Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
7 MITIGATION AD 12[M[DIATBO~ OF Old AD GAS ACT'VIT'RS I~T~ODUCT10~ This chapter describes the potential effects of selected routine petroleum industry activities and accidental events in the marine arctic environment and their mitigation and remediation. In the case of accidents the main concern is with of] spiDs~oth the effects associated with placing equipment in arctic marine waters and the special difficulties in applying countermea- sures to mitigate oil spills are discussed. To predict effects properly, one needs to know about the biological and human resources that might be affected and have some idea of the location and size of the hydrocarbon resource base. Prediction of the locations and severity of effects depends on the size, location, and distribution of the reservoirs to be developed. Loss or displacement of species harvested in traditional hunting and fishing areas must be considered in the evaluation of environmental effects. Seasonal extremes of temperature and light determine the periods of biological activity, and these factors also determine and limit Me time available for control and cleanup of a spill. There continues to be controversy, perhaps semantic in origin, about whether the arctic environment should be called "fragile." The popular news media and some special interest groups use the term to characterize He uniqueness of the Arctic on the evidence of long-lasting disturbance by human activities on land surfaces, although it is debatable if this rationale can be applied to industry effects in Be marine environment. Some scientists with long experience in the Arctic (Dunbar, 1985) do not believe that He effects of pollution in the Arctic are greater Can Hey are else 153
154 OCS DECISIONS: ALASKA where. Dubber makes the point Hat although Were is less diversity among species in the Arctic than can be found in temperate and tropical regions, numbers within species tend to be larger and distributions extend over greater geographic areas, suggesting that even severe local damage could be repaired by immigration from adjacent areas. Also, it is known that the arctic environment can be highly variable within and among seasons, and it would not be an unreasonable assumption that arctic species are adapted to short-term harm and are able to compensate for population losses in one area or year by transmigrabon and adjustments in reproduction. The danger lies in blaming human activity for a short-term loss in species caused, for example, by a heavy ice year. On the other hand, specific regions arm populadons can be extremely significant to Alaska Natives. In addition, long-term industrial presence in one area might have ecological consequences, especially on migrating species, by making this affected zone unsuitable for potentially obligatory habitat for a population. In dial case, it would be essential to know Be size of Me "footprint" of He activity and its ecological spatial and temporal relevance to draw any meaningful conclusions about its potential to cause lasting harm. A12EAS AFFECTED BY ICE Platform Deslon The task of designing and cons~uchng of} production platforms for regions such as He Chukchi and Beaufort seas should seem nearly impossi- ble given He formidable environment that must be confronted, but it is not. It is generally believed that designing a permanent production facility is less of a challenge Han is designing an exploration facility, because a production facility will be larger Han an exploration platform. The object of explora- tion is to arrive on site, drill, and leave, encountering as little ice as pos- sible. Careful dining of the dolling and strategies such as coming out of the hole or disconnechug the rig to avoid heavy ice are a common part of such operations. For production, periods of minimal ice would be chosen for moving a facility on site. However, once He structure is there, it would have to withstand He ice as it comes. The oil industry already has considerable experience, for instance, with the platforms in Cook Inlet, which have successfully operated throughout
MITIGATION AND REMEDLATION 155 the winter ice season for the past 27 years. Although the ice conditions in Cook Inlet are less severe Han in the Arctic, the operations there never~e- less are instructive. There has been extensive exploratory drilling in both the Canadian and the Alaskan portions of the Beaufort Sea as well, where the systems used to protect against ice have ranged from artificially thick- ened ice pads Hat support winter-only operations in shallow water, to grave} islands, to gravel-fi~led caisson structures, to concrete structures. Several ancillary techniques also have been used to minimize the effect of ice forces on such structures. One is to construct berms that force grounding and de- formation of the ice before it actually arrives at the offshore structure. Another is to spray water to build large, thick ice "bumpers" around drilling platforms. Although these techniques would prove useful in the development of designs for production platforms, they would be used only to minimize the general force Hat ice exerts on the platform, therefore minimizing structural fatigue. Platforms would have to be designed to deal with ice forces without the application of such techniques. Designing offshore structures that can withstand the forces exerted by moving pack ice is hardly a new problem. Similar problems with designing icebreakers (Makarov, 1901; Lewis and Edwards, 1970; Weeks and Handy, 1970), in coping with ice forces on pilings and piers exerted by river and sea ice (Blenkarn, 1970; Afanas'ev et al., 1973; Neill, 1976; Nevel et al., 1977), and in evaluating ice-bearing capacity (Hertz, 1884; Wyman, 1950; Assur, 1961; Barthelemy, 1990) have been studied for years. In addition, since He discovery of of} at Pru&oe Bay, Here has been a major industry program to advance knowledge about ice mechanics specifically related to He design of offshore structures. Some sense of He extensive literature available on tills subject can be gained by reading He work of Sanderson (1987, 1988), Cammaert and Muggeridge (1988), and Timco (1989~. However, a few points need to be mentioned. There are many aspects of He general subject of ice forces on offshore structures Hat are less well understood Can is desirable. For instance, dlere still is no generally agreed-on constitutive equation Hat describes He general mechanical behavior of sea ice. Also, He pressure- to-area curve for ice-and-structure interactions, although reasonably well documented at small scales (Sa2~erson, 1988), lacks an adequate theoretical basis to describe loads at failure as a function of He size of He failure area under consideration. High-quality biaxial and biaxial tests on sea ice are she quite rare. Tests that reset in He fidl-scale failure of ice sheets are still
156 OCSDEClSIONS: ALASKA limited, although a small number have been completed (Dykins, 1971; Vaudrey, 1977; Chen and Lee, 1986; Lee et al., 1986). Some observations have allowed estimates of the force levels reached when large, thick multiyear ice floes run into natural "towerlike" rock islands (Danielew~cz and Blanchet, 1988; Montemurro and Sykes, 1989~. An adequate explana- don for the surprisingly low forces that were observed is still being sought. Models of the ridging process have not been adequately incorporated into procedures for estimating ice forces (Vivitrat and Kreider, 1981~. However, even considering these limitations, there still are enough field and laboratory data and theory to assess threats to and design criteria of offshore production structures for the OCS areas of the Beautort and Chukchi seas. One reason that some believe that it is impossible to design production structures for the Beautort and Chukchi seas is the presence of ice islands, which are produced by the breakup of portions of the Ellesmere Ice Shelf and are the tabular icebergs of the Arctic Ocean. They typically are 40-50 m thick with lateral dimensions that range from tens of meters to tens of l~ometers (leffries, 1992~. If a large ice island embedded in drifting pack ice were to run into an offshore production facility, Be facility would probably be destroyed. The chances of this occurring are extremely small; large ice islands are rare. In addition, it is easy to identify ice islands by synthetic aperture radar (SAR) imagery Jeffries and Sackinger, 1990~. If an ice island were to drift near a production platform, the possibility of a collision would be known well in advance and proper precautions could be taken. Such rare events would, however, affect the potential economics of offshore operations and should be considered in estimating the economics of any new field. It should be noted that research in ice mechanics continues. During the winters of 1992-1993 and 1994-1995, field programs supported byindustry and by the Office of Naval Research (ONR) were and m11 be conducted. The ONR program also includes a laboratory component that is generating new test data under carefully controlled conditions. If successful, these programs should contribute new information about the scaling problem and add additional biaxial, biaxial, and fil11-ice-thickness tests. The committee does not believe that additional studies in ice mechanics supported by MMS are essential to the safe development of OCS of! in either the Chukchi or the Beaufort.
MITIGATION AND REMEDIATION 157 ICE GOUGING ED PIPELINE PACT It takes several years to collect some of the information needed for design of safe offshore facilities. One example is the time required to colDect Me data necessary to make confident estimates of safe burial depths for offshore pipelines. A significant firm Rat opens the way to development of an of] field must account for Be way of} would be brought to shore. For example, if the of} is to be shipped by tankers loaded at the offshore field, feeder pipelines might be necessary to move the of! to a central loading point. In Be Beautort Sea and at most locations in the Chukchi Sea, feeder pipelines would probably transport the of! near to the center of the field, from which it would be fed into a main pipeline for transport to the coast and ultimately to the Trans-Alaska Pipeline System. In the Chukchi Sea, it is easy to conceive of production sites requiring hundreds of kilometers of subsea pipelines that transect regions where the seafloor is extensively gouged by ice. In the Beautort Sea, the area affected by gouging is smaller, because the shelf is narrower, but the problem is no less severe; gouging is generally more intense in the Beaufort Sea. Gouging is not a problem in Be deeper Navarin Basin. Engineers designing such subsea pipelines need to be able to answer several questions, including the following: How deep should Be line be buried along each section of a proposed pipeline route to ensure that He risk of its being damaged by a moving ice mass is acceptable? · Is simple burial adequate for absolute safety in high-risk areas or is it necessary also to armor the pipeline to reduce the possibility of an under-ice spill from a ruptured line? · Are Here regions where gouges do not occur and where pipelines can be placed on He seafloor wig minimal burial or none at all? If so, where are these regions? Given the cost of deep burial and the short operating season for pipeline burial equipment such as subsea plows or cutter-suction dredges, the cost of pipeline installation could have an appreciable effect both on when a given field could be brought on line and on the overall economics of the field. In an extreme situation, pipeline costs could rule out production at
158 OCSDECISlONS: ALASKA some sites or, more likely, could delay the decision-making process while missing data thought necessary for safe design were collected. State of Knowledge As the ice pack moves over the shadower waters of the continental shelf, Be keels of pressure ridges and the lower surfaces of ice islands plow up the seafloor has been known as a scientific cunousity since the 1920s. Once offshore oil development on the margins of the Arctic Ocean became a possibility, a thorough understanding of the process became important. During the past 20 years, gouging has received considerable attention. For instance, a recent bibliography (Goodwin et al., 1985) lists 379 publications with some 101 papers dealing with the Beaufort Sea, 23 that discuss the Chukchi and Bering seas, 15 that deal win theory and modeling, and 29 that deal with protection. (There also are many published studies about gouges produced by true icebergs off the east coast of Canada. Most of this information is not particularly applicable to arctic regions.) It is well established that gouge depths can be described by a simple neg- ative exponential distribution: Small gouges are frequent; deep gouges are rare. The spacing between gouges also shows a similar distribution, sug- gesting that gouging can be approximated as a random process. Bow the frequency and the intensity of gouging are clearly related to water depth. In the U.S. portion of the Beautort Sea, most deep gouges occur in water of 30~0 m deep. The largest gouges observed on the U.S. portion of Be Beautort shelf are about 5.5 m deep (Barnes et al., 1984~. Larger gouges of up to 6 m deep have been observed in Be Canadian portion of the Beau- fort Sea, probably because of differences in the consolidation of the subsea sediments. Closer to Be Mackenzie Delta, gouge densities are extremely variable; there can be as many as 500 gouges/km2. Track samples indicat- ing more Man 100/krn2 are not rare. In Be Chukchi Sea, Be deepest gouges occur at water depths of 35-50 m and are as much as 4.5 m deep. Most gouges are oriented roughly parallel to Be coast, running generally east-west in Be Beautort Sea and nor~east-southwest in Be Chukchi Sea. In bow tile Beautort and Be Chukchi seas, where there are offshore shoals, the largest number of gouges occur on the seaward flanks of the shoals (Reimnitz and Kempema, 1984~. In general, gouging is rare in water Cat is more than 50 m deep, and it does not appear to occur in water deeper Man 60 m. Unfortunately, Be amount of data bearing on these limiting values is small.
MITIGATION AND REMEDL4nON 159 emorovInd the State of Knowledge It would be desirable to have a better estimate of Be depth at which gouging ceases to be a problem in Be Chukchi and Beautort seas. In the recent geological past, sea level was roughly 100 m lower than it is now. Therefore, gouges in deeper waters may be relics rawer than indicative of current activity. This is particularly true if gouges formed in deeper water survive for long periods. Because the Chukchi and Beautort shelves are shallow and slopes are low, small changes in the maximum water depth where active gouging occurs result in large changes in the areas where gouging must be considered a problem. Even more important is the limited availability of data that provide direct determination of gouging rates as a function of water depth and location. It is impossible to know, for example, in a region with 100 gouges/km2, whedler they were formed during the past year, during the past 10 years, or during the past ~ ,000 years. There is no absolute way to date gouges, despite the importance of the information to the decision-making process (Weeks et al., 1984; Wang, 1990~. Simulations suggest that most gouges are fined within a few years (Weeks et al., 1985~. Field observations also suggest significant sediment movement along Be arctic OCS (Barnes and Reimnitz, 1974; Barnes, 1979~. Whether these conclusions apply to water depths greater Man 45 m is not known. Measurement Program The development of a gooc! data base on gouging rates and the infilling process requires yearly replicate measurements of gouging patterns in selected regions of Be Alaskan OCS. Although some data exist (Barnes et al., 1978; Rearic et al., 1981), expansion of this data base would contribute significantly to safe pipeline design. We stress that this information cannot be quickly obtained even if major financial resources are devoted. On Be over hand, by carrying on a low-level, continuing program, excellent data can be obtained for reasonable cost. The longer the time series, the better wait be Be estimates of the gouging rates and initial gouge depths. This recommended program will not replace the site-specific measure- meets that are required once a field large enough to warrant development is found. Nevertheless, general information would allow designers to view the site-specific measurements, which undoubtedly would be of much shorter duration, in a broader context. Given that ice-induced ruptures of
160 OCS DECISIONS: ALASKA ~- . . ~.. subsea pipelines should be considered a probable cause of below-ice oil spins and that current practice assumes that the granting of offshore leases win be followed by permission to develop if adequate petroleum resources are found, the more timely and safe development that should result from of these proposed measurements would be well worth the additional cost. The importance of initiating this program promptly is supported by the toi~ow~ng onserva~aons. ARCO announced in October 1992 the discovery of oil (3,400 barrels/day) at its Kuv~um No. ~ drilling site. The well is located in 36 m of water on the Beaufort shelf, roughly halfway between Prudhoe Bay and Olivia (Bar er Island). The drill ship had to discontinue operations several times and move off the site to avoid collision with an extremely large, thick floe drifting nearby. (This procedure is normal, ef- fective, and necessary; driD ships are not designed to survive significant ice forces.) During the time the floe was in view of the drilling platform, there was a period when the floe was obviously grounded, presumably resulting in one or more gouges. The water depth at the grounding location was 30 m. The Kuvlum prospect is believed to be part of a field that could produce more than ~ billion barrels of oil. To date in the Beautort OCS, ~ of 21 wells drilled have exhibited significant quantities of oil, so we believe Tat it is desirable to know as much as possible about gouging over both the Beautort and Chukchi OCS. AQCTIC OIL SPILLS: I~CIDI FOCI ED Al SPO~SI Recent reviews of the quantity and incidence of oil discharges into We world's oceans (GESAMP, 1993) show that input of oil from anthropogenic sources has decreased during the past decade. The quantity of oil entering the ocean from tanker spills has declined to only about 20% of input. (NRC, 199Ic3. Although marine transportation accidents account for many of Be largest single accidents, Heir frequency has declined. Figures 7-l and 7-2 demonstrate tills trend, particularly in tile United States in recent years. Transportation accidents presumably will continue to decline as stronger regulatory controls are implemented. The Canada Oil and Gas Lands Administration (COGLA, 198Sb) published a review of Be question of blowout frequency associated wig petroleum driBing, based on worldwide records for 1955-1980. The study
3.5 v, y c ~ 2.5 o o . - a' .. _ ~ ~. in Cal MITIGATION AND REMEDL4TION 161 11Z A )` By Tonnage Distance ~l/\ /Av ~ _ ~\// \ in, ~ y \ By Number at Tankers ~ _ \ _ u . , O 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 10 8 6 Year FIGURE 7-1 Worldwide rates of serious casualties to oil/chemical tankers, 1974-1988 (6,000 gross tons arid above). Source: NRC, l991c. determined that the blowout frequency associated with an oil loss was ~ in 3,065 wells c riDed; the incidence of blowouts associated with a " significant" of} loss-more than 50,000 barrels-was more remote, ~ in 7,325. The accumulated incidence of blowouts at exploratory wells was about three times as high as that for development wells. The frequency of pipeline ruptures worldwide has not decreased in recent years, resulting in a greater proportional contribution to the total amount spilled (Figure 7-3) (OSTR, 1993~. SPIII Response The response to an of} spill in arctic conditions is complicated by three main factors: the common presence of ice, the distance from spill cleanup logistic support, and long periods of darkness and cold. Although the response to a spill in open water can rely on standard techniques such as mechanical containment and recovery, ice-infested waters present special
162 OCS DECISIONS: ALASKA Millions of Gallons / / / / 140 120 100 60 4D of' art- ~ , ~ rid ~ . :/ , :, . ~ _ . .. . .. . :: . - ~.:~., ~ . . . : . /~/~/~. . . :-: air:: --/ ~, -: ~ A/ / . ~ .,~.. ~ A:' ~ :: ~ _ Vl ' ~ . ~1~:. HI _ y - _ _ _ ~ _~_ ,_ 'A A. ~7 ,,,,: it: ~ At /.:- / ~ : : ~ ::: ~ ____......... ,(:::: f /" / ,~:, / I~ ::~- A - :. it ,, ..., a: :~:::i . ~ , , ,, . ---;,' I,/ 1 He, . ~ . 1 : ~ -.: 1 : :. -: : _ ::: : :: ::: ::: :~ I: - . :::: . - - :: . . . 1978 1979 1980 1981 1982 1983 1984 1985 19B6 19B7 19B8 1989 1990 1991 1992 ,<: :: :J Par i.:::::: 1 ,t- ::::: :,::,, At V 9~/ Or - i/ us incidents ray /: / ..... ~ :/ a: n-U! i incidents FIGURE 7-2 Annual amounts spilled in vessel incidents, 1978-1992 (includes tankers, barges, and other vessels). Source: OSIR, 1993. Reprinted with permission from Oil Spill Intelligence Report Newsletter; copyright 1993, Cutter Information Corp., Arlington, Mass. problems and the effectiveness of most containment and recovery systems is compromised. New techrK)Iogy is slow in being realized, although burn- ing oil on ice and in ice-enclosed water surfaces appear to be the most useful countermeasure. An ice corral can contain of! so Mat countermea- sures can be applied, but it also can prevent access. Mechanical contain- ment and recovery have limited ability to handle solids (such as broken ice). Limits on remote-sensing capacity and difficulty in modeling the fate of oil caused by inadequate detailed knowledge of under-ice currents are addi- tional problems. Finally, bioremediation attempts are likely to be slowed by the low water and shoreline temperatures. When oil is spilled, the first
MITIGATION AND REMEDL!TION 163 / Millions of Gallons 250 200 150 100 50 a ~ r ,~., _ ski ~ forage Tanksl ~/ b ~ hcilities/Rigs a_ , ~/pines /~/a~l8SlOthH Vessels 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 FIGURE 7-3 Annual amounts spilled by source type, 1978-1992. Source: OSIR, 1993. Repnnted with permission from Oil Spill Intelligence Report Newsletter; copyright 1993, Cutter Information Corp., Arlington, Mass. priority in the Arctic, as elsewhere, should be to stop the source of the release and to contain We spill to as small an area as possible (Mackay, 1985). SPIIIS In Ice and In Very Cole Water The occurrence and fate of arctic of! spills have received considerable attention in He technical literature since oil was discovered at Pru&oe Bay and in Be Canadian Arctic. The following books and papers offer an in
164 OCS DECISIONS: ALASKA troduction to this subject: Malins (1977), Clark and Finley (1977), Thomas (1983), Engelhardt (1985a,b), Mackay (1985), Payne et al. (1987, 1989), Baker et al. (1990), annual Arctic Marine Oilspill Program symposia, and biennial oil spill conferences. For arctic of! spills, two possibilities should be examined: first, that both the spill and the cleanup occur under ice-free conditions, and ice is not present during the interim period; second, that sea ice is present at some time during the life of the spill. Soills In Ice-Free Water When both the spill and the cleanup occur in ice-free water, more information and response capability are available because the expected environmental conditions will not be much different from those that have been encountered at marine spills in more temperate regions. The primary differences here are well understood and can be attributed to the fact that the sea temperature is never far from the freezing temperature of sea water (~- I.~° C). Therefore, after the spilled oil cools to ambient temperatures (- ~ to +5° C) the of! will be more viscous than at higher temperatures. In addition, any chemical reactions that lead to the degradation of oil will be slower. Typically, degradation at 5° C is 20 to 50 times slower than at 25° C (NFAC, 1978~; for instance, the loss of volatile organic compounds from Me oil will take longer (Kirstein and Redding, 1987; Payne et al., 1987), which can lengthen the period during which the oil can be burned success- fully. At Me same time, the lower temperatures make ignition more difficult and burning proceeds at a slower rate. Burning has been used with some success in test spills in the Canadian and Norwegian Arctic, but additional tests are desirable. More important is that even though cleanup techniques are more efficient in temperate regions than they would be in the Arctic, recovery of oil from very large spills even in temperate waters is unlikely to exceed 20%. SPIIIS In Sea Ice There are several scenarios for spills in regions wig sea ice. For instance, Me spill could occur in open water and new ice could develop before cleanup begins, or an ice cover could already be present at the time
MIlIGATION AND REMEDlATION 165 of the spill. Also, ice conditions might be such that cleanup could be accomplished completely in the ice or only after the ice had melted or, more likely, by a bit of both. Finally, a spill could happen near fast ice or in mobile pack ice two vastly different situations. Spills In Newt Forming Sea Ice The simplest example in which the spill occurs in open water and subsequently becomes involved in newly forming sea ice occurs during the initial fall ffeeze-up when a Din first-year ice cover starts to develop. Good examples of regions where this is possible are the Bering and southern Chukchi seas; both are commonly completely ice-free at the end of the summer. However, similar processes can occur in polynya areas or in oil- covered leads even in regions of heavy multiyear pack ice. In such situations, the presence of the of} will damp surface waves and reduce turbulence in the upper part of the water column, thereby reducing condi- tions that would favor the formation of frazi} ice, the fine-grained, granular ice that forms in supercooled, turbulent water. Once the of! becomes cold enough, sea ice starts to form beneath the slick, and most of the of} will be trapped either on the upper ice surface or in the top few centimeters of the ice (Payne et al., 1987). Any rafting, which is common during this growth stage, will presumably smear oil over both the upper and the lower ice surfaces. In the case of pressure ridging, the oil will be incorporated completely throughout the ridge. It would then move with the ice and could be tracked along with the ice. Cleanup during this time of the year (largely during fall freeze-up) would be difficult at best. Although the ice is thick enough to limit shipping to vessels specifically designed for ice operations, it is sail too thin to support heavy equipment. Also, such ice is frequently vent mobile and commonly experiences lead formation and ridge building. Drying freeze-up the weaker is commonly marginal and daylight is limited, so Me problem of deploying and guaranteeing Be safety of cleanup crews would become a major consideration. Spills us Fast Ice If a spill occurs in fast ice, some advantages can be had if Be ice is Lick enough to support cleanup equipment. For instance, if Be spill is on top of
166 OCS DECISIONS: ALASKA the ice, the oil will be restricted to a smaller area. However, the oil wait seep into the surface snow cover, and if the spill is of an appreciable size, the task of removing large volumes of oil-saturated snow is formidable. Also, in surface spills, some of the oil would invariably become trapped either within or below the ice. In each case, locating and cleaning up the fraction of the spill either within or below the ice would be difficult. In all probability, this oil would not be recovered until the ice cover melts, at which time open-water cleanup procedures could be used. By this time, the fast-ice region will have become a region of pack ice. If the spill occurs below a region of fast ice, the ice also tends to restrict the movement of the oil, because the underside of undeformed fast ice is irregular due to local variations in ice growth rate caused by variations in the thickness and properties of the surface snow (Kovacs, 1977; Barnes, 19791. A few studies suggest that the bottom roughness of typical fast ice near the margin of the Arctic Ocean is adequate to immobilize significant quantities of oil. Estimates of the volume that would be trapped by the rough underside of the ice and therefore that would drift with the ice are 10,000 to 35,000 m3/km2. The containment potential for multiyear ice is generally believed to be much larger. For Instance, Kovacs (1977) estimates a value of slightly under 300,000 m3/km2 for multiyear ice with a mean draft of 4.3 m. The exact geometry and leakiness of these under-ice pools has not been documented. If the of! remap n place for an appreciable period during the ice-growth period, subsequent ice growth below the oil layer we result in the incorpo- ration of the of} into the ice sheet. The problem is then with locating and removing He oil. We are not aware of a proven remote-sensing method for locating such oil, although it is possible that either surface-based radar systems (Kovacs and Morey, 1986) or helicopter-borne electromagnetic systems (Kovacs and Holladay, 19903 win prove useful. Drilling and coring would be effective, but these procedures are not efficient. Once He of} is located some of it can be removed by down-hole pumping techniques. There is no simple way to remove of! that is entrapped within ice, so final cleanup win be possible only after He ice has melted and open- water or on-ice procedures can be applied. Encapsulation of of} by subsequent ice grown is only one method by which of} can become trapped in the ice. If He oil is of a sufficiently low viscosity it can move upward into the ice through He ubiquitous brine drainage channels Hat occur in natural sea ice (Lake and Lewis, 19701. In 2-m-thick sea ice, the average
MITIGATION AND REMEDLATION 167 spacing between these brine channels is about 15 cm (Lewis, 1976). Brine drainage channels are particularly important in tile spring because Hey can enlarge to features win a diameter of 5 cm or more, significantly facilitat- ing the entrance of Me of! into We ice. Unfortunately, although brine drainage tubes have been studied to explore Heir use in the desalination of sea ice, Heir role as a pathway to He incorporation of of} into sea ice has received little attention. Oil can also accumulate in He much larger diameter seal-brea~ing holes, presumably filling Rem up to at least He Fiche of He ice (~e actual position and height of the of} in the holes will depend on He volume of contiguous of} under He ice surface, He density different, and He size of He opening). This clearly would Greaten seals if He spill were large. By summer, oil would be on top of the ice even if the initial spill were initially completely entrapped beneath He ice (Clark and Finley, 1982~. As increasing amounts of shortwave radiation arrived both at He ice surface and at locations within the ice, this would warm the oil relative to the ice because of oil's much lower albedo-He fraction of incident electromagnetic radiation reflected by the surface. The shortwave radiation also would speed He melting of the ice, which would speed up the time when open- water cleanup techniques could be applied. The Office of Technology Assessment (OTA) estimates that mechanical recovery has been approx~- mately 15% efficient under good conditions. Even though major advances are being made for spill response in open waters as a result of the require- ments ureter the Oil Pollution Act of 1990, it is unlikely that the capability for control and recovery of spilled oil in ice-infested waters will be easily improved Spills in Pack Ice The most difficult situation is when the spill occurs in a region of heavy, mobile pack ice of varied thickness the situation most commonly encoun- tered in the Beautort and Chukchi seas, where multiyear ice is invariably present. Also associated with this is the ubiquitous occurrence of leads and pressure ridges. Of particular concern is the nearshore lead sys- tem sometimes referred to as He flaw lead or leads-located just seaward of He outermost fast-ice edge, which during the spring serves as a seabird foraging area and marine mammal migration route. The polynya that
168 OCS DECISIONS: ALASKA commonly occurs along portions of the Chukchi coast under conditions of offshore winds is important because it is an area of significant seabird and marine mammal activity. Cleanup of all but the smallest spills in pack ice presents a major problem, even though the ice itself tends to immobilize all or a portion of Be oil, depending on the size and the nature of the spill. However, this is a small consolation: The ice in which Be of! is spilled is, in most cases, drifting to Me west at speeds of 2 to 3 km/day. During storms, drift rates on the order of 2 m/s (4 knots) have been observed by radar at Barrow, Alaska (Shapiro and Barnes, 1991~. This could make the logistical aspects of a cleanup extremely difficult. The fate of the of} in the ice is similar to what might be anticipated from our previous discussion of of} in fast-melting ice, with the addition of the complications added by extensive lead forma- tion and ridging. Leads occur not only at the fast-ice edge, they can be found anywhere within tile pack. Any of} in the water that is not trapped by Be bottom roughness of the ice has a good chance of ending up in an area where the ice is thinnest; in fact where there is little or no ice at all- in a lead. Once in a lead, the oil is effectively trapped, although it can readily move laterally in the lead. The of} can, in principle, form a Hick layer because during the winter leads typically comprise only a small portion of the area of pack ice. Once trapped in a lead, the oil can be moved by the wind to the downwind side of the lead. There, it can be incorporated into the accumulating frazi} or form a layer on top of newly forming ice as described earlier. As He Minor ice classes characteristically found in leads are also He ice most commonly found In pressure ridges, it is to be expected Hat numerous ridges will have oiled ice distributed throughout their complete thickness. It is also likely that the keels of pressure ridges, which can go as much as 50 m into the water, will act as skimmers to any of} present in He water column that is not electively fixed to the ice. This will lead to an accumu- lation of of! on the upstream side of the keel. Another difficult is caused by He presence of the thicker (2.5-5.0 m) undeformed multiyear ice, which can constitute a major percentage of He ice cover. The additional thickness increases the difficulty of locating of} beneath He ice. Also, in many cases Be ice floes are neither large enough nor flat enough to make He use of heavy equipment practical, even Bough dley might be of sufficient thickness to support the equipment. In short, He presence of ice greatly complicates Be already difficult
MITIGATION AND REMEDL4TION 169 problem of cleaning up an oil spill in the ocean. In fact, it makes cleanup so difficult that in many cases final cleanup will be possible only after Me region of ice containing the of} has melted and open-water procedures can be applied. It is therefore important to be able to track the affected region of ice until it melts. Tracking SPIIIS In Ice The direct detection and tracking of oil in and under ice is not feasible with any certainty at this time. There are several techniques available for tracking ice. Perhaps the simplest is to place one or more global positioning system (GPS) buoys on the affected ice. This would provide a continuous record of the position of the ice until final cleanup. If it is not possible to do ~is, the movement of We ice could be tracked by SAR imaging. Adequate tracking should be available after 1995, when the Radarsat satellite is scheduled to be launched. It will provide He wide-area coverage (500 km) and high resolution modes that are lacking in the current genera- tion of SAR satellites. Open-ocean oil slicks can be identified by SAR, which therefore could be used to both track the ice and to identify the final open-ocean oil slick. Undoubtedly, both GPS and SAR would be used. There is also some promise for a sensor Cat can be used to detect oil under ice; for example, a pulsed ultraviolet fluorescence sensor Eat could detect of! through ice of 0.5-l m Dick (Moir and Yetman, 19934. The desdnation of an of! spill originating in He Chukchi-Beaufort region is not predictable (Thomas, 1983; Colony, 1985; Pritchard and Hanzlick, 1987~. Depending on He location and timing of He spill, He oil could exit He Arctic Ocean via He Bering Strait and arrive at He ice edge somewhere in the Bering Sea. However, it is more likely to drift to He west past Wrange! Island (Colony, 1985), where it could either hit He Siberian coast or become involved in the drift of He pack in He Transpolar Drift Stream. This route would take He oiled ice over He Norm Pole toward Pram Strait between Greenland and Svalbard (an archipelago of Norway), allowing He of] to exit the pack some 24 years later into He Norm Atlantic off He east coast of Greenland (Thorndike, 1986~. It is also possible Hat He oiled ice could become trapped in the Beautort Gyre for I-2 rotations of 5-10 years each before ultimately joining the Transpolar Drift Stream. When produc- don areas are identified in He Chukchi and Beaufort seas, additional studies
170 OCS DECISIONS: ALASKA that use state~f-the-art ice~ynamics models should be completed and compared with data sets colRected via drift buoys to validate the forecasts of ice trajectories for these areas. [FF[CTIVI HI AS OF SPOTS M"SU~I S It is important to evaluate spin countermeasures in an ecological context. Ever since the Torrey Canyon (1967) and Santa Barbara (1969) oil spills, there have been conflicts during spill cleanup over the desire to remove visible oil so that the environment looks "clean" and We desire to minimize the spill's environmental impact and promote natural recovery, which can mean using control methods in one area to prevent oil from entering another or leaving some of} on shorelines or in sensitive habitats when further cleanup would increase the damage caused by a spill (Lindstedt-Siva, 1979a,b). Much research has been done on the effects of spilled oil and of various spill cleanup methods. A review paper prepared for a 1982 workshop of of! spill specialists cited appropriately 1,000 references (Tetra Tech, 19851. The report was used as a basis for discussions, and the recommendations developed were later published as a manual (API, 1985) that identified habitat types that are particularly sensitive to damage from spills and techniques to protect them. The manual also evaluated the relative damage caused by various shoreline cleanup me~ods compared wig Me damage caused by the actual spill. Steam cleaning and over high-impact me~ods generally were not recommended. Low-pressure flushing and over low- effect techniques were recommended. The manual also promoted Me rationale Mat it is often better to leave some of} in Me environment Man it is to use extreme measures to remove it. Response TechnoloUles A work group on oil spill cleanup (Lindstedt-Siva, 1984) divided methods into several generic types, among Hem natural cleansing, mechanical removal, use of sorbents, use of chemical dispersants, use of sinking agents, and burning. Four general categories of environmental c amage Mat can result from these me~ods also were identified. Table 7-!
MITIGATION AND REMEDLATION 171 TABLE 7-1 Potential impacts on Vanous Oi'1 Spill Cleanup Techniques In Different Habitats Clean-up Methods 1. Natural Cleansing 2. Mechanical Removal a. Containment & Recovery b. Mechanized Equipment c. Substrate Disturbance d. Beach Cleaners e. Vacuum Pumping f. High Pressure Flushing g. Low Pressure Flushing h. Steam Cleaning 3. Manual Removal a. Collection b. Vegetation Cropping 4. Sorbents 5. Chemical Dispersants 6. Sinking Agents 7. Burning + - Technique has been used or considered Potential Cleanup Impacts 1 - Potential for substrate or physical habitat disruption. 2 - Potential for loss or disruption of community structure. 3 - Potential for increase in likelihood of oil uptake potential. 4 - Potential for increase in the likelihood of oil reentry to the environment. summarizes Me areas of potential harm in the four categories by various cleanup techniques in IS habitat lopes. Substrate or habitat disruption. Some cleanup methods disrupt the integrity or change We quality of a substrate, and, consequently, associated organisms. One example is the use of heavy equipment on some beaches.
172 OCS DECISIONS: ALASKA · Loss or disruption of community structure. Cleanup techniques, such as steam cleaning or high-pressure flushing of a rocky intertidal habitat, can kill organisms that survive the spill itself. · Increased likelihood of oil exposure or uptake. This occurs when the cleanup technique, such as high-pressure, hot-water washing, exposes organisms to additional oil because of increased penetration of oil into sediments. · Increased of! reentry to the environment. This can occur with tech- niques that reduce the viscosity of the oil (hot-water treatment), disturb an oiled substrate to remobilize oil (flushing without proper recovery of free oil), or incorporate oil (steam cleaning) into sediments where, as erosion occurs, it will be reintroduced into the environment. Mechanical removal Mechanical removal methods include the use of special equipment developed to contain and recover oil on water, as well as conventional construction or other motorized equipment (vacuum trucks, front-end loaders) adapted for oil spill cleanup. Mechanical removal can remove organisms as well as sediments from shorelines. There is the potential for damage to surrounding habitats from operation of heavy equipment or from trampling by work crews. Many Alaskan beaches are so remote as to make mechanical cleanup infeasible. D.soersants There has been much research, including several experimental spills (NRC, 1989b), on the effectiveness and the effects of dispersant use. Generally, He tradeoff is between Be relatively short-term effects of a pulse of dispersed oil introduced into the water column versus the long-term effects of of} that spreads out over broad areas, strands on shorelines, and contamirntes wildlife. When applied by aircraft, dispersants, which are less toxic than oil when applied at sufficiently low concentrations, can cover a large slick relatively quicldy and are effective enough to affect the environ- men~ outcome of the spill. One National Research Council panel recom- mended Mat dispersants be considered equally with other response measures
MITIGATION AND REMEDL4TION 173 as a means to increase Me overall effectiveness of spill response (NRC, 1989b). Sergy (1985), in a review of Me results of Me Baffin Island Oil Spill project- a controlled set of experimental spills in Me Canadian Arctic concluded Mat Mere are no compelling reasons Mat dispersants should not be used in Me Arctic when Weir use would result in benefits relative to over options. The American Society for Testing and Materials (ASTM, 1986) has developed guidelines for dispersant use in Be Arctic. Burning In-situ bunting of of] on ice was first developed as a response memos for Be Arctic for of} on Be ice surface. There has been a large continuing program in recent years to evaluate burning as a major on-water response memos. Results indicate this memos is effective (Allen, 1988~. It may be Be most effective means to deal wig of} on ice, in ice leads, and in snow (Sveum and Bech, 1991; Brown and G - ciman, 1986~. For example, MMS has recently supported useful, multiple-sponsored studies of in-situ burning as a spill response (Evans, 1991~. Test burns of of} in ice were conducted in Spitsbergen, Norway, in 1992, and were planned again for May 1993. A large complex in-situ burn study on open cold waters was carried out in August 1993 off Newfoundland. Ioreme~lation Bioremediabon-enhancement of natural biodegradadon processes shows promise as a shoreline post-spill cleanup memos, but furler research is needed to assess its effectiveness under arctic conditions and to make products suitable for Be Arctic. The field of bioremediadon of of! on shorelines is currency under active research and development in bow public and private sectors. shoreline Protection and Cleanup The ~mplicabons of spill responses in various habitat types were discussed in Be 1970s (Westree, 1977; Lindstedt-Siva, 1977, 1979a, b). The major
174 OCS DECISIONS: ALASKA contributions of this work were to identify habitats that were particularly vulnerable to the effects of spilled oil and to assess the effects of various cleanup techniques, as well as the effects of oil. A 1982 workshop, which included international oil spill experts from government, industry, consulting firms, and universities, produced a manual, "Oil Spill Response-Options for Minimizing Adverse Ecological Effects" (API, 1985~. This publication evaluated response techniques in terms of their ecological effects as weld as their effectiveness, and presented recommendations. It also identified environmentally sensitive habitats recommended for high-priority protection during a spill. The manual contains a section on arctic habitats. Ecology-based response planning includes an analysis of a particular segment of coastline, identification of specific environmentally sensitive habitats, and development of plans to protect them to prevent or reduce the amount of of! entering them if a spill should occur (Hum, 1977; Lind- stedt-Siva, 1977; Miche} et al., 1978; Pavia et al., 1982~. Ecology-based response planning's goal is to minimize the environmental damage caused by spills by protecting those areas most vulnerable to longer-term effects. Much of the arctic Alaskan coastline has been analyzed by the state of Alaska and the of! industry (Alaska Department of Fish and Game, 1977; Alaska Clean Seas, 1983~. [XPerlmenial SPIIIS One of Be best ways to study Me fate, behavior, and effects of of} spills, assess the effectiveness and effects of countermeasures, and improve the capability to respond effectively is to conduct experimental spills (controlled releases). In recent years it has become increasingly difficult to obtain permits for such research in US waters. This approach has been recom- mended for basic research, evaluation of response techniques and training (Interagency Coordinating Committee, 1992; NRC, 1993b; Owens et al., 1993~. There are many advantages of conducting experimental spills over using so-called "spills of opportunibr." First, Mere is opportunity to plan the experiment to fit the site. This includes collection of prespill data (something usually not possible at spills of opportunity) and control of He dosage of of} and He exposure of ecological communities. This allows re- searchers to distinguish more subtle changes Dan would be possible during accidental spill events and allows verification of laboratory observations.
MITIGATION AND REMEDIATION 175 One of the most important advantages of well planned experimental spills is that teams of scientists and the required logistical support can be mobilized and dedicated to the project. It is difficult at best to conduct science during the emergency phase of the response to accidental spills. During a complex response operation, there are many priorities (e.g., control of Be spill at the source of the release, vessel stabilization, offshore recovery operations, protection of environmentally sensitive areas) and limited resources (e.g., aircraft, vessels). Science, understandably, will not be a priority during this phase. Nor is it always possible to mobilize the scientific personnel and equipment needed rapidly. This is certainly an important consideration in the Arctic. Experimental spills also are needed to evaluate spill- response technolo- g~es. For example, some testing of mechanical recovery equipment and chemical dispersants can be conducted in test tanks and in the laboratory. However, for full evaluation of these methods at appropriate scales, occasional tests must be conducted in the field. For reasons described above, it is usually not possible to conduct these evaluations adequately during a spill emergency, nor should such evaluation be a condition of using these methods. Finally, experimental spins are needed to train people who must respond to accidental spills. If response crews are never allowed to practice on real oil, they cannot realistically be expected to perform well during an actual spill emergency. Al STO~AT10 During the past decade, restoration ecology has become recognized as an important and integral component of a national interest in conservation (NRC, 1992c). Attempts to restore ecosystems follow environmentally damaging events or a process of degradation; they assume recognition of the effects and a desire to accelerate recovery. Effective restoration requires knowledge of biological detail and the outcome of interactions between species. It is a test of our knowledge of the important interactions that determine the form and functional relationships within communities. It is a challenging task, and often we do not know enough to be successful (NRC, 1992c). On the North Slope, efforts to vegetate disturbed areas with native grasses used by foraging caribou provide an example of research designed
176 OCS DECISIONS: ALASKA to anticipate a needed solution rather Man merely to document injury. Restoration or maintenance of valued resources could also require manipula- tions of animal populations that expand abnormally around areas of development. Given the limitations in our ability to restore damaged ecosystems, and Me difficulty in identifying aspects of marine ecosystems mat might be amenable to restoration, what should MMS do? First, it should investigate available information on Me restoration of arctic ecosystems and identify research needed to protect and restore, where possible, environments at risk of damage from OCS of} and gas activities. Such research should include the identification of those habitats and ecosystems that are unlikely to be amenable to restoration efforts; indeed, it seems safe to say that MMS should focus on nearshore and terrestrial ecosystems rather than on deep- water habitats and pelagic ecosystems. Second, MMS should take advan- tage of restoration efforts of industry on the North Slope and evaluate its experiences. Finally, although this study's scope includes only Alaska, the committee recommends Mat MMS consider restoration as part of its national framework and use experience gained in any region to inform its efforts in other regions. MITIGATING SOCIOI COSMIC I FFI CTS Although MMS-sponsored studies have often been sensitive and accurate in discussing the importance of subsistence activities, Hey have generally been weak in identifying He steps, if any, that could be taken to mitigate negative impacts on subsistence, whether Hose impacts are spill-related or take place in the absence of a spill. In general, He studies tend not to go beyond broad or overall assessments, outlining instead He significance of subsistence to Inupiat ways of life, and pointing out Hat if subsistence resources were ever affected by a spill, He consequences could be almost overwhelmingly negative. The fact that cash obtained from employment is used to complement, rawer Han substitute for, subsistence harvests suggests Hat it will not be easy to mitigate losses in subsistence activity avid monetary compensation. Similarly, many of tile studies describe tile "boom" aspect of tile oil- related cash economy, and Hey project a coming "bust." Some studies note Hat OCS development may forestall, but not prevent, He forthcoming decline. However, none of He studies systematically addresses He potential
MmGA170N AND REMEDY TION 177 for m~tigadng the impact of the coming "bust" on the communities, nor do they discuss potential options for dealing with the more gradual problem of long-term overadaptation or offer options for managing the inevitable impacts. Studies also need to be conducted that address whether the Inupiat community wail be able to revert to a primarily subsistence economy why a declining input of cash, or whether He communities will to some degree collapse. They also need to deal with the questions of whether and how such an effect can be managed. For example, Barrow has accumulated a large i~ucture and associated maintenance costs, as well as a substan- tial debt, a factor that points to the need for further consideration of pragmatic impact-m~tigation options. In order to manage impacts to the human environment, it is critical to determine how to deal wig debt payments as of} revenues begin to decline and ultimately cease. Significant numbers of Alaska Natives have been employed by the borough~ften further contributing to local betterment through the construc- tion of housing, educational, and sanitation facilities; through improvements in health and social services; and through activities that help document, strengthen, and preserve traditional cultures. All of these activities, and more, have been facilitated in part by oil revenues from Crusoe Bay. Depending on the ways in which any potential OCS development would be managed, it could contribute either to a worsening of the eventual "bust," with the cessation of activities in the region, or to the effective mitigation of those problems. As another example, it is possible that of} development in OCS waters could help to sustain oil-related activity in the area by extending Me useful life of the Trans-Alaska Pipeline Systems. Unlike the case for the Pru~hoe Bay and onshore fields, however, developments in federal waters would not be expected to provide Be same amount of income to Be borough or the state because many of the facilities will be offshore, beyond the reach of loch and state taxation. Furthermore, this is a critical component to analy- zing, quantifying, and managing potential effects from the ultimate down- mrn of oil revenues. It is particularly deserving of attention because MMS has influence over whether facilities are located onshore or offshore, and because Be effect could vary greatly depending on whether potential revenues to the North Slope Borough, and the implied effects, are consid- ered in the decision of where to locate facilities. In addidon, attention needs to be devoted to the degree to which continued of! development, if feasible, would in fact contribute to Be long-term well-being of Be region and its people or exacerbate Be difficulties created by the inevitable longer-term
178 OCS DECISIONS: ALASKA "bust." The state of Alaska has been able to establish at least a sizable permanent fund from the relatively short-term bonanza extracted from Pru~hoe Bay, but Me situation for the Norm Slope Borough might be more challenging; although tile borough has set up its own permanent fund, it currently stands at only $259 million. Even wad substantial additions during the remaining years of operation at Prubhoe Bay, the fund is likely to prove far too small to support anything like the current level of borough activities on a continuing basis. The situation is further complicated by the fact that Be borough's long-term bonded indebtedness as of July 1993 was $72X million. CO~ICLUSIO~S ED JACOBI EDITIONS The specific conclusions and recommendations for this chapter follow. For Be general and overall conclusions of this report, see Chapter 8. Conclusion I: The development of a good data base on gouging rates and the inhaling process requires yearly replicate measurements of the gouging patterns on selected representative regions of the Alaskan shelf. Although some such data exist (Barnes et al., 1978; Rearic et al., 1981), Be expan- sion of this data base would be highly desirable and would contribute significantly to safe pipeline design. Recommendation I: Collect yearly replicate measurements of gouging depths and patterns once an oil field has been identified. MMS should support a small but long-term (5 years minimums program Hat collects annual replicate measurements of gouging depths and patterns Tom selected, representative regions of the Beautort and Chukchi shelves. Of particular initial importance now is the offshore area in the vicinity of Be Kuv~um discovery in Be Beaulort Sea. Dough a Kuv~um study will not replace site-specific information from other areas, it should contribute significantly to pipeline route selection and to safe pipeline design. The personnel requirement for this study would be 3 persons per year for each 5-year study. Altemative: The alternative to adequate temporal data is sigruficant
MITIGATION AND REMEDL4TION 179 overdesign of the pipeline either by deeper burial than actually required or by armoring. Conclusion 2: Information is needed on the rate of gouge infilling and on potential scouring problems along pipeline routes as Hey are identified. Recommendation 2: Ed observanonal data base on the temporal variations u' near-sea currents. MMS should support a program that expands the observational data base on the temporal variations in near- seafloor currents and in sediment transport for both the Beautort and Chukchi shelves. Measurements should be taken at several water depths at representative locations over a 3-year period. The personnel requirement for this study would be IT person-years per year for a project total of 4~/2 person-years. The deployment and recovery of buoys could be carried out as a part of the study discussed as in Recommendation I. Akerr~ve: Estimates might be provided by attempting to model both the currents and the resulting sediment fluxes. However, in the absence of adequate field data, there would be no way to evaluate the adequacy of the estimated values. In that the oceanographic conditions along both the Beaufort and Chukchi coasts are believed to be com- plex, field data are probably essential. The ultimate consequence of inadequate offshore pipeline design could, of course, be a spill below the pack ice. The social, economic, political, and ecological conse- quences could be devastating. Conclusion 3: There is a need for further studies Hat Will contribute to the development of effective remote sensing techniques that are capable of locating oil hidden either within or, more commonly, below sea ice. Recommendation 3: Develop remote-sensing techniques to locate of! within or zanier ice. Preliminary assessment of currently available technology would require 2 years and a total of 4 person-years of effort, assuming Hat some of the instrument systems are operated by contractors. The cost of developing a selection of effective systems for detecting oil under ice might be in the range of $3-5 million,
1&0 OCS DECISIONS: ALASKA including design, construction, and testing. This is a very rough estimate, because no truly promising technologies have yet been iden- tified. Acoustic methods derived from military submarine develop- ments might be promising, now that some are potentially available for development and application in the private sector under cooperative research and development agreements. Alternative: There are no effective alternatives. Drilling pilot holes in the ice with hand equipment, which could be the only current alter- native, is ineffective over a large spill area. Conclusion 4: It is necessary to integrate all relevant available data sets so that decision makers are ideally informed. Recommmdation 4: Use data sets from the Cared North, Barents Sea, and Prince William Soundly Additional integration of data from other places, industry data, data from the Prince William Sound spill, and data from government-sponsored studies (e.g., the Anchorage Symposium on the Prince William Sound Spill sponsored by the National Oceanic and Atmospheric Administration) should be synthe- sized and evaluated. Also, data released by the Exxon Corporation for the Prince William Sound spill at the American Society for Testing and Materials Atlanta Conference in April 1993 should be evaluated. This effort would require ~ person-year. The neglect of applicable studies that have been carried out at other locations by other concerns is, at best, a waste of both time and money. Alternative: None recommended. Conclusion 5: Regardless of probability, of} spills are a major concern in the Arctic because of He perceived or real inability to clean them up or control them. A few of the many techniques used in temperate regions have been tried in experimental tanks with cold water. There is only limited experience in cold and ice-infested waters wig experimental spills. Recommendation 5: MMS should attempt to evaluate critically the relative effectiveness of different countermeasures as applied under a representative range of conditions that ought be expected in the Arctic. Countermeasures could include in-situ burning, use of cold water
MITIGATION AND REMEDIATION 181 dispersants, bioremediation, mechanical recovery, and protective procedures. This study should assist in formulating critical questions that should be answered by appropriate field tests. This study should require 2 person-years of effort, and should be cognizant of work carried out on this theme in Canadian and Norwegian waters. Alternative: None recommended. Conclusion 6: Because of He difficulty of getting permits to apply of} to shorelines or to spill of! in open water, most spill effects studies have been part of larger programs to study Me effects of accidental spills, or "spills of opportunity.'' Although much valuable information has been gained, Were frequently is no opportunity for rigorous science avid adequate control or reference sites to detect Be effects of Be spill or the control/cleanup mesons used. Also, resources needed to conduct proper scientific studies (boats, aircraft) are in demand to perform Be cleanup and control. Experimental spills are planned in advance, and can be controlled. Reference sites and tile number of replicate samples and plots are deter- mined in advance. Recommendation 6: MMS should be a major participant In the design and completion of one or more experimental of! spills in the general area of the Ch~chi and Beaufort seas. We believe Cat experimental spills are essential because they can contribute to Be scientific understanding of processes, and fill data gaps, about Be interaction of ice and oil. These tests also can contribute to accurate assessments of the abilities and limits of countermeasures, and remediation and restoration procedures. The data from recent tests should be evaluated. Carefully controlled field tests would be an in- valuable extension of such work into a realistic environmental setting. Experimental spills also are essential in the training of local industry and community spill teams. The performance of spill response teams cannot be expected to be excellent if Key are never permitted to practice using real oil. Also, preimpact data could not be collected without an experimental spill and would prevent proper evaluation of spill, control, and cleanup effects. An estimate of tile cost of an investigation of an experimental shoreline spill would be $! million to $~.5 million per year, carried
182 OCS DECISIONS: ALASKA out at years 0, 1, 2, 3, and 5. An open-water or ice spill investigation carried out for 1 year might cost $1 million to $2 million; a 2-year investigation would probably be needed. This activity would require 3 person-years per year for the shoreline investigation, and 4 person- years per year for an open water-ice spill trial. Altemative: None recommended. Conclusion 7: Outside Me U.S., the individual agencies and companies involved in arctic exploration and development are conducting experimental spill studies wig little interaction wig MMS. Recommendation 7: Institute cooperative research panning and furring of expenmental spid shies. We generally recommend fewer, more comprehensive projects rawer than a large number of small projects. SpiDs should be used for a variety of purposes simultaneous- ly (research, evaluation of countermeasures technology, training) and supported cooperatively by federal and state agencies and industry. Alternative: None recommended. Conclusion S: It is essential Cat EISs be current wig state and federal legislation. Recommendation S: Ensure thm ElSs reflect the new fedleral and state legislation. Regulatory requirements greatly influence requ~re- ments for of} spill preparedness and response. This would require l/2 person-year. Conclusion 9: MMS's approach to mitigation and remediation project selection has largely been from Me "bottom up." Useful information has been obtained from these studies. MMS should now focus its mitigation and remediabon aches on He most critical potential sources of environmental damage. The discussion in Chapter 6 clearly identifies Cat ~adidonal knowledge should be used in study design and consulted for lease decisions. Recommendation 9: Take a top~own approach, i. e., an approach in which an overall scientific vision or framework influences the
MITIGATION AND REMEDlATlON 183 studies program, in selecting mitigation am remediation studies to be corrected. It is now opportune for MMS to make a transition to "top- down" sh~ies that are critically targeted to specific industry activities. Ibis approach should optimize the chances that, with limited funding, support will be given to projects that would contribute significantly to environmentally safe, efficient, and timely development of offshore resources. These studies should be integrated with continuing long- term studies of Arctic ecosystems and processes. There should be no increment resource requirement to implement this recommendation. It will be more cost effective to fund lower-level, long-term studies using disciplinary experts than to respond sequentially to crises. This latter approach is financially exorbitant, politically costly, and does not create a positive working relationship between agencies and the affected public. Alternative: Research will be in response to crises and the resultant high-cost, short-term studies will be relatively ineffective. Conclusion 10: Most of MMS's research on oil spills associated with lease sales has focuses! on documenting potendal effects rather than on preventing them or minimizing their consequences. MMS does have a program on countermeasures that is not associated with its lease sale programs, and this program has contributed valuable information about in-situ burning and Me use of dispersants and other chemical agents. Recommendation 10: Develop specific countermeasures to minimize the effects of arctic spills. The next step is to look at particular lease- sale areas and develop specific countermeasures. More research is needed on in-situ burning on ice, in ice, and on cold water. The ecological tradeoffs involved in using cold-water dispersants need to be evaluated and compared with other spill control technologies in We Arctic. The effectiveness of mechanical recovery systems should be evaluated uglier arctic conditions. Shoreline cleanup and remediation, including bioremediation methods, also should be studied. This effort, shared with other organizations, might have a cost to MMS of $5 million and require 3-5 years to implement. Alternative: None recommended.
184 OCSDEClSIONS: ALASKA Conclusion 11: Gaps still exist in fate, behavior, and effects data bases; therefore, the accuracy of models used is dubious. Recommendation 11: Fate, behavior, aM effects studies are needed both tofid ~ gaps for arctic environments and to improve or verify models. Experimental spills are He preferred way to acquire this information, but spills of opportunity also should be studied as appropriate and a science response plan should be developed to take advantage of Be study opportunities Key provide. More work is needed on developing and verifying models Cat assess fate, behavior, and effects of spills. Models should always be compared wad field data and refined accordingly. The financial cost of this effort might be $! to $~.5 million; duration cannot be estimated because Be committee cannot know when Be opportunity might arise. Alternative: None recommended. Conclusion 12: MMS has not fully apprised Norm Slope Borough communizes and oil industry operators of its research, nor has it encour- aged Weir participation in development of research programs. Recommendation 12: North Slope Borough communities and of! industry operators should be kept informed about the progress of MMS research on of! spill response and other scientific studies and, as much as possible, participate in the development of research programs. Alternative: None recommended. Conclusion 13: The Alaska OCS oil and gas leasing process suffers from Be same sort of environmental gridlock found in over regions. Recommendation 13: MMS needs to review st~ies on "environmen- tal gridlock" and incorporate lessons from these studies into the dEecision-rr~ng process. These studies demonstrate that development of a process based on mutual trust is an integral component of tile solution. This process requires a genuine sharing of power, a sincere effort to seek accommodation with local groups, and should begin wid a recognition dial many answers about lease-stage and development- stage impacts are not yet known.
MmGAlION AND REMEDIAT1ON 185 If properly implemented, such an approach could permit MMS to work cooperatively with affected groups in a good-fai~ effort to develop relevant answers where possible and to develop mutually acceptable procedures for dealing why residual and unavoidable uncertaindes where answers are not likely within a reasonable period and at reasonable cost. Alternative: None recommended.