Deep Sea Oil Trap: An Underwater Cleanup System for Platforms and Ships to Contain and Recapture Oil Plumes



By: Sachin Narayan


Underwater oil plumes are an accumulation of oil microdroplets that, due to a number of factors (namely size, temperature, pressure, and currents), do not rise to the surface (Zhao, 2016). This research aims to better understand underwater oil plume dynamics and forces and test reusable materials and processes that effectively recapture the oil without causing further environmental damage. The work concluded that a two-part system involving containment and recapture provides an effective way to deal with this problem without excessive energy use. The first part, containment, involves enclosing the oil with a net coated with a fluorocarbon-based compound for oleophobicity. The second part, recapture, involves magnetizing the oil using a magnetite, iron (II,III) oxide, powder. The proposed design has potential to be used as both a preemptive measure at drilling sites as well as a mobile cleanup practice following an oil spill. Currently, other practices addressing deep sea oil plumes either introduce thousands of highly toxic chemicals into the ecosystem or require immense amounts of energy (Torrice, 2010). The experimental system, with an average 78.41% recapture rate, is an effective proof of concept, prompting further research and enabling potential industrial implementation of the system. Future improvements to this system should entail better dispersion methods of magnetite and use of a recently developed superoleophobic coating material with no water repellant properties (Gordon, 2015).


Offshore oil spills can be devastating to the environment and economy. Following disasters like the Deepwater Horizon and Exxon Valdez spills, millions of gallons of oil spilled into the surrounding waters, wiping out entire native marine populations (CNN Library). The Deepwater Horizon Spill is worth around $96 billion itself and that number continues to rise with the long-term impact on the fishing and tourism industry (Amadeo 2017). While most oil rises to the surface, a significant portion of oil spilled at depths remains submerged and forms massive underwater oil plumes (Figure 1). These oil plumes severely damage marine ecosystems by killing organisms ranging from larger animals such as seals and ducks to microscopic creatures such as phytoplankton which comprise the first level of the food chain (“Effects of oil on marine wildlife”).  Current cleanup practices inadequately address this problem, either requiring excessive energy or leaving the environment to cope by letting bacteria degrade the massive amounts of oil (Dell’Amore, 2010). Apart from allowing a threat to marine life to persist, the main drawback of letting bacteria deal with the oil is that it triggers a process of deoxygenation of the water. An abnormally high concentration of microbes in one area around the oil depletes the oxygen supply of the surrounding water, making the environment uninhabitable for a number of years to come (Schrope, 2010). Along with its natural formation, oil plumes also form due to the use of dispersants like Corexit, which break up surface oil into smaller droplets, allowing them to sink (Lovett, 2011). During 2010, the BP spill’s main oil plume measured around 35 kilometers long and 200 meters wide, spreading for months at depths of around 1,100 meters (Dell’Amore, 2010). It took months for scientists to detect its presence and by that time, it had impacted thousands of square miles in the Gulf. The movement of these plumes and their impact are not entirely understood, but initial studies of Gulf of Mexico in the years after 2010 lead many experts to predict potential long-term ecosystem damage.

Underwater Oil Dynamics and Plume Formation–

Unlike the majority of oil which rises in water, some droplets remain submerged due to their small size. Water deeper in the sea is much colder and saltier which results in a higher density than water closer to the surface (Lovett, 2011). Oil microdroplets end up intermixing with this deep sea water and as a result, rise very slowly with a higher density. At a point in the water column, the rising oil droplets’ density matches that of the surrounding water and thus, the oil remains suspended (Lovett, 2011). Following high pressure exertion from an underwater leak, oil can break up into 1 mm – 10 mm droplets which, depending on salinity, temperature, and currents, would rise to the surface over a period of hours or days (North, 2015). This allows for a movement of these droplets into surrounding areas causing further environmental damage. Still, they rise relatively quickly through the water column. The widespread use of dispersants worsens the problem, decreasing droplet size to 1000-5000 µm which can persist at depths of over 3,000 feet for months (North, 2015). Additionally, drilling sites extract more than just oil, as a number of gases and extraneous chemicals also exit the seafloor. Mixtures of oil and other substances cause changes in density which can further hinder the ability of the oil to rise up to the surface (North, 2015). The size of these droplets has presented many challenges to underwater cleanup practices as well as the ability to contain the movement of the plume.

Existing Cleanup Methods–

There are a number of methods currently used to contain and recapture oil on the surface which could also be applicable to underwater situations. The most common tactic is using oleophilic absorbents which attract and collect oil. Common absorbents include synthetic fibers, plant material, and hair. Currently, absorbents are most commonly used in oil booms which are cylindrical bundles of a particular oleophilic material that surrounds a surface oil slick (Adebajo, 2003). These booms contain oil to a particular area, making cleanup easier. The major drawback for its underwater usage is that once the material soaks up oil, it cannot be reused and, thus, the process is very inefficient.

Another tactic involves a vacuum system which sucks up all surrounding liquids and transports them to a factory for the separation of water and oil. This technique can be used on the surface and underwater, but due to its high energy consumption and inefficiency, it is not widely used.

Other relatively new and uncommon practices for surface cleaning include oleophobic materials, meaning oil-repellant, and ferromagnetism which involves the attempt at magnetizing oil. The advantage of these practices is their reusability because contact with oil does not degrade their effectiveness over time. Oleophobicity can be used to enclose oil, but its other uses are limited. The idea of magnetizing oil is relatively new, suggested by a physicist named Arden Warner who devised a system of electromagnets to collect oil on the surface by intermixing the droplets with iron (Gertz, 2014). Both these processes demonstrate good potential for usage underwater compared to other common practices.


This experiment aims to understand underwater oil dynamics and to evaluate the two sustainable methods of oleophobicity and ferromagnetism in their ability to contain and recapture submerged oil droplets. Oleophobicity and ferromagnetism together form a two-part system that first works to limit the movement and environmental impact of an oil plume. Following this containment, the system then magnetically constricts and recaptures the oil, extracting it from the deep sea ecosystem. This experiment evaluates both methods individually and in conjunction with one another to best evaluate how the two-part system would work when implemented in its entirety.


The total system was estimated to have an overall 70% effectiveness. Due to the fact that there are very few other underwater cleaning systems, there is not an established baseline that helps determine whether a proposed system or method is effective. Additionally, deep sea methods cannot be fairly compared to surface cleaning methods which have been heavily researched and do not face the challenges of the underwater environment. The 70% estimation was based on both the quality of the hydrophobic and oleophobic commercial coating product and initial visual evaluations of magnetizing oil. Furthermore, a removal of 70% of oil from a given area would decrease the size of a plume to a point where microbes could break down the oil without causing deoxygenation of the waters.

The individual parts were given estimations based on initial visual evaluations, commercial product quality, and the understanding that the additive effect of both parts would yield 70%+ effectiveness for the full system. Part 1, the oleophobic containment system, was estimated to have a 50% effectiveness. Although coating materials have already been tested for their oleophobicity, the coating compound was estimated with a relatively moderate efficiency because of the unknown effectiveness of commercial products underwater. Part 2, the ferromagnetic recapture system, was also estimated to have a 50% effectiveness due to the initial visual evaluation and its unprecedented use underwater. A potential challenge that arose with its use underwater is the simple challenge of getting the magnetite powder to coat the oil droplets because the powder and droplets’ movements are entirely dependent on underwater currents.

The research is designed as a proof of concept experiment to evaluate the system’s potential for underwater use. Its success in this small-scale experiment should prompt further research and development on a larger scale.


Full System–

  • 20 gallon tank
  • 8 gallons of water
  • 800 g sea salt
  • 200 mL Heavy Aromatic-Naphthenic Crude Oil
  • Centrifuge
  • 12 mL syringes

Part 1: Containment Net–

  • Ultra Ever-Dry Oleophobic Coating (commercially available)
  • Fish Tank Cleaning Net
    • 15 cm x 12 cm frame, Flow Rate ? 3 L/min

Part 2: Magnetic Recapture–

  • 25 grams iron (II,III) oxide
    • Magnetite/Fe3O4, 50-100 nm particles
  • Neodymium Magnet


Tank Setup–

  1. Dissolve 752.72 grams of sea salt into 8 gallons of water
  2. Pour saltwater into 20 gallon tank

Part 1: Containment–

  1. Coat the cleaning net by spraying it with the Ultra-Ever Dry Top and Bottom Coats
  2. Weigh the coated net
  3. Create a solution of 4 mL of oil and 6 mL of water in a syringe
  4. Shake syringe vigorously to separate oil into smaller bubbles
  5. Submerge coated net with its opening facing down
  6. Inject oil and water mixture into the mouth of the net from below its frame. The injection of the oil is designed to keep the oil within the confines of the net mimicking how the net would surround oil plume during real use
  7. Hold the net underwater for 30 seconds and then scoop up the water within the vicinity of the net
  8. Let the water drain out the side of the net as the oil beads up on the oleophobic surface
  9. Leave the net to dry, then weigh the net and record the difference between the initial weight and weight after the test. This difference in weight equals the captured amount of oil by the containment net.
  10. Repeat procedure for 4 trials

Oleophobic Coating Compounds–

Oleophobicity, otherwise known as lipophobicity, is a unique property of a handful of materials. Most coating systems are hydrophobic, meaning water repellant. Unlike these hydrophobic coatings, oleophobic materials repel hydrocarbons like oil. The specific class of materials that possess this property are mostly fluorinated materials, especially fluorocarbons. When fluorine, a highly electronegative element, bonds with carbon, it forms a very dipolar bond as a result of fluorine’s greater electron density. Once these dipolar bonds are constructed into a chain (Figure 2), the molecule forms an intensely negatively charged field, repelling foreign compounds including hydrocarbons (Hanson, 2015).

The effectiveness of an oleophobic compound is also dependent on the medium on which it coats. The type of mesh (metal or fiber), the integrity of the material, and the pore sizes all play big roles in how effectively the negatively charged field of the fluorocarbons can cover the entirety of the net’s surface area. Due to the extremely small droplet size of submerged oil, the net also needs to have extremely small pores (around 0.5 mm or less) . While the oil is not let out of the vicinity of the net, water should be able to escape through the pores, relieving much of the water pressure on the system that generates from a growing or spreading oil plume.

Part 2: Recapture (includes total system measurements)–

  1. Create a solution of 0.5 g magnetite and 5 mL water in a syringe
  2. Create a solution of 4 mL oil and 6 mL water in a syringe
  3. Shake oil and water syringe vigorously to separate oil into smaller bubbles
  4. Place magnet in the bottom of a glass tube
  5. Submerge the net underwater
  6. Inject the magnetite plume into the vicinity of the containment net (Figure 4)
  7. Immediately after, inject the oil solution into the vicinity of the net and magnetite plume
  8. Hold tube with magnet within the vicinity of the net for 30 seconds to collect up oil and magnetite
  9. Heat tube and scrape off oil, water, and magnetite into a centrifuge tube (Figure 3)
  10. Spin the recovered solution in the centrifuge on for 30 minutes at 17,000 rpm
  11. Drain out water and use magnet to extract magnetite to leave oil in the tube for weighing
  12. Repeat procedure for 4 trials

Ferromagnetism and Van der Waals Forces–

Ferromagnetism has potential to be used in underwater cleanup. Ferromagnetic objects have a high susceptibility to magnetic fields, aligning and attracting themselves towards a magnet. Oil, being a hydrocarbon, is not naturally magnetic, but when mixed with ferromagnetic particles, oil exhibits magnetic characteristics. Physicist Arden Warner originally presented this idea and demonstrated its potential for surface cleaning, but its applicability to underwater cleaning is unknown (Gertz, 2014). When oil and magnetite intermix, their interaction is not defined by strong ionic or covalent bonds, but rather by a force known as London Dispersion that exists between all molecules (Warner, 2014). This Van der Waals force results in a weak bond between the magnetite nanoparticles and the oil microdroplets. Thus, when a magnetic field is introduced and the magnetite particles attract, the weak bond drags the connected oil droplets towards the magnet as well (Figure 5). When dispersed underwater, the magnetite nanoparticles form a cloud that slowly settles down onto the underwater plume, priming the oil for attraction. In addition, the use of magnetite particles presents a number of environmental advantages. Unlike the harmful chemicals in dispersants, iron (II,III) oxide is a naturally found compound already present in the marine ecosystem. It is also widely available in many locations around the world, making the process feasible and environmentally-friendly.

Analysis / Observations–

On average, the first part of the system, the containment net, surpassed the expected percentage as predicted in the hypothesis by around 10% (standard deviation 3.87%). The second part of the system, the magnetite recapture, collected around 6% less of the oil than expected (standard deviation 0.52%). The overall system, combining the first and second part, surpassed its expected percentage by about 8% (standard deviation 0.77%). In all of the trials, relatively small amounts of crude oil, around 4 mL, were injected to form the underwater plume with the intention of limiting the amount of oil able to escape the net’s vicinity. The ratio of magnetite to oil injected was 1:8 in order to mimic an ideal situation in which the cleanup process uses substantially less magnetite than the amount of oil in the plume.

For the containment net tests, the oleophobic compound performed remarkably well, but some challenges still existed that could have impacted the accuracy of the tests. In order to account for oil that would stick to the injection syringe by lining the insides with oil, the calculations involve finding the mass of the syringe with the oil inside it before and after the experiment takes place. The difference in these masses reflects the amount of oil that was injected into the water. Following, the mass was converted into mL by the density of the specific heavy aromatic-naphthenic crude oil used in the experiment (0.94 g/mL). Similarly, the oil contained by the net beads up on the oleophobic surface. The calculations of the amount of this recovered oil comes from the difference between the mass of the coated net before the experiment and after. The difference would reflect the amount of oil contained by the plume and again, using the density, the mass would be converted to volume (mL). In both the weighing of the injection syringe and the net, sufficient time was given for the water to dry off, ensuring that only oil remained either lining the inside of the syringe or beaded up on the net’s surface.

In a similar manner to the testing in Part 1, the calculations involved the amount of oil remaining inside the injection syringe by subtracting the initial mass from the final mass. The process of separating the oil and magnetite involved a number of steps, increasing the chance of experimental error. After mixing the oil, water, and magnetite mixture in the centrifuge, in order to extract the water and magnetite from the tube, the experiment involved coating the sides of the tube with the oil. This would allow the water to drain and the magnetite to be attracted out of the tube, thus only leaving the oil in the tube, which can be weighed and converted to mL. The process of filtering out water and iron (II, III) oxide from the tube could possibly have taken some oil with it as well, thus affecting the amount of oil recaptured.

The total system test involved combining both parts and testing their efficiency. The same processes to determine the amount of captured oil for each method were repeated and the total recovered oil is equivalent to the sum from each part. The increased total effectiveness compared to that of the individual parts shows how the two parts work in tandem to improve the efficiency of the overall system. Yet, around 20% of the oil injected still escaped both the containment net as well as the magnetite recapture.

In each of the three sections tested, the amount of oil injected into the tank on different trials varied in an effort to understand if changes in the amount of oil affected the system’s effectiveness. Still, the amount of oil injected was still kept below 12 mL to prevent the buildup of large droplets of oil that would rise immediately and do not mimic the microdroplets that make up underwater plumes. Additionally, keeping the amount of oil relatively small minimized the chances for human error such as oil being injected outside of the net’s vicinity.


Overall, the results of testing indicate a successful proof of concept and possible future research into large-scale application. The results for each section were different than the hypothesized values, underperforming and overperforming in different cases. Both the containment section as well as the full system exceeded expectations by yielding 60.70% and 78.43% effectiveness, respectively. These values surpassed the hypothesized values by around 10%, suggesting that the use of oleophobicity and the two methods together is an effective method for dealing with underwater plumes. On the other hand, the magnetite recapture section performed poorer than expected, yielding a 43.79% effectiveness. This value, around 6% below the expected amount, likely resulted from a lack of intermixing between the magnetite particles and oil droplets, and it demonstrates the challenges of using a coating system underwater. In all three areas tested, the hypothesis inaccurately predicted the efficiencies, but did accurately reflect the additive impact of the oleophobic net and magnetic recapture by correctly predicting an increased result from the total system.

In order to improve the efficiency of the magnetic recapture, one could increase the ratio of magnetite to oil injected underwater, yet this would not be ideal because the system attempts to maximize its recapture while using minimal amounts of the magnetic powder. A better alternative would be devise an improved way to distribute the magnetite cloud evenly over the oil plume. Thus, with the same amount of magnetite, a larger surface area of the plume would be covered and magnetized.

In this experiment, a few important factors were not accounted for due to the constraints of testing in a lab. These factors include deep-sea pressure and underwater currents which could affect the integrity of the containment net. Further testing would need to be done on a larger scale to closer mimic a deep-sea environment and to evaluate situations when a growing oil plume exerts physical pressure on the net.

Unlike current methods treating the issue of underwater oil plumes, this two-part system is efficient and does not require excessive amounts of energy. Both the oleophobicity and ferromagnetism work in tandem, not only preventing the further spread of these submerged oil microdroplets, but also by recapturing the oil in an environmentally-friendly way. Furthermore, the oil cannot degrade the integrity of the oleophobic net as it does not come in contact with the mesh. The magnetite can also be separated from the oil after cleanup, allowing it to be reused in further cleanup efforts.

Future Research and Application–

For future improvements and application, manufactures could coat the net with a substance that is super-oleophobic while having no interaction with water. The commercial product used for this experiment, the Ultra Ever-Dry coating, was both oleophobic and hydrophobic, which can present problems on a large scale. A coating without any interaction with water would theoretically allow the net to pass water through while enclosing oil on one side. This would reduce the pressure and drag put on the system at such depths. The University of Wisconsin has recently developed a material that fits this criteria for underwater use and use of this product would likely increase the efficiency of the system (Gordon, 2015). During future research, investigation should be done on the mesh size needed at depths, accounting for both oil containment and deep-sea material that will potentially contact the net. Additionally, an AC electromagnet can be used at such depths to attract in magnetized oil. Further research is needed regarding its functionality for this purpose.


This research was made possible by the Schmahl Science Workshops ASRP, Belinda Schmahl, and mentor Sean Carroll. Special thanks to Bellarmine College Preparatory and chemistry professor Debjani Roy, Ph.D. for their support.


  1. Adcroft, Alistair, Robert Hallberg, John P. Dunne, Bonita L. Samuels, J. A. Galt, Christopher H. Barker, and Debra Payton. “Simulations of Underwater Plumes of Dissolved Oil in the Gulf of Mexico.” Geophysical Research Letters 37.18 (2010): n. pag. Web.
  2. Adebajo, M.O., R.L. Frost, J.T. Kloprogge, O. Carmody, S. Kokot. “Porous Materials for Oil Spill Cleanup: A Review of Synthesis and Absorbing Properties.” Journal of Porous Materials 10 (2003): 159-170. Web.
  3. Amadeo, Kimberly. “BP Spent $56.4 Billion on Spill So Far.” The Balance. N.p., 18 July 2017. Web. <>.
  4. Barron, Andrew R. “Compounds of Fluorine.” OpenStax CNX. N.p., 11 Feb. 2010. Web. <>.
  5. Brown, P. S., O. D. L. A. Atkinson, and J. P. S. Badyal. “Ultrafast Oleophobic–Hydrophilic Switching Surfaces for Antifogging, Self-Cleaning, and Oil–Water Separation.” ACS Applied Materials & Interfaces 6.10 (2014): 7504-511. Web.
  6. Dell’Amore, Christine. “Giant Underwater Plume Confirmed-Gulf Oil Not Degrading.” National Geographic. National Geographic Society, 20 Aug. 2011. Web. <>.
  7. “Effects of Oil on Marine Wildlife.” Global Marine Oil Pollution Information Gateway. UNEP GPA, n.d. Web. <>.
  8. Gertz, Emily. “To Clean Up Oil Spills, Magnetize The Oil First.” Popular Science. N.p., 7 Aug. 2014. Web. <>.
  9. Ge, Jin, Yin-Dong Ye, Hong-Bin Yao, Xi Zhu, Xu Wang, Liang Wu, Jin-Long Wang, Hang Ding, Ni Yong, Ling-Hui He, and Shu-Hong Yu. “Pumping through Porous Hydrophobic/Oleophilic Materials: An Alternative Technology for Oil Spill Remediation.” Angewandte Chemie 126.14 (2014): 3686-690. Web.
  10. Gordon, Scott. “New Materials Repel Oil Underwater, Could Better Clean up Oil Spills.” News. University of Wisconsin-Madison, 15 Apr. 2015. Web. <>.
  11. Hanson, Eric, Ph.D. “Ask the Expert: What Makes a Nanocoating Oleophobic?” YouTube. Aculon, 07 Oct. 2015. Web. <>.
  12. Jung, Yong Chae, and Bharat Bhushan. “Wetting Behavior of Water and Oil Droplets in Three-Phase Interfaces for Hydrophobicity/philicity and Oleophobicity/philicity.” Langmuir 25.24 (2009): 14165-4173. Web.
  13. Kota, Arun K., Gibum Kwon, Wonjae Choi, Joseph M. Mabry, and Anish Tuteja. “Hygro-responsive Membranes for Effective Oil–water Separation.” Nature Communications 3 (2012): 1025. Web.
  14. Leggett, Martin. “‘Deepwater’ Oil Snatched from Depths Helps to Map Lurking Oil Plumes.” Earth Times. N.p., 18 July 2011. Web. <>.
  15. Lin, Ling, Mingjie Liu, Li Chen, Peipei Chen, Jie Ma, Dong Han, and Lei Jiang. “Bio-Inspired Hierarchical Macromolecule-Nanoclay Hydrogels for Robust Underwater Superoleophobicity.” Advanced Materials 22.43 (2010): 4826-830. Web.
  16. Lovett, Richard A. “Why Did Huge Oil Plumes Form After the Gulf Spill?” National Geographic. National Geographic Society, 20 Apr. 2011. Web. <>.
  17. North, Elizabeth W., E. Eric Adams, Anne E. Thessen, Zachary Schlag, Ruoying He, Scott A. Socolofsky, Stephen M. Masutani, and Scott D. Peckham. “The Influence of Droplet Size and Biodegradation on the Transport of Subsurface Oil Droplets during TheDeepwater Horizonspill: A Model Sensitivity Study.” Environmental Research Letters 10.2 (2015): 024016. Web.
  18. “Oil Spills Fast Facts.” CNN Library. Cable News Network, 09 Feb. 2017. Web. <>.
  19. Schrope, Mark. “Oil Cruise Finds Deep-sea Plume.” Nature News. Nature Publishing Group, 18 May 2010. Web. <>.
  20. Torrice, Michael, and Melody Voith. “New Oil Clean-Up Technique Tested.” CEN RSS. N.p., 7 May 2010. Web. <>.
  21. “Van Der Waals Forces.” Van Der Waals Forces – A-Level Chemistry Revision. N.p., n.d. Web. <>.
  22. Warner, Arden. “The Magnet Wand That Cleans Oil Spills.” Interview by Zach Goldbaum. YouTube. Motherboard, 12 Dec. 2014. Web. <>.
  23. Yapa, Poojitha D., Li Zheng, and Kisaburo Nakata. “Modeling Underwater Oil/Gas Jets and Plumes.” Journal of Hydraulic Engineering 125.5 (1999): 481-91. Web.
  24. Zaitsev, Vladimir S., Dmitry S. Filimonov, Igor A. Presnyakov, Richard J. Gambino, and Benjamin Chu. “Physical and Chemical Properties of Magnetite and Magnetite-Polymer Nanoparticles and Their Colloidal Dispersions.” Journal of Colloid and Interface Science 212.1 (1999): 49-57. Web.
  25. Zhao, Lin, Franklin Shaffer, Brian Robinson, Thomas King, Christopher DâAmbrose, Zhong Pan, Feng Gao, Richard S. Miller, Robyn N. Conmy, and Michel C. Boufadel. “Underwater Oil Jet: Hydrodynamics and Droplet Size Distribution.” Chemical Engineering Journal 299 (2016): 292-303. Web


Please enter your comment!
Please enter your name here