SOIL REMEDIATION

Technical Field

This invention relates to remediation of contaminated soil and, more particularly, this invention relates to remediation of soil containing organic or inorganic con- taminants by sonication of the soil in aqueous media, preferably in the presence of a surfactant.

Background of the Invention

Industrialization has raised the standard of living throughout most of the world. However, manufacture of material goods results in creation of chemical waste by¬ products that have been dumped, stored on-site or moved to storage dumps. The storage of heating fuel and other hygros¬ copic hydrocarbons in underground or surface tanks made of iron over many years has resulted in deterioration and rupture of the tanks. The hydrocarbon liquids can then percolate into the adjacent soil creating a hazardous mixture or can con¬ taminate potable water supplies underlying the spill. The number of sites requiring remediation throughout the world is staggering.

Though the severity and seriousness of the soil contamin¬ ation problem has been recognized and efforts to clean up the sites has started, there is a continuing accumulation of hazardous materials from current manufacturing activities and there are fewer and fewer sites available which will accept and store such wastes. Unless the contaminated soil can be treated on-site, transportation of large amounts of con¬ taminated soil over long distances may be required and can add a substantial cost to the manufacturing process and product. The generator of the waste is also required to pay annual fees to store the waste.

On-site remediation can cost as much as $200 to $400 per

ton of soil and may only result in a remediation of a portion of the soil. Depending on the soil type and grain size distribution, the contaminants may concentrate into a fraction of the soil. This fraction may still require transportation and storage at a designated hazardous waste storage facility. Incineration can be an appropriate on-site means to remediate contaminated soil. However, fuel is expensive and the stack gases often contain undesirable air pollutants. In- situ soil cleaning with solvents has also been proposed as a means of remediation. However, solvents are expensive, may be hazardous or toxic are often flammable and the dissolved mix¬ ture can migrate into underground water supplies thereby creating a worse problem.

Soil washing is another potential on-site remediation technology. Soil washing can be defined as the ex-situ treat¬ ment of contaminated soil using water as the primary solvent. The cleaned fraction is returned to the excavated site. Over¬ size materials are mechanically removed from the soil and may be treated by spray washing to remove contaminants. Soil washing has been practiced in Europe since the mid-

1980's and since 1990 it has been approved in the United States for remediation at 17 Superfund sites. The technology is most effective in remediating coarse grained sands or gravels contaminated with organic or inorganic compounds. The fine grain clay or silt fraction below about 70 microns presents difficult problems in removing the con¬ taminants by traditional soil washing techniques. Over time, as a result of migration, weathering and degradation, soil contaminants, having a greater affinity for the fine-grained materials, will tend to accumulate and concentrate on the fine grain particles. The physical characteristics of the fine grain particles result in greater adhesion of contaminants than on the coarse grain fractions. The fine grain particles have greater surface area and greater adsorptive binding for- ces than do larger particles. The surfaces of clay particles

can be charged which contributes to adsorption and also to dispersion of. the particles as colloidal suspensions in the liquid phase. The fine particle fraction is difficult to treat and to separate from the liquid phase. Due to the difficulty in removing the contaminants from the fine fraction, soil washing is used in conjunction with other remediation operations such as incineration or soil washing is used to concentrate the contaminants in the fine fraction which is then transported to a licensed storage facility. The volume reduction of the contaminated soil produced by traditional soil washing does therefore provide a cost benefit by substantially contributing to reduction in the volume of waste. The traditional process can concentrate 70- 90% of the non-volatile organic and heavy metal residual products into the fine fraction representing 5-40% of the original soil volume. Reduction in volume can itself contribute to cost effectiveness. However, traditional soil washing does not clean all of the soil. The fine fraction, if still contaminated, must be stored permanently or until a feasible remediation technology is developed.

There have been attempts to augment the removal of con¬ taminants from soil by agitation to provide abrasive scouring and/or scrubbing action. Surfactants can be employed to increase mobility of the washing fluid by reducing surface tension and to enhance release of the hydrophobic organic contaminant from the surface of the soil particles by reducing interfacial tension (IFT) . Though recovery of contaminants is improved, a substantial amount of organic contaminant typical¬ ly remains with the fine fraction. Another disadvantage of traditional soil washing is the movement of contaminant into the wash water and the necessity and expense of treating and disposal of the wash water and the need to constantly add substantial amounts of make up water.

Statement of the Invention

It has now been discovered in accordance with the inven¬ tion that the removal of contaminants from soils containing a substantial percentage of fines can be significantly enhanced by suspending the soil in water and subjecting the suspension to sonication. The removal of contaminants is further en¬ hanced by conducting the treatment in the presence of a sur¬ factant and especially a supersurfactant containing a large range of sizes of vesicles, particularly a supersurfactant produced by sonicating a material containing anionically substituted hydrocarbon compounds in the presence of a basic salt. It is preferred to develop the supersurfactant from a sample of the contaminated soil so that the surfactant micel¬ les have a natural similarity to and affinity for the contami- nants.

The use of sonication eliminates the need to separate the fine fraction from the soil. It also can allow retention of the treated soil on site where it can be returned to the ex¬ cavation. The ultrasonic soil washing process including the surfactant can generate fresh surfactant in-situ during the process. Furthermore, the supersurfactant can be recycled several times to clean the soil which lowers cost and conser¬ ves process water. The supersurfactant is also found to work with waters having a high salt content. The supersurfactant may exhibit detergent rather than miceller surfactant activity in saline water. However, the use of ocean or brackish water can greatly reduce process cost especially when cleaning beaches contaminated by soil spills.

Moreover, the supersurfactant appears to induce fraction- ation or "cold-cracking" in which an oily fraction thicker than the contaminant itself, rises to the surface where it can be removed by skimming or overflow into a recovery vessel. The process thereby generates two products: a recovered con¬ taminant and clean soil. The system of the invention can be applied to soils con-

taminated with inorganic or organic contaminants though it is most useful in treating soils contaminated with hazardous, heavy hydrocarbons and especially, those soils in which the fine grained fraction exceeds 20-40% by weight of the soil. Up to 90% by weight or more of the contaminant is removed simultaneously from the mixture of coarse and fine grains. Typical hazardous materials that can be substantially eliminated from soils are petroleum, diesel fuels, heating oil, jet fuel, kerosene, gasoline, fuel processing residues, creosote, metal salts, solvents such as chlorinated ethylenes, halogenated hydrocarbons such as polychlorinated biphenyls, pesticides and herbicides. The system of the invention can also be used to treat mixed hazardous wastes rich in hydrocar¬ bon residues such as those materials stored at the Casmalia and Kettleman Hills sites. These wastes may contain mixtures of organic compounds or organic and inorganic compounds such as radioactive compounds. The invention can also be used to clean crude oil contaminated sands such as in Kuwait or crude from ocean spills such as the Exxon Valdez that migrate onto beaches.

The process of the invention does not require application of heat or pressure, though local heating develops during cavitation attendant to sonication of the particulate suspen¬ sions. Preheating the supersurfactant to moderate temperatures from 30"C. to 75°C. appears to increase rate and effectiveness of the treatment.

The invention significantly enhances the environment by reducing the amount of contaminated soil present at hazardous sites and returns the site to active use. The invention is cost effective since the amount of soil to be treated by other more costly methods is substantially reduced. The process can be practiced in a closed environment, controlling fugitive dusts and volatile emissions. The process can remove up to 99% or more of contaminants from fine grained soil. These and many other features and attendant advantages of

the invention will become apparent as the invention becomes better understood by reference to the following detailed description when considered in conjunction with the accom¬ panying drawings.

Brief Description of the Drawings

Figure 1 is a block diagram view of a first system for treating contaminated soil in accordance with the invention; and Figure 2 is a schematic view of a continuous flow, soil remediation system according to the invention.

Detailed Description of the Invention

Referring now to Figure 1, the system of the invention proceeds by classifying soil 16 from tank 10, in a classifier 12 to remove large particles 14 over 5mm such as tree branches, tires, metal pieces, etc. The soil 16 is then mixed with water 18 in a slurry tank 20 containing a mixer 22 to form a slurry 24. The slurry 24 is then fed into contact with transducers 26 mounted on the inside wall 28 or outside wall 30 of the soil washing tank 32. The tank 32 may also contain a mixer 34 which circulates the slurry 24 past the surface 25 of the transducers 26 during treatment.

During sonication, minute vacuum bubbles form and implode. This action creates heat and mechanical energy at the surface of the particles and softens viscous contaminants such as heavy hydrocarbons, dislodges them from the surface of the particles and can fractionate them into individual molecules which rise to the surface of the suspension. An oily film 38 which can then rise to the surface 36 of the slurry 24 where they can be skimmed off by means of a skimmer 39 and can be recovered in tank 41. The sonicated slurry 24 can then flow through outlet 43 into a settling/separation tank 40. The oily film 38 is recovered from the surface set- tling/separation tank 40 and clean soil 42 is recovered from

the outlet 44 to the settling/separation tank 40.

Various additives can be added to the slurry tank 20 such as inorganic bases, acids or salts, surfactants, detergents, metal binding agents such as chelating agents and the like. The preferred surfacta: is formed by sonication of an aqueous suspension of anionically substituted hydrocarbon in the presence of a strong base such as a Group I or II metal hydroxide, carbonate, phosphate or silicate. The addition of an oxidizing material such as hydrogen peroxide during the formation of the preferred surfactant promotes the formation of the atomically substituted hydrocarbons. The intense local turbulence and heat caused by the cavitation, causes the inor¬ ganic base to react with anionic polar groups such as car- boxylate groups on the organic contaminants to form water soluble surfactant compounds which in turn form micelles and vesicles over a wide range of sizes from very small, less than a micron up to large, about 10 microns. The supersurfactant preferably includes at least 30% micelles having a size below 70 microns. These small micelles can enter pores in the con- taminated soil thereby more effectively dislodging, dissolving or emulsifying the contaminant prior to transferring the con¬ taminants to the aqueous emulsion phase of the suspension. The base ions also can contribute to neutralizing any electrostatic or other adhesive force between the soil par- tide and the contaminant layer which surrounds it.

The action of the supersurfactant is accelerated by the presence of a small amount of a free radical agent such as those disclosed in U.S. Patent Nos. 4,765,885; 4,897,131 and 5,017,281 the disclosures of which are expressly incorporated herein by reference. Only a trace amount is necessary. As little as 10 ^" to 1.0 gram of a free radical activator such as hydrogen peroxide, benzoyl peroxide or azoisobutyronitrile per 100 grams of organic contaminant significantly decreases the time required to separate contaminants from the particles. If the contaminants do not contain sufficient polar

groups the separation reagent can be premanufactured from tar sands and inorganic base, preferably under basic conditions. A pH of at least 7 appears to be necessary for supersurfactant activity. The preferred inorganic base is sodium silicate and, especially sodium silicates having a Si0 ₂ /Na ₂ 0 ratio of from 1.6 to about 3.20.

Referring again to Figure 1, the system may contain a surfactant supply tank 50 which adds surfactant 52 to the mixing tank 10. The tank 50 may be used to manufacture super- surfactant and may contain transducers 54 to sonicate a suspension 56 of the contaminated soil or tar sand in the presence of a base such as sodium silicate. The concentration of the base in the supersurfactant make up solution is from 1 to 20 percent by weight, usually 3 to 10 percent by weight. The ratio of base to organic contaminant such as crude oil is from 1/10 to 10/1 usually about 0.5/1 to 2/1.

A continuous flow system 100 is illustrated in Figure 2. The system 100 includes a mixing hopper 102, a sonication trough 104 and a settling tank 106 connected in series. The mixing hopper 102 contains a paddle mixer 108 driven by motor 110 connected to power supply 112. A plurality of transducers 114 are disposed on the bottom wall 116 of the sonication trough 104 and are electrically connected to the power supply and controller 112. A set of adjustable baffles 118 controls the level of the layer 120 flowing over the transducers 114. An air supply 122 is connected to an air injector chamber 124 positioned at the bottom of the settling tank 106.

The system is operated by adding contaminated soil 126, surfactant 128 and water 130 to the mixing hopper 102 and rotating paddle mixer 108 to form a suspension 132. The suspension 132 flows out the outlet 120 into the sonication trough 104. The baffles 118 are adjusted to form a narrow layer 120 of suspension over the transducers 114. The transducers sonicate the layer 120 to remove contaminants from the particles. The suspension 132 flows through outlet 138

into the settling tank 106. The sonication trough 104 can be hingedly connected to the mixing hopper 102 and settling tank 106 by means of hinges 150, 152 and the pitch of the trough can be adjusted to control the rate of flow of the suspension 132.

Large particles 140 settle to the bottom of the settling tank 106 and they are recovered at outlet 142. The air bub¬ bles float the contaminant to the top of the suspension to form a layer 144 recovered through outlet 146. The surfactant and fine particles are removed through outlet 148 which can be recycled to the mixing hopper 102. Fine particles 156 can be removed from the recycled stream by floatation, filtration, centrifugation, etc. in separation unit 154.

The particles forming the suspension should be free and independent. Large objects are first removed by screening. Usually particles larger than 5mm are removed by screening. If the excavated soil contains large agglomerates the screened soil is crushed in a mill to form particles passing a 60 to 80 standard U.S. mesh screen. The crushed particles are then mixed with water to form a suspension. The amount of solid particles present in the suspension depends on the con¬ centration of separation reagent, the energy and frequency of the sonication applied to the suspension, and the depth of the suspension. Usually the particles form from 1 to 50 percent by weight of the suspension, preferably from 10 to 30 percent by weight.

The suspension is preferably subjected to a pretreatment before being fed to the sonication unit. During pretreatment the surfactant can penetrate the layer of contaminant, reduce surface tension at the interface of the soil particle and hydrocarbon layer. The surfactant also coats the outside surface of the contaminant coated particles and increases the fluidity and lowers viscosity of the suspension. The pretreatment can be conducted at ambient temperature or the suspension can be heated to 100°C. in the pretreatment vessel.

The recycled stream from the sonication vessel is usually at an elevated temperature from 40°- 60"C. as a result of the heat generated during sonication. The use of heated recycle stream will preheat and soften the contaminant layer. The frequency and power of the sonic energy applied to the suspension depends on the location of the transducers, thickness and solids content of the suspension, solubility or dispersibility of the contaminant, etc. The transducers are preferably water resistant and can be disposed within the suspension. Suitable transducers are 25 kHz immersible piezoelectric transducers manufactured by Bronson, operating at 720 watts.

The transducers can be located along one wall such as the bottom wall of the sonication trough or can also be disposed on the side walls and top wall of a closed trough. The fre¬ quency of the sonication controls the number of implosion sequences per unit time. For example, at 45 kHz ultrasonic frequency, about 90,000 bubbles form and implode per second in the suspension. The ultrasonic frequency is usually from 10 to 60, typically 20-45 kHz.

The suspension can be stationary and treated as a batch process. It is preferred to agitate or circulate the suspen¬ sion so that all contaminated particles flow in a narrow zone past the surface of the transducers. The active sonication zone can be as small as 1-2 inches and is usually ho more than 20 inches. The suspension can be recirculated and allowed to flow into the sonication zone and past the transducers a plurality of times or the layer of suspension can flow past the transducers only once in a continuous or semi-continuous process.

The sonication of an aqueous suspension of contaminated soil containing a significant fine grain fraction at these frequencies is found to remove a significant portion of the contaminant from the soil. The recovery of contaminant can be enhanced by use of additives. Chelating agents such as

ethylenediamine te acetic acid (EDTA) can be used to bind heavy metals. PH -....usting agents such as inorganic or or¬ ganic acids or bases can be added. It is preferred to main¬ tain the suspension at a pH above 7 during pretreatment and sonication usually with a sodium or potassium oxide or hydroxide.

Many types of surfactants can be added to the suspension. Nonionic and anionic surfactants are preferred. Suitable nonionic surfactants are the polyakylene oxide based deter- gents. Representative anionic surfactants are long chain C ₁₀ -

C ₂₄ akyl, alkylaryl, sulfate, carboxyl or sulfonic acids such as sodium dodecyl sulfate (SDS) . As previously discussed, surfactants formed by sonication of a mixture of large chain hydrocarbon substituted with polar groups form a very active and effective supersurfactant. The supersurfactant under the influence of cavitation contains a wide range of vesicle sizes including at last 30% of small micelles below 70 microns and preferably at least 10% of micelles below 10 microns. The small micelles are effective in removing the contaminant layer from the below 70 micron fraction of the soil. The vesicles in combination with the high anion concentration weaken the physical force binding the hydrocarbon contaminant layer to the soil particles. The sonication waves and cavitation for¬ ces drive the small micelles and multilayer vesicles into the small pores in the soil penetrating the hydrocarbon layer therein and chemically and physically disrupting the layer thereby removing it from the particle and emulsifying the hydrocarbon into the aqueous suspension. If the hydrocarbon contamination includes fractions containing polar groups the supersurfactant can incorporate the polar substituted con¬ taminant material into membranes by invasion of the micelles into vesicles and stabilization of the hydrocarbon by the resin components of the layer of the vesicle as disclosed in U.S. Patent Nos. 4,765,885; 4,891,131 and 5,017,281, the disclosures of which are expressly incorporated herein by

reference. This activity fractionates and separates a light hydrocarbon layer which floats to the top of the suspension. This cold-cracking separation process operates by means of membrane-mimetic chemistry. A reagent known as REMSOL based on tar sand can be utilized initially to start the process until enough supersur¬ factant is formed in-situ in the sonication tank or the tar sand based reagent can be solely used as the separation reagent for the separation process. REMSOL is usually prepared in the absence of organic solvent. The reagent is prepared by adding 10-35 percent by weight of tar sand and at least 0.01 percent by weight of an alkaline metal salt such as sodium silicate to water and sonicating the suspension for a time sufficient to form a sur- factant. A small amount of a peroxide such as hydrogen peroxide can be present as an adjuvant to increase the rate of separation. A light layer of hydrocarbon is removed from the surface and asphalt agglomerates and sand are removed from the bottom of the sonication vessel. Usually the aqueous phase of the first suspension contains 40-70 percent of the supersur¬ factant. The time needed for separation of contaminants is enhanced by the addition of a free radical agent as previously discussed.

Supersurfactant separation reagent was prepared according to the following procedure.

EXAMPLE 1 Approximately 50ml of sodium silicate was added to 1000ml of distilled water in a one liter glass beaker. This solution was then heated to 45°C. and stirred until all sodium silicate dissolved. 200 grams of Athabasca tar sand (14% hydrocarbon) was added to the solution to form a suspension. About 1 ml of 35% hydrogen peroxide was added to the suspension. The beaker was then placed into a sonic bath and sonicated at 45 kHz with stirring at 320 RPM for 5 to 30 minutes. The sonicator con- tained a 3 gallon bath. The bath could be sonicated with

variai e energy output up to 0.5 watts/square inch and the freque -y could be varied up to about 40 kHz. The solution becomes very dark. About 5% of the bitumen dissolved to form the supersurfactant, REMSOL. Higher concentrations of REMSOL, up to 20% are obtained by sonicating the suspension for a longer period of 1-3 hours. The solution was set aside for several hours until the solids settled out. The bitumen on top of the solution was removed and the solution of REMSOL was carefully poured off and saved. The supersurfactant of Example 1 was tested for aquatic toxicity against the Fathead Minnow for 96 hours by the State LC 50 toxicity test. REMSOL does not have an aquatic 96 - hour LC 50 less than 500 mg/L with Fathead Minnows and accor¬ ding to 22 Cal.Adm. Code, Art. 11, Sec. 6696(4) is not hazar- dous or toxic by this criterion.

EXAMPLE 2 A test of the surfactant properties of REMSOL for a high¬ ly contaminated soil was conducted. Soils obtained from a petroleum refinery contaminated with a complex mixture of a heavy vacuum gas oil, Diesel No. 2 hydrocarbon was tested by adding REMSOL to a 20% suspension of the soil in a sonication tank at 45 kHz. Within 2 minutes of stirring, a visible dark layer formed on the surface of the suspension. The con¬ taminant was removed from all soil particles including fines. The soil is a mixture of clay and degraded sandstone.

EXAMPLE 3 A standard soil was prepared by screening soil to remove large soil particles and organic matter larger than 1 mm. The soil was then heated at 400°C. in a muffle furnace to remove any other organic material. Various hydrocarbon contaminants were then added to the soil and tested under a variety of conditions.

EXAMPLE 4 500 grams of the roasted, clean sandy soil of Example 3 was mixed for 8 hours with 2.4% by weight of Diesel No. 2 in

a sealed bottle to ensure homogeneity. 10 grams of the con¬ taminated standard soil and 50 grams of the roasted soil were extracted using 1:1 v/v hexane:acetone to develop background levels of hydrocarbon contributed by the soil.

TABLE 1 Sample Concentration of Hydrocarbon Mg/kg

Clean Roasted Soil < 5 Diesel contaminated Soil 24,000

EXAMPLE 5 Suspensions of 50 grams of the standard soil of Example

4 were prepared in 100 ml of deionized (D.I.) water. A 1% dilution of REMSOL of Example 1, a 5% dilution of REMSOL of Example 1 and a 20% dilution of REMSOL of Example 1 were prepared. Suspensions of 50 grams of roasted clean soil in 100ml of D.I. water and 1% dilution of REMSOL were prepared. These suspensions and 100ml stock REMSOL from Example 1 in 250ml sealed jars were rotated for 8 hours. The samples were then sonicated for 30 minutes at 40 kHz using a Bronson Sonic 4 sonicator. The supersurfactant liquid (95-99ml) was decanted. The remaining soil was extracted with 3-20ml of (1:1 v/v) hexane:acetone solvent. The combined extracts were concentrated to 2ml and analyzed by gas chromatography/mass spectrometry. The 100ml of REMSOL was extracted with methylene chloride to determine the amount of hydrocarbon contributed by the reagent.

TABLE 2

Residual Hydrocarbon Percent Sample Concentration, mg/kg Hydrocarbon Removed

Cont. Soil β D.I. H ₂ 0 2500 90

Cont. Soil @ 1% REMSOL 1700 93

Cont. Soil @ 5% REMSOL 1100 95 Cont. Soil @ 20% REMSOL 970 97

Clean Soil @ D.I. H ₂ 0 < 5

Clean Soil @ 1% REMSOL < 5

100ml REMSOL * < 0.2 * Not applicable

Sonication in water alone was effective in removing 90% of Diesel No. 2 oil from the soil. The recovery of the Diesel oil increased from 93 to 97% as the concentration of the supersurfactant increased from 1% to 20% by volume. The clean roasted soil showed less than 5mg/kg of Diesel No 2 which is the detection limit for the measurement. The hydrocarbon contribution of the reagent to the residual soil was negligible.

EXAMPLE 6 The ability of REMSOL of Example 1 to separate hydrocar¬ bons from soil was compared to a commercial soap, surfactant, which is believed to be based on detergent alkalate surfac- tants. 100 ml of D.I. water containing 5 drops of the commer¬ cial surfactant, was added to 10 grams of the contaminated soil of Example 4. The suspension and a suspension of 100ml of D.I. water containing the 20% REMSOL dilution were sonicated for 30 minutes at the same frequency and power. The supernatant liquids were poured off and the residual soil was extracted with solvent and analyzed as in Example 5. Results follow:

TABLE 3

Hydrocarbon Concentrat i on Percent Sample In Soi l mg/kg Remova l

REMSOL ( 20%) 1800 92 . 5%

Commercial surfactant 3200 86 . 7%

The REMSOL sonicated soil is nearly twice as clean as the soil sonicated with ordinary detergent.

EXAMPLE 7

The solvation action of REMSOL without sonication on Santa Maria Crude Oil, a heavy, tarry crude, was compared to that of water and seawater. Soil was contaminated with crude, rinsed with 35ml of 5% REMSOL (Example 1) (A) , D.I. water (B) or sea water (C) . The extracted soil samples and supersurfac¬ tant solution were separately extracted with dichloromethane results follow:

TABLE 4

Sample Total Crude Percent of

Recovery (mg) Total Recovery

A

REMSOL 39 33

Soil 79 70

Total 118 100

B

Water ND 0 Soil 123 100

Total 123 100

Sea Water ND 0

Soil 163 100

Total 163 100

Neither water nor sea water shows solvation action or crude contaminated soil. REMSOL dissolved about 1/3 of the

crude without sonication.

EXAMPLE 8

A matrix that resembles the beach sand at Valdez Alaska was prepared from a mixture of coarse sand, fine sand, gravel and rocks. The matrix was contaminated with 160,000 parts per million of crude oil. The contaminated sample was mixed with the 5% REMSOL separation reagent prepared in Example 1 and sonicated under the conditions of Example 2. The untreated matrix (A) , contaminated matrix (B) , and treated matrix (C) , were extracted with methylene chloride. Results follow:

TABLE 5 Sample mg/kg Hydrocarbon

A 18

B 160,000

C 10

The treated matrix had less hydrocarbon than the untreated material. The reagent removed essentially all the crude from the beach sand. (More than 99.99% removed).

EXAMPLE 9

The following matrixes were subjected to extraction with 5% reagent of Example 1 with application of sonication at 0.5 water/sq.in. in a 3 gallon bath for 2 hours at 40 kHz. The samples were stirred. The suspensions were skimmed and stirred every 30 minutes. Results follow:

TABLE 6

Mat ri x/Contami nant Clean Matrix Matrix plus After % Contaminant contaminant treatment Removed

TAR SAND n/a 74 , 000. 140. 99 . 8%

Coarse Beach Sand 18 . 160 , 000. 10 . 99 . 99% w/CRUDE OI L

Bentonite with 24. 180,000. 1,200. 99.3% CRUDE OIL

Modeling Clay 160,000 310,000. 400. 99.9% with CRUDE OIL

The remediation reagent uniformly removed substantially all the heavy hydrocarbon contaminant from the matrixes even fine clays such as Bentonite, though the Bentonite containing sample had to be centrifuged after extraction. Modeling clay contains a high concentration of wax material as sold. It is noted that the wax material was extracted along with the Crude Oil.

EXAMPLE 10

Soil washing experiments were conducted on several dif¬ ferent matrixes with the reagent of Example 1 under the con¬ ditions of Example 2. Samples were sonicated for 2 hours. Every 30 minutes the samples were stirred and then separated hydrocarbon was skimmed from the surface. Results follow:

TABLE 7 CONTAMINANT CONCENTRATIONS In mg/kg

Mat r i x/Contami nant Replicate Matrix Plus After REMSOL X Contaminant

Number Contaminant Treatment Removed

Athabasca 1 74,000. 140. 99.8%

TAR SAND 2 74,000. 240. 99.7%

Coarse Beach 1 100,000. 50. 99.9%

Sand w/ CRUDE OIL

2 100,000. 70. 99.9%

Coarse Beach 1 100,000. 240. 99.8%

Sand w/DIESEL #2

2 100,000. 270. 99.7% ine Beach Sand 1 35,000. 200. 99.4% w/CRUDE OIL

2 35,000. 240. 99.3%

Clay Soil 1 100,000. 2,100. 97.9% w/CRUDE OIL 2 100,000. 3,600. 96.4%

3 100,000. 4,300. 95.7%

Clay Soil 1 100,000. 180. 99.8%

W/DIESEL #2

Soil Spiked w/PCB 1 600. 27. 95.5%

AROCLOR 1260

2 600. 59. 90.2%

3 600. 27. 95.5%

Coarse Beach Sand 1 100,000. 290. 99.7% w/UASTE OIL

2 100,000. 340. 99.7%

Coarse Beach Sand 1 100,000. 120. 99.9% w/HYDRAULIC OIL

2 100,000. 1,300. 98.7%

Fine Beach Sand 1 100,000. 170. 99.8% w/CRUDE OIL (time

2 100,000. 170. 99.8% study) 3 100,000. 140. 99.9%

4 100,000. 60. 99.9%

Fine Beach Sand 1 100,000. 170. 99.8% w/UASTE OIL

2 100,000. 190. 99.8%

Fine Beach Sand 1 100,000. 100. 99.9% w/HYDRAULIC OIL

2 100,000. 200. 99.8%

Clay Soil 1 200,000. 1,100. 99.5%

W/WASTE OIL 2 200,000. 1,700. 99.2%

Clay Soil 1 100,000. 880. 99.1% w/HYDRAULIC OIL

2 100,000. 1,500. 98.5%

In the time study, the remaining sand in the sample was analyzed for residual Oil after every 30 minutes, .

The use of separation reagent with sonication resulted in essentially complete remediation of light or heavy hydrocarbon contamination of fine or coarse soil matrixes.

EXAMPLE 11 In this experiment, suspensions of contaminated soil were subjected to sonic energy and constant stirring during extrac¬ tion with the reagent of Example 1. Each suspension was sonicated for two hours, during which time sub-samples were taken and analyzed at 15 minute intervals. Results follow:

TABLE 8 CONTAMINANT CONCENTRATIONS In mg/kg

Matrix/ Time Sample Starting After REMSOL X Contaminant Contaminant Interval Concentration Treatment Removed Number

Clay Soil w/ 15 min. 1 100,000. 4,200. 95.8% CRUDE OIL

30 min. 2 3,400. 96.6%

45 min. 3 3,600. 96.4%

60 min. 4 1,700. 98.3%

75 min. 5 2,000. 98.0%

90 min. 6 4,800. 95.2%

105 min. 7 3,000. 97.0%

120 min. 8 5,100. 94.9%

It appears that remediation was at a maximum in about 1 hour. Further sonication resulted in no further improvement.

EXAMPLE 12

The procedure of Example 1 was repeated substituting a like amount of crude oil by weight for the tar sand. A dark, oily suspension resulted.

EXAMPLE 13

The surfactants of Example 1 and of Example 12 were utilized to treat soil containing 100,000 ppm of crude oil contaminant according to the procedure of Example 2. A

control sample with only water was also run. All three suspensions were sonicated for 45 minutes and rinsed with deionized water 3 times. The soil samples were then analyzed for TPH using the EPA 418.1 method. Results follow: TABLE 9

Sampl e Treatment Residual crude oi l , pom X HC Removed

Surfactant of Example 1 670 99.3

Crude Oil derived surfactant 650 99.4 Water 48,000 52

These soil washing experiments were conducted in a mobile, trailer based pilot plant unit. The pilot plant con¬ tains a horizontally mounted mixing drum with internal flutes. The suspension moves down the flutes and enters a sonication trough. Three Bronson 20 mHz submersible, sonication transducers are mounted within the trough. The trough is mounted at an angel. The suspension moves down the trough by gravity, optionally aided by pumping, over the transducers as a narrow layer. The intense mixing separates the hydrocarbon from the soil. The hydrocarbon rises to the surface and is removed by skimming.

EXAMPLE 14 2 liters of sand contaminated with 10 ⁵ ppm of crude oil was formed into a suspension with 6 gallons of water con¬ taining 20% of the surfactant of Example 1 in the mixing drum of the pilot plant. The suspension was sonicated for 20 minutes. The cleaned sand was washed and analyzed. The residual hydrocarbon content was 160 ppm representing a 99.8% removal of hydrocarbon.

EXAMPLE 15 A test soil was prepared by sieving 30 mesh sand through a 200 mesh screen (75 microns) . The fine fraction collected on the pan was spiked with 10 ⁵ ppm of crude oil. The con- taminated soil was suspended (20% by weight) in the REMSOL reagent of Example 1 and sonicated under the conditions of Example 2 for 20 minutes. 95% of the fines suspended in the

reagent were essentially free of contaminant. 5% of the fines were 95% free of contaminant. The amount of contaminated fines is reduced by 19 fold.

In certain cases, it may be necessary to conduct the soil washing in salt water or in extremely hard water due to the lack of available fresh water. In such cases, it has been found to be effective to add a chelating agent such as ethylene diamine tetracetic acid or some other inexpensive complexing agent in order to prevent the REMSOL surfactant from becoming rendered insoluble by calcium, magnesium and other cationic species present in salty and/or hard waters. It is most convenient to add the chelating agent to the water supply in whatever amount is observed to just prevent cloudiness from developing. An alternate solution for salt water applications invol¬ ves the creation of a remediation surfactant or detergent reagent under strongly acidic instead of basic conditions. For example, this may be accomplished by treating the tar sand or on-site hydrocarbon mixture with concentrated sulfuric acid in place of the usual sodium silicate. The surfactant thus produced may be used in salt water or hard water without the development of cloudiness and without the need for a chelating agent.

EXAMPLE 16 Soil samples of the standard soil of Example 3 were spiked with a mixture of Aroclor 1016 and Aroclor 1254 in capacitor oil. The samples were placed in beakers which were sonicated for 20 minutes at 40 kHz in a 3 gallon bath in a Braun sonic 4 sonicator at 0.5 watts/square inch and then stirred and skimmed and then sonicated for an additional 20 minutes. The supersurfactant liquid was decanted. The remaining soil was extracted with 3-20 ml aliquots of (l.lc v/v) hexane:acetone solvent. The extracts were combined, concentrated to 2 ml and analyzed by gas chromatography mass spectrometry using an EPA 8270 instrument. Total extraction

time was 40 minutes, Results follow: TABLE 10

Matrix/ PCB Type Capacitor OiI PCB in Matrix PCB in Matrix X Con¬

Contaminant Cone, in mg/kg Before mg/kg After taminant

Matrix Treatment Treatment Removed

Soil w/Aroclor 1016 20,000. 3.9 0.54 86%

1016 and 1254

1254 20,000. 6.1 0.78 87%

Fine Beach 1016 20,000. 3.9 0.11 97%

Sand w/ 1016

1254 20,000. 6.1 0.22 96% and 1254

Soil w/ 1016 100,000. 20. 0.17 99%

Aroclor 1016

1254 100,000. 31. 0.29 99% and 1254

Fine Beach 1016 100,000. 20. 0.06 99%

Sand w/1016

1254 100,000. 31. 0.12 99% and 1254

The treatment resulted in removal of over 85% by weight of the PCT in every case. Surprisingly, the sonication-sur- factant treatment was more effective as the PCB concentration increased. Effectiveness was higher in fine beach sand than in soil.

EXAMPLE 17

The experiment of Example 16 was repeated with the analysis of soil samples at 20 minute and 40 minute intervals. Results follow:

TABLE 11

Matrix/ Time Capacitor PCB Cone. PCB Cone. X Contaminant Contaminant Interval Before After Removed Oil mg/kg

Treatment Treatment

Soil After 100,000. 20. 0.6 97% w/Aroclor 20 min. 100,000. 31. 1 97% 1016 and 1254

Soil After 100,000. 0.11 99% w/Aroclor 40 min. 100,000. 0.2 99% 1016 and 1254

The removal of contaminant increased from 97% to 99% by weight after 40 minutes.

EXAMPLE 18 Example 17 was repeated except that fresh reagent was added after 20 minutes. Results follow:

TABLE 12

Matrix/ Extraction Capacitor PCB Cone. PCB Cone. X Contaminant Contaminant Interval Before After Removed Oil mg/kg

Treatment Treatment

Soil w/PCB 20 min. 100,000. 20. 0.18 99% 1016 and 1254 100,000. 31. 0.31 99%

Soil w/PCB 20 min. 100,000. 0.18 0.56 1016 and 1254 100,000. 0.31 0.52

There was no further reduction in PCB concentration by using fresh supersurfactant after the initial treatment.

EXAMPLE 19

A soil sample was spiked with pure Aroclor 1260 (no capacitor oil) and extracted under the conditions of Example 17 (2 - 20 minute sonications with skimming and stirring at 20 and 40 minutes followed by addition of fresh supersurfactant) . Results follow:

Extraction PCB Concentration PCB Concentration X Contaminant Before Treatment After Treatment Removed Interval

Extraction #1 600. 13 98% 20 min.

Extraction #2 7 99% 40 min.

Extraction #2 7. 8 20 min.

Again essentially all the PCB was removed after 20 minutes. 1% more was removed after 40 minutes and addition of fresh reagent and further sonication was not effective in removing the last 1% of the PCB.

EXAMPLE 20

Solvent extraction with a layer of hexane was utilized to skim the separated PCB in the following experiments which were conducted under the conditions of Example 17. The soil was spiked with a mixture of Aroclor 1200 and Aroclor 1254. Results follow:

Time Capacitor O l PCB Cone, i PCB Cone, in X Contaminant Interval in Matrix mg/kg Matrix Before Matrix After Removed Treatment mg/kg Treatment mg/kg

20 min. 100,000. 20. 0.22 99% 100,000. 31. 0.41 99%

40 min. 100,000. 0.19 99% 100,000. 0.34 99%

The hexane layer was effective in removing the separated PCB from the top of the suspension.

EXAMPLE 21

The experiment of Example 20 was repeated using pure Aroclor 1260 (no capacitor oil) . Results follow:

TABLE 13

Time PCB Cone, i n Matrix PCB Cone, in Matrix X Contaminant Removed Before Treatment mg/kg After Treatment mg/kg Interval

20 Min . 600. 15 98%

40 Min . 13 99%

60 Min . . 8 99%

Again the floating layer of hexane was effective in ab¬ sorbing the PCB which separated from the soil and rose to the surface of the suspension.

EXAMPLE 22 Soil spiked with pure Aroclor 1260 was sonicated for 2 hours with stirring and skimming every 30 minutes. Results follow:

TABLE 14

Time PCB Cone, in Matrix PCB Cone, in Matrix X Contaminant Removed Before Treatment mg/kg After Treatment mg/kg Interval

30 Min. 600. 27 95 . 5%

60 Min . 600. 59 90. 2%

90 Min . 600. 27 95. 5%

Extraction appeared to be at a maximum at 30 minutes and the PCB redissolved at the 30-60 minute interval and was re- extracted at the 2 hour interval. The use of the hexane layer appears to increase recovery. (See Example 21 and Table 13)

EXAMPLE 23 The distribution of PCB between the oil layer, reagent layer and soil for the soil spiked with Arolcors 1016 and 1254 (Examples 20 and 21) is presented in the following Table:

TABLE 15

Layer PCB Type Initial PCB After Treatment PCB Distribution Cone, in Matrix

(NORMALIZED)

Oil 1016 0 11.8 mg/kg 98% 1254 0 19.6 mg/kg 98%

Reagent 1016 0 0.003 mg/1 0.025% 1254 0 0.005 mg/1 0.025%

Soil 1016 12 0.24 mg/kg 2.0% 1254 20 0.37 mg/kg 1.9%

EXAMPLE 24 The distribution of PCB in a soil sample spiked with pure Aroclor 1260 (no capacitor oil) is presented in the following Table:

TABLE 16

Layer PCB Type Initial PCB After Treatment PCB Distribution Cone, in Matrix

Reagent 1260 0 98 98%

Soil 1260 100 2.0 2.0%

There was no floating oil-emulsion layer on top of the reagent-soil suspension. The PCB was extracted directly into the hexane layer.

The invention is based on matching vesicle size of the supersurfactant with pore or grain size of the fines in the soil. The surfactant slurry usually contains from 1 to 15% organic content. Formation of liquid crystals is evident after 5-10 minutes of sonication. Presoaking in supersurfac¬ tant solution for periods of 1-20 hours before sonication treatment reduces time of treatment and generates multilayered or multicompartmented surfactant vesicles. These complex

vesicles are believed responsible for release of hydrocarbon from substrates such as fines on which they are absorbed. A supermicelle having a diameter from 80 to lOOA and a large supermicelle having a diameter of about 200A. The supermicel- les can aggregate or precipitate into micelles or floe having a size of about 2000A. Tests confirm that the surfactant vesicles are polymeric form probably due to free radical polymerization. The surfactant vesicles are stable after several years of storage. It is to be realized that only preferred embodiments of the invention have been described and that numerous substitutions, modifications and alterations are permissible without departing from the spirit and scope of the invention as defined in the following claims.