soils. This is confirmed by four field studies where in-situ ozonation was apparently success- fully used.

References

Bailey, P.S. 1982. Ozonation in Organic Chemistry, vol. 2, Nonolefinic Compounds. Academia Press, New York, pp. 30.

Cole, D.K., Davies, S.H.R., Masten, S.J., 1996. In-Situ Ozonation for the Remediation of Soils Contaminated with Petroleum Hydrocarbons, Proceedings of the IOA Pan American Conference: Applications and Optimization of Ozone for Potable Water Treatment, 8-11 September 1996.

S.J. Masten, S.H.R. Davies~Journal of Contaminant Hydrology 28 (1997) 327-335 335

Day, J.E., 1994. The Effect of Moisture on the Ozonation of Pyrene in Soils, Masters Thesis, Department of Civil and Environmental Engineering, Michigan State University, Michigan.

Day, J.E., Masten, S.J., 1992. Effect of Moisture on Removal of PAHs From Soils Using In Situ Ozonation. Proceedings of 24th Mid-Atlantic Industrial and Hazardous Waste Conference, West Virginia University, 14-17 July 1992.

Gilbert, E., 1987. Biodegradability of ozonation products as a function of COD and DOC elimination by example of substituted aromatic substances. Water Res. 21, 1273-1278.

Hsu, I.Y., 1995. The Use of Gaseous Ozone to Remediate Organic Contaminants in Unsaturated Soils. Ph.D. Dissertation. Michigan State University, East Lansing, MI.

Hsu, I.Y., Davies, S.H.R., Masten, S.J., 1993. The Use of Ozone for the Removal of Residual Trichloroethy- lene from Unsaturated Soils. Proceedings of the 48th Annual Purdue University Industrial Waste Conference, 10-12 May 1993.

Jarret, M.. Bermond, A., Ducanze, C., 1983. Applicatien de la chromatographie en phase liquide h la d6terminatien des produits form6s par ozonatien des solutiens aqueuses de ph6nol. Analusis II (4), 185-189.

Langlais, B., Reckhow, D.A., Brink, D.R., (Eds.), 1991. Ozone in Water Treatment--Application and Engineering. Lewis Publishers, Chelsea, MI.

Luthy, R.G., Dzombak, D.A., Peters, C.A,, Roy, S.B., Ramaswami, A., Nakles, D.V., Nott, B.R., 1994. Remediating tar-contaminated soils at manufactured gas plant sites. Environ. Sci. Technol. 28, 266A-276A.

Miller, R., 1994. Bioventing, Performance and Cost Summary. U.S. Air Force Center for Environmental Excellence, Brooks AFB, Texas.

Nelson, C.H.. Brown, R.A., 1994. Adapting Ozonation for Soil and Groundwater Cleanup. Environmental Engineering, a Supplement to Chemical Engineering, pp. EE20-EE24.

Upham, B., Yao, J.-J., Trosko, J.E., Masten, S.J., 1995. Determination of the Efficacy of Ozone Treatment Systems using a Gap Junction Intercellular Communication Bioassay. Environ. Sci. Technol. 29 (10), 2923-2928.

Upham, B.L., Masten, S.J., Lockwood, B.R., Trosko, J.E., 1994. Tile Effect of Ozonation on the Non-geno- toxicity of Polycyclic Aromatic Hydrocarbons using Intercellular Communication in Rat Liver Epithelial Cells. Fund. Appl. Toxicol. 23, 470-475.

US EPA, 1995. Workshop on Chemical Oxidation in the Vadose Zone. US EPA Edison Laboratory, IT Corp., Cincinnati, OH, June 14, 1995.

Yao, J.-J., Masten, S.J., 1992. Use of ozone for the oxidation of PAHs in dry soils. Proceedings of the 24th Mid-Atlantic Industrial and Hazardous Waste Conference, 14-17 July 1992, Morgantown, WV. pp. 642-651.

Yao, J.-J., 1997. The Mechanism of the Reaction of Ozone with Pyrene and Benz[a]anthracene in Acetoni- trile/water Mixture. Ph.D. Thesis Michigan State University, East Lansing M.I. 1997.