COMBUSTION OF WATER-IN-OIL EMULSIONS OF DIESEL AND FRESH AND WEATHERED CRUDE OILS FLOATING ON WATER

• Main Author: Walavalkar, Ajey Y.
• Other Authors: Dr. Anil K. Kulkarni, Dr. Savash Yavuzkurt, Dr. Ralph Webb, Dr. Sarma Pisupati
• Format: Text
• Language: English
• Published: Penn State 2001
• Subjects: Mechanical Engineering; Milne Point; Alaska North Slope; north slope; Alaska
• Online Access: http://etda.libraries.psu.edu/theses/approved/WorldWideIndex/ETD-52/index.html

Burning of spilled oil on the ocean surface or in-situ combustion is one of the techniques used in oil spill clean up operations. With the passage of time, as the crude oil stays on the ocean surface, evaporation of the lighter components of the crude oil and mixing of water with the oil due to wind and ocean turbulence create emulsions that are more dense and more viscous than oil. This makes the ignition of the oil harder to achieve. Prior studies have shown that emulsions containing more than a certain fraction of water do not burn, and thus present a difficulty in applying the in-situ combustion technique. Many normally incombustible materials can be ignited when subjected to a certain minimum heat flux, and sustained fire and flame spread can be achieved on these materials. In the present work, this principle is applied to the emulsion combustion problem so that, if successful, the window of opportunity for in-situ combustion of oil spills can be widened. It is proposed that there exists a threshold heat flux level for most of the emulsions such that, when emulsion is exposed to a heat flux equal to or greater than the threshold heat flux, sustained fire and flame spread can be achieved. A test facility was designed and built to conduct laboratory scale burn tests of water-in-oil emulsions to study the threshold heat flux values for different types of emulsions. The emulsion samples, floating on water surface, could be subjected to radiative heat flux ranging from 1.2 kW/m2 to 21 kW/m2 in steps of about 1 kW/m2. Experimental measurements of threshold heat flux values were made for emulsions of diesel, Milne Point crude oil and Alaska North Slope crude oil with water. The water fraction in the emulsion was changed from 0% to 80% by volume. The crude oil samples were also tested for effects of evaporation of the lighter fractions from the crude oil. Laboratory scale experiments clearly verified that there exists a threshold heat flux for each type of emulsion studied. For example, emulsions of fresh Milne Point crude oil containing 35% water by volume could not be ignited with external heat flux of 11 kW/m2. But the same emulsion composition could be ignited with external heat flux of 13 kW/m2. 15% weathered Milne Point crude oil could not be burned, in unemulsified state, without external heat flux of 10 kW/m2. Emulsions of 26 % weathered Alaska North Slope crude oil containing 50% water by volume could be successfully burned with the help of external heat flux of 9 kW/m2. The data indicate that higher threshold heat flux is required to cause successful burning of emulsions having higher water content or emulsions of more weathered oil. Upon correlating the threshold heat flux data for the crude oils with the density of the crude oil, it was observed that the threshold heat flux values increase with increasing density of the crude oil. Previous studies have proposed that it is the oil that separates from the emulsion that burns and not the emulsion itself. Combining this idea with the results presented, it can be argued that in order to create a sustained emulsion pool fire, the emulsion must be separated into water and oil at such a rate that the rate at which the oil is separated from the emulsion must be at least equal to the rate at which the oil is consumed by combustion. The results from the study also suggest that in order to achieve flame spread, the fire must be of such size as to impose the surrounding emulsion pool with heat flux equal to or more than the threshold heat flux. The combustion process of a water-in-oil emulsion layer floating on top of a water surface, as in case of in-situ burning of oil, spilled at sea that has turned into emulsion, was modeled using comprehensive mathematical treatment. Mathematical models are available in the literature for oil pool fires but not for emulsion pool fires. This model is unique as it incorporates the separation of emulsion into water and oil, a phenomenon that governs the emulsion pool fires. The burning process is divided into three regimes, as follows. 1. The initial regime, when the emulsion layer floating on the ocean surface receives heat flux from an external source such as an igniter or a burning oil pool; 2. The intermediate regime, from the instant when there is first appearance of oil layer on the top of the emulsion layer due to breaking of emulsion until the oil starts evaporating; and, 3. The final regime, which is characterized by the combustion of oil vapor and continues till the fire extinguishes. The model was solved numerically using finite difference method and it was validated using the data from the laboratory scale burn experiments involving emulsions of diesel with water. The model predictions were used to study the effects of emulsion composition and oil weathering on important emulsion pool fire characteristics such as, the average burning rate, total duration of the burn, the volume of the oil residue and the overall oil burning efficiency. The average oil burning rate, total duration of burn, the volume of the oil residue and the burn efficiency decreased with increasing water content of the emulsion. Average oil burn rate and the overall burn efficiency decreased with increased weathering of the oil whereas the total burn time and the volume of the oil residue increased with increased weathering of the oil. Comparisons of the model predictions with the experimental observations showed that most of the model predictions were found to be within 25% of the observed values. The model has captured the description of significant processes involved in emulsion combustion, and thus is able to describe the experimental observations with sufficient accuracy.