EPA's Alaska oil spill bioremediation project Final part of a five-part series P.Hap Pritchard U.S. Environmental Protection Anencv Environmental Research L a b o r l t o j GulfBreeze, FL 32521 Charles F. Costa US.Environmental Protection Agency Environmental Monitoring Systems Laboratory Las Vegas, NV 89293 The bioremediation of hazardous wastes, although not a new concept, h recently taken on new significance as increasingly effective and potentially i expensive cleanup technology (1, 2). contributory factor has been a mol= comprehensive understanding of the natural biodegradation potential of microbial communities in many different types of habitats (3-S). Several field demonstrations have now convincingly utilized laboratory test information to develop effective treatment strategies for bioremediation (9, 20). For similar reasons, the significance and feasibility of tl bioremediation of oil-contaminated e vironmental materials has also ii creased. In addition, major oil spills like th-. of the Exxon Valdez have galvanized public attention to alternative cleanup technologies. Oil biodegradation in aquatic (marine and freshwater), terrestrial, and groundwater environments has been extensively studied in laboratory systems over the past 20-30 years (11-25). but it is only recently that this information has been considered for large-scale bioremediation efforts ( 4 , 15-27). Definitive success in the restoration of gasoline-contaminated aquifers (18-22) and oil-contaminated soils (IS. 23, 24) has occurred because of the application of this basic research data base. Enhanced oil biodegradation Enhancing biodegradation processes to assist in the cleanup of oil spills in marine environments has been suggest372 Environ. Sci. Technoi., Vol. 25, No. 3, 1991
ed several times, with much emphasis on the treatment of oil on the sea (27, 25). Several approaches have been discussed and debated. For example, the idea of accelerating oil biodegradation rates by increasing the availability of oil to bacteria through the use of dispersants (bacteria degrade oil primarily at oil-water interfaces) has been tested with mixed results, giving a confusing picture as to the ultimate effectiveness of this treatment (26-29). The concept is complicated by potential problems of dispersant toxicity to both bacteria (30, 32) and other aquatic life (32, 33). Seeding oil-contaminated areas with
hydrocarbon-degrading bacteria has also been considered as a possible bioremediation option in both aquatic (34, 35) and terrestrial (36, 37) environments, but again the results have been mixed. Recent activities associated with the Mega Borg and Apex barge spills (38) have drawn attention to the problems associated with seeding floating oil with commercially available mixtures of hydrocarbon-degrading bacteria. The efficacy of this approach has yet to be definitively demonstrated because of the exceedingly complex logistical problems and ambiguity of open-water monitoring.
This article not subject to U S . wpyright. Published 1991 American Chemical Society.
Previous studies illustrate these difficulties; in one case seeding worked when performed in saline ponds in the Arctic (35) hut was ineffective in artificial marine enclosures (34, 39, 40). In another case, the added organisms may have produced extensive quantities of emulsifying agents, causing the oil to rapidly disperse before extensive biodegradation occurred. Reisfield et al. (41) demonstrated some years ago that the addition of a hydrocarbon-degrading Arihobucter species, with strong emulsification capabilities, caused oil in an enclosed chamber to disappear from the water surface in one day without much degradation. Despite the fact that this process produces dispersed oil droplets that are colonized with oil-degrading bacteria, it is not clear that this is a viable cleanup procedure. Seeding of oil-contaminated shorelines is also possible hut has rarely heen tested (40.42). Experiences with soil inoculation, where conditions are more amenable for this approach, have shown only marginal success, in part because of the presence of indigenous bacteria with very effective oil degradation capabilities (36,37). By far the simplest approach for enhancing oil biodegradation is the addition of nitrogen and phosphorus nutrients in a well-oxygenated environment. It is well known that enrichments of oildegrading microorganisms occur rapidly following oil spills in most environments. But in the face of a large amount of degradable oil carbon, biodegradation quickly becomes limited by nutrient and oxygen availability (10-16). Numerous laboratory and field studies show that attempts to overcome these limitations generally lead to successful enhancements of oil biodegradation (4, 12, 14, 16, 17). Surprisingly, this bioremediation approach has never before been used on a large scale to directly assist in the cleanup operations that commence following a major oil spill.
Office of Research and Development (ORD) that hioremediation was a reasonable option despite the complexity of the environment. Thus the Alaskan Oil Spill Bioremediation Project was initiated [preliminary accounts of this effort have been published (43, 48-53)]. This project resulted in the most expansive use of hioremediation yet undertaken. The possibility of using bioremediation to assist in the cleanup of oilcontaminated beaches in Prince William Sound was supported by recommendations arising from a feasibility workshop sponsored by ORD in late April 1989. As EPA already had a large research program in bioremediation (54), including the necessary technical personnel and in house contractors, it could rapidly respond to the workshop recommendations. Consequently, a research plan was develop,^ .~ determine if the addi-
-
. . . subsugace oil may only be
treatable by bioremediation . . . 111
tion of fertilizers would accelerate oil hiodegradation sufficiently to promote the consideration of bioremediation as a supplemental cleanup tool. The concept was to perform an initial field demonstration of hioremediation and then, if it was successful, make recommendations to Exxoo. In addition, EPA would provide a follow-up field study as proof of the success of large-scale application. Spill of opportunity A team of experts from the different The Erron Vuldez oil spill in Prince research laboratories within the ORD William Sound, Alaska (March 25, (see acknowledgments for key personnel and their roles in the project) was as1989). provided a unique opportunity to test the feasibility of hioremediation sembled to implement the field demon(43). Within days following the spill, stration project. It was the effort of this several million gallons of Prudhoe Bay team that set the stage for the overall success of the project. Exxon generouscrude oil, an oil that had been the focal point of several previous biodegradation ly provided, through a Technology studies in cold water environments (35, Transfer Cooperative Agreement, fman44-47), had contaminated almost 300 cial and operational support (43). withmiles of rocky coast line in Prince out which the project would not have William Sound. This presented Exxon, been possible. the state of Alaska, and the US. Coast Guard with the largest cleanup problem Test beaches and fertilizers In early May 1989 the following ORD in US.history. As a variety of cleanup options were assessed and implemented, laboratories were mobilized: Las Vegas, Cincinnati, OH; it became clear to scientists at EPA’s NV, Gulf Breeze, n,
Athens, GA; Research Triangle Park, NC; and Ada, OK. Two sites were selected, Snug Harbor and Passage Cove (50,511. These beaches were comprised mainly of large cobblestones overlying a mixed sand and gravel base. The Snug Harbor beaches had a moderate degree of oil contamination confined to a broad hand within the intertidal zone. At the initiation of the project, no oiled beaches were available that had been washed by the Exxon process (high-pressure water jets and water heated to 140 OF) to physically force the oil off the beach surface on to the water for collection. Thus, this study site was chosen to approximate beach conditions following physical washing. Passage Cove, on the other hand, was a heavily oiled beach that Exxon had physically washed, thus removing the bulk of the oil and spreading the remaining oil over beach surfaces. Both beaches had a thm layer of oil covering the surface of the cobblestone, as well as oil mixed into the sand and gravel under the cobble to varying depths. Selection of fertilizers, which had to be done quickly, was based on considerations of application strategies, logistical problems for large-scale application, commercial availability (particularly if large-scale application became reasonable), and the ability to deliver nitrogen and phosphorous nutrients to the microbial communities on the surface and the subsurface beach material for sustained periods. Three fertilizer application strategies were adopted for testing: commercially available slow-release formulations, an oleophilic fertilizer, and water-soluble fertilizer applied as a solution (51). The concept behind the use of slowrelease fertilizer formulations was to apply a commercial product with the appropriate release rate characteristics to the beach surface, and then allow tidal action to disperse the released nutrients over the contaminated area. The product had to deliver sufficient quantities of nutrients while on the beach for several weeks. The two slow-release fertilizer formulations selected were briquettes (about the size of charcoal briquettes) and granules (about 2-3 mm in diameter). The briquettes, produced by Vigor0 Industries (Fairview, IL), have an N P K ratio of 14:3:3 and release urea by the spontaneous chemical hydrolysis of isobutylidene diurea as it leaches from the inert briquette matrix. Urea is assumed to be rapidly converted to ammonia by indigenous microbial urease activity. Phosphorus was supplied by similar leaching of a citric-acid-soluble phosphatic fertilizer contained in the briquettes. Fertilizer granules (Customblen), Environ. Sci. Technol.. Vol. 25, No. 3,1991 373
produced by Sierra Chemicals (Milpitas, CA), have an N:P:K ratio of 28:8:0 and slowly release ammonia, nitrate, and phosphate from inorganic ammonium nitrate and ammonium phosphate encapsulated inside a diene-treated vegetable oil coating. Laboratory studies with both of these products (49, 50) verified that after an initial rapid release of nutrients, remaining nitrogen and phosphorus were released slowly over approximately 3 4 weeks. The fertilizer granules were broadcast on the beach surface at a concentration of 90 g/mz using a mechanical seed spreader. Their high specific gravity, propensity to adhere to the oil, and tendency to entrain under rocks and in interstitial spaces (Figure I), ensured that they would remain on most low- and moderate-energy beaches in Prince William Sound. The briquettes, on the other hand, were packaged into 26 seine net bags (30 Ib each) and placed on the beach in two rows, one at the low-tide line and the other at mid-tide. The net bags were held in place by rope attached to buried steel rods (Figure 2). The rationale behind the use of oleophilic fertilizers was to “dissolve” the nutrients into the oil by applying the material as a liquid to the oiled beach material. This sequesters nutrients in the oil phase, facilitating bacterial growth on the surface over sustained periods. Several formulations of oleophilic fertilizers have been successfully tested in both laboratory and small-scale field experiments. Most of these experiments were designed and executed with the idea of treating oil on the surface of water rather than oiled beach material. The pioneering studies of Atlas and Bartha (55) demonstrated that the addition of paraffinized urea and octylphosphate (as sources of N and P) to a crude oil on the water surface significantly enhanced the biodegradation of the oil. This enhanced degradation was further documented following the treatment of oil in arctic tundra ponds (35). Similar success has been reported for other oleophilic formulations, including the slow release of nitrogen and phosphorous from MgNH,PO, incorporated into a paraffin support base (56);a commercial organic phosphate fertilizer, Victawet 12 (2-ethylhexyl-dipolyethylene oxide phosphate) (57); several natural sources of lipophilic nitrogen and phosphorous, such as soy bean lecithin and ethyl allophanate (58); and oleophilic nutrients with mild surfactant characteristics (59). Despite uncertainties regarding specific mechanisms underlying enhanced oil degradation, the factors affecting physical release rates of the nutrients in situ, and the addition of large quantities of organic carbon to the 374 Environ. Sci. Technol., VoI. 25. No. 3. 1991
FIGURE 1
Slow-release fertilizer granules stuck to oiled beach material in Passage Cove
FIGURE 2
Bags of slow-release fertilizer briauettes positioned on an oiled test beach in Snug Harbor
oil, all products seem to enhance the rate and extent of oil biodegradation. However, in some cases it has been observed that oil degradation is not substantially greater than that resulting from the addition of inorganic nutrients alone (25. 601. Therefore, the application of oleophilic fertilizers should be dictated by the site-specific environmental conditions that prevent nitrogen and phosphorous nutrients from remaining at the site of microbial activity for sustained periods. We selected the oleophilic fertilizer Inipol EAP 22, produced by Elf
Aquitaine Company (Paris, France). It was the only commercially available oleophilic fertilizer that met our aforementioned criteria and that could be produced quickly and in large quantities. This unique product is a stable microemulsion consisting of a core of urea (the nitrogen source) surrounded by an oleic acid carrier. Laureth phosphate (a surfactant and the source of phosphorus) is incorporated as a stabilizer, and butoxy ethanol (methyl cellusolve) is added as viscosity reducer. It has shown promise in laboratory tests and in large
outdoor tanks using oil-contaminated FIGURE 3 beach material of varying consistency Application of the oleophilic fertilizer, using a back pack sprayer, and environmental location (40, 61-63). to oiled test beach in Snug Harbor Some of the best field results were obtained with beach gravel from Spitsbergen, Norway (63j. It has recently been tested on sandy beaches in Nova Scotia using in situ enclosures (601 but with little success compared to the addition of agricultural fertilizers. EPA laboratory studies showed that after application of Inipol EAF’ 22 to oiled beach material, 5 M O % of the ammonia and phosphate was released within the first few minutes, followed by a slower release over a three-week period. lnipol was applied in Prince William Sound using a backpack sprayer (Figure 3) to thinly coat the oiled beach material at an application rate of approximately 0.1 gum*. The third type of fertilizer application regimen evaluated was spray irrigation using an aqueous nutrient solution. This application was conceived as a method to introduce nutrients into the oiled beach material, particularly oil below the surface, in the most defined, coutrolled, and reproducible way. This was FIGURE 4 accomplished by dissolving commer- Application of fertilizer solutions from a sprinkler system on a cially available sources of ammonium test beach in Passage Cove nitrate ( 3 W ) and triplephosphate (0454) into seawater pumped from below the beach. The resulting fertilizer solution was then sprayed over the beach at low tide using a pump and lawn sprinkler heads (Figure 4). The fertilizer application rate was 0.4 in. of water applied over four h, which resulted in an approximate concentration of 6.9 g of Nlm’ and 1.5 g P/m2. Results from fertilizer application The first application of fertilizer occurred July 8, 1989, at the Snug Harhor site. Inipol EAP 22 and the fertilizer briquettes were applied to separate 28-m x 14-m plots. Approximately 2-3 weeks later, the cobblestones on the beach treated with the oleophilic fertilizer were visibly cleaner. This produced a striking “window” of clean surface on the oiled beach (Figure 5). Close examination of the beach showed that significant quantities of oil remained under the cobblestone as well as within the beach subsurface. However, even this oil seemed to disappear slowly over the next few weeks. ’Ibis contrasted with the untreated control areas and the beaches treated with the fertilizer briquettes, in which there was little visual change. Subsequent laboratory studies verified that Inipol was not a chemical rock washer. To obtain definitive information on the role of biodegradation in this event, beach material was sampled according to a block design (21 samples taken at each sampling time, 7 each in contigu-
ous blocks along l i e s corresponding to the high-, mid-, and low-tide zones of the beach). Samples of oil extracted from the surface of the cobblestone on the oleophilic fertilizer-treated beach showed that this disappearance of the oil was accompanied by significant decreases in total oil residues (i.e., weight of extractable material). However, decreases of total oil residues could occur because of other nonbiological processes, and thus changes in hydrocarbon composition that are indicative of biodegradation must be mea-
sured. This is done historically (11) by examining the weight ratio between hydrocarbons known to readily biodegrade (generally the C-17 and C-18 alkanes) and those that biodegrade more slowly (such as the branched alkanes, pristane and phytane) but that are very close in gas chromatographic behavior. This marker hydrocarhon concept is based on the idea that most nonbiological fate processes (physical weathering, volatilization, photolysis, etc.) will not differentially affect aliphatic hydrocarbons having similar gas chmmatopphic Environ. Sci. Technol., Vol. 25, No. 3, 1991 375
behavior. Correspondingly, Figure 6 shows the extensive decline in the C-18:phytane ratio through time for the oleophilic fertilizer-treated beach. A decrease of approximately 70% occurred during the first 4 weeks. As this extent of differential loss of marker hydrocarbons can only be caused by biodegradation, it follows that the decreases in oil residue weight on the cobblestone SUTface were also due to biodegradation. Note, however, the large error bars around the mean values in Figure 6. This large variability (that is, many samples show evidence of biodegradation while others show much less) is a function of the highly heterogenous distribution of oil on the beach. The rate of hiodegradation was probably similar in each sample, but because biodegradation takes place on the oil surface, a sample containing a large amount of oil (with less exposed surface area) will show a lower percentage of degradation. Decreases in the C-18:phytane ratio also occurred on the briquette fertilizertreated beach (data not shown) but to a lesser extent (approximately a 30% decrease from the original value in 4 weeks). Thus, the lnipol enhanced biodegradation to a greater extent than did the briquettes, accounting for the substantial visual difference in the treated beaches. We suspect that after a certain extent of oil degradation was achieved, the oil became less sticky and flakier; this innocuous degraded material was then easily scoured from the rock su1 faces by tidal action. To our surprise, however, oil samples from the control beach showed decreases in the C-18:phytane ratio that were approximately equal to those seen in the beach treated with the fertilizer briquettes. The amount of nutrients supplied by the briquettes, therefore, was probably no greater than that supplied naturally in seawater as it washed over the oiled rocks at each high tide. Monitoring data for the water overlying both beaches showed low but detectable concentrations of ammonia and phosphate. Continued observations at Snug Harbor showed that by August 1989, all of the beaches (treated and untreated) looked cleaner. Thus, the use of the fertilizer had accelerated the cleanup process by approximately two months. This could be valuable for beaches with heavier amounts of oil contamination, in that cleanup can be accomplished during the short summer when warmer water temperatures encourage greater biodegradation. Effects of the fertilizer applications on the oil-degrading microbial conununities in the beach material were also examined. Samples were periodically taken to determine the concentration of 376 Environ. Sci. Technol., Vol. 25, No. 3,1991
FIGURE .~5 ~
Effect of the oleophlllc fertilizer on the biodegradation of surface oil in Snug Harbora
=The clear "win
efines Dreciselv where the fenilizer was awlied.
I io 01 ne(C ind sure of oil bioaegaavaaion) tnagUYgh time tion of the oleophilic fertilizer June 8,
_ _ _ _ _ _ _ _ _Fresh _ _ _Prudhoe _ _ _ _Bay _ _crude - - - -oil- - - - - - - -
I
mg date 'amis repressntthe mean and standard deviaaon Of 21 t e a
oil-degrading microorganisms. This was accomplished using a dilution-to-extinction method and an endpoint based on a microhially mediated change in the consistency of oil floating on the surface of nutrient medium. Because of high variability in the numbers of bacteria in each sample it was not possible to show statistically significant increases in the
oil-degrading microbial populations as a result of the fertilizer addition. This was further complicated by the awareness that the numbers of oildegrading bacteria in the oiled beach material before exposure to fertilizers were very high, averaging 1-10% of the total bacterial population. These high numbers represented an enrichment of
oil-degrading microorganisms of approximately lo3 to lo5 as compared to beach microorganisms not exposed to oil, Hence, the beaches were well primed for the bioremediation. A monitoring program established in Snug Harbor to check for potential adverse ecological effects verified the safety of bioremediation as a cleanup tool (53).To determine the potential for eutrophication, ammonia, phosphate, chlorophyll, bacteria, and primary productivity were measured in the water column directly offshore of the fertilizer-treated beaches and in control areas distant from the test plots. Measurements in the experimental area were not significantly different from the range of values observed in control areas. In addition, ammonia, the only component of the fertilizers acutely toxic to marine animals (based on laboratory bioassays with indigenous and surrogate test species), never exceeded toxic concentrations. Mussels suspended in floating cages just offshore from the treated beaches showed no bioaccumulation of oil residues, which supports the contention that bioremediation activities did not cause any release of oil from the beaches. Finally, genotoxicity assays demonstrated that the toxicity of fresh Prudhoe Bay crude oil was ameliorated by bioremediation and that no toxic degradation products were formed.
Recommendations, subsequent testing By the middle of July 1989, results from our demonstration project convinced Exxon to consider the use of bioremediation on a large scale as part of their cleanup effort. We recommended that the oleophilic fertilizer, Inipol, be applied to surface oil on the beaches, as this appeared to be the most effective use of this fertilizer at the time. We further recommended that the combination of Inipol and the fertilizer granules (Customblen) be applied where there was both surface and subsurface oil, as the granules were a simple means of releasing nutrients that could be flushed down into the beach subsurface by tidal action (the briquettes were not recommended because of the operational and logistical problems in the preparation and application of the seine net bags on a large scale). Exxon, after gaining approval from the appropriate regulatory groups, began fertilizer application in early August to approximately 50 miles of beach in Prince William Sound that had been physically washed. Increasing the biodegradation rate of oil at this point was very important in order to achieve maximal degradation before winter conditions curtailed cleanup operations. In many cases, the results of large-
scale fertilizer application were as dramatic as our test plot observations at Snug Harbor; that is, where the oil was spread thinly over the cobble surface (as was the case on many beaches that had been physically washed), the oil had virtually disappeared in less than 20 days. With time, it was also apparent that on many beaches, even oil underneath the cobblestone had also disappeared but to varying degrees depending on the extent of oil contamination. Although it is difficult to prove experimentally, we believe that the physical cleaning process used by Exxon dispersed the oil throughout the beach material to such an extent that the exposed oil surface area was greatly increased, allowing greater bacterial colonization and subsequent biodegradation. This response may be analogous to the stimulating effect that physical agitation has on oil biodegradation rates during soil tilling procedures used in land farming treatment (15, 16). The Passage Cove study was initiated as the definitive technical support site for the large-scale application of Inipol and Customblen fertilizers. These fertilizers were applied in combination in late July 1989 to a large test beach, and samples of beach material were analyzed for changes in oil residue weight and aliphatic hydrocarbon composition. These changes were compared to those observed in an untreated control beach. In addition, a beach treated with a seawater solution of inorganic fertilizer (applied via a sprinkler system) was examined in the same way. In approximately 2 to 3 weeks, oil on the cobble surfaces of the beaches treated with a Customblen-Inipol combination and fertilizer solution had been degraded to the point of producing visibly cleaner surfaces much as we had seen in Snug Harbor. The control beach, however, did not look cleaner. Disappearance of oil from the rock surfaces on the beach treated with the fertilizer solution definitively proved that biodegradation (and not chemical washing) was responsible for the oil removal, as there was no other reasonable mechanism to explain this loss of oil. Despite sampling and interpretation complications resulting from the high variability in oil distribution on the beaches, we have been able to show statistically that oil biodegradation (as measured by changes in residue weights and oil chemistry) was significantly greater on the beach treated with the fertilizer solution than it was on the control beach. Based on this information, we projected that after 45 days approximately three to four times more oil remained on the control test beach than on the fertilizer solution-treated beach. This
corresponded to an enhanced biodegradation rate of about two- to threefold. Results were similar on the Inipol-Customblen-treated beach except statistically significant differences were more difficult to establish. However, it appeared that accelerated biodegradation (approximately a two- to threefold increase) occurred early in the test when nutrient concentrations were highest. The striking results of the fertilizer solution application strongly support the idea that oil degradation in Prince William Sound was limited by the amount of nutrients. The results also imply that fertilizer reapplication (maintaining nutrient concentrations at high levels for long periods) is important. Nutrient-monitoring studies on the fertilizer solution-treated beach in Passage Cove showed that low concentrations of ammonia and phosphate accumulated in the beach subsurface and persisted (in the absence of further application) for at least 4-5 days. This nutrient-laden water continued to expose oiled beach material to nutrients with each incoming tide. These data suggest that fertilizer solutions can be applied less frequently, thereby simplifying the fertilizer application.
Ensuing activities The long-term benefit of fertilizer application was realized during examination of the beaches in Passage Cove in November 1989 and early June 1990. Very little oil was observed on either of the treated beaches at the surface or in the subsurface (12-in. depth). However, the untreated control beach still showed areas of heavy oil contamination in the subsurface beach material. These observations provided the final definitive demonstration of the long-term success that can be expected from bioremediation of oil-contaminated beaches. In the spring of 1990, a cleanup plan for the remaining oil-contaminated shorelines in Prince William Sound was agreed upon by the U.S. Coast Guard, Exxon, EPA, the Alaskan Department of Environmental Conservation (ADEC), and other resource agencies and landowners. Bioremediation was an integral part of the plan, both as a primary and secondary method of cleanup. However, at that time, ADEC still did not feel fully confident that bioremediation was as effective as the results seemed to indicate. They consequently asked that further substantiation be provided early in the summer of 1990. The resulting information would determine their decision to continue with bioremediation as part of the cleanup plan. To that end, a joint bioremediation monitoring program was conceived and implemented by scientists from Exxon, Environ. Sci. Technol., Vol. 25, No. 3, 1991 377
EPA, ADEC, and the University of Alaska using logistical support and resources from Exxon. The concept was to select several beaches where planned cleanup operations were underway and monitor for increases in oil-degrading microbial activity and oil degrader biomass, relevant changes in oil chemistry, and any adverse ecological effects. Much of the emphasis centered on the ability of bioremediation to work effectively on oil in the beach subsurface (0.3-1.0 m depths). By July 1990, it was clear from the monitoring program that fertilizer addition had stimulated oil-biodegradation activities by three- to fourfold and that the biodegradation was affecting the removal of more than just the easily degradable aliphatic hydrocarbons from the oil. In addition, the observed sustained levels of nitrogen and phosphorous in interstitial water corresponded with stimulation of microbial activities in the beach subsurface to a depth of at least 70-80 cm. With the absence of any measurable adverse ecological effects, the ADEC agreed to continue with bioremediation as part of the cleanup plan throughout the summer. This paved the way for multiple reapplication of the ~
Get the wholepicture from the world's most cited chemical publication
I
JOURNAL OF THE AMERIG4N CHEMICAL SOCIETY
1
Editotor: Allen J. Bard University of Texas Each biweekly issue of JACS gives you authoritative, original articles (plus concise communications hot off the press!) in every key are??aof chemical research. What's more, JACS is the most competitivelyprlced journal of its kind. Order your subscription today! Call Toll Free 1-800-227-5558 Telex: 440159 ACSPUI or 892582 ACSPUB Cable Address: JEICHEM
American Chemical Society 1155 16th Street, W , Washington, D.C. 20036 202-872-4363
378 Environ. Sci. Technol., Vol. 25,No. 3, 1991
fertilizers, a necessary step in many areas where large quantities of oil remained. Results from the monitoring program and additional field research conducted by EPA (with the support of Exxon) suggested that as the summer progressed, bacterial colonization of the oil had increased to such a degree that biodegradation rates were increased even further. We believe this may have been caused, to some extent, by the increased surface area of exposed oil that occurred as a result of impregnation of the oil with fine glacial till. In addition, EPA research has shown that oil biodegradation can be enhanced and sustained over several weeks by a single exposure to fertilizer for a few hours (sprinkler system). This significantly changes our concept of bioremediation; sustained nutrient input to oil-contaminated beaches may not be necessary for maximal biodegradation of the oil. A more detailed description of these results will be published in the near future. A reassessment of the beaches in spring 1991 will further verify the significance of these observations.
Summary and conclusions The results from our oil spill bioremediation project show that the presence of oil on the beaches of Prince William Sound, Alaska, caused a significant enrichment of oil-degrading microorganisms in the beach material but their effectiveness in degrading the oil was limited by the availability of nitrogen and phosphorus nutrients. Our field study program has demonstrated convincingly that fertilizers can be applied to oiled beaches to overcome these nutrient limitations, thereby enhancing biodegradation of the oil. In Prince William Sound, the natural biodegradation rate of oil on the beaches was found to be quite high, primarily because of small concentrations of ammonia and phosphate in seawater that are introduced into the beach material with each tide. However, the addition of fertilizers was capable of increasing this biodegradation as much as two- to threefold. In addition, the extent of enhanced degradation was such that beaches appeared much cleaner. This project also produced advances in the technology and science of applying fertilizers to complex beach terrain. Our work showed, as anticipated, that the most effective means to enhance oil biodegradation was to ensure that the oil-degrading microbial communities were exposed to high concentrations of nutrients for sustained periods of time (days to weeks). An oleophilic fertilizer appeared to be very effective in this regard, and its application was straightfor-
ward. It produced a long-term positive effect, suggesting that an oleophilic fertilizer does sequester nutrients at the oil-water interface in such a way that biodegradation is promoted. When subsurface oil was treated by oleophilic fertilizer in combination with a slow-release fertilizer (6-12 in. deep in the mixed sand and gravel), the oil was degraded to a greater extent compared to untreated reference beaches. However, the use of fertilizer solutions was the most effective nutrient application technique in terms of rate and extent of oil degradation, but it was more operationally complicated. Results from the use of fertilizer solutions unequivocally demonstrate that oil biodegradation rates in Prince William Sound were limited by the availability of nitrogen and phosphorous and that the clean appearance of rock surfaces was directly caused by biodegradation. Bioremediation of oil-contaminated beaches was shown to be a safe cleanup technology; no adverse ecological effects were observed. The fertilizers caused no eutrophication, were not acutely toxic to sensitive marine test species, and did not cause the release of undegraded oil residues from the beaches. The success of our field demonstration program has now set the stage for the consideration of bioremediation as a key component (but not the sole component) in any cleanup strategy developed for future oil spills. Its use and effectiveness will depend on the amount of oil present in the contaminated environmental matrix; that is, a longer time will be required for degradation of high concentrations of oil and consequently a longer period of fertilizer application will also be required. In addition, location of the oil and the acceptability of other cleanup options must be considered. In most aquatic environments, enrichments of oil-degrading microbial communities occur soon after oil contaminates shorelines. It is unlikely that natural sources of nitrogen and phosphorus will be sufficient to give maximal degradation rates in light of the available degradable organic carbon from the oil. Thus, the application of fertilizers in these situations should enhance degradation and result in oil removal. Although the amount of oxygen may become limiting in certain situations (e.g., finegrain sandy beaches), the high porosity and large tidal fluxes characteristic of Prince William Sound beaches precluded this as a limitation.
Acknowledgments The success of this project would not have been possible without the collaboration of
the following EPA scientists: J o h n Rogers (microbial analyses), AI Venosa and J o h n Glaser (fertilizer evaluation), Fran Kremer (field site coordination). Dan Heggem (quality assurance), J i m C l a r k and Rick C o f f i n ( e c o l o g i c a l a s s e s s m e n t ) . Larry Claxton (genotoxicity), and J o h n Haines (fenilizer application). We wish t o also acknowledge the special contributions from John Baker, Steve McCutcheon, Dick Valentinetti, Dennis Millar, a n d Steve Safferman. Without the r e s o u x e s and support of Exxon this project would not h a v e been possible; w e a c k n o w l e d g e t h e valuable scientific interaction with S t e v e Hinton, Roger Prince, and Russ Chianelli at Exxon.
Environmental Protection Service Canada. 1982: publ. no. EPS-3-EC-82-2. pp. 4750.
Focht. J. M.: Westlake. D.W.S. Can. J. Microhiol. 1982, 28. 117-22. Bunch. J. N. et al. Conodion Monurcripr Rcporr of Fisheries and Aquoric Srieme.v 1983, NO. 1708. Rabichaux, T.: Myrick, H. J. P m d . Tech. 1972,24. 16-20. Griffths. R. P. et al. Mar. Environ. Res. 1981.5, 83-91. Parsons. T. R. Mor. Eni'iron. Rts. 1985. 13, 265-76. Doe. K. G.: Wells, P. G. In Chemicol Disprrsonrs f i r rhr Conrrol of Oil Spills, ASTM STP 6 5 9 McCanhy, L. T. et al..
Eds.. American Society for Testing and Materials, 1978, pp. 5 M 5 . (33) Duke, T. W.: Petrazruolo. G., Eds. Oil
References (I) Atlas. R.: Pramer. D. Am. Sor. Mirrohiol. NWS 1990, 56.352-53. (2) Lipkin. R. lnriphf 1989, 11. 5CL51. (3) Bedard. D. L. et al. Appl. Emiron. Mionh i d . 1987.53. 1094-1 102. (4) Nelson. M.J.K. et al. Appl. Envirun. Microhiol. 1987, 53. 949-54. (5) Sutlita, I. M. et al. Science 1982, 218. 1115-17. (6) Quensen. J. F.: Tiedje. J. M.: Boyd, S. A. Science 1988.242. 752-54. (7) Kilbane. I. J. et al. Appl. Environ. Microh i d 1982,44. 72-78. (8) Mueller. J. G.: Chapman. P. 1.: Pritchard. P. H. A d . Eniirun. Mirrohiol. 1990. 56. 1079-8k: (9) Bourquin. A. W. Ho:ord. Morrr. Conrrol
(11) Atlas. R. M. Microhiol. Rev 1981, 45,
W - ?.M ..n.. .4 7.
(34) Tagger. S. et al. Mor. B i d 1983, 78. 1320. (35) Atlas. R.: Busdosh, M. In ProcorPedinpr of Ihe 3rd lnrernarionol Biodqrodorion Symposium. Kingston. RL: Applied Science Publishers: London. 1976: OD. 79-86. Jobsan. A. et al. Appl.'Mirrohinl. 1974, 27. 166-71. Westlake. D. et al. Con. J. Microhiol. 1978.24. 2 5 4 4 0 .
..
(40)
1985. (14) Leahy. J. G.: Colwell. R. R. Microhiol. nev. 1990, 54.305-15. (IS) Bartha. R. Microhid. E d . 1986, 12. 15s-72~ ~.. ~.
(49)
401.
(21) Yaniga. P. et al. Proce'eedinRs5rh Narionol
(23) (24) (25)
(50)
(SI)
(52)
Conference. Los Angeles. CA, American
Petroleum Institute: Washington. DC, 1985; publ. no. 4385. pp. 531-37. (26) Westlake. D.W.S. In Oil and Dispersoms in Conodion S e a s 4 c x o r r h Appraisal ond Rerommmdarionr: Sprague. I. B.: Vandermulen. J. H.: Wells. P. G.. Edr.:
S. P.. Eds.: Louisiana State University: Baton Rouge. 1973. pp. 291-95. Lee. K.: Levy, E. M. PI.orPedinRs 1987 Oil Soil1 Conference. Baltimore. MD, American Petr&um Institute: Washington. DC, publ. no 4452. pp. 411-16. Reisfeld. A. et al. Appl. Microhiol. 1972, 24. 3 6 3 4 8 . Venom A. EPA Engineering Research Laboratory. Cincinnati; OH, ihpress. "Alaskan Oil Spill Bioremediation F'm,~ iect": . Office of Research and Development. US. Environmental Protection Aeency: Washington, DC. 1989 €PA/ ~ / 8 - 8 9 / 0 7 3 .pp.-l -I6. Fedorak, P.: Westlake. D. Con. J. Microh i d 1981. 27. 43243. Horowitz. A,:Atlas. R. A w l . Environ. Micmhiol. 1977.33. 647-53: Atlas. R. et al. J. Fish. Res. Boord Con. 1978., . 3. 5 .. . SRS-YO. .. . .. C&k, F.: Westlake, D. Task Forre on Norrhern Oil Dewlopmm: 1974 Repon NO. 74-1.#R72-12774. "Alaskan Oil Spill Bioremediation Project, Update": Office of Research and Develooment. U S . Environmental Protection Agency: Washington. DC, 1 9 9 0 EPAI~OOB-S~ pp./ ~1-20. ~~. Pritchard. P. et al. Inrerim Report: Oil SDill Biorcmediorion Proiecr; Office of Research and Developmen;. U:S. Environmental Protection Agency: Gulf Breeze. FL. 1990. Kremer. F. et al. Proceedinps of the 83rd Annual Mwrinp ond Exhihirion. Air and Wosre Monopcmenr Asmciorion. Pittsburgh. P A 1990. 90-22.1. Rogers. J. et al. Pr0cedinp.t of the 83rd Anuaol Meviinp ond Exhihirion. Air and WOSIP Monopcmenr Arrociolion. Pittsburgh. P A 1990. 90-22.2. Venosa. A. et al. ProcPedinRs of the 83rd Annaol Meerin~and Exhihirion. Air and Wasre Monapmmr Assoociorion. Pittrburgh, PA, 1990. 90-22.3. Clark. J. et al. Proceedidps of rhr 83rd Atmud Meeviq and E.vhihirion. Air ond Wasre Manogemmi As.soriorion. Pittsburgh. PA. 1990. 90-22.4. ~~~
(16) Morgan, P.: Warkinson, R. 1. CRC Crir. Rev. Biortch. 1989,8. 305-33. (17) Lee. K.: Levy. E. M. In Aquatic Toxidogy and Water Qualiry Mottagemeni: Nriagu. I. A,. Ed.: Wiley & Sons: New York. 1989. pp. 2 1 7 4 3 . (18) Raymond, R. L.: Jamison. V. W.; Hudson, J. 0. AlChE Symp. Ser. 1976, 73. 3 ~ 0 2 . (19) Raymond. R. et 81. American Petroleum Institute: Washington. DC, 1978: Project No. 307-77. (20) Minugh. E. et al. Prmecdinps 1983 Oil Spill Confcwnce. San Antonio. TX:
(22)
Atlas. R.: Banha, R. Environ. Sri. Techno/. 1973, 7. 53841. Olivieri. R.. et al. Appl. Emiron. Microhi01. 1976,31, 629-34. Bergstein, P.: Vestal. J. Arcrir 1978, 31. I50-hO ..., .,,.
(58) Olivieri. R. et al. Mor. Pollm Brdl. 1978, 9, 217-20. (59) Bronchan. R.. et al. Procwdinyr 1985 Oil Spill ConfPrrnce. Lor Angeler. CA:
American Petroleum Institute: Warhington. DC. 1985. publ. no. 4385. pp. 453GI)
"A.
(64)Lee, K.: Levy. E. Proceedings 1989 Oil Spill Confermw, San Antonio. TX:
American Petroleum Institute: Warhington. DC. 1989 publ. no. 4479. pp. 479(61) Tramier, B.: Sirvins. A. Plurccdinpr 1983 Oil Spill Confwnrc. San Antonio, T X American Petroleum Institute: Warhington. DC. 1983: publ. no. 4356. pp. 11519. (62) Sirvins. A, Angles. M. NATO AS1 S ~ r i e i 1986, ~ 9 . 3 5 7 4 0 4 . (63) S w u m , P.: Ladausse. A. Prorecdinps I 9 8 9 Oil Spill Confcrmc~.San Antonio. TX. American Petroleum Institute: Washington, DC, 1989: publ. no. 4479. pp. 43946.
"Bugs t w k big bite of Gulf oil spills":
(12) Atlas. R. M., Ed.: Peroleurn Mirrohiolo~ y Macmillan: : New York. 1984. (13) National Academy of Sciences. Oil in the S m : Inpurs. Fares, ond Eflerrs: National Academy of Sciences: Washington DC.
Symposium on Aquifer Resrororion and Ground Worer Monirorinp. National Water Well Association: Wonhington. OH, 1985: pp. 51CL16. Brown. R. et al. Pollur. Enp. 1985. 25. 234-39. Rittmann, B. E.: Johnson, N. M. Wafer Sci. Tech. 1989,21. 209-19. Dibble. J.: Banh, R. Soil Sri. 1919, 125, 56-50, Halmo, G. Proredinps 1985 Oil Spill
600/9-90/041.
Enrironmenr T d q . 1990, 1. 1. Miget. R. In The Mirr.ohio1 Dqradorion
