Torrey Canyon

In 1967, the supertanker Torrey Canyon leaked oil onto the English coastline.^[5] Alkylphenol surfactants were primarily used to break up the oil, but proved very toxic in the marine environment; all types of marine life were killed. This led to a reformulation of dispersants to be more environmentally sensitive.^[when?] After the Torrey Canyon spill, new boat-spraying systems were developed.^[5] Later reformulations allowed more dispersant to be contained (at a higher concentration) to be aerosolized.

Exxon Valdez

Alaska had fewer than 4,000 gallons of dispersants available at the time of the Exxon Valdez oil spill, and no aircraft with which to dispense them. The dispersants introduced were relatively ineffective due to insufficient wave action to mix the oil and water, and their use was shortly abandoned.^[6]

A report by David Kirby for TakePart found that the main component of the Corexit 9527 formulation used during Exxon Valdez cleanup, 2-butoxyethanol, was identified as "one of the agents that caused liver, kidney, lung, nervous system, and blood disorders among cleanup crews in Alaska following the 1989 Exxon Valdez spill."^[7]

Early use (by volume)

Dispersants were applied to a number of oil spills between the years 1967 and 1989.^[8]

Deepwater Horizon

During the Deep water Horizon oil spill, an estimated 1.84 million gallons of Corexit was used in an attempt to increase the amount of surface oil and mitigate the damage to coastal habitat. BP purchased all of the world's supply of Corexit soon after the spill began.^[10] Nearly half (771,000 gallons) of the dispersants were applied directly at the wellhead.^[11] The primary dispersant used were Corexit 9527 and 9500, which were controversial due to toxicity.

In 2012, a study found that Corexit made the oil up to 52 times more toxic than oil alone,^[12] and that the dispersant's emulsifying effect makes oil droplets more bio-available to plankton.^[13] The Georgia Institute of Technology found that "Mixing oil with dispersant increased toxicity to ecosystems" and made the gulf oil spill worse.^[14]

In 2013, in response to the growing body of laboratory-derived toxicity data, some researchers address the scrutiny that should be used when evaluating laboratory test results that have been extrapolated using procedures that are not fully reliable for environmental assessments.^[15][16] Since then, guidance has been published that improves the comparability and relevance of oil toxicity tests.^[17]

Rena oil spill

Maritime New Zealand used the oil dispersant Corexit 9500 to help in the cleanup process.^[18] The dispersant was applied for only a week, after results proved inconclusive.^[19]

Overview

Surfactants reduce oil-water interfacial tension, which helps waves break oil into small droplets. A mixture of oil and water is normally unstable, but can be stabilized with the addition of surfactants; these surfactants can prevent coalescence of dispersed oil droplets. The effectiveness of the dispersant depends on the weathering of the oil, sea energy (waves), salinity of the water, temperature and the type of oil.^[20] Dispersion is unlikely to occur if the oil spreads into a thin layer, because the dispersant requires a particular thickness to work; otherwise, the dispersant will interact with both the water and the oil. More dispersant may be required if the sea energy is low. The salinity of the water is more important for ionic-surfactant dispersants, as salt screens electrostatic interactions between molecules. The viscosity of the oil is another important factor; viscosity can retard dispersant migration to the oil-water interface and also increase the energy required to shear a drop from the slick. Viscosities below 2,000 centipoise are optimal for dispersants. If the viscosity is above 10,000 centipoise, no dispersion is possible.^[21]

Requirements

There are five requirements for surfactants to successfully disperse oil:^[5]

• Dispersant must be on the oil's surface in the proper concentration
• Dispersant must penetrate (mix with) the oil
• Surfactant molecules must orient at the oil-water interface (hydrophobic in oil and hydrophilic in water)
• Oil-water interfacial tension must be lowered (so the oil can be broken up).
• Energy must be applied to the mix (for example, by waves)

Effectiveness

The effectiveness of a dispersant may be analyzed with the following equations.^[22] The Area refers to the area under the absorbance/wavelength curve, which is determined using the trapezoidal rule. The absorbances are measured at 340, 370, and 400 nm.

Area = 30(Abs₃₄₀ + Abs₃₇₀)/2 + 30(Abs₃₄₀ + Abs₄₀₀)/2 (1)

The dispersant effectiveness may then be calculated using the equation below.

Effectiveness (%) = Total oil dispersed x 100/(ρ_oilV_oil)

• ρ_oil = density of the test oil (g/L)
• V_oil = volume of oil added to test flask (L)
• Total oil dispersed = mass of oil x 120mL/30mL
• Mass of oil = concentration oil x V_DCM
• V_DCM = final volume of DCM-extract of water sample (0.020 L)
• Concentration of oil = area determined by Equation (1) / slope of calibration curve

Dispersion models

Developing well-constructed models (accounting for variables such as oil type, salinity and surfactant) are necessary to select the appropriate dispersant in a given situation. Two models exist which integrate the use of dispersants: Mackay's model and Johansen's model.^[23] There are several parameters which must be considered when creating a dispersion model, including oil-slick thickness, advection, resurfacing and wave action.^[23] A general problem in modeling dispersants is that they change several of these parameters; surfactants lower the thickness of the film, increase the amount of diffusion into the water column and increase the amount of breakup caused by wave action. This causes the oil slick's behavior to be more dominated by vertical diffusion than horizontal advection.^[23]

One equation for the modeling of oil spills is:^[24]

${\displaystyle {\frac {\partial h}{\partial t}}+{\vec {\nabla }}\left(h\left({\vec {U}}+{\frac {\vec {\tau }}{f}}\right)\right)-{\vec {\nabla }}(E{\vec {\nabla }}h)=R}$

where

• h is the oil-slick thickness
• ${\displaystyle {\vec {U}}}$ is the velocity of ocean currents in the mixing layer of the water column (where oil and water mix together)
• ${\displaystyle {\vec {\tau }}}$ is the wind-driven shear stress
• f is the oil-water friction coefficient
• E is the relative difference in densities between the oil and water
• R is the rate of spill propagation

Mackay's model predicts an increasing dispersion rate, as the slick becomes thinner in one dimension. The model predicts that thin slicks will disperse faster than thick slicks for several reasons. Thin slicks are less effective at dampening waves and other sources of turbidity. Additionally, droplets formed upon dispersion are expected to be smaller in a thin slick and thus easier to disperse in water. The model also includes:^[23]

• An expression for the diameter of the oil drop
• Temperature dependence of oil movement
• An expression for the resurfacing of oil
• Calibrations based on data from experimental spills

The model is lacking in several areas: it does not account for evaporation, the topography of the ocean floor or the geography of the spill zone.^[23]

Johansen's model is more complex than Mackay's model. It considers particles to be in one of three states: at the surface, entrained in the water column or evaporated. The empirically based model uses probabilistic variables to determine where the dispersant will move and where it will go after it breaks up oil slicks. The drift of each particle is determined by the state of that particle; this means that a particle in the vapor state will travel much further than a particle on the surface (or under the surface) of the ocean.^[23] This model improves on Mackay's model in several key areas, including terms for:^[23]

• Probability of entrainment – depends on wind
• Probability of resurfacing – depends on density, droplet size, time submerged and wind
• Probability of evaporation – matched with empirical data

Oil dispersants are modeled by Johansen using a different set of entrainment and resurfacing parameters for treated versus untreated oil. This allows areas of the oil slick to be modeled differently, to better understand how oil spreads along the water's surface.

HLB values

Surfactant behavior is highly dependent on the hydrophilic-lipophilic balance (HLB) value. The HLB is a coding scale from 0 to 20 for non-ionic surfactants, and takes into account the chemical structure of the surfactant molecule. A zero value corresponds to the most lipophilic and a value of 20 is the most hydrophilic for a non-ionic surfactant.^[5] In general, compounds with an HLB between one and four will not mix with water. Compounds with an HLB value above 13 will form a clear solution in water.^[25] Oil dispersants usually have HLB values from 8–18.^[25]