The primary processes that affect the fate of spilled oil are spreading, evaporation, dispersion, dissolution, and emulsification (Payne et al., 1987; Boehm, 1987; Lehr, 2001). These processes - called weathering - dominate during the first few days to weeks of a spill, and, except for dissolution, can dramatically change the nature of the oil. A number of longer-term processes also occur, including photo- and biodegradation, auto-oxidation, and sedimentation. These longer-term processes are less important than the five above for the initial fate of spilled oil. Longer-term processes are more important in the later stages of weathering and usually determine the ultimate fate of the spilled oil.

The chemical and physical composition of oil changes with weathering. Some oils weather rapidly and undergo extensive changes in character, whereas others remain relatively unchanged over long periods of time. As a result of evaporation, the effects of weathering are generally rapid (1 to 2 days) for hydrocarbons with lower molecular weights. Degradation of the higher weight fractions is slower and occurs primarily through microbial degradation and chemical oxidation. The weathering or fate of spilled oil depends on the oil properties and on environmental conditions. It is important to recognize the dynamic nature of spilled oil and the fact that the properties of spilled oil can change over time.

Spreading reduces the bulk quantity of oil present in the vicinity of the spill but increases the spatial area over which adverse effects may occur. Thus oil in flowing systems, as opposed to contained systems, will be less concentrated in any given location, but may cause impacts, albeit reduced in intensity, over a much larger area. Spreading and thinning of spilled oil also increase the surface area of the slick, enhancing surface-dependent fate processes such as evaporation, degradation, and dissolution.

Evaporation is the primary mechanism for loss of low molecular weight constituents and light oil products. As lighter components evaporate, the remaining petroleum product becomes denser and more viscous. Evaporation tends to reduce oil toxicity but enhance persistence. Hydrocarbons that volatilize into the atmosphere are broken down by sunlight into smaller compounds. This process, referred to as photodegradation, occurs rapidly in air, and the rate of photodegradation increases as molecular weight increases.

Dispersion of oil increases with increasing surface turbulence. The dispersion of oil into water may serve to increase the surface area of oil susceptible to dissolution and degradation processes and thereby limit the potential for physical impacts.

Dissolution of oil in water is not a significant process controlling the oil's fate in the environment. It is one of the primary processes affecting the toxic effects of a spill, especially in confined water bodies. Dissolution increases with 1) decreasing molecular weight, 2) increasing temperature, 3) decreasing salinity, and 4) increasing concentration of dissolved organic matter.

Emulsification is the incorporation of water into oil and is the opposite of dispersion. Small drops of water become surrounded by oil. External energy from wave action is needed to emulsify oil. In general, heavier oils emulsify more rapidly than lighter oils. The oil may remain in a slick, which can contain as much as 70 percent water by weight and can have a viscosity a hundred to a thousand times greater than the original oil. Water-in-oil emulsions often are referred to as "mousse."

Photodegradation of oil increases with greater solar intensity. It can be a significant factor controlling the disappearance of a slick, especially of lighter products and constituents; but it will be less important during cloudy days and may be nonexistent in winter months on the North Slope. Photodegraded petroleum product constituents tend to be more soluble and more toxic than parent compounds. Extensive photodegradation, like dissolution, may thus increase the biological impacts of a spill event.

In the immediate aftermath of a spill, natural biodegradation of oil will not tend to be a significant process controlling the fate of oil in water bodies previously unexposed to oil. Microbial populations must become established before biodegradation can proceed at any appreciable rate.

Overall, the environmental fate of released oil is controlled by many factors and persistence is difficult to predict with great accuracy. Major factors affecting the environmental fate include the type of product, spill volume, spill rate, temperature of the oil, terrain, receiving environment, time of year, and weather. Crude oil will weather differently from diesel or refined oil in that both diesel and refined oil will evaporate at a significantly faster rate than crude oil.

The characteristics of the receiving environment, such as type of land, the surface gradient, marine or freshwater, surface or subsurface, spring ice overflow, summer open water, winter under ice, or winter broken ice, will affect how the spill behaves. In ice-covered waters, many of the same weathering processes are in effect as with open water; however, the ice changes the rates and relative importance of these processes (Payne, McNabb, and Clayton, 1991).

The time of year that a spill occurs has a significant effect on the fate of the crude oil. The time of year controls climatic factors such as temperature of the air, water or soil; depth of snow cover; whether there is ice or open water; and the depth of the active layer. During winter the air temperature can be so cold as to modify the viscosity of the oil so it will spread less and may even cause it to solidify. The lower the ambient temperature, the less crude oil evaporates. Both Prudhoe Bay and Endicott crudes have experimentally followed this pattern (Fingas, 1996). Frozen ground will limit the depth of penetration of any spill.

Oil movement over the ground surface follows the topography of the land (oil flows downhill). In general, oil will flow until it reaches a surface water body or a depression, or until absorption prevents further movement. Oil flowing over land can infiltrate vegetation cover, soil and snow. The rate of oil movement and depth of penetration are dependent on a variety of factors. If released onto tundra, oil can penetrate the soil as a result of the effects of gravity and capillary action. The rate of penetration will depend on the season, nature of the soil and the type of petroleum product. In summer, spills penetrate the active layer and then spread laterally on the frozen subsurface, accumulating in local downturns. From there the oil can penetrate into the permafrost (Collins et al., 1993). Precipitation may increase penetration into thawed soils (Solntseva, 1998 as cited in Chuvilin et al., 1999). If groundwater becomes contaminated, contaminants generally remain concentrated in plumes. Because ground water moves relatively slowly, contaminants do not mix or spread rapidly. Contaminated ground water may eventually migrate and appear in surface waters.

In winter spreading is controlled by the snow cover or frozen soil. Snow cover can act as an absorbent, slowing the spread of oil or preventing the spill from reaching the tundra surface. During winter, oil spreads on the surface of the frozen soil and penetration of oil into the soil is generally limited. Pore space in the soils that is not filled with ice may allow spilled oil to move into the frozen soil (Yershov et al., 1997; Chuvilin et al., 1999).

Tundra relief on the coastal plain of the North Slope is low enough to severely limit the spread of spills. During summer, flat coastal tundra develops a dead-storage capacity averaging 0.5 to 2.3 inches deep (Miller, Prentki, and Barsdate, 1980), which would retain 300 to 1,500 bbl of oil per acre. Even at high-water levels, the tundra vegetation tends to act as a boom, with both vegetation and peat functioning as sorbents that allow water to filter through, trapping the more viscous oil (e.g., Barsdate et al., 1980) - and also making recovery of the oil more difficult. On the other hand, even small spills can be spread over large areas if the spill event includes aerial, pressured discharge. With the high-velocity, bi-directional winds on the North Slope, oil can be misted miles downwind of a leak (Knowles, 1998). For example, in December 1993, an ARCO drill site line failed, and 1 to 4 bbl of crude oil misted over an estimated 100 to 145 acres (Ott, 1997).

Weathering processes generally would be similar in NPR-A freshwater and coastal marine regimes. Seasonal ice cover can greatly slow weathering in both regimes.

Oil spreading on the water surface (but not necessarily the transport of oil by moving water) would be restricted in most NPR-A waters. Because of the increased viscosity of oil in cold water, oil spills in NPR-A lake, river, and marine waters would spread less than in temperate fresh or marine waters. The exception to this would be a spill in shallow, marshy or ponded tundra or flooded lake margins in summer, which could spread similarly to a temperate spill. The exception is possible because these shallower waters can reach temperatures up to 18 °F - warmer than other tundra waters (Miller, Prentki, and Barsdate, 1980), and warm enough to lower oil slick viscosity.

Oil spills spread less in cold water than in temperate water because of the increased oil viscosity. This property will reduce spreading. An oil spill in broken ice would spread less and would spread between ice floes into any gaps greater than about 8 to 15 centimeters (cm) (Free, Cox, and Shultz, 1982).

An oil spill under ice would follow the general manner described below:

The spread of oil under the ice may be affected by the presence of currents, if the magnitude of those currents is large enough. A field study near Cape Parry in the Northwest Territories reported currents up to 10 cm/sec were present. This current was insufficient to strip oil from under the ice sheet after the oil had ceased to spread (NORCOR, 1975). Laboratory tests have shown that currents in excess of 15 to 25 cm/sec are required to strip oil from under-ice depressions (Cammaert, 1980; Cox et al., 1980). Current speeds in the nearshore Beaufort generally are less than 10 cm/sec during the winter (Weingartner and Okkonen, 2001). The area of contamination for oil under ice could increase if the ice were to move. Because the nearshore Beaufort is in the landfast ice area, the spread of oil due to ice movement would not be anticipated until spring breakup.

Evaporation of oil generally correlates to temperature (Fingas, 1996). The lower the temperature, the slower crude oil evaporates. Both Prudhoe Bay and Endicott crudes have this pattern (Fingas, 1996). Oil between or on ice is subject to normal evaporation. Oil that is frozen into the underside of ice is unlikely to undergo any evaporation until its release in spring. In spring as the multi-year ice deteriorates, the encapsulated oil will rise to the surface through brine channels in the ice. As oil is released to the surface, evaporation will occur. Because freshwater and first year ice do not have enough salts to form brine channels, the oil would be released only as the ice surface ablated to the level of the encapsulated oil. For freshwater ice, this would be when the ice became porous within about 2 weeks of meltout, from May to July, depending on weather, ice thickness, and location of the oil in the ice. In multi-year ice, surfacing of the oil probably would not occur until August, and some oil would not be released until the following summer.

Dispersion of oil spills occurs from wind, waves, currents, or ice. Any waves within the ice pack tend to pump oil onto the ice. Some additional oil dispersion occurs in dense, broken ice through floe-grinding action. More viscous and/or weathered crudes may adhere to porous ice floes, essentially concentrating oil within the floe field and limiting the oil dispersion. Alaska North Slope crude oil will readily emulsify to form stable emulsions. Emulsification of some crude oils is increased in the presence of ice. With floe grinding, Prudhoe Bay crude forms a mousse within a few hours - an order of magnitude more rapidly than in open water.

The weathering processes acting on oil in and along streams or rivers are in most cases similar to those described above for freshwater or marine spills. The dynamics of a river or stream environment, however, have additional effects on the fate and behavior of spilled oil. Oil entering rivers and streams will begin to spread as in freshwater or marine spills, but the spreading motion will be rapidly overcome by the surface current, at which point an elongated slick will form. The oil will flow downstream at the speed of the current in the absence of wind effects. In general, oil will tend to accumulate in areas of quiet water or eddies at the inside of river bends on a meandering river or stream, or in other pools where velocities are slower. Pools of oil may also accumulate behind log or debris jams. Water near the center of a stream channel will flow faster than water near the banks or bottom of the channel where the retarding forces of friction with the channel are greater. This difference in current speed and the resulting shearing forces between water layers is typically the major mixing mechanism that spreads a slick out as it moves downstream. The resulting smearing of the oil distribution along the axis of flow controls the plume shape and size, and the distance over which the oil concentration will remain above a particular level of concern. The leading edge of the slick may move as a relatively sharp front (at the mid-channel current speed) however, mixing will continuously exchange water and oil between the slower, near-bank regions and the faster-flowing, center regions of the river. From a practical point of view, this means that, although it might be possible to predict the initial arrival of oil at a point along the river, it will be considerably more difficult to estimate when the threat is past, since the areas of slower currents may continue to supply oil to the main stream channel, even after the leading edge is past (Overstreet and Galt, 1995)..

Shear-dominated flows cause another effect that characterizes river spills. Shear in currents along the banks and river bottom is typically the major source of turbulence in rivers, in contrast to surface-wave activity in oceans. Mixing and dispersion caused by the interaction of the shear and the turbulence can move significant amounts of oil below the surface (particularly if it is relatively dense, such as a heavy No. 6 oil; or if it is finely distributed as droplets). The shear-dominated river regimes tend to produce spill distributions having higher subsurface oil concentrations than would be expected in marine spills (Overstreet and Galt, 1995). This turbulence increases with increased velocity of flow and bed roughness.

On July 16, 1970, 5 bbl (Prentki, 1997, pers. comm.) of Prudhoe Bay crude was experimentally spilled in a 0.07-acre tundra Pond E in the NPR-A near Barrow (Miller, Alexander, and Barsdate, 1978; Barsdate et al., 1980; Hobbie, 1982). The general behavior of this experimental spill is instructive about what to expect for a small spill in the Planning Area during the summer or for a winter spill that melts out during thaw.

In this experimental spill, the oil spread over the water surface within a few hours to a 0.06-inch thickness. Within 24 hours, the slick thickened, as lighter hydrocarbons evaporated, and shrank into a 10- to 16-ft band on the downwind side of the pond. For about a month, the oil moved back and forth across the pond, shifting sides with changes in wind direction. Gradually, the oil worked part way into the pond's vegetated margins. By the end of summer, all of the oil was trapped along the pond margins either on the water's surface or on the bottom. No oil left the pond during the next spring runoff, despite significant water throughflow. Half of the oil was estimated to have evaporated or degraded within a year, but the rest of the oil remained with little change for at least 5 years.