DOWNLOAD PAPER

Nutrient transport during bioremediation of crude oil contaminated beaches Brian A. Wrenn (Environmental Technologies & Solutions, Rochester, NY) Michel C. Boufadel and Makram T. Suidan (Univ. of Cincinnati, Cincinnati, OH) Albert D. Venosa (U.S. EPA, Cincinnati, OH)

ABSTRACT

The effect of wave energy on transport of dissolved nutrients in the intertidal zone of sandy beaches was studied by comparing the washout rates of a conservative tracer (lithium) on two beaches in Maine. The physical characteristics of the two beaches were similar, and they were subjected to the same tidal influences, but the wave energies were very different. Scarborough Beach is a high energy beach that faces southeast toward the Atlantic Ocean, whereas Ferry Beach is in a protected harbor. This difference in wave energy caused lithium to be washed out of Scarborough Beach much more rapidly than from Ferry Beach. The higher wave energy at Scarborough Beach also appears to have increased the amount of lithium that was diluted directly into the water column. These differences in transport rate and mechanism have important implications for the feasibility of bioremediation for cleanup of oil­contaminated shorelines.

INTRODUCTION

The growth rate of oil­degrading bacteria on contaminated shorelines is often limited by the availability of nutrients, such as nitrogen and phosphorus (Pritchard and Costa, 1991; Bragg et al., 1993; Lee et al., 1993; Venosa et al., 1996). Effective bioremediation requires nutrients to remain in contact with the oiled beach material, and the concentrations should be sufficient to support the maximal growth rate of the oil­degrading bacteria throughout the cleanup operation. Contamination of coastal areas by oil from offshore spills usually occurs in the intertidal zone, where the washout of dissolved nutrients can be extremely rapid. Lipophilic and slow­release formulations have been developed to maintain nutrients in contact with the oil (Atlas and Bartha, 1992), but most of these rely on dissolution of the nutrients into the aqueous phase before they can be used by hydrocarbon degraders (Safferman, 1991). Therefore, design of effective oil bioremediation strategies and nutrient delivery systems requires an understanding of the transport of dissolved nutrients in the intertidal zone.

Transport through the porous matrix of a beach is driven by a combination of three main factors: tide, waves, and the flow of freshwater from coastal aquifers. The focus of this research was on the effects of tide and wave activity. Tidal influences cause the groundwater elevation in the beach, as well as the resulting hydraulic gradients, to fluctuate rapidly (Nielsen, 1990; Wrenn et al., 1997). Wave activity affects groundwater flow through two main mechanisms. First, when waves run up the beach face ahead of the tide, some of the water percolates vertically through the sand above the water line and flows horizontally when it reaches the water table (Riedl and Machan, 1972). Waves can also affect groundwater movement in the submerged areas of beaches by a pumping mechanism that is driven by differences in head between wave crests and troughs (Riedletal. 1972).

The relative effects of tide and waves on nutrient transport in the intertidal zone of sandy beaches was investigated by comparing the washout of a conservative tracer, lithium, on two beaches in southern Maine. Scarborough Beach is a high energy beach that faces the Atlantic Ocean, whereas Ferry Beach is in a sheltered harbor at the mouth of the Scarborough Marsh. Lithium transport at Ferry Beach was driven almost exclusively by tidal effects, whereas tide and waves both affected transport at Scarborough Beach.

EXPERIMENTAL DESIGN

Site Description. The two beaches used in this study are subjected to very different wave energies, but in other respects they are quite similar. Both are composed primarily of medium to fine sand with relatively narrow particle size distributions. Differences in the composition of the two beaches suggest that the hydraulic conductivity of Scarborough Beach might be slightly larger than Ferry Beach, but the small permeability differences were expected to have much less influence on solute transport than the differences in wave energy. The tide was identical at both sites.

Plot Setup and Sample Collection. The tracer was applied to the beach in discrete areas called "plots." Each plot was 5 m wide (i.e., parallel to the shoreline), and they were either 10 m (Ferry Beach) or 12 m (Scarborough Beach) long (i.e., perpendicular to the shoreline). Although the plots on Ferry Beach were shorter than those on Scarborough Beach, the difference in elevation between the tops (i.e., the landward edges) and the bottoms (i.e., the seaward edges) of the plots was approximately the same on both beaches. The plots were set up such that the landward edges were at the elevation that was expected for the highest tide that would occur during the study.

A transect consisting of six multi­port sample wells was installed perpendicular to the shoreline through the center of each plot. The layout of these transects and the elevations of the tops and bottoms of the plots on both beaches are shown in Figure 1. Three of the six sample wells were installed inside the plots, one well was installed landward of the plots, and two were installed seaward of the plots. Figure 1 also shows the locations of the sample ports for each well.

Sprinklers were used to apply the tracer to the beach surface inside the plot boundaries at low tide. Lithium nitrate (>99.7%; Cyprus Foote Mineral Co., Kings Mountain, NC) was dissolved in 100 gallons of fresh water to a final concentration of 33 g/L, which gave it a density approximately equal to the local seawater. Water samples were collected from the multi­port wells periodically for about two weeks.

Water Level Measurement. The water levels in the beaches were measured with transects of six piezometer wells that were installed perpendicular to the shoreline. Piezometer wells were installed at the top, bottom, and middle of the plots. One well was landward of the top, and two were seaward of the bottom of the plots. The most seaward well, which was screened over a four­foot interval above the beach surface, was used primarily to measure the level of the tide whenever it was high enough to submerge any part of the sample well transects. Vibrating wire piezometers (RocTest, Inc., Plattsburgh, NY) were used to measure the water level at each well position. Three readings were usually taken for each piezometer every 15 minutes. These three readings were averaged to smooth out the effect of waves on the water level measurements.

RESULTS AND DISCUSSION

Hydraulic Gradients. The two main forces that drive solute transport in sandy beaches are waves and tidally induced hydraulic gradients. Although no quantitative measurements of the wave activity at the two beaches used in these studies are available at this time, a qualitative comparison can be made by inspection of Figure 1. Whereas the water level changed fairly smoothly at Ferry Beach in response to the tide, the response was quite jagged at Scarborough Beach. Although multiple readings were taken whenever water level measurements were made, it was not possible to completely eliminate variations due to waves from the Scarborough Beach data.

FIGURE 1: Beach profiles showing well positions and the elevations of the tops and bottoms of the experimental plots (i.e., the areas to which the tracer was applied). The circles on each well mark the depths of the sample ports. All elevations were measured relative to a benchmark, but the absolute elevations of the benchmarks on the two beaches were not the same. The tide measurements show that the absolute elevations of the plots were similar on both beaches. Time is measured relative to the beginning of the experiment (i.e., when the tracer was applied).

The effects of waves can also be seen in Figure 2, which shows the hydraulic gradients in the bottom (seaward) half of the plots for both beaches. The response at Ferry Beach was relatively smooth, whereas the gradient fluctuated rapidly at Scarborough Beach. Wave run up and subtidal pumping probably both contributed to these abrupt changes in the hydraulic gradient. In general, the responses of the hydraulic gradients to the tide were similar in both beaches. For example, landward­directed (i.e., positive) hydraulic gradients developed only briefly in this region of both beaches. (Landward­directed gradients persisted much longer in the top half of the plots, however.) Most of the time, the hydraulic gradients were directed seaward (i.e., negative), which is consistent with previous observations (Nielsen, 1990; Wrenn et al., 1997).

FIGURE 2: Hydraulic gradients in the bottom half of the plots at Ferry and Scarborough Beaches. Positive values indicate landward-directed gradients and negative values indicate gradients that are directed seaward. The time is measured relative to the beginning of the experiment, and the time scales for the two beaches are offset by 6 hours to improve readability.

Tracer Washout. Lithium was removed from Scarborough Beach much more rapidly than from Ferry Beach. At Scarborough Beach, less than 15% of the added lithium was recovered at the first high tide following tracer application, whereas about 60% was recovered at Ferry Beach. Tracer was essentially completely removed from the experimental domain at Scarborough Beach within four days, but at Ferry Beach tracer was still present after two weeks. It seems likely that most of the tracer that was lost during the first high tide was diluted into the water column. Following the initial large decrease in tracer concentration, washout proceeded more slowly as the lithium moved through the beach subsurface. Therefore, one of the main effects of wave activity appears to be to change the amount of tracer (or nutrient) that is immediately diluted into the water column relative to that which penetrates into the beach to be transported by subsurface flow.

FIGURE 3: Washout of lithium from high energy (Scarborough) and low energy (Ferry) beaches following application of the tracer to the beach surface at low tide during the full-moon spring tide. The mass of lithium remaining was determined using water samples collected from all six multi-port sample wells in each plot.

These results suggest that surface application of nutrients will be ineffective on high energy beaches, because most of the nutrients will be lost to dilution at high tide. On low energy beaches, however, this might be an effective and economical bioremediation strategy. Nutrients that are released from slow­release or lipophilic formulations will probably behave similarly to the dissolved tracer that was used in this study. Therefore, they will not be effective on high energy beaches unless the release rate is high enough to achieve adequate nutrient concentrations while the tide is out. Subsurface application of nutrients might be more effective on high energy beaches. Since crude oil does not penetrate very deeply into most beach matrices (Gundlach, 1987), however, nutrients must be present near the beach surface to effectively stimulate bioremediation. Nutrients move downward and seaward during transport through the intertidal zone of sandy beaches (Wrenn et al., 1997). Therefore, nutrient application strategies that rely on subsurface introduction must provide some mechanism for insuring that the nutrients reach the oil­contaminated area near the surface.

ACKNOWLEDGMENTS

We are grateful for the cooperation of the citizens and town council of Scarborough, ME for allowing us to use Ferry Beach for this research and to the Sprague Corporation for granting us permission to use Scarborough Beach. The cooperation and assistance of Greg Wilfert and the employees of Scarboro Beach Park is also gratefully acknowledged. Finally, we are indebted to Dave Corbeau, Marine Resources Officer for the Town of Scarborough, and the Scarborough Police and Fire Departments for their valuable assistance during this study.

REFERENCES

Atlas, R. M. and R. Bartha. 1992. "Hydrocarbon biodegradation and oil spill bioremediation." In K. C. Marshall(Ed.) Advances in Microbial Ecology, Vol. 12, pp. 287­339. Plenum Press, New York, NY.

Bragg, J. R., R. C. Prince, E. J. Harner, and R. M. Atlas. 1993. "Bioremediation effectiveness following the Exxon Valdez spill." Proc. 1993 international Oil Spill Conference, pp. 435­447. American Petroleum Institute, Washington, DC.

Gundlach, E. R. 1987. "Oil­holding capacities and removal coefficients for different shoreline types to computer simulate spills in coastal waters." Proc. 1987 International Oil Spill Conference, pp. 451­457. American Petroleum Institute, Washington, DC.

Lee, K., G. H. Tremblay, and E. M. Levy. 1993. "Bioremediation: Application of slow­release fertilizers on low­energy shorelines." Proc. 1993 International Oil Spill Conference, pp. 449­454. American Petroleum Institute, Washington, DC.

Nielsen, P. 1990. "Tidal dynamics of the water table in beaches." Water Resources Research 26: 2127­2134.

Pritchard, P. H. and C. F. Costa. 1991. "EPA's Alaska oil spill bioremediation project." Environmental Science and Technology 25: 372­379.

Riedl, R. J. and R. Machan. 1972. "Hydrodynamic patterns in lotic intertidal sands and their bioclimatological implications." Marine Biology 13: 179­209.

Riedl, R. J., N. Huang, and R. Machan. 1972. "The subtidal pump: A mechanism of interstitial water exchange by wave action." Marine Biology 13: 210­221.

Safferman, S. I. 1991. Selection of nutrients to enhance biodegradation for the remediation of oil spilled on beaches. Proc. 1991 international Oil Spill Conference, pp. 571­576. American Petroleum Institute, Washington, DC.

Venosa, A. D., M. T. Suidan, B. A. Wrenn, K. L. Strohmeier, J. R. Haines, B. L. Eberhart, D. King, and E. L. Holder. 1996. "Bioremediation of an experimental oil spill on the shoreline of Delaware Bay." Environmental Science and Technology 30: 1764­ 1775.

Wrenn, B. A., M. T. Suidan, K. L. Strohmeier, B. L. Eberhart, G. J. Wilson, and A. D. Venosa. 1997. "Nutrient transport during bioremediation of contaminated beaches: Evaluation with lithium as a conservative tracer." Water Research 31: 515­524.