WO2002038504A1

WO2002038504A1 - Apparatus and method for in situ burning of oil spills__________

Info

Publication number: WO2002038504A1
Application number: PCT/US2001/041835
Authority: WO
Inventors: Gary A. Breitenbeck
Original assignee: The Board Of Supervisiors Of Louisiana State University And Agricultural And Mechanical College
Priority date: 2000-11-09
Filing date: 2001-08-22
Publication date: 2002-05-16
Also published as: AU2001283581A1

Abstract

A method for in-situ burning of oil slicks on water, including the steps of: 1) providing a plurality of individual floating wicks; 2) positioning the wicks in the oil slick; and 3) igniting the wicks. Additionally, a method for manufacturing floating wicks for use in burning oil spills, including the steps of: 1) providing a wick material having a dry density less than water; 2) applying to the wick material an oleophilic, hydrophobic coating such that the coated wick material has an oil uptake capacity of at least 2 g/g and a water uptake capacity of no greater than 5 g/g; and 3) drying the wick material.

Description

APPARATUS & METHOD FOR IN SITU BURNING OF OIL SPILLS BACKGROUND OF THE INVENTION

This application claims priority to U.S. Provisional Application No. 60/247,868 filed on November 9, 2000. Technical Field

The present invention relates to methods of cleaning up oil spills on bodies of water, hi particular, the present invention relates to a novel method of in situ burning of oil on bodies of water. Background Art In situ burning has received considerable attention in recent years as a cleanup method for oil in both upland ecosystems and on water. In situ burning involves the ignition and burning of oil spilled on water, vegetation or soils. The use of burning to address large scale oil spills at sea dates back to the Torrey Canyon incident in 1967. The failure of that and similar efforts discouraged the use of in situ burning until interest was renewed by the successful burn during the Exxon Naldez incident (1989). Since then laboratory tests, the Mobile Al burn tests (1991- 1994), the Alaska Emulsion Burn Experiments, and the Newfoundland Offshore Burn Experiment showed that when properly managed in situ burning is a rapid, effective and environmentally safe technology for removing large quantities of floating oil. As a result, many authorities now consider burning as a valuable and effective tool rather than a method of last resort in the event of a major spill.

The ability of burning to quickly remove spilled oil and prevent its spread to sensitive sites or larger areas is perhaps burning's greatest advantage as a response strategy. Because oil is destroyed rather than collected, burning is attractive where transportation and disposal options are limited. It may be the only viable alternative in many remote locations where mechanical, dispersant and no-cleanup options are more damaging to the environment. Burning merits special consideration for remediation of wetland environments or other oil-contaminated sites where access is limited or at sites where other methods prove ineffective or excessively intrusive. Several studies suggest that burning of spilled oil on open water and in upland environments may be more effective and more environmentally benign than intrusive mechanical and chemical treatments.

The ability of oil slicks to sustain combustion on open water depends largely on the thickness of the oil film. Oils will ignite if they are at least 2 -3 mm thick and will burn down to slicks 1-2 mm thick. As oil is burned, the slick thins to a point where sufficient heat is lost to the underlying water to lower temperatures below that required to sustain combustion, hi addition, "boilover" may occur when underlying water layers reach boiling temperatures, quenching the burn. Because oil on open water rapidly spreads to equilibrium thickness that is frequently less than the 2-3 mm needed for sustained combustion, heat-resistant booms are sometimes employed to entrap oil and maintain adequate thickness for efficient burning. Tins technique is costly and cumbersome to deploy, and therefore its use has been restricted to spills in remote yet assessable areas on relatively calm waters. Timeliness is far more critical to the success of marine burns than to inland burns where spilled oil has been successfully burned months and even years after impact. Dispersion rapidly thins most oil slicks to a degree that will not maintain combustion without confining booms. Spilled oil can rapidly emulsify when left to float on water and emulsification is accelerated by wave action. Attempts to burn heavily emulsified oil have not been successful. Evaporative losses of hydrocarbons, or weathering, results in the loss of the more easily ignited and burned volatile and semi-volatile components. Burning of heavier oils has proven difficult.

Because of the tendency of spilled oil to disperse on water, containment is required not only to prevent spreading but also to concentrate oil so that slicks are of sufficient thickness to ignite and burn efficiently. Without the benefit of containment booms, burning can only be accomplished within the first few hours after a spill event because oil rapidly spreads to equilibrium thickness. This thickness ranges between 0.01 to 0.1 mm for light crudes and fuel oils and 0.05-0.5 mm for heavy crudes and oils. In recent years, fire-resistant booms have been developed to facilitate burning and more are under development. Ideally, booms for burning should not only be fire-resistant, but lightweight, sufficiently flexible to accommodate waves, and easily deployed. During the early development of marine burning of oil, techniques for ignition of the slick were a principal focus, no doubt because of the difficulty encountered when attempting to ignite slicks. It is now well established that slick thickness and oil type and condition are the principal factors influencing ignition and continued burning. Many devices have been devised to supply sufficient heat to ignite contained slicks with sufficient thickness for burning. They may be as simple as a roll of toilet-paper or rag soaked in diesel fuel and tossed into the slick, or as sophisticated as the "heliotorch", a device suspended from a helicopter that drops burning packets of gelled gasoline. Similar devices are used by forestry companies and agencies to create back fires and therefore are available in most localities.

In the early 1970's, prior art attempts were made to design non-combustible silicate-based foams to be applied to oil slicks as absorbent material to support burning. See U.S. Patent Nos. 3,698,850 to Sparlin; 3,843,306 to Whittington et al.; and 3,696,051 to McGuire et al, 1972). It is believed that no field evaluation of these materials was ever performed. Failure to test these materials may have been the result of unfortunate timing since at the time these patents were issued, burning oil at sea was considered a problem, not a solution. Also, because these materials were not biodegradable, they posed a potential hazard to sea life. Additionally, recovery of these silicate materials by skimming apparently was not easily accomplished. What is needed in the art is a small, floating device to concentrate and support burning of thin fihns of floating oil, but which is readily recoverable and is not a danger to marine life forms.

OBJECTS OF THE INVENTION The present invention provides a method for in-situ burning of oil slicks on water. The method includes the steps of: 1) providing a plurality of individual floating wicks; 2) positioning the wicks in the oil slick; and 3) igniting the wicks.

The invention also includes a method for manufacturing floating wicks for use in burning oil spills. This method includes the steps of: 1) providing a wick material having a dry density less than water; 2) applying to the wick material an oleophilic, hydrophobic coating such that the coated wick material has an oil uptake capacity of at least 2 g/g and a water uptake capacity of no greater than 5 g/g; and 3) drying the wick material.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a chart illustrating water uptake of various hydrophobic coatings. Figure 2 is a chart illustrating the hydrophobic effect of ethyl cellulose. Figure 3 illustrates a typical elliptical shape of a wick of the present invention. Figure 4 illustrates a mold used to form the elliptical shape seen in Figure 3.

Figure 5 is a chart illustrating the long term bouncy of the wicks. Figure 6 is a chart illustrating the results of burning emulsified oils. Figure 7 is a diagram illustrating one method of deploying wicks within a oil slick. BEST MODE FOR CARRYING OUT THE INVENTION The present invention comprises methods of manufacturing and employing effective floating wick-like devices to support the ignition and efficient burning of spilled oil on water. Ideally, these devices should be lightweight and suitable for aerial delivery by fixed- wing aircraft so that distant spills can be addressed in a timely fashion. They should remain buoyant for several days after the burn to allow recovery. Because complete recovery of the devices is unlikely, they should be completely biodegradable. To minimize hazards to sea animals and fish, they should be non-toxic when ingested and retain as little absorbed oil as possible after use. To minimize hazards to workers, all components should burn cleanly without the formation of toxic gases. Moreover, the size and shape should allow rapid, sustained burning in the presence of waves yet permit optimal aeration to ensure clean burning of oil with minimal smoke.

Four elements are required to burn oil on water: fuel, oxygen, sufficient heat to cause volatilization and a source of ignition to initiate combustion. Oil in its liquid phase does not burn, and therefore oil must be heated to a temperature to vaporize hydrocarbons to support combustion in the air above the slick. The firepoint is the temperature a few degrees above the flash point (temperature where oil can be ignited) where evaporation of hydrocarbons is sufficient to support continuous burning. Even when ignition is achieved, the temperature at the oil- water interface is never more than the boiling point of water and is usually less, and therefore the thickness of the slick is critical in maintaining a steep thermal gradient between the surface and the underlying water, contrast, a simple oil lamp or smudge pot achieves firepoint not by heating the oil reservoir, but by causing oil to rise continuously by capillary action up an absorbent where a small mass of oil encounters a high-temperature flame resulting in volatilization and combustion. Because the absorbent wick is typically also a thermal insulator, significant heat is not lost to the underlying oil reservoir and combustion is maintained until the fuel supply is depleted.

The problem of burning oil on water is somewhat more complicated than using a wick to burn a reservoir of pure oil. The hydrophobic nature of oil causes a slick to rapidly expand until it reaches equilibrium thickness determined by the oil's composition. By applying a floating absorbent to serve as a wick, spilled oil is offered an alternative avenue of escape from the water's surface. Oil accumulates on the wick, providing a thick insulating layer similar to that obtained by containing oil in booms. The oiled environment surrounding a floating wick can be viewed as a gradient where a thick oil layer accumulates immediately adjacent to the wick, thinning outward to equilibrium thickness. Therefore, once ignition is achieved, the burning area exceeds that of the wick itself and is determined by the nature of the oil and the size and absorbency of the wick. Where the thickness of a slick is sufficient, ignition of the entire surface would conceivably occur. Various materials could be used to form the wick. It is preferable that the material have some oleophilic and hydrophobic properties and be comparatively inexpensive. Oleophilic means having a maximum affinity for oils. In terms of the present invention, a material is oleophilic where it absorbs (has an oil uptake capacity of) at least about 2 grams of oil for every gram of the material (g/g). Hydrophobic is used herein to mean having a minimum affinity for water. A material will be considered hydrophobic where is has a water uptake capacity of less than 5 g/g. Cellulosic materials typically are readily available and may form the base material for the wick. Cellulosic materials will typically be derived from plants. Illustrative examples of cellulosic materials which could be used in the present invention include bagasse (the rinds of sugar cane remaining after processing), corn cobs, kenaf (a fiber crop), peanut hulls, rice hulls and rice and wheat straws. These materials will typically be dried (e.g. at 45°C for 48 hrs.) and then ground in a small hammer mill or like machinery. The ground material was then passed through a 3 -mm screen and only material smaller than 3 -mm was used. Other materials included sawdust, dried and passed through a 3-mm screen and woodchips dried and passed through a 25.4-mm screen. Newspaper and brown paper pulp can be employed by partially hydrolyzing shredded paper in a boiling solution of 0.2 NaOH. The pulp then is neutralized with HC1, dried at 45°C, and pulverized. Another material comprises various forms of cotton, including woven and braided material as well as fibers and partially hydrolyzed fibers. As seen with the experimental examples disclosed below, kenaf proves to be a preferred material from which to construct the wicks. While the above disclosure describes using cellulosic materials, it is envisioned that other materials could be used to construct the wicks. It is only necessary that these materials be buoyant (have a density less than water), be acceptably oleophilic and hydrophobic, and be able to support burning for several hours without losing structural integrity. As discussed, it is also desirable for the material to be nontoxic and biodegradable. The material used to construct the wick must be held together with some type of binding agent. A suitable binding agent should have several properties. Binding agents should maintain the structural integrity of the floating wicks for sufficient time to allow recovery after burning. Because of the likelihood that the wicks will be distributed within the environment, binding agents should be both biodegradable and food-safe. Also, because of the high-temperatures anticipated, binding agents that included cyanate or significant numbers of C-N bonds should be avoided to eliminate the possible formation of cyanide or other toxic gases during burning. Among the various agents tested were pastes of wheat and corn flour, gum arabic, and polymers of cellulose acetate and cellulose butyrate. Cellulose polymers of various densities were obtained by dissolving different amounts of cellulose acetate and ethyl cellulose butyrate in organic solvents prior to mixing with the absorbents. Commercially available latex adhesives and aliphatic emulsions (woodworking adhesives) were also tested. While these may act as acceptable binding agents, a preferred polymer comprises a crosslinked polyvinyl acetate ("PVA") such as available under the tradename Tite Bond H from Franklin International, Columbus OH and which complies with ASTM D4236 (designating the polymer as "food- safe"). This liquid formulation is both hydrophobic and oleophilic and provided good adhesion when diluted with two to three parts water. A number of experiments were undertaken to evaluate the ability of various coatings to minimize water uptake when applied to the exterior of the wicks. The beneficial effects of the more effective coatings are shown in Figure 1. Figure 1 represents the water uptake of 29 cm³ hemispheres of recycled newspaper after being coated with the substances indicated. The water uptake is measured in grams of water per gram of material (g/g). All of the coatings shown in Figure 8 are rated as "food-safe." Ethyl cellulose and cellulose acetate butyrate were suspended in toluene: acetone before application. Figure 2 shows that very little ethyl cellulose was required to obtain optimal reduction in water uptake of these small hemispheres. This polymer is used extensively in the production of cellophane and other similar biodegradable products used in food packaging. The cross-linked PNC polymer used is also rated as "food-safe" and is easily handled with no volatile emissions.

One preferred method of producing the wicks starts with providing a loose cellulosic material, such as kenaf ground to the 3mm size as discussed above. The cellulosic material will then be saturated in a binder substance such as a solution of 3:1 water to PNA. The saturated material is then pressed into the ultimate wick shape. The wicks could be formed into various shapes, including geometric shapes such as spheres, hemispheres, cones, cylinders, ellipsoids, and rectangular solids of various dimensions. When dealing with ellipsoids, such as seen in Figure 3, the size could vary from a height of about 1 cm to about 5cm with a corresponding diameters of about 2.8 cm to about 14 cm. However, one preferred embodiment comprised an ellipsoid shape having a height of about 2.5 cm and a diameter of about 7 cm. One manner of forming wicks is to place a predetermined amount of the cellulose material, still wet from mixture with the binding agent, into a mold as seen in Figure 4. The molds seen in Figure 4 are two cylindrical sections of PNC stock which have had a half ellipsoid formed into each end. The molds are then pressed together with sufficient force to insure the mixture is completely contained within the mold and the excess binding agent is pressed out of the material. After allowing a short period of time sufficient for the binding agents to "set," the wicks may be removed and placed on an appropriate surface (such as a Teflon sheet) for drying. The wicks will then be dried, preferably at about 105°C, until no further significant loss in weight occurs (typically less than 5% water content). The wicks will then be dipped in an oleophilic, hydrophobic coating substance. In a preferred embodiment, the coating substance is a 3 : 1 water to PNA solution. The dry wicks are submerged in the solution and then pulled from the solution. After dipping the wicks, it is beneficial to place them in a wire cage sufficiently flat that the wicks are fixed in place when the cage is closed. The cage is then slowly rotated until the coating becomes tacky. This rotation prevents the coating (while in the non-tacky state) from accumulating on one surface of the wick. The wicks will again be dried until weight loss ceases. For example, drying at 200 C for two hours was sufficient for the preferred wick size described above. Experimental Examples.

Several criteria were used to initially assess the performance of various designs. These criteria included: (1) uptake of oil and water while floating on an oil- water interface and during burning, (2) tolerance to burning, (3) ability to maintain structural integrity and float for 7 days, and (4) efficacy in burning a 1-mm oil slick. The methods for measuring these criteria are as follows.

1. Measuring Basic Criteria.

1.1 Uptake of oil and water. To measure the relative uptake of oil and water, wicks were weighted and then placed in a 600 ml beaker containing 400 ml tap water and sufficient South Louisiana Sweet crude oil or diesel fuel to provide a slick thickness of 2-mm. After 2 hours, the devices were removed, placed in tared 500 ml widemouth jars, and the weight after adsorption measured to determine uptake of both oil and water. To determine the amount of oil absorbed, a wick was then chopped into a number of pieces and mixed with 250 ml of tetracholorethylene

(TCE). The jar was fitted with an air-tight, teflon-lined lid, and placed in a sonifier for 2 hours.

An aliquot was quantitatively diluted for analysis of total petroleum hydrocarbons (TPH) using a Buck Scientific Model HC-404 Total Petroleum Analyzer (Westport, Conn) standardized using known dilutions of the oil used in the test. To determine the weight of the absorbent after burning, the TCE remaining after TPH analyses was decanted, the absorbent washed with 50 ml TCE, and dried in an oven 65°C for 4 hours, and then 125°C for 4 additional hours. The total amount of oil absorbed was calculated and subtracted from the weight gain after absorption to determine water uptake. The results seen in the following Table 1 represent the average uptake of water and oil by three replicate hemispheres comprised of various absorbents bonded and coated with a 3:1 mix of water and an aliphatic adhesive emulsion. The volume of the wicks was about 30 cm³.

Table 1.

Initial wt Coating Water uptake Oil uptake

Absorbent^1' (g) (g) (g) (g/g) (g) (g g) bagasse 8.00 0.57 13.3 1.7 21.5 2.7 com cobs 6.55 0.83 18.6 2.8 19.4 3.0 kenaf 6.55 1.4 16.1 2.5 17.6 3.9 recycled paper 5.29 0.33 20.8 3.9 15.5 2.9 cotton 5.30 0.32 6.9 1.3 26.5 5.0

Blend 1 7.58 0.84 14.0 1.8 21.0 2.8

Blend 2 5.78 0.81 12.9 2.2 19.4 3.4

Blend 3 5.49 0.91 20.1 3.7 17.5 3.2

Blend 4 4.91 0.72 15.74 3.2 16.0 3.3 lsd (p<0.05) 0.21 0.13 1.1 0.4 2.9 0.4 ^tBlendl : equal weights of corn cobs, recycled paper and cotton; Blend 2: equal weights of kenaf, bagasse and cotton; Blend 3 : equal weights of com cobs, bagasse and recycled paper; Blend 4: equal weights kenaf, bagasse and recycled paper.

1.2 Tolerance to burning. Excessive combustion of the absorbent material reduced the effectiveness of some materials. To test their tolerance to burning, the weight loss of the wick upon burning in a 1-mm slick was determined (see 1.1 above), h some cases excessive burning of the wick was clearly evident, and precise measurements were unnecessary. To test the capacity of the final designs to tolerate burning, three wicks were placed in a diesel slick with an initial thickness of 1ml. A peristaltic pump was adjusted to deliver 250 ml of diesel per hour. The pump was momentary switched off when it appears that the slick had developed sufficient thickness to support combustion of the entire surface. Once surface oil was partially depleted, the pump was restarted. The wicks were igmted and the experiment terminated after the combustion of 5 L of fuel. It was apparent from the burning tests that rice and peanut hulls, along with rice and wheat straws, burned to an excessive degree and would be less preferred materials for the wicks. Bagasse, com cobs, kenaf, recycled paper and cotton withstood burning sufficiently well.

1.3. Structural integrity and buoyancy. Designs that easily crumbled or disintegrated during burning were immediately rej ected. Those that demonstrated promise were floated on water for 7 days. Sinking or disintegration during that time caused the design to be rej ected. From various tests, kenaf appeared to be a preferred material. To evaluate the buoyancy of the final material, eight wicks of this material constructed as described above were placed in a water bath containing a 1-mm slick of Louisiana sweet cmde oil, and ignited. Once burning was complete, the spent wicks were left floating and placed in a fume hood for 21 days at room temperature (22- 24 °C). The number floating after various intervals was recorded. The results of this experiment are seen in Figure 5.

1.4 Efficacy in burning oil. The capacity of various wicks to bum oil was initially assessed in the laboratory using oil slicks contained in stainless steel rectangular pans (23.5 x 30.0 cm) 10- cm in height. 4.5 liters of water was placed in the pan and a volume of oil added to establish the specified slick thickness. Placing these wicks in fuel oil (diesel) slicks of >1.5 mm invariably ignited the entire slick surface, resulting in a violent fire that caused boiling over of the underlying water. A 1-mm initial slick thickness was selected for routine use not only to minimize hazards, but because it more typically represents the thickness in an unconstrained slick after a few hours on water. To determine the volume of oil resulting in a specific slick thickness, pans were initially calibrated by calculating the average increase in fluid level resulting from replicate additions of 200 ml aliquots to pans containing 4.5 liters of water at 22-23 °C. These calculations indicated that 70.5 ml of oil was needed to establish a 1-mm layer of oil floating on the water surface.

After the addition of the oil, the pans were allowed to stand in a fume hood for 30 minutes to allow the oil to evenly spread over the water surface. The window of the hood was adjusted to provide an simulated average wind speed of 17 mph. Wicks were dropped onto the oil surface and allowed to absorb oil for 5 minutes before ignition using a Bunsen burner supplied with natural gas. The bum times of the initial lighting and after additional relights were recorded. When more than one wick was used, relighting frequently occurred spontaneously when an extinguished wick collided with a burning wick. After testing, the wicks were removed and the amounts of oil retained and remaimng on the water surface measured. The amount of oil burned was calculated as the difference between the amount of oil initially present minus that recovered as absorbed or from the water. The amount burned was divided by the total time between the initial lighting until the end of the burning.

The volume of oil retained by the wicks after burning was determined by the TCE extraction procedure described in Section 1.1, except in these experiments the TPH Analyzer was calibrated using oil similar to that recovered from the water surface after burning. After removal of extinguished wicks, approximately 3 L of water was drained from the bottom of test pans. The remaining 1.5 L was transferred to a 2-L Erlenmeyer flask with a narrow neck, using a wash bottle to recover film on the sides of the pan. Water was then added to the flask to raise the oil level about half-way up the neck of the flask. Cleaning of the flask between samples with TCE was necessary to prevent beads of oil from adhering to the flooded sides of the flasks. After transfer, the flask was capped and allowed to stand for 2 hours to allow any emulsified during the transfer process to separate and rise. The flask was then placed on an adjustable j ack stand, and the oil layer withdrawn using a 50-ml beret fitted with a stainless steel blunt needle tip and a hand vacuum pump. Once all of the oil was taken into the beret, the difference between uppermost and bottom minisci of the oil column was taken as a measure of the volume of oil recovered. Careful manipulation of the j ack stand and hand pump resulted in excellent recovery and quantification of remaining oil. 10-ml samples of oil pipetted on to the water surface could be recovered with less than 2% error. The diesel fuel oil was dyed with Sudan IN to facilitate its recovery using this technique. Table 2 shows the average burning characteristics of three replicate hemispheres comprised of different absorbents and coated by dipping in a 3:1 mix of water and an aliphatic emulsion. The wicks were ignited in a 1-mm slick of diesel fuel. The volume of all devices was 30 cm³.

Table 2.

Oil retained Avg. relights Final slick Total burn after burn required thickness time Burn rate

Absorbentf ml mm min ml/h bagasse 9.6 4.0 0.35 83.9 26.0 corn cobs 8.4 2.3 0.34 66.8 34.9 kenaf 5.4 2.0 0.15 104.0 31.7 recycled paper 7.7 2.0 0.47 71.1 25.4 cotton 10.1 3.7 0.46 12.1 105.5

Blend 1 8 1.6 0.29 77.2 33.2

Blend 2 6.3 2.0 0.29 98.7 26.9

Blend 3 8.8 3.3 0.25 112.8 23.7

Blend 4 7.2 3.3 0.35 103.0 22.5 lsd (p<0.05) 3.9 0.6 0.17 19.5 12.4 Blends are the same as specified for Table 1. 1.5 Burning emulsified oil. Wind and waves can rapidly mix spilled oil and water to form an emulsion. Current guidelines do not recommend in situ burning when oil contains more than 25% water. Preliminary experiments using various coating on kenaf ellipsoids indicated that the affinity of oil for kenaf was sufficient to break oil- water emulsions provided a hydrophobic coating was applied. To assess the ability of wicks to bum emulsified oil, 400 ml of water and 100 ml of diesel or crude oil were placed in a blender and mixed at high speed for 10 minutes. The emulsion was immediately poured to pans containing 4 L water. Wicks were placed in the emulsion, and after 5 minutes, ignited with a Bunsen burner. After burning, the aqueous phase was transferred to 500 ml centrifuge bottles and centrifuged at 5000 g (gravity) for 30 minutes. A preliminary study indicated that this procedure effectively separated the oil and water, though the water phase remained slightly turbid even after prolonged centrifuging. After separation, the oil was quantified using the beret technique described above.

Fig. 6 shows that portion of oil burned in these extreme instances of emulsification. Emulsified diesel burned somewhat slower but as completely as non-emulsified diesel. A substantially smaller portion of emulsified cmde oil was burned, however. These differences may have been due to the fact that diesel emulsions more readily dissociate. A thin layer on non- emulsified diesel was observed within minutes of preparing the emulsion and adding it to a water bath. Over half of the emulsified cmde burned during these experiments, and nearly 90% of emulsified diesel burned. Since there are few highly effective tools for cleanup of emulsified oil on water, use of these wick appear to offer at least a partial solution.

1.6 Evaluation of Wick Shape

A large number of shapes and sizes were tested. Preliminary testing eliminated many inappropriate designs. Table 3 shows the results of a replicated study to determine the more effective of several promising designs. The composition and density of each design were similar. These data suggest that an ellipsoid shape with a volume of 88 cm³ was the most effective even though it required an average of at least two relightings to complete burning. Table 3 illustrates the efficacy of various shapes used to ignite and burn a 1-mm diesel oil slick. Table 3.

Oil retained Final

Volume Dry wt. after burning Relights slick thickness

Shape¹ cm³ Z Z mm cone 16.5 4.4 3.2 3.67 0.34 cone 50.0 14.6 7.8 2.67 0.36 tapered cylinder 28.3 8.3 5.5 0.67 0.33 hemisphere 27.9 7.5 4.3 2.67 0.45 hemisphere 43.2 13.1 6.7 4.33 0.24 hemisphere 72.5 17.6 9.8 2.67 0.21 sphere 139.0 13.8 14.9 1 0.19 ellipsoid 36.3 5.0 3.7 2.33 0.18 ellipsoid 88.3 13.4 6.1 3.33 0.08 ellipsoid 127.4 16.4 13.1 4.67 0.14 lsd (pθ.05) 1.1 3.4 .87 .10 ^fAll devices were comprised of kenaf using a PNA polymer as a binding agent and exterior coating. Three replicates of each design were tested.

These designs resulted in the combustion of more than 80%. of the oil initially present and an average final slick thickness of only 0.08 mm . A sphere also performed well and early studies focused primarily on spherical devices because they supported rapid and clean burning. A curved or dome shaped surface rising above the water level tends to both remove the flame from the water surface and provide a greater surface area for burning. This greater surface area allows the burning oil access to oxygen, resulting in amore complete (i.e. cleaner) bum. However, spheres have a tendency to roll, extinguishing the flame. Hemispheres generally outperformed spheres of the same diameter when the flat surface faced downward, but performed more poorly when placed with the flat surface facing upward. A simple experiment involving the random dropping of hemispheres from a height of 2 m into water showed that these objects had an equal chance of floating in either position. Because we envisioned aerial dropping of these devices, this property reduced the value of hemispherically shaped devices.

The dome-shaped surface of a sphere or hemisphere that contributes to the effectiveness of these shapes is a characteristic shared by the ellipsoid. Ellipsoids have an additional advantage in that regardless of how they are applied to an oil slick, a dome-shaped surface is ensured of floating in an upward position. The diameters of ellipsoids with volumes of ~90 cm³ and heights of 2.5 cm were only about 7 cm. in diameter. Much larger ellipsoids were tested but found to be less effective largely because they retained a greater amount of absorbed oil after burning, and because oil diffusion and oxygen availability appeared to limit the size of the flame and the efficiency of the burn. Several smaller ellipsoids burning in close proximity were capable of burning a greater percentage of oil in thin slicks than a single large device. For slicks with a thickness of more than 1.5-2 mm, size is less important because the entire slick surface is quickly ignited by both large and small devices. 1.6 Burning Slicks of Various Thickness The capacity of one preferred embodiment, 90 cm³ ellipsoid devices comprised of kenaf bonded and coated with PNA polymer, to bum slicks of various thicknesses was evaluated using diesel as a fuel source (Table 4). Diesel and related lightweight fuels are frequently spilled in marine environments, but because of the difficulty in igniting these fuels, in situ burning is not generally considered an appropriate response. Lightweight fuels disperse rapidly on the water's surface. Even when confined, applying a continuous heat source to a diesel slick will not result in the entire slick's ignition unless confined to establish a thickness of ~2mm. In comparison, mildly weathered crude oil is far easier to burn. The wicks used in the experiments reported in Table 4 were dipped in a flammable liquid prior to use. The flammable liquid could comprise several alternative light oils (e.g. shorter aliphatic chains of 8, 10 or 12 carbon atoms). One preferred flammable liquid is mineral spirits. Normally the wicks would be submerged in the flammable liquid for about 5 seconds and then allowed to drain. While other experiments show that these devices will effectively burn slicks as thin as 0.15 without a flammable starter liquid, dipping these devices in mineral spirits facilitates their ignition, accelerates bum rate, and decreases the total amount of petroleum hydrocarbons remaining on the water's surface. From 39-93%o of the diesel spilled on to the water's surface was burned. The efficiency of the bum tended to increase with increasing slick thickness. When a slick 2-mm thick was ignited, the entire surface burned as evident by the short total burn time. Even so, the slick thickness after the bum was among the least. This finding suggests that boom confinement where feasible will increase both the rate and efficiency of burning. Table 4 illustrates the ability of kenaf ellipsoids to burn diesel slicks of various thicknesses on water (at a temperature of 25 °C). Table 4.

Initial slick thickness Total burn Oil retained Slick thickness %> Burned or time after burning after bum volatilized mm min ml mm

0.25 15.6 4.9 0.08 38.9%

0.50 16.6 6.9 0.12 57.6%

0.75 23.2 1.8 0.10 83.6%

1.00 27.9 4.5 0.14 80.1%

1.50 20.9 4.8 0.19 82.5%

2.00 8.0 4.9 0.08 92.6% lsd (p<0.05) 3.2 2.8 0.05 9.8

The longest bum time (27.9 min) occurred in the 1-mm slick. With exception of the 1.5 and 2.0 mm slicks where the oil surface partially or completely burned, bum time was strongly correlated to the amount of oil burnt. This finding supports other observations suggesting that increasing the density of the devices results in a corresponding reduction in time of bum. No clear relationship between the amount of oil retained by the devices after the burn and the initial slick thickness was evident. While the water temperature of the Northern Gulf Coasts is usually moderate, water temperatures in other oceans are frigid. Temperature influences the viscosity of oil, and conceivably low water temperature could reduce the ability of oil to flow to these devices. To determine the effect of water temperature on the efficacy of burns, experiments were conducted using an initial water temperature of 5°C (Table 5). As in the experiments reported in Table 4, diesel was used as the source of fuel.

The most noteworthy effect of cold water was a lengthening of total burn time. The amount of oil retained by the ellipsoids after burning was similar to bums in warm water. The percentages of oil burned and final slick thicknesses were similar to the corresponding slick thickness burned at an initial water temperature of 25°C. These finding suggests that water temperature is not a principal determinant of whether in situ burning will be successful. Water temperature may have had a more significant effect if heavier crudes or oils had been used in place of diesel. Table 5 illustrates the ability of kenaf ellipsoids to burn diesel slicks of various thicknesses on cold water (i.e. 5°c). Table 5.

Initial slick thickness Total burn Oil retained Slick thickness % Burned or time after burning after burn volatilized mm min ml mm

0.25 32.5 4.6 0.08 38.9%

0.50 30.3 5.7 0.12 57.6%

0.75 33.0 6.3 0.10 83.6%

1.00 31.4 4.5 0.14 80.1%

1.50 31.2 4.4 0.19 82.5% lsd (p<0.05) 4.8 2.4 0.04 6.9

Burning of Heavier Oils.

Guidelines established to aid on-site coordinators in determining whether in situ burning offers an appropriate response to a spill specify that spilled oil must be weathered less than 25%. These same guidelines also require a minimum slick thickness of 2-3 mm, depending on the type of oil. An experiment was performed to determine if these devices could facilitate the ignition and burning of 1-mm thick slicks of heavily weathered South Louisiana cmde and diesel oil (Table 6). Motor oil (SAE 30) was also tested as a worst-case material, since this refined oil contains few volatile and semi- volatile compounds to support ignition and burning. Cmde weathered to reduce it's mass by 10% and 30% burned as readily as similar slicks of unweathered cmde. However, weathered diesel did not burn as completely as did unweathered diesel. Final slick thicknesses for weathered diesel ranged from 0.12 to 0.21 mm, though no differences in final thickness were observed between diesel weathered to 10% and 30%). Motor oil burned poorly and with considerable emission of soot. Nevertheless, over 36% of motor oil applied as a thin 1-mm slick was consumed. A slick with sufficient thickness to support surface combustion conceivably would have elevated temperatures and resulted in a more satisfactory bum of the longer-chained hydrocarbons in motor oil. Table 6 illustrates the ability of kenaf ellipsoids to bum motor oil and weathered cmde and diesel oil. Table 6.

Fuel source Total burn Oil retained Slick thickness % Burned or time after burning after burn volatilized min ml mm

Crude 10% weathered 36.3 5.9 0.19 72.0%

Diesel 10% weathered 37.1 5.7 0.07 84.5%

Crude 30% weathered 34.1 6.5 0.14 76.0%

Crude 30% weathered 33.2 7.1 0.09 81.1%

Motor oil 16.6 9.8 0.48 36.6%

(SAE 30)

1.8 Maximum Capaci tv to Burn Oil

A simple experiment was performed to determine the capacity of 90 cm³ devices to burn cmde oil. The wicks tested were comprised of kenaf, and bound and coated with a PNA polymer. Three of these devices were placed in 1-mm thick slick of diesel and ignited. A pump then was adjusted to deliver a continuous supply of fuel to maintain a slick sufficiently thin to avoid ignition of the entire surface. Under these conditions, no relighting of the wicks was required. This observation supports the suggestion that the need to relight is the result of reduced oil diffusion across a weakening gradient within the devices as the surface oil is depleted. The experiment was terminated after combustion of 5 L of diesel because the devices showed no significant signs of deterioration. After extraction of oil and drying, the wicks appeared to have lost less than 1.5 g, or less than 8% of their initial weight. This loss is only slightly greater than the 0.7-0.8 g loss observed when these devices are used to bum only 70-100 ml of oil. This finding suggests that devices of kenaf have a very large capacity to burn oil even when the entire slick surface is not ignited. 1.9 Potential Applications

The present invention has utility in many circumstances where traditional in situ burning was not possible or not practical. As one example, the rapid spreading of oil on water complicates traditional approaches to in situ marine burning. However, the fact that within a few hours most slicks will not support sustained combustion suggests a technique for burning slicks in close proximity to a damaged vessel or other object. If wicks are used in this situation, the fire is limited to the wick surface and does not spread to the slick as a whole. To explore this possibility, a demonstration was conducted using a small (3.7 m x 3.7 m) pool constmcted for this purpose. Two gallons of South Louisiana cmde were poured onto the water's surface and a small plastic boat anchored nearby. The purpose of this demonstration was to determine whether the oil could be burnt without endangering the boat. When the demonstration was carried out, the majority of the oil was absorbed onto the wicks and burnt without damage to the boat.

When blowouts and other inland spills occur, rainfall often causes free-phase oil on vegetation and soils to migrate into small slow moving or stagnant tributaries where its recovery is difficult. A demonstration was conducted to determine if these wick devices could be used to address such conditions. Two quarts of Louisiana Sweet Cmde were introduced into a small tributary. Several wicks were placed on the small slick and the wicks ignited. A significant maj ority of the oil was removed by burning. The most significant problem was keeping the wicks in contact with the oil as the slick was burnt and/or dispersed. One potential solution could be using a small boom or blower to consolidate oil for burning. When a wick floats into an area where it threatens to catch vegetation or shore debris on fire, the wick could easily be extinguished by pushing it below the water surface. When it resurfaced, the wick would be relit by moving it adjacent to a burning wick. An interesting finding of these demonstrations was that these wicks need to only come into contact with oil a small percentage of the time to maintain the bum. When in contact, the wick is quick recharged with oil for several more minutes of burning.

When spills occur near shore, rapid containment of oil with booms to prevent spreading is a principal priority. However, not all environments and conditions permit rapid deployment of equipment to recover and store the spilled oil once it is contained. Figure 7 diagrams a simple technique employing a floating "cage" 14 to contain these wicks and to permit burning of spilled oil 10 using conventional booms 12. After confirming that the uppermost edge of the contained slick is not sufficiently thick to support combustion of the entire oil surface, the floating cage 14 could be deployed, filled with wick devices and maneuvered via cables 16 into the slick where it is ignited. As the slick is consumed, the cage 14 maybe repositioned to maintain contact with the oil 10.

A simulation was conducted to test the viability of this concept and to determine the rate at which these wicks could burn diesel oil when contained by a "cage." Using a 950 cm² (~1 sq. ft) cage containing 16 wicks, diesel fuel was metered onto a 2560 cm² surface using a peristaltic pump. The rate was adjusted to obtain a steady rate of burn within the cage while avoiding a build up of slick thickness. If an excessive rate was used, the thickness of the slick would increase to a point where the entire surface would combust, emitting large volumes of smoke and condensate. A flow rate of 50 ml diesel min maintained constant combustion. When the burn was maintained at this rate, oil supply rather than oxygen appeared to limit combustion rate and very little particulate emission was observed. Oil addition was terminated after 5 L of diesel were added. After the burn, only 34 ml of unburnt oil were recovered from the water's surface. These findings suggest that each square foot occupied by these wicks will burn about 0.8 of oil gallons per hour. However, this small scale study should not be viewed as a prediction of the behavior of burns of several hundred or several thousand square feet. Oil diffusion rates, temperatures, winds generated by the fire and oxygen supply are likely to be significantly different when burning large areas.

The present invention provides a novel and effective method for in situ burning of oil spills on water. It may seem counter-intuitive that devices only 7 cm in diameter can be effective in burning oil slicks on the open sea. Certainly when winds are sufficient to cause blowing spray and large, breaking waves, successful burning of an oil slick with these devices would seem difficult as would any other method of oil cleanup. However, the presence of oil on the surface tends to smooth the sea and prevent waves from breaking. There is no reason to doubt these buoyant, small devices would be effective in sea conditions that can support an oil slick at the surface, including large swells. Compared to a typical oil slick of less than 1-mm, these devices are large. Because these devices are inexpensive to manufacture, light-weight and easily handled, large numbers could be deployed to cause widespread burning when rapid oil removal is mandated. While the foregoing description has been of specific embodiments of the present invention, it will be understood that numerous alternate embodiments will be apparent to those skilled in the art and such alternate embodiments are intended to come within the scope of the following claims.

Claims

CLAIMSI claim:

1. A method for in-situ burning of liquid hydrocarbons on water comprising the steps of: a. providing a plurality of individual floating wicks; b. positioning said wicks in a area of liquid hydrocarbons floating on a body of water; c. igniting said wicks.

2. The method of claim 1, wherein said step of providing wicks includes providing wicks with a height of between about 1 cm and about 5 cm.

3. The method of claim 2, wherein said step of providing wicks includes providing wicks with a height of about 2.5 cm.

4. The method of claim 1, wherein said step of providing wicks includes providing wicks having a dome shaped top portion.

5. The method of claim 4, wherein said step of providing wicks includes providing wicks having an elliptical shape.

6. The method of claim 1, wherein said step of providing wicks includes providing wicks which are formed of a cellulosic material.

7 The method of claim 6, wherein said step of providing wicks includes providing wicks wherein said cellulosic material is one of the group comprising bagasse, com cob, kenaf, recycled paper, or cotton.

8. The method of claim 1, wherein said step of providing wicks includes providing wicks which are formed of a non-toxic, biodegradable material.

9. The method of claim 1, wherein said step of providing wicks includes providing wicks which are formed of an oleophilic material.

10. The method of claim 9, wherein said step of providing wicks includes providing wicks wherein said oleophilic material has an oil uptake capacity of at least about 2 g/g.

11. The method of claim 9, wherein said step of providing wicks includes providing wicks with a water uptake capacity of no more than about 5 g/g.

12. The method of claim 10, wherein said step of providing wicks includes providing wicks with a water uptake capacity of no more than about 5 g/g.

13. The method of claim 1, wherein said step of providing wicks includes providing wicks which have a coating applied thereto which is both hydrophobic and oleophilic.

14. The method of claim 13, wherein said step of providing wicks includes providing wicks wherein said coating is cross-linked polyvinyl acetate.

15. The method of claim 1 , wherein said step of providing wicks includes coating said wicks with a flammable liquid.

16. The method of claim 13, wherein said step of providing wicks includes coating said wicks with a flammable liquid.

17. The method of claim 15, wherein said step of providing wicks includes coating said wicks with a flammable liquid comprising a light oil.

18. A method for manufacturing a floating wick comprising the steps of: a. providing a wick material having a dry density less than water; b. applying to said wick material an oleophilic, hydrophobic coating such that said coated wick material has an oil uptake capacity of at least 2 g/g and a water uptake capacity of no greater than 5 g/g; and c. drying said wick material.

19. The method of claim 18, wherein said step of providing wick a material includes providing a loose wick material and further including the steps of: d. applying a binder substance to said loose wick material; e. molding said loose wick material into a predefined shape; and f. drying said molded wick material prior to applying said oleophilic, hydrophobic coating.

20. The method of claim 19, wherein said step of providing loose wick material includes providing a cellulosic material.

21. The method of claim 18, wherein said step of applying an oleophilic, hydrophobic coating includes applying a coating comprising a polyvinyl acetate solution.

22. The method of claim 19, wherein said step of applying said binder substance includes applying a polyvinyl acetate solution.